lsf!Kl African INSTITUT SENEGALAIS ...
lsf!Kl
African
INSTITUT SENEGALAIS
Association for
AABN
DE
Biological
q
I!l RECHERCHES AGRICOLES
Nitrogen
B
Fixation
A C T E S
MAXIMISER LA FIXATION
BIOLOGIQUE DE L’AZOTE
POUR LA PRODUCTION
A G R I C O L E
ET FORESTIÈRE
EN AFRIQUE
IIIème Conférence de I’AARNF
7-12 Novembre 1988, Dakay Sénégal
Mamadou GUEYE
Kalemani MULONGOY
Yvon DOMMERGUES
ISSN 0850-072 X
VOL 2
N 02
1988
.
.
.,.
.
I
ISRA
Institut Si-n+lais de Recherches Agricoles
Document réalis? par
La Direction des Recherches sur les Productions \\‘égi-tbies
Route du Front de Terre
B.P 2057 DAKAR-HANN
SENEGAL
= 32.53.03
.<
...->
I
Fixation Hioiogiqui: .iv I’A;. -1: pur ia l’rod~~lion Agricole el Fo~eslièrc cn
Airique : Papicrs prtSscn:cs a la 3e confkence de I’AAEINF à Dakar,
7-12 Novembre 1988 Collcc:on Actes de I’ISRA. Vol. 2, nQ 2.
0c ISRA, 1 9 9 0
Concrpmn et RCalisaticm LJNIVALISRA
--
. .
--
SOMMAIRE
_- - _
.L
.F,..<
. . .
.,
-.* L‘ :
.G’
‘i;;~k!i
;:,
>
‘
>,’
HI Pages
2..
0_’
Les persptxtives de la fmtion ‘biologique de
GUEYE,M.
1
l’azote en Afrique
FIXATION BIOLOGIQUE DE L’AZOTE
I. .
$,
-j,
.a
:
.“’
CHEZ LES LEGUMINEUSES A GRAINES
ferrallitique~forteme désature en bas;~, C&e
d’ivoire
r
. .
,.._ 1.n
3.,:&.” .a L‘:.‘:: ..;.
‘ - _,’
_ I _,,
,x,. ,. i ‘1 ‘i i, ^ ,-/
‘23
,:.
Influence of rhitibii ir&ula~on’on so$ean
AWONAIKB, K.O., OLUFA-
15
(Glycine max L.) performance mNige% ’
: JQ, 0.0. md mu, 1.~.
i
Nodulahon and growth of bambam groundnut : MULONGOY, K and,
21
(Vigna subterranea) in tree soils.
GOLI, A.E. .
.
Effet de l’inoculation avec des souches delihizo-
GUEYE, M.
34
bium et de la fertilisation azotée sur le rende-
, <
,.: II..> ,.
mentengrainsduVoandzou(Vignasubt&?ànea
(L.) Tbouars) au S&t@al.
’ .’
HXAtiON BId&)GIQ* ~gp&~*&hcA ZI.: *i%iv j($
;
f-t.3 *~~!‘~~:‘~!ii~:~’
fnfji&itffrjf$,
,
CHEZ LES LEGUMINEUSES!:F&JR-
RAGERES
Nitrncrn fïu;llinn nf lhc fornm lcc:llrncs 2nd
r!:\\Qr’E, 1
.;,i*
therr resrdual el’kcts on whcat growth and yield.
FLXATION BIOLOGIQUE DE L’AZOTE
CHEZ LES ARBRES
. . . I
Effect of the addition of phosphoks and&o-
BADJI, S ; THOEN, D. ;
54
bium inoculum on thenodulation and thegrowth DU,CGUSSO, M; and ,
of two species of gum arabic trwAcacia senegal
COLONNA, J.P. i
L. Willd and Acacia luetu R. Brl cx Benth.
:.
Nodulation survey of leguminous and non kgU-'
&J&, p-A. ; AKUNDA,
6 3
minous trees growing at ICRAF’s field station.
E.M. and WAMBUGU, P.N.
Effect of inoculation withe Rhizobium, P appli-
GICHURU, M. and
.
12
cation and liming on early growtb of leucaena
MULONGOY, K. ‘r-
(Lewuena leucocephulu Lam. de wit).
:i..:;
,.
, -+
:
.,:.
A preliminary study on the compatibility bet-
ODEE, D.W.
8 1
ween indigenous rhizobia and some tree legu-
mes in Kenya
FIXATION D’AZOTE ASSOCIATIVE
Effect of inoculation on the growth and yield of
MWAIJRA, F.B. and
92
three maize cultivars.
WIDDOWSON, D.
Population of nitrogen fixing bacteria in sweet
HORTENSE, W.D. ; WAL-
99
potato fibrous roots.
TER,A.H.;MULONGOY.
K. ; ADEYEYE, S.O. and
HAHN, S.K.
ASSOCIATION AZOLLA / ANABAENA
La recherche sur Azolla et ses applications :
VAN HOVE, C. et DIARA,
1 0 6
évolution et tendances actuelles.
H.F.
Phosphorus needs and accumulation potential
DESMADRYL, D. ; GO-
1 1 8
of various Azoh species and strains.
DARD, P. ; WAUTHELET,
M. and VAN HOVE, C.
Sensitivity to Aluminium of various Azolfu _ ,WAUTHELET,,M. ; GO-
1 2 6
species and strains.
‘DARD, P. ; DESMADRYL,
D. and VAN HOVE, C.
INOCULATION :
PRODUCTION D’INOCULUM
.-. L_ .,.>. ,;:, ii’i:\\\\,cLL.;; ;\\,:\\
Production and uses of rhizobium inoculants.
BORDELEAU, L.
1 3 2
Caractkkation et choix de supports pour ino-
BEUNARD, P. et
1 4 4
culum.
SAINT MACARY, H.
Progrès récents dans la technologie des inocu-
DIEM, H.G. ; BEN KHALI-
1 5 3
lums utilises en agriculture et en foresterie.
FA, K. ; NEYRA, M. and
DOMMERGUES, Y.R.
ECONOMIE DE L’AZOTE DANS LES
SYTEMES DE CULTURE EN ASSOCIA-
TION ET EN AGROFORESTERIE
Biological nitrogen fixation and nitrogen trans-
MULONGOY, K. and
1 7 4
fer in multiple cropping in tropical Afiica.
EHUI, S.K.
AMELIORATION DE LA FIXATION
SYMBIOTIQUE DE L’AZOTE PAR LA
SELECTION DES PLANTES
.:
Importance de la plante-hôte et amélioration
DUHOUX, E.
193
génétique de la fixation biologique de N,,chez
les arbres tropicaux.
kralysis of the plant genes involved in the De LAJUDIE, P.
.
ti
207
nitrogen fting symbiosis. Sesbania rostrata as
a model.
.
Sesbania rostrata :
TOMEKPE, K. ; .,
213
un modèle pour la biologie mol6culaire de la
HOLSTERS, M. ;
nodulation
GOETHALS, K. ; VAN
DEN EEDE, G. ;
De LAJUDIE, P. ; TRAN, P.
and DREYFUS, B.
AMELIORATION DE LA FIXATION
SYMBIOTIQUE DE L’AZOTE PAR LES
PRATIQUES CULTURALES
Amélioration de la Fixation Biologique de
GANRY, F.
221
SAiste (NJ par les pratiques culturales.
:
Response of Phaseolus vulgaris L. to inocula-
AMIJEE, F. ; EDJE, 0-T. ;
228
tion with Rhizobium and fertilization with
KOINANGE, E.K. ; BlTA-
nitrogen and phosphorus in Northem Tanzania
NYL, HF. ; BRODRICK,
S.J. and GlLLER, K.E.
The influence of some parameters of soi1 fertili-
LIYA, S.M. ; MULONGOY,
250
ty on early growth of Leucaena leucocephala
K. ; ODU, C.T.I. and
and Cassia siamea
AGBOOLA, A.A.
MESURE DE LA FIXATION D’AZOTE
Evaluation of biological nitrogen fixation in
DANSO, S.K.A.
258
plants.
Influence du déficit hydrique sur la fixation
SALL, K. ; DREVON, J.J. et
212
symbiotique de l’azote atmospherique chez le
OBATON, M.
soja
Effect of ‘5N-labekd mineral-N and strains of
LUYINDULA,
N. and
2 8 8
Bradyrhizobiumon biological nitrogen and
N- WEAVER, R.W.
partitioning in cowpea (Vigo wguiculazu) (L..)
Walp.
Quantifïcation of the contribution of BNF to fïeld
BODDEY, R.M. and
298
grown plants : the use of the rsN isotope dilution
URQUIAGA, S.
technique - problems and some solutions.
Utilisation de la méthode 15N pour estimer la
DOMENACH, A.M. ;
317
fixation symbiotique de l’azote chez les plantes
KURDALI, F. et
herbacées et ligneuses.
BARDIN, R.
Nitrogen fixation in tropical trees : estimations
SAGINGA, N. ; ZAPATA
3 3 7
based on lsN techniques.
and DANSO, S.K.A.
CONCLUSIONS
ET RECOMMANDATIONS
Les recherches sur la fixation biologique de N,
DOMMERGUES, Y. ;
352
et leur impact présent et futur sur la production
BORDELEAU, L.M. et
végétale et forestière en Afrique.
GUEYE, M.
PREFACE
Aprèslesréunions~Nairobien 1984etauCaireen 1986,l’InstitutS~nnégalaisdeRecherches
Agricoles (ISRA) a organisé la troisième confërencede l’Association Afïicaine pour laFixation
Biologique de l’Azote (AABNFQ. Le thème g&kral de cette troisième confkrence était «
Maximiser la Fixation Biologique de l’Azote pour la production agricole et forestikre en
Afrique ».
Maximiser la « Fixation Biologique de 1’Azote » pour la production agricole et foresti&re
est sans nul doute une des voies que les pays en voie de développement, particulièrement
ceux d’Afrique, doivent adopter pour relancer leur production agricole dont la tendance est
2 la décroissance et sauvegarder leur patrimoine forestier exposé dangereusement a la
dégradation.
En effet, dans un environnement à forte propension aux bouleversements kologiques
et aux déséquilibres d’&osyst&mes naturellement fragiles, les pays africains ont beaucoup
de défis à relever parmi lesquels :
0 la sécurisation de la production agricole et la couverture de la demande alimentaire,
l la sauvegarde du milieu naturel.
11 s’agit la d’enjeux de toute importance exigeant la mise en œuvre d’actions correctrices
permanentes liées aux opkations agricoles. Deux actions paraissent prioritaires à l’heure
actuelle pour la sauvegarde des agrosystèmes : le maintien et l’amélioration de la fertilité
des sols (pour la plupart potentiellement pauvres et fragiles) et la lutte contre la désertification
par un effort soutenu de reboisement.
Dans ce cadre, la biotechnologie en général et la fixation biologique de l’azote en particulier
offrent d’excellentes perspectives pour les pays africains. La mise en œuvre effective de cette
technologie nécessite :
* l’isolement et la culture en quantités industrielles de souches de microorganismes
fixateurs d’azote;
l la mise au point de méthodes pratiques et efficaces d’inoculation à la portée du
développement;
fl
‘,:
,’
.:,lj’i/ >Jjb;.:..:‘. ,’ i. ,,. i ::
,hL ;:, ;.:lL’.% L’:
:.i!ii L:< icur aptitude à l
a
nodulation et à la fixation de l’azote;
l
lkolement et l’identification de souches de champignons micorhiziens et la mise en
œuvre de méthodes pratiques d’inoculation facilement utilisables en pépini&e.
L’Association Africaine pour la Fixation Biologique de l’Azote devra aida les pays
africains à atteindre ces objectifs pour l’accroissement du niveau de vie des populations rurales
par la promotion d’opkrations agricoles intégrant l’utilisation des microorganismes fixateurs
d’azote.
Je remercie toutes les institutions nationales et internationales ainsi que la communauté
scientifiquemondialequiontaimablementassisté1’ISRAdansl’organisationdecetteconf~ence
de I’AABNF.
Mouhamudou El Habib LY
Directeur Général de 1’I.S.R.A.
AVANT PROPOS
troisième conférence de l’Association Africaine pour la Fixation
Biologique de l’Azote (AABNF) a et6 organisée ti Dakar, Senégal,
par l’Institut Sénégalais de Recherches Agricoles (ISRA) du 7 au
12 Novembre 1988.
L’ISRA, établissement public chargé de concevoir, orga-
niser et mener à bien tout= la recherches relatives au secteur mal
au Sénégal, bénéficie de la collaboration très étroite de l’Université Cheickh Anta Diop
de DakartJJCAD) et de celle de l’Institut Français de Recherche Scientifique pour le Dé-
veloppement en Coopkration (ORSTOM) au Sénégal.
Plusieurs agences internationales ont contribué financièrement à l’organisation de
la 3éme conférence AABNF :
* l’Organisation des Nations Unies pour I’Education, la Science et la Culture
(UNESCO) à Paris par ses divisions : SER, ABN, ROSTA et MAB.
LUNESCO centralise le réseau mondial des MIRCENs ;
l le Programme des Nations Unies pour l’Environnement à Nairobi ;
l l’Organisation des Nations Unies pour 1’Alimentation et l’Agriculture
(FAO) a Rome ;
l le Programme des Nations pour le Développement (UNDP) à Dakar ;
l le Centre Technique de Cooptkation Agricole (CTA), installé depuis 1983
à Ede/Wageningen au titre de la convention de Lame entre les états du
groupe ACP. Le CTA est à la disposition des états ACP pour leur permettre
un meilleur accés à l’information, à la recherche, à la formation ainsi
qn’aux innovatinns dans Icc w-trur9 de r-b~rln~~cmcnt ?&r-nl- nt nlrqt rf
dc ia vuigmsdtwr ,
* le Centre de Recherche pour le Développement International (CRDI) à
Ottawa ;
l la Fondation Internationale pour la Science (FIS) à Stockholm ;
l YUnion Internationale des Sociétés de Microbiologie (IUMS).
La troisième conférence AABNF a réuni 80 participants venant de : Angleterre,
Angola, Autriche, Belgique, Brésil, Burkina Faso, Canada, Congo, Côte d’ivoire, Egypte,
Etats Unis d’Amérique, Ethiopie, France, Ghana, Haïti, Italie, Kenya, Liberia, Mali,
Maroc, Nigéria, Ouganda, Pays Bas, Rwanda, Sénégal, Sierra Leone, Tanzanie, Tunisie,
Zaïre et Zimbabwé.
Comité d’organisation
Le comité d’organisation de la 3è conférence AABNF comprenait :
BA A.
ORSTOM, Dakar, Sénégal
BA A.T.
UCAD, Dakar, Sénégal
BADJI S.
Eaux et Forêts, Dakar, ShZgal
BARRETO M.M.S. (Mme)
UCAD, Dakar, Sénégal
BORDELEAU L.M.
Canada Agriculture, Quebec, Canada
DIAGNE 0.
ISRA, Dakar, Sénégal
DIARA H.
ADRAO, St-Louis, Sénégal
DOMMERGUES B.
BSSFWCTFI’, Nogent sur Marne
DREYFUS B.
ORSTOM, Dakar, Sénégal
DUHOUX E.
BSSFI’/CTFT, Nogent sur Marne
GANRY F.
ISRA/ClRAD, Bambey, Sénégal
GUEYE M.
ISRA, Bambey, Sénégal
MBAYE D.F.
ISRA, Bambey, Sénégal
NDOYE 1.
ORSTOM, Dakar, Sénégal
SADIO s.
ISRA, Dakar, Sénkgal
SEYE Y.K.G.(Mme)
UCAD, Dakar, Sénégal
SOUGOUbRA B.
Eaux et Forêts, Dakar, Sénégal
MAMADOU NGUER
ISRA/CNRA/Bamtxy BP.53, Bambey,
SENEGAL
ABDOURAHMANE DIOM
.>
KHOUDIA NDIAYE
n
Mme DIOR FALL MBODJ
<I
ROBERT DIOKH
”
MBAYE DIOUF
91
.
r
-,
*
‘,
:...
,
..,
.- .
b!
<
.<.
:+,y;
-3i.P
!X,‘,
‘J’a:#i
0’
4,. ., . ..*
GYYE,
M.
i.1
,,‘;
,“.,
I;,
:;’
‘.
.’
,..
:.
.‘.
i
:,
.’
‘i
‘:
:.
..
,
;:>;,i
~.
;t;Qk
i:;;.
!&
MIRCEhKNRA
BP. 53 Bambey., SknkgaI
:
..,
‘.
-.,
~‘~~:Yy~:,“. I!t:’
!‘<
:”
;j?{.‘: . ..‘ ;\\;II?: !“‘!
,
.;
.\\
‘ ) 2.“‘!:, i ‘. i,
. .
ri’. ,
:.
I
,
,‘$:J . .:y*. >i(>- ,.;,
,.’ ,.. .- “.1 .:.
: : ; .. \\.>
.’
:
..?. :, .<_. . . . . . . .:_ ,..- .~Xfi&Fi[i: m ~XiSI
-.z ,:<‘i
:.-
5 ; i i .i
i,:rf;!“i,,;: of rl: *‘(;:j
1
.
&y$j+;&~*,c
‘&&&j UsJrt i.& .3$x3 d &a’:J : LTm+l$&
.$i$:
)“!.
h &c&&-jm ,&&&.x&&nnfnfl~;& fiútitihhe+j- &?jj&@&$$&’ afri-
cainesidQend .en premier, lieu et en ti2.s @-ande Pa&ie ++@$&@~~~a&
culture, qui sont eux-mêmes handicapés par la pa&reti d& &ls peb I&les,
carencés en CEments majeurs (N, P, K) et en oligdléments (zn, CU). POU~
obtenir des rendements
nrricoler suftkmts dans de tclc SOIS, il est nfccq-
sue de recourir à un emploi massif d’engrais chimiques. Or, les cultivateurs
afïicains, dans leur grande majorité. n’utilisent pas ces engrais parce qu’ils
sont importés, et donc trop coûteux. Ainsi, pour am6liorer la production
agricole en. Afrique, il faudra obligatoirement faire appel~à”d$tio&el&
I.
t&nol@& &,Ja fo~&ficaces @. mat&&&. sp~~~&$&&*b~?
I’introduction-sy~~matique
,des plantes’ fixatnces; ~~t~:~i.~~’ -~-’ . ~
d
agricoles devrak~permettre le maintien et l’an&tior&on dé ~.‘f&&&&$
sols. L’ag&&.ure.itin&ante~en &t un ezémj$e. fis’p~tes&@&$@&~
jouent un..rôle prépqnd&ant dans Ile cadk ‘de. syst&mes de&& &.@&-&$
ou à~.~v&r& la’ n-&&e orgaiq&, ‘&n$, & ’ &&O~S &*~?&Q~$
les cultures en -association ou en couloir.
j:, o”2*z;,’ ” !**
‘S
:
,._
. . .
;
a-
‘.
.‘-
.; (’
t ,
.’
; 2 . . É.‘..:i. +::J:
9,.
,.
I
-I
‘:.
‘s .>.. .,!! ..<1,
2
Importance des plantes fixatrices d’azote dans l’agriculture et la foresterie
,)
,..,.
en Afrique
Les l@wGneuw a graines ont fait l’objet de nombreuses études. Elles occupent
une place pr@ond&ante dans l’agriculture et y jouent un rôle V&S important aussi bien
dans la satisfaction des besoins alimentaires des hommes et des animaux que dans le
maintien de la fertiliti des sols. En effet, grâce à leur capaciti de fmer l’azote atmosphé-
rique en association avec les Rhizubizmz, les légumineuses à graines permettent d’&cono-
miser l’azote du sol. L’utilisation des variétés a haut potentiel fïxatew d’azote dans les
sys&mes culturaux accroit la teneur en azote des sols et en assure la fertilité. D’autre
part, dans le cadre d’une rotation de culture l&umineuse-cMale, au cours des processus
de décomposition, l’azote des racines et des nodules peut être transf&6 a la &Sale suivante.
En Afi-ique~l~e dizaine de/-légumineuses $ gra$es.soFt cultiv&s~~,.$s wf6rentes
zones 6colog1q&%%&&!r&an-i~‘&~ le .&$ (G&ne max) le haric$ &aseol& vuigaris),
.
le ni&5 (Vigna zuzgzziczdata), l’arachide (Arachis hyp@$z&],&&s d&e ‘&%&i&zz&z)
et le pois bambara ou voandzou (vigne subterranea).
Les espèces ligneuses fixatrices d’azote sont appelées elles aussi 2 jouer un rôle
de plus en plus important en Afrique pour plusieurs raisons. D’abord, elles peuvent pousser
sur des sols tr&s pauvres ou totalement dépourvus d’azote fournissant ainsi du bois ou
du fourrage pour l’alimentation du bétail : c’est le cas de.Acacid &da et d’a&es acacias
introduits en Afrique sahélienne ainsi que de Medicago arborea et Acacia cyanophylla
en Afrique du Nord. D’autre part, beaucoup d’espèces ligneuses contribuent à la n5g6nnémtion
des sols en reconstimant le stock de matière organique dans le cadre des cultures intercal-
laires en sys&~es agroforestier (Leucaena leucocephala, Glirictia sepizmz) ou dans le
cas de la rizicula (Sesbania rostrata). Les esp&es ligneuses contribuent également &
la fixation d&dunes : c’est le cas du ftio (Caszzarina eqzzisetifolia).
., ,,I~;~Jw&I<&~c; &J%tape des “ii%herche?actiellésd& t&&lê fnhimum de profit
‘&& ‘~nnaiss&&~,~quises pour, V&iser au inieux la fiXarion$biologique de l’azote
dans le but d’améliorer la production agricole et forestière en Afrique.
Amélioratiori de la fixation biologique de l’azote par action sur les micro-
organismes fixateurs d’azote
<
..-., ‘.
La pren$&déanarche pour augmenter la fixation biologique de l’azote par action
.
sur les microorg+n@es ,COIXXW.~ .s&.ctionner .judicieusement pour chaque espèce de
légumin&$ ~@-e~L~~c~aque~vari&5 .de l@mineu& ftitrice ‘d’&t&‘u6e‘Ou des souches
“’ ‘i’
de Rhzrobizm c.apal$ewde@~.@& effectivement -l’azote~atmosph&ique4 la suite de leur
in?aUation &r%. p&&-hôte, considérée. Le schéma de s&ction a ét6 4r&s bien décrit
k b@&$$%.&& ‘@&Cent, 1970). Il repose sûr ‘la connaissance de critères bien
précis : (ij infect&&& :&iffecGvit6 dans la fixation biologique de l’azote ; (ii).comp&itivité
pourlafo&&~&&&&&.set(i&) adaptationauxconditionsdel’enviroimement:temp&ature,
.) ,;: .J
pH, concentration en ‘~1, &pe de sol, taux d’azote ~$/OU de matière organique du sol,
antagonistes (certains actinomycètes...).
L’inoculation des plantes fnatrices d’azote avec des souches sélection&s de mi-
croorganismessymbiodquesagénéralementdonnédebonsrésultatsdanslecasdesl~gumineu~s
suivantes : soja @mon, 1976). haricot (Keya et aL, 1982), luzerne (Medicago sativa),
(Bordeleau et al., 1977). stylosanthes (Stylosanthes guineansis) leucaena (Skerman, 1982)
4
‘;‘. : : Ii; cL;$ii “ii’I
..<#Y? ‘;-.:lt x ..‘ial ‘.f ,“,‘.-.
Actuellement, on commence & &r&ppel-‘aux techmques de .g&ie ‘g&néhqtie fi&
cn5e.r des souches ~pluQerformhntes parles ~proceksus de transformation, de traduction
et.de conjugair8n-b~ac~~~~éntiLiue l’information g&&ique qui déter-
mine la fixation biologiquei~W~ le!~%k&ies”& cO&5iue dans une séquence
de. genes~d6nornmés N&dorlt la~ta;‘strucmre:et le ~fonc%ionne&nt kont décrits avec précision
(Sasson, 1984). hs gènes Nif de Klebsiellaqneumoniae
ont tu5 transf6res avec succès
très peu effective et n
Am6lioration de la fixation biologique de l’azote par la sblectlon des plantes
RL!i‘I?i‘!OI!~ ijlii: IA i)‘JllbliISC:
2SL L)X Ci:I’,:a:Li<,J, “’
u18~ d!ovCid~JOi;l Liiilb L1CU.i ~);LLi~lI;liIi~
pour un bénéfice reciproque. Il est donc logique de chercher a maximiser la fixation biolo-
gique de l’azote en agikanf’non ‘s&lement sur le-partenaire bactérien, mais aussi en agis-
a-: Q3
A h fm de h ‘d&$&&‘~~@i), .uùe.nouvell~~~~no~ogie s’est &,&pp& :‘i;l~~git
de la multiplication vég&a&ve in vitro qui permet une exploitation raison& de la’!&&+
bilit6 intraspkifique chez les ‘phuites. Dans le cas des esp&es fixat.rices d-te; cette
technologie permet d’accroître substantiellement la quantité d’tite fixé ‘en permettant
la multiplication desr clones iden$fï~ comme les plus, performants.
‘, ‘; ,,&) J1::~,
Cette démarche’peut’facikment s~a@iqi.$r’&x:ri.rb~~, fiiateu&d’&ote’&r~i@ pt-&
sentent une variabilité très importante et leur densite a l’hectare est comprise entre 2ooi)
4
et 5000 plants. En revanche, on ne peut envisager de recourir à cette méthodologie dans
le cas des plantes furatrices d’azote annuelles dont la densité est de l’ordre de 100.000
plants à l’hectare (Dommergues, 1987).
Les centres de ressources microbiologiques (MIRCENs)
L’amélioration de la fixation biologique de l’azote ktant une priorité pour les pays
africains, puisqu’elle contribue à l’autosuffisance alimentaire avec un minimum d’engrais
azotes, il y a lieu de collecter et de préserver les microorganismes fixateurs d’azote capables
de s’associer de faç~~~esfective
avec les plantes amtliorées. C’est la un des rôles majeurs
des h4IRCENs. Le rôle de ces MlRCENs est de :
fournir l’infrastructure d’un réseau mondial de laboratoires travaillant en
l
collaboration au niveau régional et interr6gional à la distribution et ci l’uti-
lisation des pools de gènes des microorganismes ;
renforcer les efforts relatifs à la préservation des microorganismes d’intérêt
l
économique ;
promouvoir l’application de la microbiologie a l’agriculture ;
l
servir dé, centre de formation pour la diifusion des connaissances sur la
l
1 microbiologie.
Actuellement, il y a trois MlRCENs en Afrique. Le MIRCEN de Nairobi avec
une collection de 500 souches de Rhizobium, produit de l’inoculum sur un support de
boue fïltréc pour la culture du haricot. Le MlRCEN du Caire renferme une collection
de 200 souches de Rhizobjum et conduit plusieurs experiences sur l’inoculation de la
fève (Viciafubu). La‘collection du MIRCEN de Bambey renferme 300 souches de Rhiwbium
pour les exp&i~,~~~s@ !a fi@ion biologique de l’azote chez l’arachide, le niébé, le
;)‘:: .,;<
.’
<
V~&&U, ‘ie soja et lis &a&&
L’association africaine pour la fixation biologique de l’azote (AABNF), les
institutions internationales
Dans la plupart des pays ahcains, existe une institution nationale, privée ou semi-
privée, travaillant sur un aspect de la fixation biologique de l’azote. Cependant, ces institutions
nationales ont des moyens très limites qu’il conviendrait de rassembler sous la forme
d’un réseau. C’est encore là un des objectifs des MIRCENs et par delà un objectif de
I’AABNF. II n’est.donc pas étonnant que les MIFKENs d’Afrique aient soutenu et supporté
les premières activk% de I’AABNF. A côté des institutions nationales, il y a les institu-
tions internationales.
L’impact dè l’International Institut of Tropical Agriculture @TA) de Ibadan, Nigeria
dans les programmes de recherches sous régionaux n’est plus à démontrer : depuis les
recherches sur le ni&6 en particulier jusqu’aux r6cents travaux sur l’exploitation des arbres
fixateurs d’azote dans les systemes de culture en couloir. L’International Council of
Research in AgroForestry (ICRAF) a démontre, dans les régions au Sud du Sahara, l’impor-
tance des arbres fixateurs d’azote et leur intégration dans les systèmes agroforestiers.
Il y a tgalemnt les nombreuses agences internationales oeuvrant pour le développe-
ment de l’agriculture en Afrique a travers des programmes axes sur la fixation biologique
de l’azote : CRDI, CTA, FAO, FIS, NAS-BOSTID, UNDP, UNEP, UNESCO.
5
11 apparait donc indispensable d’&.aIuer dès a présent, le potentiel humain dont
dispose actuellement l’Afrique pour mener à bien les recherches envisagées. C’est pourquoi,
PAABNF, par l’intermediaire de I’IITA à Ibadan, dresse un annuaire des microbiolo-
gistes en Afrique travaillant dans le domaine de la fixation biologique de l’azote. Il faut
maintenant raisonnablement mesurer ce qui reste a faire pour que tous les pays africains
puissent utiliser avec le maximum de profit toute la technologie de la fixation biologique
de l’azote. Cet objectif requiert au prt%able la formation d’un personnel qualifié et une
plus large diffusion des connaissances acquises sur la fixation biologique de l’azote en
Afrique.
Rf$f&ences
BORDELEAU, L.M., ANTOUN, H. et LACHANCHE, R.A. 1977 Effets des souches
de Rhizobim meliloti et des coupes successives de la luzerne (Medicago sativa) sur
la fixation symbiotique de l’azote. Can. .J. Plant Sci. 57, 433-439.
BURTON, J. 1976 Problems in obtaining adequate inoculation of soybcans. In: World
soybean research. Pmceedmgs of tlte world soybean research conference. Ed. by Lowell
D. Hill. The Interstate’Printers and Publishers, Inc. Danville, Illinois p. 170-179.DIEM,
H.G., GAUTHIER, D. et DGMMERGUES, Y. 1983 An effective strain of Frankia
from Casuarina sp. Can. J. Bot. 61, 2815-2821.
DOMMERGUES, Y. 1987 Comment accroître la fixation symbiotique de l’azote par les
arbres en milieu tropical. In : Actes des semina& sur les arbres fixateurs d’azote
et l’amélioration biologique de la fertilité du sol. Editions de l’ORSTOM, Paris p.
18-27.
GUEYE, M. and BORDELEAU, L.M 1988 Nitrogen fixation in bambara groundnut,
Voandzeia subterranea (L.) Thouars. Mircen J. 4, 365-375.
KEYA, SO., BALASUNDARAM, V.R., SSALI, H. et MUGANE, C. 1982 Multiloca-
tional field responses of Phase&s vulgaris to inoculation in eastem Africa. In :
BNF Technology for tropical agriculture. Papers presented at a workshop held at
fT,AT \\fnrch 4.12 JO‘?j C~>~~ITY PM, y!,.1 TI,.....' c p '- ' r, 1. .r' ' : ,!
,
23 l-234.
PGSTAGATE, J. 1987 Nitrogen fixation, 2nd edition. Edward Amold (Publishers) Lui,
London 73~.
SASSON, A. 1984 Biotechnologies : challenges and promises. Eds. UNESCO, Paris. P.
109-155.
SKERMAN, PJ. 1982 Les légumineuses fourrageres tropicales. Collection FAO produc-
tions veggétales et protection des plantes. Rome. 666 p.
VINCENT, J.M. 1970 A manual for the practical study of root nodule bacteria, IBP
n* 15. Blackwell Scient& Publications, Oxford.
FIXATION BIOLOGIQUE DE L’AZOTE
CHEZ LES LEGUMINEUSES A GRAINES
Maximiser la FBA pour la Production Agricole et Forestière en Afrique
x-.
..; I 1
,r;i.‘, t’ :,<y ,,.. g*;&; .::;.
_,
b
‘.
:‘,‘T
.,
.f,6..Lu1$e-.h’ d;*$*ja Gly&i& m& ‘&) fie,*i , ” j_<
et du niéb& Vigna unguiculata (L.) Walp.)
sur sol feryallitique fortement désaturé
..a .’
.,..<.:..
.en .basse&~te &Ivoire.
. . .
L$xmztoire de Microbiologie et des Industries Agro-alimentaires. EhTA, OS BP 35 Abidjan
08,Côtcd’Ivoùe.
,.‘..
Des essais d’inoculation du soja (Glycine max) et du ni6bk (Vigna wzguiculata) ont
td conduits en serre et en plein champ sur un sol ferrallitique fortement désature
en .Mai, .Juin et Juillet 1987. Les résultats obtenus indiquent d’une part qu’en sol
.I
.;!
;, 1 - &’ &&:(PH 4,8),‘3 ~t+po&iblë $a&&&& &&e &&&&a~~*&&~s
&&&a-
tions symbiotiques (L&rmineuses-Rhizobium)
bien adaptées aux conditions extrêmes
de culture. D’autre part, l’inoculation du soja et du nitW peut être bénéfique en
Côte d’ivoire. du moins dans la @on où l’expérience a Cte conduite.
La nutrition arot& des légumineuses est un important facteur limitant du rendement
I
cn graiilzj du SOJ;I ((J/JC& mm) CL du nlcbk (l-iztid bryucuidld) CI dc leur I”;L.
en tant que précédent cultural. or, cette nutrition azotée est assurée par deux voies :
l’assimilation de l’azote combiné du sol ou des engrais azotes et la fixation biolo-
gique de l’azote atmosphérique. Ces deux voies sont influenckes différemment par
les facteurs de l’environnement si bien qu’elles peuvent être concurrentes ou
compl6mentaires suivant les stades physiologiques de la plante ou les conditions du
milieu (Decau et al.. 1975 ; Obaton et Kimou. 1985).
La modification & l’dquilibre entre les contributions de ces deux sources peut avoir
d’imporumtes implications agronomiques et économiques. En effet, des gains sub-
stentielsdefixatonetderendementsontpossibleschezcertainese~del~gumineuses
par inoculation avec des souches trés efficientes. Mais l’eflïcacité de cette inoculation
peut être compromise par des conditions pkloclimatiques
inadéquates. Dans le présent
travail, nous avons cherché à estimer l’influence des
modalitts de fourniture d’azote
par le sol ou par la fumure sur la participation relative de l’une ou l’autre source
et à en évaluer les cons6quences agronomiques en sol fcrrallitique fortement désa-
tur6 très typique des zones forestiks de la basse Côte d’ivoire.
,.
.,,..,
~)
.._
_>.
I
,.,-<. ..-.,.
.
.
.
-,..-
c.
,I
MatWels et mbthodes
Le niébé cv. V, et le soja cv. Tropical ont été cultivds parallèlement en champ
et en serre en pr&ence de 0 et 90 kg N/ha. L’azote est appport& sous forme d’ur&.
Pour chaque type de IQnrninenses. les ,@nes ont ét.6 s@il&+ parJrempage pendant
30 mn dans une solution d’hypochlofité- dè calcium (66 g/l) püi ,$@@mrnent-rinu&
à l’eau distillk stirile et ensuite répar& eti deux lots. Un lot ,#$noc& avec une suspen-
sion de Rhizobium japonicum G49 @WA-Franc@ paw le ,yJa et,(e:R@zobium tropical
N7 (ENSA-Côte d’ivoire) pour le &~‘~~‘i&culation ,a &té ‘rt?aWë8 une dose forte
(lOg Rhizobium par graine) juste avant le semis; L’ati&e’lot SI~ @nec est Seme sans
inoculation.
L’expérience au champ a été conduite sur la parcelle exp&imentale de 1’Ecole
Nationale Agronomique d’Abidjan en une seule saison culturale de Mai a Juillet 1987.
Cette période correspond ZI la grande saison detipluies&&ti
&?&hüre en Côte’d’Ivoire.
Les sols sont du type sablo-argileux contc+mt en moyenne .2;84JWde matière organique
et 1080 ppm d’azote total ; 17,s ppm de phosphore assimilable ; complexe absorbant
25,ll % ; Mg : 0,23 meq/lOO g ; Ca : 0,84 me@00 g ; K : 0,05 me@00 g et pH
eau 4,8. Le site chosi pour l’exp&ience est relativement homog&ne et n’a jamais porté
de culture de soja et de ni&.
En serre, les plantes ont été cuItivées en pots de 5 1 sur des échantillons de sol
provenant de la parcelle d’expérience au champ. La serre est un abri avec un toit en
verre transparent qui n’arrête que les eaus de pluies. I-es murs sont constitués de grillage
de 0,5 cm de maille. Ainsi les plantes en serre bénéficient des mêmes conditions d’&zlaire-
ment naturel (durée d’ensoleillement var&nt entre 10 et 12 heures), de’temp&ature (tempéra-
ture ,dii$;,: .g,O. à, ,30.?&; ,~~ra&n&wcmme S?4-&26930)&~&fiuhUmiditi (humidité
relative pendant la p&iode de.culIturec95 WXl~:%): quec&ItWSpletichtip.
Les plantes
sont arros&s au fur et à mesure des -besoins, en solution non coulante à l’eau distill&.
En pleine floraison, 45 jours après le semis pour le niébt! et 60 jours après le
semis pour le soja, quelques plantes sont déterrées pour le dénombrement des nodosités
formées. La partie aérienne de chaque plante est sectionnée au noeud cotylédonnaire et
i I i .;; Il i>L!:~i ;i . L...,<’ .i e;
.
I
L jLiicid:ii -3~1 11. LC +ds dz. lidikc: sAc: JC: ia p.ut~c
aérienne est déterminé par pesée. Les rendements en graines sont mesurés après la récolte
?+ la maturité et après séchage au soleil jusqu’à 15 % d’humidité.
Le dispositif expérimental utilisé en champ et en serre est en blocs de Fisher entière-
ment randomisés. Il comporte 4 traitements .(t&noin absol& témoin! avec inoculation,
90 kg N/ha et 90 kg N/ba avec inoculation.
_.
;.
-,
Rkwltats et discussion
Les &ultats de l’inoculation sont indiqués dans les Tableaux 1 et 2 respective-
ment pour le soja et pour le Niét&
Inoculation du soja - Les effets de l’inoculation du soja, cv. Tropical avec
la souche G49 sont très nets. Au champ, en l’absence de fumure azotée, l’inoculation
permet de passer de.O, t/ha (témoin absolu) à 2.62 t/ba soit un accroissement du rendement
en graines de 23 % par rapport au témoin non fumC et non inocul& (Tableau 1A). En
l’absence d’inoculation, l’apport de 90 kg N/ha augmente le rendement en graines, d’environ
9
67 % mais ne provoque aucune différence significative sur la production de matière séche
des parties aériennes. En présence d’inoculation, l’apport de 90 kg N/ha augmente
significativement la production de matière sèche des parties aériennes de 13.9 % et provoquent
dans le même temps une diminution de 21 96 du rendement en. gra$tes~ Cette baisse
du rendement est due à un avortement elevé:des fleurs snite à un developpement vég&atif
excessif des plantes suivi de leur verse. Ces difft%ntes observations faites au champ sont
confirmées par les resultats en serre (Tableau 1B).
Tableau 1A : Essai d'inoculation de Glycine max (cv. Tropical) avec
l'inoculum 649 (INRA-France) : résultats au champ.
Nombre de no-
Poids de ma-
Rendement en
Traitements
dosités par
tiére sèche
graines
plante
t/ha.
t/ha
Témoin non
inoculé
0 kg N/ha
O b
5,69 c
0,78 c
90 kg N/ha
O b
5,92 c
1.30 b
Inoculé
0 kg/ha
72 a
7,97 b
2,62 a
90 kg/ha
69 a
9,08 a
2,16 a
Dans chaque colonne les valeurs suivies de la m@me lettre ne dif-
fèrent pas au seuil de 5 % (méthode Newman-Keuls)
10
Tropical)avec‘
z.
T'inoculum G49.SINRd-France) : résultats en serre
P,
. . . .
Nombre de no-
Poids de ma-
Rendement en
Traitements
dosités par
tiëre sèche
graines
plante
gfplante
g/plante
Témoin non
inoculé
0 kgfha
O b
18,28 d
3,57 d
9 0 kg/ha
O b
24,32 c
8,90 c
Inoculé
0 kg/ha
5 0 a
31,14
b
15,04 a
9 0 kg/ha
5 6 a
41,68 a
Il,83 b
- 90 kg N/ha correspond à 33 mg N/lOO g du sol utilisé en serre
- Chaque valeur est une moyenne sur 16 plantes (4 plantes x 4 répëti-
tionsl
Inoculation du niébé - Les observations faites sur l’inoculation du soja, cv.
Tropical sont identiques à celles faites sur l’inoculation du niébé cv. V,. Cependant, l’effet
de cette inoculation semble moins spectaculaire sur le niébé comparé aux résultats obtenus
sur le soja. Au champ, en l’absence de fumure azotée, l’inoculation augmente le rendement
en graines de 32,75 % par rapport au témoin non inoculé (Tableau 2A). Pour les autres
traitements, aucune différence significative n’est observée. La différence de résultats observée
entre le soja et le niébé pourrait être justifiée par le fait que le niébé non inoculé porte
des nodosités qui possèdent une certaine aptitude à furer l’azote atmosphére (section rouge
des nodosités). Un traitement témoin non nodulé aurait été intéressant pour infirmer ou
confirmer cette hypothèse.
_
*
,
..,
;.~,.ti’;,’ -..
.
.
.
..:,
‘.‘.
:
_.
Tableau 2A : Essai d'inoculaiion de Vignatunguiculata (Vl) avec
.,.<'i,'.. 1 ../-
l'inoculum'local N 7 (ENSA-Côte-d'Ivoire) : résultats au
champ.
-Nombre de nodo-
Poids de ma-
Rendement en
Traitements
sités par
tière sèche
graines
plante
t/ha
t/ha
Témoin non
!: :>.:
<. .'
i n o c u l é
..-.;,F'.
0 kg N/ha
8 b
6,OO b
0,76 b
9 0 k g N/ha
7 b
6,65 a b
0,85 a b
.nnr,:lé
0 k g N/ha
20 a
5,86 b
1,OOla
9 0 k g N/ha
1 5 ab
7,08 a
0,86 ab
Dans chaque colonne les valeurs suivies de la mëme lettre ne diffèrent
pas au seuil de 5 4 (Méthode Newman-Keuls).
I
,
12
Tableau 28 : Essai d'inoculation de Vigna unguiculata (Vl) avec l'ino-
culum local N7 (ENSA-Côte-d'Ivoire) : Résultats en serre
Nombre de nodo-
Poids de ma-
Rendement en
Traitements
sités par
tiëre g/par
graines
plante
plante
g/pl ante
Témoin non
inoculé
0 kg N/ha
8 b
23,08 b
3,75 c
90 kg N/ha
8 b
3,95 c
Inoculé
90 kg N/ha
27 a
24,90 a
4,99 b
- 90 N/ha correspond â 3 mg N/lOO g du sol utilisé en serre
- Chaque valeur est une moyenne sur 16 plantes (4 plantes x 4 répéti-
tions.
- Dans chaque colonne, les valeurs suivies de la mëme lettre ne dif-
fèrent pas au seuil de 5 49 (Méthode Newman-Keulsl.
13
Conclusion
De ce travail deux points essentiels semblent retenir particulièrement notre atten-
tion :
en sol tr?s acide, il est possible d’avoir une bonne nodulation avec certaines
l
associations symbiotiques «légumineuse-Rhizobium» bien adaptées aux
conditions extrêmes.
9 l’expérience conduite sur le domaine exp&imental de 1’Ecole Nationale
Sup&ieure Agronomique d’Abidjan montre que l’inoculation du Soja cv.
Tropical et du Ni&6 cv. V, peut être bénéfique. Bien que l’inoculation
du niébé soit rarement pratiquée dans le monde, cette augmentation de
rendement d’environ 33 % dans le sol présentant de telles caractkis-
tiques nécessiterait d’effectuer des expérimentations analogues en d’autres
points du pays. La confirmation de tel effet bénéfique pourrait conduire
à une application pratique (inoculation du soja et du niébé) en milieu
Pw=n.
Remerciements
Ce travail n’aurait certainement pas eu lieu sans l’aide financière de la Communauté
Economique Européenne (Contrat CEE 02.087/TSD-A-180) et la franche collaboration
des chercheurs de 1’INRA de Montpellier en particulier M. Obaton et J.J. Drevon. Qu’ils
rqoivent ainsi que tous les membres de la CEE nos sincères remerciements.
RQft$rences
DECAU, M-J., BOUNIOLS, A., LENCREROT, P. et PUECH, J. 1975 : Concurrence
ou complémentarité de l’alimentation azotée non symbiotique et de la fixation bactérien-
ne chez le Soja (Glycine max. L. Merr.). CR. Acad. Sci.. Paris 281, 535-538.
<.,_
..\\.....~.. z< LL .i...IDC
nase au cours du cycle végétatif chez le Soja. In : Nutrition azotée des légumineuses,
Versailles, 19-21. Ed. INRA, Paris, 1987 (les colloques de I’INRA, N* 37).
Maximiser la FBA pour la Production Agricole et ForestZre en Afrique
I,
t.;.:::;’
‘.
.;.
f.j,d
.
I.
.i,.
. . ;
I, ,,d
.1,:
;
:
,.
‘.,I.
,
.
.:
.
,.
-(
.
.
1
.
:.
In.flu&nce of rhiz6bia inoculation
on soybean -(GZ~cine .max L.)
perfbrmtitice’ ‘iti : NigeGa
AWONAIKE, tiO.,“’ .~LUFAJO, 0.0.‘2’
and. Ah,, J.K.(2)
: htitutc of Agricdhual Research and Training Obrgemi Awolowo University
PMB. 5029, Ibadan. Nigeria;
(2):. Institute for Agricultural Research Ahmadu Belle University Samaru, tiia, Nigeria
’
The response of soybean (Glycine max L.) to inoculation with exotic and indigenous
rhizobia (no inoculation) and inorganic nitrogen fertilization was studied at Ilorin
and Samaru, Nigeria, during the 1985 cropping season In the two locations. nodula-
tion was enhanced by all inoculants. while inorganic nitrogen fertilkation significan-
tly enhanced vegetative dry mattex production. Location effect was observed for the
respmse of soybean to inoculation with regard to seed yield. While in Ilorin, grain
yield was not enhanced by inoculation, in Samaru, only strain IRj 2133 significan-
tly’ influenced the seed yield of some soybean cuhivars.
j
A generaliiation 04 the need or otherwise to inoculate soybean for enhanced perfor-
mm% in Nigeria is not possible from this study. However, a more justifiable conclu-
sion may be reached if other parameters, such as amount of nitrogen fixed and the
proportion of it which is available to a subsequent grown cmp, are assayed.
16
lntroductlon
The potential contribution of soybean to the improved nutrition of most Nigerians
that can ill-afford animal proteins cannot be overemphasized. Although it has been culti-
vated for sometime in traditionally growing areas of Nigeria, its wide national acceptance
and cultivation is relatively recent.
Soybeans derive N directly from the soil as inorganic residuals or as mineral&d
organic soil fractions and indirectly by the symbiotic nodule relationship where the bacte-
rium Bradyrhizubim juponicwn fix N present in the atmosphere. Harper (1974). confirmed
that both symbiotically fixed and inorganic N are required for highest yields of soybean,
while estimates of the fared N to the total N range from 25 to 60% (Vest et al., 1973).
Thus, yield increases from N fertilization would be expected to be relatively common.
Yet only a few reports of measured yield increases from N fert&ation are available
in the literature (Bhangoo and Albritton, 1972 ; johnson and Hume, 1972) and these
increases were often very small (Sorrensen and Penas, 1978). Somme other investigations
however found no effect of N fertilization on soybean field (Wagher, 1962, Beard and
Hoover, 1971).
Ayanaba (1977) reported that the tropical soils are devoid of indigenous B. japo+un,
although the cowpea type Rhizobizm strain (which is prevalent in these soils) readily
form effective nodules on the roots of soybean varieties of Asian origin (Nangju. 1980).
It is generally believed that when soybean is cultivated in a field with no previous history
of soybean cultivation, artificial inoculation with B.japonicum
is essential to ensure high
seed and protein production (Graham, 1985). Although several investigators have reported
beneficial effect of inoculation on soybean (Rang, 1975 ; Pal and Saxena, 1975 ; Chowdhury,
1977 ; Singh and Tilak, 1977), others have reported inoculation failnres (Salema and
Chowdhtuy, 1980 ; Awai, 1981 ; Awonaike, 1988).
Addition of combined nitrogen and inoculation have been shown to respectively
enhance vegetative matter yield and nodulation in some grain legumes (Ayanaba and Nangju,
1973). Their combined effect on seed yield of some tropical grain legumes under field
conditions is inconclusive (Ofori, 1973 ; Luse er al., 1975 ; Pal et al., 1985).
The objectives of this study were therefore to ascertain the need or otherwise to
inoculate soybean for enhanced performance in Nigeria and if necessary, to identify strains
of B. juponicum that are adapted to the environmental conditions.
Materials and methods
,
Field experiments were conducted on newly cleared land (without a history of
soybean cultivation) in 1985 at the Ballah substation of the Institute of Agricultural Research
and Training in Ilorin (8”3O”N, 4O35’E) and on the Institute for Agricnltural Research’s
experimental farm at Samatu, Nigeria (ll”ll”NN, 7”38’E). The soil at Ilorin belongs to
the Apomu series and that of Samaru is a well-drained, leached ferrogenous tropical sandy
loam.
(i) : Treatments of the llorin trial - Eleven soybean cuhivars were subjected
to three nitrogen nutrition regimes (viz., N, : uninoculated control treatment without nitrogen
fertilization ; N2 : 150 kg urea per ha provided in three split applications at 25 days
interval commencing at planting ; N3 : inoculation with a mixture of B. juponicum strains
IRj 2114 and IRj 2133). A randomized complete block design incorporating four replicates
17
of the 33 treatments was used. A basal application of 400 kg single superphosphate
pe.r hectare was applied before planting because of the low P status of the soil (3.2 ppm
p>-
(ii) : Treatments of the Samaru trial - Treatments consisted of five soybean
cultivars-grown with four single IRj strain peat inoctits (2114,2123. 2133 and 2144)
and a mixture containing equal proportions of the four strains. An uninoculated control
treatment without nitrogen fertilizer and another uninoculated treatment with 100 kg
N/ha applied as calcium ammonium nitrate (CAN) (26 96 N) were also included. A randomi-
ml complete block design incorporating three nql@es of each of the 35 treatments
I
i.‘.
was used.
Each plot size in the both trials was 6 m by 2.25 m made up of four rows plants,
75 cm between the rows and ‘5 cm within the row.
Symbiqtic effectiveness was evaluated at 9 weeks after planting (WAP) by asses-
sing fresh weight of nodules produced and dry weight of the shoots (Ilorin trial) and
tie nodules number and dry weight (Samaru tial) of five competitive plants per plot,
The remaining plants were left to mature when the seeds were harvested.
AI1 the data collected were subjected to statistical analysis of variance.
Results
Nodulation and vegetative matter production. The nodulation data as expres-
sed by the fresh weight of nodules per plant (Ilorin trial) and number and dry weight
of nodules per plant (Samaru trial) at 9 WAP arepresented in Tables 1.2 and 3, respectively.
Averaged over cultivars, all inoculants significantly enhanced nod&ion when compared
to the uninoculated control in both locations. Also, nitrogen f-on decreased nodula-
tion, but not significantly. Among the inoculants at Samaru, strain IRj 2133 gave the
hig&t nodules number and weight while strain IRj 2123 performed poorest. The cultivars
also differed significantly in their ability to nodulate. As expected at Ilorin, Bossier cultivar
nodulated very poorly with the indigenous rhizobia but nodulated highly with the mix-
ture of inoculant strains, while cultivars TGx 539-!%, Samsoy 2 and M-90 nodulated
well with both the indigenous rhizobia and the inoculant mixture. At Samaru however,
cultivar TGx 814-26D produced the highest number and weight of nodules while the
Samsoy cultivars produced the least number and weight of nodules. Significant interaction
between cultivar and inoculant strain was observed at Samaru in that while cultivar TGx
88849C responded to all inoculants except strain IRj 2144, cultivars TGx 536-02D, Sam-
soy 1 and Samsoy 2 responded td all inoculants ex&pt strain kj 2123. TGx 814-26D
responded to all inoculants except the mixture of strains.
At Ilorin (Table 4) and Samaru (result not shown) inorganic nitrogen fertilisation
significantly enhanced vegetative matter production. There was however no significant
difference in the mean vegetative matter production of the inoculated and uninoculated
control plants. Varietal differences were also observed.
Seed yield. The seed yield data are presented in Tables 5 and 6.
Despite the significant effects of source of nitrogen on nodulation and vegetative
matter production at Ilorin, these were not reflected in the seed yield (Table 5). Significant
varietal effects were however observed. TGx 814-26D and TGx 539-5F significantly out
yielded the rest of cultivars. The very poor yield of TGx 573-209D was due to poor
18
Table 1. Effect of inoculation wi'th B. japonicum strains on nodulation
-
of soybean cultivars (Ilorin location).
Fresh weight nodules (ng per plant)
Cultivars
Uninoculated
Uninoculated
Mixture of
(indigenous
Mean
+ 150 kg N/ha strains
rhizobia)
1Rj 2114
and 2133
_-
-
Samsoy 1
0.06
0.05
0.24
0.12
TGx 297-10F
0.05
0.03
0.14
0.08
TGx 539-5E
0.10
0.03
0.38
0 . 1 7
M-90
0.05
0.03
0.11
0.06
TGx 814-26D
0.04
0.04
0.16
0.08
TGx 136-0211
0.02
0.01
0.15
0.06
TGx 813-6D
0.01
0.05
0.08
0.05
Samsoy
2
0.08,
0.01
0.16
0.08 ;.
TGx 573-209,
':
0.02
0.02
,0.13
0:oi )-
TGx 888-49C
0.05
0.05
0.19
0.10 '
Bossier
0.01
0.01
0.46
0.16
Mean
0.04
0.03
0.20
C.V. % 32.5
LSD (P = 0.01) N = 0.01
LSD !P = 0.01) V = 0.02
LSD (P. = 0105) N x V = 0.07
~ -
Table 2 : Effett'of inoculation with B_,,;japonicum strains on the noduIation of
" sdy'#an cultivars (Samaru location),
,:. :I
Number of nodules per plant
TGx
TGx
TGx
Inoculant
88-49C
536-02D
Samsoy 1
Samsoy 2
814-26D
Mean
1Rj 2114
70.2 (b-h)
43.5 (c-i) 67.9 (b-h)
14.5(hi)
98.9 (a-c) 59.0 (tu)
1Rj 2123
51.3 (b-i)
22.3 (f-i) 29.1 (e-i)
16.7
g-i)
93.0 (a-d) 42.5 (uv)
1Rj 2133
79.7 (b c)
87.3 (b-d) 72.5 (b-g)
88.4
a-d)
140.3 (3)
93.6 (s)
1Rj 2144
23.7 (f-l)
102.7 (a b) 69.7 (b-h)
75.0
b-f)
104.1 iab)
75.1 (st)
Mixture
59.8 (b-h)
63.7 (b-h) 43.3 (c-i)
59.2
b-i)
82.3 (0-e) 62.3 (tu)
Uninoculated
23.3 (f-i)
6.8 (i)
17.'3 (g-i)
23.1
f-i)
41.2 (d-i) 22.3 (VW)
Uninoculated+
.
100 kg N/ha
9.4 (i)
7.3 (i)
20.2 (f-i)
18.3 (g-i)
49.1 !I)-i) 20.9 (w)
Mean
45.4
48.2
45.7
42.2
87.0
SE inoculant means = 7.26; SE cultivar means = 6.13 ; SE inoculant x (.liltivar mean =
16.23. Mean followed by the same letter or letters are not significant at the 5 per
cent level, according to Duncan's multiple range test. Means followed lly 2 letters
connected with a hyphen are included in all groupings between the two letters in
the alphabet.
Table 3 : Effect of inoculation with B japonicum strains on the nod .!ation of
soybean cultivars (Samaru location).
Nodule dry weight (mg per plant)
TGx
TGx
TGx
Inoculant
88-49C
536-020
Samsoy 1
Samsoy 2
814-26D
14 c a n
-
-
-
--
1Rj 2114
382.9
222.4
353.9
55.6
254.7
253.9 (bc)
1Rj 2123
267.6
100.5
99.7
83.3
344.1
179.0
cd)
1Rj 2133
429.8
472.7
304.7
524.5
486.5
4ti3.6
a)
1Rj 2144
139.2
364.0
327.3
336.3
460.3
3i',.4
b)
Mixture
472.4
315.5
159.4
320.3
299.1
31 j-3
b)
Uninoculated
153.4
43.8
64.4
88.1,
173.7
104.7
d)
Uninoculated+
100 kg N/ha
52.2
45.1
91.3
68.8
218.9
94.7 (d)
Mean
271.1
223.4
200.1
210.5
319.6
----
__-
S.E.D. of inoculant means = 44.4 ; S.E.D. of variety means = 37.5 ; ' .F:.D. of
inoculant x variety means = 99.25.
Values in column followed by the same letter are not significantly di' ferent at
the 5 percent level, according to Duncan's multiple range test.
21
establishment of the plants. At Samaru, the relatively earlier maturing cultivars
TGx 888-49C and TGx 536-02D yielded significantly higher than the other three cultivars
which did not differ from each other. Also, the interactive effect of cultivar and strain
on seed yield was significant (F’ = 0.05) at Samaru. Inoculants did not increase the yields
of cultivars TGx 88849C. TGx 53602D. Samsoy 1 and TGx 814-26D, while the yield
of Samsoy 2 was improved by strain IRj 2133.
Discussion
Nangju (1980) postulated that the local cult&us can form effective nodules with
.indigenous rhizobia and thkrefore can be grown without seed inoculation with R. juponicum.
All the cultivars used in this study (except cultivar Bossier) were selected under the
local conditions and were found to respond to inoculation. The fact that the strains differed
in their ability to form effective nodules with the different cultivars indicated that the
most important
is the use of cfllcicnt strains of R. jupunicum that can compete
with the indigenous rhizobia. The comparative advantage of higher number of organisms
in the itioculank than in the soil may also account for the response. Other researchers
have reported differential performance of soybean inoculants (Kang, 1975 ; Chowdhury,
1977 ; Chowdhury er al., 1983).
The positive re&nses of all the cultivars to inorganic N fertilization (with regard
to vegetative matter production) and to inoculation (with regard to nodulation) suggest
that the indigenous rhizobia do not adequately supply the N needs of the plants. It is
however not clear why these (increased matter production and nodulation) do not enhance
seed yield. Further studies on the distribution of absorbed and fixed N to the parts of
an actively fixing soyv plant especially at its reproductive phase should help to explain
‘.’ ‘the situation.
.
Until this is done however, the inoculant requirement situation in Nigeria for enhan-
ced soybean production is uncertain.
Acknowledgement
The authors are grateful to Dr. K. Mulongoy (IITA, Ibadan) for supplying the
peat inoculants. Also a grant from the Federal Ministry of Science and Technology of
the Federal Government of Nigeria, for the National Co-ordinated Research Projects on
Soybean under which auspices this study was conducted, is gratefully acknowledged.
22
Table 4 : Effect of inoculation with B. japonicum strains on vegetative dry
matter yield hg per plant) of soybean cultivars (Ilorin location).
Uninoculated
Uninoculated+
Mixture of strains
Cultivars
(indigenous)
150 kg N/ha
1Rj 2114 and 2133
Mean
Samsoy 1
10.5
13.2
9.3
11.0
TGx 297-1OF
12.3
14.1
11.4
12.6
TGx 539-5E
10.6
15.8
7.6
11.4
M - 90
8.5
12.8
7.0
9.4
TGx 814-26D
10.8
15.9
7.9
11.5
TGx.536020
12.6
16.9
9.6
.13.0
:,
TGx 813-6D
13.6
14.9
11.1.
13.0
Samsoy 2
9.5
18.2
10.7
12.8
TGx 573-208D
12.2
15.6
11.0
13.0
TGx 888-49C
10.7
16.8
11.5
13.0
Bossier
7.2
22.4
9.2
12.9
Mean
10.8
16.1
9.7
C.V. (“i, ) = .19.5
LSD (P = O-01) N = 3.6
LSD (P = 0.05) v = 1.0
Table 5 : Effect of inoculation with 8. japonicum strains on tl., seed yield
'a:+(kg per ha) of so$bean~~ultivar^s (Ilorin location).
'
! ! i',..
I
: LL:I:
;;'*
-*,(>q
;>E
I
\\I
j c: j
i
. :
., ,.
!:rq,
!I
Uninoculated
Uninoculated+
Mixture of c.Irains
Mean
Cultivars
(indigenous
+ 150 kg
1Rj 2114 ant:
Mean
rhizobia)
N/ha
2133
--
--
-
Samsoy 1
1941
2499
2335
2285.3
TGx 297:1OF
2260
1665
2173
2032.7
TGx 539-5E
3648
3203
3242
3364.7
M - 9 0
2508
2320
1930
2286.0
TGx 814-260
3608
3143
3170
3307.0
TGx 536-021)
1578
2008
2903
2163.0
Samsoy 2
1917
1685
1814
TGx 573-209D
860
218
337
471.7
TGx 888-491:
2812
2588
2773
2724.3
Bossier
1615
2092
1908
1871.7
Mean
2208.1
2118.0
2201.8
C.V. (%I 15.8
q
LSO (p = 0.01) v = 268.3
-v..--I
=
e---n..-.--
-A
-I----
A
.-
Y A--
I8
Y
&.A
a
‘3
Y
a,
Y
IX---
‘3
--......
‘3
6
24
bb
Y
Y
‘3
A
Y
25
References
I
AWAI, J. 1981 Inoculation of soybean (Glycine mar L. Merr.) in trinidad. Trap. Agric.
;‘”
(Trin.).
~58,?,3~“;~18.
::
‘_,
“&&&&,. K;ij’. .i(j8i Respon;iof $y&, ~&~~ & ,L*) u, indigenous *izobia
and other sources of nitrogen in south-western Nigeria. Nig. J. Agronomy (in press).
AYANABA, k. 1977 Towards better use of inoculants in the humid tropics. In : Biological
nitrogen fixation in farming systems of the tropics. (Ayanaba, A-and Dart, P.J. eds.),
pp. 18f-187. .Jdhn Wiley and Sons, New York.
AYANABA, A. and FANGJU, D. 1973 Nodulation and nitrogen fixation in six grain
’ iegdipe&‘& 1 Pro&dings of the first IITA grain legume improvement workshop.
pp. 198-204. IFA, Ibadan publ.
BEARD, B.H. akd HOOVER, R.M. 1971 Effect of notrogen on nodulation and yield
o f irrig,qt& soybeans. Apron. J. 6 3 , 815-816.
BHANGHOO, MS. and ALBRITTON, DJ. 1972 Effect of fertilizer nitrogen, phosphorus
and potassium on yield and nutrient content of Lee soybeans. Agron. J. 64, 743-
746.
CHOWDHURY, M.S. 1977 Response of soybean to Rhizobium inoculation at Momgoro,
Tanzania. In : Biological nitrogen fmation in farming systems of the tropics. (Ayanaba,
A. and Dart, PJ. eds.) pp. 245-253. John Wiley and Sons. New York.
CHOWDHURY, M-S, MSUMALI, G-B. and MALEKELA. G.P. 1983 Need for seasonal
inoculation of soybeans with rhizobia at Morogoro, Tanzania Biol. Agric. Hart. 1,
219-228.
GRAHAM, P.H. 1985 Problems of soybean inoculation in the tropics. In : World soybean
research conference III. mgs (Shibles, R. ed.) pp. 951-959. Western Press,
Inc.
HARPER, J.E. 1974 Soiland symbiotic nitrogenrequirementforoptimum soybean production.
Crop Sci. 14, 255-260.
JOHNSON, H.S. and HUME, D.J. 1972 Effects of nitrogen sources and organic matter
on nitrogen fixation and yield of soybean. Can. J. Plant. Sci. 52, 991-996.
KANG, B.T. 1975 Effect of inoculation and nitrogen fertilizer on soybean in Western
Nigeria. Expl. Agric. 11, 23-31.
LUSE. RL., KANG, B.T., FOX, R.L. and NANGJU. D. 1975 Protein quality in grain
legumes grown in the lowland humid tropics with special reference to West Africa.
In : Fertilizer use and protein production. XI. Collegium International Potash Institute,
Denmark.
NANGJU, D. 1980. Soybean response to indigenous rhizobia as influenced by cultivar
origin. Agron. J. 72, 403406.
OFORI, C.S. 1973 The importance of fertilizcr N in grain legume production on soils
of granitic origin in the upper region of Ghana. In : Proceedings of the Fit IITA
grain legume improvement workshop. IITA Ibadan publ.
PAL, U.R., OLUBAJO, 0.0. and ADU, J-K. 1985 Recent findings on the cultural and
nutritional requirements of soybean in the Guinea savanna zone of Nigeria. In : Pro-
ceedings of the 5th annual workshop of the nigerian soybean scientists. pp. 52-64.
26
PAL, U.R. and SAXENA, MC. 1975 Response of soybean to symbiosis and nitrogen
fertilization under humid sub-tropical conditions. Expl. Agric. 11, 221-226.
SALEMA, M.P. and CHOWDHURY, M.S. 1980 Rhizobial inoculation and fertilizer effect
on soybean in the presence of native rhizobia at Morogoro, Tanzania. Beitz Tropiqh.
a
Landwirt. Veterinamed. 18, 245-250.
SINGH. H.P. and TlLAK, K.Y.B.R. 1977 Response of soybean to inoculation with various
commercial inoculants of Rhizobium. Indian J. Agron. 22, 57-59.
SORRENSEN, R.C. and PENAS, EJ. 1978 Nitrogen fertilization of soybeans. Agron.
J. 70, 213-216.
VEST, G., WEBER, D.F. and SLOGER, C. 1973 Nodulation and nitrogen fixation.
In : Soybeans : improvement, production and uses (Caldwell, B.F. ; Howell, R.W.and
Johnson, H.W. eds.) Amer. Sot. of Agron., Madison, Wisconsin.
WAGNER, G.H. 1962 Nitrogen fertilization of soybeans. Missouri Agrir. EXP. Starim
ices. b’uu. IY I.
I.‘__
Maximiser fa FBA
la Production Agricole et Forestitre en Afrique
Nodulation and growth
o f baididra &dunddut (Vigrtu su7jterrLd,l,eu>
in th&e soils
.,:
:.
._
MULONCiOY, K ‘and’ GOLI; A.E.
Inter~ional htitute of Tropical Agriculture (II’EA)
PiUB. 5320 Oyo Road, Ibaakn. Nigeria
-1~
.
Ab’stract
L
:.r
:,.;.A:
.:
.,I.
. .
Nodulation and dry matter production of ten cultivars of bambara groundnut differing
in growth habit, maturity period and yield potential were assessed in potted soils
from three locations in southern Nigeria. Nodulation occurred in all three soils with
the least number of nodules plant-’ in the acid soil (Ultisol) from Onne. near Port
Harcourt
Urea did not affect nodulation in general. Differences were observed between cultivars
. .‘bS I.* on the:basis “6f ntkhrk number
:>f-> (1 ,a.
but not of nodule’ dry..weight Relative symbiotic
(_
. -:::,. $3 ..
.,.
~~‘~effectiven&~vahresbased
c-k phmt dry ‘tieight av&rage$W % for all cultivars grown
in the three soils. This indicated that bambara groundnut
$xablishes effective symbioses
with iitdig&ot&‘iG&a in thkse ‘soils. Only cultivar TVsu 1231 responded to urea
i
N. signifrc~tly (at P g 0.05) in soils from the International Institute of Tropical
.*’ Agriculture (ETA) and Pashola (Alfiiols). There &s’no correlation between growth
habit or yield p&x&l and plant dry matter’pmduction.
28
Introduction
A few years back, bambara groundnut (Vigna subterranea
(L.) Verdcourt) was
one of the most widely cultivated leguminous crops in Africa (Doku and Karikari, 1971).
Nowadays, with the expansion of groundnut production for oil. bambara groundnut is
only sparsely grown in Africa (Goli, 1987) in spite of its pest and disease tolerance both
in fields and during storage, its nutritional value and its adaptation to a wide range of
soils (Doku and Karikari, 1971).
Few studies have been conducted to document the ability of bambara groundnut
to fix nitrogen and grow on low-N soils. Duke et al. (1977) reported in their review
that high soil N promotes vegetative growth of bambara groundnut at the expense of
seed production. Stanton (1968) found that inoculation of new fields of bambara ground-
nut with soil from plots where the crop grew well improved crop yields. But Johnson
(1966) and Denarit et al. (1968) showed that inoculation with rhizobia did not improve
yici& signil’ictiltiy. Marc inucuiauurr studits are being carried in pars oi Senegal where
bambara groundnut does not nodulate freely (Gueye, M. pers. comm.).
The germplasm unit of the International Institute of Tropical Agriculture @TA),
Ibadan, Nigeria, possesses an important collection of bambara groundnut with 2027 ac-
cessions. In view of the potential of this legume in tropical agriculture, we assessed its
growth and nitrogen-fming ability in three low-N soils comprising an Ultisol from the
humid region, and two Alfisols from the subhumid forest-savanna transition zone. Cultivars
used in this study represented a range of growth habits, maturity periods and yield poten-
tials of bambara groundnut
Materials and methods
Ten cultivars of bambara groundnut were compared on the basis of their nodu-
lation, dry matter production and response to urea to assess their nitrogen-fixing ability.
The experiment was conducted in pots at IITA. Growth characteristics and yield potential
of the cultivars used are presented in Table 1. The completely randomized design was
used with five replications and a factorial combination of 10 bambara groundnut cultivars,
three soils low in N from IITA, Fashola (at about 70 km North of Ibadan) and Onne
near Port Harcourt, all in Nigeria, and two rates of urea applied at 0 and 100 mg
N kg-’ soil. Some properties of the experimental soils are presented in Table 2. Urea
.y,as applied as a water,, solution at one and two weeks after plant emergence. All pots
” received a basal application of single superphosphate (25 mg P kg-’ soil) and muriate
of potash (15 mg K kg’ soil).
Five seeds of each cultivar were sown in each pot containing 3 kg of soil previ-
ously passed through ‘a 10 mm sieve to remove stones and large plant parts. Seedlings
were thinned to three at one week after emergence. Plants were harvested at 8 weeks
after planting for nodule count and mass, and plant dry weight.
Results and Discussion
Nodulation occurred in all the soils with an average of 30 nodules plant-’ in Fashola
soil (Table 3). Nodule number was 21 % and 44 % less in IITA and Onne soils respec-
tively ; but plant nodule dry weight was the same in the three soils indicating that nodules
29
Table 1. Growth characteristics and yield potential of cultivars of
Bambara groundnut studied.
Cultivars
Yield jqtential
'Growth hab4t'
(TVsuI
Days to maturity
(g plant-')
5
spread
123
4
9a
bunch
113
27
9b
spread
115
1 5
132
spread
131
1 8
409
intermediate
105
5
465
b u n c h
105
1 5
2"
790
b u n c h '
127
094
1 055
bunch
150
5
1 061
intermediate
157
5
1 231
intermediate
123
87
30
Table 2. Some properties of the experimental soils collected from
IITA, Fashola and Onne.
-
Parameter
IITA
Fashola
Onne
P H
5.8
6.2
4.6
Organic C, St
0.53
0.36
1 .Ol
Total N, %
0.11
0.04
0.13
Extractable P (Bray-l),
mg kg-'
21.0
7.8
54.8
CECa,
meq 1OOg -1
1.89
1.21
2.32
Total acidity (Al + H), meq 1OOg -'
0.02
0.01
1.72
aCation e x c h a n g e c a p a c i t y
were relatively smaller in IITA and Fashola soils. Thompson and Dennis (1977) reported
numbers ranging between 92 and 318 nodules plant-’ in Ghana. Free modulation was also
reported in Madagascar (DenariC et al.. 1968) and in Zaire (Rassel, 1960).
Nodulation in this experiment was not affected by fertilization with urea except
in Fashola soil (results not presented) where nodule number of cultivars TVsu 9b and
TVsu 409 decreased and that of cultivar TVsu 9a increased in the N treatment These
differences were also observed when the average nodule number for all three soils was
calculated (Table 3). In general, there were in both N treatments differences between
cultivars of bambara groundnut on the basis of nodule number but not of nodule dry
weight (Table 4). Correlations between nodule. numbers or weights and shoot, root or
plant weights were poor at P = 0.05 (r < 0.253) in the no-fertilizer N treatement. For
instance, plant dry matter production was highest in Onne soil where nodulation was
the poorest (Table 3). Also, there was apparently no relationship between growth habit,
maturity period or yield potential and nodulation, relative effectiveness or dry matter pro-
. :
. . ..__ :;...
Table 3. Nodulation and growth of Bambara groundnut in three
different soils (1)
.:
Plant dry'
fiodule
Nodule
Single nodule
Soil
weight '.
number
dry weight
dry weight
(g plant-')
plant -1
(mg plant-')
(mg)
Ibadan
4.2
2 3
8 7
3.8
Fashola
3.2
41
100
2.4
Onne
5.3
_
16
80
5.0
LSD (0.05)
0.89
4.3
NS
1.6
(1) Means of 10 cultivars in the no-fertilizer N treatment.
32
Table 4.
Response of ten cultivars of bambara groundnut to fertilizer Nl
and their relative effectiveness (R.E.)2 in nitrogen fixation
based on plant dry weight.
..,>i,“l’
r>O,:L
“‘y
~*liuii
1 t ItUlilbcr
I.UliU I c til y
>lhglc
II”UVI’
:..:
weight
weight
dry weight
(TVsu)
N O
Nl NO
N1 NO
N1
NO
N1
__- -
-
-
-
_
g plant -1
No. plant -1
mg plant -1
-----w-----
?z
5
4.3 5.3
3 0
27 168 98
5.6
3.6
8 2
9a
3.5 4.0
1 8
31 54
71
3.0
2.3
8 8
9b
3.7 4.4
2 2
16 80
126
3.6
7.9
8 4
132
5.3 4.0
2 3
27 88
59
3.‘8
2.2
133
409
3.3 4.0
3 4
21 111
146
3.3
7.0
a 3
465
5.1 4.7
28
32 92 97
3.3
3.0
109
790
5.5 6.3
2 5
32 83
7 3
3.3
2.3
8 7
1055
4.5 3.8
2 2
33 88
132
4.0
4.0
118
1061
5.7 6.0
2 0
13 69
5 6
3.5
4.3
9 5
1231
4.3 6.7
11
16 60 71
5.5
4.4
6 4
LSD(S%)
1.48
11.1
96
4.32
ND3
1NO = without fertilizer N; N1 = with 100 mg urea N kg-l soil
2R.E. : Relative effectiveness = Plant dry weight at NO * lOO/Plant dry
weight at NT
3ND
: not determined as R.E. values were obtained from computed values.
3 3
Dry matter production was only slightly increased by N fertilization. Except for
cultivar TVsu 1231. relative effectiveness based on plant dry weight ranged between 82
and 132 %. This was an indication of effective symbiosis between bambara groundnut
and indigenous rhizobii present in the three soils studied. Cultivar TVsu 1231 had relative
effectiveness values of 421.64% and 87% in Fashola, IITA and Onne soil respectively,
with an average relative effectiveness of 64%. Thus cultivar TVsu 1231 had the least
effective symbiosis with native populations of rhizobia in BTA and Fashola soils and
would require inoculation with the appropriate rhizobia to fix nitrogen adequately in these
soils.
Aknowledgement
The autors thank KE. Dashiell for his suggestions and W.O. Oyekanmi and E.I.
Jegede for technical assistance.
References
DENARIE, J., ANDRIAMANATENA, S. and RAMONJY, J. 1968 L’inoculation des 16-
gumineuses a Madagascar. R&nltats de I’expkimentation. Agr.Trop. 23, 925-966.
DOKU, E.V. and KARIKARI, S.K. 1971 Bambara groundnut. EconBot. 25, 255-262.
DUKE, J.A., OKIGBO, BN., REED, CF. and WEDER, J.K.P. 1977 Voandzeiu subferranea
Q Thouars. Tropical Grain Legume Bulletin. 10, 8-l 1.
GOLI, A.E. 1987 Plant exploration in Senegal. G.R.U. Exploration 1986. BTA, Ihadan,
Nigeria.
JOHNSON, D.F. 1966 Bambara groundnut. A review. Rhod. Agric. .I. 65, 1-4.
RASSEL, A. 1960 Le voandzou (Voandzeiu subterranea L. Thouars) et sa culture au
Kwango. Bulletin Agricole du Congo Belge. 51, l-26.
STANTON, WR. 1968 Grain Legumes in Africa. Rome : FAO.
TOMPSON, EJ. and DENNIS, E.A. 1977 Studies on nodulation and nitrogen fixation
by selected legumes. Proceedings of the University of Ghana Council for Scientific
and Industrial Research : symposium on grain legumes in Ghana. University of Ghana
p. 85-102.
35
Très rtkemment, les essais d’inoculation du voandzou (Vigna subferranea (L.)
Thouars) ont débuté en station expérimentale. Le voandzou est une lt5gumineuse à graines
de haute qualiti nutritive (les graines contiennent 17% de protkiies), r&&ante aux
maladies et aux insecks, et parfaitement adaptée aux conditions semi arides des rkgions
tropicales (NA& 1979). Le voandzou est cultivé au swkst du Séneg+ dans les petites
exploitations agricoles, seul ou en association avec une c&%le.
Ce prkent rapport dkrit les premiers rkwltats obtenus sur l’effet de l’inoculation
avec des souches de Rhiwbium préalablement s6kctio~6es (Guéye et Bordeleau, 1988)
sur le rendement en grains de deux variCt&s de voandzou.
Matbriels et mbthodes
Un .&sai d’inoculation du voandzou au champ a &5 conduit en station e.xp&-@e&e
au sudest .du %n&al & Sinthiou h4aRme en 1986 sur lé %e A (pluviom&rie : 794 km)
et en 1988 sur le site B @luviom&rie : 873 mm). Les caractkistiques physicochimiques
dti sol des deux sites sont indiquées au Tableau 1.
. c
,. . .
Tableau 1
: Caractéristiques physico chimiques du sol de Sinthiou Ma]$c
-~
wm
. .
Comppsifion (5;)
,.
Passimilabk '
Sites
C total
N total
P total- fOlsen*) Argile Limon Sable pH (H20
---y
-
-
-
-
A
4380
450
557
156,5
4,8
?,5
97.7
6.40
E?
4540
470
402
10995
4,3
2,3
93,4
6,64
*Olsen e t .a?< (19541
,_,_,
‘,
.>.
’
”
;
hxirniser la FBA pour la production agricole et forestière en Afrique
.
:1.
<. .’
i
2~. . . : 2.,
.’ St S:*I!, ~ “,’
_;,
‘“:x
‘2.
Effet de l’inoculation avec des souches
de Rhizobium et de 15, fyC$sation azotée
sw;. i&‘A,2dé&&ti *é;;;dgrains
” “,
du Voandzou (Vignci subterranea (L.) ‘Thouars)
.._
au Sénégal
GUEYE, M
.
MIRCEhKNRA, BP. 53. Bamhq. S&dgal
:
_.._I’
I-..
.
_
_.._.
-
_____
.-
.--,,-,.I1._
_-.
I
..
--
:,;y,
.-
.,‘.
RBsunlé
Desttssaisauchampontétécondllitc PU S&+~~RI m !Wf;ct 10PR c-T~T~~:P- r..~L-:z:.~~:z’.:
pour étudier la réponse du voandzou (Vigo subterranea (L.) Thouars) à l’inoculation
avec deux souches de Rhizobiwn : MAO 113 et TAL 2.2. L’inoculation des deux
vari&% étudiées a augmenté leur rendement en grains de 27 à 58%. L’amélioration
de. rendement la plus significative (58%) a kté obtenue en 1986 avec la variété
83-131 inocul6e avec la souche TAL 22 En 1988, l’ion avkc la-rsouche
MAO 113 a amklioi le rendement de la vari&? 79-l de 52%.
Introduction
L’inoculation des légumineuses A graines avec des souches de Rhizobium n’est
pas une pratique courante au SMgal bien que ses effets btkéfiques y aient été mis en
évidence sur le soja en particulier. L’inoculation du soja au champ avec la souche de
Rhizobium USDA 138 conditionnée dans de la tourbe a augmenti significativement le
rendement en grain du soja @archer et al., 1984) ; au laboratoire, Dommergues er al.
(1979) avaient egalement obtenu une augmentation tr&s significative du poids sec des
nodules et de l’azote total des parties aériennes du soja inocul6 avec la souche de Rhizobim
USDA 138 incluse dans un polymkre de polyacrylamide.
Deux va&&-de voandzou (79-l et 83-131) ont 6t6 inocuEes avec deux souches
de Rhizobiwn : MAO 113 isol6e du voandzou au S&&gal (Guèye et Bordeleau, 1988)
et TAL 22 fournie par le projet Niftal à HawaiL De plus, l’essai comportait deux traite-
ments supplémentaires sans inoculation avec une souche de Rhiwbium : M traitement
consstantenrmeapplicationd’~~azo~a~de50kg~etuntraitement
sans azote. L’essai a 6t.6 conduit en blocs compl&tement randomisés avec quatre @&itions.
Chaque panAle Unenta& (2,4 m x 15 m) avait reçu une fertil&tion de fond de
60 kg P,o/ha sous forme de supertriple et de 120 kg KCJIha. L’inoculum a 6t6 apport.&
au sol sous forme de tourbe (5g/poquet) contenant approximativement 109 Rhizo-
bium/g. Le voandmu a été sem6 à raison d*,F”gr@e par poquet selon un Ment
de 30 cm entre les lignes et 15 cm F .la ligne.
Le rendement en grai& du voandiou a & Cvalué à la &olte. Les anaIyses statis-
tiques ont &fZ effectu&s en utilisant le test de .Duncan (1955). .
.I i .
.
:.
’
;. . .
.i.,
_,
Rb~Rats
e
t
dls&sslon : -
:
:
:’
En comparaison avec les parcelles t6moins (Tableau 2). le rendement en grains
du voandzou n’a pas été significativement affect& par un ap~~~rt de 50 kg ur6e/ha bien
que sur le site B en 1988, il.y.aït eu 30% d’augmentation avec la vari&6 83-131. Par
contre, l’inoculation du voandzm avec les souches de Rhizobium MAO 113 et TAL 22
a augmenté le rendement en grains des vari&?s 79-1 et 83-131:En 1986, l’inoculation
avec lasoucheh4AO 113 aaugmenté IerendementdelavarH 83-131 de51%; l’inoculation
aveclasoucbeTAL22aaugmentélerendementdesdeuxvaritWs79-1
et83-131 respectivement
de 50 et 58%. En.1988, l’inoculation avec la souche MAO 113 a am6lionZ le rendément
des vari&+ 79-l” et 83-131 ~tivement de 53% $ 33% ; avec la. souche TAL 22,
les m~@at$yq@ * .+$%&r la varit% .:ZJ-llet 27% :F ~puZt6 X3-131:
Ces &ultats. montrent que ,la difE.rence g&&ique entre les deux va&& uti&s
influence fortement la r@onse ‘au champ du voandzou a l’inoculation avec les souches
de Rhizobium : la Aponse de la variété 83-131 est plus stable que celle de la vari&&
79-l. Cela était pr&isible car une exmence précédente avait mont& que l’indice de
nodulation de ces deux variétés révClait une symbiose plus efficiente de la variété
83-131 en association avec les souches MAO 113 et TAL 22 (Guèye et Bordeleau, 1988).
Ces résultats confhment donc les travaux de Dadson er al. (1988) qui avaient mis en
évidence l’influence du g&otype des variétés de voandzou à l’inoculation avec des souches
de Rhizobium. L’effet bénéfique de !‘,noc@tion~,+~~ 1~. ,so+~ de &$yrhfzobium
sur
le rendement en grains.du &andzou était @lemenbpr&i@ble cardes &des antieunzs
(résultats non publ&) avaient montré que le sol de sinthiou MalémeÏ&ferme t.r& peu
de Rhiwbium (c 2OO/g) et que le voandzou a souvent besoin d’&re inoculé (Johnson,
1968). On pourra3 donc recommander l’inoculation du voandzou dans les sols du StMgal
renfermant t&s peu de Brdyrhizobizm. Cependant, on doit garda pr&ent A l’esprit le
fait que l’efficience d’une association Rhizobim x plante-hôte varie considérablement
d’un site 2 un autre.
3 7
:
f
‘..;i..
.:
:
&fF-.?,, -:>.:L2<- ,*..a. -’ J,.’ 1.. . . - .
.; .:. i.-~~~~~~~~~~~~~~~~
;<,, : .-
;---
.<
__c-
-----
----..
-
-.__-----
- .
-.
.-----
---
I
.Tableau 2 : Effet de-l'azote et-de-l'inoculation avec les souches de
BradyrhizobiÙm:yD:-113 et TAL 22 sur le rendement en grain
du.vo?uizou ~varj.~t~-79-1 et 83-131.) cultivé au champ, à
]
-
la station ~e&&'iÏn&ale de Sinthiou MaliMe.
l
:
_-
I
.-
.: .:
:
,
:
:.-:/:
:.
~ Y.
.,
.1-p.. - 3.
Rendement en.grain- {kg/ha)
/..,.
.'.
*
Varietés: Traitements
Site A (1986) Site B (1988)
. . ;.
_-
.3062,50 b‘
Témoin
'
3080,30 Ii
3555.40 b
Azote (50 kg urée&)
3832;90 ab
4624,lO a
r';r?"
_I
."_
,.~
-->
TAL 22
4863,90 a
4524,40 a
'... Daqs.:chaque colonne et?poÛr.chaque variété les valeur suivies d'une:
'
.
:: &jnê+:&'ttr~ ne ront.jign~~iCaMveiaent- diffbentes 'd'après. le test de
ii.- ~Duncan.:31955) '
a
u
seuil:d$~,5 ?h:- ',: i
..r_
A;
_ I
-- L. U-L I_ ___ --._ ..1.- -.a--
.y,
_ --~i...,.-eT
._ ::.-:
_-
3 8
Remerciements
Ces travaux ont étk réalises d’une part grâce au support financier & 1’U.S. BOSTID-
NAS par la bourse no BNF-SN-2-84-21 et d’autre part grâce au financement du projet
UNEP-FAO CP/FP 6106-84-02. L’auteur remercie Alioune Gning, Cheikh Samb et Omar
Tour6 pour leur assistance technique.
Bibliographie
DADSON, R.B., BROOK, C.B. and WUT.OH, J.G. 1988 Evaluation of sekcted rhizobial
strains in association with bambara groundnut (Voandzeiu subferreneu (L.) Thouars).
Trop. Agric. flrinidad). 65, 254-256.
DOMMERGUES, Y.R., DIEM, H.G. and DIVIES, C. 1979 Polyacrylamide-entrapped
Rhizobizm as an inoculant for legumes. Appl. Environ. Microbiol. 37, 779-781.
DUNCAN, D.B. 1955 Multiple range and multiple test. Biomefricu 11, l-42.
GUEYE, M. and BORDELEAU, L.M 1988 Nitrogen fixation in bambara groundnut, Vound-
zeiu subterruneu (L.) T~OUTS. Mircen J. 4, 365-375.
JOHNSON, D.T. 1968 The bambara groundnut : a review. Rhodesiu Agric. J. 65,
l-4.
LARCHER, J., WEY, J. et GANRY, F. 1984 Recherches sur le soja de 1978 a 1983.
Rapport ISRA, CNRA Bambey, 86 p.
NATIONAL ACADEMY OF SCIENCES. 1979 Puises. In : Tropical Legumes : ressources
for the future. pp. 47-53. National Academy of Sciences, Washington D.C.
OLSEN, S.R., COLE, L.V., WATANABE, F.S. and DEAN; L.A. 1954 Estimation of
availablephosphate in soils by extraction with sodium carbonate. Circzdur of US DepurtmeM
of Agriculture No 939.
FIXATION BIOLOGIQUE DE L’AZOTE
CHEZ LES LEGUMINEUSES FOURRAGERES
Uaximiser la FBA pour la Production Agricole et ForestiZre en Afiique
Nitrogen fixation of the forage legumes
and their -residual effects on wheat growth
and yield
Soils rmd Plant Nutrition Section, h&national Livestock Centre for Af&a (ILCA),
P.O. Box 5689. AaZis Ababa. Ethiopia
Abstract
The symbiotically fixed nitrogen of various forage legumes was evaluated for two
seasons un% upland soil conditions in Ethiopian highlands using ‘5N technique and
Avena sativa (oats) as a non nodulating reference crop. Forage oats gave higher
dry matter yield than Vi& darycxzrpu (vetch), however, the latter gave feeding stuff
of highex N concentration and yield (169 kg N/ha&ear for vetch vs. 73 kg N/ha/
year for oats). In 1985. 78, 78 and 73% of the total N in vicia, clover (Trifoliwn
steudneri)andmedics(Mcdicagopdymorpha)p;iginatedfnrm~~.Vi~medics
and trifolium fixed 131, 77 ‘and 61 kg N/ha nz&uively. In 1986. the dry matter
and grain yield of wheat following vi&, medics and trifolium was significantly higher
than that following oars. Between 12 and 31 kg of additional N was taken up by
wheat straw and grain following legumes and fallow as CompBTed with those following
Oats.
Introduction
Of the many possible nutrient deficiencies, lack of N imposes the most widespread
and strongest restrictions on plant and animal production in sub-Saharan Africa (Nnadi
and Haque, 1988a). While the N status of soils can be improved by the addition of N
1:
.
.
:
^- ; : ,-
41
fertilizer, it is expensive input and is reflected in its low consumption in the region @‘DC,
1986).
In Ethiopia, the average yields of most of the cereals are below 1000 kg/ha. In
order to improve the productivity of feed and food in various production systems and
to sustain it at reasonable level, there is an urgent need for alternate sources to increase
the influx of N in the system. Use of legumes in various production systems to benefit
from biological nitrogen fmation @W) contribution to following cereal crops may attract
quick adoption by resource poor farmers. Legumes also contribute to soil productivity
by improving soil structure (Andrew, 1965 ; Stoveman, 1973) and can control pests and
diseases of cereals in legumecereal rotations (Moore et al. 1982).
Owing to the multiple facets of the contribution of legumes to cereals and the
difficulty of isolating these factors, the contribution of legumes is often studied by measu-
ring the yield of following crops (Muss and Burhan, 1974 ; Papastylianou et al.
1981 ; Singh et al. 1985 ; Ayoub, 1986 ; Osman and Nersoyan, 1986 ; Mohamed-Saleem
et al. 1986 ; Papastylianou, 1987 ; Papastylianou and Samios, 1987 ; Keatinge et al.
1988).
This communication reports BNF by annual forage legumes and their residual effect
on subsequent wheat crop on an upland soil in Ethiopian highlands.
Materials and methods
Climate
Rainfall, potential evapottmspirattion (PE) and water balance.
The rainfall exhibits strong scasonality and of the total annual rainfall of 866 mm, 84%
is recorded in the rainy season extending from June to September (Fig. 1). Short season
is from March to May, when a total of about 140 mm rainfall is received. PE is 1320
mm (Fig. 1) and values are lower (2-3 mm/day) during the rainy season as compared
with short rainy season (34 mm/day).
The water budget (Fig. 1) shows that in the months of July and August there
is a surplus of about 300 mm of water. Daniel (1977) reported that potential runoff at
Debre Zeit, in an average year, is likely to be 100 mm which could be collected and
reuse for crop and livestock production in the post rainy season.
Ten?perature and radiation. Temperature and radiation are two factors
that limit crop and livestock production in Ethiopian highlands. Mean maximum, minimum
and average monthly and an&al temperatures. are shovvn in Fig. 2 Relative sun&ine
hours are about 5O%‘d&ng the rainy months and are 70 to 80% during the rest of the
year. Total radiation is around 500 cal/cm2/day and somewhat lower during the rainy
months (300450 cal/cm2/day) and higher in the dry period from February to May (500-
550 cal/cm2/day).
Soils
Debre Zeit is situated 8” 44’W, 39” 02”E at an altitude of about 1850 m above
sea level, about 50 km south-east of Addis Ababa in Ethiopian highlands. The soils are
mainly vertisols and some alfisols. The experiment was conducted on an alfiiL
Along the toposequence, alfisol, soil with vertic properties, and vertisol stored 113,
187 and 278 mm water per meter depth respectively. During the maturity stages of forages
:
,.
.l‘i..‘;
;.::..t,
:
‘.
._ ,.... ‘> .,., i,;:
,’
:
.
.
._.
./
42
Debre Zeit, Ethiopia
8044'N. 3g002'E, 1850 as1
e Rainfall
m m
- - - - pE
240
Annual roinfoli - 866mm
A n n u a l P E
- 1319mm
1 6 0
A
--A-,
/
/
C-
I
i’
-8C
Figure 1:
Rainfall,
pan
evapotranspiration
(PE) and water
budget at Debre Zeit for five years.
. .
43
Debre Zeit. Ethiopia
as1
30
8O44'N, 39O02'E, 1850
Maxlmum
26.0%
;3
0
-g 20
A-g=. - _ - ,
-.
/.A’
---2.
.
-.-. -._.
18.7%
3
I-
-
-
-
0
I
I
I
I
I
I
I
JFMAMJJAS () i D Annual
Month
Figure 2:
Maximum, minimum and average temperatures at Debre
Zeit.
“.
‘/‘...
and crops, xrtisol stored 57% more moisture than alfisol and remained moist throughout
the year below 50 cm depth (Kamara and Haque, 1988). Some of the physico-chemical
properties of experimental soil are given in Table 1.
Agronomy
A total of five treatments were tested along with a &llom and the non-nodulating
reference (oats) treatments (fable 2). A randomized complete block design with four re-
plications was adopted. The plot size was 6 x 4.50 m and there were 2 and 1 m paths
between replications and plots respectively. AU the legume treatments as well as non-
nodnlating stada-d crop recdd 20 kg W F+ _~r~~zq%nting~ except in- ~mplots.
Within each of these plots, a n&oplot of 2 x%!5 rir rcce$d 20 kg N/ha of 5 % 15N
a.e. in the form of urea solution.
All these treatments received a uniform application of 30 kg P/ha as triple su-
perphosphate. Medics and clover were planted at 25 ,cm apart at a seed rate of 10 kg/
ha.Vetchwasplantedat5Ocmapartat.25kg/ha.whileoats,25cmapartat1OOkg/
ha. The experiment was plantcd on July- 11, -1985 and 2Okg N/ha of 5 96 ‘%I ae. urea
was applied on July 15, 1985.
Plants were harvested 98 days after planting (DAP) for dry matter estimation and
1 mz was sampled from ea$.of ?-%zlabelled microplots
,.:;.: ‘Z.. :..<. .,-,\\t-T:., .j
.‘..L,< & :: (1 ‘r
for total N and %I analysis. The
praporpon of ~~t.N .&.g?&..m ,~bi~tic,N.~~~~.~:;~~~‘to
the
procedure given by Hardarson (1984). The legume residues were removed from the plots.
The following year, wheat cv. Buhe at 25 cm apart was planted on 18 June, 1986
atseedingmteof1OOKg/ha.TwoNrates(Oand6OkgN/ha)intheformof~ureawere
superimposed on each plot before planting. Wheat was barvested from 2 x 2.5 m at
113 ,DAP for dry matter and grain e&nation. Composite samples of straw and grain
I.
w&eanalysedfortotalNtoestimateNuptake.-
I*.
‘{ 2. I 7. p ;f ,7 f *:: I-:.. ;:‘T i.
: . .
:
.;.
.-
,_:..
.:
:
45
’
. Table'1 : Physico-chemical properties of experimental soil
Properties
Values
‘_I
.’
.,...
“
.:
‘-
5,
..
s.
Sand
(%),
49
Silt 1%)
19
j 'Clay 1%)
32
pH (H$I)
.
-: b;iganic matter (%)
‘;_ 6.!
.i' :I :
.;
,. 2 . 2
Nitrogen (%)
0.05 :
Available P (Bray II - ppm)
7
Exch. cations (me/100 g)
K
0.65
Ca
13.20
k.l
9.76
N
a
0.41
I
.’
rs
.I
:’ : . . : .‘.
.. 46
Results and discussion
Dry matter and nitrogen yield of legumes or oats
There were no signifi--t diffpk am&ig,.pats, yetch and medics with respect
to dry matter yield ‘and thi: l@~&%*ced $0&x .dry: mattex than oats (Table 2).
Si~tlyhigherNyieldwasob~~frdin~~asw’nparedwithothertreatments
and there were no difference$$$J:jyields among the others (Table 2). Papastylianou
and Samios (1987) reported that wooly p&l vetch gave higher N concentration and yield
(86 kg N/ha&ear for vetch vs 55 G N/ha per ye&) ‘than barley.
:,
:,
6lologlcal nirogen fix&n ! : _
Table 2 shows the 15N a&?&derived from &xation and N flied by various
legumes l%J enri&ment of oats w& si&ificantl~higher~as compared with that of legumes.
However, there was no significant differences among the legunies though lowest ‘W enrichment
was noticed in vetdh (Table 2).
Nitrogen derived from ftion varied fi-om 73 to 78% in three legumes and no
significant differences among various.trealments were observed (Table 2). Total N fixed
was 131,77 and 61 kg N/ha in vet&me$ics and clove resptively. Vetch significantly
fixed hi@@,, as ‘Compared ivit6:&ve&(Tabl~Table; ?). .&&~ti&e,& al :.(1988) reported that
.;,&m*:;:-$ i‘..
vetch’-m .#). to 4g tig Nb at ,dy md Td &ii .g$-v~+&y & No*m syria
They further observed that vetch fixed do$Ie the amount of N in kiproved management
as Compared with traditional .qp, Higher N fixation by vetch in Ethiopian highland
conditions might be due io mor( reJiab1~ rainfa& prkence of indigenous rhizobii and
higher fertility of upland soils as timpared with no&&n Syrian’conditions.
-;.. *
- :
i
. Residual &f&t bf &g&i& ,$ $#$$
i
ic
i’
.py “-..’ j.
“u ::.
Dry matter atid gn$e y@!~.The+nxid~.effem oQ985 season cropping
treatments on dry matter and grain yield of wheht at two N rate~-ire given in Tables
3 and 4. The dry matter yield of whe~~~following v&h was significantly higher than
that following oats or natural faIIow. However, no significant difference in dry matter
was ohserved -following v&h, medics and clover.
The grain yield of wheat following ‘vetch was sigriificantly higher than that following
oats. No significant difference in wheat grain yield was observed in;plots previously cropped
to, the legumes. Yields in fallow. * oats plots w? jthe lowest Fable 4).
s. r:
-2.
e s
., .)
..%.s ; 1
.*
::
“ .
i
Table 2 : Dry matter yield, total N yield, percent T5N atomic excess, percent N derived from
fertilizer, N fertilizer yield,
Ndf fixation and N fixed by forage legumes,
Debre Zeit, 1985.
Dry matter
N yield
15N
Ndf
N fert
Ndf N
Treatments
(kg/ha)
a.e
fert.
yield
.'fix
fixed
non
Isotopic
%
%
(kg/ha)
%
.'
(kg/ha)
' :
Isotopic
;
----
Vicia dasycarpa
4531b*
49DOa* 169a*
O.O99b* 1.852b* 3.015b*
77.93a* 13la*
Medicago polymorpha
4163b
39D;Oab 106b
0.123b 2.306b 2.425b
72.73a
77ab
Trifolium'steudneri
2150~
2625b
78b
0.102b 1.903b 1.475b
77.93a
61b
Avena sativa (oats)
5813a
5325a
73b
0.471a
8.827a'
6.300a
-
-
*Values withincolumns followed by the same letter are not significantly different at 5% level according
to Duncan's Multiple Range Test.
I
UJ 0
Y
\\ -c
(7,
ca
I
I
u ccr 2: 2: L h 7-u
48
L aJ
- IA
!
_
--
-
.--
.._
Table 4 : Effect of previous cropping and nitrogen application
on grain yield of wheat, Debre Zeit, 1986. .
Previous cropping
kgN/ha
1985
-
Vicia dasycarpa
2612a*
2810a*
Medicago polymorpha
2531a
2280bc
Trifolium steudneri
2600a
2661ab
Fallow
2090ab
2046~
Avena sativa
1753b
2119c
*Valueswithin columns by"the same letter are not significantly
different at 5% level according to Duncan's Multiple Range Test.
;.,
,,
.
,l_
.-
.,
I
( /
I,
t
.:\\
:..
I,.
:
:
:
j,
t 1 I ,,.. ;,rz+>-
. . ‘,. . . .
.
.
., i..;<.. : .:l ,. i :
.,,
:
,.
.
:
.,
50
Nitrogen yield of wheat straw and graln. In this investigation, we
wish to quantify the residual benefit in N derived from legume crops. It can be seen
from Fig. 3 that between 1 and 8 kg/ha of additional N are taken up by wheat straw
following legumes aud fallow compared with those following oats. Pmceding vetch crop
resulted iu 53% higher N uptake by wheat as compared with amtroL
ItisclearfromPig.3thatbetween
11 and23kglhaofadditionalNaretaken
up by wheat grain following legumesand fallow compared with following oats.
Vetch
produced 58% higher N uptake as compared with control. Nnadi and Haque (1988b) reported
that vetch species could corm&me .a considerable amount of N to following crops even
when tops are not~retumed to the:‘&iL Minemkation of root N occured in vetch species
where root N content exceeded 296.
-..
Conclusion
.-
The results of the present study could be of importance wherefarmers cannot afford
to apply fertihzers which may often be unavailable, The choice ,+ to which is the most
beneficial forage to be grown should depend on feed production, nuuitive value for animals,
N economy and rotational value for following crops.
.:, : ‘z
Recommendation for cereals in Ethiopian highlaudsjs ;2l&l$j;.j (EPID, 1975a).
OFF, @Js y$ ~XXEJS on +mf=’ fields W
show np t.0 80%
increak in yield over current recommendations with applicatioqof q.kg N/ha (Christensen,
1973 ; EPlD, 1975b). The amount of N fixed and Contributed by vetch used in this study
could raise substantially the yield of cereal crops and livestockproduction on upland
soils ill Ethiopian highlands.
::
.-:
Acknpwledgements
.
.-I
-..
‘Ihe author express his appreciation to Mr,,Peter Ibara, Ato Yirsaw Wubete, Ato
Muhommed Yasin, Wzt. Abeba G&tom. and WzC Tsehay~Aberra and other staff of Soil
ScieIlceandPlantNuhritionSectionf~technicalassistance,andtotheSel’beEsdorfL,
International Atomic Energy Agency, Vienna for the ‘?N analysis.
:
_
.
.
._
.:
.,.
.
_’
..
,.
;.i,..
.!
”
.:
Straw
Grain ',
I
I
fzzrzm-
N yield (kg ha')
100-r
V.D = Vicia dasvcarcta
M.P = Medi,cago boly'morpha
,.
“..
.
. 1
1 j
= Avena sativa
60
40
20
0
Pb.Fceding crops
Figure 3:
Effect of preceding Crops on nitrogen yield of wheat straw end grain
in unfertilized plots, Debre Zeit, 1 9 8 6 .
51
References
ANDREW, E.D.1965 The effect of some pastme species on soil structure. Aust.‘.T. Expl.
Agric. Anim. Hush. 5. 133-136.
.:
AYOUB, A.T. 1986 The potential contriiution of some f&age legum&oth~~nitrogen
budget and animal feed in the Sudan Gezira farming system. -In : Potential~of forage
legumes in farming systems of sub-Saharan Africa.. Eds. Hague, L, Jutzi, .S.C. and
Neate, PJ.H. p. 5868. Pmczdings of a conference held at ILCA, Addis Ababa,
Ethiopia, 16-19 September, 1985.
.:
G&, . ,/.1 I ., I .,,i;
CHRISTENSEN, II. 1973 Fertihxr and variety triaIs~&idfdemonstmtions:in~Ethiopia,
’
197241973. Extension and Project Implementation deprtment (EPID) Pd. No. 10.
Minstry of Agriculture, Addis Ababa, Ethiopia.
DANIEL, G. 1977 Aspect of climate and water budget in Ethiopa. Addis Ababa University
Press. 71 p.
EXTENSION and PROJECT IMPLEMENTATION DEPARTMENT @PID). 1975a Hand-
book for Agronomy of Crops. EPID Pd. No. 10. hdinistry of Agriculture, Addis
Ababa, Ethiopia.
:
^
/
EXTENSION and PROJECT IMPLEMENTATIONDEPARTMENT@PID). 1975bResuhs
,. .; ‘?“.<’ ‘I, j:.7i. i *’
of EPID PubL No. 31. Minis&y, of Agriculture:~ ~~<~Abjb$~~Ethiopia$$?~~~-
HAQUE, I. and TOTHILL, J.C. 1988 Forages and pastmes-in mixed &mingsystems
of sub&haran Africa. In I Land management and managementofacidsoilsinAf.iica.
Eds. Latham, M. and Ahn, P. p. 107-13 1. Pnxxdings ofthe Second Region&Workshop
on Land Development and Management of Acid Soils in Africa held in Lusalm and
Kasama, Zambii 9-16 April, 1987.
HARDARSON, G. 1984 Biological nitrogen fixation-of grain-legumes : The use of =N
methodology to assess nitrogen fixation and guidelines’ for improvement of nitrogen
fmtion in grain legumes. FAO/IAEA Agricultural Biotechnology Laboratory, Sei-
bersdorf, Austria.
INTERNATIONAL FERTILIZER DEVELOPMENT CENTER (IFDC).1986, Fertilizer si-
tuation : Africa. IFDC Special Publication, Alabama, USA.
KAMARA, C.S. and HAQIJE, I. 1988 Soil moisture storage along a toposequence in
the Ethiopian highlands. Iti : Management of vertisols in sub-Saharan Africa Eds.
Jutzi,S.C.,Haque,I.,McIntire,J.andStares,JE.p.183-20O&oceedmgsofaconference
held at ILCA, Addis Ababa, Ethiopia. 31 August to 4 September, 1987.
KEATINGE, J.D.H.. CHAPANIAN, ‘N. and SAXENA, MC. 1988 Effect of improved
management of legumes in legume-cereal rotation on field e&mates of crop nitrogen
uptake and symbiotic nitrogen fixation in Northern Syria. J. Agric. Sci. (Cumb.). 110,
651-659.
MOHAMED-SALEEM, MA, SULEIMAN, IL and OTSYINA, RM. 1986 Fodder banks
for pastorahsts or farmers. Zn : Management of vertisols in sub-Saharan Africa. Eds,
Jutzi. SC., Haque, I., McIntire. J. and Stares, JE. p. 420-437. Proceedings of a
conference held at ILCA., Addis Ababa, Ethiopia. 31 August to 4 September, 1987.
MOORE, KJ., HERRIDGE, D.F., STARR, G. and DOYLE, AD. 1982 Evidence for
disease control as a factor in improved wheat yield in a 1upincereaI rotation. Proceedings
of the 2nd Australian Agronomy Conference, Wagga, Working paper. p. 327.
.:
:
z.
,:
.I_
:.
..),.’
;
....-..
j_
..,
,: ._: . . . .
:
52
h4USA. UM. and BURHAN, H.O. 1974 The relative performance of forage legumes
as rotational crops in the Gezira ExpZ. Agric. 10, 131-140.
NNADI, LA. and HAQUE. I. 1988a Forage legumes in African croplivestock systems.
.,;%r JL(X B~u&.$$: IO+: ._ :.
NNADI, LA. and HAQUE, I. 1988b Root nitrogen transformation and mineral composition
in selected forage legumes. J. Agric. Sci. (Camb.). 111, 513-518
OSMAN, AE: and NERSOYAN, N. 1986 Effect of the proportion of species on the
yield and quality of forage mixtures and on the yield of barley in the follckng year.
:Expl..tyl@;~22;z:r;5-351;,
PAPASTYLtiOU, I. 1987 Effect of preceding legume or cereal on barley grain and
nitrogen yield. J. Agric. Sci. (Camb.). 108, 623-626.
PAPASTYLIAN~~, I. and SAMIOS, Th. 1987 Comparison of rotations in which barley
for grain folloqrs wollypod vetch or forage barley. J. Agric. Sci. (Cumb.). 108, 609-
6 1 5 .
PAPASTYLIANOU, I.,‘lkXRIDGE, D.W. and CARTER, E.D 1981 Nitrogen nutrition
of cereals in short term rotation. I. Single SeaSOn treatments as a source of nitrogen
‘.
for sul@eq~$~$&xeal ,xrops. Au&. J.’ Agric. Res. 32, 703-712.
<.
S~G&C&$&;f&$& RB: 1985 Contr&tion of legume & he fern nitrogen
economy in a maize based cropping system. J. Agric.Sci. (Camb.). 105, 491-493.
STOVlZ+AN, T.C. 1973 Soil structure changes under wheat belt farming systems. J. Dept.
Agric. K A&. 94, 209-214.
:
1,
FIXATION. BIOLOGIQUE DE L’AZOTE
CHEZ LES ARBR+ES
.
hkhiser la FBA pour la production agricole et foresti&e en Afiique.
Effect of the addition of phosphorus
and Rhizobium inoculum on the nodulation
and the growth of two species of gum arabic
trees Acacia Senegal L. Willd
and Acacia la&a R. Br. ex Benth
BADJI, SP, THOEN, DP,
DUCOUSSO, M.0’ and COLONNA, J.P.a’
mLabo&oire & Microbiologic des Sols, ORSTOiU B.P. 1386, Dakm, Sknegal ;
m Fondion Luembourgwise ;
w ISRA:DRPF, BP 2312. Dakar, Senkgal.
Abstract
Four trea$~ were applied to seedlings of Acacia senegal and Acacia laeta
grown in containers filled with sandy soil poor in available phosphorus (11 ppm.).
Compared .to the con@ Rhizobium inocdum treatmmt induced rm incrwxq not
always statistically s&&ant, of the measured parameters of growth and nodulation.
Thcfoliarandnodulebiomasses
of Acacia Senegal were increased upto 132% and
283% nsptctively. and upto 21% and 66% for Acacia &eta
‘;i.
Rhizobium~andaddirionof30or60ppmPenhaficedvays~~.
antheparrrme~ofgrowthandnodulationoftfietwogumarabicspacies.Inthc
ease of Acacia senegal, the addition of phosphorus W the growth 7
~5~to2oo4b.tbefoliarbiomassabout4oo46.thenodulatianptaametcrJfrom
1100% to 1400%. In ampmkm, forAcacia kzeru, the addition of phosphmns increa-
sed the growth parameters from 40% to 131%. the folk biomass about 15?% and
the nodulation parameters i?om 250% to 350%. However, there was no difference
between these both treatments. Addition of 30 ppm P per plant seems to be suffi-
cient to assure optimal response.
Further research is planned to determine the complete response curve of both
trees to phosphorus in the presence or absence of Rhizobium.
.’
._..
.‘:
.-‘.
55
Introduction
Among the roughly one thousand acacia species (New, 1984), Acacia Senegal
was used, like many other woody leguminous trees, to reforestation in the Sahelian zone.
(Badji et aI., 1989).
The Sahalian ecosystem has been severely degraded due to the combined effects
of an extended period of drought and the excessive exploitation of natural resources by
man and his livesto&. The nahd population of Acacia Senegal has decreased dramat-
ically. This species is an ecologically beneficial tree because of its important role in the
reconstitution of the ‘vegetative cover of the zone.
Acacia senegul is able to fix atmospheric nitrogen through rhizobial nodulation
(Basak and Goyal, 1980 ; Dreyfus and Dommergues, 1981), and hence may contribute
to the improvement-of the soil. Found natumlly throughout the semi-arid Sahel, this species
seems to be well adapted to the edaphic and climatic conditions of the region (Giffard,
1975). The tree has many useful products for local populations, such as aerial fodder,
firewood, wood for tools and construction. Moreover, the tree exudes gum arabic, a product
of considerable economic value.
It may lx possiile to introdue the Acacia laeta in Senegal, onother gum arabic
producing species commonly found m t$ Sudan ; the major gum arabicexporting countries.
Because a pdsitive effect of p;hbSjhoruson Acacia holoseticeaand &zcia ra&&na
has been shown earlier in Senegal (Comet and Diem, 1982 ; Cornet er al. 1982). the
aim of this study is to determine the influence of the addition of phosphorus with R&o-
bium inoculum on the nodulation and the growth of these two acacia species grown on
sandy phosphorus poor soils found in the SaheL
Materials and methods
The seeds of Acacia Senegal belonged to the Mlight grey>, phenotype (Dione, M.
pers. comm). As with the seeds of Acacia &ta, they were furnished by the acDirection
des Recherches sur les Productions Forestieres~ (DRPF) of the &stitut Senegalais de
Recherches Agricoles~ (ERA). The seeds were putted in concentrated sulphuric acid for
14 mn, rinsed thoroughly in sterile water, pregerminated in sterile vermiculite, and after-
wards transferred into individual containers filled with’soil sampled from sand dunes Ioca-
ted near Kebemer. This soil has a sandy texture (94,3% sand) and a low available phos-
phorus level (11 ppm,). Each container was filled with a 1.5 kg mixy of soil and polys-
t y r e n e m (2..,‘v/v)s
:: .3F& :,“i..’
..,
- The factorial experiment, with randomimtion within the blocks~ had three replicates
.and three contmlled factors. The first factor is species with two levels : the Acacia senegal
and the Acacia ha. The second is soil sterihzation, also divided into two levels : non-
sterilised soil and autoclaved soil (120°C 60 mn). The third control factor is fertilization
with four levels : 1) no inoculation .; 2) rhizobii inoculation ; 3) rhizobial inoculation
plus 30 ppm of phosphorus ; 4) rhizobial inoculation plus 60 ppm of phosphorus. Phos-
phorus is supplied as KHQO, in single application. Each elementary plot contained fours
plants. The mean values for each plot was used in the statistical analysis.
At the transfer to the greenhouse, the seedlings wexe inoculated with 2 ml ofRhizobium
suspension (109 cells/ml). A fast growing strain (ORS 1007) selected previously Q3adji
er al., 1988) was employed. It was isolated from nodules of Acacia Zaeta and it is effective
on the two acacia species. The cross nodulation was demonstrated (Badji et al. 1988).
56
Twenty five weeks after planting (WAP) nodulation and growth were recorded
using the following parameters : diamqr at the collar (A); height of the main stem (B);
total length of branches (C!); number of shoot nodes (D); dq biomass of the leaves Q;
number (Fj and dry weight (G) of the nodules.
;-
:
I
.,
R e s u l t s ahd’,‘d&ss& ’ ”
I
Species effect ‘.
Table, 1 shows that .dming :the fkst months of growth and development, Acacia
ZaetuandAcaciasenegalpresentedsomesignificantdifferenceswith~~totheparameters
A, B, C, E, and F. One non significant increase occm-ed for number of shoot nodes (D)
and dry weight of nodules (G).
Table 1 : Differences in growth and nodulation between 25 weeek old seedlings
of Acacia Senegal and Acacia Laeta
i”
Measured Parameters*
Growth
Nodulation
SPECIES
‘,. A.
6
ci
:c
D
E.
F
G
A'!- - - - .~
Acacia
4.85
24.50 40.90 -35.38 1558
14.06
307
Senegal
b
b.
b
-
b
b
-
(AS)
Acacia
5.61
35.85
63.25 38.17 2356
17.09
309
Senegal
a
a
a
-
a
a
(AS-AL) 100
e----------_
--t16
t46
t55
+ 8 +51
+22
+l
AS
. ..
* per plant, A = collar diameter in mm ; B = height of plant in cm ;
C = total length of stem and branches in cm ; D = number of shoot nodes ;
E = dry weight of leaves in mg ; F = number of nodules ; G = dry weight of
nodules in mg. For each parameter data in each column without letter or
followed by the same letter are not significantly different (P f 0.05 -
Newman Keuls test).
57
Soil sterilization effect
For both species and for all growth parameters (A to E), the values were higher
on non-sterilized soil than on sterilized soil (Table 2). However, these differences are
not significant for the collar diameter (A) and the height (B). They are significant for
total length of stem and branches (C) and, in the case of Acacia Zueta, for the number
of shoot nodes (D) and the dry folk biomass Q. The differences were not significant
for the number of nodules (F) and for the nodule dry weight (G). To conclude, the effect
of soil sterilization was rather depressive but not significant (Newman and Keuls test,
p = 0.05). Soil sklixation lolled the native rhizobium and the endomycorhizal propagules;
this might explain the ‘depressive effect.
Fertilization effect
The statistid analysis showed that generally, the effects of the fertilizing treatments
were very significant (Table 3) on the growth parameters (A to E) and the noduIation
parameters (F and G.). Acacia senegal responded better than Acacia Zaetu. There was
no interaction between soil sterilization and fertilizing treatments.
Inoculation with rbirobium, in comparison with the control induced an increase
of both growth and nodulation parameters with the exception of parameter B for Acacia
laeta. However, these increases were not always significant. The significance threshold
was reached only for the pammetexs A, F and G by Acacia Zaetu and for B, C, D, and
E by Acacia Senegal. Bhizobial inoculation plus 30 ppm or 60 ppm of phosphorus induced
important increases for all the parameters. The values reached were significantly different
from those of the other treatments. No significant differences occurred between the phosphorus
treatments for both acacia species. One may suppose that in this case, the economically
optimum level for the application of phosphorus is 30 ppm.
Table 4 shows the percentage of increase of all parameters for the treatments 2,
3, and 4 in comparison with the control treatment. They were low for treatment 2, they
ranged from 20% to 132% for the growth parameter for Acacia senegd and no more
than 21% for Acacia Zuetu. For the number and dry weight of nodules, the increase ranged
respectively from 190% to 283% and 66% to 142%.
For treatments 3 and 4, with phosphorus supplements to the rhizobial inoculation,
these increases were higher ; for the growth parameters (A to D) of Acacia senegul they
ranged from 50% to 200% and 35% to 131% for Acacia Zueta. The gain for the dry
foliar biomass was 506% for Acacia senegal and 177% for Acacia Zueta. For the number
(F) and dry weight (G) of nodules, the mean increase was approximately 1250% for Acacia
Senegal and 300% for the others species. Given what is lmown about the role of phosphorus
in the structure of essential molecules, in the energy process, and in the metabolism of
the plant and recent new information (Bielesky, 1973 ; Heldt et aZ., 1977; Preiss, 1982 ;
Mengel, 1984), this effect of phosphorus is not surprising. On the phosphorus deficient
soil from Kebemer, P fertikation combined with by Rhiwbium inoculation had a positive
effect on the growth of both gum arabic trees and probably on N, fixation. Indeed the
phosphorus treatments largely increased the nodulation. As nodule formation and nitrogen
fixation require energy and thus ATP, one might assume that the increase of the available
phosphorus level in the deficient soils induces a better functioning of the rhizobial symbiosis.
Table 2 : Effect of sol1 sterilization on growth and nodulation parameters of Acacia Senegal and Acacia Taeta
Measured Parameters *
Growth :
Nodulation
Species
Soil sterilization
(factor 1)
(factor 2)
A
B
C
D
E
F
G
'--
a. sterilized soil
4.69
24.10
3'7.86b
33.50
1462
14.32
281
3
Acacia
b. non sterilized soil
5.01
24.99.
43.93a
37 .'27
1654
13.80
328
Senegal
( a -b). 100 / a
tll
to4
+16
tll
+13
- 0 4t17
a. sterilized soil
5.56
35.13
59.13b
34,66b
1958b
18.64
304
Acacia
b. non sterilized soil
5.66
36.56
67.36a
41.6ga
2755a
15.53
307
laeta
(a-b).lOO/a
+02
to4
t14
t20
t41
-17
to1
* As in Table 1
; each data for each species and in each column without letter or with the same letter are not
significantly different (P = 0,05 - Newman Keuls test).
,,
_.
.
.
.
.
.
...:.
,-
. . __ .
.
.)
:
~.
_
59
Table 3 : Effect of fertilizing treatments on the growth and nodulation
of Acacia-Senegal and Acacia Laeta
Measured Parameters*
Species
Fertilizing
treatments
A B
C
.D
E
F G
-
-
-
-
-
-
-
A. Senegal
No inoculation
3.63b 17.00~ 19.5Oc 16.83~ 423~
1.67b 39.6b
Rhizobial ino-
4.35b 22.97b 30.93b 25.87b 981b
4.85b 151.5b
culation
Rhizobial ino-
5.61a 28.55a 56.93a 50.90a 2232a 25.85a 475.5a
culation + 30
PpmP
Rhizobial ino-
5.73a 29.07a 55.28a 47.27a 2566a 23.43a 552.4a
culation + 60
Ppm p.
A. laeta
No inoculation
4.52~ 30.17b 40.17b 23.17~ 1310~
5.5oc 117.4c
Rhizobial ino-
5.14b 29.35b 47.13b 28.17~ 1580~ 13.32b 195.4b
culation
Rhizobial ino-
6.34a 42.75a 79.40a 47.38b 2840b 26.82a 462.6a
culation + 30
tvm P
Rhizobial ino-
6.35a 40.85a 85.42a 53.43a 3631a 22.37a 444.9a
culation + 60
Ppm p
* Asin Table 1 ; for each species, in each column, data follow by the same
.
letter are not significantly different (P = 0,05 - Newman Keuls test).
i‘“‘.
1 :
~
-
__
. , I . ,
_,,
,
_
, .
,?.
”
- .
,.
::-
.;;:..
.(,
.:
:
.
.:
60
-
--
.,_- ---- :. ._. ~.
_
; Table 4 : Variations in percentage for each parameter due to fertilizing
treatments (2, 3, 4) versus control (11
Measured Parameters*
j
::
j' Species
Fertilizing
treatments**
A
8
C
D
E
F
G
-
-
-
-
-
-
' A. Senegal
2
20
35 59 54 132 190
283
3..
55
68 192 202 428 1447 1100
4
58
71 183 181 506 1302 1294
A. laeta
2
14
-3
17
22
21
142
66
3
40
42
98
‘1 04
117
388
294
4
40
35
113
131
177
307
279
* Parameters A.to G as in iable 1
;
**Fertiiizing treatments as in Table 3.
)
_.
I
.:.
.
..’
”
y”
..,-
.-*
61
Further investigations are planned to determine the influence of lower phosphorus
levels on the growth of acacia in the presence or absence of rhizobii inoculation. Such
experiments should help to solve the problem of costs and benefits of phosphorus appli-
cation during afforestation programmes.
References
BADJI, S., DUCOUSSO, M., GUEYE, M. et COLONNA, JP. 1988 Fixation biologique.
de l’azote et possibii de nodulation croisk chez les deux espkes d’acacias producteurs
de gomme dure : Acacia senegal L. Willd. et Acacia laeta R. Br. ex Benth. CR.
Acud. Sci., S&. IJI. 307 (1 l), 663668.
BADJI, S., DUCOUSSO, M., GUEYE M, et COLONNA, JP. 1989 Effets de l’inoculation
par diverses souches de Rhizobizun du S&r&al sur les deux principaux Acacias gom-
miers en culture semi-aseptique. In : &I&w Symposium SUF le Go&r et la Gomme
Arubique (SYGGA III), 25-2-8 Oct. 1988, St. Louis du Senegal. Collections Actes de
I’ISRA, p. 171-179.
BASAK, UK. et GOYAL, SK 1980 Studies on tree legumes. II. Further additions to
the list of nodulating tree legumes. Plant and soil. 56, 33-37.
BIBLESKI, R.L. 1973 Phosphate pools phosphate transport and phosphate availability.
Ann. Rev. Plant Physiol. 24,225252.cENTRE DU COMMERCE IN’IFRNATIONAL
CNUCED. 1978 Le marche de la gomme arabique, CNUCED Ed. Geneve, 181 p.
CORNET, F. et DIEM, H.G. 1982 Etude comparative & l’effkaciu5 des souches de Rhizobium
d’ Acacia isolees de sols du Senegal et effet de la double symbiose Rhizobium - Glomus
mossae SUT la croissance d’Acacia holosericea et A. raaWana. Bois et For&s &s
Tropiques. 198, 3-15.
CORNET, F., DIEM, H.G. et DOhIMBRGUES, Y.R. 1982 Effet de l’inoculation avec
Glomus mossae sur la croissance d’Acacia holosericea en p&pin&e et apr&s trans-
plantation sur le terrain. In I Les mycorhizes : biologie et utilisatior~ Colloque de
I’INRA n”13, 5-6 Mai 1982, Dijon, Ed. INRA Publ. pp. 287-293.
DREYFUS, B.L. et DOMMERGUES, Y.R. 1981 Nodulation of Acacia species by fast
and slow growing tropical strains of Rhiibium. AppIEnvironBioZ. 41, 97-99.
GIFFARD, P.L. 1975 Les gommiers, essences de reboisement pour les regions sah&ennes.
Bois et Fon%s des Tropiques. 161, 3-21.
-HELDT, RW., JA CHONG, C., MARONDE, D., HEROLD, A., STANKOVIC, Z.S.,
WALKER, D.A., KRAMlNGER, A., KIRK, MR. et HEBER, U. 1977 Role of or-
thophosphate and other factors in the regulation of starch formation in leaves and
isolated chloroplasts. Plant physiol. 59, 1146-1155.
MENGEL, K. 1984 Uptake of mineral nutrients and their biological functions. In. VIeme
Collocpe International pour l’Opt.imisation de la Nutrition des Plantes. 2-8 Sept 1984,
Montpellier, public par AIONP/CIRAD et ACCT, pp. 1495-1524.
NEW, TR. 1984 A biology of Acacia. Oxford University Press, Melbourne, 153 p.
6 2
PREISS, J. 1982 Regulation of the biosynthesis and degradation of starch. Ann.
Rev. Pfmr
Physid. 33. 431454.
WALKER, D.A. 1980 Regulation of starch synthesis in leaves : @e role of orthophosphak
In. Physiological aspects of crop productivity. Intern. Potash Institute, Beane, pp.
195-207.
&icole et Foredre en Afrique
I
_:
i.’
.
Nodulation survey ~~6~:~~&i&1$nous
and non legum&&i trees’ growing
at ICRAF’s field station
ODUOL, PA, AKUNDA, E.M.
and WAMBUGU, PX.
ICRAF P-0. Box 30677 Nairobi, Kenya
Introduction
Substained productivity, of &d is very much dependent on nitrogen input to the
soil from both biologiqal and non-biological sources. Legiunes have been used in building
and c3mmving soil fertilitj since thk beginning of agriculbnz because of their ability
to NIX nitrogen from the atmosphere. It is estimated that this symbiotic process between
bacteria and host plant accounts for approximately 20% of the total nitrogen fmed an-
nually in the world (Dazzo and Hubbel 1974).
Most legumes grow as wee+ (shrubs and trees), thus very little knowledge in
the extent of nodulation is known. The presence of nodules is the prelinkury and visual
confirmation of nitrogen fixation in legumes. Nodulation among legumes is affected by
64
several factors among which include soil temperature (Santo and Dobereiner, 1970 ; Habiih,
1970 ; Gibson, 1977), soil moisture (Habish. 1970), absence or presence of excess minerals
(Jutse and Haque, 1984). All these factors and the presence of the right Rhizobium affect
the rate and position of nodulation.
About 49 trees (Table 1) grow at ICRAF’s f=ld station (some indigenous and
introduced). Most of these species are likely to be used in agroforestry interventions which
focus on fertility restoration. Up to date, no quantitative or qualitative information on
the nodulating ability of these species has been given under the sub-humid, moderately
acidic condition of Machakos field station, the center of ICRAF’s field trials.
This paper reports he findings of the noduktiug survey of some leguminous (47)
and non-leguminous (2) trees being gr&n ‘at the field station in a sub-humid and serni-
arid environment.
Materials and methods
The study site lies between l”33”S latitude and 37”lS’E longitude at an altitude
of 156Om in the sub-humid to semi-arid zone with bimodal annual rainfall of about 700mm.
Mean annual air temperature of 22”C, with deep soil profile, well drained, moderately
leached, weakly acid soils with pH rate 6.0-6.5 and medium nutrient levels. Soils gene-
rally classified as al&ols.
Forty nine plants were examined for the presence of nodules. These plants were
leguminous and non-leguminous woody shrubs and trees Some of these 49 species have
been reported to nodulate in various reports (Table 1) (Corby, 1974 ; Allen and Allen
1981). The species examined were both indigenous as well as introduced exotics
(Table 2).
Periodic sampling,,for nodules was carried,out during the wet active growing sea-
son (and’dry season) of the ‘kees/sbrubs. This was done on diverse dates during April
and May in 1986, 1987 and 1988 after digging up a hole near the tree trunk to expose
the root system. The excavated roots were washed under runuing tap water and the root
systems were floated in trays of water to facilitate detection of nodules. The presence
or absence and degree of nodulation on the root systems were noted. The morphology
of mature nodules was described and their sizes (diameter) measured in mm. Special
care was taken to distinguish root nodules from root malformations, such as those caused
by nematodes, insects or other root inhabiting pathogenic microorganisms, by disecting
the nodules to establish whether they are red, brown or green iu internal colour.
Results
Observations on nod&ion are presented in Table 2, which also lists whether the
plants are indigenous or introduced species The mcorded presence of root nodules on
the examined species was checked against published lists of Allen and Allen (1961,1981),
Corby (1974), Lim (1979) and Mohammad and Mahmoud (1985). Table 2 also presents
species which nod&ted during the three seasons of sampling from May 1986 to May
1988. Nodulation octm-ed in 24 species, represented by : Mimosoideae, Caesalpinoideae,
Papilionoideae, and non-legumes representing Simaroubaceae and Casuarinaceae. Majority
of the species were moderately to heavily nodulated, some were very heavily nodulakd,
specially Gliricidia sepium, Leucaena leucocephala, Samanea sanuq Sesbania se&an,
65
- - .
-I__--_
_...._
-__-
_
&=.a-.--
__---.-.~
.._
.,.
-.-
-_-
Table 1.; .-
List of Nitrogen fixing trees growing at ICRAF's Field Station
Machakos - Kenya
Nodulation
Species
Present
Previous Record
Study
(in literature1
Acacia albida
Acacia deamii
y'
Acacia farnesiana
Y
Acacia holosericea
NI
Acacia mangium
;
Y
Acacia mellifera
Y
Acacia paniculata
N I
Acacia pennatula
Y
Acacia stuhlmanii
t
Y
Acacia saligna
t -
Y
Acacia tortilis
+
Y
Acacia victoriae
Y
Acacia xanthoplea
Y
Albizzia longepedata
**Alvarodoa amorphoides
t
!
Ateleia herbert-smithii
NI
Caesalpinia velutina
+
NI
Caesalpinia coriaria
-
NI
Caesalpinia eriostachys
t
\\’
Crescentia olata
;:-. ..-;::, ;:- ,
Cajanus cajan
-I-
Y
Calliandra calothyrsus
t
Y
Cassia alata
B
Cassia siamea
Y
*jCasuarina equisetifolia
Y
i
Cordeauxia edulis
Y
Enterolobium cyclocarpmum
Erythrina abyssinica
Gliricidia sepium
Haematoxylon brasilleto
Leucaena diversifolia
Leucaena shannoni
Leucaena leucocephala
(Cunningham).
Leucaena leucocephala (Peru)
+
Leucaena leucocephala
t
(Hawaiian giant - K81
Mimosa tenuiflora
Myrospermum frutescens
Parkinsonia aculeata
Pithecellobiun? dulce
Prosopis juliflora
Prosopis nigra
Table 1 : (continued)
Nodulation
Species
Present .i
Previous Record
Study
(in literature)
Prosopis pallida
Y
Prosopis alba
Y
Prosopis chilensis
Samanea saman
+
::
Sesbaniamacrantha
+
Y
Sesbania grandiflora
+
Y
I
Sesbania sesban
+
i
Senna atomeria
NI
r* iWi~ieguminous
+ Presence of nodules
- absence of nodules
Y Nodulation previously observed in other studies
B Species previously investigated but nodulation never observed
NI No information available.
68
Sesbania macrantha and Sesbania grandiflora. Heavy nodulation ocurred in species Alvarodoa
amorphides, Casuarina equisetifolia, Calliandra Calothrysus, Erythrina abyssinica,
Enterolobium cyclocarpomum and Pithecelobium dulce. Sparse to moderate nodulation
oaxm-4 in species of Acacia saligna, Acacia tortilis, Acacia stuhlmanii, Prosopis jul#lora.
&e.&pinia velutina, h?Sa@da eriostachys, ittxcae~ diversifo~ia. kucaena shamoni,
provenanees of Lacae~ leucocephala (Hawaiian giant K.8) and Cunningham and Mimosa I
tenufflora and Myrospemum frutescem
Most of the nodules were located on the upper parts of central tap root systems
in clusters as well as on lateml roots mg.1). Most of the young nodules were rcknd,
white and smooth-surf&xl whi@ .older nodules were of various shapes, mostly brown
in internal colour and rough-surfaced. Nodule shapes were classiied into different types
as described by the drawings of Lim and Ng (1977). The nodde shapes in Mimosoideae
were f&n shaped, elongated, lobed, globose and bifurcate. cardlloid types were found
in Sbnaroubaceae and Caskeae. In Caesalpinoideae, the nodules were globose. and
in Papi~Yonoideae the nodule shapes were globose and fan-&p&d (Table 2). Fig. 2 shows
globose shaped nodules for &sbania sesban occuring in clusters.
Table 2 also shows average diameters for mature nodules. The nodule diameters
are way variable, ranging from 1.7 mm in Lmcaem leucocephala Peru) to large diameters
of 4 mm in Mikosa ten@orti .and Sesbania granal.@ora.
.:
,~
:
, ~
;,;..,
:,.:::::
,,
.
.
. . .
.
.
‘,’
I
.,.:
68
Sesbania macrantha and Sesbania grandifora. Heavy nodulation ocurred in species Alvar&a
amorphides, Casuarina equisetifolia, Calliandra Calothrysus, Erythrina abyssinica,
Enterolobium cyclocarpomum and Pithecelobium aklce. Sparse to moderate nodulation
occurred in species of Acacia saligna, Acacia tortilis, Acacia stuhlmank Prosopis julipora,
Caesalpinia veluthuz, Caesalpinia eriostachys, L+eucaena diversifolia. Leucaena shannoni,
provenances ofLeucaena leucocephala (Hawaiian giant K.8) and Cunningham andMimosa
tenuiflora a n d Myrospermum jktescens.
Most of the nodules wexe located on the upper parts of central tap root systems
in clusters as well as on lateral roots (Fig.l). Most of the young nodules were rotid,
white and smooth-smfaced while older nodules were of various shapes, mostly brown
in internal colour and rough-surfaced. Nodule shapes were classified into different types
as described by the drawings of Lim and Ng (1977). The nodule shapes in Mimosoideae
were fan shaped, elongated, lobed, globose and bifin-cate. Carolloid types were found
in Simaroubaceae and Casuarinaceae. In CaeSalpinoiakae, the rqdules were globose, and
in Papilionoideae the nodule shapes were globose and fan-shaped (Table 2). Fig. 2 shows
globose shaped nodules for Sesbania sesban occuring in clusters.
Table 2 also shows average diameters for mature nodules. The nodule diameters
are very variable, ranging from 1.7 mm in Leucaena leucocephala (Peru) to large diameters
of 4 mm in Mimosa tenuijlora and Sesbania grandiJora.
Figure 1 : Exposed root system of Sesbania bispirwsa showing nodules confined to the upper parts
of the tap root and in late&l roots.
69
Figure 2 : Sesbaniu sesban globose shaped nodules in clusters
69
;,.
i.
i,
i
;
,.
,’
‘,
.-.
. . -,
1.
‘-.
‘.
;
,$.
i
I
. .._
*
.;
*
;;
.,.
.”
.
.
”
- .
:.
70
Discussion
In this survey, 24 out of the 49 species were recorded for presence of nodules.
However, the absence of nodules on other species is not a conclusive evidence for lack
of nodulating ability. The findings of this survey are in broad-agreement with the surmnary
of the world-index of nodulation by Allen and Allen (1961). However, some of these
species which were not observed to nodulate in this study have been reported elsewhere
to be nodulating (Table 1). Successful nodulation in these plants indicates the presence
of potential infective autochthonous rhizobial strains in the soil. The degree of nodulation
was sparse in some species ; thii can be attributed to the fact that most of these species
were not inoculated at the time of planting (Table 2). Those nodulating species somehow
formed a symbiotic relationship with the indigenous rhizobia in the soil. Conclusive evidence
cannot be drawn on whether there is correlation between those introduced and indigenous
species in nodulation. There was a general absence of nodules on these trees/shrubs
during the dry season. This confirms the inability of species to nodulate under extreme
or adverse environmental conditions. Nodules were generally active during the wet seasons
when there was a general active vegetative and reproductive growth in these species.
However, the nodules during the dry season were generally inactive. Nodulated legumes
and non-legumes can contribute significantly to the nitrogen status of the soil. In order
to exploit the potential of fixed nitrogen, it is necessary to successfully identify nitrogen
fming nodulated legumes and non-legumes.
In this study, the non-legumes Alvaroaba amorphoides and Casuarina equisetifolia
nodulated prolifically at this site, thus indicating a great potential of non-leguminous plants
with nodules possessing nitrogen fixing capacity. In future, more and more nitrogen-fixing
tree species with multiple benefits will be used in farming systems . So, more information
is required on the agronomic and agricultural aspects of nodulated plants to enable the
symbiotic system to be .exploited to its maximum. What is presented in this paper is
an account of the potential contribution of these nodulated plants to the nitrogen cycle
in the biosphere.
The study shows prolific nodulation which varies among species. The next stage
is to select the most prolific nodulating species for quantitative nitrogen fixation studies
to be carried out at different periods of the year.
References
ALLEN, E.K. and ALLEN, 0-N. 1961 The scope of nodulation in theLegminosae. Recent
Advances in Botany. Vol. 1, University of Toronto Press, Canada.
CORBY, H.D.L. 1974 Systematic implications of nodulation among Rhodesian 1egume.s.
Kirlia. 9. 301-329.
DAZZO, R.B. and HUBBEL, D.H. 1974 Biological nitrogen fixation. Soil and Crop Science
Sot. Florida Proc. 34, 71-74.
GIBSON, A.H. 1977 In : A Treatise on dinitrogen fixation. IV. Agronomy and Ecology
(Hardy, R.W.F. and Gibon, A.H. eds.). pp. 350-393. Wiley Interscience, New York,
U.S.A.
JUTSE, S. and HAQUE, I. 1984 Soil plant nutrition and forage agronomy research (F’ro-
ject F’rotocols). Highland Programme, ILCA, Addis Ababa, Ethiopia
7 1
HABISH, D.A. 1970 Effect of certain soil conditions on nodulation of Acacia spp. Plant
and Soil. 33, l-6.
LIM, G. and NG, HL. 1979 Root nodules of some tropical legumes in Singapore. Plant
und SoiI. 46, 317-327.
MOHAMMAD, A. and MAHMOOD, A. 1985 Qualitative study of the nodulating abiity
of legumes of Pakistan. List 3. Trap. Agric. 62 (IQ 49-51.
SANTO, D.M. and DOBBRBINER, J. 1970 Perquisa Agrqxcuaria. Brmileira. 5,
365-371.
Maxim&r la FBA pour la Protfuction Agricok et Foresti&e en Aj?ique
Effect of inoculation- with Rhizobium,
P application and liming on early growth
of Ieucaena (Leucaena Zeucocephalci Lam. de wit).
. .
GICHURU, M. and.MULONGOY, K.
InternatM Ikstitti of Tropical Agricuhre. PMB 5320. lbaahn, Nigtfria.
Abstract
The .effects of P applioation (0 and 50 mg kg’ aoil), inooul&on with soil from
a field with well nodulated plants of LeucaeM Ieu~~~ephaIa, and with a water sus-
pension of a Rhizobiwn strain IRC 1045 specific for leueaena, on establishment of
leueaena were studied in four soils. two Psammentic Ustorthenta, one Oxic Paleustalf
and one Typic Paleudult. Application of P resulted in imProved leueaena growth
and nodulation in all four soils. The sharpest response was obtained in the Psammentic
Ustorthent from a farm where leucaena established Poorly. Soil inoculation with rhi-
zobia resulted in inqxoved leueaena growth increased dry matter yield and nodulation
except ou the Oxic Paleuatalf from the International Institute of Tropical Agriculture
(IlTA) on which leueaena normally paforms welL The aoil inoarkmt performed better
than the water auqension of atrain IRC 1045. Dicaleium phosphate affected the early
growth of leueaena on the TyPic Paleudult because of its liming properties.
Introduction
Alley cropping is an agroforcshy system in which food crops are grown in alleys
formed by hedgerows of suitable trees or shrubs which can be pruned periodically to
prevent excessive shading and which provide mulch and green manure for companion
food crops. Leucacna (Leucaena feztcocephdu var. K 28) appears to bc a suitable hedge
row tree for alley cropping on Alfiils and Inccptisols
of southwestern Nigeria (Kang
and Wilson, 1987). The problems of establishment and of slow initial growth of leucacna
7 3
remain however a major limitation for adaptation alley cropping by farmers in some areas,
particularly when the direct seeding method is used. The poor establishment experienced
in farmers’ fields has been attributed to competition with companion crops or weeds (Jones
and Bray, 1983 ; Atta-Krah and Kolawole, 1987) to lack of effective strains of Rhitobium
for nitrogen fixation (Ruaysoongnem,et al., 1984 ; Sanginga et al., 1988) to P deficiency
(Sanginga et al., 1988) and to acidity in highly weathered soils (Hutton and Andrew,
I
1978 ; Olverra et al., 1982).
The purpose of this study was to evaluate some of the soil factors that are respon:
sible for the poor establishment of leucaena in southern Nigeria. The effect of lime and
dicalcium phosphate on leucaena early growth in an acid soil was also compared
Materials and methods
A greenhouse pot experiment was conducted at the InternationaI Institute of Ti-o-
pical Agriculture (ETA), Ibadan, Nigeria. The experimental design was a split-plot with
four soils, two from on-farm sites near Alabata village, approximately 15 km Northwest
of Ibadan (Psammeutic Ustorthent), one from the IITA main station (Oxic Paleustalf)
Ibadan, southwestern Nigeria, and one from the IITA high rainfall station (Typic Paleudult)
at Onne, southeastern Nigeria, as main plot treatments. Leucaena establishment was rela-
tively good at one on-farm site (farm-l) and poor at the other site @m-2). The soil
fi-om IlTA station usually supports good leucaena growth whereas leucaena performance
in the acid soil is normally very poor. Chemical properties of these soils are shown iu
Table 1. The subplot treatments were two rates of dicalcium phosphate (0 and 50 mg
P kg’ soil) and three inoculation treatments (uninoculated control, inoculation with a water
Table 1 : Selected chemical properties of the soils used for the pot experiment
Soil
PH
Organic Total Bray-l Ca Mg K
Total
C
N
P
acidity
--v-P
- -
%
%
kg -1 ------
w
meq/loog-------------
IITA
6.3 1.49 0.121 11.2
0.74 0.32 0.29 0.16
(Oiic Paleustalf)
Farm-l
5.8 0.65 0.052 7.2
0.53
0.17 0.30 0.05
(Psammentic Ustorthent)
.
Farm-2
5.6 0.51 0.052
2.2 0.52 0.15 0.24 0.00
(Psammentic Ustorthent)
Onne
4.9 1.14 0.089 62.2. 0.73 0.25 0.30 0.36
(Typic Paleudult)
,.-
,f.”
:‘:
.<.’
74
suspension of Rhizobium strain IRC 1045 and inoculation with a soil from a field with
well n&dated leum) in a factorial arrangement, Rhiwbium strain IK 1045 is an
elite strain isolated from .no+les of leucaena grown on an Alfisol in southwestern
Nigeria (S.anginga ef al, 1988). Ten mihiliters of the water suspension of this strain (109
cells ml-l)’ ixp‘g: of the’ ,++l :inoculano, (10’ leucaena rhizobia g-l soil) were mixed
,withtbesurfaci:soil~inthepots,Thefinaweigfitofsoilineachpotwas3kg.
A second pot experiment was carried out to study the effect of lime (CaCOJ and
dicalcium phosphate fextiker on early growth of leucaena in the acid soiL The treatments
wereOand2000mgCaCO,kg’soil,andOand50mgPkg1soilinafactorialarran-
gemIlL
I ”
Six seeds of leucaena (Leucaena leucocephala var. K-28) previously scarified in sul-
furic acid we sown in each pot Seedlings were thinned tb three after one week. The
pots were miformly watered with 100 ml of deionised water daily for the first 6 weeks
and twice daily af&wa&. Bi-weekly height measurements were made until 14 weeks
after planting (WAP). At harvest (14 WAP), above-ground dry matter was determined.
Roots wea-e washed in tap water over a sieve to miuimi~ losses. Nodules were separated
from the roots, counted and dried.
The Statistical Analysis Systems (SAS) package was used for statistical evaluation
of variances. Least significant differences (LSD) are given only when the overall F-test
is signiticant @ = -0.05).
Results and discusslon
Height- measurements
Th~slxqz@al effect of.,P appbcation on leucaena growth was obseaved from 4 WAP
and the effe&w sign&c&t for; the rest of the experiment (Table 2). Application
of 50 mg -P kg’ as dicalcium phosphate increased leucaena height by 18 - 42%. The
largest effect was observed at 6 and 8 WAP.
Table 2 : Effect of P application on leucaena height at different times
(Means of four soils, 3 inoculation treatments and four replications)
Weeks after planting
. .
,. ‘;
J
.
Phosphorus‘,
rate
2
4
6
a
10
12
14
--P---P
-1
“‘g kg
-----------------_-_ (cm) --------------------------- :
0
7
9
13
17
22
27 34
j'
50
7
11
17
24
28
33
40
Ls0(0.051
NS
0.4
0.4
0.8
1.2
1.6
2.6
75
Response of leucaena to inoculation is shown in Table 3. Leucaena growth on the
IITA station soil showed no reponse to inoculation except at 4 WAP. This indicates that
effective rhiiia capable of infecting leucaena were not limiting at this site. In conrast,
growth on other soils showed variable response to iuoculation. ‘I$ two on$$n showed
a delayed response to inoculation compared to ‘the Omie soil but the in$ial gt6vth of
leucaena in the latter soil which is acidic was much slower than in the other soils. The
soil inoculant perform better than Rhizobizun strah IRc 1045 on all soil types. Soil ino-
culauts are also known to supply other useful microorganisms like endomycorrhi# fungi
and some soil nutrients.
,.
The better response to soil inoculant can explain the improved establishment of’tians-
planted leucaena compared with directly seeded plants in areas where leucaeua is imro-
duced for the first time. Soils used in nurseries are likely to be from sites with favorable
nutrieut aud microbiological properties.
Dry matter yield and nodulation
Above ground and root dry matter were significantly enhanced by P application with
sign&am interactions between soil type and P iu root dry matter yield (Table 4). Without
phosphorus application, leucaena growing on Farm-2 soil produced significantly less bio-
mass compared with the other three soils. The sharp response to P by leucaena on Farm-
2 soil indicates that available P was limiting leucaena growth on this soil. This was confir-
mtxl by the soil analytical results (Table 1).
Nodulation was not significantly enhanced by P application except on the acid soil.
Sign&am interactions between soil type and inoculation were noted iu dry matter yield
and nodulation with the acid soil producing much sharper responses compared to the
other soils (Table 5). The llTA soil did not show response to iuoculation There was
a general tendency for increased nodulation with inoculation although the effects were
significant only in the acid soil. Nodule counts were highly variable particularly in the
sandy soils due to the difficulty of collecting the small nodules.
Effect of lime and dicalcium phosphate
In the aid soil, leucaena dry matter yields and nodulation showed significant inter-
actions between lime and P (Table 6). Growth and nodulation were increased by the
application of either lime or dicalcium phosphate. Lime produced greater increases. Extrac-
table P is high in this soil (Table 1). With the application of lime, phosphorus produced
no additional increase in native leucaena growth and nodulation. The explanation for this
is that application of lime to this soil makes the native soil phosphorus more available.
76
Table 3 : Effect of inoculation with a soil inoculant and a water-
suspension of Rhizobium strain IRc 1045 on height of leucaena
on four soils at different times.
Inoculation
Weeks after planting
treatment
2
4
6
8
10
12
14
-----~
- ----_- --------------(cm)---------
---_-_---------_____
I ITA
Control
7
12
16 22
27
33
3 9
Soil inoculant
7
9
15 21
27
34
39
IRc 1045
7
12
16 23
30
35
42
LSD(O.05)
NS
1.8
NS NS
NS
NS
NS
Farm 1
Control
7
11
15 21
25
31
39
Soil inoculant
7
11
15 23
28
33
40
IRc 1045
7
10
14 20
24
27
34
LSD(O.051
NS
NS NS NS
2.6
3.3
NS
Farm 2
Control
7
11
14 17
20
23
28
Soil inoculant
7
10
14 19
24
28
33
IRc 1045
7
11
14 18
22
26
30
LSD(0.05)
NS
NS
NS 1.6
2.3
3.6
NS
Onne
Control
5
8
13 18
22
27
35
Soil inoculant
6
9
15 23
29
34
42
IRc 1045
5
9
14 21
25
31
40
LSD(O.051
0.5
0.8
1.4 2.1
3.1
4.5
NS
77
Table 4 : Effect of P application on dry matter production and
nodulation of leucaena on four soils at 14 weeks after
planting.
Phosphorus
applied
Top
Root
Nodules
Nodule
(mg kg-l)
weight
weight
number
weight
g plant -1
g plant -'
No. plant -1
q plant -1
IITA
0
5
5
34
107
50
7
7
42
100
LSD(O.05)
0.5
0.6
NS
NS
Farm 1
0
5
5
36
_
140
50
6
6
39
110
LSD(O.05)
0.6
0.7
NS
NS
Farm 2
0
3
3
28
60
50
5
5
39
a3
LSD(0.05)
0.3
0.5
NS
NS
Onne
0
5
5
42
a3
50
7
9
74
157
Lso(o.o5)
0.8
1.0
25.0
60.0
78
Table 5 : Effect of inoculation with a soil inoculant and a water-suspension:
of Rhizobium strain IRc 1045 on leucaena dry matter yield and
nodulation on four soils, at 14 weeks#after planting.
Inoculation
Top
Root
Nodule
Nodule
treatment
weight
weight
number
weight
g plant -1
g plant -1
No: plant-'
mg plant -1
,,
IITA
Control
6
6
32
120
Soil inoculant
6
6
42
90
IRc 1045
6
6
40
97
I
LSD(0.05)
NS
NS
NS
NS
Farm 1
Control
6
5
28
87
;
Soil inoculant
6
6
43
157
I
IRc 1045
5 .
5 .
42
130
I
LSD(O.05)
0.7
NS
NS
5
3
Farm 2
Control
3
3
19
53
Soil inoculant
4
4
39
87
IRc 1045
4
4
42
73
LSD(O.05)
0.3
NS
NS
NS
Onne
Control
5
5
30
167
Soil inoculant
7
7
97
153
IRc 1045
5
5
50
97
LSD(O.05)
1.0
1.2
30.6
NS
79
.
I/
I
Table 6 : Effects of lime and dicalcium phosphate on the dry matter yield
,
and nodulation of Leucaena leucocephala in a pot experiment with
i
an acid soil from Onne, at 14 weeks after planting.
P applied
CaC03 applied, mg kg-l
(mg kg-')
0
2 0 0 0
Means
TOPS, (g plant-')
0
4
10
7.0
50
6
10
8.0 SE _f 0.54
Means
5.0
10.0
SE + 0.54
SE (1 cell) = k 0.76 :
CY 21%
t
Roots, (g plant-')
0
4
7
5.5
50
6
6
6.0 SE + 5.53
SE + 0.55
SE (1 cell) = 0.78
CV 27 %
Nodule number (No.plant-')
0
18
112
65.0 SE + 10.0.
50
62
97
79.5
Means
40.0
104.5
+
10.0
E
SE (1 cell) = 14.2
39 %
80
Acknowledgements
The authors thank J. Pleysier and AN. Atta-Krah for their suggestions and E.I. Jegede
for the technical assistance.
References
A’ITA-KRAH, AN. and KOLAWOLE, G.O. 1987 Establishment and growth of leucaena
and gliricidia alley cropped with pepper and sorghum. Leucaena Res. Reports. 8,46-
47.
HUTTON, E.M. and ANDRBW C.S. 1978 Comparative effects of calcium carbonate
on growth, nodulation and chemical composition of four Leucaena leucocephala lines,
Mucroptilium Iathyroides and L&on&s buinesii. Aust. J. Exp. Agric. and An. Hush.
18, 81-88.
JONES, RJ. and BRAY, R.A. 1983 Agronomic research in the development of leucaena
as a pasture legume in Australia. In : Leucaena Research in Asian Pacific Region,
pp. 4148, IDRC, Ottawa, Canada.
KANG, B.T. and WILSON, G.F. 1987 The development of alley cropping as a promising
agroforestry technology. In : Steppler, H.A. and Nair, P.K.R. (eds.) Agroforestry
a Decade in Development, pp. 227-243. ICRAF, Nairobi, Kenya.
OLVERRA, E., WEST, S.H. and BRUE, W.G. 1982Establishment ofkucaena leucocephula
in acid soils. Leucaena Res. Rep. 3, 84-85.
RUAYSOONGNERN, S., SHELTON, H&L and EDWARDS, D.G. 1984 ‘Ihe influence
of lime, combined nitrogen supply and Rhizobixun sbxin on growth, nodulation and
nitrogen status of I!.+?UCU~?M kucocephakz cv. Cunningham. In : I.R. Kennedy and
L. Copeland (eds) The Seventh Australian Legume Nodulation Conference, Aust. Inst.
of Agric. Sci. Occusionul Pub. W 12, 21-22
SANGINGA, N., MULONGOY, K. and AYANABA, A. 1988 Response of L.eucueunu/
Rhizobim symbiosis to mineral nutrients in southwestern Nigeria. Plant and Soil.
112, 121-127.
Maxbnisa la FBA pour la Pmduction Agricok et Fore&&e tn Afrique
_ I
’
.,.
A prelimkxry study on the compatjbility
between indi&nou.s rhizsbia
and some tree? Ieguixiqs in Kenya
ODEE, D. W. r
Kenya Fomstry Rescaxh Indtute. Nairobi, P.OBox 20412, Kenya.
..’
.,.
.
‘,..(
: -
..i
cInm’- stldy with indigenous rhizobia was carried out on seval host s&X-
tics of isolation : Acacia a&&z, Acacia meamrii, Callkmdra caldhyrsuq Lencaena
leucocephala. Prosopis jul$lora. Sesbania graruiifrora ad Sesbania se&an. The Rhi-
zobium strains were highly infective and effectiveness was variable across the host-
strain combinations. There occured four cross-inoculation groups within the range
of Rhizobium strains and host species studied. The two members of the genus Se.sbania
were more specific in their rhimbial partna-s in terms of beneficial ass&a&n Nodule
shape and clieutio? ,was kmktent’onthcrootsystemofcachhostspeciesin
.I. _.
i 1
: effectivc : symtnosis
:*
Introduction
Acacia albida, Acacia meanki. Calliandra calothyrsus. L.eucaena leucocephala,
Prosopis julijlora, Scsbania grandifIora and Sesbania se&an are grown in Kenya to provide
green manure and mulch in intercropping systems, browse or fodder for livestock and
fuelwood. These nitrogen fming trees (NJ?lJ nodulate sporadically with indigenous rhi-
mbia in the forest nurse&s and field conditions (Odee, unpub$shed data).
Despite the inaeasin g use of these NET in the country, little is known about their
symbiosis with the naturally occurring rhizobia. Nodulation in both exotic and indigenous
82
legumes by natural rhizobii has been described as erratic (Keya and Van Eijnatten,
1975 ; Keya, 1977). but most of the work reported on natural nodulation have been of
grain and pasture legumes (McDonald, 1935 ; Bumpus, 1957 : Morrison, 1966 ; Souza,
1969).
This papet reports on the results of cross-inoculation study of selected Rhiwbium
strains on the given host spectrum and dkusses the salient features.
Materials and methods
Eight strains of Rhiwbiwn isolated from geographically different areas were se-
lected for the study. Strain NUM 777 was acquired from the Nairobi Rhizobium h4tRCEN.
Each strain was charactetized by streaking on plates of yeast extract mannitol (YEM)
agar. The composition of the medium was as described by Ber&nsen (1980) as follows
(g 1-l) : manttitol, 10 ; yeast eXtraCl, 0.4 ; K.$IPGd, 0.5 -i MgSOi; 7H.+I, 0.2 ; NaCl,
0.1 and distilled water. 0.5% alcoholic solution of bromothymolblue(BTB) was incorporated
into the medium at the rate of 5 ml per liter for. pH reaction of the Rhiwbium strains.
Streaked plates were incubated at 26°C for 10 days. Growth rates were described
by cultural and colony size (Vincent, 1970). Fast growers achieved moderate to abundant
growth with colony sizes of equal or greater than 2 mm within 5 days. Slow growers
achieved slight to moderate growth with colony s&es of less than 2 mm after 5 days.
The characteristics of the strains are shown in Table 1.
Leonard jar assembks (Vincent, 1970) with vermiculite as the growth medium
were used to grow seedlings. The composition of the nutrient solution for the assemblies
was as described by Somasegaran and Hoben (1985) as follows (pM) : CaC4,2%0 :
1000 ; K.$ZPG, : 500 ; CJXsOrFe, %O : 10 ; MgSO,, 740 : 250 ; K$04 : 250 ;
M&O,, II-&O : 1 ; qB0, : 2.; Z&O,, WO : 0.5 ; CuSO,, 5H.&#~ ; .CoSO,, TO:
0.1 ; Na@G, 2qO : 0.1 and 0.05% KNO, (w/v) was added to the nutrient solution
in the plus nitrogen (+N) control assemblies. The pH was adjusted to 6.8 with 1N NaOH.
The entire assemblies with nutrient solution and vermiculite medium were autoclaved at
121°C for 1 h.
Seed of the tree legumes obtained from the Kenya Forestry Seed Center were scari-
fied with concentrated sulphuric acid for the following times (mn) : Acucica albida and
Acacia mearnsii, 20 ; L.eucaeim leucocephala, Calliandra calolhyrsus &d Sesbania sesban,
15 ; Prosopis julijlora and Sesbania grandjlora, 10. After treatment, the seeds were washed
in several changes of sterile distilled water until all %;.$f the, acid were removed.
The seeds were then aseptically placed onto 0.75% (w/v) wather agar plates and incubated
at 28°C for 72 h.
Two pm-germinated seedlings of similar size and radicle length were grown per
assembly. The seedlings were inoculated by dispensing 1 ml of a fully grown test culture
around theii roots using a fresh pipette for each strain of Rhiwbizm. Sterilized dry gravel
was spread over the surface of the vermiculite. Plants were thinned to one per assembly
by excising the shoot of the unwanted plant with sterilized scissors.
Each treatment of host-strain combination and uninoculated plus nitrogen (+N) and
minus nitrogen (-N) controls were replicated four times and grouped according to host
species. The plants were then grown in the glasshouse for 10 weeks. Replenishment of
water and nutrient solution was done once every fortnight.
.-
Table 1 : Growth characteristics of eight. I<hizobium straiis isolated from sev'en tree legumes
Rhlzdbium
Locality
Host [II' Isolation
Growth
Reactton on
Colonial morphology
strain
Bromothymol
Blue (BTB)
1Bb
Magarini
Sesbil,lla randiflora
Slow
Alkaline
Flat, dull, cream coloursd'with
-
-
(Coast Province)
translucent margin differentiating
into greater opacity at the center
of the colonies
lla
Turbo
Sesballl a sesban
Fast
-
-
-
Acidic
Flat, shiny, white coloured and
(Rift Valley)
evenly opaque colontes
2c
Mbits point
Leucaena leuceupheld.,; Fast
Acidic
domed,; shiny, cream coloured with
(Nyanza Province)
tran$lucent margin differentiating
,..
into greater opacity at the center
,I
of the colonies
6b
Kisumu
Leucaf:na leuceu hala!'
-
-
Fast
Acidic
Domed, shiny, cream toured with
'(Nyanza Province)
', z.,
translucent margln dtfferentlatlnQ
fnto greater opacity at thecente'r
of the colonies.
15b
Katangi
Proso;)is
jultflord
- .-
Fast
Acidtc
Flat, shiny, white coloured and
(EasternProvince)
evenly,opaque colonies.
16~
W u n d a n y i
Calli,lridra calothyrs&
~ . ..-
Fast
Acidic
Domed, shiny: white coloured wit);
(Coast Province)
t,ranslucerit margin differentiating
'i.
into greater,opacity at the center?.
of the colonies
e%r
,
;- :
22
hbita Point
Acacl albida
'
-
-
Fast
Acidic
(Nyanza Province)
.?--
5-L
/..
NM 7?7*
Eldoret wattle
Acacl,!
mernsii
'[;.
-.. -
Fast
Neutral
Flat, dull, white Colau&d and evhnly
e;;;;ctin f ctory
T..?
opaque colonies ::
al ev
e
* Source : The Nairobi Rhizdbitxn MIRCEN.
,! !
..: .:
.,i
, -.?
.’
:.
.
.
.
: :
84
Infectiveness and effectiveness of each host-strain combination was described by
visual growth ratings. An effective association (B) produced nodulated dark green healthy
plants showing excellent growth ; partially effective association (F’E) produced nodulated
plants with light green colour and restricted growth ; ineffective association (e) produced
nodmated plants which were pale green and stunted ; and plants which were not nodulated
were designated non-infective (ni). Nodule shape and distribution were described for each
host-strain association according to Somasegaran and Hoben (1985). The dry weight of
the plant shoots were determined after oven-drying at 70°C for 48 h.
Results and Qiscu,+@n
The results of the cross-inoculation tests are given in Table 2. Uninoculated control
treatments remained free of nodules in all host species tested.
The Rhiwbium strahs exhibited 8 high degree of infectiveness with the test plants.
All strains were infectiv& iin their own host of isolation. Strains isolated from CaNziudra
calothyrsus, Leucaena leucocephala and Prosopis jul$i’ora nodulated all the hosts whereas
those from Sesbania grandi$lora, Sesbania sedan and Acacia mearnsii variably nodulated
five host species each.
There were four cross-inoculation or infective groups within the scope of this study.
These were as follows : (i) first group : Acacia mearnsii and Prosopis jzdij7ora ;
(ii) second group : Leucaena leucocephala. Callzkdra calothyrsus and Prosopis Julifzora ;
(iii) third group : Acacia albida. Acacia mearnsii, Sesbania grandiflora and Sesbunia
sesban ; (iv) fourth group : Acacia Albida, Leucaena leucocephala, Sesbania grand@lora
and Sesbania sesban. I-,
Rhizobii in each infective group were mutually interchangeable between or among
the plant host members as defined by Graham (1976). Given the narrow range of Rhiwhm
strains and host species, ‘and the fact that strains +olated from a given tree can also
be different, the apparent infective groups may be regarded as atypical because of the
evident overlap. Howevertheconcept of ctiective group>> is not to be emphasized particularly
in the tropics because even in the recognized groups, infection across the boundaries have
been shown to be prevalent (Lange, 1961 ; Habish and Khairi, 1968 ; Trinick, 1968).
Symbiotic effectiveness as rated by visual appearance of test plants gave the following
own host associations : effective symbioses in Sesbania sedan, Sesbania grandiflora,
Calliandra calothyrsus and Acacia albida ; partially effective symbioses in Leucaena leu-
cocephala and Acacia mearnsii ; ineffective asskation in Prosopis julijlora. Effective
cross-inoculation symbioses were obtained by Prosopis julij7ora strain 131 on Lt~aena
leuwcephala and Calliandra calothyrsus ; Calliandra calothyrsus strain 16~ on Leucaena
leucocephala, and Acacia mearnsii stxGn NUM 777 on Acacia albida.
On the basis of beneficial symbiosis, Sesbania sesban and Sesbania grandflora
were more specific in their rhizobial requirement. Others were less specific because in
addition to forming effective or partially effective own host symbioses, they also exhibited
varying degrees of effectiveness in cross-inoculation with the exception of Prosopis jul$ora.
The ineffective own host nodulation in Prosopis jdifora is not an isolated case. Mettinen
et al (1968) working in Bum (Eastern Kenya) reported a similar occurrence and attributed
it to cross-inoculation with indigenous soil rhizobia compatible with the indigenous Acacia
Senegal. This would seem possible because Prosopis jdflora is not endemic or as natur&zed
as the other species in Kenya and thus may merely be acting as trap-host for indigenous
fable 2 : Ineffectiveness and effectiveness ratlqgs of host-straln combinations
Rhizobtum
Host of
Sesbania
Sesbania
Leucaena
Calllandra
Acacia
Acacia
strain
Isolation
grandlflora
SeJblln
-hala
calothyrsus m
-
liizarnst I
-
18b
Sesbanla
PE(2) ':
E(1)
:r”
e
mora
ni.(?:)
,,(JI
:
lla
Sesbanla
5eSbln
e
e
ni
e
2C
Leucaena
leucbcephala
e
e
PE
e
. ,.
P E
6b
Leucaena
-hala
e
.) e
PE
e
e
E,
15b
e
‘- e
E
E
PE '
PE
ii!iEs?
:,.
16c
Calllandra
calothyrsus
e
e
E
E
e
e
22
Acacia
m
e
.'
e
e
nf
IE
NUM 777
Acacia
E
PE
,e
,’,:‘.
(1) E = Effective, (2) PE = Partially effective, t31e = Ineffective, (4) nl = non infective.
,:
.
86
rhizobii Although the Rhiwbium strains were isolated from effective (pink) nodules, the
partiahy effective or ineffective own host associations reported in this study emphasizes
the need to re-authenticate strain effectiveness on host of isolation prior to use in cross-
inoculation studies. The response to inocuIation in terms of shoot dry weight on own
host and cross-inoculation combinations gave similar trend as in visual ratings. Own host
association of strains lla, 18b, 16c and 22 (Fig. la, b, c, e) gave shoot dry weights
that were signihcantly higher (p = 0.05) than the tminoculated -N control. Shoot dry
weights of cross-inocuIation association of strains 15b on Leucaena Zeucocephafa and
Calliandra calothyrsus, 16c on Lmcaena leucocephala and NUM 777 on Acacia nlbida
(Fig. lc, f. e) were aIso significantIy higher than the uninoculated -N control other beneficial
symbioses of shin 16c on Prosopis juliflora (Fii. Id) and strains 6b, 18b, NUM 777,
2c, 15b, 22 on Acacia meamsii (Fig. lg) were not significant. The insignificant response
of the Iatter two host species could be attributed to their growth rates which were observed
to be relatively slower than the other host species. There also occurred a marked increase
of shoot dry weight by the following symbioses and magnitudes relative to +N control :
strains lla on Sehmia sesban, x 6.0 ; strains 16c and 15b on Leucaena leucocephala,
x 4.7 and 3.8 respectively ; strain NUM 777 on Acacia alIda, x 1.6 ; and strain 15b
on Calhndra calothyms, x 3.0. Despite the fact that the amount of inorganic N applied
in +N control may not have been optimum, the magnitudes of the comparisons indicate
some potential in biological nitrogen fucation of these IWT with the indigenous rhizobia.
The plant host determined the shape and distribution of nodules on the root system
in each host species irrespective of host of isolation of the Rhizobium strain forming
the association (Table 3). However, consistency in nodule shape and distribution was main-
tained only in effective symbioses.
The resuhs presented in this paper have shown varying degrees of infectiveness
and compatibihty of indigenous rhizobii with the NFT. These have in effect posed a
great challeng& The chaIlenge to assess the need of inoculation, select effective symbioses
under simuIated environmentaI conditions in the laboratory/gIasshouse and evaluate the
selected symbioses in the field sites where they are to be grown. Kenya Forestry Research
Institute is addressing itself to these studies.
.”
‘i
Table 3 : Nodule shape and distribution of effective host-strain associations 10 weeks after planting
.'
Test plant
Rhizoblum rtraln and host
of isolation (In brackets)
Nodule shape
Nodule $strlbutlon
Sesbania grandlflora
lab
(2. grandlflora)
Se&globase
Prolfflc t&-root and bcca-
,, -
;.
sionpl lateral root
._.,.
ribdulatlon
'f
.
Sesbanla sesban 3;
-
-
Globose':
5.:z? Occassional tap-root and large
humber of root nodulation
.
.
e
:
,(
Leucaena leucecophala
16~ (C. cdloth rsus)
U;$ate and
"^
"'
Occasstonpl'thp-root and
15b (F T+
lateral root nodulation
I
Prosopts juliflora
16c (C. calothyrsus)
Semi-globose and
Occasslonal tap-root and
lobed
lateral root nodulation
Calliandra calothyrsus
lbc IC. calothyrsus)
Semi-globose
Large number of tap-root and
15b (p. juliflor.aT
lateral root nodulation
Acacia alblda
-
-
22 (A. albfda)
Semt-glabose
Prolific tap-root and occas-.
NUM 777 ~arns~l)
-
-
stonal lateral root nodulatlon
88
.‘
(al sesbd;jia &
(b) Sesbania grandifkxx
(Cl Callii~ cabthyrsus
-
-
T r e a t m e n t s
Fig.1. a b c 6 d. Effect of inoculation
with strains of m and application of
inorgrjc N on shoot dry weights d various tree species in
Leonards jars at 10 weeks after p4ant-q.
/..,
.:,.t,
3;.
. ..c.:
:
-
‘:
I
II
‘..I
;?
L
i. ‘.:.. .
__
9 0
Acknowledgements
I am very graMid to Professor S.O. Keya and Nairobi Rhiwbiwn MIRCEN staff
for technical assistance, and IDRC for financial support to attend the meeting.
References
BERGENSEN, FJ. 1980 Methods for evaluating biological nitrogen fixation. John Wiley
and sons, Ltd.
BUMPUS, ED. 1957 Legume nodulation iu Kenya. East Afi. Agric. For. J. 23, 91-99.
GRAHAM, P.H. 1976 Identification and classification of root nodule bacteria-In : Symbiotic
nitrogen fixation in plants. (Nutman, P.S. ed.). pp. 99-l 12. Cambridge University Press,
Cambridge.
HABISH, HA. and KHAJRI. S.M 1968 Nodulation of legumes in the Sudan : cross-
inoculation groups and the associated Rhiwbium s&aim. Exp. Agric. 4, 227-234.
KEYA, LO.1977 Nod&ion and nitrogen fucaton in legumes in East Africa. In : Biological
nitrogen fixation in farming systems of the Tropics. (Ayan&a, A. and Dart, PJ. eds.).
pp 233-243. John Wiley and Sons, Chichester.
KEYA, N.C.O. and Van EIJNA’ITEN, C.L.M. 1975 Studies on oversowing of natural
grassland in Kenya. 2. Effects of seed inoculation and pelleting on the nodulation
of Desmodirun uncinatum (JACQ.) DC. East Afk. Agric. For. J, 40, 351-358.
LANGE, R-T. 1961 Nodule bacteria associated with indigenous Leguminosae of South-
Western Australia. J. Gen. Microbial. 26, 351-359.
MCDONALD, J. 1935 The inoculation of leguminous crops. &sf Afr. Agric. For. J. 1,
8-13.
MORRISSON, J. 1966 Productivity of grass and grass/legume swards in Kenya Highlands.
East Ajr. Agric. For. J. 32, 19-24.
SOMASEGARAN, P. and HOBEN, H.J. 1985 Methods in Legume - Rhizobium technology.
University of Havaii, NiiAL
Project and MIRCEN.
SOUZA, D.I.A. de 1969. Legume noduhuion and notrogen fixation studies in Kenya.
East Aj-. Agric. For. J. 34, 299-305.
TRINICK, MJ. 1%8 Nodulation of tropical legumes. 1. Specificity in the Rhizobium
symbhsis of Leucaena leucpcephala.
Eap. Agric. 4, 243-253.
VINCENT, J.M. 1970 Amanual for the practical study of root nodule bacteria. IBP Handbook
n* 15. Oxford, Blackwell Scientific Publications.
FIXATION D’AZOTE ASSOCIATIVE
Maximiser la FBA pout la Production Agricole et Forestih en Aftiqve
Effect of inoculation on the growth
and yield of three maize cultivars
IMWAURA, F.B. and WIDDOWSON, D.
Dqwtment of Botany, University of Nairobi, Nairobi, Kkya
Summary
Illocallation exp+mexlra were con&ted using three lnaizc cultivars (z?u &yiL)
and three N,-fixing plant-growth-lvomotinging-rhizobacteria
The plants WQC grown to
maturity and various vegetative and reproductive m were determined. The
results ohtaincd indicated significanr increases in the shoot dry mat&s yield (342%)
of one maize arltivar with only a small inaease in seed dry weight Significant
increases in seed dry weight (38.4%-495%) were mcorded for inoculated plants of
two cultivars. No si&ficaat increase in the cnrde protein of the grain was observed
in the. inoculaled tTeatments.
lntroductlon
It has been suggested that increasing the proportion of rhizosphere w in
grass-bacteria associations, through inoculation, may result in &eased N, fixation and
plant yield (Pan-iquin, 1982 ; Okon, 1984). Such increases in the numbers of N,-hxing
rhizobacteria may be achieved by application of highly competitive bacterial strains, heavy
inoculation or by the introduction of N,-fixing genes to other root associated bacteria
(Kleeberger and Khngmuller, 1980 ; Postgate and Kent, 1987). However results obtained
in inoculation stuclies of cereal crops and forage grasses are very varied ranging from
no response (Schank et al. 1980 ; Wright and Weaver, 1982) to highly significant responses
(Kapulnik et al., 1981 ; Mertens and Hess, 1984 ; Smith et al., 1984). Differences in
inoculation responses attributable to plant genotype effects have previously been demon-
.
.
I
93
trated in wheat and maize among other cereal crops (Kapulnik et al., 1981 ; Reynders
and Vlassak, 1982 ; Millet et al., 1984).
In this study we report the response of three maize cultivars to inoculation with
N,-fixing plant-growth-promoting-rhizobacteria in a tropical soil.
Materials and Methods
Three facultatively anaerobic diazotmtrophs C, J and L previously isolated from
maize plant roots were grown under static conditions at 24°C in a liquid nitrogen deficient
glucose medium (L&Berg and Granhall, 1984 ; Mwaura and Granhall, 1986). The cultures
were grown to stationary phase, centrifuged and washed twice in 0.05 M sterile phosphate
buffer pH 7.0 @erg et al., 1980). The cells were finally resuspended in buffer and trans-
ferred into sterile petri dishes.
Seeds of three maize cultivars LG 11, Makueni and Katumani were washed under
running water and in sterile distilled water to remove the fungicides. The seeds were
germinated on moistened filter paper in the dark at room temperature. The seedling lot
of each cultivar was divided into four batches. For each cult&r, one batch of seeds
was used as a control while the other three batches were inoculated with the three bacterial
strains C, J and L respectively. The seedlings were inoculated by placing them in petri
dishes containing the bacterial suspensions for 30 mn. Control seeds were immersed in
sterile buffer.
Four maize seedlings were sown in 30 1 plastic troughs containing well moistened
soil from a weeded plot. Great care was taken to avoid cross contamination between
the containers. Each treatment was replicated four times and all containers were randomly
arranged in an open space but under a roof of clear plastic sheeting to prevent flooding
of the containers by rain water. The soil in each container was covered with washed
vermiculite lo millhisalgal growth:All plants were watered every two or three days
as necessary. The plants were thinned to three in each container 10 days after planting.
After growing for 110 days, the plants were harvested and their shoot lengths measured.
The shoot, ear and lOOO-seed dry weights were determined in each treatment.
The plant materials were dried to constant weight in a forced-air oven at 80°C.
The materials were then milled and the nitrogen content was determined by the Kjeldahl
method (Hesse, 1971). The crude protein content of the maize grain was estimated by
multiplying the nitrogen content with a factor of 6.25 (AOAC, 1960).
Results
The results are summarised in Tables 1 and 2. Improved vegetative growth was
particularly evident in maize plants of cultivars LG 11 and Katumani inoculated with
the three bacterial strains Substantial increases in plant height (11.6 - 43%) and shoot
dry weights (3-l- 342%) were recorded over the uninoculated plants (Table 1). However,
little or no improvement in vegetative growth was observed in inoculated maize plants
of the Makueni cultivar. Among the three inoculants tested, bacterial strain C was the
most effective in increasing the shoot dry matter yield. Inoculation however did not increase
the nitrogen content of maize plant shoots except for plants of cultivar LG 11 inoculated
with bacterial strain C (Table 1).
Table 1. Effect of jnoculation on the vegetative growth and N-yield of three maize cult'ivars.
Cultivar
Inoculant
Plant height
Shoot d,wt./plant
% N
Ndylel d/plant
(cm)
(g)
(mg)
LG 11
control
130.2 +
12.1
39.1 +’
4.9
0.35
136.9
i_
152.1 7
5.4
51.1 7
2.8*
0.34
J'
186.0 7
5.6
":'
47.7 T
2.5
0.37
;:;*;
,C
180.3 z
9.1
:
52.5 z
4.4*
0.54
283:5.
.:
P
Katumani
control
184.7 t
14.3
:,
77.6 2
7.0
0.65
504.4
206.2 7
6.6
8.1
0.64
512.0
!i
225.5 7
15.4
,.'
0.61
C
244.9 7
9.2
9712
;i.;';2' ii::*
0.57
Makueni
control
175.1 +
8.9
75.4 t
6.6
0.41
309.1
J'
196.8 187.7 7 11.9 10
80.0 76.1 7
7.3
oi.39
296.8
L
8.1
0.38
304.0
C
162.7 z 11.5
1.
74.2 z
6.0
q.39
289.4
* significant increases (P = 0.05').
Results are mean of six replicates plants t S.E.
_,
-
.
._ _
.
.
I_-. --
. ..-.........
_ ..^. ..-..--
_.
..--&-
--..
-.
_-_-
__
.
_
95
With regard to the reproductive parameters, ear dry weight and the mean seed
dry weight increased in inoculated plants compared to those of uninocuhued treatments.
‘Ihe most significant responses were observed in Kammani and Makueni cultivars for
which the lOOO-seed dry weight increased by up to 49.5% and 38.4% respectively frable
2). The results indicate that bacterial strains J and L were the most effective in promoting
reproductive growth of the maize plants. Higher ear and seed dry weights were observed
in maize plants inoculated with the two strains compared to those of plants inoculated
with bacterial strain C. The crude protein content in the grain was however unaffected
by inoculation except for the Makueni cuhivar in which higher crude protein levels were
recorded in inoculated than in the uninocuJated treatment (Table 2). It was also noted
that the inoculated maize plants flowered and matured about 10 days earlier than the
control plants.
Discussion
In addition to the inoculated plants maturing slightly earlier than the controls, ino-
culation increased plant height and dry matter yield Earlier heading and flowering has
similarly been observed in wheat, Setaria iralicu, sorghum and Punicum sp. in the green-
house and in the field after inoculation with N,-fiig bacteria (Kapuhtik et al., 1981
; Kapulnik et al., 1983 ; Lin et al., 1983 ; Yaholon et al., 1984 ; Sarig er al., 1984).
Jnocnlation of the Makueni cultivar did not promote vegetative growth of the maize plants
but the seed dry weight increased by up to 38.4%. Large increases in ear dry weight
(31.9%) and seed dry weight (49.5%) were also recorded for the Katumani maize cuhivar
inoculated with the test bacteria. The observed differences in make cultivar response to
inoculation with the N,-fixing rhizobacteria were rennuI&le in that the third cultivar LG
11 responded best in terms of plant dry matter yield increases but recorded the lowest
increase in seed dry weight among the three cultivars. Similar genotype effects have been
established for maize (Kapuhuk et al., 1981 ; He@ and Monib, 1983) and other cereals
(.YMillet et al., 1984).
In Israel, Kapuhtik et &(1981) reported seed and cob dry weight increases of
28.2% and 34.4% respectively in inoculated maize grown under field conditions. Few
published reports on inoculation effects in maize and other cereals in Africa are presently
available. In Egypt, Hegazi and Monib (1983) reported increases of 150-170%
180-270%, 120-130% for straw yield, grain yield and total N yield respectively in maize
inoculated with AzospiriUum
sp. under field conditions. More concerted efforts need to
be made to study the potential for increasin g grain and forage yield of cereal crops
through inoculation with growth-promoting-rhizobacteria.
‘.
.: ..*
. .
-
.-
:_
Table 2.
Effect of inoculation on yield parameters of three maize cultjvars.
Cultivar
Inoculant
ear d.wt,
lOOO-seed
mean seed
X crude
/plant 19)
d.wt. (g)
d. wt. (mg)
protein
-
I
LG 11
control
57.3 f 5.2
*I
129.8
129.8
13.1
J'
56.7 54.0 7 0.5 4.8
._
130.8 135.2
130.8
12.8
-..
',
135.2
C
67.7 2.13.2z .:',
141.7
',
141.7
1;::
Katumani
control.
32.9 ,+ 4.3
74.1
74.1
-
13.2
:
43.4 38.4 7 7 7.2 3.7,
i
110.8
9660
96.0
:
10.5
11.6
c
"
39.2 2 4.1
:
76.5
':E
.
14.1
Makueni
control
4442 t 1.9 i"
81;6
81.6
46.4 7 10.7
112.9
112.9*
109*:
:
49.8 7 6.6 +.
c
.._
42.8? 2,.5 :"'.-
'07K
'%
.'
10:5
.
.F.
.
11.4
",e #
: ,. *
* signlflcant'increases (P = 0.05).
, ,..I.'
,, .,
Results are means of SIX replicate plants 2 S.E.:,Ears of replicate plants were pooled, shelled and the
grain dried to constant welght.
'
.
.
.
+.
.’
:.
.-
.+
97
Acknowledgements
We wish to thank the International Foundation for Science and the Kenya National
Council for Science and Technology for financial support to one of us (Mwaura, B.)
References
AOAC. 1960 Methods of Analysis. Washington D.C.
BERG, R.H., TYLER, ME., NOVICK, NJ., VASIL, V. and VASIL, Ix. 1980 Biology
of Azospbillum - sugarcane association : Enhancement of nitrogenase activity. Appl.
Enrivon. Microbial. 39, 642-649.
HEGAZI, N.A. and MONIB, M. 1983 Response of maize plants to inoculation with azos-
pirilla and (or) straw amendment in Egypt. Can. J. Microbial. 29, 888-894
HESSE, PP. 1971 -A textbook of soil chemical analysis. John Murray Publishers.
KAPULNIK, Y., SARIG, S., NUR, L and OKON, Y. 1983 Effect of Azospirillum inocu-
lation on yield of field-grown wheat Can. J. Microbial. 29, 895-944.
KAPULNIK, Y., SARIG, S., NUR, I., OKON, Y., KIGEL, J. and HENIS, Y. 1981 Yield
immases in summer cexeal crops in Israeli fields inocdakd with Azospirillm Expl
Agric. 17, 179-187.
KLEEBERGER, A. and KLINGMULLER, W. 1980 Plasmid mediated transfer of nitrogen-
fixing capability to bacteria from the rhizosphere of grasses. Mol. Gen. Gem. 180,
621-627.
LlN, W., OKON, Y. and HARDY, R.W.F. 1983 Enhanced mineral uptake by Zea mays
ad Sorghum bicolor mm inoculated with Azosptillwn bradense. Appl. Environ.
Microbial. 45, 1775-1779.
LINDBERG, T. and GRANHAIL,, U. 1984 Isolation and characterisation of dinitrogen-
fixing bacterial from the rhizosphere of temperate cereals and forage grasses. Appl.
Environ. Microbial. 48, 683-689.
MERTENS, T. and HESS, D. 1984 Yield increases in spring wheat (Triticum aestivum
L.) inoculated with Azospirillm lipoferum under greenhouse and field conditions of
a temperate region. Plant ana’ Soil. 82, 87-99.
MILLET, E.. AVM, Y. and FELDMAN, M. 1984 Yield response of various wheat geno-
types to inoculation with Azospirillum bradense. Plant and Soil. 80, 261-266.
MWAURA, F. and GRANHALL, U. 1986 Nitrogen fixation (CA reduction) associated
with Maize (Za mays L.) in a Swedish soil. Swedish J. agric. Res. 16, 49-56.
OKON, Y. 1984 Response of cereal and forage grasses to inoculation with N2-fixing
bacteria. In : Veeger, C. and Newton, W. E. (eds.). Advances in nitrogen fixation
research, pp. 303-309. Martinus Nijhoff/Dr. W. Junk Publishers, The Hague.
PATRIQUIN, D.G. 1982 New developments in grass-bacteria associations. In : Subba
Rae. N.S. (ed.). Advances in agricultural microbiology. pp. 139-190. Butterworth Scien-
tific, London.
POSTGATE, J.R. and KENT, H.M. 1987 Qualitative evidence for expression of Klebsiella
pnewnoniae Nif in Pseudomonas put&. J. Gen. Microbial. 133, 2563-2566.
98
REYNDERS, L. and VLASSAK, K. 1982 Use of Azospirillwn brasilense as biofertilizer
in intensive wheat cropping. Plant and Soil. 66. 217-223.
SARIG, S., KAPULNIK, Y., NUR, I. and OKON, Y. 1984 Response of non irrigated
Sorghum bicolor to Azospirilium inoculation. Expl. Agric. 20, 59-66.
SWANK, S.C., SMlTH, R.L. and WEISER, G.C. 1980 Responses of two Pearl Millets
grown in vitro after inoculation with Azospirillum brasilense. Soil Crop Sci. Sot. Fla.
Proc. 39, 112-115.
SMITH,R.L., SCHANK, S.C.,MILAN, JR. andBALDENSPERGER, A.A. 1984Responses
of Sorghum and Pennisetwn species to the N,-fixing bacterium Azospirillwn bradense.
Appl. Environ. Microbial. 47, 1331-1336.WRIGHT, S.F. and WEAVER, R.W. 1982
Inoculation of forage grasses with N,-fixing Enterobacteriaceae. Plant and Soil. 65,
415-419.
YAHALON, E., KAPULNIK, Y. and OKON, Y. 1984 Response of Setaria italica to
inoculation with Azospidlum brasilense as compared to Azotobacter chrooccocum.
Plant Soil. 82, 77-85.
MW la FBA pour la P&don Agricofe cl For&e en Ajkiquc
Population of-~i&ro~e&ixhg bacteria
in sweet potato fibrou& ri3ots
HORTENSE, W.DP’, WALTER, A.HF,
IMULONGOY, K.t2’, ADEYEYE, S.0.‘2’,
and HAHN, SX.“‘. ”
(1) Tiuk-egee University. Alabama,VSA;
_,
{2). I- Znstit+ .~..T~~~.:~~~~=,,~.Niguin.
:
. .
Abstract
Fibrous e’s of six sweet ~q~~<ul$~ars w~,ev@ated
for rhizosphese popula-
ti. of N$.&y&.;~
JuWAikc.,~~k‘~aamplingdates.PopllatiomOf
the bacteria were de&mix& using culture tubes (pellicle formation and acetylene
reduction activity), and colony plate counts. Plant growth parameters evahuted
were : fresh and dry weights of storage roots. foliage, fibrous roots and total
biomass. The bacterial population ranged from 10’ to 106 g per dry root weight
depending on the counting method used- The N$xing bacterial populations were
positively correlated (r = 0.67 to 0.88) with foliage arid total biomass dry weight
for cultivars TIS 70357 and TIS 8441 ; for the same parameters, negative corre-
lations (r = -0.68 to -0.80) were found for cultivars TIS 9265 and Tlb 4 and corm-
lations were not significant for aikivars TIS 2498 and TLS 8504.
100
Introduction
Sweet potato cultivars from the breeding program at the International Institute of
Tropical Agriculture @TA) produce fi-om 40 to 50 mt/ha of storage roots and up to
31 mt/ha foliage when grown in soils with low soil N indexes (IIahn et al.. 1984 ; Hill
et al., 1985). Azospidum-like bacteria am widely distributed in tropical and temperate
region soils and plants, and have been suggested as important associative N,-fming bac-
teria under the direct influence of plants. Several strains of A. brudense and A. Zipoferum
associated with the roots of different sweet potato varieties have been isolated and characte-
rized (Hill ef al., 1983 ; Hill and Bacon, 1984). Reviews of Azospirillum inoculation
studies by Boddey and Dobereiner (1982) and Dart (1986) show inconsistent responses
of cereals and grasses for several parameters including mtrogenase activity, nutrient
uptake and top dry matter. When sweet potato was inocul& with Azospidhm bra-
dense, the root yield .and N ‘content were increased (Crossman and Hill, 1986).
The purpose of this study was to quantify naturally occuring nitrogen fixing ba-
cteria in sweet potato rhizosphere and determine the effect of fertilizer nitrogen addition
on growth and microbial parameters, and to evaluate relationships between sweet potato
growth parameters and parameters used to measure associative N, fixation.
Material and Methods
Six sweet potato cultivars were selected from the IITA breeding program : TIb4,
TIS 2498, TIS 70357, TIS 8441, TIS 8504 and TIS 9265 were planted in loamy sand
at the IlTA Ihadan, Nigeria. The experimental design was a split plot with two rates
of fertilizer N (0 and 50 kg of calcium ammonium nitrate N ha-‘) as main plots and
cultivars as subplot treatrmzirts. Each treatment con&ted of four rows of a sweet potato
cultivar, 13.2 m long and Im wide each, with plants spaced 0.5 m within rows. At 6,
10 and 14 weeks after planting (WAR), four plants were collected at random f?om the
left center row, and storage roots, foliage and fibrous root weights and associative N,
fmation parameters were evaluated. At each sampling date, sub-samples of 2 g of fibrous
roots plus adhering soil were analysed for number of Azospirillum-like bacteria, using
the serial dilution method described by Wollum (1982) for colony, plate count and the
most probable number method described by Knowles (1984). Malate N free media was
used for both solid plates and semi-solid tubes. Acetylene reduction assay (ARA) was
carried out on all tubes -showing pellicle formation, using a Carle Age 211 Gas Chro-
matograph’ equiped with a flame ionisation detector.
Results and Discussions ..
Storage root weight was higher for TIS 8504 and 8441 than for the other cultivars
at 6, 10 and 14 WAP irrespective of fertilizer treatment (Table 1). Differences in total
dry biomass and storage root weights were most apparent at 14 weeks after transplanting
for both 0 and 50 kg N ha-‘. Fertilizer N addition tended to depress storage root yield
and foliage weight of most cultivars. Acetylene reduction activity was detected in culture
tubes at 10’ dilution for all six cultivars at all three sample dates suggesting the presence
of N,-fixing bacteria in the rhizosphere of all the cultivars studied. The most probable
numbers (MPN) of N,-fixing bacteria were low, and significant differences were not apparent.
101
Table 1 : Effect of fertilizer N (50 kg N ha-') on sweet potato storage
roots and foliage growth.
Tuber wt
Foliage wt
Cultivars
(kg ha-l)
(kg hi-11
N = 50
N = O
N = 50
N = O
TIS 8504
824 a
635 a
1822 b
2162 a
TIS 8441
492 b
548 b
2291 ab
2075 b
TIS 2498
319 cd
330 c
1547 bc
1439 bc
TIb 4
390 c
374 c
1322 c
1172 c
TIS 9265
156 d
147 d
1967 b
1705 bc
TIS 70357
188 d
151 d
2502 a
2227 a
Means with the same letter in the same column are not significantly
different at the 5 X level by the Duncan's multiple range test.
The lowest yielding cultivar TlS 70357 had a significantly higher MPN and percentage
of N,-fixing bacteria (Table 2). Significant correlations (r = 0.67 to 0.88) were found
for foliage and total dry biomass weight, with MPN of AzospitiZZIun-like bacteria in the
rhizosphere (Table 3).
Correlations between Nz-fixing bacterial population and foliage weight or total biomass
were positive for TlS 70357 and 8441 suggesting that Azospirilhm-like bacteria were
responsible for the increase in foliage production. Correlation for the same parameters
were negative for TIb 4 and TlS 9265 (r= -0.68 to -0.80) and not significant for cultivar
TIS 2498 and 8504.
These results suggest that associative N2 fixation in sweet potato is cuhivar dependent,
and that foliage and total biomass production of some sweet potato cultivars on soils
with low N indexes may be related to the presence of N,-furing bacteria However the
association between Azospidlum-like bacteria and sweet potato fibrous roots seems to
be regulated by complex mechanisms that need to be more investigated.
102
Table 2 : Number of.Azospirillum bacteria in sweet potato fibrous roots
by the acetylene'reduction assay (ARAB and the most probable
number (MPN)
Cultivars
ARA
Log ARA
MPN
Log MPN
105 /g roots
105 /g roots
TIS 8504
8.20 b
3.25 abc
0.090 a
3.75 a
TIS 8441
9.86 b
3.17 bc
0.067 a
3.62 a
TIS 2498
11.79 b
2.63 c
0.042 a
3.50 a
TIb 4
5.77 b
3.34 abc
0.055 a
3.67 a
TIS 9265
7.07 b
4.09 ab
0.089 a
3.72 a
TIS 70357
88.31
a
4.57 a
0.010 a
3.72 a
Means with the same letter in the same column are not significantly
different at the 5 % level by Duncan's multiple range test.
103
Table 3. Correlation coefficients between selected growth.para&ters and -
rhitosphere Np-fixing bacterial count for six sweet potato -'
I,
cultivars. -
Log
Log
Log -. ;
CULTIVAR
MPN-A'
HPN-N2
CPC3
HPN-A
WPN-N
CPC
TIS 8504
FDW4
N S
NS
NS
NS NS
NS
TD85
N S
NS
NS
NS
NS
NS
TIS 8441
FDW
NS
0.75*
0.84***
NS
NS
NS
-
TDB
NS
0.77**
0.88***
NS
NS
NS
TIS 2498
FDW
NS
N S
NS
NS
NS
NS
I
FDB
NS
N S
NS
NS
NS
NS !
Jib 4
* FDW
-0.80**
N S
NS
-0.7&
-0.76*
. I
-0.70*
TDB
-0.72*
N S
NS
-0.78** N S
NS
TIS 9265
FDW
N S
N S
NS
NS
NS
NS
TDB
NS
N S
NS
-0.75**
-0.68**
NS
TIS 70357
FDW
0.71*
N S
0.75*
0.72*
0.67*
NS
TD8
N S
N S
NS
NS
NS
NS
1. Most probable numbers of Azospirillum
2..Most probable numbers of Ricronitropliilic
bacteria
3. Colony plate counts
4. Foliage dry weight
5. Total dry biomass
*Significant at the 5% level
l +Significant at the 1% -level
***Significant at the 0.1% level
‘..‘.
.’
104
Acknowledgements
This paper is contribution No. PSOO8 of the George Washington Carver AgricuI-
turaI Experiment station, Tuskegee Univexsity. The research was supported by funds fi-om
the U.S. Agency for International Development (project No. DAN-5053-G-55-6062-00)
and the I$SDA/CSRS @ant No. ALX-SP-1).
References
BODDEY, RIkL and DOBEREINER,J. 1982 Association of AzospirifZum and other dia-
zotrophs with tropical gmminae. In : Non symbiotic nitrogen fixation and organic
matter in the tropics. pp. 28-47. Symposium Papers. I. Transactions
12th Int. Cong.
Soil Sci New Delhi, India.
CROSSMAN, S.M. and HILL, WA.. 1986 Inoculation of sweet potato with Azospidlum
brasilense. Hart. Sci. 18, 169.
DART, PJ. 1986 Nitrogen fixation ass&a&d with non-legumes in agriculture. Plant and
Soil. 90, 303-334.
HAHN, SK, ALVAREZ, UN., CAVENESS. F.E. and NG, S.Y. 1984 Sweet potato
improvement at the International Institute of Tropical Agriculture. Proc. 1st Ckibean
Regional Workshop in Tropical Root Crops. pp. 93-97. University of the West Indies,
Kingston. Jamaica.
HILL, W.A. and BACON, P. 1984 Fertilizer N use efficiency and associative N, fixation
with non legumes. Proc 6th Int’l. Symp. Tropical Root Crops. CUP. Lima, Peru, Feb.
21-26, 1983.
HILL, WA, BACON, P., CROSSMAN, S.M and STEVENS, C. 1983 Characterkation
of N2-fixing bacteria associated with sweet potato roots. Can. J. Microbial. 29,
860-862.
HILL, W.A., HAHN, S.K. and MIJLONGOY K. 1985 Most probable numbers of nitrogen-
fming bacteria associated with sweet potato roots. Proc. 3rd Symposium Int. Sot.
Trop. Root Crops, African branch, Owerri, Nigeria, 16-23, August 1986. IDRC, Ottawa
Canada.
KNOWLES, R. 1984 Free living dinitrogen bacteria. In : Page, AL., Miller, RH. and
Keeney, D.L. (E&z.) Methods of soil analysis, part 2, Chemical and Microbiological
properties.p. 1070-1090 Agronomy Monograph Np 9, American society of Agronomy,
Madison, WI.
WOLLUM, A.G. 1982 Cultural methods for soil microorganisms. In : Page, A.L., Miller,
R.H. and Keeney, D.L. (&is.) Methods of soil analysis, Part 2, Chemical and Microbio-
logical properties. p. 781-802. Agronomy Monograph Np 9, American society of Agro-
nomy, Madison, WI.
ASSOCIATION AZOLLA / ANABAENA
MaruKrer la FBA pour la production Agricole et Forestière en Afrique
La recherche sur AzoZZa et ses applications :
évolution et tendances actuelles
VAN HOVE, C.(l) et DIARA, H.F.@)
(1) Ldzboruto~ de physiologie v&&ale, hiversit~ Catholique de haùh
Place croix ak sud, 4. B-I348 LmvahJa2deuvc. Belgïum.
(2) Association pour le Dtfveloppemmt
de la Rizicuhre
en Afrique de I’OM (ADRAO), BS. 96, Saint-buis, Sknpgal
i
Summaty
AwIIa uses as biofertilizer and food, traditional in some par& of China and Vietnam,
havebeenignoredarneglectedinothcrcounhies(andeveninwideareasoftbese
cmntrks) uutil the si~ties, when research programmes were launched with a view
to inueasing areas under Awlla cukivation and impmving cultural pra&e.s adap
ted to various mmmnes. Sincc thcn resxch, pure and applied, has blossomed, and
effurtstojntroduceAwZiammanykriancountriesbutalsomotha~ofthe
“
world~veinrreasod,withvarying~sts.ThepuIposeofthispapaistooutline
the major steps of ouf lcnowledge an AwZ& to bring out the main prt&nt re-
search trends, to descrîble thc various possible AzoZZu uses and theii m,
withspecialemphasisOn~SUgg~~~finanypnsenttd,~atpromoting
Azolkrinoroductioninrhiscontmen5~yf~søanfarmers,wbo’arethemain
potaltid beneficiaries.
Introduction
L’objectif du prbent document est de dresser un bref historique des principales
connaissances acquises sur la fougkre aquatique Azollu, de degager les grands axes de
recherche en cours, de dkrire les diverses possi%ilitt% de valorisation d’AzolZu en agriculture
:
107
et les problemes poses par son utilisation, l’accent étant mis sur le cas particulier de
l’Afrique. Des suggestions sont enfim émises en vue de promouvoir l’azolliculture dans
ce continent, spécialement aup&s des petits paysans qui en sont les principaux b&&-
ciaires potentiels.
L%volutior) de la Recherche
Des origines ii la crise p&roll&e
La valeur fertilisante et ahme&re de la fouragere aquatique AzoZZu est connue
et mise à profit depuis de nombreux siècles dans certaines figions tri2 lirnitks de Chine
et du Vietnam ; situ6es le long de la côte entre le 18ème et le 3Oi?me degre de latitude
Nord, ces r&gions se caractkisent ‘par un climat subtropical humide, à variations saison-
nières de temp&ature souvent consiMles. Jusqu’en 1960, la maîtrise d’Azollu y &ait
essentiellement empirique.
Dans le monde occidental par ailleurs, dès le début du dix neuvième siècle appa-
raissent quelques études fondamentales sur AzoZ2u ; ce n’est cependant qu’au début du
vingtième siècle, en grande partie gr& B l’intervention de chercheurs travaillant en Jndo-
chine, que l’int&& pratique d’AzolZu est signal& C’est ainsi qu’il est fait mention, sans
d’ailleurs que des arguments probants soient. toujours pr&ent&, & l’effet d@essif d’une
couverture de plans d’eau par Azolla, sur le &veloppement des moustiques (Smith, 1910),
de la capaciti de rkluction de l’azote (Ces, 1913), .du file fertïkant d’AzolZu en riziculture
et de sa valeur alimentaire pour des animaux d’blevage (Chevalier, 1926), de l’action
inhibitrice d’AzoZZu sur le développement des adventices (Braemer, 1927) ou enfin de
son influence negative sur les pertes d’eau par évapotranspiration (Nguyen, 1930). On
avait donc, des cette époque, relevé l’essentiel des proprit& pratiques d’AzoZlu conuues
aujourd’hui, mais il faut bien recormaît~~ que ces informaticms ne suscit&rent pas grand
int&êc
En Crient par contre, dks la fin des annees 50, les gouvernements vietnamien et
chinois décidèrent de promouvoir la recherche scientifique ayant trait a Azollu. J.l en re-
sulta un foisonnement & travaux, en g&u?ral a orientation appliquée. En Chine, Lumpkin
et Plucknett (1982),par la mise au point & techniques diverses de conservation en saison
froide, permit d’étendre consid&ablement l’aire de culture vers le Nord, jusqu’au delà
de Beijing ; dans le Sud par ailleurs des techniques & protection contre les tempkatures
élevées de l’été permirent~i’utilisation d’Azollu pour les cultures automnales.
L’étude des principaux .mseUes..@@tetns ,d’Aw&x (qui constituent un autre pro-
blktne majeur en Asie) et des modes &,lutte contre ceux-ci, ainsi que le développement
de techniques d’enfo uissement d’AzoZlu entre les Lignes de riz am&ior&ent quant à eux
I’effkacit6 de l’azolliculture.
Du Côt6 vietnamien (Dao ami Thuyet, 1979). les efforts ‘Jxntkent sur la s6lection
d’écotypes (indig&nes) d’A. pinnuru var. irnbkuta et sur l’am&oration des pratiques cul-
turales en association avec la culture du riz d’hiver, avec pour r&uhat un accroissement
considéxable des surfaces couvertes. lXs 1965, se développa la monoculture hivernale
d’AzolZu, dans divers systkmes de rotation. En même temps, de nombreuses recherches
portant sur les prkiateurs, les exigences en phosphore, la conservation d’Azolfu en saison
chaude, etc..., ttaient entreprises.
108
L’explosion des années 1970
c’gst durant les annees ‘1970, en particulier grâce aux publications vietnamïemres
plus ais&nent accessibles que celles provenant de Chine, que les rnilkux scientifïques
extkaiy a ces deux.~~ys pr$ent rkllement conscience de l’inter& potentiel d’Adla
eti igri~m :; @~Ymonographi~,de Moore (1969) joua certakment un r& knportant
acet~~.La,aupointen1%7&latechniquede mesure de sactiti nitrog&
aisue QJJ pstchromatographie gazeuse suscita par ailleurs un regain g&.&al d’intfZzi?t
pour l’&ude de la fixation biologique de l’azote. La crise pétroliere de 1973 enfin fut
un stimulant puissant a la recherche sur les engrais biologiques A cette @que com-
mencent dès @rs B se multipliex les programmes de, recherche sur ,&oZZu de par le monde.
Parmi les th&nes’de~recherche initiés et les acquis majeurs des ann&s 1970, on
retiendra particuli&ement :
‘k &f&n&o~ de la localisation de la nitrogtkase au sein d’Anubuenu
l
(Peters and .Mayne, 1974) ;
les preniières applications de la méthode isotopique au lsN pour l’étude
l
des relations entre les partekes de la symbiose (Pe&rs et al., 1977) ;
la description d&aillee du cycle de ~développemént d’AzoZZa (Konar and
l
lbpoc=, 1974) ;
l’am&xation des connaissance sur les relations morphologiques et phy-
l
siologiques entre Azolla et Anabama @Sers, 1984) ;
les tentative de culture in vitro d’Anabaena ~azollae (Bai et al., 1978 ;
l
Newton and Herman, 1979) ;
la confirmation de l’importance du phosphore (Subudhi and Watanahe,
l
1979).; . : ‘,
,
‘.d -“:9 la’conf&n&ti~~~~de l’effet «herbicide» d’Aha .(N$, 1973 ; Talley et
az., 1977) ;
. le développement des recherches sur la composition chimique d’AzolZa
et sur son utilisation comme aliment (Buckingham et al., 1977, 1978;
Subudhi and Singh, 1978, 1979) ;
l’étude de la d6composition d’AzoZla dans le sol (Brotonegoro and Ah-
l
,
$dkadir, 1978) ;
les essais de compostage d’AzolZa (Tran and Dao. 1973) ;
l
l%mod&on et la diffusion d’Azolla jiliculglaes:en Chine en 1977, et
l
. .
,.! des autres .esp&=.+xr 1979 &un#n ,a& mucloleta, 1982)
l’organisation des premiers essais coordonnb de l’International Network
l
on Soil Fertility and Fert&er Evaluation for Rice (INSEEER) sur AzolIa
(Watanahe, 1987) ;
l’@iation des pmmi&res grandes collections de &erence d’AzolZa 31 l’in-
l
temational Riche Research Institute, aux Philippines et a I’Unviversiti
Catholique de Louvain, en Belgique.
Les ann6es 1990 : I’Age d’Or 3
La stimulation de la recherche sur AzoZla pendant les anntks 1930 et les espoirs
qu’elle suscite sur le plan des applications donnent lieu à l’organisation de conf6rences
109
internationales CI Porto Rico (Silver and Schroder, 1984) et en Chine @IRI, 1987) et
de cours de formation sur Azollu. au P&ou et au S&I&~ en 1983, en Thailande en 1984.
en Chine en 1987.
En 1982, la FAO et l’Agence Internationale de I’Energie Atomique développent
un progmnme de recherche coordonm? sur Azolla, tandis que paraît une excellente et
tr&s compl&e monographie (Lumpkin and Pluknett, 1982). qui fait encore autoriti à ce
jour. En Chine, le CentreNational de Recherches sur Azolla (NARC) est inaugun5 2
Furhou, en 1985. Le rythme des publications continue a s’acc&%er et des prog&s décisifs
emq$str&s ,:
‘* <
le prob&ne~de l’identification d’AzoZZu au niveau spécifique et subspe-
cifique, crucial pour les milisateurs, s’il n’est pas rt%olu, fait l’objet d’études
approfondies basées sur les méthodes classiques (Pe&ins et al., 1985 ;
D&am and fowler, 1987), et de chimiotaxonomie (Mai Kodomi et al.,
1984 ; Zimmerman et al., 1984 ; Zimmerman et al., 1988) ;
la cara&risation des Anabaena pr&ents dans les diverses e~pkes d’Azolla
l
progresse grike à la g&r&ique mol&rlaire (Cohen - Bazire and Franche,
1985 ; Franche and Cohen - Bazire, 1985), et aux techniques des anticorps
monoclonaux (Gates et al., 1981 ; Ladha and Watanabe. 1982 ; Tang
et az, 1987) ;
l’intervention d’un troisième partenaire, probablement obligé, de la sym-
l
biose, a savoir une bactérie (du genre Arrhrobacter ?) semble se confirmer
(Gg et al., 1980 ; Wallace and Gates, 1986 ; Formi and Grilli Caiola,
.
- ,
l’insensibïïtt! relative de la nitrog@ase du complexe Azolla - Anabaena
l
&IX ,compo& azptés est mise en évidence, avec de fortes diff&ences
sp&ifiques’ et’ subsp6cifiques (Okoronkwo et aZ., 1989) ;
des différence considdrables quant a la valeur alimentaire des diverses
l
espèces d’AzoIZu sont démontrrees (Antoine et al., 1986 ; Sanginga and
Van Hove, 1989) ;
les $remih hybridations entre esp&es d’AzolZa sont r6alis4e.s (Wei et
l
af., 1986) ; dans le même ordre d’idée, de nouvelles combinaisons sont
obtenues par transfert d’Anabuenu provenant d’une espèce d’AzolZa 21 une
autre- esp&% pn%lablement dt%aras& de son propre symbionte (Lin et
al.;388);:; L
;
&.-&t&.h G!@j&&on de biogàz a p& #holIa sont entrepris
-
l
(charlier et az.. 1983) ;
un systhe de cultm-e intiti Riz-Azolla-poisson est testé, apparemment
l
avec des r7%uhats t& positifs (L.iu, 1987) ;
le pouvoir ~accumulateur de potassium d’AzoZZu est mis en évidence
l
ainsi que divers avantages de la culture d’AzoZZa sur sol humide (Liu,
E%n ;
enfin, les Uanges de souches entre centres de recherche et les essais
l
d’adaptation dans diverses 6coxones s’intensifîent.
‘. . .
:
:
’
:
.,.i
.._
:..
.
.
‘.’
_
.’
Y
.”
..;
.:
,Y.,._,
.,
.,.
.
.
.
-....
110
Les principaux modes actuels d’utilisation d’Azo!la
Azolla comme engrais vert
C’est principalement en culture h-r&& qu’Azo&z est utilis6 comme engraisvert.
nyestc~tiv~enassociatianaveclerizetenfolrigImeou&~reprisesau~du
&veloppement de ce dernier ; lorsque le calendrier cultnral et la dïspom%iW en eau
le permettent, une mokcuhure d’tlzolla et un enfotiat avant repiqkge du riz peuvent
~JII%&X la dam associée.
ces pratiques foumisent de I’amte au riz mais a outre inhibent les adve&ces,
am&oreIlt la structure du sol et diminuent les pertes d’ea&l par kipotmnspi&œl.
D’autres cultures inigukes, telles que celles du Lotus et de la Colocase (ta@,
sont Cgalement fert&&s de cette manike.
Azollu peut -enfin servir d’engrais vert en cultures maraîchères ou horticoles de
pleine terre (par exemple jardin potager jouxtaut les p&ni&res de maintenance) ; il-peut
y être incorpo~ au sol & l’&at frais ou sec (en vue de facilitez le transport ou de différer
son emploi), ou apri3s compostage, seul ou en m&nge avec d’autres composants tels
que la paille de riz.
Azolla comme aliment
:,..
L’amBioration des cormaissances sur la valeur alimentaire d’Azont.l, G lltkmsité
de valorisez la biomasse produite dans les pépiniikes en dehors des saisons,d’utili&ion
en r%re, l’extention de l’azolliculture dans le nord de la Chine, r&ion à traditiond’élevage,
ont eu pour effet de stimuler l’emploi d’AzoZlu comme aliment pour divers animaux,
principalement porcs et volaille mais aussi ruminants et plusieurs esp&Xs & poissons
(Su~hudhi and Singh, 1978 ; FAO, 1985, 1985b, ; Que@ii et al., 1986)2n &gle gt%-
rale, il ,semble qu’AzoZZa puisse se su~tituer à en@o~..2@% des raisons classiqks, mais
il faut reconnaître que des 6tudes compl~mentaiks sont indispensables pou+ &ner ces
estimations. Quoi.qu’il en soit cette situation d’AwZZa est en forte expansion actuellement.
Azolla en systbmes d’exploitation complexes
Comme signal6 plus haut une technique promeueuse consistant a associer riz, pois-
sons et Azolla a 6t6 rtkemment mise au point, chacun des partenaks b&&iciant de la
prt5sence des autres. Cette--technique se développe actuellement dans plusieurs pmvinces
de Chine (L~U Chung Chu, Comm. Pers.).
+
Un systkme Mgr6 impliquaut culture d’Azo&z,, &wage de volailles en voli&res
1 ti piloti su$om&nt des étangs a Tilapia et &eva&‘de porcs est &alernent B l’essai
en Côte dIvok
La production de biogaz enfin, par fermentation anakobie d’un m6lange d’AzoZZu
et de paille de riz, et la rkup&ation de l’effluent du digesteur pour fertiker les cultures
voisines semblent donner des &suk& positifs aux Philippines (Johnson, 1988).
La situation en Afrique
Les premiem rtisultats portant sur le potentiel d’AzoZZu en agriculture africaine (Roger
et Reynaud, 1979) suggkaient qu’Atolla &icanu (A. pimuta var. pùmata) pourrait être
valoris6 dans la val& du fleuve S&@al pour autant que le probEme du manque d’eau
111
en saison sMe soit rksolu. Des essais d’adaptation de cette souche en riziere mont&ent
cependant sa sensibilite aux basses temperamms de décembre a f&ier (Ton That, 1980).
En 1980, I’ADRAO lance, en collaboration avec Wniversid Catholique de Louvain
(Belgique), un programme derecherche ayant pour objectif de tester le potentiel de valori-
sation d’Azdla dans:diverses conditious kologiques d’Afrique. de I’Ouest. Les premiers
essais sont rtMis6s a Richard Toll ‘(Sénégal) et B Rokupr (Sierra Leone) ; ils portent
sur l’adaptation d’une-vingtaine d’kotypes reprkentant les sept espkes et deux vari&%
d’AmIla. 11 apparaît rapidement que certains écotypes s’adaptent bien dans chacune des
stations, et que l’introduction d’espkes &rangeres est essentielle, les souches d’origine
loeale~&ant..en.-gMral p e u performanks..:
‘_
Suite ii ces r6sultats encourageants, la collection de refere.nce est progmssivement
enrichie en vue d’identifier des souches a haute productivité ; des m6thodologies adaptées
à l’étude compark de ces souches en conditions contrôhks et en champs sont mises
au point
“, . .
Les premiers essais de fertilisation -des rizieres sont entrepris en 1981 ; les effets
escomptes (apport d’N, inhibition des adventices) sont manifestes
Un cours et des stages de formation sont organises a l’intention des chercheurs
des pays membres de I’ADRAO et une plaquette sur Azolla est difkst% dans les milieux
intétessés (Van Hove er al., 1983).
Ces actions, ainsi que di- iutementions de la FAO, sont suivies de la mise
en place de programmes de reche,rche dans plusieurs pays d’Afrique de l’Ouest, ainsi
qu’au Burundi, au Rwanda et en Tanzanie.
Plus rkemment enfin des acti&s de vulgarisation et d’encadrement en milieu
paysan ‘ont d6buté.
De l’expérience ahsi acquise (Diara and Vand Hove, 1983 ; Diara et al., 1984a,
1984b. 1987 ‘i Van ‘Hove,: 1987, ;:;..Vji iuove and Diara, 1987a, 1987b ;. ,Van Hove er
al., 1983, ‘1987)“Se dégagent. les kon&isions suivantes :
’
les conditions climatiques de nombreuses r&ions d’Afrique sont fâvo-
l
rables a une bonne, ou même à une excellente productivité d’AzoZZa, et
ce pendant toute l’année, pour autant qu’un choix approprié d’kotypes
soit effectue ;
les prédateurs d’Azq& qui empêchent ou contrarient souvent l’utili-
l
sation â’Azolla en Asie, ne provoquent pas de dégât sérieux en Afrique.
Cet& situation n’est cependant sans doute que momentan&z et il faut s’at-
I
tendre, s~.yt~$t+y$y .y!? tr@=Ja hqi$y. B =: que. œmh~ de
ces pt-édatéurs’ tiiiiumyu, &ja pr&entS en A~&U< posent pkbli?me au
fur et à me.s&$re I’azolliculture se d&eIopptG ;
le manque de maîb+e de l’eau. dans une proporGon élevée des zones
l
agricoles africaines. constitue probablement le kteur physique limitant
le plus habituel ;. on doit esp6rer cependant une am&ration progres-
sive de la situation dans ce domaine, amt%oration d’ailleurs requise pour
une meilleure production & riz, ind@endamment d’AzoZla.
pour autant que la culture d’Azolla soit bien pratiquée, ses principaux
l
effets en riziculture sont manifestes : augmentation sensrbledelaproductivin5
(chaque enfouissement d’AzoZZa &ant 6quivalent a l’application de 30 a
40 unités d’azot&a), et diminution du développement des adventices (et
112
donc des travaux de d6sherbage) ; il faut y ajouter l’konomie en eau
-
i
et l’amt%oration de la structure du sol ;
les tentatives d’introduction de l’azolliculture en milieu paysan sont ac-
l
a-.
:.,:. ic: :; i.’ ,~c’&$,@ diversement selon la structure des exploitations agricoles. Dans
-.; ~ ; .,,. y.:, * ! - 5i, &$ iii@ à foa.,e++raUent technologique (distribmion des intrants ,as-
, ,~“.s$e? mr+mwtion pousA...) elles semblent avoir peu de chance d’aboutir,
la +otivation des agriculteurs &ant jnwfkinte pour soutenir un effort
pr$+ig6 de .maîhise des pratiques culturales reqkes. Ailleurs par contre,
.
_ ,:,,,ief x-t%ctio~ sont souvent positives, et dans quelques ca& tles lime
-
,.
I., ,.,jl,~~~l$+, actuelle,iLest vrai-,les paysans prakpmt dt5jZt effectivement
-.
‘.
et $e leur propre inihative I’azoUiculture.
Dkcusslorr .ti &nc!usions
. . ,,a. ,
‘.::’ <’ 7 .
:.
?
&es ipz.heFhes sur Azollu ont considkablement progre& ces 20 dernieres années
en m&& temps que s’am6lioraient et se diversifîaient les pratiques ~ulturales destinées
ii ‘tirer profit de cet+ plaqte et que s’accroissaient les superficies cultivk.
Depui.4 quelques temps cependant, un courant d’opinion peu favorable à la poursuite
: des !+F!qkF: œ ,@n+ne ig dessine dans œrtains milieux, les arguments invoqué
plus ou moms .exp@ement &ant l’effondrement du prix des engrais de synthèse sur
.,._
le marche mondial, la diminution de l’emploi d’AzoIla comme engrais vert en Chine,
ainsi que le faible pourcentage de Assite dans les tentatives d’introduction de l’azolli-
culture dans d’autres pays.
Si l’on consid&re le cas particulier de l’Afrique, il faut cependant reconnaître que
ces arguments~,sont~peu convaincauts. Même au prix actuel (qu’il serait sans doute im-
pN*;%@*a.q mm?: immgble)~4J~qqpis,commexcialxesteraxans doute encore
longtemps imxcesible au petit pays& 6loignC des centres de distribution et daus Ce cas
AzoZZa peut repn%enter une aknative,,.i
‘~‘Q&tu’lar&ession
de l’&ohkuhure enkine, il faut signaler qu’elle effecte
principalement 1’utiUation d’AzoZZa en monoculture, et dans une moindre mesure en culture
assoc& avec le riz; tandis que la production pour lWmentation animale et en systkme
intkgre est au contra& en expansion. Les raisons justifiant cette r6gression partielle sont
l’am&ration de la situation konomique qui facilite l’ac&s aux engrais commerciaux,
la diminution de la main d’oeuvre disponible dans les campagnes et la nt!cessité de consacrer
la total& ducalendrier 8 la production alimemare, compte tenu de la forte densit6 de
population ; c’est ainsi que la motroc&m5hiveanale d’Azo2ki’y’êst de plus en plus souvent
_ remplack par des Vc#ures maraîchkes ou & ble d’hiver. De teks raisons ne peuvent
pas’être invoq+ dans les conditions g&k%ales actuelles de lkgriculture atricaine.
Ilestvraienfînqueles
pmgrammes de recherche t+lant Zt introduire l’azolliculture
dans plusieurs pays‘dYAsie n’ont pas souvent (mais parfois) 6tt5 couro~ de suc&.
Les causes en sont sans doute diverses~Dans cer&s cas, des conditions similaires a
celles prdvalant actuellement en Chine peuvent être invoqu&s ; dans d’autres, les condi-
tions du milieu sont effectivement défavorables au dkloppement d’AzoZla (abondance
de prklateurs, temp&ature et humidit6 excessives, alcabuite du milieu favorable à la prolife-
ration des algues...), dans d’autres, enfim c’est l’insuffisance des moyens mis en œuvre
pour la recherche...), la formation et/ou l’encadrement qui est en cause. Pour de nom-
breuses r&ions d’Afrique les deux premières raisons ne peuvent certes être invoquées,
; .,_. :
113
et les n%uhats de la recherche sont en outre dés & prksent suffisamment positifs pour
y justifm l’introductiou de l’azolliculture.
La question se pose d&s lors de savoir quels sont les moyens a mettre en œuvre
pour que le paysan africain tire effectivement parti & ces acquis. L’expkence montre
qu’un encadrement suivi est essentiel; au moins dans un premier temps. Un personnel
compkent doit donc &re forme et des conditions de travail lui permettant & se tenir
en permanence a la disposition des paysans doivent lui &re fournies. La prise en charge
de ce personnel sera assur& selon les cas par des n%eaux nationaux de vul~on,
des coop&atives, des organisations non gouvernementales. Si on se re&re aux pays où
Azollu est effdvetit valc+rist! ii i’heure actuelle; ‘force est de constater que le dkvelop
pement de l’azolliculture y a toujours &e promu de marriere tri?s directe par les pouvoirs
publics qui y ont vu un investksem em, non négligeable, sans doute, mais rentable à terme
; en cas de rkssite de l’opkation les d@enses consenties prkentent en effet le triple
avantage de ne pas requérir de devises &rangères, de n’être pas récurrentes (contrairement
aux dépenses de subsidiation des engrais) et de toucher une tranche de paysannat particu-
lièrement d&vorisee. ll faut noter enfii qu’une telle stratégie ne comporte pas de risques
élevés, compte tenu de ce qu’elle peut fort bien être testée. à titre probatoire, dans des
régions limitées. Son succks dépendra du choix des zones d’introduction ; certaines sont
en effet plus prometteuses que d’autres (maîtrise de l’eau, conditions climatiques, stmc&ms
du paysannat) ; il d@endra peut-&-e surtout de la formation du personnel’d’encadmment
et de sa disponibilité sur le terrain.
Rbfbrences
ANTOINE, TS., CARRARO, S., MJCHA, JC. and VAN HOVB, C. 1986 Comparative
,‘, :- appetency for Azolla of Cîchksoma and Or-eochrornis (Tiipia). Aqua&ure ‘53,
95-99.
’
’
BAI, K.Z., YU, S.L., CHBN, WL. and YANG, S.Y., 1978 Cultivation of Alga-free
Azolla, isolation of Anabaena azollae and preliminary attempt at their association.
In : Proc. Symp. on Plant Tiie ~Culture, Peking, 23-27 May 1978, Sciences Press
PubL Peking. pp. 455458.
BRABMER, P., 1927 La culhm des Azolla au Tonkin. Rev. ‘Bot. App. Agric. Col. 7,
816819.
_
BROTONEGORO, ST and ~ABDULKADIR, S. 1978 The decomposition of Azolfu pinnata
.e in moist and$ooded.,,it Ann.. Bogor., 6. 169-175.
: i
~BtiCKJNG~~ Kti., ELA, S.W., MORRIS, J.G. and GOLDMAN, C.R. 1977Protein
quality evaluation sud proximate analysis of AzoZZa fiZicnZoiifes Fed. Proc. 36, 1179.
BUCKJNGHAM, KW., ELA, S-W., MORRIS, J.G. and GOLDMAN, CR. 1978 Nutri-
tive vatue of the nitrogen-fixulg aquatic fern Azolla jïlîculoïdes. J. jlgnc. Food Chem.
26, 1230-1234.
CHARLIER,S.,LEGROS,A.ASJNARJDXSANMARZAN O,C.M,NAVEAU,H.,NYNS,
EJ., VAN HOVB, C. and GODARD, P. 1983 Biomethanation potentialities of Azolla.
In : 3d Intem. Symposium on Anaerobic Digestion, 14-20 Aug 1983 Boston, USA
PP- 7.
114
CHEVALIER, A. 1926 La culture d’Azollu pour la nourriture des animanx de basse-
cour et comme engrais vert pour les rizi&res. Rev. Bot. App. Agrk Col. 6.350-360.
COHEN-BAZIRE, G. and FRANCHE, C. 1985 Restriction sites in the nif H, D, K gene
region qf habaena uwllae endosymbionts from nine different Atolla qxcies.:In :
fitrogen fixation research progress. Evans, H.J., Bottomley PJ. and. NemqW.E.
Eds. M&inus Nijhof Publ. Dordrecht, p. 146.
DAO ‘Ihe Tuan and TRAN QUANG THUYET 1979 Use of AzolZa in Vietnam. pp.
395405. In : Nitrogen and rice, IRRI, Ed-, Los Banos.
DIARA, H.F. and VAN HOVE, C. 1983 ,The influence of AzolZa on rice productivity
. .
Pdmmary results obtained at WARDA Richard Toll project - Senegal. WAkDA
tecticaI NewsZetter 4, 7-8.
DIARA, HF., DD[oN, C.A. and VAN HOVE, C. 1984a Adla in West Africa : first
results IÏom the WARDA program. In : Developments in plant and soil sciences :
practical application of Azolla for rice production. Silver, W.S. and Schroder, E.C..
Eds., Martinus Nijhof Publ. Dordrecht, pp 202-207.
DIARA, H.F., DIXON, CA and VAN HOVE, C. 1984b Influence d’Azolla sur la produc-
tivité du riz à Rokupr, en Sierra Leone. Bulletin technique de PADRAO 1, 3-4.
DIA&& m., VAN BRANDT, &, DIOP, A.M. and VAN HOVE, C. 1987 Azolla and
its mie in riœ culture in West Africa. In: AzolZu utilizatï~ IRRI, Ed., I.oS.Banos
pp. 147-152.
DUNHAM, D.G. and FOWLER, K. 1987 Taxonomy and species recognition in AzoZlu
Lam. In: holIa utilisation, IRRI, Ed, Los Banos pp. 7-16.
FAO 1985a Propagation and agricultural use of Azollu in Viehxun, ZRCN. 34,257~263.
FAO l!$b Utilizat@ of Azollu in China. IRCN. 34, i65-273.
.)1
FORNI,’ C’and GRILL1 CAIOLA, M. 1988 Physiological cbaractekation of the bacteria
isohaed f.kom the aquatic fem Azollu Lam. p. 235. In : Nitrogen fixation : Himdred
years afta. Bombe, De Bruyn, Newton, Eds., Gustav Fi&er, Stuttgart.
FRANCHE, C. and COHEN-BAZIRE, G. 1985 The structural nif genes of four symbio-
tic Anubuena uzollue show a highly conserved physical arrangement Plant Sci. 39,
125-131.
GATES, JB., FISHER, R-W, and CANDLER, R.A. 1980 The occurence of coryneform
bxteria in the, leaf cavity of Azolldrch. Microbiol. 127, 163-165.
GAT@. 3.E.. FISHER, R-W.. GOGGIN, T.W. and AZROLAN, N.T.. 1981 A ukparïson
: Of @é kfkx antigedicy of the N-fïxing cyanobacterial symbionts of Ada pinnuta
A carolinianu and A. microphilla. Plant Physiol. 67, 134.
IRRI 1987 Azolla utilization. Proœedings of the Workshop on Azolkà Use, Fuzhou, Fujian,
chine. 31 Mar - 21 Apr 1985, IRRI, Ed. Los Banos, 2% pp.
JOHNSON, P. 1988 A smala farmer reaps honor and succes frorn Green Gold, Agribusiness
Weekly. 5, 4-5.
KONAR, R.N. and KAPOOR, RK 1974 Embryology of Azollu pinnutu. Phytwnorphology.
24, 228-261.
LADHA, J.K. and WATANABE, 1.1982 Antigenic similarity among Azollu uwllae separaH
from different species of Azolla. Biochem. Biophys. Res. Commun 109, 675-682.
: .. ., .:..
115
LIN, C.,LIU, C.C.,ZHENG, D-Y., TANG, LF: and WATANABE, 1.1988 Re-establishment
of symbiosis to Anabama - Ce& Azolla (in chinese). Acta Scientia sinica (B). 31,
700.708.
LIU, C.C. 1987 Reevaluation of Azolla utiliz&k In : Azolla utikation, IRRI, Ed., Los
Banos, pp- 67-76 ‘...,‘( ii;
:: *., i ,’ .;‘i:“’ ‘28 :
*
LUMPKIN, TA and PLUCKNE’IT, I&I,. 19~2-Az&& a green manure : use and manag-
ement in trop pmduction. Westviey Press, Boulder, 230 pp.
MAI KODOMI, Y., GODARD, P. and VAN JjIOk, C. 1984 F%vonoid analysis in the
genus awlla: Arck: In& Physùik Bfkx~~~2: 34-35.
MOORE, kW. 1969 %ok i && and @$&ok&&nikance.
Bot. Rev. 35, 17-34.
NEWTON, J.W. and HERMAN, AL .197?, 1solat.i~ of cyanobacteria from the aquatic
fcrn Azolla. Ar~h. Microbiol. 120, :161-165.
1
NGO-GIA-DINH ‘i973 The effect of k Azolla caver ‘on the germination of Barnyard
gras (Echinochloa crus-galli (L.) Beau~.). In ~‘Proc. 2nd Indonesian Weed Sci. COI&,
Yogyakarta, 2-5 Apr 1973. p. 201.
NGUYEN, CT. 1930 L’Azolla cultiv6e comme e&ra& vext Bull. Econ. Indochin. (Hanoi).
33, 335-350.
OES,A.l913Uebexdi&ssim&iondesfrei~s&&ffsdurchAzoZ~BotnnischesCentralbl
Il, 164-165,
OKORONKWO, N. VAN HOVE, C. and ESKEW, DL. 1989 Evaluation of Nitrogen
fixation by different stmins of adla &abaena symbiosis in the presence of a high
level of ammonium. Biot. Fert. Soik 7, 275-278.
PERKINS, S., PETERS! 94, LUMP~,~X,awi C+&VERT, H. 1985 Scanning elec@n
microscopy of perine ar&t.e&ure~ “aS a%&&&~ &l k the genus Azolla Gmark.
Scanning Electron MicrOs. 40, 1719-1734.
PETER!$ GA. 1984 AzoZZa ‘&nubaenu symbiksës: basic biology, use and prospects for
the fumure. In : Developments in p@nt and soil Sciences: practical appkation of Azolla
for ricc production. Silver, W.S. and Schr&ier, E-C., Fxls., Martinus Nijhof PubL Dor-
drecht, 1-14.
PETER$ GA. and MAYNE, B.C. 1974 The Azolla-Anabaena azollae relationship. II:
Lmalmtion of nitrogenase activaty as assayed by acetylene reduction. Plant Physiol.
53, 820-824.
ship. V : lsN, fixation, acetylene reducti&‘ kd H, production Plant Physiol. 59,
1021-1025.
QUERUBIN, LJ. ALCANTARA. P.F. and PRINCESA, AO. 1986 Chen&al composition
of three Azolla spekes (A. caroliniana, A. rnicrophilla and A. pinnata) and feeding
value of Azolla meal (A. microphilla) in broilex ration Phil. Agric. 69, 479490.
ROGER, P. and RBYNAUD, PA. 1979 Premiks données sur l’kologie d’Azolla africana
en zone sah&ienne (SCn6gal). Uecol. Plant. 14, 75-84.
SANGINGA, N. and VAN HOVB, C. 1989 Amino-acid composition of Azolla as af-
fected by strains and population density. Plant Soil. 117, 263-267.
:
‘.
,.
-.:
.:_.
.i
,..,,_
116
SILVER, W.S. and SeOIDER, E.C. 1984 Practical application of Azolla for rice pro-
duction. Pmceedhgs of an International Workshop, Mayaguez, Puerto Rico, Nov. 17-
19, 1982. 227 pp.
SMITH, JB. 1910 Azollu vs. mosquitoes. Enromol. New. 21, 437-440.
SUBUDHI. B.P.R. ad SINGH, S. 1978 Nutritive value of the water fem Azolla for chicks.
Poult. sci. 57, 318-m.
SUBUDHI, BSR. and SINGH, P.K. 1979 Effect of macronutients and pH on the growth,
nitrogen fixation and soluble sugar content of water fem Azolla pinnata. Biol. Plant.
21, 66-70.
SUBUDHI, B.P.R. and WATANABE, 1. 1979 Minimum level of phosphate in water for
growth of Azolla determined by continuous flow culture. CUIT. Sci. 48, 1065-1066.
TALLBY, S.N., TALLEY, BJ. and mS, WD. 1977 Nitrogen fixation by Azolla in
rice fields. In : Gene& engineering for nitrogen fuation, Alexander Hollaender, Bd.,
Plenurn Press, New York, pp. 259-281.
TANG, LP., ZHBN, Q. ZHEN, D.Y. and LIU, C.C. 1987 A fluorescent antibody assay
for detecting antibodies to surface antigen on Anabaena-azollae : a prehmiuary report
p. 290. In : Azolla utilization, IRRI, Bd. Los Banos.
TON THAT TRINH 1980. Intmductory notes on Azolla &ricana, an atmospheric nitrogen-
fixing fem in irrigated rice culture; importance and prospects of its use in OMVS
(organisation pour la mise en valeur du fleuve Sénegal) irrigated amas. IRCN 29,23-
27.
TRAN QUANG THUEY and DAO THE TUAN. 1973 Azolla : a green compost. Vief.
StKd. 38, 119-127.
VAN HOVE, C. 1987 Potentiel de valorisation d’Azolla en agriculture afrkxiue. Cahors
dkthologie appliquée. 7, 31-42.
VAN HOVE, C. de WAHA BAILLONVILLE, T. DIABA, H-F. GODARD, P.. MAI KO-
DOMI, Y. and SANGINGA, N. 1987 Azolla collection and selection pp. 77-87.
In I Azolla utilimtion. lRRL, Bd., Los Banos.
VAN HOVE, C. and DIARA, HF. 1987a Azolla introduction in afiican agriculture :
progress and pmblems. IRCN 36, l-4.
VAN HOVE, C. and-DI@A, B.F. 1987b Bxplanatory gui& of a series of slide pictures
illustrating the methods of Azolla utilization in agriculture. 9 pp. 40 slides, C. Van
Hove, Ed Lonvain-La-Neuve, Belgium.
-VAN HOVB, C DIARA, H.F. and GODARD, P. 1983 Azolla en Afrique de l’Ouest
- in West Africa, 56 pp. 11 illmons. Van Hove, C. Ed. Louvain-La-Neuve, Belgium.
WALLACE, W.H. and GATES, J.E. 1986 Identification of Eubacteria isolated fiom leaf
cavïties of four species of the N-Mg Azolla fem as Arthrobacter Conn and Dinmick.
Appl. Environ. Microbiol. 52, 425-429.
WATANABE, 1. 1987 Summary report of the Azolla program of the intematioual net-
work on soil fertility and fertilizer evaluation for rice. pp. 197-205. In : AzoZZa utili-
zation, IRRI, Bd. Los Banos.
WBI, W.X., JlN, G.Y. and ZHUAN, N. 1986. Primat-y report of Azollu hy&idi&on
study (in chinese). Bull. Fujian Ad. Asrie. Sci. 1, 73-79.
;
a>.‘,,.
:
:->
.,,..;.
:
.a :
117
ZIMMERMANN, W. LUMPKIN, T. and WATANABE, 1. 1988 Biochemical taxonomy
of Azolla Anabaena, section Azollu. p. 240. In : Nitrogen fixation: Hundred years after.
Bothe, De Bruyn, Newton, Eds., Gustav Fisher, Stuttgart.
:
_i_’
.:’
.
.
.
.
.
.,.
h4mimi.uz la FBA pour la Production Agricole et ForestiSre en Afrique
Phosphorous needs and accumulation
potential of various AzoZZa species and strains
DESMADRYL, D., GODARD, P.,
WAUTHELET, IMa and VAN HOVE, C.
Laboratory sf Phmt Physiology, Catholic Universiry of Lmvuin.
B-1348 Louvain-la-Neuve,
Belgium.
It is know-n that Apda are able to accumulate phosphorus when grown on
P-rich media, and it has been suggested to take advantage of this luxury consump-
tion, loading up Awlla with P in the nurseries before mating them into the rice
field, ww they could continue to multiply using their excess P. The purpose of
the present styly was to identify AwZla ecotypes with high accumulation power
,ad/or eveutually low P needs, comparing 20 high m tolerant strains. Pour
A. caroliniana (ACA). 4 A. microphylla (Ah@, 6 A. pimata var. imbrica& (API)
and 6 A. pinnafa var. pinnafa (APP) have been grown first on N-free Hoagland
solutions (dihlted 2J5) containing respectively 15.30.45 and 60 ppm P for 18 days.
None. or only dight effects of the treatments are observed on productivity, but Awl[a
P content increases line&y with external P concenhation. strongly in ACA and AhJI,
only slightly in API and APP, and varies widely from one strain to another, even
from the same species. When such variously P-loaded Azolla are transferred
(60 g/n?) on a P-free medium, the biomass they math at the plateau phase is stmn-
gly dependent on the pre%matrnent for ACA and AhJI, not for API and APP ; at
this stage Awl& P content varies fi-om 0.023 to 0.106% (DM.). but is not correlated
withfmalbiomassnorwithspecies.
i
*
.:..,
:
,
I
.-..
119
Introduction
Phosphorous is frequently the limiting factor of AzoZZu productivity (Tung and
Watanabe, 1983 ; Lumpkin, 1987) ; on the other hand Azollu is able to make some luxu-
rious P consumption (Lumpkin, 1987 ; Reddy and Debusk, 1984 ; Subudhi and Singh,
1979 ; Sub&i and Watanabe, 1981), and it has been suggested to make advantage from
this property, enriching AzoZZu in the nurseries before field inoculation (Jumpkin, 1987).
Few data are nevertheless available on the growth potential of P-preloaded Azollu
when cultivated on P deficient media
The purpose of the present research is to identify AzoZZu ecotypes with high ac-
cumulation power and/or low P needs ; the investigation is restricted to a group of high
temperature tolerant strains.
Material and Methods
Twenty Azollu straius from the ADUL collection (Van Hove el al. 1987). previ-
ously selected among 150 for their tolerance to high temperatures (unpublished results)
are tested in this experiment : four A. cmolitiunu (ACA) : ADUL 8, 61, 109, 113 ;
Four A. microphyZZu (AMI) : ADUL 65, $6, 104,175 ; six A. pinnatu var. pinnufu (APP) :
ADUL 98, 99, 111, 129, 136, 144 and six A; pikufu var. imbricufu (API) : ADUL 6,
26, 87, 92, 100, 137.
Experiment 1 : four culture baths containing 40 plastic boxes with the bottom
replaced by a mosquito net (Van Hove et al. 1987) are filled with 120 1 of diluted
(4/10) Hoagland solution in which nitrates are replaced by chlorides and phosphorus.is
added as KH.$O, at respective concentrations of 15, 30,45 and 60 ppm P (initial pH :
6.00, 5.29, 5.19,4.89). Each box is inoculakd with lg (F.W.) AzoZZu, every strain being
duplicated, at random, in the four culture baths. Culture media are replaced after 9 days.
The experiment is realized in a greenhouse in which conditions are : light intensity :
natural + 65 pE.m-2S.-1; air to : 35 f 4°C ; water to : 33 f 4°C ; relative humidity :
47 Ik 5%.
On day 19, AzolZu samples are harvested and their fresh weight measured after
1 mn centrifugation (16 g} of the AzoZZu containing boxes, followed by 20 s drainage
on absorbent paper ;- 1 g of each sample is set apart as inoculum for the second experi-
ment ; the remaining biomass is used for dry weight measnrement and phosphorous ana-
lysis.
r.F.1 : \\ :
..’
j
Experiment 2 : The device is similar to the pmceding one except that the culture
medium (initial pH : 6.05) is P-free, and is renewed every 7 days ; air t!’ : 38 f 5°C ;
water to 37 zk 4°C ; relative humidity : 51 f 5%. Each box is inoculated with 1 g of
variously P-enriched AzoZZu from the first experiment (60 gm-9.
Fresh weight is measnred as above, every 2, 3 or 7 days according to the growth
stage until the plateau phase is reached (two successive not mcreasing biomass values).
Azollu P content is measured at the end of the experiment.
Chemical analysls : Total phosphorous is determined calorimetrically as its re-
duced phosphomolybdate complex, after miueralization in diluted nitric acid.
;.
I
120
Results
Experiment 1 : Effect of P concentration in the medium on Azollu growth and
P cQntent.
Probucfiv~fy - Under the prevailing experime&l conditions A. carolinium
and A. microphylla are the most productive species, followed by the A. pinnata varieties
(Table 1). Increasing P concentration Erom 15 to 60 ppm has no or only slight effects
on productivity.
Table 1 : Mean productivity and standard deviation (gFWm2.day) of 4 Azolla
species or varieties as a function of the external P concentration
(Ppm PI.
Species
15 ppmP
30 ppmP
45 PpmP
60 ppmP
ACA
122.2 2 4.8a
120.8.: 3.7a
118.8 f 5.0a
116.8 + 5.3a
I.,
AM1
' 116;g +, G;.;;'
118.0 2 7.6a
11955 f 4.4a
111.0 _+ 9.0a'
API
75.9 i 4.lb
'70.1 _+ 5.5b
72.7 _+ 8.8b
70.4 _+ 7.6b
APP
62.7 +14.9c
63.9 f 3.0b -
68.7 +10.3b
66.3 + 9.5b
In a column, values with the same letter are not significantly different at
P = 0.05 (Duncan test).
Phosphoms confent - Internal P concentration at the end of the experiment
(Fig. 1) varies considerably Corn one strain to another in a given species or variety. At
low (15 ppm) external P concentration, there is nevertheless no specific or varietal diffe-
rence (Table 2) but P content increases linearly with increasing external concentration,
slightly in the two A. pinnata varieties. for which angular coefficients of the regression
line (internal P content (%)/medium P content (ppm)) vary fmm 0.003 to 0.009, strongly
in A. caroliniana and A. microphylla (angular coefticients : 0.01 to 0.02).
.
.
.
:,‘-:
121
Figure 1 : Three typical behavior of 20 strains of Azolla species or varieties growing on a P deficient medium after a pretreatment on 4 P concentrations : ACA
& AMI (a), best APP (b) and API or other APP (c).
Growth curves as biomass in kgFM/m2 related to time in day after pretreatment at 15 (I) ), 30 ( + ), 45 ( s ) and 60 ppm P (a ).
Maximum biomass in kgFM/m’ ( 0 ) and final P content in %DM (% ) related to P concentration in ppm P of the pretreatment medium.
122
Table 2 : Final P content and standard dev!ation (XDM) of 4 Azolla species
or varieties as a function of the external P concentration
(ppm P).
Species
1 5 ppm P
30 ppm P
45 Ppm P
60 ppm P
ACA
0.94 + 0.14a
1.05 + 0.07b
1.32 + 0.12a
1.53 + 0.16a
AM1
1.00 + 0.19a
1.25 -r 0.17a
1.39 + 0.18a
1.59 + 0.15a
API
0.93 + 0.16a
1.01 + 0.18b
1.11 + 0.20b
1.20 -I
16b
APP
0.88 2 0.16a
0.96 f 0.176
0.99 f 0.13b
1.11 f O.llb
In a column, values with the same letter are not significantly different
at p = 0.05 (Duncan test)
Experiment 2 : AzoZla growth in P-free media as influenced by initial P content.
Maximum biomass - Maximum biomass for a given pretreatment (Table
3) is highly strain dependent in A. pinnutu var. pinnata (high standard deviation), not
in A. pinnata var. imbricata which always shows low plateau values, or in the two other
species. As expected from results obtained in the first experiment, A. carolinianu and
A. microphyllu final biomasses are highly influenced by the pretreatment, whereas A. pinnatu
are practically not. Fig. la and lb illustrate these two typical growth patterns. Also for
A. caroliniana and A. microphylla there is a good correlation, at the species level, between
initial P content and maximum biomass. For A. pinnatu on the other hand such a correlation
only exists at the strain level as illustrated in Fig. lb and c.
Final Azolla P content - Azolla P content at the plateau phase (Table
4) is generally higher in A. microphylla and A. caroliniana than in A. pinmta, which
presents very high intraspecific variations. It must be mentioned that the observed values
are higher than the calculated values (initial P content. initial biomass. final biomass’) ;
this has to be attributed to traces of P brought in the system by demineralized water
compensating evapotranspiration.
123
Table 4 : Final internal P content and standard deviation (% DM)
of 4 Azolla species or varieties cultured in a P deficient
medium after a pretreatment at different P concentration
(Ppm P)
Species
Pf(%DM)
ACA
0.076 + 0.009 b
AMI
0.087 + 0.013 a
API
0.056 + 0.022 c
APP
0.041 + 0.018 d
In a column, values with the same letter are not significantly
different at p = 0.05 (Duncan test).
Table 3 : Maximum biomass yield and standard deviation (kgFM/mZ) of 4 Azolla
species or varieties cultured in a P deficient medium after a
pretreatment at different P concentration (ppm P).
Species
15 ppm P
30 ppm P
45 ppm P
60 ppm P
ACA
2.4 + 0.4ab
2.7 + 0.3a
3.4 + 0.3a
4.1 + 0.4a
AMI
2.0 + 0.3bc
2.4 + 0.3a
2.9 k 0.4b
3.6 + 0.5b
API
1.6 + 0.3~
1.8 + 0.2b
2.1 + 0.3~
2.0 + 0.3d
APP
2.5 5 0.8a
2.6 2 0.6a
2.7 + 0.6b
2.9 _+ 0.6~
In each column, values with the same letter are not significantly different
at p = 0.05 (Duncan test)
124
Dlscusslon
Subudhi and Watanabe (1981) have shown that A. pinnuta growth does not improve
when external P concentration increases from 5 to 10 ppm, whereas Singh ef al (1984)
obtained growth increases up to 15 ppm P (the highest concentration tested) forA.fiZicuZoiaks
and A. mexicana, to 10 ppm for me A. pinnata strain, and to 5 ppm for another. Cary
and Weerts (1982,1984) observed better A. jXcu2oide.s growth at 20 ppm than at 5 ppm ;
for A. pinnata optimal growth was reached in the 5 to 20 ppm range, with a significant
decrease at 40 ppm.
In the present study, the P concentration range was extended to 60 ppm, which
is very high according to agricultural practice standards, with the view of testing phos-
phorus accumulation potential of various AzoZla species and strains. Results are generally
in agreement with previous information, except for the fact that no depressive effects
of high P concentration were generally noticed
Subudhi and Watanabe (1981) also found that P content in A. pinnata increases
linearly with increasing external P concentration between 0 and 10 ppm. We observed
the same behaviour for the 20 strains tested between 15 and 60 ppm P ; the fact that
productivity is not affected in this range of concentration confirms the luxury consumption
capacity of Azollu (Lumpkin and Pluchnett, 1982). Significant interspecific differences
nevertheless exist, A. caroliniana and A. microphylla having a higher accumulation power
than A. pinnutu ; considering that they are also characterized by a good correlation bet-
ween P content and maximal biomass, over fertilization of these two species at the nur-
sery level seems justified. If P fertihzer availability is restricted on the other hand, some
A. pinnata var. pinnata should be preferred.
Looking at the Azolla P content at the plateau phase in P deficient culture media,
it appears once more that A. caroliniana and A. microphylla differ from A. pinnata, with
a higher P content and a lower intraspecific variability. The extreme values are 0.023%
and 0.106% ; assuming a 1.2% P content in the inocuhtm, which is easily obtained with
a good P fertilization in the nursery, this means biomass multiplication potentials, in a
P-free medium, of respectively 52 and 11. Azolla strains with low minimal P content
L::y..c ‘l7t.p
. r
r
,.
mmpq:-1.. :,T b,z ;a,nbiJ.l’- ’ ;..
L . . . . . . . . ..I. a.,
_ cu ALA ,&c;iol~ j;iUt;itiII.
I\\i~l’i; iikuiuduu* MUUIU IK,VU
theless be necessary concerning N content of Azolla in these extreme conditions since
it is known (Watanabe et al., 1980) that P and N contents are correlated.
Furthermore, the capacity of a given strain to reach a high biomass in P deficient
culture media is not attributable to lower P needs, final biomass and final P content being
poorly correlated (r = 0.35); this means that strains having similar initial and final P
contents can differ by their fmal biomass, due probably to differential capacities of their
root system to extract traces of phosphorus present in the culture medium, as suggested
by Subudhi and Watanabe (1981).
Conclusion
The aim of the present research was to estimate the range of phosphorus needs
and phosphorus accumulation power of various Azollu species and strains, with the view
of introducing, eventually, such criteria in Azolla selection program.
125
Based on a limited number of strains, the conclusion is that the two criteria have
to be considered, somme strains from the A. pinnutu group being characterized by their
low treshold P level, some others, from the A. caroliniana and A. microphylla group
having a high phosphorus accumulation power.
References
CARY, P.R. and WEERTS, P.G.J. 1982 Nutritional and water temperature factors affec-
ting growth of aquatic plants. Rex Rep., CSIRO Div. Irrig. Res. 1981-1982, 37-40.
CARRY, P.R. and WEERTS P.G.J. 1984 Factors affecting growth of Azollu spp. Res.
Rep., CSIRO Div. Irrig. Res. 1983-1984, 43-45.
LUMPKIN, T.A. and PLUCKNEIT, D.L. 1982 Azolla as a green manure : use and ma-
nagment in crop production. West View Press, Boulder, Colorado. 230 pp.
LUMPKIN, T.A. 1987 Environmental requirements for successful Azolla growth, p.
89-97. In : Azollu Utilization. IntemationaI Rice Research Institute, Manila.
REDDY, KR. and DEBUSK, W.F. 1984 Phosphorus removal by Azollu curoliniana
cuhured in nutrient enriched waters. p. 151-162. In : Practical application of Azollu
for rice production. Silver, W.S. and Schroder, E.C. (eds). Martinus Nijhom W.
Junk, Dordrecht.
SINGH, P.K., PATRA, R.N. and NAYAK, S.K. 1984 Sporocarp germination, cytology
and mineral nutrition of AzolIa, p.5572. In : Practical application of Azolla for rice
production. Silver W.S. and SchrUder E.C. teds). Martinus Nijhofm W. Junk, Dor-
drecht.
SUBUDHI, B.P.R. and SINGH, P.K. 1979 Effect of phosphrus and nitrogen on growth,
chlorophyll, amino nitrogen, soluble sugar contents and algal heterocysts of water fern
Azolla pinnata. Biol. Plantarum 21, 401-406.
SUBUDHI, B.P.R. and WATANABE, I. 1981 Differential phosphorus requirements of
Azolla species and strains in phosphorus-limited continuous culture. Soil Sci. Plant
Nutr. 27, 2??-347.
TUNG, H.F. and WATANABE, I. 1983 Differential response of Azollu -Anubaena as-
sociations to high temperature and minus phosphorus treatment. New Phytol. 93,
423-431.
VAN HOVE, C., DE WAHA BAILLONVILLE, T., DIARA, HF., GODARD, P. MAI
KODOMI, Y. and SANGINGA, N. 1987 Azollu collection and selection, p.77-87.
In : Azolla utilization. International Rice Research Institute, Manila.
WATANABE, I., BERJA, N.S. and DEL ROSARIO, D. 1980 Growth of Azolla in paddy
field as affected by phosphorus fertilizer. Soil Sci. Plant Nub-. 26, 301-307.
Maximisez la FBA pour la Production Agricok et Forectit?re en Afrique
Sensitivity to Aluminium of various AzoZZa species
and strains
WAUTHELET, M., GODARD, P.,
DESMADRXL, D. and VAN HOVE, C.
L&oratory of Plant Physiology, Catholic University of Louvain,
B-1348 L.ouvain-la-Neuve,
Belgium.
Abstract
High alumivillm rnnwntTrt:nn iq 2 +r?j” Fnhl-- ir: mz~)r ~ci~.Ic tr-;i;::! 2 -i!:,
where it interferes among others with phosphorus availability, an element especially
important for Azolla. Information on Azolla sensitivity to Al is nevertheless mis-
sing, and it was the purpose of this preliminary research to estimate the range of
Al concentrations allowing proper Azollo growth and to select, eventually, Al tole-
rant straim. Twenty high temperature tolerant strains, four A. caroIiniona (AC&.
” four A. microphylla (Ah@, six A. pinnata var. imbrkata (API) and six A. pinnata
var. pinnuta (APP), were grown on N-free 2/5 Hoagland solutions with respectively
0.2.5 and 8 ppm Al, and their biomass (fresh matter) was measured after 11 days.
Results show that, according to the strain productivities decrease with increasing
Al concentration either +_ linearly or exponentially. At the species level API is the
most sensitive, followed by ACA and AMI and then by APP. Intraspecific diffe-
rences arenevertheless high, except for API strains. According to the strain, productivities
at 2 ppm vary from 100 to 50% and at 8 ppm from 73 to 16% as compared to
the control.
127
Introduction
Aluminium toxicity is an important growth limiting factor for plants in many aci-
dic soils, in which Al is easily solubilized. Al toxicity can induce mineral nutrition distur-
bances, such as P (Randall and Vosc, 1963) and Ca (Awad et al., 1976), uptake decrease ;
it also inhibits cell division (Horst et al., 1983).
As far as the aquatic fern Azolla and its endosymbiont Anabaena are concerned,
only scarce information is available (Lumpkin and Plucknett, 1982). The purpose of the
present research is to estimate the range of Al concentrations allowing satisfactory Azolla
growth, and to select eventually Al tolerant ecotypes (among a group of high tempe-
rature tolerant strains).
Material and Methods
AZ&Z strains as well as the experimental device are identical to those described
by Desmadryl et al., (page 118 in this volume). Culture media contain 0, 2, 5 and
8 ppm Al as Al, (S0,),.18H,O), their respective pH being 4.80 f 0.41, 4.14 + 0.11,
3.72 f 0.04 and 3.59 f 0.06 during all the experiment. Each box is inoculated with
5 g (FW) Azda (& 300 g/m”) and each strain is duplicated. The conditions in the green-
house are : light intensity : natural + 65 pE.m-2.s~1; photoperiod 12/12 h; air to :
31 * 2T; water to : 30 rt 2°C; relative humidity : 55 + 5%.
After 11 days AzoZZa samples are harvested and their fresh weight measured after
1 mn centrifugation (16 g) and 20 s drainage on absorbent paper.
Results
In the conditions of the experiment, the productivity of the controls (0 ppm Al)
varies strongly between strains, but also between species and varieties (Pig. 1) A. caroliniana
and A. microphylla are very similar and the most productive, followed by A. pinnata
var. pinnatu and finally by A. pinnata var. imbricata. Productivity of all the strains de-
creases with increasing Alummium concentration, eilher exponentially, for the most sensi-
tive strains, or more or less linearly for the others. Concentrations of 2 and 5 ppm Al
only slightly affect growth of tolerant strains, while for the sensitive ones, growth is
already strongly depressed. At 8 ppm Al only one strain (ADUL 136 PP) grows very
poorly. A. pinnutu var. pinnata is the most tolerant variety (Table l), A. pinnuta var.
imbricatu the most sensitive, the two other species being intermediate. Intraspecific and
intravarietal differences (except for A. pinnatu var imbricata) are nevertheless high.
Discussion and Conclusion
Sensitivity to Al varies widely among plants. Gossypium hirsutum root growth
for example stops above 0.5 ppm Al (Rios and Pearson, 1964), other plants are affected
at 1.5 ppm Al (Pay and Brown, 1964) whereas Vigna unguiculuta sustains 2 - 4 ppm
(Horst et al., 1983), Sorghum bicolor 8 ppm (Purlani and Clark, 1981). Pennisetum typhoides,
16 ppm (Long et al., 1973) and Cumelia Sinensis up to 27 ppm at pH 6 (Matsumoto
128
et al., 1976). Or-pa sativa which is of special concern here, is not depressed at 10 ppm
at the adult stage (Howeler and Cadavid, 1976), whereas seedlings are quite sensitive
(Thawomwong and Van Die&, 1974).
Compared to these plants, Azolla in general thus seems moderately tolerant to Al.
It must be mentioned that growth inhibition observed in this experiment at high Al con-
centrations could be partly attributed to pH effects ; nevertheless it is well known that
Azolla is able to grow properly in a very wide range of pH (Nickell, 1961 ; Lumpkin
and Plucknett, 1982 ; Van Hove et al., 1983).
In view of the wide Azolla species and strains Aluminium sensitivity differences,
selection based on this characteristic has to be considered for Azolfa use on Al rich soils.
Table 1 : Productivity as compared to the control and standard deviation
(%I of Azolla species or varieties as influenced by the external
Al concentration (ppm Al).
Species
2 ppm A l
5 ppm A l
8 ppm A l
APP
90.51 + 11.81a
63.46 !' 16.60a
46.52 & 15.27a
ACA
80.42 + 13.52ab
45.07 i 16.16b
33.76 _f 14.19b
AM1
76.20 + 13.50b
3:.33 + ;J.Sib
23.01 :
/.6/bc
API
61.74 + 8.70~
33.06.+ 7.77b 21.70 _+ 5.09~
In a column, values whith the same lettre are not significantly
different at P = 0.05 (Duncan test)
129
\\
1OC
\\
\\
a
b
p IC
r
t
u
c a0
:v
1
t 40
J
a0
0
100
d
P 80
I-
:
u
c oo
t
I
I
i 40
t
J
a0
0’
I
t
t
0
a
4
0
8 0
a
4
a
I
PPm Al
PP= Al
129
Figure 1 : Productivity (g/m’. day) of 20 Azolla strains as influenced by the external Al concen-
tration @pm Al). Strains are characterised by their entry number in the ADUL colle&on
(a) ACA
:
6851
6 1
109
113.
(b) Ah4I
:
66:
104:
175.
(c) +PI
:
6,
26,
87.
92,
100,
137.
(d) APP
:
98,
99.
111.
129,
136.
144.
130
References
AWAD, AS., EDWARDS, D.G. and MILHAM, P.J. 1975 Effect of pH and phosphate
on soluble soil alumimium and on growth and composition of Kikuyu grass. Plant
Soil 45, 531-542.
FOY, C.D. and BROWN, J.C. 1964 Toxic factors in acid soils : II. Differential alumi-
nium tolerance of plant species. Soil Sci. Sot. Am. Proc. 28, 27-32.
FURLANI, P.R. and CLARK, R.B. 1981 Screening sorghum for aluminnium tolerance
in nutrient solutions. Agron. J. 73, 587-594.
HORST, W.J., WAGNER, A. and MARSCHNER, H. 1983 Effect of aluminium on root
growth, cell-division rate and mineral element contents in roots of Vigna unguiculatu
Genotypes. Pflanzenphysiol. 109, 95-103.
HOWELER, R.H. and CADAVID, L.F. 1976 Screening of rice cultivars for tolerance
to Al-toxicity in nutrient solutions as compared with a field screening method. Agron.
J. 68, 551-555.
LONG, F.L., LANGDALE, G.W. and MYHRE, D.L. 1973 Response of an Al-tolerant
and Al-sensitive genotype to lime, P, and K on three atlantic coast flatwoods soils.
Agron. J. 65, 30-34.
LUMPKIN, T.A. and PLUCKNE’IT, D.L. 1982 AzoIZa as a green manure : Use and ma-
nagement in crop production, Westview Press, Boulder, 230 pp.
MATSUMOTO, H., HIRASAWA, E., MORIMURA, S. and TAKAHASHI, E. 1976 Lo-
calization of aluminium in tea leaves. Plant and Cell Physiol. 17, 627-631.
NICKELL, L.G. 1961 Physiological studies with AzolIa under aseptic conditions. II. Nu-
tritional studies and the effects of the chemicals on growth. Phyton. 17, 49-54.
RANDALL, P.J. and VOSE, P.B. 1963 Effect of aluminum on uptake and translocation
of phosphorus 32 by perennial Ryegrass. Plant Physiol. 38, 403-409.
RIOS, M.A. and PEARSON, R.W. 1964 The effect of some chemical environmental
factors on cotton root behavior. Soil Sci. Sot. Am. Proc. 28, 232-235.
THAWORNWONG, N. and VAN DIEST, A. 1974 Influences of high acidity and alumi-
nium on the growth of lowland rice. Plant Soil 41, 141-159.
VAN HOVE, C., DIARA, H.F. and GODARD, P. 1983 Azolla en Afrique de 1’Ouest
- in West Africa. 56 pp. 11 illustrations. C. Van Hove, Ed., Louvain-la-Neuve, Belgium.
INOCULATION :
PRODUCTION D’INOCULUM
ET ESSAIS D’INOCULATION
.
. . (
I
. “ . ,
.
.
.
. : . .
Umimisa la FBA pour la Production Agriwle et Foresti& en Afiiqu~
.I.
:
Prbd&tioti x&i uses of rhizobium inoculants
BORDELEAU, L.
Research Sation, A@ult~ Canada,
2.560 Hochelaga Bhd. Sbinte-Fey,
Qdbec. Canada GIV 213.
Abstract
‘,
‘.
The inoculation of legumes with Rhizobiwn inoadants is one of the few cases
of man’s successful exploitation of microorganisms fqr avp action. This bio-
technology is outstanding in its wide applicability in diffmt agricultural envimn-
merits. We outline here background information on the various aspects of legume
inoculant technology, namely : strain selection and characcetistics, preparation and
properties of carriers, growth and survival of rhizobirg quality control program from
strain selection tbmugh preparation and distribution of final products. Criteria used
in the selection of Rhizobium strains for inoculation include prompt effective nodu-
lath over a wide range of field conditions and root tempaatures, ability to grow
wellinculture.togrowandiveincarrierandto‘~~onthe~Other
characters are competi-tiveness in nodule formation and survival and multiplication
inthefoilpH~l~pesticidetoleranceand~inthepesenceofcom-
bined nitrogen The sxxxss of legume inoculation~depauis on both the quality of
inoculantp-eparationQldthe~~~usedto~~~rhizobiaintosailatsowing.
lnoculant carriers must be selected accord+ to availability of material and t&no-
log&. and must permit the introduction of the highest possible number of infective
cells in the soil where the host plant will first accept an infection Finely ground
peat is the rhizobial carrier most extensively used. New technologies such as polymer-
entrapped microorganisms, and concentrated freezedried cells powder inoculants are
arriving on the market A quality control program is an essential part of inoculation
technology.
133
Introduction
Food and fibers production is often limited by low soil fertility and nitrogen
appears to be one of the most limiting factors. Industrially manufactured nitrogenous ferti-
lizeds are not afordable for most of developing countries that want to intensify their agri-
cultural production. Therefore, the alternative of using some microorganisms to convert
atmospheric nitrogen to a form available for the nutrition of higher plants was extensively
exploited
The nitrogen-fixing symbiosis between leguminous hosts and their rhizobia is an
outstanding biotechnology because of its wide applicability in diverse environments and
agricultural systmns. The inoculation of legumes with Rhiwbium inoculants is one of
the few cases of man’s successful exploitation of microorganisms for agricultural pur-
poses. ibis biological system has attracted considerable attention of scientists having
diverse backgrounds because of the complexity of the phenomenon which demands the
application of many skills for its understanding and more successful utilization. However,
there is a real risk when the enthusiastic non-microbiologist moves in to invest&ate a
system in which a bacterium, which may be more difficult to control, plays its own major
role in an intimate association with the more tractable host. Whether he is plant physiologist,
biochemist, geneticist, soil scientist, or even a microbiologist unfamiliar with rhizobium,
a worker coming into this field cannot afford to take this microorganism too lightly. Diffi-
culties can be encountered by the relatively slow development of even the faster-growing
rhiiobia, leaving them open to overgrowth by a contaminant, by the gummy nature of
its growth, increasing the likelihood of an associated organism remaining undetected even
with single colony isolation. Furthermore, the difficulty one has to definitively distinguish
rhizobia from some other genera, apart from demonstration of nodulating capacity, adds
up to the confusion due to contamination or wrong identification and may lead to a whole
body of developmental work which is bias to the practice of legume inoculation. In such
a case, either the newcomer takes the trouble to fanGnize himself sufficiently with the
nature of the microorganism he is dealing with, and the technique needed for its purifi-
cation and maintenance, or he works cooperatively with someone who has this knowledge
and technical capacity.
In this presentation, we will outline background information on the various aspects
of legume inoculant technology, including factors affecting quality and technique for
quality control
Selection of rhkobk for inoculants
Many of the failures to establish rhizobia in soil successfully may be attriiuted
either to insufficient or unsuitable rhizobia in the inoculant. As agriculture extends further
and further into regions which are marginal because of low and/or erratic rainfall and
poor fertility, rhizobia will be called upon to perform with a wider range of host plants
and environmental and edaphic conditions. The selection of suitable strains of rhiiobia
is the first step in the produciton of legumes inoculants. At one time, nitrogen fvtaton
efficiency was the only criterion seriously considered for a good inoculant strain, but
now there are many that should be considered as well : (i) competitive ability with other
strains, including indigenous population, for prompt effective nodulation of the host legumes ;
(ii) nitrogen fixing ability over a wide range of environmental conditions ; (iii) nodule-
134
forming and nitrogen-fixing abilities in the presence of soil nitrogen ; (iv) ability to multiply
in broth and to survive in carriers ; (v) ability to survive on the seed ; (vi) ability to
survive in adverse physical and physiological conditions such as desiccation, heat, free-
zing, low soil fertily ; (vii) tolerance to pH changes, pesticides ; (viii) genetic stability
during-w~growth;
In the context of inoculant production, strain competitiveness and specifkity refer
to effectiveness in forming rapidly an effective association with a particular host (Antoun
et al., 1979). The host legume exerts a strong influence on the effectiveness of the sym-
biotic association in fixing N,, and strain specificity exists for particular hosts (Gueye
and 3ordeleau, 1988). while intemctions between the symbiotic association and the environ-
ment have been also described @rockwell, 1980 ; McLoughIin et al., 1984). l&se create
a dilemma in deciding wether to produce numerous inoculants with highly effective
strainsforindividualhostorU,usea~~~strainsthatvaryfromgoodto
excellent with a range of legumes. The use of multiple strain iuoculants should be avoided
due to possible antagonistic and competitive effects witbin the culture and the likelihood
of competition in nodule formation from less effective strains in the inoculant (Bordeleau
and Antoun, 1977). This problem poses a delicate situation for microbiologists who must
decide between practical aspects of inoculant production and marketing, and scientific
excellence in providing the best available mater2 for inoculant production.
As a rule, foreach legume that we want to improve yields through nitrogen fixa-
tion, it is preferable to select strains of rhixobia directly from the cultivar we intend to
use and then check the efficiency. The main criterion used in selection is the amount
of N, fixed, either assayed directly, or indirectly by determining plant dry weight @or-
deleau et al., 1977). The latter method integrates all the factors that influence the symbiotic
system. With grain legumes, pediodical m easurements of biomass and nitrogen content
of plant parts during the development and at maturity will pezmit to calculate t&harvest
index of a strain, since it is hown that plant inoculation with Rhiwbim can modify
the nitrogen distribution in the plant. Tests in controlled environment provide reliable
comparative information on the ability of a number of strains to fix N,, but they do
not evaluate ability to colonize the rhizosphere, to compete for nodule sites, or to persist
in the soil. In the other hand, field ~CS~.S cm CV~USZ only a Llii-ALcd liu~~~k~ of strains
because of demands on time, labor and equipment. Both systems, therefore, are commonly
used in sequence, the former to select a set of strains with above average N,-fixing ca-
pacity and the latter for an overall evaluation of the strains’ ability to fm N2 under a
variety of field situations. Specific separate tests can be conducted if a particular cha-
racter is bemg evaluated. However, field trials provide the ultimate test of a strain’s perfor-
mance, since N2 fixation and the proportion of nodules formed by the inoculum are the
end result to the interactions of all the factors involved.
Large-scale production of rhizobla
The first step in the production of legume inoculant is massive production of a
selected Rhiwbium species in liquid medium. Rhizobia are heterotrophic and generally
not highly demanding in their nutrient requirements.
Different media have been proposed, they are composed of a carbon source, a
source of nitrogen and growth factors, and different minerals such as potassium phos-
phate, magnesium sulfate, and sodium chloride. Mannitol and sucrose are the conventional
135
carbon sources used in routine media, although growth of slow growers on mtitol is
variable. The most satisfactory carbohydrate is glycerol, on which slow growers have
shorter generation tunes than on glucose, mannitol, galactose, or sucrose (Stowens, 1985).
Yeast extract can be used as the sole carbon and nitrogen source ; however, high concen-
trations cause cell d&or&m and even growth inhibition of certain Rhizobium; which can
be co~tmcted by calcium. The economy of largescale production is largely governed
by the price and availability of a suitable carbon source. Corn-steep liquor, proteolysed
pea husks, malt sprouts, whey and industrial-grade yeast extract have been pmposed as
media for Rhiwbium production at the industrial level (Meade et al., 1985 ; B&OM&&
et al., 1986). All of these materials ate industrial by-products which contain growth factors,
nitrogen and carbon. Efficient uulization of a carbon source often depends on aeration,
method of ster%zation and equipment available. For the production of inoculants, a high-
count broth is required to ensm a highquality inoculant ; increased numbers of viable
cells may be achieved by varying the source and concentration of materials acconhng
to the individual strain requirements. A- short fermentation time for the broth culture is
desirable in order to reduce contamination risks. A large inocuhun, providing 106-10’
rhizobia/ml culture medium at the beginning of fermentation, reduces the time required
to reach maximum viable numbers, and, therefore, reduces the chance and effects of
contamination. Fast-growing strains reach maximum viable number iu 30-40 hours, com-
pared with 80-120 hours for slow-growing strains. Temperature of incubation for rhii
bia is generally kept at 2628”C, although few strains can be produced at 32-35°C
without affecting their nod&ion ability.
Facilities for multiplication of rhizobia in liquid culture vary from elaborate
large-capacity (10002000 liters) industrial fermentors. to simple flasks or drums
(10-100 liters) (Date and Roughley, 1977). Good fermenters can be fabricated from
steel drums fitted with brass inlet, outlet, inoculation and sampling ports Tbe air inlet
extends to the bottom to provide stirring of the broth. Proper filtration systems st&lize
the air from an external supply. The vessel with filters and outlet tubes attached and
two-thirds filled with medium is sterilized by autoclaving. Maximum infective rhizobia
after incubation should be in the order of 109-KY0 cells/ml.
Carriers of rhizobia and their preparation
High-quality legume inoculants is &pendent on selection of suitable carrier ma-
terials and their pre-treatment (Kremer and Peterson. 1983). An assessment of the literature
on carriers of rhizobia leaves the net impression that peat is still unchallenged as a carrier
and that it is relatively easy to devise a substrate from a variety of matea%& that would
support satisfactory growth and survival of rhizobia (Burton, 1976 ; Brockwell, 1980).
As a rule, carriers must have a high water-holding capacity, provide a nutritive medium
for growth of rhizobii and enhance survival during storage, distribution and when ino-
culated onto seed Various carriers of organic origin have been investigated These include
: soil and peat mixtures, soil plus charcoal, Nile silt plus nutrients, soil plus coir dust,
coal with and without nutrients, composted maize cobs, bagasse, filter mud from sugar
cane, cellulose powder, rice husk, bentonite, ground common talc.
The most important factors to consider before selecting a particular carrier are
the availability and characteristic of the material. Whereas the choice may be aided by
a chemical analysis, actual multiplication and survival studies of Rhizobium strains in
136
it are essential. Pinely ground peat is the rhizobial carrier used most extensively in conven-
tional leguine inoculant pxodllctial. Afm pmper neutmnsation with taco,, and milling
to as fine a particle size as possiile, broth culture of rhizobial strain containing at least
109rhizobidmlaremixedwithdryunsterilizcdpeatinbulkandthenmaauedinshallow
trays for 7296 hours at temperatures near WC. Inoculants pmpared with unsterilized
peat usualIy contain up to lOO-fold fewer rhizobia than skriked peat Packaging is done
in polytbene bags, moisture content cormsponds to 3540 % on a wet weight basis. Rhi-
zobia g,row and survive better in sterile than in tuwfxile peat. The extent of this dif-
ference is depeknt on the suitability of the peat and the temperature and duration of
storage. ‘Ihe choice of a method for steriWng peat depends on the type of container
inwhichtheculMeistobe~~,thenumberofculturestobeprepared,andthe
availability of steriking facilities, Autoclaving, ethylene oxide treatment, and gamma irra-
diation are used. Stklized peat in sealed polythene bags is inoculated by injecting enough
broth culture to raise the moisture content to 50-60 %. The puncture is sealed with
adhesive tape, the peat and broth are blended by manipulation and then incubated at
26°C for 2 weeks followed by storage at 4°C. High quality inoculants containing up to
lOlo rhizobia/g can be prepared by this method (Roughley and P&ford, 1982).
Alternative can&s to prepare Rhizobium inoculant are polymer gels ma& of po-
lyacrylamide, alginate, or xauthan. They were successfully tested 10 years ago by Dom-
mergues et al. (1979). Alginatebased polymer-entrapped Rhizobizun inoculant contain 109
to 10” ceils/g dry material of high survival ability Rhizobium (Diem et al., 1988).
However, this technology requires specialized equipment and a certain level of expertise
that are not yet adapted for use in developing countries.
Improved technologies for dried microbial cells preparation through free;tedrying
processus, in which various protectants are added, were recently applied for Rhizobium
inoculauts. Thus;high number of rhizobia can be added to seed because of the very high
count of the inoculant so prepared. Rhizobium can sukve for a very long time in these
concentrated inocuWs (E. Brochu, Rosell Institute, Montreal, Canada, pers. comm.).
Survival of rhkobla
Survival ability of rhizobia throughout the manufacturation and uses of inoculants
am very important, since the inoculum must contain very high number of rhizobia if
success of nodulation is to occur in field situation. Survival of Rhiwbium has been
studied in culture media (Bordekau and McLoughling, 1982), in carriers &renter and
Peterson, 1983), in the seed environment (Burton, 1986) and in soils (Bushby and Mar-
chall, 1977). Cultural factors and conditions may affect survival ability ; it is generally
recagnized that yotmgex cells of Rhiwbiwn survive better when subjected to stress con-
dition because they have more endogenous energy available to them to overcome the
adverse conditions (Bordeleau and McLaughlin. 1982). In liquid medium, survival
abiity of fast growing Rhizobium is not significantly affected by limited amounts of
carbon, phosphorus and potassium in growing medium. However, under desiccation, sur-
vival is poor unless cells are protected by agents such as lactose. It appears that cell
leakages, and therefore cell damages, are limited by lactose and the exact nature of this
protection is unknown. Commercial production of fast-growing rhizobia on lactose rich
by-product such as whey is now becoming a practice (Eissonnette et al., 1986). not
only because this medium permits the production of high concentrated broth (lOlo rhi-
137
zobia/ml), but also acts as a good protectant and contributes to high survival. Strains
of tibia varied in their susceptibility to desiccation, and they have to be selected ~CCOT-
ding1 y.
Survival of organisms plays an important part in microbial ecology, especially
when the environmental stress may adversely affect the cell. These environmental
stresses are generally the lack of nutrients (energy) in the environment. la& of moisture,
presence of other organisms, changes in pH, Presence of chemicals, etc. When applied
to the rhizobii, the ability of legumes to be properly infected with desirable strain depeuds
on the rhizobia to survive in the carrier, on the seed, and in the soil under adverse conditions
until it is ready to infect the host. With the current practice of. inoaihnmg the seeds,
ability of the rhizobia to survive throughout the process is of the first importance. As
mentioned earlier, carriers can be manipulated to sustain high number of cells. However,
exposure of inoculants to hot and dry conditions during shipping, storage, and planting
often results in decreased numbers of N,-fixing effectiveness of the rbizobia. Variable
inoculant quality is a common problem in the tropics and subtropics where imxulants
are often subjected to unpredictable handling and storage during distribution and use.
Therefore, inoculant formulation and packaging cannot be overlooked from a practical
point of view.
The survival of rhizobia on seed is of the greatest impormnce since seed pelleting
as a method of legume inoculation has received widespread attention and it is now a
standard agricultural practice. Rhizobia die quickly on freshly inoculated seed particu-
larly with broth inocula (Burton, 1976 ; Brockwell, 1980). Therefore, protective agent
assuring survival of rhizobia and a durable coating the seeds without affecting their genni-
nation must be used Lost of humidity is an important factor for the death of rhizobia
on the seeds (Salema et al., 1982) and, on the other hand, the coating has to be dried
enough to stay stable and to equilibrate with the moisture level of the seed. Other l&or-s
must be taken into consideration such as toxicity of the coating, pH and nutrient value.
Pelleting of leguminous seeds with organic based inocula and pulverized limestone and
nutritive sticking agents such as lactose, gum arabic and cheesewhey provides both a
good survival of rhizobia and uniformity of coating. Pelleted seeds are often treated with
fungicides to protect them against fungal diseases. Problems arise when legumes ino-
culants are used in conjunction with pesticide treatments (Tu, 1981). Such treatments may
affect the survival and/or the infectivity of Rhizobium imculants, unless resistant strahs
and compatible doses of pesticides are used (Odeyemi and Alexander, 1977).
Quality control program
The quality of legume inocula depends on both the number of rhizobia they
contain and their effectiveness in fling nitrogen with the intended host. Quality control
tests measure the number (quantitative) and identify the type (qualitative) of rhizobia in
the incculant during its recommended useful life (Roughley and Pulsford, 1982). Quality,
therefore, is determined by the various strain attributes mentioned earlier. The standards
of quality adopted need to reflect both the number of rhizobia required for successful
nodulation and the ability of manufacturers to produce such inoculants. The principal
reason for quality control is consumer protection, a function normally performed by go-
vernment agencies. However, such control applies only to the fmal products. Quality control
during preparation and before sale is usually the responsibility of manufacturers.
138
It can be safely stated that the highex the number of infective rhizobia contained
in the inodant, provided the strain is efficient, the better the inoculant. All investiga-
tors recogn& the importance of inoculant having at least a lo-fold more rhizobia when
produced than required at the time of usage to compensate for die-off. Ideally the standard
for an inocuh@ is based on the mm&r of infective rhizobia/seed necessary to initiate
prompt and effective nod&Con. This number will vary considerably from the favcrable
conditions of aseptic culture in greenhouse to the highly competitive situation in the field.
In practice, standards are strongly infhrenced by the level of rhizobii that manufacturers
can consistently attain and maintain in their inoculants with available technology -Many
investigators recommend an inoculum which would provide approximately l(Y to 1V rhi-
zobia per seed for a good nodulation. Even at this level of inoculation and where indi-
genous soil rhixobial population might be as high as lWcells/g soil, the introduced
Rhiwbium is so heavily outnumM that it needs special advantages in order to become
established. In Cauada, quality standards of inoculants were established at levels that
provide at least103 infective rh&bia/seed for small seeded legumes, and up to 10s infec-
tive rhizobia/seed for larger seeds such as soybean and peanuts (Bordeleau and
Prevost, 1981).
Basically any program should involve both qualitative aud quantitative tests of
the broth culture before impregnation of the carrier, fmal carrier culture immediately after
manuf&tu&on and before release, and inoculants obtained from the retail outlets. Tests
on broths and carriers before release am essential for all batches of culture. Broth cultures
should be examined for authenticity by appropriate immunological tests to ensure the
identity of the inoculaut stmin, and for freedom from contamination by absence of growth
on glucose-peptone agard and gram-stained smear free of positive cells. Results of these
tests~aMilablewithin24hours.Thestrainintbeb~thshouldalsobecheckedfor
infectiv’~ess and effectivemss by inoculating plants growing on N-free medium in an
aseptic environment.
After maturation of the inoculant in the carrier and before distribution, plate counts
on sterile carriers should be made to determine the number of rhiibii present and to
ensure that contamination is low. Where unsterile peats are used, plate counts am not
feasible because no specific media exist to differentiate rhizobial cells, plant-infection
tests should be made (Bordeleau and Prevost, 1981). Although, labor tune, space and
material consuming, the plant-infection test provides a useful tool to assess identity and
number of infective rhizobia in non-sterile carriers. Only after completion of these tests
should the inocuhnt be released for use and an expiry date assigned. The length of the
useful life of the inoculant will depend on the chamcteristics of the packaging and carrier
materials, storage conditions, methods of distribution and sale. Therefore, it is impossible
to set general expiry date.
The quality control program should include testing of inoculants sampled from
retails outlets. Such a program was established in Canada and it contributed greatly to
improve the quality of legume inoculant products offered to the farmers. This program
provided information on the effect of storage conditions, on the performance of particular
inoculant and a check on tire labelling and claims by manufacturers. It also provided
maximum consumer protection aud more direct assistance to manufacturers by early de-
tection of problems such as poor survival in carrier culture and on pm-inoculated seeds,
or t%hxre to produce nodulation on the recommended host-plant.
139
Field applications of legume inoculants
The act of inoculation concentrates the inoculum close to the position where tbe
host plant will first accept an infection whereas competing soil organisms, sometimes
much more numerous, are distributed randomly. A major advantage enjoyed by .inocu-
lant rhizobii is proximity to a developing legume rhizosphe.re where it can multiply.
Moreover, the time between sowing and ger&nation is a critical period. If the inoculant
mortality rate is high due to a soil environm ent that is physically, chemically, or microbio-
logically unfavourable, delayed gemGnation can lead to a situation where the &z&ii
popdation is too low to colonize the rhizosphere. The likelihood of nod&ion can be
increased by reducing the mortality rate of the inoculant, by raising the rate of inoculation,
or by decreasing the time to seedling germination. Increased oi; massive rates of inoculation,
may lead to the establishment of the inoculant in the soil micro-habitat by weight of
numbers. Alternatively, seed pelkting procedures protect the rhizobia against deleterious
environmental factors. We will &sent here a few methods in usage for inoculant appli-
cation in the field, with a relative apprakal of each method.
Seed Inoculation.
Dusting. At one time some manufacturers recommended that ino-
culant be merely mixed with the seed immediately before sowing or sprinkled on to the
seed in the seed box. Some culture adheres to the seed by lodging in the micmpyle and
in scratches and irregularities on the seed coat, and by electrostatic attraction. However,
much of it is shed particularly during passage of the seed through machinery. Dusting
is certainly the simplest method of inoculation but it is also the least effkient with conven-
tional products, and unless break-through in this method proves a better efficiency. by
;.
the
use of new products, it is not a recOmmended method.
Slumy inoculation. In order to increase the amount of inoculant
adbering to the seed, the culture is applied as a water suspension ; alternatively the ino-
culant is mixed with moistened seed. The seed must .be dried before sowing (not in direct
sunlight), but as it dries a proportion of the inoculant is lost. More inoculant can be
attached to seed by using an adhesive in the slurry ; e.g. a 10 96 solution of household
sugar (but sugar lacks tenacity). Adhesives such as gum arabic and substituted cellulose
compounds are more tenacious. Caution must be exercised to avoid samples of gum arabic
which contain preservatives k&al to rhizobia. The celluloses tend to be rathe~ cheaper
and more readiIy available than gum ambic.
Seed pelkting. Coating legume seeds with lime to promote no-
dulation apparently had its origin in Maine, U.S.A., more than 40 years ago. The development
of the process as a technique for establishing pasture species on acid soils was begun
in Australia in the 1950”s (Roughley and Pulsford, 1982). Its use is now widespread
extending beyond more protection against soil acidity. Advantages include protection of
the inoculant rhizobii against rhizobiotoxic substances contained in some legume seed
coats, unfavourable physical and chemical conditions in soil, competition from the soil
microflora, the effects of acid fertilii, and against seed-harvesting ants. Pelleting makes
aerial sowing of inoculated seed a practical proposition and ensures better survival of
the rhizobia when delays between inoculation and sowing are unavoidable.
140
Materials used for pellet coatings include calcium carbonate in many forms, dolo-
mite, various grades of gypsum, bentonite and other clay minerals, rock phosphate and
other phosphorus compounds, inert materials such as titanium dioxide and talc, soil and
humus, and activated &u-coal. The main requirements for a good coating mate&l are
tbatitsbouldbe~v~yc~toneutralityiapHand~lygrwnd(9o%atleast
passing 300 mesh). Adhesives used for attaching the coating material to seed include
synthetic glues, glues of vegetable or animal origin, gelatin, various sugars, and honey.
The adhesive should have sufficient tenacity so that the coating material does not slough
off, but should not be so tenacious that the cotyledons are damaged during geimina-
tion.. Adhesives must be free of preservatives.
Seed bed inoculation
There are several situations in which seed application of rhizobia may be an inef-
ficient means of inoculation : (i) preemergence diseases- or insect attack may make it
necessary to use seed dressings of fungicides or insecticides all of which have some effect
on the viability of rhizobii and some of which are very toxic indeed (Odeyemi and Alexan-
der, 1977 ; Tu, 1981) ; (ii) inoculation for broad-acre sowings of crop legumes with
high seeding rates is a major task, limiting the speed of the sowing operation ; (iii) some
seeds are too fragile to be inoculated, e.g. shelled peanuts (huch hypoguea L.) ; (iv)
some legumes such as soybean (GZycti mur (L.) Merr.) and subterranean clover (nijb-
lium subterranewn~
lift the seed coat out of the soil during emergence so that the rhizobia
on the seed coat are not deposited in the soil ; (v) the dimensions of the seed surface
place a test&ion upon the number of rhizobia which may be applied ; sometimes, .espe-
dally when small-seeded IegMles are used or where naturally occurring competitive rbi-
zobia are present, it is necessq that rates of inoculant application should exceed this
Limit ; (vi) the seed coats of some~species contain substances toxic to rhizobia. In these
circumstances, almtive means of inoculation are needed.
Solide inoculant - Solid inoculant is made by coating solid, granu-
lated material with peat inoculant in an adhesive. Suitable adhesives include a 25%
aqueous solution of gum arabic (no preservatives) or a 5% solution of methyl cellulose.
Tenacity of the adhesive solution can be improved by chilling overnight Peat inocu-
lant is thoroughly stirred into the adhesive and this suspension poured on to the beads
and mixed together until all beads appear evenly coated. The beads should be dried by
~ginathinlayer.Whendryandanylumpshavebeenbrokenup,thematerial
is ready for use. other inert particulate, freely flowing material such as marble chips,
coarsegraded sand, polyethylene beads, clay gram& can be used for preparing solid
inoculant. Availability and economic factors are the main points to consider in choosing
material. For large-scale application, solid inoculant can be applied through the fertilizer
box or insecticide attzzhment of a seed drill.
Liquid (spray) inoculation - Excellent nodulation can be obtained
by spraying inoculum into the row beside or beneath the seed. A peat culture of rhizobia
or fi-ozen, concentrated broth cultures, is mixed into a paste with water, diluted to a slurry,
then added to a water-filled tank prior to spray application. Any spray equipment is satis-
factory provided it has not been used previously for toxic chemicals. This method provides
very good results because high number of rhizobia can bc applied. However, marketing
141
problems exist especially with frozen concentrated cultures which deteriorate during
handling if proper conditions are not given to avoid thawing of the preparation
Pre-inoculation before sale
pre-inocnlationistfieinoculationofseedpriartosaleor~utionforfarmers;
it implies a ,peaiod of storage of theinoculamd seed before it is sown. The main objec-
tiveis~allowseedmetchants~buildupas~~~ofinoculatedseedino~to
meet predictable heavy demand by farmers at time when sowing deadlines are imposed
by climate. Bacteriologically, the objectives should be to use culuues containing very
high populations of rhizobia and to preserve their viability so that, at sowing, the level
of inoculum on pre-inoculated seed is equal to that on freshly inoculated seed. New tech-
nology exists to produce pelleted, pre-inoculated seed with excellent mechanical proper-
ties and large population of viable rhizobia. Pm-inoculamd seeds should meet the same
standard in the quautitative terms of infective rhizobii by seed as that provided by ino-
culation at sowing time. Seed pelleting represents the major aud probably the only real
advance in the technology of inoculant application, and appears to offer the best prospects
for the development of a reliable inoculatior~
Concluding remarks
The introduction of a legume seed inoculant strain into soil aud its establishment as a
permanent member of the soil community is an exercice in applied microbii ecology.
Quantitatively, the number of rhizobii associated with iuoculamd seed mt only a
way small fraction of the total soil microflora. The choice of a strain, the quality of
the inoculant and its method of application are impor%+ to insure that the mtmduced
Rhiwbium strain survives and establishes an effective symbiosis with its designatkd ho%
Unless proper methods of inoculaut manufacturation and application are used, inoculant
will fail to meet the objectives of its practice, regardless of the strain. Quality control
programm is an essential part of inoculation technology.
References
ANTOUN, R, BORDELBAU, L.M. and LACHANCE, R.A. 1979 Rendement de la hrzerne
(cultivar Saranac) inocul6e avec une souche t&s effkace de Rhiwbizun meldoti en
presence d’autres eqhes de Rhiwbium. Can. J. Plant Sci. 59, 521-523.
BISSONNETIE, N., LALANDE, R and BORDELBAU, L.U 1986 Large-scale produc-
tion of Rhiwbium meliloti on whey. Appl. Environ. Microbial. 52, 838-841.
BORDELEAU, L.M. and ANTOUN, H. 1977 Effets de l’inoculation mixte avec des
souches de R. meliloti sur le rendement de la luzeme. Can. J. Plant Sci. 57,
1071-1075.
BORDELEAU, L.M. aud McLOUGHLIN,‘TJ. 1982 Survival studies with Rhizobium
species. Can. Sot. Microbial. Ann. Meeting Proc. EM-I.
142
BORDEIXAU, L.M. and PREVOST, D. 1981 Quality control program by federal go-
vernment agencies and method of testing legume inoculant and preinoculated seed
pmducts in Canada. In : Clark, KW. and Stephens, J.H.G. (eds.), Pmuxdings of
the 8th North Amexican Rhiwbium Confea~~ce, Winnipeg, Canada, pp. 566-579. The
university of - F+iirhg stices, wilmipeg.
BORDELEAU, LM. ANTOUN, II. and LACHANCE, RA. 1977 Effets &s souches
de Rhiwbium melihi et des coupes successives de la hrzerne (Medicago sclrivp) sur
la fvration symbiotique d’azote. Can J. Plant Sci. 57, 433-439.
BROCKWELL. J. 1980 Experiments with crop and pasture legumes - Principles and Prac-
tice. In : Bergersen, FJ. (ed). Methods for evaluating biological nitrogen fixation,
pp. 417488. John Wiley and Sons, N.Y.
BURTON, J.C. 1976 Methods of inoculating seeds and the effect on survival of Rhizobia.
In : Nutman P.S. (ed). Symbiotic nitrogen fixation in plants, IBP 7, pp. 175190.
Cambridge University Press, Cambridge.
BUSHBY, lXV.A and MARCHALL,, KC. 1977 Some factors affecting the survival of
root-nodule bacteria on desiccation. Soil Biol. Biochem. 9, 144-147.
DATE, R.A. and ROUGHLEY, RJ. 1977 Preparation of legume seed inoculants.
In : Hardy, R.WF and Gibson, AH. (e&z.). A tmatise on dinitrogen fucaton, section
IV, pp. 243-275. John Wiley and Sons, N.Y.
DIEM, H.G., BEN KHALIFA, K., NEYRA, M. and DOMMERGUES, Y.R. 1988 Recent
advances in the inoculant technology with special emphasii on plant symbiotic mi-
croorganisms. Workshop on advanced technologies for increased agricultuml produc-
tion, Santa Margherita Ligure, Italia, 25-29 Sept. 1988.
DOMMEKC;tiES, Y.K., DIEM, H.G. and DIVLES, C. 1979 Polyacrylamide-entrapped
Rhizobium as an inoculant for legumes. Appl. Environ. Microbial. 37, 779-781.
GUEYE, M. and BORDELEAU, L.M. 1988 Nitrogen fixation in bambara groundnut, Voand-
z&a subterranea (L.) l%ouars. Mircen J. 4, 365-375.
KREMER, RJ. and PETERSON, ILL. 1983 Effects of carrier and temperature on sur-
vival of Rhiwbium spp. in legume inocula : Development of an improved type of
iddant. Appl. Environ. Microbial. 45, 1790-1794.
McLOUGHLIN, TJ., BORDELEAU, L.M. and DUNICAN, LX. 1984 Competition stu-
dies with Rhizobium mfolii in a field experiment. J. Appl. Bacterial. 56, 131-135.
WE, J., HIGGINS, P. and GGARA, F. 1985 Production and storage of Rhizobium
leguminosarum cell concentrates for use as inoculants. J. Appl. Bacterial. 58,
5 17-524.
‘_.
.:
.’
1,
‘!...
..--.-
:
143
ODEYEMI, 0. and ALEXANDER, M. 1977 Resistance of Rhiwbium strains to Phy-
gon, Spexgon, and Thiram. Appl. Environ. Microbial. 33, 784-790.
ROUGHLEY, RJ. and PULSFORD, DJ. 1982 Production and control of legume
ino-
culants. In : Vincent, J&i. (ed.), Nitrogen Fixation in Legumes, pp. 193-209. Act+
demic Press, Sydney.
SALEMA, M.P., PARKER, CA. and KEDBY, DJL 1982 Death of Rhizobia on inocu-
lated seeds. Soil Biol. Biochem. 14, 13-14.
STOWENS, M.D. 1985 Carbon metabolism in Rhiwbium species. Ann. Rev. Microbial.
39, 89-108.
. _
TU, CM. 1981 Effect of fungicidal seed treatments on alfalfa growth and nodulation.
Chemosphere 10, 127-134.
Mmimisez la FBA pour la Production Agricole et Forestière en Afrique
Caractérisation et choix de supports
pour inoculum
BEUNARD, P. et &UN!I’ MAURY, H.
IRAT~CIRALI - BP. 5035. 34032 MONTPELLIER cedex France.
L’étude réalis& avait pour but de tenter de relier quelques unes des carackk-
tiques Dhvsicn-chimiques de SU~JXX?.Q possihlw d’innmlllm WCC leur cayGt6 r;rllr
à permettre la survie des rhimbiums.
Les modalitis de conditionnement des inocuhuns ont tgakment fait l’objet de mises
au point visant B l’obtention de produits de qualité Llev&
Les supports utilisks proviennent de pays où des unit& de production d’inoculum
pour l@mineuses ont &5 installh dans le cadre de programme de coop&tion
technique de la FAO : Madagascar, Tunisie, Zaïre, Bhutan, Nicaragua, Turquie., France.
Les rhultats obtenus ont permis de montrer : (i) L’importance du pH initial du
mpport, l’optimum se situant entre 6 et 6.5. Lorsque œ n’est pas le cas, une neutra-
lisation peut êh~ effecde avant l’utilisation ; (ii) les conditions & sttkilisation des
apports ont des cons&wnces primordiales sur la survie des rhk9knns ; (iii) I’exis-
tente dans la majoritk des cas d’une Ikg&re multiplication des rhizobiums au cours
des premiks semaines suivant l’ensachage puis une dkroissanœ
plus ou moins
rapide selon les conditions de conservation (te.mp&at.ure de stockage).
Ces r&ultats ont conduit & des recommandations sur les mkhodes de production
des irmxulums dans des unités de petite taille. Leur mise en œuvre fait tgalement
robjef d’une jxckerltation.
145
Introduction
Le nombre de rhizobiums contenus dans un inoculum conditionne en grande partie
son efficacitk. Lors de la production 2 l’tkhelle pilote, pour les besoins de l’expkimen-
tation en milieu paysan, la qualiti du produit utilis5 joue un rôle primordial et doit être
aussi élevée que possible. La multiplication des rhizobiums dans des petits fer-menteurs
simplifiés mis au point par I’IRAT et baptises UPIL (unité de Production d’Inoculums
pour Légumineuses) est maintenant bien maitriske et adoptée dans plusieurs pays (Saint
Macary et al, 1986). L’étape suivante de la production, dans laquelle la culture liquide
est mélangée a un support nkessite la recherche, au niveau local, de composés permettant
une bonne survie des rhizobiums. Dans le cadre de projets de coopération technique de
la FAO, de tels supports ont été recherches et leur aptitude & être utilises dans les unités
a et& évaluée.
Matbriel et m&hodes
Les principales caractéristiques physiques et chimiques des supports utilises sont
prksentks dans le Tableau 1.
L’ensemble des supports a &.k séché a l’air puis à l’étude a 50°C pour obtenir
une humidité finale de 12 à 15%. Ils ont ensuite été broyés. passes au tamis de 0,5 mm
puis conditionnés dans des sachets en plastique autoclavables (PolytWrylène moyenne den-
site ; épaisseur 150 p.) a raison de 80 g par sachet. L’ensemble des sachets contenant
les supports a été stérilise par trois passages B l’autoclave (120°C, 20 mn), les autoclaves
étant espacds de 24 h
Les cultures de rhizobium utilisées pour injection dans les sachets ont ktk rkalisks
en fioles d’erlenmeyers, agitées & 30°C contenant du milieu yeast extract mannitol
(YEhî), (Vincent, 1970). Les inoculums ont été préparés en injectant en conditions skiles
dans les sachets une quantité de culture liquide correspondant ?t une valeur situke entre
40% et 60% de la capacité maximale de rétention en eau du support (Date et Roughley,
1977).
Api2 JiiLeLdib L~I~~S, CL, szion ics ui~~kmmth ;ippqucs ~~LCIIII~~UUII
uu non,
neutralisation ou non, température de stokage), le nombre de rhizobiums contenus dans
les sachets a été déterminé selon les méthodes de suspension dilution puis étalement en
boite de p&ri (FAO, 1984).
RBsultats et Discussions
Neutralitb des supports
Les résultats obtenus sur la tourbe de Betafo (Madagascar), représentés dans la
Fig. 1 montrent clairement que la neutralisation par du carbonate de calcium d’un support
tri?s acide peut considerablement améliorer sa capacité a maintenir une population micro-
bienne en vie. La tourbe neutralis&, actuellement utilisée pour la fabrication d’inoculums
permet la survie de Brudyrhizobium japonicum pendant pres de deux mois, ce qui permet
d’envisager un délai entre la fabrication des produits et leur utilisation.
Tableau 1 : Caracteristiques chimiques des supports utilises
Lieu de
Identification
Conductivi-
Pays d'origine
Nature
prélèvement
sur Fig. 4
pH (eau)
Carbone (%)
C/N
té è 25'C
(micromhos)
Madagascar
Tourbe
Betafo
3,78
15,86
16
1180
Turquie
Tourbe
Bol~
5,60
41,06
23
1650
Trabzon
4,60
24,65
2s
600
Kayseri
7,60
9,90
17
1700
B h u t a n
Tourbe
Par0
4,60
21,80
13
-
Nicaragua
Tourbe
Las Playitas
6,35
14,70
64
176
6,68
6,38
36
106
6,90
7,56
34
116
San Tarlos
5,90
19,02
20
104
Balle de riz
5,14
27,70
69
397
Ecume de canne
5,92
37,62
51
2750
Bagasse de canne
5,54
14,74
79
350
Bagacillo de canne
6,18
15,60
80
335
Pulpe de caf@
7,37
46,94
42
2330
0
0
147
i
.
148
SMlIté des supports
Le traitement par la chaleur humide de tourbes de Turquie (Fig. 2) amkliore la
conservation des rhizobiums. Il est naturel, en effet, qu’en l’absence d’autres microorga-
nismes, les rhizobiums ne rencontrent aucune comp&ition pour les sources d’éikments
nutritifs contenus dans le support. L’hypothèse souvent exprimk d’un risque de dégra-
dation du support pendant la stkilisation et d’une libération de composb toxiques pour
les rhizobiums ne se trouve cependant v&if%e pour aucun des deux supports comparés
dans cette &ude. Entre ces deux supports cependant, des difftkences notables existent
et la tourbe extraite de Bolu fait aujourd’hui l’objet d’une utilisation par l’industriel de
fabrication d’inoculum turque.
Tempbature de conservation des supports
L’étude r&lis& sur une tourbe de Paro (Bhutan) comparant l’effet du stockage
d’inocnlnms à 4’C et 25OC a permis de montrer (Fig. 3) qu’il n’existait pas, pour la
p&iode consid&& de diffkrence importante entre les traitements. D’autres r6sultat.s (non
présentés ici) ont montré que la conservation des inoculums & tempkrature ambiante ou
& 30°C pouvait rkwlter. dans un premier temps en une multiplication des rhizobiums
et l’obtention de meilleurs inoculums que ceux conservés dès leur conditionnement à
basse tempkature.
Relations pH, matibre organique, survie
La survie de la souche de Bradyrhizobium japonicum IRAT FA3, apr6s 50 jours
de conservation dans différents supports est repr&ent& dans la Fig. 4, où sont égale-
ment indiquées deux caractkristiques chimiques des supports : le pH et la teneur en
carbone. On constate qu’il n’existe pas de règle gtkkale, au moins pour les deux facteurs
pH et C, permettant de pr6dire la valeur d’un support, Les meilleurs rkwltats ont Cd
obtenus avec des supports pmches de la neutralité (pH = 5 à 7) et pour des teneurs
en C moyennement élevks. Ces conditions peuvent semb1e.r nécessaires mais ne sont
pas suffites puisque certains supports, comme la bagasse de canne (2) ou la tourbe
de Las Playitas (C) ne donnent que des r&ultats mtiocres. On peut remarquer cependant
cjti: kS xillcurs rCstiki1;s 51;iii SOUVCiii &&;;ds a\\&. &s cornpusk i>rg:ili~;qu~;b LUIIIIIIC
la tourbe (Uantillons D, B, K, N), les matières organiques fraîches (Z, R, S) étant
souvent de mauvais supports d’inoculums.
Conclusion
L’ensemble des rkwltats obtenus dans ces &udes et dans d’autres essais non pré-
sentés ici, montrent la difficulté qu’il y a à prévoir, par une simple analyse chimique,
le comportement d’un support vis-à-vis des rhizobiums. Parmi les rkgles pouvant être
tkonckes pour la fabrication d’inoculums dans lesquels les rhizobiums pos&ient une pro-
babifti de survie tlevtk, celles de la neutralité et de la st&ilid du support semblent
primordiales. Par la suite, une &Valuation bas& sur des tests biologiques de survie des
microorganismes semble obligatoire pour dCterminer les meilleures conditions de prc$a-
ration et de conservation des produits. En la matière, une bonne connaissance de l’&o-
lution du nombre de rhizobiums dans le support permet de déterminer la dur& maximale
d’utilisation des inoculums, de prkvoir leur dur& de production en conskquence, et de
prkiser les conditions de leur distribution, la tifrigération ne constituant pas toujours
une obligation.
149
Nombre de rhizobiums
,
vivants par gramme d’inoculum
(Log)
Sauche employ6e :
R. lquminowum biovar phnsd PIXC 40?4
8 i--.
>
0
8
15
2 2
3 6
50
Temps (jours)
Figure 3 : Effet dc! la tempkature de conservation des inoculums sur la survie des rhtzobiums
151
152
Cette étude n’a pris en compte, comme critère de qualité des inoculums, que le
nombre de Rhizobium vivants. Il n’est pas certain que dans tous les cas des modifica-
tions de l’efficacite des Rhizobium ne se soient pas produites. L’obtention et l’applica-
tion d’un nombre aussi élevé que possible de Rhizobium viables constitue cependant un
prt%lable indispensable a une bonne qualité des inoculums. C’est pourquoi le crX?re de
nombre a été seul consid6re dans l’etude.
Remerciements
Le auteurs remercient T. DORJI, S. MARTIN DU PONT, M. NEYRA, M. OCHOA,
D. PINOIT, P. RAKOTONDRAMASY et J.A. SCAGLIA pour leur participation aux ex-
p&imentations décrites. Ils remercient également le Service des engrais et de la nutrition
des plantes (AGLF) de la FAO pour la fourniture des différents supports.
Rbfbrences
DATE, R.A. and ROUGHLEY, R.J. 1977 Préparation of legume seed inoculants.
In : A treatise on dinitrogen fixation. Section IV, Edited by Hardy, R.W.F. and Gibson,
A.H., John Willey and Sons. p. 243-275.
FAO 1984 Fichier technique de la fixation symbiotique de l’azote, Organisation des
Nations Unies pour I’Alimentation et l’Agriculture. Rome.
SAINT MACARY, H., BEUNARD, P., MNTANGE, D., TRANCHANT, J.P. et
VERNIAU, S. 1986 Setting and diffusion of a production system for legume Rhizo-
bium inoculants. Symbiosis. 2, 363-366.
VINCENT, J.M. 1970 A manual for the practical study of root nodule bacteria.
IBP Handbook N*15. Blackwell Scient& Publications, Oxford and Edinburgh.
Maumiser
lo FBA pour la production agricole et fortxtière en Afrique
Progrès récents dans la technologie
des inocülums utilisés en agrictiture
et en foresterie
DIEMI, H.GP’, BEN KHALIF’A, KP’,
NEYRA, M.@) and DOMMERGUES, Y.R.“‘.
(1) : BSSlT (ORSTOMICTTFTICNRS)
45 bis, avenue de la Belle Gabrielle 94736 Nogent-sur Marne cedex (France) ;
(2) : SocU Calliope, 17, rue de Shastopol,
34500 Bhziers (France) ;
(3) : ORSTOM, BP. I386, Dakar (Sénégal).
Summaty
The preparation of inoculants requires (i) mass production of the target micro-
organism to be introduced in the agrosystem, and (ii) an appropriate formulation
of the microbial culture ensuring long survival during storage and case of appli-
cation
The pxinciples of mass culture methods for gmwing rhizobia, ectomycorrhizal
fungi, Frankiae, and VAM fungi are briefly presented. Emphasis is specially piaced
on the di!Sulties encountenzd when gmwing Frmkiue and on the solutions adopted
to increase its bîomass production, which includes biphasic culture, two stage culture,
ancl gmwth in a polymeric matrix.
The modem technique of processing the inoculants is based on the inclusion of
bacteria in a polyrneric matrix (generally alginate). Two types of polymeric it~~~hmts
cari be prepared : PEM inoculants made of microbial cells entrapped in the polymeric
matrix ; PEC inoculants made of miaobîal colonies entrapped in the matrix (the
colonks resulting from the growth of entrapped isolated cells). PEM and PEC ino-
culants are usually dried, then applied as a powder onto seeds or soil. or rehydrated
before beiig used. Polymeric ir~~hmts were shown to be quite effective and reliable
in field expeziments with rhizabia, VAM mycorrhîzae and frankiae.
154
Introduction
On peut classer en deux catkgories les microorganismes rhizosphériques et
phyllosph6rique.s qui améliorent la production végétale.
Le premier groupe est constitué par les microorganismes qui agissent directement
sur le veg&al en contribuant ZI sa nutrition azotée ; par ex. Rhiwbium ou Frankia, en
am&nanf l’absorption des min&aux et de l’eau, par ex. Awspidzun (Okon, 1984 ; Nayak
et al., 1986 ; Boddey et Dobereiner, 1988), champignons mycorhiziens (Menge, 1983;
Gianinazzi-Pearson e t G i i , 1986) ; en solubilisant certains 616ments. par ex. bac-
tkries solubiit les phosphates (Banza et al., 1%7 ; Azcon et af.. 1976) ; en synthe-
tisant des substances de croissance,* par ex. Awtobacter (Brown, 1982), Awspirillum
(Okon, 1987 ; Boddey et Do~re&r, 1988).
L.e deuxième groupe affecte indirectement la croissance du végétal en luttant contre
les microorganismes pathogènes, par ex. Trichoak-ma SP., Agrobacterizun radiobacter ;
en luttant contre les arthopodes, par ex. Beauveria bassiana ; en luttant contre les mauvaises
herbes (microorganismes pathogknes de ces plantes notamment mycoherbicides), par ex.
Phytophthora palmivora ; en pr6servant des effets du gel, par ex. Pseudomonas
syringae (Okon et Hadar, 1987).
En plus des microorganismes rhizophénques et phyllosph&iques, des microorga-
nismes endophytiques (c’est-à-dire vivant dans le xylème des plantes) pourraient être utili&s
à l’avenir, apAs manipulation génétique, comme agent de lutte biologique contre dif-
férents ennemis des plantes, spécialement des insectes (Carlson, 1988).
On pense que l’inoculation des plantes effectuée dans le but d’améliorer les r&oltes
est un type de biotechnoiogie &Zmentaire ; en fait il s’agit d’une op&ation relativement
compliquée (Bashan, 1986). Les techniques conventionnelles donnent des r&ultats in&
guliers et les m&hodes d’inoculation sont souvent primitives. C’est ainsi que dans certains
pays, on inocule encore les plantes actinorhiziennes avec des bmyats & nodules. On-cIoit
interdire une telle pratique en raison du risque d’introduction de souches ineffectives (non
furatrices d’azote) et surtout du danger de contamination des semis ou des boutures avec
des agents pathogènes, tels que Rhizoctonia solani ou Pseudomonas solanacearum dans
Ic cas dc Cuslrarina cqtisefifolia (Liang Zizh, l986) ou dcb nhu~des lorsque l’on
a affaire à des plantes sensibles à ces derniers parasites, telles que les acacias australiens
introduits en Afrique de I’Ouest..
Cette communication fait le point des progr&s les plus marquants réalisés r&em-
ment (i) dans la culture des quatres types les plus importants de microorganismes symbio-
tiques, Rhizobium, Frankiu, champignons ectomycorhiziens (EMC) et vésiculo-
arbosculaires (VAM), et (ii) dans le conditionnement du microorganisme utilisé pour
l’inoculation, Op&tion fond& sur l’inclusion du micmorganisme dans une matrice de
polymére, par exemple alginate. On donnera enfin les r6sultats d’essais d’inoculation au
champ portant sur ce nouveau type d’inoculum.
Culture en masse du microorganisme
La production d’inoculum repose sur la culture en masse du microorganisme que
l’on souhaite introduire dans l’agrosystème. Alors que l’on peut cultiver facilement certains
microorganismes in vitro (rhizobiums, champignons ectomycorhiziens), il est beaucoup
plus difficile de produire des quantit&s importantes d’autres microorganismes (par ex. Fran-
155
kia). En outre on n’est pas encore parvenu à cultiver in vitro un certain nombre de
microorganismes, en particulier les champignons VAh4.
Rhizobiums
On peut cultiver facilement les rhizobiums dans de nombreux types de fermen-
teurs, du type &atch» ou semi-continu (!Vii, 1984).
On peut aussi cultiver ces bactiries B l’état inclus dans une matrice polyémrique
suivant le pro&% pour les frankias (cf: infï-a). Contrairement aux lka&ias dont les CO--
lonies sont régulièrement distribuées à l’intkrieur des billes, les rhizobinm& comme les
autres microorganismes aérobies tels que Buciilus subtiZis (Baudet et al., 1983), poussent
seulement dans la couche externe des billes, dont le diam&re.ne devrait donc pas dépas-
ser 1 mm.
On sait maintenant que, dans la symbiose rhizobienne, nombre de composés appa-
rentés aux flavones stimulent l’expression des gi?nes nod, y compris les génes nod ABCEF
(seul le g&ne nod D est constitutif chez les rhizobiums ZI l’ttat libre), et acAlèrent la
nodulation, ce qui rend plus compétitifs les rhizobiums incubés en prksence de ces subs-
tances. Ces composés se trouvent dans les exudats des plantes-hôtes, mais la quantiti
exsudée peut ne pas être suffisante pour GAer la nodulation (Halverson et Stacey, 1986 ;
Redmond et al., 1986 ; Phillips et al., 1988). L’addition d’exsudats ou de substances
inductrices réduit le temps nécessaire à la nodulation (Rolfe et al., 1987 ; Wijffelman
et al., 1987). Ce traitement sera probablement commercialii dans un proche avenir.
Champignons ectomycorhiziens
On sait cultiver les champignons ectomycorhiiens (par ex. Stillus granr&tz~ ,
Rhizopogon luteus, Thelephora terrestris. Pisoüthus tinctorius) soit en milieu liquide,
dans difftkents types de fermenteurs y compris de grands fermenteurs industriels (Marx
et Kenney, 1982), soit sur substratum solide imprQné de milieu nutritif. Dans le second
cas, on met 2 incuber le substratum inoculk avec le champignon dans des sacs de plas-
tique de 5 à 10 litres dot& sur une face d’une membrane permettant les échanges gazeux
tout en empêchant des conramirwtinn?
Pwvtuelles (Le Taccr! c! Uxbr;:, 1936 ; Gsg-,~n
et ai., 1987).
Frankias
La culture in vitro de Frankia soulkve plusieurs problèmes @km et al., 1988)
dûs aux caractkistiques de cet actinomyc&e : (i) une croissance tri% lente, le temps de
doublement &ant de 50 A 60 h, (ii) l’inhibition de la croissance en batch lorsque la bio-
masse de Frankia dans le milieu atteint le seuil anormalement bas de 10 pg de prot&nes
cytoplasmiques par ml c’est-Mire 70-100 1.18 de protéines totales par ml de milieu,
(iii) une diminution II& rapide de la biomasse physiologiquement active à la fin de la
phase active de croissante (Fig.l), cette phase de dUin étant caract&is& par une lyse
marquée des diffkrentes structures de Frunkia, (iv) une sporulation souvent irr&uliére
et instable. Ce dernier point mérite d’être ttudib car on peut penser qu’en obtenant une
sporulatîon plus abondante, on améliorera la viabilité de Frankia dans l’inoculum
pendant le stockage.
Des expériences rikentes effectuks dans notre laboratoire suggèrent que l’on peut
stimuler la sporulatîon de Frunkiu en utilisant un milieu de culture limité en N ou en
apportant certains acides aminés (leucine, alanine, phénylalanine, arginine et valine).
L . E
h
156
Figure 1 : Courbe de croissance de la souche de fkankia ORS 020607. La courbe en fonction du temps de la teneur de la culture en protéines
comporte trois phases : (i) phase de latente (jusqu’au Sème jour), (ii) phase de croissance active, que l’on suppose de type linbaire, du 5kme
au 15ème jour et (iii) une phase plus ou moins stationnaire jusqu’au 962me jour.
157
Jusqu’à p&ent, on a utilise surtout les cultures en batch, avec ou sans agitation,
avec ou sans a&ation. Comme on ne peut utiliser ces méthodes pour produire les quantitks
importantes de Fr& nkessaires à la pr@aration des inoculums, nous cherchons ac-
tuellement a mettre au point trois nouvelles techniques de production en masse.
Culture blphaslque
Etant doMe que Frunkia pousse bien lorsqu’il est attaché à une surface solide,
Diem et Dommergues (1985) ont mis au point une m&hode de culture biphasique dans
laquelle on inocule le milieu liquide avec des fragments de gélose provenant de l’krasement
d’une culture de Fr& ayant pousse sur un milieu nutritif gt%&, les fragments de
gélose constituant des matrices où sont incluses de nombreuses structures de Frunkiu.
Cette méthode permet d’obtenir une croissance satistàisante de FrÜnkiu 21 l’interface entre
la courbe de gelose 6cras& (phase solide) qui tombe au fond du flacon de culture con-
tenant le milieu liquide (phase liquide). On obtient ainsi la formation, à l’interface solide-
liquide, d’un enchevêtrement homogène d’hyphes de Fr& ayant 2 mm d’6$akeur.
Une telle am&ioration de ht croissance au niveau d’une interface liquidesolide a aussi
et15 observée dans le cas d’autres bactt%ies (Bauder et al., 1983). Ce syst.&me & culture
biphasique permet en outre (i) de rktiver la croissance de vieillesoul~ de FrankG
qui ne pouvaient redm A la suite du simple transfert sur les milieux solides ou liquides
habituels, (iii d’assurer le développement des microcolonies obtenues au cours dé l’iso-
lement de Frmkiu iI partir des nodules, la biomasse initiale de ces microcolonies étant
trop faible pour en assurer leur croissance uMrieure dans les milieux solides ou lîq&les
habituels.
Culture ii deux atapes
Neyra a mis au point dans notre laboratoire une methode de culture a deux &apes
(resultats non publiés). Dans la premikre Ctape, on cultive Frunkia dans un chkmostat
conçu par Belaich, J.P. (comm. pers.) pour la culture des microorganismes a croissance
lente. Dans la deuxième ttape, la culture de Fr& qui sort du chemostat est utilisée
pour inoculer des batchs ancillaires contenant du milieu frais.
Culture en billes d’alginate
Le principe de la mtthode est le suivant : les celhrles de Frankîa sont incluses
aseptïquement dans des billes d’a@nate (qui constituent lamatrice polym&ique) obtenues
conform6ment au protocole dkrit au paragraphe consacre au conditionnement des îno-
culums et leg&ement modifie. Cette modification consiste a remplacer Yeau distilltk utilis&
pour faire la suspension d’alginate a 5 5% par un milieu de cuhure renfermant moins
de phosphate que le milieu usuel. Ces billes sont ensuite lavtks dans l’eau stkik et incubees
pendant 10-15 jours dans le milieu de culture. Les colonies de Frmkia qui se d&elop-
pentu à l’inteneur des billes (Fig. 5) renferment de nombreux sporanges au milieu de
chacune de ces colonies. Lorsque les colonies form6e.s a l’intérieur des billes sont suffkm-
ment développées, les billes sont &ch&s et constituent ainsi l’inoculation PEC en poudre
(cf : infra).
Dans le cas des plantes actinorhiziennes, l’expression des gks twd est proba-
blement induite par des substances du même type que ceBes qui interviennent dans la
symbiose Rhizobium-16gumineuse. En effet, Prin et Rougier (1987) ont montré que les
exsudat.s de 1’Alnus sont impliqub dans le processus d’infection.
-
158
Champignons mycorhiziens v&ico-arbosculaires (VAM)
Ainsi qu’on l’a signalé anterieurement, les champignons VAM ne peuvent être
cultivb in vitro, c’est-a-dire en l’absence de la plante-hôte. Contrairement aux rhizo-
biums et fknkias, ces champignons sont si peu spécifiques que presque toutes les espèces
végétales dotées d’un système racinaire abondant peuvent être utiliskes comme
plantes-hôtes.
La méthode habituelle, encore largement utilisée, consiste à cultiver la plante-hôte
dans un sol ou un substratmn stérile place dans un rtkipient ou directement sur le plancher
de la serre. On inocule le sol ou le substratum avec des spores ou des racines deja in-
fectks avec le champignon VAM à multiplier, provenant de cultures en pot effectuées
anterieurement. Linoculum ainsi prépare est volumineux et.. difficile 21 manipuler-
Récemment, Mosse et Thompson (1984) ont mis au point une méthode appelée
NET (Nutrient Film Technique). Cette methode consiste a faire pousser la plante-hôte
dont les racines ont et6 préalablement infectées avec le champignon VAM à multiplier
dans un canal étroit où coule la solution nutritive pour le vegétal. Le réseau dense de
racines formé dans ces conditions constitue un milieu idéal pour la multiplication du
champignon VAM. L’avantage de l’inoculum ainsi obtenu est qu’il suffit de 60 kg de
racines infectées pour traiter 1 ha, poids bien infkrieur à celui de l’inoculum classique
dkrit ci-dessus. En admettant que le rendement en racines soit de 300 g (poids frais)
par m de canal, on peut considérer qu’avec 200 m de canal on peut inoculer 1 ha @mes
et al., 1983). Il est 6vident que le matiriel ainsi obtenu n’est pas exempt de contami-
riants, les plus fi@uents ktant des protozoaires ou des nkmatodes saprophytes (Masse
et Thompson, 1984). La méthode NET a éte expkimentke avec succès sur haricot (Phu-
seolus vulgaris), maïs (Zea mays) et d’autres hôtes infestés avec GIomus mosseae,
G. fasciculatus, G. caleahium et Acadospora luevis @mes et Mosse, 1984).
Une autre. méthode est fond& sur l’emploi de plantes-hôtes cultivtks aeroponi-
quement. Hung et Sylvia (1988) ont confirmé des études antérieures montrant que cette
methode peut être utilisk avec SUC~S pour multiplier les champignons VAM. car une
bonne aération de la rhizosphère stimule la sporulation.
Enfin, Mugnier et Mosse (1987) ont proposé une approche originale pour amé-
liorer la culture des champignons VAM : au lieu d’utiliser la plante entière, on utilise
des racines transforrkes par le T-DNA cSAgrobacferium rhizogenes. Les racines trans-
formées de Convolvufw sepium, que l’on peut cultiver dans des fermenteurs usuels, sont
infectées par des spores de Glomus mosseae et assurent une bonne croissance du cham-
pignon. Ce pro&&, qui donne satisfaction au niveau du laboratoire, n’a pas encore été
commercialise.
Inoculants constitu& par des mtcroorganismes inclus dans une matrice
polym&ique
Le conditionnement des inoculums microbiens utilisés en agriculture est très im-
portant. Le succès de ces inoculums dépend en effet non seulement de la survie du plus
grand nombre possible de cellules microbiennes pendant le laps de temps s’écoulant entre
la fabrication et l’application, mais aussi de la formulation qui doit être simple mais per-
mettant un transport facile et une application aisée. Le conditionnement classique est fondé
sur l’adsorption de la culture microbienne sur un support protecteur, en g&&ral la tourbe
(Date 1977 ; Burton, 1979 ; Williams, 1984). Des supports de remplacement ont et& pro-
159
posés et utilisés tels que lignite, charbon, paille, cellulose, bagasse, résidus de récolte
broyés en poudre et sol (Chao et Alexanda, 1984). Dans certains cas, la culture et le
conditionnement du microorganisme sont effectues en même temps. C’est ainsi que les
inocuhrms d’ectomycorhizes sont faits d’un melange d’hyphes fongiques et du substrat
de vermiculite irnpr&ne du milieu spécifique sur lequel le champignon a pousse (Max
et Kenney, 1982).
Di+s 1977 et 1979, Dommergues et aL, ont suggeti que l’on pouvait utiliser des
gels de polym&es pour inclure les microorganismes b&%ques pour @parer des ino-
cuh.uns a usage agricole. Les premiers inocuhnns faits de microorganismes inchrs’ dans
des matrices polym&iques ont et6 tes& sur des rhizobiums il y a 10 ans. Ce concept
est maintenant appliqué avec succès, sur une &helle expt?rimentale, à d’autres groupes
de microorganismes, par ex. champignons VAM (Ganry ët 42.; 1982), champignons
ectomycorhiiiens (Jung et al., 1981 ; le Tacon et d., 1985), AzOspirilIum (Bashan, 1986)
et aussi divers chamignons utilises en lutte biologique (Lewis et. Papavizas, 1985). Nombre
de substances utili&s pour I’inunobilisation des cellules, enzymes, anticorps et autres
protéines (Scott, 1987) ont éte test&s pour preparer des inoculums polymériques :
polyacrylamide, alginate, xanthane (Jung et al., 1982) et de nombreux autres gels na-
turels tels que le carrageenan.
Cellules mlcroblennes
incluses dans une matrice polymerique
(PEM)
On peut utiliser divers polymeres pour inclure les cellules microbiennes dans une
matrice polyémerique, le produit obtenu étant désigné sous le terme d’inoculum poly-
merique PEM ou plus brièvement inoculum PEM (Polymer-entrapped microbial ceps).
Dans le cas de I’alginate de sodium (par ex. Alginate S 170 fabrique par SATIA, 15
avenue d’Eylau, 75116 Paris, France), qui est un polymère t&s largement employé, on
prepare une solution B 5% (poids/volume) de ce compose dans l’eau distillée et on mélange
cette solution avec le même vohnne de la culture microbienne à la fm de sa phase active
de croissance. Gn laisse tomber goutte Li goutte le melange constitue par la solution d’al-
ginate et la culture microbienne dans une solution O,l-0,2 M de CaCl, sous agitation
magnétique. En entrant en contact avec la solution de CaCI-. chaque gmtte forme une
bille sphérique indépendante (diamètre 14 mm) résultant du pontage de I’afginate par
les ions Ca2+ (Fig. 2). On maintient les billes dans la solution de CaCl, pendant 45 mn,
puis on les récolte et on les lave plusieurs fois dans l’eau st&ile ; finalement on les
sèche pendant 24 h sous hotte a flux laminaire (Fig. 3). On peut broyer les billes s&ches
pour en f%re une poudre de 50-200 prn. L’inoculum ainsi fabriqué est stocke dans des
récipients fermes. Dans le cas du rhizobium, chaque g d’inocuhrm-PEM séché renferme
10g a 101r unités formant des colonies (CFU).
i
160
160
Figure 2 : Inoculum PEh4 : billes d’alginate fraîches (non séchées) où sont inclus des rhizobiums
(diamètre : envinm 3 mm)
Figure 3 : Inoculm PEh4 : billes d’alginate séchées où sont inclus des frankias (longueur : environ
1.5mm)
161
Dans le cas des champignons VAM, le materie microbien inclus dans la matrice
polymérique n’est pas seulement constitué de structures fongiques (spores, hyphes, vé-
sicules) mais aussi de fragments de racines infect& par le champignon (Fig. 4). Toutes
ces structures doivent être soigneusement lavees avant d’être mélangées avec la solu-
tion d’alginate. Les demiks phases de la préparation de ces inoculums sont les mêmes
que celles concernant la préparation des inoculums bactériens (Jung er al., 1981 ; Gamy
et al., 1982).
Figure 4 : Inoculum PEN : billes d’alginate fraîches (non séchées) où sont inclues des structures
de champignon VAh4 et des fragments de racine (diamètre : environ 3 mm)
L’iuoculum PEM est appliqué aux plants (généralement sur les graines ou dans
le sol de la raie de semis) soit sous forme skchke, soit sous forme d’une pseudo-solution
obtenue en mettant l’inoculnm seché dans une solution tampon (0,03 M KH,PO, ;
0,17 M K.$PO, ; pH 7,4) pendant 4-8 h.
L’inoculum sec peut être faxé sur les graines avec une substance adhtsive (Elegba
et Rennie, 1984), telle que la gomme arabique (40% poids/volume). On peut ainsi fixer
entre 1 et 2 mg d’inoculum PEM sec sur une graine de soja, soit plus de 106 à
2 x 10” CFU par graine.
Les principaux avantages des inoculums PEM à base d’alginate (avantages pro-
bablement partages avec les autres inoculums PEM) sont les suivants : (i) leur faible
poids qui les rend faciles à manipuler et à transporter de l’unité de fermentation au site
d’application ; (ii) ils peuvent facilement être melanges les uns aux autres ou avec dif-
ferents produits chimiques tels que des pesticides (Dreyfus et QI., 1983) ou des subs-
tances protectrices ; (iii) apres un skchage convenable de l’inoculum, de nombreuses
espkes de microorganismes peuvent survivre longtemps : On peut conserver Frankia et
les champignons VAM pendant 2-3 ans à 20-25 C sans qu’ils perdent leur pouvoir infectif.
161
Dans le cas des champignons VAM, le matkiel microbien inclus dans la matrice
polymérique n’est pas seulement constitué de structures fongiques (spores, hyphes, ve-
sicules) mais aussi de fragments de racines infectés par le champignon (Fig. 4). Toutes
ces structures doivent être soigneusement lavées avant d’être mélang&es avec la solu-
tion d’alginate. Les dernières phases de la ptiparation de ces inoculums sont les mêmes
que celles concernant la préparation des inoculums bactériens (Jung et al., 1981 ; Ganry
et al., 1982).
L’inoculum PEM est appliqué aux plants (généralement sur les graines ou dans
le sol de la raie de semis) soit sous forme séchée, soit sous forme d’une pseudo-solution
obtenue en mettant l’inoculum séché dans une solution tampon (0,03 M KH,PO, ;
0,17 M K$PO, ; pH 7,4) pendant 4-8 h.
L’inoculum sec peut être fixé sur les graines avec une substance adhésive (Elegba
et Rennie, 1984), telle que la gomme arabique (40% poids/volume). On peut ainsi fixer
entre 1 et 2 mg d’inoculum PEM sec sur une graine de soja, soit plus de 106 à
2 x 108 CFU par graine.
Les principaux avantages des inoculums PEM à base d’alginate (avantages pro-
bablement partagés avec les autres inoculums PEM) sont les suivants : (i) leur faible
poids qui les rend faciles à manipuler et à transporter de l’unité de fermentation au site
d’application ; (ii) ils peuvent facilement être mClangés les uns aux autres ou avec dif-
férents produits chimiques tels que des pesticides (Dreyfus et al., 1983) ou des subs-
tances protectrices ; (iii) après un séchage convenable de l’inoculum, de nombreuses
espèces de microorganismes peuvent survivre longtemps : On peut conserver Frankia et
les champignons VAM pendant 2-3 ans à 20-25°C sans qu’ils perdent leur pouvoir infectif.
162
De même il est clair que certains champignons ectomycorhixiens béneficient gran-
dement de leur inclusion dans l’alginate (Jung et al., 1981 ; le Tacon et al., 1985).
La survie des rhizobiums (Jung ef al., 1982). et plus gén&akment des bactkies
ne formant pas de spores, depend de la uSrance des souches au processus d’inclusion
(g&fkation du polyr&re), de la nature des carbohydrates pr6sents dans la matrice et
de la teneur f&e en eau (en fait l’activité de l’eau : a> de l’inoculum seché. Mugnier
et Jung (1985) ont montre que les carbohydmtes à faible poids mohZculaire (C, a C$
avaient un effet défavorable sur la survie, alors que les composes de poids molkculaire
plus klev6 (C, à C,d protégaient les microorganismes. Avec les souches test& par ces
auteurs, le nombre de cellules vivantes (10’0 par g d’inwulum .sec) reste constant pendant
un stockage d’une dur& sup&ieur.e B 3 ans a 28°C quand a, de l’inoculum est maintenue
au dessous de 0,069. Dans I’iutervalle de aw 0,069-0,830, le taux de survie diminue d’autant
plus vite que a, est élevée. La tokkance des rhizobiums a la dessiccation varie en fonction
de la souche considérée et d’autres facteurs tels que la phase de croissance (Mary et
al., 1985). Lorsque a” de I’inoculum tombe au-dessous d’un certain seuil, la perte d’eau
des cellules peut devenir ir&versible de sorte que les bactkies incluses meurent
Jl est ensentiel que le nombre de bactéries viables dans l’inoculum se maintienne
à un niveau Hevt! entre le moment de la production et celui de l’application. Une bonne
conservation de la viabiité pendant le stockage permet au fabricant de pnZparer les mo-
culums bien avant qu’ils ne soient r&lam& par l’utilisateur, ce qui est tout particulie-
rement important lorsqu’on doit disposer d’un large éventail de souches.
Des recherches sont encore nt%ssaiw pour améliorer la survie des microorga-
nismes dans les inoculums PEM et PEC. Differentes approches peuvent être employées
pour assurer la survie des microorgauïsmes pendant l’inclusion, la dessiccation et le
stockage :
(i) utilisation d’un agent g&fiant approprie, par ex., dans le ‘cas de l’al-
ginate, remplacement de CaC& par le gluconate de Ca (C,$J&IaO,S
(Fravel et ai., 1985) ;
(ii) incorporation dans la matrice de substances protectrices telles que les
argiles (Jung et ul., 1982 ; Fravel et ul., 1985) ou des carbohydrates
de poids moléculaire élevé comme on l’a signalé plus haut ;
(iii) stimulation de la production in vitro de structures de rf%istance teks
que les sporanges (Fr&), ascospores et conidies (champignons) ou
sclerotes (par ex. dans le cas de Phlebopus soudanensis, (Ihoen, D.
Comm. pers.) ;
(iv) stockage à basse tempt%ature.
Colonies incluses dans une matrice polymbrlque (PEC)
Ce type nouveau d’inoculum est obtenu à partir de billes d’alginate dans lesquelles
on a laissé croître les microorganismes, qui forment donc des colonies à l’intérieur des
billes @iem et Dommergues, 1985). On désigne cet inoculum sous le terme d’inoculum
polymétrique PEC ou plus brièvement inoculum PEC (Polymer-entrapped colonies)
(Fig. 5). La préparation de ces inoculums a Cté dkrite ci-dessus. D’après les recherches
pr&ninaires en cours, les inoculums PEC renferment un plus grand nombre de CFU
que les inoculums PEM, chez lesquels il n’y a pas eu de croissance des microorganismes
*-’ f.. :. - - _... . ,._ ,.
-
__
-_
__
._._-_ _ _, _ ___ .____ .-e^.i.i-i.~:=
,.
*-. . .
.-
-.-
--.
_,
_____
163
après l’inclusion dans la matrice polymérique. Des expériences au champ sont actuel-
lement en cours pour comparer la réponse des legumineuses et des plantes actinorhizien-
nes à l’application des inoculums PEC et PEh4.
Figure 5 : Inoculum PEC : colonies de fkmkia poussant dans des billes d’alginate (diamètre : environ
3mm)
Expériences au champ
Inoculum polymérique PEM de rhizobium
En 1979, on a commencé des expériences portant sur l’utilisarion d’inoculums
polymériques PEM renfermant des rhizobiums, aussi bien pour l’application au sol dans
la raie de semis que pour l’enrobage des graines. Une de ces expériences a été mise
en place à Guérina au Sud du Sénégal, dans un sol sableux déficient en N (0,026% N)
pendant la saison des pluies, c’est-à-dire de Juillet à Octobre, 1980. L’objectif de cette
expérience était de comparer la réponse du soja A l’inoculation avec trois types d’ino-
culums, deux inoculums polymériques (PEM-alginate et PEM-xanthane) et un inoculum
liquide (culture liquide de rhizobium). Le sol utilisé n’avait pas été cultivé en soja aupa-
ravant de sorte qu’il ne renfermait pas le Bradyrhizobium spécifique du soja. La fertili-
sation avait consiste dans l’apport de 30 kg P (superphosphate) et de 100 kg de K (KCl)
par ha.
On a comparé cinq traitements : pas d’inoculation (témoin) ; inoculation des graines
enrobees avec la poudre d’inoculum PEM-alginate ; inoculation des graines enrobées avec
la poudre d’inoculum PEM-xanthane ; inoculation du sol avec une culture liquide de Rhizo-
bium appliquée à une dose très élevée ; application d’engrais azoté à la dose de
300 kg N par ha. Le dispositif expérimental comprenait quatre blocs randomisés constitués
de parcelles élémentaires de 8 m x 8 m. Le Tableau 1 montre que les inoculums PEM
163
après l’inclusion dans la matrice polymerique. Des expériences au champ sont actuel-
lement en cours pour comparer la r@onse des légumineuses et des plantes actinorhiiien-
nes à l’application des inoculums PEC et PEM.
Exptkiences au champ
Inoculum polym&ique PEM de rhizobium
En 1979, on a commencé des expériences portant sur l’utilisation d’inoculums
polyrnériques PEM renfermant des rhizobiums, aussi bien pour l’application au sol dans
la raie de semis que pour l’enrobage des graines. Une de ces exp&iences a Cté mise
en place a Gu&i~ au Sud du Sénégal, dans un sol sableux déficient en N (0,026% N)
pendant la saison des pluies, c’est-a-dire de Juillet à Octobre, 1980. L’objectif de cette
exp&ience etait de comparer la @onse du soja à I’inocuhtion avec trois types d’ino-
culums, deux inoculums polymériques (PEM-alginate et PEM-xanthane) et un inoculum
- liquide (culture liquide de rhizobium). Le sol utilise n’avait pas eté cultivé en soja aupa-
ravant de sorte qu’il ne renfermait pas le Brudyrhizobium spécifique du soja. La fertili-
sation avait consiste dans l’apport de 30 kg P (superphosphate) et de 100 kg de K (KCI)
par ha.
On a compare cinq traitements : pas d’inoculation (témoin) ; inoculation des graines
enrobees avec la poudre d’inoculum PEM-alginate ; inoculation des graines enrobees avec
la poudre d’inoculum PEM-xanthane ; inoculation du sol avec une culture liquide de Rhizo-
bium appliquee à une dose très élevée ; application d’engrais azoté à la dose de
300 kg N par ha Le dispositif exp&imentaJ comprenait quatre blocs randomisés constitues
de parcelles élémentaires de 8 m x 8 m. Le Tableau 1 montre que les inoculums PEM
164
sont aussi efficaces que l’inoculum liquide appliqué directement 3 trks forte dose. En
ce qui concerne la nodulation, I’inoculum PEM-xanthane est apparu supbrieur & I’ino-
culum PEM-alginate ; mais pour les rendements en graines, il n’y a pas eu de diffhence
significative entre ces deux inoculums (Jung et aI., 1982).
On a utilisé ultkrieurement avec succès les inoculums PEM dans le cas des Iégu-
mineuses à nodules de tige, plus particuli&ment celui de Sesbuniu rosfrafa (Dreyfus
et al., 1983).
Tableau 1 : Réponse du soja à l'inoculation des graines avec l'inoculum PEM-
alginate et l'inoculum PEM-xanthane et à l'inoculation du sol avec
une culture liquide. Expérience effectuée à Guérina, Sénégal
(Jung et al., 1982).
Traitements
Rendement en grain (kg/ha)
Poids sec des
nodules
(g/plantel
Poids sec
N total
Pas d'inoculation
1919 c
107 b
0,08 c
'
Inoculation des graines
r i
PEM-alginate
2752 ab
190-a
1,6b
"'
PEM-xanthane
2773 a
176 a
5,0 a .
~
Inoculation du sol
culture liquide
2445 ab
16G a
2,2 b
Engrais azotë
2366 bc
130 ab
0,02 d
(300 kg N/hal
Les chiffres suivisdes mémes lettres ne diffërent pas significativement pour
P = 0,Ol.
165
Inoculum polym&ique PEM de Frankia
On a mis en place au Sénégal une experimentation sur une échelle relativement
grande pour évaluer la réponse de Casuarina equisetifolia 2 l’inoculation avec un ino-
culum PEM de Fr&.
Le 15 mai 1984,‘on a plante des semis âgés d’un mois dans des gaines de polyé-
tbylene remplies d’un sol tr& pauvre (OP% C ; 0,026% N) st&ilisé au pliable au bromure
de méthyle. La moiti6 des semis a été inoculee ; l’autre moitié @moins) ne l’a pas IX
L’inoculum PEM de Frankiu a été fabriqué avec la souche de Frunkzb ORS 021001 (Diem
et al., 1983) incluse dans des billes d’alginate prepardes comme il est indiqué ci-dessus,
ces billes ayant été séchées a l’air et conservées pendant 2 ans a la temp&ature du labo-
ratoire (20-25°C) avant d’être utilisées. On a procédé à I’inoculation en introduisant
dans chaque gaine une suspension de Frankia dans un tampon phosphate pendant
4-8 h (15 g d’inoculum PEM-alginate sec dans 1,5 1 de tampon) et en homogén&ant
la gelée liquide ainsi obtenue. Pendant tout leur séjour en serre, chacun des semis de
Cusuarinu equisetifolia a rqu une fois par semaine 100 ml d’une solution Hoagland
sans azote diluée au 1/4. Le 23 juillet 1984, lorsque les plantes avaient environ 20 cm
de haut, elles ont ti mises en place a Notto, à 80 km au Nord de Dakar et a 5 km
de la Côte. Le sol qui était un sol dunaire sableux très pauvre (0,047-0,121% N), n’avait
jamais eté plante auparavant en Casuarina et ne contenait pas de Fmnkia. La planta-
tion a été faite apres que les pluies eussent mouille le sol jusqu’à une profondeur de
30-40 cm. Le dispositif exp&imental comportait huit parcelles elémentaires de 416 m2
(quatre temoins, quatre inoculées) avec 84 plants par parcelle, ce qui correspond à environ
2 000 plants par ha. Une fertilisation phosphatde a ti appliquée aux doses suivantes ;
premiere année, 5 g K.$PO, ,* deuxieme année 10 g ; troisieme année 15 g. Les pluies
ont et& respectivement de 215, 375 et 473 mm en 1985, 1986 et 1987. Au cours de
la Premiere année, on a apporte 200 1 d’eau .a chaque plant, mais cette irrigation n’a
pas Cté répétée au cours des deux années suivantes.
Pendant trois ann&s consécutives, soit en Juillet 1985, Juillet 1986 et Juillet 1987,
dans chacune des huit parcelles, on a arraché cinq arbres repérés au hasard. Apres les
avoir mesures et pesés, on a divise chacun des arbres en quatre fractions : rameaux assimi-
lateurs, branches,’ racines et nodules. Chaque fraction a éte séchée et broyee. Après
echantillonnage, on a déterminb N total. Le Tableau 2 montre que la reportse de Cusuarina
equisetifoliu à l’application d’inoculum PEM de Frankia est significative (P < 0,05) ou
très significative (P c 0,Ol) sauf en ce qui concerne le paramètre hauteur. L’effet favorable
de l’inoculum a porte sur la biomasse des différentes fractions des arbres exprimée en
poids sec ou N total. Le rapport parues a&iennes/racines calculé sur la base du poids
sec ou de N total a tte également favorablement modifié par l’inoculation (Sougoufara
et al., 1989).
Les résultats indiquent clairement que l’inoculum PEM de Fnmkia est un ino-
culum efficace et qu’il peut être stocke pendant 2 ans sans perdre ni son inefectivité
ni son effectivité.
Inoculum polymérique (PEM) de champignon VAM
En 1982, Ganry et al. ont mis en place au Sénegal une exp&ience destinée a
evaluer la rt$onse du soja à l’application d’un inoculum PEM de Glomus mosseue fait
de spores, mycelium et racines mycorhiennes homog&&&es, ces différentes structures
166
Tableau 2 : Réponse de Casuarina equisetifolia à l'inoculation avec l'inoculum
PEH-alginate de frankia 1, 2 et 3 ans après la mise en place au
champ. Expérience effectuee à Notto. Sénégal (Sougoufara et al., 1989).
Poids sec tg/arbre)
N total (g/arbre)
Rapport(l)
(3)
P.A./Rac.
basé sur
br.
ras.
rat.
nod.
br.
ras.
roc.
nod.
PS N
-~ -
-
-
-
-
-
-
-
-
Juillet 1985 (1 an après la transplantation au champ)
0
151
163
89
266
0
0.63
1.46
1.24
0
0.95
1,69
+
182
324
119
309
2.55
1.99
2.17
1.60
0.04
1.42
2,54
*
*t
**
*
**
**
*
*
**
Juillet
986 (2 ans aprës la transplantation au champ)
0
291
890
628
507
0
5,64 -9,03
2,52
0
2,99'
5,82
+
313
1349
871
626
12.18
9.39
16.24
3.21
0.20
3,48
7.5;
NS
*t
**
**
**
**
**
* *
Juillet
987 (3 ans après la transplantation au champ) (5)
0
500
1451
1000
840
0
7,32
15,57
4.74
0
2,91
4,85
+
575
2220
1356
1031
22.75
12,20
23.35
5,12
0.40
3,39
6.47
NS
**
**
**
**
**
NS
**
*
Seuil de signification : ** P = 0, 01 ; * P = 0, 05 ; NS = non significatif
(1) 1. = inoculation. 0 : pas d'inoculation (témoins) ; + : inoculation avec l'ino-
culum PEM-alginate de la souche de frankia ORS021001.
(2) Hauteur moyenne des arbres en cm.
(3) br. = branches y compris les troncs ; ras = rameaux assimilateurs ; rat. = racines;
nod. = nodules.
(4) Rapport "Parties aëriennes/Racines" exprimé en poids sec (PS) ou N total (NI.
(5) Au cours de la 3ëme année des attaques de borers ont très sensiblement ralenti
la croissance des arbres.
167
ayant été incluses dans des billes d’alginate. Glomus mosseae avait été au préalable mul-
tiplie sur ni&& (Vigne unguicdafu). Le champ d’expérience, situe à la station ISRA (IIE-
titut Sertégalais de Recherches Agricoles) de Séfa, au Sud du Sénégal, avait été choisi
en raison du faible potentiel infectieux de son sol en ce qui concerne les champignons
VAM. Ce sol deficient en N (environ 0.04% N) n’avait jamais éd cultivé auparavant
en soja Le dispositif exp&imental utilise était du type split-plot avec huit répétitions.
Les trois traitements principaux (chaque parcelle principale avait 40 rnq ont été les sui-
vants : T (témoin), inoculation avec une souche ineffective de Bradyrhiwbium sur support
tourbe ; B, inoculation avec une souche effective de Bradyrihiwbium sur support tourbe ;
BG, inoculation avec la souche effective de Bradyrhizobiwn sur support tourbe (comme
dans le traitement B) et inoculation avec inoculum PEM-alginate de Glomus mosseae
(conserve à la chambre froide pendant plus d’un mois) appliqué a la dose de 15-20 billes
fraîches (non s&h&s) à 34 cm de profondeur autour de chaque semis de soja, lorsque
ceux-ci avaient 15 jours. Les deux sous-traitements ont été les suivants : OP, pas d’appli-
cation de fertilisation phosphatée ; P, application de superphosphate à la dose de 20 kg
par ha. L’azote marqué (‘-?Y) a été appliqué sur des microparcelles de 1.65m2 pour
mesurer la fixation de N, par la methode de la valeur A. On a analyse séparément
feuilles, tiges, gousses et graines. Par souci de simplification, les donnees concernant les
diff&eme.s parties a&iennes des plantes, à l’exception des graines ont été regroupées sous
le terme de «parties aeriennew. Les chutes de pluie ont Cté de 170 mm avant le semis
(ler Mai - 17 Juillet) et de 692 mm pendant le cycle cultural (17 Juillet - 10 Octobre).
La distribution des pluies a été assez regulière, sans période sèche marquée.
Le Tableau 3 montre clairement que l’inoculum PEM-alginate de Glomus mosseae
peut être utilise avec suceb au champ, en tant que traitement complementaire de l’ino-
culum de Bradyrhizobium, ce qui confirme les experiences publiees ant&ieurement par
Ganry et 42.. (1982).
Quand on a appliqué une fertilisation phosphatée legère (20 kg P/ha) I’inoculum
PEM-alginate de Glomus mosseae a augmenté la fixation de N, de 109 (inoculation avec
Bradyrhizobium seul) B 139 kg N$a, ce qui constitue une amélioration substantielle.
Dans les parcelles ayant reçu la fertilisation phosphatée, l’inoculum PEM-alginate de
Glomus mosseae a augmente significativement le rendement en grain exprime en poids
sec, et N total, et en outre améliore l’indice de récolte. Ces r&ultat.s positifs revêtent
une grande importance pour le fermier.
Conclusion
II est maintenant clair que l’on peut ameliorer considerablement la réponse des
plantes à l’inoculation en faisant appel à la technologie fondée sur l’utilisation des
microorganismes inclus dans une matrice polym&ique.
Etant donne que la production de ces inoculums perfectionnes exige un certain
niveau d’expertise, les usines de fabrication d’inoculums polymeriques devraient étre ins-
tall&s clans un nombre limite de centres d’excellence. En outre, les exp&iences au
champ, qui sont capitales pour toutes les souches, devraient être mises en place dans
les diff&-entes regions écologiques où l’inoculation est reconnue indispensable. Ceci est
particulièrement vrai pour les arbres fixateurs de N,
On peut penser que le developpement de l’utilisation d’inoculum de haute qualité
n’accroîtra pas seulement la fixation de N2 et la croissance des plantes, mais aussi ren-
168
Tableau 3 : Réponse du soja à l'inoculation avec une souche non effective
(non fixatrice de N2) de Bradyrhizobium (traitement T), d'une
souche effective de Bradyrhizobium seule (traitement BI OU
associée à Glomus mosseae, ce dernier inoculum étant un ino-
culum PEM-alginate (traitement BG). Dans la moitié des parcelles
on n'a pas apporté de P (traitement PO), dans l'autre moitié
on n‘a apporté P sous forme de superph0sphate.à la dose de
22 kg P/ha (traitement Super). Expérience effectuée à Séfa,
Sénégal (Ganry et al., 1985)
'Traitement
Rendement en grain
Grains et parties
Indice de
N2 fixé
(kg/ha)
aériennes (kg/ha)
récolte
sur la base (kg'ha'
PS
N
N.
P
de N total
T OP
Super
1093
1725
65,7
101,6
84,2
127,2
1;::
3,55
3,98
0
B OP
1423
90,3
4,18
73,1
B Super
2017
133,8
4,71
109,o
.
BG OP
1431
98,2
120,o
699
4,42
80,2
BG Super
2290
154,7
183,7
11,8
5,33
139.3
PPDS entre les traitements'
T,
B et BG
192
12,8
OP et Super
197
13,5
169
forcera leur rt%iitance aux parasites et pathogenes, ce qui permettra une diminution des
besoins en engrais et pesticides, d’où une r&htction des sources de pollution.
Actuellement, on met au point en France la fabrication à l’echelle industrielle
d’inoculums PEM et PEC pour les rhizobiums, Frunkziu et differents microorganismes
impliques dans la lutte biologique. Cette technologie devra être adaptée aux pays en voie
de développement dans un proche avenir.
Remerciements
Ce travail a &6 financé en partie par le contrat ANVAR no A 87 03 133 Q
AL 1340. Nous sommes reconnaissants au Professeur J.P. Belaich des conseils qu’il
nous a don& -pour le lancemem de la culture continue de Frankia.
AZCON, R., BAREA, J.M. and HAYMAN, D.S. 1976 Utilization of rock phosphate in
alcaline soils by plants inoculated with mycorhizal fungi and phosphatesolubiliziig
bacteria. Soi1 Biol. Biochem. 8, 135-138.
BAREA, J.M. NAVARRO, E. and MONTAYA,iE; t1976 Production of plant growth
regulators by rhizosphere phosphate-solubilizing bacteria. J. Appl. Bacteriol. 40,
129-134.
BASHAN, Y. 1986 Alginate beads as synthetic inoculant carriers for slow release of
bacteria that affect plant growth. Appl. Environ. Microbiol. 51, 1089-1098.
BAUDET, C., BARBOTIN, JN. and GUESPIN-MICHE!L, J. 1983 Growth and sporu-
lation of entrapped Bacilles subtilis cdis. Appl. Environ. Microbiol. 45, 297-301.
BODDEY, R.M. and DGBEREJNER, J. 1988 Nitrogen fmtion associated with grasses
and cemals : Recent results and perspectives for future research. Plant and Soil. 108,
53-65.
BROWN, M.E. 1982 Nitrogen fixation by free-living bacteria associated with plants -
fact or fiction ? In : Rhodes-Roberts M. and Skinners F.A. (eds.) Bacteria and plants,
pp. 2541. Academic Press, London.
BURTON, J. 1979 New developments in inoculating legumes. In : Subba Rao NS (ed.)
Recent advances in Biological nitrogen furation, pp. 380-405. Oxford and IBH Pu-
blishers, N e w D e h l i .
CARLSON, P.S. 1988 Microorganismes endophytiques : un syst&me r6volutio~ de
protection biologique des plantes. In : Abstracts «Advanced technology and plant bree-
ding strategy. Potential applications to woody species» meeting ,organized by the
Mo& Hennesy. Louis Vuittorm group, 19-21 Sept. 1988, Versailles.
CHAO, WL. and ALEXANDER, M. 1984 Mineral soils as carriers for Rhîzobium ino-
culant.~. Appl. Environ. Microbiol. 47, 94-97.
DATE, R.A. 1977 The development and use of legume incculants. In : Ayanaba, A.,
Dart, P.S. and Gibson, A.H., (cds.) Biological nitrogen fixation in farming systems
of the tropics, pp. 243-276 John Wiley, New York
170
DIEM, H.G. and DOMMERGUES, YR., 1985 Procede de culture de microorganismes,
notamment du groupe des Frankia et pr@ration d’inocnlums bacteriens. Demande
de brevet no 85-17933 du 04-12-85 (Method for cultivating microorganisms parti-
cularly of the Frankia group and preparation of bacterial inoculums. Patent appli-
cation PCT/FR86/00417 December 4, 1986).
DIEM, H.G., GAUTHIER, D. and DOMMERGUES, YR. 1983 An effective strain of
Frankia from Casuarina sp. Can.J. Bot. 61, 28152821.
DIEM, H.G., DUHOUX, E., SIMONET, P. and DOMMERGUES, YR. 1988 Actinorhizal
symbiosis biotechnology : the present aud the future. In : Durand G., Bobichon L.,
and Florent J. (eds.) 8th International Biotechnology Symposium Paris, 1988. Procee-
dings, Vol. II, pp. 984-995. Soc%% Française de Microbiologie, Paris.
DOMMERGUES, YR., DIEM, H.G. and DIVIES, C. 1977 procédé microbiologique
destine à maîtriser la productivit& des plantes cultivées. Demande de brevet no 77-
10254 du 05-04-77 (Microbiological process for comrolling the productivity of culti-
vated plants, US patent no 4 155-737, May 22, 1979).
DOMMERGUES, Y.R., DIEM, H.G. and DIVIES, C. 1979 Polyacrylamide-entrapped Rhi-
zobium as an inoculant for legumes. Appl. Environ. Mcrobiol. 37, 779-781.
DREYFUS, B., RINAUDO, G. and DOMMERGUES, YR. 1983 Pro&& d’inoculation
des plantes, notamment des I&umineuses à nodules caulinaires. Demande de brevet
no 83 02847 du 22.02.83.
ELEGBA, M.S. and RENNIE, R.J. 1984 Effect of different inoculant adhesive agents
on rhizobial survival, nodulation, ,and nitrogenase (acetylene reducing) activity of
soybeau (Glycine max (L.) (h!terill.). Can. J. Soi1 Sci. 64, 631-636.
ELMES, RP. and MOSSE, B. 1984 Vesicular-aruscular endomycorrhizal inoculum pro-
duction. II. Experiments with maize and other hosts in nutrient flow culture. Can.
J. Bot. 62, 1531-1536.
ELMES, R.P., HEPPER, C.M., HAYMAN, D.S. and O’SHEA, J. 1983 The use of vesi-
cuhrr-arbuscular mycorrhizal roots grown by the nutrient film technique as inoculum
for field sites. Ann. Appl. Biol. 104, 437-441.
FRAVEL, D.R., MAROIS, JJ., LUMSDEN, RD. and CONNICK, Jr. WJ. 1985 En-
capsulation of potential biocontrol agents in a alginak-clay mauix. Phytopathology.
75, 774-777.
GAGNON, J., LANGLOIS, C.G. and FORTIN, J.A 1987 Growth of containerized jack
pine seedhngs inoculated with different ectomycorrhizal fungi under a controlled fer-
tiliiation schedule. Cari. J. For. Research. 17, 840-845.
GANRY, F., DIEM, H.G. and DOMMERGUES, YR. 1982 Effect of inoculation with
Glomus mosseae on nitrogen fixation by field-grown soybean. Plant and Soil. 69,
321-329.
GANRY, F., DIEM, H-G., WEY, J. and DOMMERGUES, Y.R. 1985 Inoculation with
Glomus mosseae improves N, fixation by field-grown soybean. Biol. Fert. Soifs. 1,
15-23.
GIANINAZZI-PEARSON. V. and GIANINAZZI, S. (eds) 1986 Physiological and gene-
tical aspects of mycorhizae. INRA, Paris.
.
..i
171
HALVERSON, LJ. and STACEY, G. 1986 Signal exchange in plant-microbe interactions.
Microbiol. Reviews, June, 193-225.
HUNG, L.L. and SYLVIA, D.M. 1988 Production of vesicuhu-arbuscular mycorrhizal
fungus inoculurn in aeroponic culture. AppJ. Environ. Microkiol. 54, 353-357.
JUNG. G., MUGNIER, J., DOMMERGUES, Y.R. and DIEM, H.G. 1981 ProcedC d’inclusion
de microorganismes du groupe des mycorhizes et des actinorhizes. Demande de brevet
no 81-04474 du 06-03-81 (Process for inclusion of mycorrhizae and actinorhizae in
a matrix, US patent pending).
JUNG, G., MUGNIER, J., DIEM. H.G. and DOMMERGUES, Y .R. 1982 Polymer-entrapped
Rhiwbium as an inoculant for Iegumes. Plant and Soi1 65, 219-231.
LE TACON, F. and GARBAYE, J. 1986 La maîtrise des associations mycorhiziennes
en pépinière forestière. Rev. For. Fr. 38, 249-257.
LE TACON, F., JUNG, G.. MUGNIER, J., MICHELOT, P. and MAUPERlN, C. 1985
Efficiency in a forest nursery of an ectomycorhizal fungus inoculum produced in a
fermentor and entrapped in polyméric gels. Cun. J. Bot. 63, 1664-1668.
LEWIS, J.A. and PAPAVIZAS, G.C., 1985 Characteristics of alginate pellets formulated
with Trichoderma and Gliocladium on their effect on the proliferation of the fungi
in the soiL Plant pathology, 34, 571-577.
LIANG. ZICHAO 1986 Vegetative propagation and selection on Casuarina for resistance
to bacterial wilt. Tropical Forestry (Science and Technology) Guangzhou. 2, l-6 (In
chinesse, summary in english).
MARY, P., OCHIN, D. and TAILLEZ, R. 1985 Rates of drying and survival of Rhizobium
meliloti strains during storage at different relative humidities. App. Environ. Micro-
biol. 50, 207-211.
MARX, D.H. and KENNEY, D.S. 1982 Production of ectomycorhizal fungi inoculants.
In : Schenck N.C. (ed.) Methods and principles of mycorrhizal research, pp. 131-
146. The American phytopathological Society, Saint-Paul, Min.
MENGE, J.A. 1983 Utihzation of vesicular-arbuscular mycorhizal fungi in agriculture.
Can. J. Bot. 61, 1015-1024.
MOSSE, B. and THOMPSON, JP. 1984 Vesicular-arbusctdar endomycorrhizal inoculum
production. 1. Exploratory experiments with bean (Phaseolzu vulgaris) in nutrient flow
culture. Can. J. Bot. 62, 1523-1530.
MUGNIER, J. and MOSSE, B. 1987 Vesicular-arbuscular mycorrhizal infection in trans-
formed mot-inducing T-DNA roots grown axenically. Phytopathology. 77, 1045-1050.
MUGNIER, J. and JUNG, G. 1985 Survival of bacteria and fungi in relation to water
activity and the solvant properties of water in biopolymer gels. Appf. Environ. Microbiol.
5 0 , 1 0 8 - 1 1 4 .
NAYAK. D.N., LADHA, J.K. and WATANABE, 1. 1986 The fate of marker Azospirillum
lipoferum inoculated into rice and its effect on growth, yield and N, fixation of plants
studied by acetylene reduction, “N feeding and “N dilution techniques. Biol. Fert.
Soils. 2, 7-14.
OKON, Y. 1984 Response of cereal and forage grasses to inoculation witb N,-fixing
bacteira. In : Veeger C. and Newton W.E. (eds.) Advances in nitrogen fixation
research, pp. 303-309. NijhoffBung, the Hague.
172
OKON, Y. and HADAR, Y.A. 1987 Microbial inoculants as trop-yield enhancers. CRC
Critical Reviews in Biotechnology. 6, 61-85.
PHILLIPS, D.A., JOSEPH, C.M. and MAXWELL, C.A. 1988 Flavonoid nodulation signais
in alfalfa. In : Abstracts 7th Congress on nitrogen fixation, 13-20 March 1988, Cologne.
PRIN, Y. and ROUGlER, M. 1987 Preinfection events in the establishment of Alnus-
Frankt% symbiosis : study of the root hair deformation step. Plant Physiol (Life Sci.
Adv.). 6, 99-106.
PRIN, Y., NEYRA, M. and DIEM, H.G. 1989 Evaluation of Frankia growth and viavility
using Bradford protein and INT reduction tests. Plant and Soi1 (in press).
REDMOND, J.W., BATLEY, M., DJORDJEVIC, M.A., lNNES, R.W., KUEMPEL, P.L.
and ROLFE, B.G. 1986 Flavones induce expression of nodulation genes in Rhizobium.
Nature. 323, 632-635.
ROLFE, B.R., REDMOND, J.W., BATLEY, M. and DJORDEJEVIC, M.A. 1987 Nodu-
lation inducing factors. European patent application no 87302621.5, date of
filing 26.03.87.
SCO’IT, C.D. 1987 Immobihzed cells : a review literature. Enzyme Microb. Technol. 9,
66-73.
SOUGOUFARA, B., DIEM, H.G. and DOMMERGUES, Y.R. 1989 Response of field-
grown Casuarina equisetifolia to inoculation with Frankia str’ain ORS 021001 entrap
ped in alginate beads. Plant and Soil (in press).
WIJFFELMAN, C.A., SPAINK, H.P., OKKER, R.J.K., VAN BRUSSEL, A.A.N., ZAAT,
S.A.T. and LUGTENBERG, EJ.J. 1987 Process for activating rhizobium nodulation
promoters European patent application no 87200348.8. date of filing 26.02.87.
WILLIAMS, P.M. 1984 Current use of legume inoculant technology. In : Alexander,
M. (ed.) BiologicaI nitrogen fixation, pp. 173-200. Plenum Press, New York.
ECONOMIE DE L’AZOTE
DANS LES SYSTEMES
DE CULTURE EN ASSOCIXTION
ET EN AGROFORESTERIE
h4aximisez Ia FBA pour la Production Agricole et Forest&e en Afrique
Biological nitrogen fixation
and nitrogen transfer in multiple
cropping in tropical Africa
MIJLONGOY, K. and EHUI, S.K.
International Institute of Tropical Agriculture,
PMB 5320, Ibadan, Nigeria
Abstract
Traditional farming systems throughout the humid and subhumid tropics are cha-
racterized by multiple cropping, the association of crops .in time and space on a
piece of land during one calendar year. These systems usually involve leguminous
crops and trees. Legumes can harness atmospheric nitrogen (NJ into the systems
and benefrt non-leguminous crops through the transfer of nitrogen (N) in the root
systems anddecomposition of their residues. Biologicalmechanisms such as competition
for light, water and nutrients that determine the yield levels for the individual crops
can affect the proee~~es involved in N, fiiation. Legumes can improve soil N status
with resulting N and grain yield benefits to the succeeding nonleguminous crops.
Even in the absence of net N benefit, intercropped nonlegumes have been shown
to derive important fractions of their N from N, fmed by leguminous component(s).
Only few studies have been published on economic profits of using legumes in cropping
systems in the tropical Africa. Although they show interesting returns relative to
the use of chemical fertilizers, they usually do not consider risk factors. A quadratic
programming approach is thus proposed as a more appropriate method to appraise
the economic potentials of legumes in multiple cropping in sub-saharan Africa.
175
Introduction
Traditional agricultnre in tropical Africa is based on shifting cultivation and related
bush fallow systems. The restorative success of these systems depends largely on the
litter cover from the fallow vegetation and on the nutrient recycling function of the deep-
rooted trees and shrubs (Sanchez,, 1976). In recent years, however, due to increasing po-
pulations and/or the scarcity of labor in areas with low population densities, fallow periods
have been systematically reduced or suppressed, leading to more intensive cultivation of
marginal lands. Continuous cultivation of the same fields has resulted in soil erosion,
increased weed populations, and decreasing soil fertility and crop yields (Rang et al.,
1985 ; Lal, 1987).
Earlier research approaches towards alleviating the food deficit problem in tropical
Africa were based on the Green Revolution which was associated with high-yielding va-
rieties, sole cropping, monocuhure and enormous inputs of non-renewable resources to
extend the cropping season, increase soil productivity and control pests. The failure of
the Green Revolution, some fifteen years ago, led udevelopers, to revise their approach
for the promotion of food production in the tropics. They began to emphasize the needs
for crop production stability, resource conservation, reduction of external inputs and more
complete use of natural resources. This tendency has stimulated a lot of research effort
in the area of multiple cropping aiming at defining plant populations of the component
crops, planting patterns, plant structure, relative duration of the life cycles of the component
crops, and fertihzer managements that would result in combined yield advantages (Okigbo
and Greenland, 1976).
Legumes are usually included in multiple cropping systems for three basic reasons.
Low-growing legumes can be integrated into the architecture of tall cereals without exces-
sive reduction in yields of component crops, but with significant increases in protein and
crop yields per hectare (Norman, 1982). Legume-cereal mixtures are advantageous because
legumes can fm atmospheric N, symbiotically and transfer this additional N to other crops
in the system. Finally, farmers are risk averters who base their farming decisions not
on profit maximization but on the judicious choice and arrangement in time and space
of various crops to suit their microecologies and conditions and ensure crop production
with minimum risk.
There is however a dearth of information about how much N is fmed ; how this
process is affected by the various plant interactions such as shading in the cropping sys-
tem ; how leguminous plants and, in general, N,-furing systems contribute to the yield
advantages and the maintenance of soil fertility in multiple cropping. Economic evalua-
tion of Nz fixation and N transfer is also lacking partly because of the poor knowledge
of complex N transfer pmcesses in multiple cropping systems. Although, a great deal
of research work on legumes in multiple cropping has been conducted in tropical Asia,
comparatively little work has been conducted on this topic in tropical Africa.
In this paper, we briefly review the major ecological factors determining yield
levels in multiple cropping systems and their influence on biological N, fixation. We
then discuss N contribution of legumes in selected types of multiple cropping. Our dis-
cussion does not include associative N, fixation and pasture systems to simplify the presen-
tation of the paper. However, principles described for other systems are also valid for
these systems. Finally, we briefly review the economics of using legumes in multiple
cropping and consider the quadratic programming approach that includes the risk mini-
176
mixing attitude of farmers as a possible tool for economic assessment of the potential
of legumes in sub-saharau Africa.
Some definitions
Before addressing these issues, it is useful to define certain concepts. Multiple
cropping is the general term which is used to describe systems where two or more crops
are grown on the same field in a year. The system is said sequential when the crops
am grown in sequence. In this case, the intensification of crop production is only in the
time dimension. Sequential cropping is rare in traditional afiican farming systems. It charac-
terizes areas where on-farm agricultural technology is geared towards commercial pro-
duction (Andrews and Kassam, 1976).
Intercropping, the growing of two or more crops simultaneously on the same field,
is widespread in tropical Africa with the highest complexity in compound gardens of
the rainforest region where are intercropped annual staples, vegetables and perennial fruit
trees (Okigbo and Greenland, 1976). Mixed intercropping, the simultaneous growing of
two or more crops without distinct TOW arrangement, and relay intercropping, the growing
of two or more crops simultaneously during part of the life cycle of each component,
are the predominant practices in the tropics. In row intercropping, two or more crops
an3 grown simultaneously with one or more crops planted in rows. Strip intercropping
is the growing of two or more crops simultaneously in strips. Both row and strips intercrop-
ping are rare in Africa except where animals or tractors are used in cultivation (Okigbo
and Greenland, 1976 ; Steiner, 1982). In the live-mulch system, food crops are grown
in an establish field of a low-growing perennial leguminous cover crop (Akobundu, 1980).
The live mulch prevents soil degradation and erosion, suppresses weeds and provides
biologically fixed Nz
Trees such as oilpalm, coconut, cocoa are also often intercropped to a range of
annual crops in the early years of plantation (Okigbo and Greenland 1976). This mul-
tistorey cropping is considered as the ideal farmiug system in the humid tropics as it
resembles the natural vegetation. Related to the multistorey cropping system is alley crop
ping, the growing of food crops between hedgerows of woody species that are perio-
dically pruned to prevent shading of the associated crops and to make available mulch
materials and the nutrients locked up in the prunings (Rang et uL, 1981).
Multiple cropping is generally more complex in the forest regions than in the sa-
vanna areas. In the forest region, crops are more diverse, farmers’ holdings arc relatively
small and the growing season is longer (Okigbo and Greenland, 1976). Within the same
ecological zone, the diversity in component crops and their spatial arrangement increases
with soil fertility which in turn improves with the proximity to the homestead (e.g. corn-
pound gardens).
Factors determining yield levels in multiple cropping systems
The major factors determining yield levels in multiple cropping systems are the
interactions of component crops during the use of growth factors, the N,-fxing ability
of intercrops and stresses interfering with biological N,-fixation. We discuss each of these
processes subsequently.
177
Interactions of component crops
When plants are intercropped or grown in sequence, they interact through various
forms of modifications imposed on the environment. These interactions and their effects
on crop yields were reviewed by Trenbath (1976). Willey (1979a) and Palaniappan (198.3.
The interactions are said non-competitive when the associated crops share growth factors
that are present in adequate quantities, such that some positive level of output from one
crop is possrble without any reduction in the output of another crop. Competition occurs
when component crops share a limited pool of growth resources. In this case, one output
must be forgone in order to produce more of the other output.
In an intercropping situation, ‘when the component species are able to exploit the
supply of growth factors in different ways in time or space, or if one species is able
to help the other in the supply of a factor, the interaction is said to be complementary.
The pmcess by which growth factors are used is referred to as annidation. Other synonyms
found in the literature include facilitation, cooperation aud compensation. Another form
of interaction observed among plants grown in intercropping is the influence of one com-
ponent crop on pests that may affect the performance of other crops in the system.
Plant growth factors that are likely to play important role in multiple cropping
are light, water and mineral nutrients. Light is absorbed by the leaves, and water and
nutrients mainly by the roots. In healthy young plants, growth is responsive to the rate
of absorption of any one of the growth factors when it is in relatively short supply (Tren-
bath, 1976). Competition for light occurs when the photosynthetic canopy of one com-
ponent is set higher than that of another. Thus taller components compete more effectively
for light. But leaf area and leaf inclination are likely to determine the magnitude of com-
petition. If shading is partial, the shaded plants can adapt to low light conditions by re-
ducing their rate of dark respiration, lowering their root/shoot ratio, or increasing the
leaf area/leaf weight ratio. These adaptative changes can be part of selection traits in
breeding genotypes suitable for multiple cropping. Intense shading will reduce growth
considerably and induce early senescence of leaves. Light is different from other growth
factors as it cannot be much influenced by man, Therefore, by maximixing light inter-
ception, multiple cropping can suppress weeds and reduce evapotranspiration losses from
the soil surface effectively.
The uptake of water, dissolved nutrients and oxygen (03 by a root surface esta-
blishes a concentration gradient in the vicinity of the root The zone of depletion of water
and mobile ions such as nitrates will expand rapidly as compared to that of P and cations
like ammonium, calcium and potassium that are strongly adsorbed onto the surfaces of
soil particles. Thus in multiple cropping particularly in intercropping, a component crop
can induce water and nutrient stress for the other crops. Experiments on competition for
0, have not clearly demonstrated the magnitude and consequences of this interaction.
Competition for carbon dioxide (CC2) has been demonstrated in sealed containers, but
it is unlikely to occur in the fields owing to the turbulence of air within canopies (l&n-
bath, 1976).
Yield advanfages of nitrogen-fixing intercrops
Nitrogen-fixing intercrops affect crop yields through annidation in space, in time
and with respect to nitrogen. Spatial differences between crops are related to shoot height,
plant architecture, and root depth and distribution. Willey and Rao (1981) demonstrated
178
the importance of spatial interactions in intercropping by comparing the performance of
17 genotypes of pigeon pea in association with sorghum. None of the genotypes affected
sorghum yields adversely but their own yields ranged between 36 and 73% of sole crop
yields. Genotypes that were compact in the early growth stage and that spread later to
fully utilize the resources after sorghum harvest minimized competition with sorghum
significantly. They were found to be most suitable for high combined yields.
Another example of annidation in space is the multistorey cropping involving
planting leguminous shade trees or low-growing cover crops such as Pueraria or Centrcy
sema in cocoa and tea plantations. Trenbath (1976) noted that where the taller intercrop-
ped component has steeply inclined leaves and the shorter component has prostrate leaves,
there is a possibility of a land equivalent ratio (LER) > 1. In alley cropping, FIemingia
congesta is one of the few nitrogen-fixing shrubs that perform well under the shade of
plantain. Its N-rich pruning can improve plantain (mtemational Institute of Tropical Agri-
culture, 1985) and maize gram (Table 1) yields . The principle of exploiting different
vertical space layers is also found in live-mulch system. However, cover crop species
currently used in the system such as Psophocarpus palumis and Centrosema pubescent
tend to climb and strangle the associated crop if they are not sprayed with growth retardants.
-
Table 1.
Effect of prunings from Cassia siamea, Flemingia congesta,
Gliricidia sepium and Leucaena leucocephala on maize' N
content and grain yield (Mulongoy, unpublished data).
Pruning N
Maize grain
Pruning
applied*
yield
Maize N
-------------------kg ha-' ------------______-__________
Cassia
168
16.7
2430
Flemingia
90
13.9
T540
Gliricidia
180
21.8
-
2870
Control
0
8.0
-
1100
1 On the basis of 40000 plants ha-'. Nitrogen content was determined at 8
weeks after planting.
2 Prunings were applied only at maize planting
179
Some degree of root stratification must be common in intercrops of dissimilar com-
ponents on deep soils (‘frenbath, 1976). But only few reports have been published on
this topic. In alley cropping, Kang et al. (1985) showed that both maize and leucaena
had distinct root feeding zones with leucaena extracting moisture and nutrients from deeper
layers. However, hedgerow trees established from cuttings may have shallow nx)t systems
(AttaXrah et al. 1985) and thus compete with the associated crops.
Two intercrops of widely different maturities will likely reach their peak demands
for growth factors at different times. This dissimilarity will reduce interference between
them. A minimum maturity difference of 30-40 days for yield advantage has been sug-
gested (Palaniappan, 1985). After the early crop is harvested, the other crop will fully
exploit the growth factors. Yield advantages in relay cropping are partly based on this
principle. Another temporal effect occurs in situations where senescent plant materials,
litter or prunings could make nutrients available to another crop. Norman (1982) notes
for instance that up to 70% of the aboveground materials of guar (Cyamopsis tetrago-
doba) returns naturally to soil as leaf fall. Charreau and Vidal (1965) also noted the
leaf fall of Acacia ulbiuir and its contribution to soil N and N uptake of cereals planted
around it.
In an intercropping system involving Njfiing and non-N,-fixing systems, the
N,-freer will obtain part of its N from biological fixation of atmospheric N,, while the
other component(s) will utiIize exclusively soil mineral N. If the non-N,f=ing component
is very competitive for soil N, a depleted soil in N will favour nodulation and nitrogen
fixation of the N,-fixer intercrop. Pigeon pea seems to nodulate better where the roots
intermingle with those of intercropped sorghum (Willey, 1979a). Ennik cited by Trenbath
(1976) reports that white clover intercropped with grass on N-poor soil gives a LER up
to 6.7 due to the interactions between the component crops. Some investigators explain
the positive effect of added legumes on LER by some transfer of the N, fixed to the
non-leguminous component (Trenbath, 1976). But more data are needed to corroborate
this explanation.
interference stresses on biological nitrogen fixation
Various stresses can impair biological N, fixation in multiple cropping systems.
Biological N, fmation is markedly dependent on current photosynthesis regulated by light
This is shown by diurnal variations in nitrogenase activity (Sloger et al., 1975) and by
shading effect on NZ fixation. Gibson (1976) reported that nitrogenase of Trifolium sub-
terraneum last 40% of its activity within five hours when the plants were transferred
from an ilhnninance of 620 ttBrn-%-i to one of 165 l.t.Brn-W. Nutman (1976) reported
also that 50% shading of soybean plants at the end of flowering decreased fixation from
125 to 91 kg N ha-‘. He concluded in his review that there is a direct relationship between
light intensity and N, fixation through the amount of photosynthate produced. Similarly,
growth and CO, uptake of intercropped azolla declines in response to a developing rice
canopy (Lumpkin and Plucknett, 1982). Photosynthesis by leaves of legumes seems satu-
rated at 30-50% of maximum sunlight (Ludlow cited by Henzell, 1981). As stated earlier,
plants exhibit various phenotypical adaptations to low light to ensure maximum light inter-
ception.
The detrimental effect of water stress on the host-plant, the rhizobia and symbiotic
effectiveness is well documented (Sprent, 1971, 1972 ; Eaglesham and Ayanaba.. 1984).
180
Fthizobial populations are reduced in size when soils become desiccated. Slow-growing
rhizobia tend to survive desiccation better than fast-growing rhizobia Soil type influ-
ences the desiccation tolerance of rhizobia. Nodulation tends to fail at low soil moisture
levels not only as a result of death of the rhizobia, but essentially because root hairs
assume a configuration incompatible with the infection process (Eagle&m and Ayanaba,
1984). The water potential levels allowing nodulation of legumes should be defined.
Sprent (1971) noted that soil moisture loss below 80% of the maximum resulted in irre-
versible damage of soybean nodules. Some legumes such as clovers can recover more
rapidly from drought than others like soybeans that shed nodules under dry conditions.
Water stress can also reduce photosynthesis by reducing leaf area and inducing leaf
fall ; this in turn will reduce symbiotic N, fixation.
It is accepted that N,-fixing plants require more P than plants supplied with mineral
N. Phosphorus deficiency will limit nodule development, nitrogenase acrivity and host-
plant growth. A number of experiments have demonstrated the important role of vesicular-
arbuscular mycorrhiza in increasing P uptake of infected plants. Also, dual inoculation
with effective strains of rhizobia and mycorrhizal fungi shows synergistic effects on no-
dulation and reduction of atmospheric N, in Pdefrcient soils. Recent studies reveal ge-
netic diversities within plant species in tolerance to low soil P (Sanginga et al.. 1990).
The effect of K on the Rhizobium - lugume symbiosis seems to be indirect and
to act through the physiology of the plant (Jardim Freire, 1984). There are however reports
showing nodulation response to K under field conditions (Ayanaba and Dart, 1977).
Danso et al. (1987) observed a reduction in dry matter yield, total N and N2 fixed
by fababeans intercropped with barley. Mulongoy (1986a) found a similar reduction in
dry matter of intercropped cowpea and maize that resulted in reduced total plant N. With
increasing plant population densities of either fababeans in the sole crop or barley in
the intercrop, Danso et al. (1987) obeserved increases in the proportion of N derived
by fababeans from fixation. Greater Nz fixation may either increase the total N yield
in the legume or just the proportion of N from fixation as compared to N derived from
mineral sources by the legume. Danso et aZ. (1987) found in their study that the presence
of barley decreased soil and fertilizer N uptake by fababeans and enhanced symbiotic
N, fixation of the legume component. The mechanisms governing these processes were
not described. However, since the reduced soil N uptake by the legume was not reflec-
ted in increased soil or total N in barley, there was therefore no evidence of N-sparing
effect by the legume for the associated non legume. Data presented by Danso et al. (1987)
do not support the hypothesis of N transfer by the legume to maize.
Mulongoy and Nkwiine (1987) observed that populations of rhizobia decreased
when maize was included in a sequence of soybean monoculture (Table 2). Whether the
reduction in rhizobia populations was due to allelopathic effects was not investigated.
Reduction in the size of rhizobial population can result in poor nodulation and nitrogen
fixation. It is important to note that poorly or not nodulated legumes can deplete soil
N just like cereals. Mulongoy and Nkwiine showed in the same experiments that crop
ping cowpea or soybean after maize increased the population of Bdyrhizobizm jupo-
nicum. It is important to note that poorly or not nodulated legumes can deplete soil N
just like cereals.
From this discussion, it is evident that various types of interferences occurring
in multiple cropping can reduce the number of rhizobia and affect nodulation and nitro-
gen fixation adversely. Few investigations have been carried out to monitor changes in
181
these factors and how they affect biological nitrogen fixation of leguminous components
in multiple cropping.
Table 2.
Effect of cropping sequences of soybean(s) and maize on the
lo910 of the numbers of soybean rhizobia per gram of soil
recorded after the last cropping season, on a loamy sand.
(Mulongoy and Nkwiine, 1986a)
Cropping sequence
Bradyrhizobium japonicum
(log10 number)
Soybean - Soybean - Soybean
3.710 a
Soybean - Soybean - Maize
3.492 b
Soybean - Maize
- Soybean
3.710 a
Soybean - Maize
- Maize
3.118 c
Nitrogen contribution of legumes in multiple cropping
Nitrogen contribution of legumes and, in general, of N,-tixing systems in multiple
cropping can occur through (1) exudation or excretion of nitrogenous substances by active
nodules and roots, (2) leaching of soluble N from living leaves and plant parts during
rainfalls, (3) litter fall and decomposition, (4) root and nodule decay, (5) decomposition
of crop residues and prunings, and (6) animal manure. All these modes of N transfer
are possible in an intercropping situation, but only residual effects will be involved in
sequential cropping. In this section, we examine successively N contribution in legume-
non-legume intercropping, livemulch systems, alley cropping and sequential farming.
Nitrogen contribution in legume-nonlegume intercropping
Various reports give evidence that the legumes benefit the associated crops in
intercropping systems (Evans, 1960 ; Ckiru and Willey, 1972). But Trenbath (1976) com-
piled data (607 values) from various sources and showed that the effect of legumes on
LER was not necessarily better than adding nonleguminous intercrops. He concluded that
the main advantage from the legume~components was in the saving of N-fertilizer. He
also noted that while competition for N would be miuimized in mixtures involving legumes,
competition for other growth factors could occur.
Early scientific reports on active N tranfer from the root nodules to associated
crops were published in the 193Os, in pot experiments (Willey, 1979a). It was noted that
shading favours N excretion from the legumes. This may however be significant only
when the legumes have already fmed N,, as shading will reduce N, fixation. Defoliation
can also promote N release in the root system. Whitney and Kanehim (1967) found N
182
contribution of 0.1 to 0.6 kg ha-’ from the root system of defoliated Desmodilun inrortum
to pangola grass.
Although direct evidence of active N excretion from the root nodules of-legumes
is lacking, other data suggesting N transfer from the legumes include the lack of response
of cereals to fertiliser N when they are intercropped with legumes (Agboola and Fayemi,
1972 ; Remison, 1978), higher N uptake of the cereals in intercropping than when they
are grown as sole crops (De, 1980 ; Eaglesham et al., 1981) and ‘%I studies (Eaglesham,
1982 ; Patra et al., 1986). Short cycle legumes seem to contribute more nitrogen to the
associate4l crops than late maturing species. Maize derived more benefits from Vigmz i&a
than it did from cowpea (Agboola and Fayemi, 1972), and from L.uthyrus sp. and early
maturing cowpea or groundnut than it did from Phuseoh vulgaris (Waghmare and Singh,
1984). This suggests that N contributions might occur mostly through senescence and
decomposition of the short cycle legumes.
Nitrogen transfer from legumes to nonlegumes in intercropping is not always trans-
lated in N (Fig. 1) or grain’yield (Danso et al., 1987) advantages. Nitrogen-15 studies
provide the most reliable data on the occurence and magnitude of N transfer in intercrop-
ping involving legumes. Eaglesham (1982) found that intercropped maize had access to
a source of N not available to sole maize crop. He presented 15N data consistent with
N transfer from cowpea to the intercropped maize. In experiments conducted by Patra
ef al. (1986), 28% of maize N representing 21.2 kg N ha’ was obtained from the atmos-
phere through transfer of fmed N by cowpea grown in association with maize. They con-
firmed their data in pots with either maize or wheat intercropped with cop
N content. kg ha-l
Cbwpoa (LSD 5% - 6.4)
I-
Maize (LSD 5% - 4.2)
Figure 1: Response of wwpea and ma&N
content to intezcmpping. Cowpea v&&s
are average for 6 wwpea cultivi3rs ; for
maize, average of maize intezcropped with
I-
6 wwpea cultivars (Adapted fium Ma&m-
gay. 1986a).
sole cropping
lnb3rcropping
183
Nitrogen contribution in livemulch systems
Live-mulch is a crop production system in which food crops are planted into a
i’
low-growing cover crop with minimum soil disturbance (Akobundu, 1980). The cover
crop smothers weeds and protects the soil. Akobundu (1980) observed that after five sea-
sons of continuous cropping, maize did not respond to fertihzer N in live-mulch plots
of Psophocarpus pahris and Centrosema pubescens whereas maize yields in bare plots
under minimum or conventional tillage increased when N fertilixer (60 kg ha-‘) was given.
Mulongoy (1986b) compared N contributions to maize of newly established and 3:years
old live-mulches of Psophocarpus, Centrosema and Arachis repens. In the newly esta-
blished field, cover crops and maize competed for N to the disadvantage of maize. In
the longteim plot, P. p&.mis was the best live-mulch with N contributions averaging
30.7 kg ha-‘.
In another experiment, nitrogen uptake by maize was monitored for five cropping
seasons in a live-mulch of P. palustris. During the first four cropping seasons, P. paLlcstrii
and maize competed for soil N in the live-mulch, and maize grown in the live-mulch
contained less N than maize grown by itself. In the fifth season, the live-mulch contributed
15 kg N ha-l to maize in the plots that were cropped continuously for five seasons. Maise
grown in a psopho livemulch that had aheady been established for four cropping seasons
contained 36 kg ha-’ more N than maize grown by itself. This positive N contribution
came, at least in par& from increases in microbii biomass and in organic matter derived
from the cover crop in the form of litter and earthworm casts (Mulongoy, 1986b).
The live-mulch system, however, needs refinement because, at present, a net contri-
bution of N and an increase in food crop yield occur only after the cover crop has been
established for some cropping seasons. This lag in benefits has to be shortened if the
live-mulch system is to be accepted by farmers. One way to shorten the lag is to minimize
competition for soil N between the cover crop and the food crop during live-mulch esta-
blishement. Also, if the pathways of N transfer from the live-mulch to the food crop
were understood, it might be possible to manage the system so as to increase the amount
of N contributed by the legume.
Simpson (1976) and Ladd er al. (1981) also found little N transfer from low-growing
legumes to associated crops, in pastures. The main N value from the legume seems to
be the building-up of soil organic N for adequate N supply to subsequent crops. Milller
and Sundman (1988) who monitored the fate of iw released during decomposition of
leguminous plant materials buried in the field found that 2746% of input N was re-
tained in the soil. A succeeding crop of barley took up 6-254s of input.
Nitrogen contribution in alley cropping
In the alley cropping, leguminous trees are preferred to monleguminous trees for
the hedgerows. They can bring atmospheric N, into the soil system. Well-nodulated trees
or shrubs can yield more than 500 kg N ha-l in a year (Guevarra, 1976) owing to their
high N,-fixing ability. Estimation of N, fixation of trees is difficult. But current data
indicate that nodulated trees can fix symbiotically more than 50% of their N (Roskoski,
1981 ; Hogberg and Kvarnstrom, 1982 ;~Rinaudo ef al., 1982 ; Sanginga et al., 1986;
MuIongoy et al., 1988 ; Ndoye and Dreyfus, 1988). The mojor proportion of hedgerow
tree N is found in the prunings consisting of leaves, twigs and succulent plant parts inclu-
ding immature pods. Application of legume prunings increases N and grain yields in the
184
associated crops uable 1 ; Kang, 1987). However, N contribution of the prunings to
the associated crop may not always be very efficient. Kang (1987) and Mulongoy and
Van der Meersch (1988) reported it in the range of 10 and 40% of the input N, and
Kang et al. (1985) found that up to 80 kg fertilizer N was needed for optimum maize
grain yields. Some results suggest that part of the N from prunings is retained in the
soil organic-N pool (Mulongoy and Van der Meersch, 1988). In a hydromorphic soil crop
ped to rice, N contribution of Sesbmiu rostrutu was estimated equivalent to 40 kg N
ha-’ representing about 30% of the N applied in the prunings (Mulongoy, 1986c). Mana-
gement practices such as incorporation of prunings in the soil can increase the efficiency
with which N is transferred from prunings to the associated crops (Read et al., 1985 ;
Wilson et d., 1986).
Nitrogen contribution Corn belowground parts can be considerable. Sangiga et al.
(1986, 1988) estimated it at about 30% of total N contribution of leucaena to maize
in a rotation experiment. Nitrogen contribution from roots of hedgerow trees needs to
be investigated.
Nitrogen contribution in sequential cropping
The residual effect of leguminous crops in a sequence will greatly depend on
several fixton including the legume N at harvest time, N harvest index, management
of the residues (Whether retained on the soil surface, incorporated in the soil or partly
removed for domestic use or for feeding animals) and on the time interval between crops
in the sequence. Kang (1987) discussed some examples of N returned to soil in crop
residues of pigeon pea, cowpea and soybean. The following generalizations can be
drawn : (1) legume N yield i.e. the sum of atmospheric nitrogen fixed and N absorbed
from soil, is high, varying in these examples between 73 and 205 kg N ha-* ;
(2) N in fallen leaves and root system is usually low ; but it can be as high as
40 kg N ha-’ in well nodulated shrubs at harvest ; (3) N removed in crop residues, pods
and seeds can be relatively low ; it ranged in these examples between 26 and
85 kg N ha-’ ; (4) N balance can be very low even negative indicating that the legume
can deplete soil N particularly when N, fixation is low and/or N harvest index is high.
The N retmned to the soil in the residues will then be mineralized before it is
used by the succeeding crops. Plowing the residues under may reduce the probability
of N loss by volatibzation and erosion, and may increase N contribution to the succee-
ding crops.
Kang (1987) reports a number of experimental data showing that the legumes im-
proved soil N status, with resulting N and grain yield benefits to the succeeding non
legtmrinous crops. There were differences among plant species in the level of N contri-
bution. Pnzeding non-leguminous crops with a legume increased their yields more than
intercropping. Some investigators however reported no or negative N contribution from
legumes in sequential cropping. A good knowledge of the N cycle in the system will
allow to devise practices that can maximize the resources for low-input agriculture
suitable for the smallholder farmers.
185
Methods for evaluating the economic potential of legumes In multiple cropping
systems
The purpose of this section is to examine methods by which to appraise the eco-
nomic potential of legumes in multiple cropping systems. The basic question is what
are the best systems (with or without legumes) capable of raising or maintaining the level
of soil fertility while satisfying the farmer’s goals and resource limitations. The need
to answer this question is pressing as results from past agronomic studies show that
multiple cropping systems (especially when legumes are associated with non
leguminous food crops) are much more efficient than sole cropping (Willey, 1979a and
1979b ; Steiner, 1982 ; Horwith, 1985 ; Spencer, 1985). Leguminous crops have been
shown to improve soil fertility (through nitrogen fmation and transfer) thus contributing
to yield stability (risk reduction) over different seasons and higher yields in a given season.
Few economic studies are available which evaluate the role of leguminous crops
in multiple cropping systems. Mittal er aI. (1985) conducted a study to determine yield,
total production and net returns by intercropping legumes (black gram, green gram, cowpea
and groundnut) with maize. Their results show that maize in- with groundnuts
is the most profitable and suitable system for rainfed conditions in Shiwalik region, India.
Singh and Chand (1980) also studied the income effect of intercropping grain legumes
with maize. Their results show that maize intercropped with soybean is the most remu-
nerative system with 120 kg N ha-l.
Although these studies show interesting results, they are limited for a number of
reasons. First, they were conducted essentially in tropical Asia and therefore their results
may not be applicable to sub-saharan Africa where the rate of demographic change is
higher and the physical conditions which determine technical potential are more difficult
(Matlon and Spencer, 1984). Most of these studies use the partial budgeting approach
in which the market value of land and labor are unknown or included in an ad-hoc manner.
Yet the real economic prices 0.e. shadow prices) of these resources are essential for eva-
luating the profitability of a cropping system and how for example labor and different
land types are allocated in the cropping seasons and among crops. Partial budgeting ap
preach ignores the substitutability of inputs and how they are allocated based on fixed
endowments and the implicit prices of the resources. Finally, these studies do not consider
risk factors. Farmers in sub-sahamn Africa are risk-averse. They hedge against risk by
diversifying into as many crops as possible through multiple cropping. Crop yields and
prices vary from one growing season to another thus affecting the expected income of
the farmer. If this is true, then an economic evaluation of multiple cropping systems using
riskless models is inappropriate.
For appraising the economic potential of legumes in multiple cropping, a moderne
theory based on utility maxim&ration and risk programming is proposed. This is a much
more appropriate model to describe farmers behaviour in sub-sahamn Africa than the standard
profit maximization models. Farmers practice diversified farming not only to take ad-
vantage of the complementary nature of the various crops but also to minimize risk in
aggregate farm income given a level of expected income. Quadratic programming which
is based on the expected-income variance criterion is a practical way of selecting income-
risk efficient plans. It assumes that the farmer’s preference among alternative farm plans
is based solely on expected net farm income, E and variance, V (Freund, 1956). Quadratic
programming model was first applied to agriculture by Freund (1956) and has since been
186
recommended for analysis of uncertainties in gross margins, gross returns and net variable
costs in farm planning (Heraedy and Candler, 1958).
The problem of resource allocation under risk can be mathematically formu-
ItiaS:
h&x EV(X) = RX - (aLI)XQX
Subject to AX-3, where X is an (n x 1) vector of activities (e.g. hectares planted to
maiz&mpea ; sole maize, maiz4cassava/leucaena etc.)
R
=
(n x 1) vector of expected net returns associated with the vector of acti-
vities
t
=
trampos of R and X
A =
(n x n) matrix of technical coefficients (e.g. labor, rotational or land
requirement etc.)
B
=
(n x 1) vector of resource constraint level (e.g. available labor or capital,
or land etc.)
=
is the risk averse coefficient
;
=
(n x n) matrix of tbe variance-covariance matrix of gross returns.
The special feature of the quadratic programming approach is the variauce cova-
riance component of the objective function, EV. When the risk aversion coefficient cc
is zero, the problem collapses to a standard linear programming problem indicating that
farmers are risk-neutral. A positive (negative) value of a indicates the farmer is risk averse
(lover). Although there is no clear method by which to estimate a the latter is usually
obtained by eliciting farmers’ preferences. Application of the quadratic programming ap-
proach also mquires that expected returns and variance and covariauce of returns for all
enterprises be known. ~Covariances am fundamental for efficient diversification among
farm enterprises as a means of hedging against risk (Heady, 1952). For example combi-
nation of activities that have negative covariate gross margins will usually have a more
stable aggregate return that the return from more speciabzed strategies. A case in print
in sub-saharan Africa is the intercropping versus the monocropping system. Returns of
crops in an intercropping may covariate negatively thus providing in the aggregate a stable
return, compared to monocropping systems.
The quadratic programming approach assumes that the farmer is rational and thus
restricts his choice to those farm plans for which the associated income variances are
minimum for given expected income levels. This assumption is based on the premkes
that individuals like increases iu expected returns and dislike increases in variance of
returns. The variance and covariance of return can be obtained using standard statistical
packages or a culcu@ion using the variance formula
V = cj(Rj - r)2/n-l
where Rj denotes the gross margin of activity j, and r is the average return on activity
j over n cropping seasons. Other data needed depend on the problem specification and
are usually obtained by means of survey techniques, direct measurements and secondary
sources of information.
187
Conclusions
Multiple cropping is typical of agricultural practices in the tropics. Depending on
the level of agricultural technology, multiple cropping involves the growing of sole crops
in sequence or various mixtures of rain-fed crops. Intercropping is the predominant prac-
tice by small-holder farmers producing relatively low crop yields. It does however make
better use of the resources than sole cropping, and help minimize variability in returns.
Legumes are usually included in multiple cropping as sources of protein. This is
because they fit the temporal and spatial requirements of the system. It is indeed believed
that legumes competeonly little with other crops for N, and that they can reduce the
need for fertilizer N in the system since they can fix atmospheric N,
There is a dearth of information in multiple cropping’.on the dynamics of rhizobia
and other nitrogen-fixing microorganisms when they are living in the rhizosphere of non-
host plants, on how nodulation is affected by the various interactions observed between
component crops, and on how much N, legumes and other N,-fixing systems fn in mixtures
or in sequence with non-leguminous crops. The response of legumes and other crops
such as maize to crop mixture and high population densities have been widely studied
utilizing the aeplacemenb series (De Wit, 1965) or, most commonly, the u additive >
technique (Chang, 1981). The same techniques can be used to monitor the changes in
biological N, fijration parameters. Also, only little quantitative data is available on the
process of N cycling in multiple cropping. Nitrogen-15 studies will provide the most
accurate information about N, fixation and N transfer. They allow a distinction on one
hand between N derived from atmosphere, soil and fertilizer and, on the other hand, bet-
ween N transfer and competition (Pa&a et al., 1986).
Although, research results have demonstrated the N and grain yield advantages
of including legumes in multiple cropping, analyses are laking which determine the eco-
nomic potential of nitrogen fixation and transfer of legumes in multiple cropping systems
in Africa. In this paper, a method of economic analysis is highlighted which takes into
account the risk minimizing behaviour of farmers in sub-saharan Africa.
References
AGBGGLA, A.A. and FAYEMI, A.A. 1972 Fixation and excretion of nitrogen by tropical
legumes. Agron. J. 44, 409412.
AKOBUNDU, Lo.1980 Live mulch : a new approach to weed control and crop pro-
duction in the tropics. Proc. British Crop Protection CoyT - Wee&. pp. 377-382.
ANDREWS, D.T. and KASSAM, A.H. 1976 The importance of multiple cropping in
increasing world food supplies. In : Papendick RI., Sanchez, PA. and Triplett G.B.
(eds). Multiple Cropping, pp. l-10 ASA Spec. Publ. 27 Madison.
A-I-TA-KRAH, AN., SUMBERG, J.E. and REYNOLDS, L.1985 Leguminous fodder trees
in the farming system. An overview of research at the Humid Zone Programme of
lLCA in southwestern Nigeria. In : Haque, L., Jutzi, S. and Neate P.S.H.(eds.). Potentials
of Forage Legumes in Farming Systems of Sub-Saharan Africa, pp. 307-329 ILCA,
Addis Ababa, Ethiopia.
AYANABA, A. and DART, P.J. 1977 Biological Nitrogen Fixation in Farming Systems
of the Tropics. John Willey and Sons, Chichester.
188
CHANG, JE. 1981 An analysis of competition between intercropped cowpea and maize.
Ph.D. Thesis, lowa State University.
CHARREAU, C. and VIDAL. P. 1965 Influence de I’Acacia albida De1 sur le sol, nutrition
minhale et rendements des miIs Pennisetum au Senegal. Agronomie Tropicale. 20,
600-625.
DANSO, SKA, ZAPATA, F., HARDARSON, G., and FRIED, M. 1987 Nitrogen fixation
in fababeans as affected by plant population density in sole or intercropped systems
with barley. Soil Biol. Biochem. 19, 411415.
DE, R. 1980 Role of legumes in intercropping systems. In : IAEA. Nuclear Techniques
in the Development of Management Practices for Multiple Cropping Systems. pp.
73-84. IAEA, Vienna.
DE WIT, CT. 1%5 Competition among herbage plants. Netherlands JOIUM~ of Agricul-
tural Science. 13, 212221.
EAGLESHAM, ARJ., AYANABA A., RANGA, RAO, V. and ESKEW. D.L 1981 Impro-
ving the nitrogen nutrition of maize by intercropping with cowpea. Soil Biol. Biochem.
W, 169-171.
EAGLESHAM, A&J. 1982 Assessing the nitrogen contribution of cowpea (vigna ZUI-
guhlatu) in monoculture and intercropping. In : Graham, PH. and Harris, S.C. (eds).
Biological Nitrogen Fixation Technology for Tropical Agriculture. pp. 641-646. CIAT,
Cali; Colombia.
EAGLESHAM, ARJ. and AYANABA, A. 1984 Tropical stress ecology of rhizobii, root
nodulation and legume fixation. In : Subba Rao N.S. (ed.). Current developments
in biological nitrogen fixation pp. l-35. Oxford and IBH Publishing Co. New Delhi.
EVANS, AC. 1960 Studies of intercropping. I. Maize or sorghum with groundnuts. East
African Agricultural and Forestry Journal. -26, I-10.
FREUND, RJ. 1956 The introduction of risk into a programming model. Econometrica.
34, 253-263.
GIBSON, AH. 1976 Recovery and compensation by nodulated legumes to envrionmental
stress. In : Nutman, P.S. (ed.). Symbiotic nitrogen fixation in plants, pp. 385-403.
Cambridge University Press, Cambridge.
GUEVARRA, A.B. 1976 Management of L.eucaena leucocephala (Lam.) de Wit for maxi-
mum yield and nitrogen contribution to intercropped corn. PhD. Thesis, University
of Hawaii.
HEADY, E.O. 1952 Economics of agriculturaI production and ressource use. Englewood
cliffs, NJ : Prentice-Hall.
HENZELL, E.F. 1981 Forage legumes. In : Broughton WJ. (ed). Nitrogen fixation. Volume
1 : Ecology, pp. 264-289 Clarendon Press, Oxford.
HOGBERG, P. and KVARNSTROM, M. 1982 Nitrogen fu;ation by the woody legume
Leucaena leucocephala in Tanzania. Plant and Soil. 66. 21-28.
HORWlTI-I, B. 1985 A role for intercropping modem agriculture. Bioscience. 35, 286-
291.
INTERNATIONAL INSTlTUTE of TROPICAL AGRICULTURE. 1985 Annual Report
for 1984, pp. 182-183. Ibadan, Nigeria.
189
JARDIM F’REIRFL JR. 1984 Important limiting factors in soil for the RHZZOBZUM -
legume symbiosis. In : Alexander, M.. (ed.). Biological nitrogen fixation. Ecology,
technology, and physiology, pp. 51-74. Plenum Press. New York and London.
KANG, B.T., WILSON, GF. and SIPKENS, L. 1981 Alley cropping maize (Z&z mays
L.) and Leucaena leucocephulu &am) in southern Nigeria. Plant and Soil. 63, 165
1 7 9 .
KANG, B.T., GRIMMB, H. and LAWSON, T.L. 1985 Alley cropping sequentially crop-
ped maize and cowpea with leucaena on a sandy soil in southern Nigeria. Pht arid
Soil. 63, 165-179.
KANG, B.T. 1987 Nitrogen cycling in multiple cropping systems. Paper presented at the
International symposium on advances in nitrogen cycling in agricultural ecosystems.
Brisbane, Australia.
LADD, J.N., OADBS, J.M. and AMATO, M. 1981 Distribution and recovery of nitrogen
from legume residues decomposing in soils sown to wheat in the field. Soil Biol.
Biochem. 13, 251-256.
LAL, R. 1987 Managing the soils of sub-sahara Africa Science. 236, 1069-1076.
LUMPKIN, T.A. andPLUKNB’IT, D.L. 1982AzoZZu as a green manure : use and management
in crop production. Westview Press, Boulder,. Colorado.
MATLON, PJ. and SPENCER, D.S.C. 1984 Increasing food production in Sub-saharan
African : Environmental problems and inadequate technological solution. Amer. .Z.
of Agric. Economics. 66, 671-676.
MITTAL, SP., GREWAL, S.S., AGNIHOTRI, Y. and PRATAP SING 1985 Economics
of intercropping legumes in maize under rainfed condition. Intern. J. Trop. Agric.
3, 187-191.
MULLER, M.M. and SUNDMAN, V. 1988 The fate of nitrogen (%) released from dif-
ferent plant materials during decomposition under field conditions. Plant and Soil.
105, 133-139.
MULONGOY, K. 1986a Nitrogen fixation and N contribution of cowpeas in a maize-
cowpea intercropping system. Paper presented at the Second AABNP Conference,
Cairo, Egypt Dec. 15-19.
MULONGOY* K. 1986b Microbial biomass and maize nitrogen uptake under a Psopho-
carpus palustris live-mulch grown on a tropical Alfkol. Soil Biol. Biochem. 18,
395-398.
MULONGOY, K. 1986c Potential of Sesbuniu rostrutu (Brem.) as nitrogen source in
alley cropping systems. Biol. Agric. and Hortic. 3, 341-346.
MULONGOY, K. and NKWIINB, C. 1987 Effect of cropping on rhizobial pbpulation
and nodulation of soybean. In : BTA resource and crop management Program. Annual
Report for 1986, pp. 77-78. Ibadan, Nigeria.
MIJLONGOY, K. and VAN DER MBBRSCH, M.K. 1988 Nitrogen contribution by leu-
caena (Leucaena leucocephulu) prunings to maize in an alley cropping system. Biof.
Fert. Soils. 6, 282-285.
190
MlJTXlNGOY, K., OWOAJE, B.T., IBEWIRO, E.B. and OKOGUN, J.A. 1988 Nitro-
gen fixation and N contribution of selected trees in alley cropping system. In : Bothe,
de Bruijn and Newton @is.). Nitrogen fixation : hundred years after Gustav Fisher,
p. 824. Stuttgart, Germany.
NDOYE, I. and DREXFUS, B. 1988 N2 fmtion by Sesbania rosfruta and Sestmniu sesban
estimated using 15N and total N difference methods. Soil Biol. Biochem. 20, 2091
213.
NORMAN, MJ.T. 1982 A role for legumes in tropical agriculture. In : Graham, P.H.
and Hank, S.C. (eds.). Biological nitrogen fixation technology for tropical agriculture,
pp. 9-26, CIAT, Cali, Colombia.
NUTMAN, P.S. 1976 Symbiotic Nitrogen fnation in plants. Cambridge University Press,
Cambridge.
OKIGBO, BN. and GREENLAND, DJ. 1976 Intercropping systems in tropical Africa
In : Paper&k, RI., Sanchez, P.A. and Triple& BB. (eds.). Multiple cropping, pp.
63-101. ASA Spec. Publ. 27. Madison.
OSlRU, D.S.O. and WILLEY, R.V. 1972 Studies on mixtures of dwarf sorghum and
beans (Phuseolus vulgaris) with particular reference. to plant population. Journal of
Agricultural Science, Cambridge. 79, 531-540.
PALANIAPPAN, S.P. 1985 Cropping systems in the tropics : principles and management,
Wiley Eastern Ltd., New Delhi.
PATRA, D.C. SACHDER, M.S. and SUBBIAH, B.V. 1986 lsN studies on the transfer
of legume fmed nitrogen to associated cereals in intercropping system. Biol. Fertil.
Soils. 2, 165171.
READ, M.D., KANG, B.T. and WILSON, GF. 1985 Use of kucaena Zeucocephala (Lam.)
de Wit leaves as a nitrogen source for crop production. Fert. Res. 8, 107-116.
REMISON, S.U. 1978 Neighbour effects between maize and cowpea at various levels
of N and P. Erpl. Agric. 14, 205-211.
RlNAUDO, G., DREYFUS, B. and DOMMERGUES, Y.R. 1982 Intkence of Sesbania
rostrata green manure on the nitrogen content of rice crop and soil. Soil Biol. Biochem.
15, 111-113.
ROSKOSKI, JP. 1981 Nodulation and N,-fixation by Inga junicuil, a woody legume
in coffee plantations. Measurements of nodule biomass and field CA reduction rates.
Plant and Soil. 59, 201-206.
SANCHEZ, P. 1976 Properties and management of soils in the tropics. John Wiley &
Sons, New York, USA.
SANGINGA, N., MULGNGOY, K. and AYANABA, A. 1986 Inoculation of Leucuena
Zeucocephula (Lam.) de wit with Rhizobium and its nitrogen contriburion to a subsequent
maize crop. Biol. Agric. Hortic. 3, 347-352.
SANGINGA, N., MULONGOY, K. and AYANABA, A. 1988 Nitrogen contribution of
L.eucaenalRhizobium symbiosis to soil and subsequent maize crop. Plant and Soil.
112, 137-141.
SANGINGA, N., BOWEN, G.D. and DANSO, S.K.A. 1990 Irma-specific variation in
growth and N,-futation of L.eucaena leucocephala and Giiricidia sepium at low levels
of soil phosphorus. Plant and Soil. (In press).
191
SIMPSON, JR. 1976 Transfer of nitrogen from three pasture legumes under periodic
defoliation in a field environment. Aust. J. Exp. Agric. Anim. Hub. 16, 863-870.
SINGH, C.M. and CHAND, P. 1980 Note on economics of grain legume intercropping
and N fertilization in maize. Indian J. Agric. Res. 14, 62-64.
SLOGER, C., BEZDICEK, D., MILBERG, R. and BOONKERD. N. 1975 Seasonal and
diurnal variations in N, (C,H@ixi.ng activity in field soybeans. In : Stewart W.D.P.
(ed.). Nitrogen fixation in free-living micxorganisms. pp. 271-284. CambridgeUniversity
Press, London.
SPENCER, D.S.C. 1985 Aresearch strategy to develop appropriate agricuhural technologies
for small farm developement in sub-saharan Africa. In :--Ohm H. and Nagy. J. @is.).
Appropriate technologies for farmers in semi-arid West Africa;pp. 308-325. J. Intema-
tional Programs in Agriculture, Purdue University, West Lafayette, Indiana, USA.
SPRENT, J-1. 1971 The effects of water stress on nitrogen-fixing root nodules. I. Effects
on the physiology of detached soybean nodules. New Phytol. 70, 9-17.
SPRENT, J-1. 1972 The effects of water stress on nitrogen-fixing root nodules. IV. Effects
on whole plants of Vi& f&a and Glycine max. New Phytol. 71, 603611.
STEINER, KG. 1982 Intercropping in tropical smallholder agriculture with special reference
to West Africa. German Agency for Technical Cooperation. Germany.
TRENBATH, BR. 1976 Plant interactions in mixed crop communities. In : Papendick,
R.I., Sanchez, P.A. and Triple& G.B. (eds.). Multiple Cropping, pp. 129-169. ASA
Spec. Pub. Np 27. ASA, Madison, Wisconsin.
WAGHMARE, and SINGH. 1984 Sorghum-legume intercropping and the effect of nitrogen
fertilization. I. Yield and nitrogen uptake by crops. ExpZ. Agric. 20. 251-259.
WHITNEY, A.S. and KANEHIRO, Y. 1967 Pathways of nitrogen transfer in some tropical
legume - grass associations. Agron. J. 59, 585-588.
WILLEY,R.W. 1979aIntercropping - its importanceandreseamh needs. Part 1. Competition
and yield advantages. FieZd Crop Abstracts. 32, l-10.
WILLEY, R.W. 1979b Intercropping - its importance and research needs. Part 2. Agronomic
and research approaches. Field Crop Abstracts. 32, 73-85.
WILLEY, R. W. and RAO, M.R. 198 1 Genotype studies at ICRISAT. In : Proc. Int. Workshop
on Intercropping. 10-13 Jan. 1979. ICRISAT, Hyde&ad. pp. 117-127.
WILSON, G.F., KANG, B.T. and MULONGGY, K. 1986 Alley cropping: Trees as sources
of green-manure and mulch in the tropics. BioZ. Agric. Hort. 3. 251-267.
Maximiser la FBA pour la Production Agricole et Forestière en Afrique
Importance de la plante-hôte et amélioration
génétique de la fixation biologique de NS
chez les arbres tropicaux
DUHOUX E.
Université Paris VII et BSSmlCTFT (CIRADIORSTOM),
4.5 bis, Av. de la Belle Gabrielle. 94736. Nogent sur Marne France.
Summaty.
The host plant plays a most important role in the pmce~s of symbiotic N2 fixa-
tion Some of host gene products involved in the development and maintenance of
the symbiosis have already been identified. But further studies are necessary before
attempting the transfer of the symbiotic genes to novel plants.
Improvement of the host plant performances to increase its N,-fixing potential
and tolerance to environmental stresses could be achkved by : exploiting natm=al
genetic variability, expanding genetic variabiity through somaclonal variation and
introducing new genes in the plant genome by genetic transformation.
Theresultsdiscussedinthepapersupportthe recomlnwdation that selectian of host
plant should take into account various traits, specially N2-fixing potential, nitrate
assimilation, and tolerance to combined nitmgen inhibition.
Introduction
Les stratigies d’amt%oration des arbres forestiers reposent, comme chez les plantes
cultiv&s annuelles, sur une amélioration génétique des individus et également sur I’utili-
sation de mCthodes culturales adapteeS. Ce qui diff&re chez les ligneux, c’est tout d’abord,
la dur& de la vie de la plante qui peut être parfois consid&able et ensuite, la grande
variabilité g&&ique naturelle qui existe dans les peuplements, par suite de l’absence de
194
travaux de g&tique’sur ces végétaux a cycle de gMration tr&s long. De surcroît, dans
les associations symbiotiques. l’etude du partenaire veggétal a été longtemps d&tiss& au
profit de celle du microorganisme. Il n’est donc pas étonnant que l’arbre, en tant que
plante-hôte des associations symbiotiques forestières, ait été peu utilise en tant que syst&me
fixateur de N2
Les récents progr&s obtenus ces dernieres années dans les technologies avancees
de transfert de gènes aussi bien chez les microorganismes que chez les plantes herbacees
ont permis d’obtenir des r&sultats tout à fait remarquables (Shah et al., 1986 ; De. Blok
et al., 1987). Qu’en est-il des systbmes forestiers fixateurs de N2 ?
Avant de faire l’état actuel de nos coxuxGmnces, nous rappellerons l’importance
du partenaire végétal en dégageant quelques-uns des rôles connus de la plante-hôte dans
le fonctionnement de la symbiose furatrice de N2. Nous envisagerons ensuite les approches
possibles que l’on peut mettre en œuvre pour ameliorer la fixation de N2 en faisant appel
A la plante-hôte, et les premiers r&wltats concernan t cette stratégie seront présentés. Enfin,
l’amélioration gennétique de la plante-h& sera discutée en fonction des autres approches
Utilisées.
Importance du rble de la plante-hôte dans la fixation de N,
Qn sait que la planteh& fournit 1’6nergie snffiite au métabolisme de la fixation
de N, et à la survie du microorganisme symbiotique. Elle contrôle l’activité de I’hydrog&tase
et participe aussi à la protection de la nitrogénase. Depuis un demi-siècle, de nombreux
travaux ont demontre abondamment que chacune des etapes du processus de l’infection
et de la nodulation est contrôlt5e par la plante, en relation étroite avec le microorganisme
symbiotique. Le contrôle porte en particulier sur le nombre, la taille, la morphog&se
du nodule et la potentialit6 de forer N2’ On peut mettre en évidence ce rôle, non seulement
dans l’expression phénotypique, mais aussi, de plus en plus, au niveau génomique.
L’infection
L’existence des groupes d’inoculation croisée entre l’hôte et la bactérie symbiotique
est une manifestation évidente du rôle de la plante-hôte dans la symbiose. Une espèce
donnée de Rhizobium ou de Fnmliiu ne nodule pas toutes les espèces de lt5gurnineuses
ou de plantes actinorhiziennes mais seulement un ensemble d’espèces vegetales appartenant
a un groupe dit wl’inoculation crois&». Ce groupe peut comprendre un grand nombre
d’especes (qui sont dites non spt?cifïques) ou un petit nombre d’espèces ou même de
varie (groupe spécifique).
Sion s’en tientau syst&meL&umineuse-Rhizobiumquiconstituelemodelesymbiotique
le mieux connu actuellement, on peut souligner les rôles du genome de la plante qui
interviennent lors de l’infection de la plante par le Rhizobium :
émission de substances chimiques excr&es par la racine.
Ces substances jouent le r61e de signaux pour l’activation ou non des g&nes nod de la
bact&ie. Certaines de ces substances, hydroxyflavones et isoflavones (Fiiin et aL. 1986 ;
Peters et al., 1986) seraient stimuIatrices, tandis que d’autres, coumarines et isoflavones,
reprimeraient l’expression de ces mêmes genes (Djordjevic et al., 1987a). Les substances
stimulatrices seraient secretées dans la zone d’émergence des poiis absorbants, les substances
inhibitrices proviendraient des autres zones de la racine (Djorvdjevic et aZ., 1987b).
195
actlvatlon des syst&mes de reconnaissance. Sous l’influence
de polymères polysaccharidiques varies, exopolysaccharides, lipopolysaccharides, polysac-
charides capsulaires (Leigh et aL, 1985) synth&ises dans la paroi bacterienne, le génôme
de la plante neconnait ou non le Rhizobium spkcifique. En cas de reconnaissance il participe
a la courbure du poil absorbant et a la formation du cordon d’infection. Dans le cas
contraire il induit une réaction d’incompatibilité.
mlse en place de f’autorégulation de la nodulation. Ce phé-
nomène déclenché par le g6nome de la plante permet de limiter le nombre des nodules
et définit leur r@artition sur la racine (Nutman, 1952 ; Pierce et Bauer, 1983). La zone
de la racine susceptible d’&re infect& est Emit& à la région d’émergence des poils absor-
bants. Par ailleurs, une premiere infection inhibe un deuxième point d’inoculation situe
à proximité (Rolfe et Gresshoff, 1988).
Etablissement de la symbiose
La variabilité dans les réponses obtenues concernant la nodulation est le rbultat
de l’interaction du génome des deux partenaires. Par l’interm&liaire des mutants obtenus
à Ia fois chez la batterie et Ia plante-hôte on a mis en évidence une serie d’anomalies
dans la nodulation qui ont comme support le g&tôme des deux parknaires. Nous nous
attacherons ici a pr6sente.r celui de la plante-hôte.
On résumera dans le Tableau 1 quelques exemples de caractkstiques contrôlées
par la plante en indiquant, lorsqu’ils sont connus, les gènes responsables.
Fixation de N,
Chez de nombreuses associations L&umineuses-Rhizobium, Pisum (Cousin et al.,
1985), Ttifolium (Nutman, 1971). Meukugo (Viands et al., 1981 ; Barr~es et ai., 1986)
on a pu mettre en Cvidence une large variabilid génétique dans l’activité fvratrice. Une
part importante de cette variabilité revient a la plan-hôte.
Râble de I’hBte dans la quantité de N2 ~IX&. On choisira comme
exemple, une méthode d’amélioration de la fixation biologique de N, développée chez
la luzerne (Phillips et al., 1985) bas& sur la sélection des génotypes de la plante. Deux
génotypes HP et HP32, sont inocules avec quatre souches de Rhizobium. Il apparait (Fig.
1) que la quantité de N2 fixé est de 22 à 53% plus elev6e dans le génotype HP32. Les
performances symbiotiques sont donc nettement améliorees si on choisit le génotype de
Ia plante-hôte.
Influencede l’hôtedanslavariabilitédepotentlalité
de fixation
de A/, En utilisant le modele expkimental dtkrit dans la Fig.1. Phillips et ol., (1985)
ont montre que plus de 70 96 de la variabilité relative au poids sec des tiges, au poids
sec total, a à l’azote totzd peut être attribuée aux génotypes de la plante. Les souches
de Rhiwbium n’interviennent que pour 10 à 15 % dans la variabilite.
Au cours de la fmtion de N2, on sait que le g6ndme de la plante intervient &alement
Des gènes vegétaux sont exprimes uniquement dans le nodule conduisant à la formation
de substances spécifiques appeks nodulines. Dans les symbioses L&umineuses-Rhizobium
on en connait au moins une vingtaine (Govers et al. 1985) qui interviennent dans le
fonctionnement de la symbiose.
Tableau 1 : Rôles de l'hôte dans la nodulation
Caracteris-
hôte
g&ie de
Phénotype
Références
tique de la
l'hôte
observe
symbiose
responsable
pas de nodule
Glycine max
rjl
pas de cordon
CALDWELL, 1966
d'infection
pas de nodule
Pisum sativum
syml
pas de cordon
LIE, 1971
taille et
Trifolium sp. -
complexe
in NUTMAN, 1981
nombre des
nodules
nombre des
Glycine max
nombre x 10 NOS-
CARROLL et al.,
nodules
tolérant = super-
1985
nodulants
structure et
Lupinus sp.(L.) -
t R. Lupini no-
KIDBY et
forme des
Ornithopus SP.(O)
dules en anneau
GOODCHILD, 1966
nodules
(L) 1 bacte-
roTde/mb nodules
cylind. (0) pl.
bactérofdes/mb
mode d'in-
Eleagnus
Frankia
MILLER et
fection
Myrica
infection
BAKER, 1986
interc. (E)
infect. poil
abs. (Ml
1 9 7
102F28
lOîf65
414
445
.sescbes k RM2wbram mermtr
Fiiure 1 : N2 total fïxC par deux génotypes HP 32 et HP de luzerne inocul& avec quatre souches de
Rhizobium mdiloti. (Phillips et coll.. 1985).
,-
Les approches pour I’amMoratlon g&Gtique de la plante-hbte chez les arbres
Exploitation de la variabilit6 g&Mique naturelle
Approches conventionnelles. Une technique simple et efficace
consiste à utiliser les variations g&r&iques spontan@ obtenues dans un lot de graines
d’une même provenance (c’est à dire, d’une même aire geographique) ou de provenances
différentes. L’identification des individus ayant un potentiel fixateur de N, élevé doit être
bas& sur des tests non destructeurs, comme de la num&ation des nodules ou des tests
de réduction d’ac&yEne. Ensuite, les individus s&ctionnés sont propages par des techniques
de multiplication végétative classique (Sougoufara et al., 1987) et testes au champ pour
leur aptitude a fixer l’azote. Les auteurs ont pu ainsi identifier un clone 6 qui fixait 1.8
fois plus d’azote que le clone a inoculC avec la même souche de Frunkiu.
Approches modernes. Dans ce cas, l’ensemble des op&ations est
effectut5e au laboratoire. L’inoculation et la nodulation des plantes est obtenue in vitro
dans des systèmes expérimentaux adaptes aux arbres (Dhawan et Bhojwani, 1987 ; Alabarce,
1987). L’inoculation experirnentale en conditions aseptiques permet d’assurer le contrôle
de nombreux paramètres et d’optimiser la nodulation.
La s&ction des génotypes repose sur les comparaisons des tests d’effectivitt5 et
les mesures de biomasse des nodules et des plantes. Les’conditions d’asepsie permettent
de réaliser une micropropagation efficace des plantes s&ction&es. La multiplication vCg&
tative est assurée par les techniques de culture in vitro, en .proc&ant au developpement
des bourgeons axillaires vCg&atifs chez AZnus (Tremblay et Lalonde, 1984 ; Simon et
al., 1987), Eleugnus (Bertrand et Lalonde. 1986), Casuurin~ (Sougoufara et al., 1986).
Acacia ulbidu (Gassama et Duhoux, 1986), Acuciu nmngium @aha, com. pers.), ou
au developpement de bourgeons axiBaires d’inflorescencesimmaturesdecmumina@uhoux
et ul., 1986).
Induction par culture de tissus d’une nouvelle variabiltt6 g&étique par
modification globale du gBn6me
La variation somaclonale. La culture in vitro des cellules veggétales
s’accompagne tres souvent d’anomalies gén&.iques. Plus la phase de désorganisation tis-
sulaire (cal) est longue, plus fnQuents sont observes les variants lors de la I-ég&u%ation.
Ces anomalies relevent de plusieurs catégories : chromosomique (aneuploïdes, polyploïdes),
Cpigertique (d’origine nucléaire et/ou cytoplasmique) (Mestre et Benbadis, 1985).
La variation somaclonale repr&ente ainsi une methode pour produire une variation
gennétique en Cvitant la voie de la reproduction sexuel. Ces dernières annees, on a pu
appliquer ce proc&le pour l’obtention de divers variants chez les plantes herbacees (.Y&
lection de plantes tolerant aux toxines de pathogènes, sélection de plantes r&Aames aux
herbicides, Larkin. 1987).
Chez les arbres fixateurs de N, cette technique ne peut-être mise en œuvre que
chez les plantes capables de regtnérer a partir de cals : Alnus (Hutinen et ul., 1982 ;
Tremblay et ul.. 1985) ; Acacia kou (Skolmen et Mapes, 1976) ; Dulbergiu lutifoliu (Rao,
1986) ; Allocusuurina, (Duhoux, n%ultats non publiés).
199
Hybridation sumafique. Avec la fusion de protoplastes, deux
partem+s (es+ ou genres diff&ems) peuvent proceder à un échange d’information
génétique (nucleaire et/ou cytoplasmique) en contournant aussi une partie des obstacles
de la reproduction sexu6e.
Cette technique pourmit conduire 3 la cr&tion d’hybrides somatiques pos&dant
de nouvelles cam&&iques quant Zr leur nodulation, ou pr6sentant mie tol&ance accrue
à certaines contraintes de l’environnement. Actuellement, l’obtention de protoplastes d’arbres
fixateurs de NZ demeure rare ; Alnus ghtinosa. A. incana (Hutinen et al., 1982) et A.
incana (Tremblay et al., 1985) et I’absence fréquente de leur pouvoir de division et de
régtn&ation demeure encore une limitation de l’emploi de ces techniques.
introduction dans le gbnome de la plante d’un ou de quelques caract&es
nouveaux.
Les r6cents succès en biotechnologie ont montre des r&tssites remarquables pour
des plantes modeles comme le tabac avec laquelle Shah et al., (1986), De Block et al.,
(1987) ont obtenu des plantes transg4niques n5sisumte.s B un herbicide, le glyphosate. L’ob-
tention d’arbres transg&tiques rt%istants B watains herbicides est envisagee. Des telles
plantes possederaient ainsi un avantage sélectif certain, vis à vis de plantes herbacées
qui entrent en compétition au moment de la mise en place des plantations.
Les techniques de transfert de génes utilisent soit, des vecteurs biologiques
(Agrobacterium fum&acien.s, A. rhizogenes, le virus de la mosaïque du chou fleur), soit
le transfert direct (&ctroporation, liposomes, microinjection, microprojectiles enrobes de
DNA...). Le nombre de plantes transgeniques obtenues chez les arbres est encore faible
; un gène de tol&ance à un herbicide a été transf6r6 et exprimé chez un peuplier (Fillati
er al., 1987). Chez les arbres fixateurs de N,, des travaux ont débuté chez 1’Aulne
(Mackay et d., 1987a) et les Caswriw (Phelep, 1988).
Il est a noter que si les travaux..de gt%%que moEculaire concernan t la fixation
de N, sont maintenant bien avancés pour Ie microorganisme (Rhizobiwn), il n’en est pas
de même pour la plante-hôte (16gumineuses et plantes actinorhiziennes). Il en r6sulte que
nos co~aissances actuelles de la biologie mol6culaire de la plante-h&e ne sont pas encore
suff=antes pour envisager l’utilisation du génie g6nt%que dans l’amélioration de la poten-
tialité de fixer Nz.
D’autres strategies, comme celle de transf&er les g&nes de la nodulation zi des
arbres (le bouleau) non fixateurs de Nz, mais proches d’un point de vue systematique
des plantes actinorhiziermes (aulnes), sont envisagées (Tremblay et al., 1985).
Appllcatlon de I’exploltatlon de la varlabillt6 naturelle a l’am6lioratlon de la
flxatlon de N, chez les arbres
Les r&ultats expérimentaux
Les travaux concernant les arbres sont peu nombreux. Ils seront illustres ici par
un exemple d’am&ioration basée sur l’exploitation de la variabilité naturelle chez une
plante actinorhizienne (Sougoufara et al., 1987, 1989).
Un criblage a et.6 effectué sur de jeunes plants issus de semis de C. equisetifolia
et inocules avec la souche de Fmnkia ORS 021001 selon la methode déjà dt?crite (Sou-
goufara et ol., 1987). Trois individus a. 6 et S qui pn%entaient des differences signifi-
200
catives en ce qui concerne la biomasse de leurs parties aériennes, leur nodulation, et leur
aptitude à fixer N, ont été retenus. Des boutures âgées de 3 mois de ces individus ont
et.6 plantées dans un sol stérile déficient en N, marqué avec de l’urée 15N afin de comparer
les quantités d’azote de la plante provenant de la fixation de Nz (Ndfa) et du sol @kifs).
Une moitié des plantes a Cté inoculée avec la même souche de Fmnkàa, l’autre n’a pas
1505 inocul&. Les clones inocules de 12 mois diffktient entre eux, non seulement par
le poids de leurs nodules mais aussi par leur aptitude à fucer N, (Ndfa) et a absorber
N du sol (Ndfs), (Sougoufara et al., 1989). Les clones non inocules differaient Cgalement
entre eux par leur absorption spkfique de N du sol (Fig. 2).
10
8
6
4
2
Figure 3 : Biomasse exprim6.e en g/N/plante de trois clones de Casucwinu equisefifoia (S. U, $) 12 mois
apr&s leur transplantation (SOUGOUFARA et al., 1989).
1: plants inocul6-s avec la souche de
ORS 021001. La contribution du sol est repkentie par les
rectangles vides et celle de la fixation de N2 par les rectangles avec l’indication “Nz”.
201
On a pu ainsi mesurer, dans le cadre d’une variabilite intraspkifique de plantes-
hôtes de C. equisetzfolia, difft%ents critikes, Ndfa, Ndfs, biomasse des plantes (exprimée
en N total), qui sont indépendants des expkimentations. Les differences observées sont
uniquement le reflet de caractikes phenotypiques de la plante. Autrement dit, chaque clone
posskde des caracteristiques propres en ce qui concerne la potentialite & fmer NZ, d’absorber
Ndusoloudepmduiresabiomasse.
Dans le cadre de cette experience, les augmen-
tations de rendement consécutives à l’inoculation sont spectaculaires : +200% pour le
clone S, +158% pour le clone a, et +157% pour le clone B. La micropropagation des
clones s&c-tionn6s et leur transfert au champ peut donc amener des gains de produc-
tivité consid&able.
Choix des crit&es de s&ection
L’arrklioration de la plante-hôte doit tenir compte & la potentialitk de fixer N,
et Cgalement des crit&es d’adaptation a l’environnement (conditions climatiques, &a-
phiques, biotiques).
Pf~prï&&? symbiotiques. Les. mesures de biomasse de nodules
constituent M premier Grnent d’apprkmtion. Les techniques de mesure directe de la
fixation de N, étant difficiles à mettre en oeuvre chez les arbres âges, on se contentera
d’individus de quelques années, v@re même de jeunes plants. Les mesures sont réalisées
avec le test de réduction de l’ac&yli?ne (ARA) ou mieux, par les méthodes isotopiques
(Dommergutx et a/., 1988).
Adaptation $ l’environnement. Un des problemes importants
pourlamiseenplaced’un
pmgramme de sklection est de d&erminer, afin de les éviter,
les facteurs Limitant la croissance de la plante tek que la prknce d’azote combine du
sol, les contraintes hydriques, salines, et pathogenes. La t&rance aux herbicides peut
constituer pour ces vegetaux une caractkistique gknétique à introduire quand on sait, qu’en
zone aride, la compktition de la couverture herbacee est teks forte pendant les premières
ann&s de la vie de l’arbre. Parmi les facteurs biotiques, il faut inclure l’influence de
la comp&ition avec les souches sauvages de Rhizobium et de Frankia du sol. Considkant
ce dernier point, il est recommandé (Dommergues et ~2.. 1988) de s&ctionner des plantes-
hôtes capables de noduler seulement avec des souches trks effectives de microorganismes
spécifiques, pour eviter I’&abhssement de la symbiose par des souches sauvages pkentes
dans le sol.
Critéres de sélectlon propres aux arbres
(i) Utilisation de la multiplication v&&ative comme technique &
propagation des plantes sklectionnkes : il apparait peu realiste d’envisager un programme
d’amélioration bas& sur des expkiences d’hybridation sexuée chez les arbres tropicaux
fixateurs de N, qui prtkntent une trks forte hkkozygotie. La multiplication végttative
par les techniques conventionnelles et les techniques de culture in vitro s’avexent être
pour le moment, les seules mttbodes pratiques de propagation des individus s&ctio~&.
On connait la perte progressive de l’aptitude au developpement orthotrope et de la capacitt!
rhizogene des arbres âges, il s’avère donc preferable d’utiliser les techniques de culture
in vitro dont la puissance de multiplication est par ailleurs remarquable.
202
(ii) Sélection précoce des individus : la micropropagation est beaucoup
plus aisée a partir de jeunes individus (meilleure rkactivité, infections réduites...) que sur
des sujets matures. Si on peut détecter. par ailleurs, les carackkistiques physiologiques
des jeunes plants comme dans l’exemple du Cusuurinu (Sougoufara et al., 1989) la s&ec-
tion prkoce des individus réduira la dur& du processus d’amélioration de la plante.
(iii) Di@xltk du Eh& s’es critéres aé sélection : compte tenu des
résultats pr&minaires que nous posskdons, il est n6cessair-e de d&erminer des parambtres
simples comme la biomasse des plantes (exprirntk en N total et tenant compte a la fois
du potentiel fixatem de N, et de l’aptitude de l’arbre a utiliser N du sol) et de la biomasse
des nodules (exprimke en N total) qui semble assez bien corr&.s avec la fixation de
N, (Sougoufara et al., 1989).
Pour identifier les individus caract&ist% par un bon potentiel fixateur de N,, il
est indispensable de faire appel à une méthode kotopique qui permet en outre, de mesurer
l’aptitude des plantes a utiliser l’azote du sol. La s6lection des clones selon leur aptitude
à absorver N du sol, mais aussi selon leur susceptibilité a ce facteur extkieur semble
être un paramere important. Par ailleurs, Mackay et-d., (1987b) ont montre que la perfor-
mance des génotypes d’Alnw glz&wsu inoculés avec des souches de Frunkia &aient dépen-
dants de l’apport en N combiné du substrat, certains clones ttant plus d6pendants que
d’autres.
(iv) L’ éiectrophort?se 6 isozymes pmrmit-elle permettre d’ identjfier
des ghotypes bons jùateurs de IV, ? L’Ctude des profüs électrophorétiques (protéines,
DNA) constitue un outil puissant pour identifier et reconnaître a un stade precoce des
clones (Feldmann et al., 1983), pour contiler l’efficaciu5 d’un brassage génétique et en
déduire le type de fécondation, auto ou allogamie, pour estimer la variabilid génétique
(Vigneron, 1984). Cette technique peut-elle être utilis& pour l’identification de gknotypes
bons furateurs de N, ? Nous sommes tentés d’en douter, car on ne connait pas encore
le support gén6tique de la potentialiti plus ou moins grande de fuer NZ qui est complexe
et vmissemblablement polyg6nique. Par ailleurs, la variation quantitative d’une fonction
génique traduit la variabilité des gènes de r&ulation affectant les quantités de prot&nes
et la variation qualitative correspond en majorité & la variabilité des genes de structure
(Damerval et af., 19%). Une étude &ctrophor&ique pourrait donc, sans doute, identifier
et repérer des génotypes non fkateurs par comparaison avec des individus futant NZ- Par
contre, il semble plus improbable de mettre en evidence une liaison entre une variation
dans le degre de fixation de N, chez un génotype et une variation de la nature des marqueurs
enzymatiques ; les techniques &ctrophor&iques n’appr&ant que des variations quali-
tatives des molkcules (protkii, ADN) analyseeS. C’est d’ailleurs la raison pour laquelle
les caractkres enzymatiques ne sont que tres rarement lies a la valeur agronomique des
individus.
Conclusions
L’amélioration de la fixation biologique de N, à travers la sélection et la gennttique
est un probleme complexe car ce phenomène relève des deux partenaires. Les travaux
de recherche dCvelopp& ces dernières annt%s ont largement contribué à l’amelioration
de nos connaissances sur la complexité des mtkanismes mis en jeu. Les premières recherhes
ont surtout été orientées vers le microorganisme et une première moisson de résultats
t&s importants concernant le Rhizobium a dejà Cté récoltée.
203
Nous avons prkenté ici l’etat de nos connaissances concernant la plante-hôte dont
le rôle joué dans la symbiose commence a être de mieux en mieux connu, même s’il
reste encore beaucoup à faire en gédtique et en biologie mol~ulaire. Le s&ctionneur
peut améliorer efficacement la biomasse produite en utilisant la grande variabilite géné-
tique naturelle des arbres. Selon les rksultats dont nous disposons actuellement, il apparait
que l’objectif des travaux de s&ction à venir serait d’identifier des gknotypes p&entant
les trois caract.&istiques suivantes : grand pouvoir de fixation de N,. faible susceptibilité
de la symbiose a l’egard de N combiné du sol, et accroissement de l’assimilation de
N du SOL Ces trois caractkristiques sont des caractères indQe&am.s sur le plan g&rt%iqué ;
elles peuvent vraissemblablement agir en synergie, et sont pour quelques cas connus, le
Cas~ (Sougoufara et al., 1989) et le soja (Imsande, 1985) corr&es avec une biomasse
importante de l’appareil vég&atif de la plante. Il serait Cgalement utile de sélectionner
des arbres capables de continuer a fixer N, en présence de N combine du sol en particulier
dans les peuplements denses (forêts) où N du sol s’accumule au cours du temps.
Rfifbre nces
ALABARCE, J., 1987 Localisation et developpement des poils absorbants de Casuarina
stictu et Acacia mangium : mise au point d’un dispositif expkimental et r&ultats
prr%minaires. Mém. D.E.S.U. Université Paris VII, 62~.
BARNEiS, DK. JESSEN, D.L., HEICHEL, GI-I. et VANCE, CP. 1986 Selection for
multiple traits increases alfalfa N, fixation and yield. In : Advances in nitrogen fixa-
tion research, 7-9.
BEDMAR, EJ., EDIE, S.A. et PHILLIPS, D.A. 1983
Host plant cultivar effects on
hydrogen evolution by Rhizobim legumhwsarum. Plant Physiol. 72, 1011-1015.
BERTRAND, LJ. et LALONDE, M. 1985 In vitro propagation and nodulation by Frun-
kiu of actinorhizal Russian Olive (Eleugnus ungustijiiliu L.). Plant und Sd. 87, 143-
152
CALDWELL, B.E. 1966 Inheritance of a strain-specific ineffective nodulation in soybeans.
Crop. Sci. 6, 427-428.
CARROLL, B.J., McNEIL, DL. et GRESSHOFF, PM. 1985 Isolation and properties
of soybean (Glycine max) mutants that nodulate in the presence of high nitrate concen-
trations. Pro~. Nat. Acad. Sci. (USA). 82, 41624166.
COUSIN, R., MESSAGER, A., PICARD, J., De LAREMBERGUE, H., et MARGUET,
P. 1987 Variabiliti g&tique pour la fixation de l’azote chez le pois. In : Nutrition
azotée des @umineuses, Versailles, 19-21 Novembre 1985, &l., INRA, PARIS, 333-
342.
DAMMERVAL, C., De VIENNE, D. et HEBERT, Y. 1986 Divergence morphologique
et divergence mesur& par électrophorese bidimensionelle des protemes d&tatur&.s.
Sbminaire sur les distance.s génétiques. M&ibel, 9-16 Mars 1986, lp.
De BLOCK, M., BOTTERMAN, J., VANDEWIEIE, M., DOCKY, J., CTHOEN, V.G.,
RAO MOWA, N., THOMPSON, C., VAN MONTAGU, M. et LEEMANS, J. 1987
Engineering herbicide resistance in plants by expression of a detoxifying enzyme.
Embo .l. 6, 2513-2518.
204
DHAWAN, V. and BHOJWANI, S.S., 1987 In vitro nodulation of micropropagated plants
of Leucaena leucocephala by Rhizobium. Plant ami Soil. 103, 274-276.
DJORDJEVlC, M-A., REDMOND, J.W., BATLEY, M. et ROLFE, B.G. 1987a Clovers
secrete specific phenolic compounds which stimulate or repress nod gene expression
in Rhizobium tri&olii. Embo J. 6, 1173-1179.
DJORDJEVlC, M.A., GABRIEL, D.W. et ROLFE, B.G. 1987b Rhiwbium : t.k refmed
parasite of legume. Ann. Rev. Phytopathol. 25, 145168.
DOMMERGUES, YR., DUHOUX, E. et DIEM, RG. 1988 Les arbres fixateurs de N,
en foresterie et agroforesterie tropicales. Bull. Sc. Bot. Actual. Bot., (sous presse).
DUHOUX, E., SOUGOUFARA, B. et DOMMERGUES, Y. 1986 Propagation of Canrmina
equisetifolia through axillary buds of immature female inflorescences cultured in vitro.
Plant Ce11 Rep. 3, 161-164.
FELDMANN, P., LACROIX, J.L. et GLASZMANN, JC. 1983 L’électrophorèse d’iso-
zymes, outil du dlectionneur et du généticien. Bulletin d’information et de liaison
de I’ZRAT. 1-2, 33-38.
FlLLATI, J., SELLMER, J., McCOWN, B., HAISSIG, B. et COMAI. L. 1987 Agrobacte-
rium mediated transformation and regeneration of Populus. Mol. Gen. Genet. 206,
192-199.
FIRMIN,J.L., WESON, KE.,ROSSEN, L., JOHNSTON, A.W.B. 1986 Flavonoidactivation
of nodulation genes in Rhizobium reversed by other compotmds present in plants.
Nature. 324, 90-92.
GASSAMA, Y.K. et DUHOUX, E. 1986 Micropropagation d’arbres adultes d’Acacia
albida @.eguminosae). In : Les arbres fucateurs d’azote. L’amelioration biologique
de la fertilité du SOL Acte des S&ninaires 17-25 Mars 1986, Dakar, Bd. ORSTOM
PARIS, 216224.
GOVERS, F., GLOUDEMANS, T., MOERMAN, M, VAN KAMMEN, A. et BISSELING,
T. 1985 Expression of plant genes during the development of pea root nodules. Embo
J 4, 861-867.
HUTlNEN, O., HONKANEN, J. et SIMOLA, L.K. 1982 Ornithine and putrescine supported
divisions aud cell colony formation in leaf protoplasts of alders (Ahus glutinosa and
A. incana). Plant Sci. Lett. 28, 3-9.
IMSANDE, J. 1985 Plant genotype and the control of nitrogen fixation on soybean.
In : Nitrogen fixation reseamh progmss. Evans, HJ., Bottomley, P.J. et Newton,
W.E. &l., Martinus Nijhoff Publisbers, 34.
KIDBY, D.K. and GOODCHILD, DJ. 1966 Host influence in tbe ultrastructure of root
nodules of Lupinus luteus and Ornithopus sativus. J. Gen. Microbiol. 45, 147-152.
LAFKIN, PJ., 1987 Somaclonal variation : history, method, and meaning. Zona State
J. Res. 61. 393-434.
LEIGH, J.A., SIGNER, E.R., WALKER, G.C. 1985 Exopolysaccharide deficient mutants
of Rhizobium meliloti that form ineffective nodules. Proc. Natl. AIX& Sci. USA. 82
62316235.
LIE, T.A. 1971 Symbiotic nitrogen fixation under stress conditions. Plant and Soi1 Spec.
vol. 117-127.
h4aGmiser la FBA pour la Production Agricok et Forestihe en Afrique
Analysis of the plant genes involved
in the nitrogen fixing symbiosis.
Sesbania rostrata as a model
De hWUDIE l?
Laboratoire & Biologic ah Sok, ORSi”OM, B.P. I386. Dakm. SMgal.
Summary
During nitrogen ftig nodule formation on ieguminous plants, several plant
genes. so-called nodulin gene& iue specifically expressed. In this paper the current
knowledge on the expression, induction and regulation of noduh genes in the diffe-
rent symbiosises studied so far is summarized, together with main results obtained
in the case of Sesbania rostra&z on nodule-specific plant gene expression.
Introduction
Each stage in the Rhizobium-legume symbiosis is characterized by a series of
developmental events concerning both bacteria and plans resulting in a complex, well
organizedand wellcoordinatedplantorgan.Thishasbeen welldescriifromamorphological
point of view (Bauer, 1981).
By classical genetic experiments, several plant genes involved in nod&ion and
symbiotic nitrogen fixation have been identified in pea, soybean, clover and alfalfa (Nutman,
1981 ; la Rue et al., 1985). Mutations in the plant genome can result in disturbed nodule
development, varying tiom the absence of nodules to the developement of wild-type-
like but ineffective nodules (Vincent, 1980).
Theknowledge of nodule formation has made rapid progress since the rise of molecular
biological techniques. It has been found that a number of genes in both plant and bacterium
are only expressed in nodules. An effective symbiosis is accomplished by different&ion
of bacteria into bactero&ls on the one hand, and differentiation of plant cells into a
root nodule on the other hand. The major part of research activity has concentrated on
the nitrogen fixing bacteria, which are more easily accessible to gehetic manipulation
in comparison with legume plants. Most of the bacterial symbiotic genes (genes for nodu-
lation and nitrogen fixation) have been identified and are well documented. The plant
is equally important, providing the right environment, the energy and using the fixed
nitrogen for its growth and development. Over the last few years, interest in the role
of the plant in the symbiosis has considerably intensified, with the evidence of plant
genome-encoded, nodulespecific proteins, called nodulins &ego&i and Vern& 1980)
::,:
differentially expressed during nodule development.
Compared with other plant differemiation processes, root nodule development is
unique in the involvement of a prokaryote in the induction and control of plant development
_
Developmental program leading to root nodules, as an organ, is as complex as other deve-
lopmental programs in plants and involves numerous genes. Nodulin gene expression is
one of the most specific aspects of nodule differentiation and its study may be a useful
approach to understand root nodule development.
Nodulin definition. By definition, nodulins are plant gene-encoded proteins, which
are found only in root nodules and not in uninfected roots nor in other part of the host
plant (Van Kammen, 1984). Nodulin genes are, by consequence, plant genes exclusively
expressed during the development of the symbiosis. Around 10 to 40 different nodulins
have been found in the various systems studied so far, like soybean (Legocki and Verma
1980). pea (Bisseling et al., 1983), Medicago sativa (Lullien et al., 1987), Phaseohs
vulgaris (Campus et al., 1987), Sesbania rostrata (De Lajudie and Huguet, 1988). Nodulin
genes are differentially expressed during nodule development (Bisseling et al., 1983 ;
Gloudemans et al., 1987). The majority of nodulin genes is expressed around the onset
of nitrogen fixation, like leghemoglobin, and are called nodulin genes. They most p&a-
bly function in establishing and maintaining a proper environment within the nodule that
allows nitrogen fixation and ammonium assimilation to occur. Few nodulins are detec-
table at earlier stages of development when the nodule structure is being formed, which
are called early nodulin genes.
Nodulin functions. In all systems studied so far, the use of mutated or engi-
neered bacterial strains that arrest nodule development at different stages has enabled
to establish parallels events at the levels of nodule structure and nodulin gene.expression
by coupling histological and molecular biological data. Very few nodulins have been iden-
tified Moreover, nodulins must play roles at every stage of nodule developement and
function : (i) for plant-microbe recognition : lectins, enzymes of host origin capable of
degrading the capsular polysaccharides of the bacteria (Batter, 1981) ; (ii) during infection
process : root hair curling, infection thread formation, development of a meristem from
cortex cells, release of bacteria in the meristematic cells and differentiation of host cells
and bacteria into a nitrogen fixing nodule ; (iii) for the transport of metabolites including
organic acids, amino acids and sugars, as well as N, and 0, ; (iv) for the assimilation
of fixed nitrogen produced by bacteroyds ; (v) roles in all the morphological, cytological
and physiological changes that take place in the infected cell, like modification of endo-
membrane systems, increased abundance of free ribosomes. polyribosomes. proplastids
and mitochondria ; carbon, nitrogen and oxygen metabolism changes dramatically during
the differentiation of the me&tern into a root nodule, as indicated by the increase in
activities of several enzymes such as PEP carboxylase, malate deshydrogenase, glutamine
synthetase and uricase (for a review see Verma and Long, 1983).
A few early nodulin genes have been identified. Among them only one has been
well character&d : ENODZ, a nodulin of soybean (Glycine mar.), which has been also
found in other legumes like pea, vetch, alfalfa and clover (Govers et al., 1987). The
strucW features of this nodulin, like its hydroxyproline-rich sequence, suggest that it
is a cell wall protein closely related to the extensin (Nap, 1988).
Late nodulins are more documented The most famous is leghemoglobin, which
constitutes 30.? of the total soluble proteins. Leghemoglobin is a hemoprotein with high
oxygen affinity, which provides high yield of oxygen to bacteroids at a low partial tension,
compatiile with nitrogen fixation. It is generally detectable just before nitrogenase activity
can be measured. Leghemoglobin is a true symbiotic protein since the heme is a product
of the bacterold and protein is plant-encoded. In alI legumes studied so far, there are
several leghemoglobins, encoded by several genes (Marcker et al., 1984). Other late nodulius
have been identifield. One is a soybean nodule-specific uricase, a key enzyme in the
ureide biosynthetic pathway for ammonia assimilation (Bergman ef al., 1983). Two enzymes
of glutamine synthetase. which catalyses the first reaction in the assim&tion of ammonia
into organic nitrogen, are present in nodules of Phaseoh vulgaris, one of which is nodule-
specific (Cullimore et al., 1983). A sucrose synthase, an enzyme which cataIyzes the
degradation of sucrose into fructose and glucose, has been found specific to nodules of
soybean (Thummler and Verma, 1987). Other nodule-specific forms of enzymes that differ
in physical, kinetic and immunochemical properties from the corresponding enzymes in
roots have been found : phosphoenolpyruvate carboxylase, choline kinase, xanthine
deshydrogenase, purine nucleosidase, malate deshydrogenase (Nap, 1988). But there is
still no evidence that they derive from nodule-specific gene expression or from nodule-
specific modifications of root enzymes.
Several other nodulins, whose functions are yet unknown, have been found as-
sociated with peribactero~d membrane, the site of closest interaction and exchanges between
the bacteroRls and the host cell.
Molecular studies on nodulin genes have progressed a lot since the last few years :
gene cloning, organization in the genome, regulation, sequence analysis (for a review,
see Bissehng et al., 1986).
Sesbanla rostmta
Sesbaniu rosrrata is a tropical legume which forms nitrogeu’ fixing uodules on
stem and roots. This makes Sesbania a very efficient system for nitrogen fixation and
a real potential as green manure in tropical agriculture. Two bacterial genera are capable
to induce nodules on S. rostruta : Azorhizobium caulinodans and Rhizobium sp. (Dreyfus
et al., 1988). Due to its particular properties (in vitro nitrogen fixation), A.cudinodam
has been extensively studied. Opposite to this, few data are available on the plant itself
(E3ogmz er al., 1987 ; De Lajudie and Huguet, 1988, 1989). We focussed on S. rosrrutu
to investigate plant genes activated during nodule development and function. S. rostrata
constitutes a good model in such a study since, opposite to that of root nodules, age
of stem nodules can precisely be known : stem nodules are homogenous in age, opposite
to root nodules that form continuously.
210
Our approach consisted in measurin g the level of translatable mRNA present in
each tissue by comparing one-and-two dimensional polyacrylamide gel electrophoretic
patterns of in vitro translation products of poly(A)+ RNA from uninfected stems and stem
nodules, uninfected roots and root nodules.
Plant gene expression in stem nodules appears very similar to that in root nodules,
except some quantitative differences in the intensity of certain spots which could be ex-
plained by the heterogenous age of root nodules. We found 16 nodule-specific polypep-
tides, in addition to around 20 other present at different relative intensities in nodules
and in uninfected tissues. During stem nodule development, majority of nodule-specific
genes are expressed 12 days after inoculation, concommitantly with leghemogt’cibin and
initiation of nitrogen fixation. These could be correlated with transport of substrates
towards bacteroyds and assimilation of fmed nitrogen. Some other genes are expressed
either during the early stages of infection, or transiently, suggesting a role in plant-microbe
recognition, infection process, and nodule morphogenesis. Some others, expressed later
(day 18), could have a role in maintaining nodule structure and function, or in nodule
senescence. All these observations indicate that there is a sequence in the activation of
nodule-specific genes during nodule life, like in other descrii systems.
Among the nodule-specific polypeptides, we identified six in vitro translation
products that cross reacted with a serum anti-leghemoglobin of SesbunLz rostrata. During
nodule development, these leghemoglobin components are differently expressed and their
relative proportions vary : some of them are fmt major to become minor during the
late stages of nodule development.
Root nodules induced either by Rhizobium sp. (ORS 51) or A. caulinodans
(ORS 571) are very similar except for the expression of one more specific gene in the
case of ORS 51. One of the great differences pointed out between stem and root nodules
is the presence of chloroplasts and photosynthesis in stem nodules. We found interes-
ting to compare plant gene expression in stem nodules developed either in the dark or
under light : they show only minor differences ; moreover, stem nodules in the dark
and root nodules show some striking similarities. This, together with other arguments
like the same seven leghemoglobin isomeres (Bogusz er aI., 1987), the same infection
process in stem and root nodules (Ndoye and Truchet, in preparation), or the fact that
root nodules developed under light are photosynthetic, suggests that stem and root nodules
are not of completely different nature.
We used different nitrogen fixation deficient mutant strains of A. caz&odans
ORS 571 and observed that induction of nodule-specific gene expression is not depen-
dant on nitrogen fixation. However the level of expression of leghemoglobin and several
other nodulin genes is lowered when there is no nitrogen fming activity in nodule. This
is in agreement with all systems described so far.
Conclusion
Besides the agronomic aspect, nitrogen fming symbiosises are a tools towards elu-
cidating several fundamental biological problems. They constitute a model for the study
of plant microbe interactions, which are diverse and numerous in nature, especially of
parasitism and pathogenicity.
Opposite to Agrobacterium tumefaciens, the other genus of the Rhizobiaceae
family, which induces tumor formation, a set of undifferentiated cells, Rhizobium is able
211
to bypass the plant defense mechanisms and to induce a de-differentiation of plant cells
and the development of a nodule, a true organ specialized for nitrogen fixation. Therefore,
Rhizobium-Legume symbiosis with Frankia actinorhii plants are an interesting and rare
example for plant molecular biology of the induction by a prokaryote of an organoge-
nesis in a eulmryote.
In less than a decade, our knowledge on plant gene expression during symbiosis
has progressed a lot, and we begin to know more about the complex mechanisms of
symbiosis, bacterial genes implied in the induction of nodulin genes, which is a first
step in identifying bacterial signals to the plant that induces expression of the successive
classes ofnodulin,genes. A better understanding of alI the mechanisms implied in sym-
biosis will help in a better domestication for a practical use in the field
R&&ewes
APPLEBY. CA. 1984 Leghemoglobin and Rhizobium respiration. In : Briggs WR. (ed).
Ann. Rev. Plant Physiol. 35. 443-478.
BAUER, WD. 1981 Infection of legumes by rhizobia. Ann. Rev.. Plant Physiol. 32,407-
449.
BERGMANN, IX, PREDDIE, E. and, VERMA, D.P.S. 1983 Nodulin 35 : a subunit
micase (uricase II) induced and locahzed in the uninfected cells of soybean nodules.
Embo J. 2 (12), 2333-2339
BISSELING, T., BEEN, C., KLUGKIST, J., VAN KIMMEN, A., NADLER, K. 1983
Nodule-specific host proteins in effective and ineffective root nodules of Pisum sa-
tivwn. Embo J. 2, 961-966.
BISSELING, T., VAN DEN BOS. R.C. and VAN KAMMEN, A. 1986 Host specific
gene expression in legume root noduks. In : Nitrogen fixation. Vol 4 : Molecular
Biology. Broughton WJ. and Piihler A. cds. Clarendon Press, Oxford, 280-312.
BOGUS& D., KORTI’, kk, APPLEBY, CA. 1987 Sesbania roslrata stem and root
nodule leghemoglobiu : purification and relationships amongst the seven major com-
ponents. Arch. Biochem. Biophys. 254 (l), 263-271.
CAMPOS,F.,PADILLA,J.,VAZQUEZ,M.,OR-IEGA,J.L.,ENRIQUEZ,C.,andSANCHEZ,
F. 1987 Expression of nodule-specific getles in Phareolus vulgaris L. Plant Mol.
Biol. 2 , 521-532.
CULLIMORE, J.V., LARA, M, LEA, PJ., and MIFLIN, BJ. 1983 Purification and
properties of two forms of glutamine synthetase from the plant fraction of Phaseolus
root nodules. Planta.157, 245-253.
DREYFUS, B.L., GARCIA, JL. and GILLIS, M. 1988 Characterization of Atorhizo&un
caulin0dan.s gene nav. sp. nov., a stem-nodulating nitrogen fixing bacterium isolated
from Sesbania rostrata. Int. J. Syst. Bacterial. 38, 89-98.
GLOUDEMANS, T. De VRIES, S., BUSSINK, HJ., MALIK, N.S.A., FRANSSEN, H.J.,
LOUWERSE. J. and BISSELING, T. 1987 Nodulin gene expression during soybean
(Glycine mux) nodule development. Plant Mol. Biol. 8, 395-403.
GOVERS, F., NAP, J.P., MOERMAN. M. FRANSSEN, HJ., VAN KAMMEN, A., and
BISSELING T. 1987 cDNA cloning and developmental expression of pea nodulin
genes. Plant Mol. Biol. 8, 425-435.
212
LAJUDIE (De), P. and HUGUET, T. 1988 Plant gene expression during effective and
ineffective nodule development of the tropical stem nodulated legume Sesbania ros-
trata. Plant Mol. Bioi. 10, 537-548.
LAJUDIE (De), P. and HUGUET, T. 1989 Mise en evidence de nodulines chez Sesbania
rostra& In : Sesbania rostrata : fundamental research and application to agriculture.
Inpress-
LA RUE, T.A., KNEEN, BE. and GARTSIDE, E. 1985 Plant mutants defective in sym-
biotic nitrogen fixation. In : Analysis of the plant genes involved in the legume-
Rhizobium symbiosis. OECD, Paris, pp. 39-48.
LEGOCKI, R.P. and VERMA, D.P.S. 1980 I&ntification of nodulespecific host pro-
teins (nodulins) involved in the development of Rhizo.bium-legume symbiosis. Cell.
20, 153-163.
LULLIEN, V., BARKER, D.G., DE LAJUDIE, P. and HUGUET, T. 1987 Plant gene
expression in effective and ineffective root nodules of alfalfa (Medicago sativa). Plant
Mol. Biol. 9, 469-478.
MARCKER, A., LUND, M., JENSEN, E.O. and MARCKER, K.A., 1984 Transcrip-
tion of the soybean leghemoglobin genes during nodule development. Embo J. 3 (8),
1691-1695.
NAP, J.P. 1988 Nodulins in root nodule development. Ph. D. Thesis. Wageningen University
(Nederland).
NUTMAN, P.S. 1981 Hereditary host factors affecting nodulation and nitrogen fixation.
In : Gibson A.H. and Newton WE. (eds). Current perspectives in nitrogen fixation.
Elsevier, North Holland, pp. 194-204.
THUMMLER, F. and VERMA, DP.S. 1987 Nodulin 100 of soybean is the subunit of
sucrose synthase regulated by the availability of free heme in nodules. J. Biol. Chem.
262 (30), 14730-14736.
VAN KAMMEN, A 1984 Suggested nomenclature for plant genes involved in nodulation
and symbiosis. Plant Mol. Biol. Rep. 2, 43-45.
VERMA, D.P.S. and LONG, S. 1983 The molecular biology of Rhizobium - legume
symbiosis. Int. Rev. Cytol. Suppl. 14, 211-245.
VINCENT, J.M. 1980 Factors controlling the legume - Rhizobium symbiosis. In : Nitrogen
Fixation II, Newton W.E. and Grme Johnson W.H. eds, University Park Press, Bal-
timore, 103-129.
la FBA pour la production agricole et forestière en Afrique
Sesbania rostrata :
un modèle pour la biologie moléculaire
de la nodulation
TOMEKPE, Kil), HOLSTERS, M.*,
GOETHALS, a(2), VAN DEN EEDE, ~9,
De LAJUDIE, P.(l), TRAN, P.(l)
and DREYFUS, B.0)
(1) : aratoire de Microbiologie des Sols ORSTOM.
B.P. 1386, Dakar, Senegal ;
(2) : Luboratorium $znetikz, Uniwsiti de Gent 9000, Belgique.
Sesbunia rostrata forme des nodules aussi bien sur les tiges que sur les racines.
Sesmicro-symbiotesappam~~8dwxgenresf~ts:AzorhuobiumetRhizobiwn.
Les souches telle que ORS 571 qui induiint des nodules de tige et de racine et
qui fixent l’azote atmosphkrique en culture pure appartiument a la nouvelle espèce
Azwhizobium cauIino&ms~ Les .s~~&e.s isolkes des racines de S. rartrata et qui
forment des nodules effectifs sur d’autres espèces de Se.&mia appartiennent au genre
Rhùobium.
La mutagenèse de la souche d’Azorhimbim caulinodanr ORS 571 a permis d’iden-
tifier deux loti Nodl et Nod2 qui sont impliqu& dans la double nodulation racikre
et caulinaire. L’hybridation molkulaire et l’analyse des s6quences du locus Nodl
ont n%lé l’existence de trois gknes ayant une forte homologie avec les gènes nod
ABC de Rhizobiwn meliloti. L’analyse génktique a mont& que l’expression de ces
g&nes est induite en pr6sence des racines et des primordia racinaires situ6.s sur la
tige de S. rostrata ; elle a par ailleurs permis de d6couvrir. dans le gknome de
ORS 571, un troisAne locus dont la pr6sence est essentielle pour l’induction des
gènes du locus Nodl. Ce troisAne locus semble réguler l’expression des gènes de
nodulation racinaire et caulinaire.
214
Chez S. rostrufu, un mutant incapable de former des nodules de tige a été isolé.
Ce mutant a une croissance atkknne normale mais sa racineprincipale est plus vigou-
reuse et ses racines secondaires moins abondantes que celles de la plante de type
sauvage. Des diffkrences dans la composition des protiines ont été mises en hidence
entre le mutant et la plante normale. Une caracttkisation mol6culaire approfondie
devrait permettre de rechercher chez S. rostrufa les gi?nes impliquks
dans la nodu-
lation caulinaire.
Introduction
Sesbaniu rostrutu est une ltgumineuse que l’on rencontre sur les sols hydro-
morphes de l’Ouest-africain. Son aptitude à former des nodules caulinaires lui confkre
un potentiel de fixation d’azote exceptionnel et lui permet de noduler et de fixer l’azote
sur des sols inondks ou renfermant des quantités elevées d’azote combine (Dreyfus et
id., 1984).
Les nodules caulinaircs se développent à partir de sites pr6d&ermin~ (Duhoux
et Dreyfus, 1982 ; Tsien et ul., 1983). Ces sites sur les tiges sont une caractdristique
particuliière de Sesbuniu rostruta, de certaines espkces d’Aeschynomene et & Neptuniu
(Dreyfus et ul., 1984) ; ils sont constitues d’ebauches racinaires adventices qui se dévelop-
pent en nodules lorsqu’elles sont infectées par des symbiotes spkcifïques (Duhoux et
Dreyfus, 1982).
Les bactéries qui sont associées à Sesbunia rostrutu appartiennent .à deux genres
différents : Azorhizobium et Rhizobiwn.
Les souches d’Azorhizobium cuulinodans du type ORS 571 (Dreyfus et ul., 1988)
induisent des nodules fixateurs d’azote aussi bien sur les tiges que sur les racines. Ces
souches sont capables de fixer l’azote en culture pure. Bien que les souches d’Azorhizobium
soient capables de former des nodules sur un grand nombre d’espèces du genre Sesbania,
elles ne fixent l’azote qu’avec Sesbuniu rostruta.
Les souches de Rhizobium nodulent effectivement les racines de Sesbania rostrata
et des autres espèces du genre Sesbuniu ; la plupart d’entre elles sont capables de former
des nodules fixateurs d’azote sur les tiges de Sesbania romura, certaines y induisent
des pseudonodules ou des nodules ineffectifs et quelques unes ne forment pas du tout
de nodules sur les tiges.
Chez Azorhizobium cuulinoduns, la mutagen&se par transposon Tn5 de la souche
ORS 571 a permis d’identifier et de cloner deux loti Noël et Nota essentiels pour la
nodulation racinaire et caulinaire (Van Den Eede et al., 1987). La technique d’hybrida-
tion moléculaire a permis de mettre en Cvidence une homologie de kquence entre le
locus Noël et les genes de nodulation nodABC de Rhizobium meliloti (Kondorosi et al.,
1984). Les gènes homologues no&IBC de la souche ORS 571 ont ensuite étk isoles et
caracti& par le skquençage du locus NodI.
Le pr6sent article est principalement consacré a l’étude de l’expression des gènes
nod de la souche ORS 571 ; nous nous intkresserons par ailleurs aux résultats prt%mi-
naires concernant la caractérisation d’un mutant de Sesbunia rosrrutu &pourvu de sites
de nodulation de tige.
215
Induction et txigulatlon des genes de nodulatlon de la souche ORS 571
Chez certains Rhizobium à croissance rapide tel que R. meliloti et R. Legumino-
sarum, les gènes nod ne s’expriment qu’en pr&ence d’exsudats racinaires issus de la
plante hôte (John et al., 1985). Différents flavonoïdes isoles a partir de ces exsudats notam-
ment la lutMine chez la luzerne, sont capables d’induire l’expression de ces g&res. Par
ailleurs, il a & mis en 6vidence dans le gtkome de ces Rhizobium, des gènes régulateurs
indispensables à l’induction des genes nod par des flavonoïdes de la plante hôte (John
et al., 1985).
Pour étudier l’induction de l’expression des gènes nod de la souche ORS 571,
nous avons construit plusieurs fusions entre le locus Nodl et le gène LucZ de la
B-galactosidase de E. coli à l’aide d’un transposon (Stachel et al., 1985). Le principe
de ces fusions consiste à insérer dans un fragment d’ADN le gene LacZ sans son pro-
moteur afin que son expression soit sous la seule dépendance de l’expression du gène
dans lequel il s’est inske ; il faut pour cela que la fusion soit orientée dans la direction
uanscriptionnelle et qu’elle soit en phase de lecture. Les fusions construites dans le locus
Nodl étaient ensuite transf&ees dans une population de ORS 571 et testées pour l’ac-
_ tivité de la &galactosidase en présence et en absence de Sesbania rostrata.
Quelques unes des nombreuses fusions testées étaient induites par les fragments
de tige et de racine de Sesbania iostrata. Ces fusions inductibles sont toutes localisées
dans les genes nod et transcrites dans la même direction ; de plus, ces gènes semblent
dépendre de la même unité transcriptionnelle, ce qui indique une organisation sous forme
d’opéron comme pour les gknes noaXBC de R. meliloti.
Pour isoler la substance qui induit l’expression des g&es de la souche ORS 571,
des exudats racinaires de Sesbania rostrata ont été analysés par chromatographie liquide.
Le test d’activite de la l3-galactosidase à partir des diff&entes fractions issues de ces
.
exudats a revele l’existence d’un inducteur de type flavonoïde. Nous caract&isons actuel-
lement cette substance. ParalElement, nous p&parons des extraits de tige afin d’isoler
et de comparer la substance inductrice de tige à celle des racines. Nous avons par ailleurs
teste les fusions inductibles en prksence d’une s&ie de flavonoïdes de provenance com-
merciale ; des niveaux elevés d’induction ont été observés avec la naring&tine (Nar).
Quelques mesures de l’activité de la &galactosidase sont prksentkes dans le Tableau 1.
Etude de la rbgulation de l’induction des génes Nod de la souche ORS 571
Comme l’induction des genes de R. meüloti par la lutt!oline est sous la d@endance
d’un gkne r&ulateur no&, nous avons pense à une situation similaire chez ORS 571.
Deux approches ont été choisies pour rechercher le gène régulateur homologue du g&ne
nodD dans le g&rome de ORS 571.
(i) Après avoir introduit une fusion inductible dam Agrobacterium tumefaciens,
nous avons recherche dans la banque d’ADN de ORS 571 les clones qui permettent son
induction par la naringtnine. Un clone ayant un fragment de 16 kilobases (clone pRG910-
16) a ainsi Cté sklectionné.
(ii) Dans une deuxième approche, nous avons isolé un mutant spoman6 (ORS 571-
IN-) qui ne permet plus l’induction des fusions ZacZ par les fragments de Sesbania ros-
trata. Inocule aux racines et aux tiges de Sesbania rostrata, ce mutant induit des nodules
anormaux et ir&uliers. Lorsqu’on introduit dans ce mutant le clone pRG910-16 qui per-
216
Table 1
: Activités de la
-galactosidase avec et sans induction des
souches ORS 571 portant des fusions 1acZ dans le locus Nodl
ORS 571 avec différentes
Activités de la pgalactosidase
fusions lac Z
(unité!-gai)
-Nar
+Nar
ORS571-Mu63
22
650
ORS571-Mu3
11
205
ORS571-Mu22
1 7
240
ORS571-Mu21
1 8
470
ORS571-T20
1680
Les unités de b-galactosidase ont été mesurêes (Miller 1972) 12 heures
après l'addition de 10pM de naringénine (Nar) à des cultures de ORS571
en phase logarithmique.
met l’induction des fusions lu& dans A. twnefaciem, il est complémenté pour tous ses
phénotypes mutants.
Ces deux approches indiquent la présence sur le fragment de 16 kilobases d’une
fonction r6guMrice ntkessaire à l’expression des gènes de nodulation du locus Nodl.
Pour isoler cette fonction, une matagenèse par transposon a été effectuée dans le clone
pRG910-16 ; une population de clones mutés a été introduite dans le mutant ORS 571-
IN- et les colonies qui n’étaient plus induites ont été sélectionnées. Les fragments mutes
ont kti extraits de ces colonies et sont actuellement cartographiks afii d’isoler les g&nes
impliques dans la rt5gulation de la nodulation.
Caract&isation du mutant de Sesbanh rostrata sans site de nodulation de
tige
Chez Sesbania rosrrafa, les nodules de tige apparaissent sur des sites pl-édéter-
mines qui correspondent a des primordia racinaires. Nous avons rkemment obtenu par
mutagen& chimique un mutant de Sesbania rosfraru ne possédant plus de sites de nodu-
lation sur sa tige.
Ce mutant ne prksente aucune anomalie dans sa croissance a&ienne ni dans la
nodulation racinaire. Il a par contre une racine principale plus vigoureuse et des racines
secondaires moins abondantes que colles de la plante normale. Dans les sols inondes,
217
la quantité d’azote fixé par le mutant décroît alors que celle de la plante normale est
augmentke suite a la prtknce de nodules de tige.
Nous avons compare par &ctrophor&se en une dimension les protéines totales du
mutant et celles de la plante normale. D’après les profds &ctrophor&iques, il existerait
une ou deux prot&nes supplementaires chez le mutant Nous proc&ons actuellement a
une analyse approfondie qui devrait nous permettre de rechercher et d’identifier les fonc-
tions qui gouvernent l’initiation et le développement des sites de nodulation de tige chez
Sesbania rostrata.
Discussion et conclusion
L’analyse génétique et moléculaire de la souche d’Azh%izobim ORS 571 a permis
d’identifier et de caract&iser des gènes impliques dans la nodulation des racines et des
tiges de Sesbania rostrata. Il a été montré que ces gènes ne sont induits qu’en présence
de fragments racinaires ou caulinaires de Sesbania rostruta et que cette induction est
sous la dépendance d’un gbne régulateur. Comme chez Rhizobium meliloti, l’interaction
entre ce r&ulateur et la substance inductrice pourrait determiner l’amorce de la nodulation.
Nous pensons que la capacité de ORS 571 à exprimer pleinement ses genes md dépend
de l’affinité entre la molécule inductrice et ce gène r&ulateur~ Il se pourrait que les g&nes
nod des Rhizobium de racine soient induits par une substaricc pr6sente dans les racines
et pas dans les sites de nodulation de tige.
La découverte de quelques souches de Rhizobium qui n’induisent que des nodules
non fixateurs d’azote sur les tiges de S. rostrata suggère que Azorhizobium ORS 571,
contrairement a ces Rhizobium, possederait des gènes ou des mkanismes spécifiques de
la fixation ou de la transformation de l’azote atrnosphkrique dans les nodules de tige
(g&nes I@, ntr et&). La mise en évidence et l’étude de l’expression de nombreux gknes
fut et nif d’une part (Danèfle et al., 1987 ; Pawlowski et al., 1987 ; Kaminski et al.,
1988) et 1’6tude des mutants de la r@lation de ces gènes d’autre part (de Bruijn et
al., 1988) devraient permettre de rechercher et de comprendre les mkcanismes spécifiques
de la fixation d’azote dans les nodules de tige.
Parrallèlement, de nombreuses recherches ont été entreprises sur S. rostrata afii
de mieux comprendre l’initiation des 6vènement.s prkoces de ht nodulation et le déve-
loppement des nodules chez la plante hôte. Une &ude r6cente (de Lajudie et Huguet,
1988) a rt%le une différence entre les protiines végétales induites dans les nodules racinaires
par la souche ORS 571 et celles induites par une souche de racine. Bogusz et al. (1987)
ont par ailleurs montre qu’il n’y avait pas de difft%e-nce signifîcative entre les composants
de la leghemoglobine extraits des nodules racinaires et ceux issus des nodules caulinaims.
Les travaux de Trinh (com. pers.) ont permis de mettre au point un système de nodulation
in vitro pour suivre avec précision les difftkentes Ctapes de l’infection a la morphogenkse
des nodules
En ce qui concerne le mutant de S. rostrata sans site de nodulation caulinaire,
nous espkons en continuant sa caracttisation mol&xlaire, parvenir à blucider les mdcanismes
de l’induction et du développement des nodules de tige et a isoler les gènes correspondants.
Ces mécanismes seront n&essaires pour crkr des nodules akiens sur d’autres espi?ces
et particuli~ement celles du genre Sesbania utilisées en agriculture et en agroforesterie.
L’association S. rostrata-Azorhïzobiw et Rhizobium apparait comme l’une des
interactions symbiotiques pr4sentant un interêt majeur pour les études moléculaires. Le
218
développement des recherches fondamentales initiées dans plusieurs laboratoires sur cette
association aura certainement des retombées sur la connaissance de l’ensemble des inter-
actions symbiotiques. Par ses caractkistiques particulieres, ce système apparait de plus
en plus comme un modele pour la biologie moléculaire de la fixation d’azote.
Réfkrences
BOGUSZ, D., KORTT, A.A. and APPLEBY,C.A. 1987 Sesbania rostruta root and Stern
nodules leghemoglobins : purification andrelationship among the seven majorcomponents.
Arch. Biochem. Biophs. 254, 263-211.
De BRUlJN, FJ., PAWLOWSKI, K., RATET, P., HILGERT, U. and SCHELL, J. 1987
The unusual symbiosis between the nitrogen fixing bacterium ORS 571 and its host
Sesbuniu rostrata : regulation of nitrogen fixation and assimilation genes in the free-
living versus symbiotic state. In : Molecular genetics of plant-microbe interactions
(D.P.Y. Verma and N. Brisson, Eds).
De LAJUDLE, P. and HUGUET, T. 1988 Plant gene expression during effective and inef-
fective nodule development of the tropical stem nodulated legume Sesbunia rostruta.
Plant Mol. Biol. 10, 537-548.
DENEFLE, P.. KUSH, A., NOREL, F., PAQUELIN, A. and ELMERICH, C. 1987 Bioche-
mical and genetic analysis of the nijHDKE region of Rhizobium ORS 571. Mol. Gen.
Genet. 207, 280-287.
DREYFUS, B., ALAZARD, D. and DOMMERGUES, Y.R. 1984 Stern-nodulating rhizobia.
Current perspectives in microbial ecology (MJ. Klug and C.A. Reddy Eds.) p. 161.
DREYFUS, B., GARCIA, J.L. and GILLIS, M. 1988 Characterization of Azorhizobium
cuulinodans gen., sp. nov., a stem nodulating nitrogen fixing bacterium isolated fiom
Sesbania rostrata. Znt. J. Syst. Bacteriol. 38, 89-98.
DUHOUX, E. and DREYFUS, B. 1982 Nature des sites d’infection par le Rhizobium
de la tige de la légumineuse Sesbania rostrutu . CR. Acad. Sci. Paris. 294, 407-
411.
JOHN, M., SCHMIDT, J., WIENEKE, U., KONDOROSI, E., KONDOROSI, A. and
SCHELL, J. 1985 Expression of nodulation gene nodC of Rhizobium meliloti in Esche-
richia coli : role of rhe nodC gene product in nodulation. Embo J. 4, 2425-2430.
KAMINSKI, PA., NOREL, F., DESNOUES, N., KUSH, M., SALZANO, G. and ELME-
RICH, C. 1988 Characterization of the frxABC region of Azorhizobium caulinoduns
ORS 571 and identification of a.new nitrogen fixation gene. Mol. Gen. Genet. 214,
496-502.
KONDOROSI, E., BANFALVI, Z.S. and KONDOROSI, A. 1984 Physical and genetic
analysis of a symbiotic region of Rhizobium meliloti : identification of nodulation
genes. Mol. Gen. Genet. 193, 445-452.
PAWLOWSKI, K., RATET, P., SCHELL, J. and De BRUIJN, F.J. 1987 Cloning and
characterization of ni$A and ntrC genes of the Stern nodulating bacterium ORS 571,
the nitrogen fixing symbiont of Sesbuniu rostrutu : regulation of nitrogen futation
(nit) genes in the free living versus symbiotic state. Mol. Gen. Genet. 206, 207-219.
219
STACHEL, S., AN, G., FLORES, C. and NFSlXR, E.W. 1985 A Tn3ZucZ transposon
for the random generation of B-galactosidase gene fusions : applications to the analysis
of gene expression in Agrobacterium. Embo J. 4, 891-898.
TSIEN, H.C., DREYFUS, B.L. and SCHMIDT, EL. 1983 Initial stages in the morphoge-
nesis of nitrogen fming stem nodules of Sesbania rostrata . J. bacteriol. 156, 888-
897.
VAN DEN EEDE, G., DREYFUS, B., GOETHALS, K., VAN MONTAGU, M. and
HOLSTERS, M. 1987 Identification and cloning of nodulation genes kom the Stern
nodulating bacterimn ORS 571. Mol. Gen. Genet. 206, 291-299.
AMELIOR.ATION
DE LA FIXATION SYMBIOTIQUE
DE L’AZOTE
PAR LES PRATIQUES CULTURALES
~Uaxbnisez la FBA pour la Production Agricole et Forestière en AfXque
Amélioration de la Fixation Biologique
de 1’Azote (N,) par les pratiques culturales
GANRY F.
Ingttkiew aé Recherche à IRAWIRAD
L’augmentation de la production d’une 16gumineuse annuelle, telle que l’arachide, fait
appel a des pratiques culturales appropriees qui doivent aboutir à une augmentation du
rendement 21 l’hectare et à un maintien du niveau de fixation symbiotique de l’azote (NJ
le plus élevé possible. C’est en vue d’améliorer la quantité d’azote frxé par la l&umineuse,
qui se differencie du premier objectif (l’augmentation du rendement) en cc sens qu’il
place la fixation de l’azote en priorite, que nous raiso~exons les diff&wes techniques
CUltieS.
La diminution de la quamit. d’azote fmt5 par une culture de 1Qurnineuse r&wlte de
l’effet de contraintes qui affectent M des paramètres suivants ou les deux à la fois :
(i) la croissance vegetative avec une action indirecte sur la fucation de l’azote et (ii) la
fmtion de l’azote avec ou sans action dépressive sur le rendement. Pour lever ou éviter
ces diff&entes contraintes, les agronomes doivent être en mesure de proposer des techniques
culturales applicables dans l’agriculture du pays, et capables de maintenir ou de redmsser
le niveau de fixation de l’azote dans les conditions d’environnement de la légumineuse
cultiv6e. Ceci pose le probleme de l’applicatioon des m6thodes de quantifîcation de la
fixation de l’azote in situ. Ces contraintes ont t% relativement bien analysées dans les
principales zones écologiques du S6n6gal pour l’arachide et le soja (Dommergues et Ganry,
1986 ; Ganry, 1987).
Les m&hodes de quantitïcation in situ se r6sument a deux types : les méthodes fondees
snr l’analyse chimique de l’azote seulement, et les m&.hodes fondees sur l’analyse chimique
et le marquage isotopique. Les premi&res prennent en compte sur une p6riode donn&
de t ann&s la variation du stock d’azote total du sol (AN=NO- NJ ainsi que les inputs
et outputs connus d’azote dans I’agrosysteme : elles permettent seulement de savoir si
le système est le siege de pertes ou de gains d’azote, les gains d’azote (inputs non connus)
Ctaut dus essentiellement a la fixation de l’azote. Les secondes méthodes permettent la
222
quantification pr6ci.w de la fixation de l’azote dans les parties ah-iennes et partant, une
quantikation de cette fixation dans les racines.
Analyse des contraintes pouvant affecter la quantit6 d’azote fixe
Ces contraintes peuvent affecter directement la fixation de l’azote, ou indirecte-
ment celle-ci par l’intermediaire de la plante hôte, essentiellement par r&duction de l’en-
racinement et de l’activit6 photosynthétique. Des techniques culturales appropriees devront
être appliquees pour lever ces contraintes.
La contrainte climatique
Cette contrainte est essentiellement hydrique. Trois voies d’étude sont poursuivies
pour résoudre le problème du deficit hydrique dont l’importance au Sénegal est crois-
sante du sud ou nord :
(i)
Choix des variétés à cycle végétatif adapté à la longueur utile de la
saison des pluies (Dancette 1979) : Il s’agit là d’une action indirecte
sur la fixation de l’azote.
(ii) Amélioration de la résistance a la &cheresse de la I@umineuse, d’autant
plus importante que la du& du cycle est raccourcie (Bockelee-Morvan
er af., 1974 ; Gautreau, 1977) : Il s’agit la également d’une action
indirecte sur la fixation de l’azote.
(iii) Amélioration des techniques permettant une meilleure alimentation hy-
drique (Kuo and Boersma, 1971 ; Sprent, 1972 ; Pankhurst and Sprent,
1975 ; Pate 1975) : Il s’agit Et d’actions directe et indirecte sur la
fixation de l’azote. A titre d’exemple, nous analyserons l’effet d’un
stress hydrique en cours de cycle sur une arachide. On sait d’une part
qu’un stress hydrique en saison de culture provient davantage de l’épui-
sement du stock d’eau disponible dans le sol que de l’accroissement
de la demande Cvaporative, de sorte que ce stress va affecter d’abord
la zone racinaire et les nodosit&. D’autre part, on sait que lorsque les
conditions environnememales ne permettent plus a la plante de satis-
faire à la demande évaporative, elle peut s’adapter en fermant ses sto-
mates mais ce faisant, elle interrompt sa photosynthèse et prive
particuli&rement de chaines carbonn& les 16gions les plus 6loigneeS
de la source, notamment les nodosités. Ainsi, lorsque des symptômes
de fl&rissement temporaire apparaissent au champ a la p&iode la plus
chaude de la journée, la fixation de l’azote a deja 6t6 fortement affectée
par carence hydrique et par carence énergétique. Les interventions
peuvent poursuivre deux objectifs complementaires : l’accroissement
du stock d’eau dans le sol en supprimant les pertes par ruissellement,
par évaporation et par transpiration des adventices et l’amelioration de
la capacité! de la plante a utiliser cette eau par amélioration du système
racinaire
223
La contrainte fertilit6
En zone soudano-sahélienne, on observe depuis une quinzaine d’ann6es une baisse
inquiktante des rendements de l’arachide, imputable en grande partie a la baisse de la
fertilité des sols. Nous savons que le phénomene de dQ5rissement de l’arachide est cause
par une in&Ganw de la fixation de l’azote liée vraisemblablement à une dégradation
de l’environnement edaphique qui n’est pas sans rapport avec cette baisse de fertilité.
Sans prbendre que toute la baisse de fertilité soit due aux pertes min&ales, on peut les
considerer comme essentielles (Blondel, 1970 ; Pieri. 1976 ; Wey et al., 1977).
La contrainte parasitisme
Elle est la charnière entre la contrainte climatique et la contrainte «techniques cul-
turales~. En effet, la virulence d’une maladie donnt% dépend a la fois du climat et des
techniques culturales adopt&s sur l’exploitation (rotation, association d’esp&es ou de variM.s)
favorisant le cycle des parasites. Par exemple, le climat humide de la Casamance dans
le sud du SCn&al favorise la rosette de l’arachide ; de plus, les semis tkhelo~és qui
y sont possibles, accentuent la maladie.
La contrainte *date de semis.
La n&essit& & semer à la date voulue se heurte souvent à un probleme de calendrier
cultural sur l’exploitation ou le manque de mat&iel et de Mail de traction, conduisant
a reporter une partie des semis au delà de cette date la plus favorable. Il en résulte des
pertes de rendement dues essentiellement à un defaut d’enracinement Les pertes de rende-
ment sur l’arachide ont pu être estim6e.s dans le Centre Sud Sénégal (Sine-Saloum) à
50 kg/hzv’jour de gousses dans la première quinzaine de juillet (Tourte, 1974)
La contrainte microbiologique
Parmi les microorganismes impliqués dans la fixation de l’azote des l&umineuses,
il y a bien évidemment les rhizobiums, partenaires obligatoires dans la symbiose, mais
aussi, nous faisons l’hypoth&se que dans les conditions édaphiques des sols pauvres 6tu-
diés, les champignons endomycorhiziens à vésicules et a arbuscules (h4VA) pourraient
jouer un r6le important (Dommergues, 1977).
L’intervention sur la biomasse rhizobiale du sol peut être r6alis6.e pour trois
raisons :
a) Les sols n’ont pas de rhizobiums : l’inoculation est alors nécessaire dans
ces sols ; c’est le cas du soja au St%gal 6tudi6 par Wey (1983).
b) Les sols sont d&àvorables a la nodulation et a la fixation de l’azote bien
que poss&ant encore des rhizobiums : on doit y remedier afin d’bviter
le d@%issement des cultures de 1Cgumineuses ; c’est le cas du d@&is-
sement de l’arachide au Sénégal sur laquelle l’inoculation a Cté 6tudiée
par Wey (1974) et par Drevon (1981).
c) Les sots sont bien pourvus en rhizobiums : dans ces sols où la fixation
de l’azote est normale, on peut theoriquement améliorer I’efficience du
syst2me fixateur en substituant à ces souches autochtones une autre souche
224
à condition de résoudre le problème de la comp&ition : c’est le cas de
l’arachide au Sénégal &udiee par Ganry et Wey (1975) et par Ganry
et NDiaye (1977).
L’intervention sur fa biomasse mycorhizienne peut être envisag6e dans les sols
où le potentiel d’infection mycorhizienne est réduit (Ganry et ~2. 1982. 1985).
L’Appllcatlon de techniques culturales approprkes permet de rbdulre les
effets néfastes des dlffbrentes contraintes
Ces techniques sont essentiellement la culture de vari&% a cycle v6ggétatif adapte,
le semis a une date optimale, la fertilisation minkale, le travail du sol, l’amendement
organique et cakique, l’inoculation avec des rhizobiums et éventuellement des MYA, la
~tatiOll.
Comment mesurer quantitatlvement l’effet des techniques culturales sur la
fixation de l’azote et plus g6nbalement sur I’économle de l’azote ?
Comme nous l’avons tkqué dans l’introduction, seules les méthodes isotopiques
(traçage isotopique et abondance naturelle) permettent d’estimer avec une relative prkision
in situ, la fixation de l’azote. Pour atteindre l’objectif visé d’économie des engrais azotes,
il faut considérer le système de culture dans lequel participe une légumineuse et mettre
en oeuvre la démarche suivante : utiliser au maximum la legumineuse pour enrichir le
pool d’azote du sol et d’autre part déplacer le flux de consommation de l’azote du sol,
de la @umineuse (en Lui assurant cependant le minimum vital pour sa productiviti) vers
la c&eale qui doit exploiter au mieux ce pool mobilisable d’azote du sol en vue d’écono-
miser les engrais azou%. Par ailleurs, il faut chercher à tiuire le plus possible les pertes
d’azote provenant de l’engrais. Cette stratkgie s’applique au systkme de culture composé
de la c&t!ale et de la 16gummeuse en rotation.
Pour d&ïnii ce systkme de culture kconome en engrais azoté, en l’occurrence la
rotation légumineuse &dale, nous devons alors poursuivre trois objectifs :
a) Pour la IQumineuse, il faut réduire le plus possible la fourniture d’azote
par le sol durant la phase où elle est capable de fixer l’azote (NJ.
b) Pour la c&%le, il faut : (i) accroître la fourniture d’azote par le sol
incluant l’azote fixé (en vue de rkhtire l’apport d’engrais azoté) et (ii)
r6duù-e les pertes d’azote engrais.
Cette démarche mkthodologique requiert la quantification des flux d’azote dans
l’agrosystéme, en particulier les pertes d’azote engrais, la fixation de l’azote et la quantid
d’azote sol prélevé par la plante. Les methodes isotopiques 15N permettent ces trois
quantifications. Au Sénégal, cette démarche a 6te appliquke aux systèmes mil-arachide
et maïs-soja (Ganry, 1987).
Conclusion
En guise de conclusion, nous prendrons l’exemple de la culture arachidikre.
L’analyse des contraintes agronomiques et climatiques pouvant affecter la produc-
tion arachidière met en relief : (i) l’importance de la contrainte hydrique, a caractkre
225
non prévisible. avec comme premii%e conskquence un defrcit prononce de la fixation de
l’azote ; (iii la contrainte fertilité, à carackre pr&isible et graduellement menaçante en
culture traditionnelle : la manifestation la plus importante de cette baisse de fertilite est
la carence en azote, attribuée aux faibles rkserves azotées des sols sableux tropicaux
(faible teneur en matik organique) mais aussi et surtout à l’acidit6 du sol dont l’effet
inhibiteur sur la fixation de !’ azote en de@ de pH = $5 se manifeste aussi par une carence
en azote.
Il apparait donc que dans des conditions de stress hydrique et d’acidité du sol
(pH = 55). la quantiti d’azote fixe par une culture d’arachide est doublement r&h&
en raison de l’action conjugwk de ces facteurs d@ressifs sur la croissance vegetative
et sur la fixation de l’azote.
Les interventions culturales appropriees devront d’abord rechercher des types de
plantes aptes a n%ister ou a Cviter la sécheresse et éventuellement r&ister à la toxicité
ahrminique. Une fois le choix variCtal fait, un certain nombre de techniques devront être
appliquées au sol : travail du sol et amendement et éventuellement inoculation avec des
microorganismes. Il importe de savoir maintenant quel est le degr6 d’application et d’ap-
plicabilid de ces techniques en milieu paysan .
En ce qui concerne le choix des vari&% adaptks et de la date de semis la plus
prkoce, il ne soulkve pas de problemes majeurs car chaque agriculteur mesure bien les
risques qu’il encourt en ne respectant pas ces choix.
En ce qui concerne les techniques de labour, d’amen&ment organique et d’amen-
dement calcique, leur application est capitale puisqu’elle conditionne la productivité du
sol et le maintien du patrimoine foncier. Malheureusement, cette application se heurte
a de nombreuses difficultks. En effet, trks peu de paysans labourent leurs champs en
fm de cycle cultural et y apportent de la matik?re organique par enfouissement des pailles
ou apport de fumier. Ce probleme doit être examine dans le cadre du systkme agraire
ayant pour objectif l’int&ation du troupeau dans l’exploitation.
Quant aux diftïcultks d’amendement calcique des sols, elles sont dues à un pr-
oblème de coût et d’approvisionnement en chaux, qui, au SCnégal, malgrk la pr6sence
d’une usine de chaux demeurent un obstacle a son utilisation par l’agriculteur.
En ce qui concerne l’inoculation par le rhizobium, elle n’a pas dkpass6 le stade
exp&imental pour l’instant au SénCgal. Des recherches sont menées en vue de son ap-
plication par l’agriculteur. Neanmoins, nous consid&ons que l’inoculation ne devrait être
envisagee, dans le cas de l’arachide que lorsque les techniques culturales prkonisks
auront Cti rationnellement appliquées.
Rc!f&ences
BLONDEL, D. 1970 Relation entre le « nanisme jaune » de l’arachide en sol sableux
(Dior) et le pH. DKrnition d’un se& de l’activit6 du Rhizobiwn. Agron. Trop. 25,
589-595.
BOCKELEE MORVAN, A., GAUTREAU, J., MORTREUIL, JC. et ROUSSEL, 0. 1974
R&&ats obtenus avec les variCt& d’arachide rksistantes à la sécheresse eu Afrique
de I’0ue.s~ Oléagineux. 29, 309-314.
DANCE’ITE, C.. 1979 Agroclimatologie appliquée à l’économie de l’eau en zone
Soudano-Sah&ienne. Agron. Trop. 34, 225235.
226
DOMMERGUES, Y. and GANRY, F. 1986 Biological nitrogen fixation and soi1 fertility
maintenance. In : Management of nitrogen and phosphorns fertiliiers in sub-saharan
Afiica. Uzo Morwunye, A. and Paul Vlek L.G. (eds.), Martinus Nijhoff Publishers,
Dordrecht, pp. 95-115.
DREVON, JJ. 1981 A deficiency of the symbiotic nitrogen fixation in a dry tropical
agrosystem. The nitrogen chlorosis of groundnut (Arachis hypogaeu) in Senegal.
In : Nitrogen cycling in West Africal ecosystem. Rosswall, T. (ed.), Proc. of a work-
shop held at BTA Ibadan Nigeria 11-15 Dec., Stockholm, SCOPE-UNEP, pp: 209-
213.
GANRY, F. 1987 Application de la méthode isotopique dans la recherche de systèmes
culturaux c&%le-légumineuse économes en azote In : Evaluation des réserves en a-
zote utilisables par les plantes dans les agrosystèmes tempérés et tropicaux~. Egoume-
nides, C. (ed.), GEMOS (Groupe d’étude de la matiere organique des sols), 8ème
r&mion Montpellier les 6 et 7 mars 1987, IRAT - CIRAD, Montpellier, France.
GANRY, F. et NDIAYE, M. 1977 Action de l’inoculation et action de l’@mdage foliaire
d’une solution nutritive (N-P-K-S) sur la fixation symbiotique et le rendement de
l’arachide. Coordinated research on the use of isotopes in fertihxer efficiency studies
on grain legumes. Research contract No RC/1296SEN of the joint FAOZIAEA Division.
In : Report presented at workshop held in Vienna in December 1977, 16~.
GANRY, F. et WEY. J. 1975 Cordinated research program on the use of isotope in studies
on biological dinitrogen fixation. Research contract No RC/1296 - SEN of the joint
FAOZIAEA Division. In : Report presented at workshop held in Vienna. « Action
de l’inoculation et de la fumure azotée sur le rendement et la fixation symbiotique
de l’arachide ». ISRAZAIEA, 20 p.
GANRY, F., DIEM, H.G. and DOMMERGUES, YR. 1982 Effect of inoculation with
Glomus mosseae on nitrogen fixation by field grown soybean. Plant and Soi1.68,421-
429.
GANRY, F., WEY, J., DIEM, H.G. and DOMMERGUES, YR. Inoculation with Glomus
mosseae improves N, fixation by field-grovn soybeans. Biol. Fert. Soils.1, 15-23.
GAUTREAU, J. 1977 Niveaux de potentiels foliaires intervari&aux et adaptation de l’ara-
chide a la sécheresse au Sedgal. Oleagineux. 32, 324-332.
KUO, T. and BOERSMA, L. 1971 Soil water suction and root temperature effects on
nitrogen fixation in soybeans. Agron. J. 63, 901-904.
PATE, J.S. 1975 Physiological studies on the reaction of nodulated legumes to environ-
mental stress. In : Nitrogen fixation and the biosphere. Stewart, W.D.P. (ed). Cambridge
University press.
PANKHURST, CE. and SPRENT, J.U. 1975 Effects of water on the respiratory and
nitrogen fming activity of soybean mot nodules. .Z. Exp. Bot. 26, 287-304.
PIERI, C. 1976 I) L’acidification d’un sol dior cultivé du Sénegal et ses consequences
agronomiques. II) L’acidification des terres cultivées exondées au S&rnégal. Agron.
Trop. 31, 339-368.
SPRENT, J.U. 1972 Nitrogen fixation by legume subjected to water and high stresses
In : Symbiotic nitrogen fixation in pIants. P.S. Nutman (Ed.) Cambridge University
Press, p 385-405
227
SPRENT, J.U. 1972 The effects of water stress on nitrogen-fixing nodules 4. Effects on
whole plants of Vicia faba and Glycine max. New Physiologist. 71, 603-611.
SIBAND, P. et NICOU, R. 1975 Réflexion d’ensemble sur le problème de l’engrais minéral
dans le bassin arachidier. Rapport ronéo. ISRA-CNRA Bambey, 111 pages.
TOURTE, R. 1974 Les Recherches de I’IRAT au Setrégal. Synth&se 1973. Rapport ronéo
ISRA-CNRA de Bambey, 125 p.
WEY, J. 1974 Inoculation bacterienne des légumineuses au Sénégal. Memoire ENITA
Dijon, France, 47 p.
WEY, J., SIBAND, P., EGOUMENIDES, C. et GANRY, F. 1987 Essai de régénkration
d’un sol de la zone arachidikre du Centre-Nord du Sénégal. Agron. Trop. 42, 2%
268.
WEY, J. 1983 (RI) Inoculation du soja par Bradyrhizobium japonicwn. (II) Production d’ino-
culum : mise au point d’une mtthode simplifitk Agron. Trop. 38, 129-136.
hhhiser la FBA pour la production agricole et forestihe en Afrique
Response of Phaseolus vzdgaris I+ to inoculation
with Rhizobium and fertilization with nitrogen
and phosphorus in Northern Tanzania.
AMIJEE, FP’, EDJE, O.T.@‘,
KOINANGE, E.K.c3’, BITANYI, H.F.f4’,
BRODRICK, S.JP), and GILLER, ICE.“’
(I) : Wye College, Univenify of London, Wye.
word, Kent, TN25 5AH. U.K.
(2) : SADCC-CLAT Regicmal Bean Programme.
P.O. Box 2704. Arushn. Tanzania.
(3) : TAR0 Lyanumgu, P.O. Box 3OM. Uoshi, Tanzania.
(4) : INSRC, P.O. Box 4302, Dar es Salaam. Tanzania
Abstract
In this study we aim to identify soil factors limiting nodulation and bean (Ph-
seoZus vulgaris L) production in Nothern T anmnia Investigations car&d out in
1987 indicated that large numbers (10 cells g-’ soil) of Rhizobium legumimsarm
bv. phaseoli were present in most soils, and rwponse to inoculation alone was poor.
Chemical analyses of soil samples collected from these arens showed that pH, or-
ganic matter content and cation exchange capacity were reasonable, but amount of
soil available P was very small (l-2 mg P kg-‘). In 1988, results based on two on-
farm experiments showed that inoculation with Rhizobiwn (strain CIAT 899) and
application of triple supcaphosphate significantly increased nodulation and yield.
Application on Minjingu rock phosphate also increased nodulation and yield but to
a much lesser extent-
2 2 9
Introduction
Bean (Phaseolus vulgaris L.) is an important grain legume widely grown in the
highland areas (1000 meters above sea level) of East and Central Africa The main bean
producers are small farmers for whom bean consumption contributes up to 50% of dietary
protein (Allen, 1986). The majority of these farmers grey the crop on low fertility soils
with no fertilizer input. Generally under these conditions, effective nodulation with Rhizo-
_ biwn is rarely observed and amount of atmospheric nitrogen fixed is probably small.
Many factors (e.g. genetic variability, eclaphic and agronomic factors) are known to-affeci
nodulation in beans (Graham, 1981). Our objective has been to identify soil factors which
limit nodulation and nitrogen fixation in the field. The investigations reported here were
carried out during two growing seasons (March to August 1987 and 1988) in Northern
Tanzania, one of the principal areas for bean production.
Materials and methods
Experiment I (1987)
On station, experiments on two contrasting soil types (volcanic soil at &lien, and
brown earth at Lambo) were done to test response to nitrogen application and Rhiwbium
inoculation. Two bean varieties of large seeded dwarf (Type 1) were chosen : Canadian
Wonder, a popular variety grown throughout Northern Tanzania, and Lyamungu 85, a
variety recently released by the Tanzania Agricultural Research Organisation (TARO)
Bean Programme. Nitrogen was applied at 0 or 30 kg N ha-’ as (NHJ,SO, at time of
planting, and seeds were uninoculated or inoculated (106 cells per seed) with peat-based
mixed Rhizobium inoculant of strains CIAT 632 and 899. A basal dose of
60 kg P,O, ha-’ as triple superphosphate (TSP) was applied to all treatments. Each treat-
ment had five replicates, and a split plot design was used to reduce the risk of conta-
mination between uninoculated and inoculated treatments. Plot size was 4 m x 6 m con-
taining eight rows 0.5 m apart with one seed sown at 0.1 m interval to give
200,000 plants hx’. Nodulation (nodule number, nodule diameter and score for effec-
tiveness) was recorded up to flowering, and seed yield was measured at harvest. Bean
fields owned by research stations and farmers were visited to assess nodulation and plant
vigour. Soil samples were collected for chemical analyses (MAFF/ADAS, 1986) to deter-
mine variation in soil fertility and to identify factors limiting the establishment of ef-
fective nodulation.
Experiment II (1988)
Two sites @rente, Lushoto and Tropical Pesticide Research Institute (TPRI), Amsha)
which had no record of previous fertilizer application were selected for an experiment
to test response to Rhizobium inoculation, nitrogen application and phosphate fertilization.
Lyamungu 85 variety was either uninoculated or inoculated with peat based inoculant
(CIAT 899) to give 106 cells per seed and N applied at 0 or 30 kg ha-’ as nitrochalk
(21% N). Two sources of P were used ; Minjingu rock phosphate (MRP) containing ap-
proximately 30% PZO, or triple super phosphate (TSP) containing 45% PZO, at equivalent
application rates of 0, 60, 120 or 180 kg PZO, ha-‘. The rock phosphate is a sedimentary
guano deposit located at South West of Arusha. N. Tanzania (Scaife, 1968). and has
230
been recommended as an alternative source of P costing a thiid of the price of TSP.
A split plot with two replicates was used to avoid contamination between uninoculated
and inoculated treatments. Plot size was 4 m x 4 m containing eight rows 0.5 m apart
with two seeds sown at 0.2 m interval to give 200,000 plants ha-‘. Measurements included
nodule number before flowering, vesicular arbuscular mycor&zal infection (Phillips and
Hayman, 1970) at pod development, and final seed yield at harvest. Soil pH and bicar-
bonate soluble P (available P) were also determined at final harvest Rainfall was poor
at TPRI, and moisture stress experienced after flowering reduced final yield and may
offset potential effects of the treatments.
Results and discussion
Experiment I (1987)
At &lien, nodulation was excellent (up to 100 nodules per root system, 2-3 mm
in diameter and effective, i.e. pink in colour) in all treatments including uninoculated
plants which nodulated well with indigenous Rhiwbim Earlier on, nodulation was sup-
pressed in treatment receiving N. However this effect was not found at later harvests.
Mean seed yield was 1350 kg ha-’ with no increase due to N application or inoculation,
suggesting that some other factors were limiting yield. A nutrient disorder resemblig po-
tash deficiency was observed, and work is in progress to identify the cause. At Lambo,
nodulation was poor in all treatment (up to 20 nodules, 1 mm in diameter and white
in colour). Mean seed yield was 840 kg ha-’ and did not differ with treatments. This
site experience prolonged moisture stress which would have offset any potential effect
of N application or inoculation. In 1988, Lamb0 received good rainfall, and nodulation
was exceellent.
Enumeration of Phaseolus rhizobia (Vincent, 1970) showed that up to 104 cells
g1 soil were present at both sites. Singleton and Tavares (1986) have shown that unless
the native population of Rhiwbium is less than 102 cells g’, response to inoculation is
not common. Hence it is not surprising to record no response to inoculation in our studies
at Salien or Lamb0 where indigenous population exceeded this number using the strains
CIAT 632 and 899 which had not been shown to have exceptional competitive ability.
Generally, bean plants had good nodulation and plant vigour when raised on fertile
volcanic soils (e.g. Salien, Table 1). In contrast, nodulation and plant vigour were poor
in the more common reddish brown soils. Chemical analysis of these soils (Table 1)
showed that concentration of available P (bicarbona~soluble) was small (l-2 mg kg
i) and that this might be a limiting factor for plant growth and nodulation. No other
single factor could be identified as the cause for pocx nodulation, Soil acidity has been
shown to depress nod&ion, however our data indicated that most soils from this region
were not very acidic @H 5.5, Table 1). organic matter content was average (3%) and
cation exchange capacity also reasonnable (30 meq/lOO g), and there was no indication
that nodulation was related to either of these (Table 1).
In addition to low concentration of soil P, the data clearly showed that soils from
research stations had unrealistically high levels of fertility and did not resemble those
Ii-am farmers’ fields (Table 1). Hence the results of Experiment I, done on sites belonging
to research stations, are not comparable wth the situation in the farmers’ field. It was
also evident that although large numbers of PhaseoZus rhizobia were often present in
the farmers’ fields, nod&&ion and plant vigour were poor and this might be due to the
small amount of available soil phosphate.
232
Experiment II (1988)
At both Erente and TPRI, inoculation with Rhizobium and P application as TSP
markedly (P < 0.01) increased nodulation (Tables 2 and 3), and yield (Tables 4 and 5).
Inoculation alone also increased nodulation and yield, but it was not statistically signi-
ficant Nitrogen application suppressed nodulation (Tables 2 and 3) and sometimes in-
creased yield (Tables 4 and 5). Phosphorus when applied in the form’ of TSP increased
nodulation and yield (Tables 2,3,4 and 5). Similar findings have been reported in Kenya
(Ssali and Keya, 1983). MRP also increased nodulation and yield but to a lesser extent
(Tables 2,3,4 and 5). The highest level of P applied (180 kg P,O, ha-i) gave the maximum
yield (Tables 4 and 5) thus potentially yield might be increased with further applica-
tion of P, although such amounts would not be economical and rarely used by the farmer.
There was no change in pH, but amount of available P had increased following increa-
sing rates of P application (Table 6), more so at TPRI than at Erente. On average, only
20% of the root system was colonized by vesicular arbuscular mycorrhizas, a fungal root
symbiosis which plays an important role in uptake of P from the soil. There was no
real difference in the amount of colonization between sites or treatments. The small extent
of colonization is surprising. For bean, roots are highly susceptible to mycorrhizal infec-
tion (Halley and Peterson, 1979). When the soil was examined, a few spores of Glomus
monosporum were present. It would be justifiable to test response to mycorrhizal inocu-
lation, especially in association with rock phosphate, if such small spore populations are
found to be common.
The preliminary results presented here are now part of a long-term investigation
to study effects of phosphorus, in particular the residual effect of MFW, an important
soil factor limiting bean production.
Conclusions
Large populations of indigenous Rhizobium phaseoli are present and yield responses
due to Rhizobium inoculation are not common in these soils, even though nodulation
may be increased.
Soil phosphate is a major factor limiting nod&ion and Phuseolus production.
By raising the soil available P, it may be possible to form effective nod&ion
from the existing indigenous population of Rhizobium, which appear to be equally effi-
cient as the inoculation strain used here.
In this fmt year of study, TSP was a more effective source of P fertilizer in compa-
rison to h4RP. Future studies will show whether the residual effect from MRP can give
similar yields obtained with TSP.
Acknowledgement
This collaborative project is funded by the U.K. Overseas Development Adminis-
tration (R4279A).
233
Table 2 :
Nodule number (m-2) at Erente (28 OAP) comparing inoculation
(+ Inoc), application of nitrogen (+N = 30 kg N ha -1) and
Minjingu rock phosphate (MRP) or triple super phosphate (TSP).
Level of P
MRP
TSP
Kg P2O5 ha-l
Inoc
-N
+N
-N
+N
-
-
0
721
189
721
189
60
558
319
1031
412
120
1119
281
1812
357
180
779
476
1544
1235
0
+
759
300
759
300
6 0
+
584
256
1970
906
120
t
830
111
1248
596
180
+
952
241
1259
819
Standard Error
82.8
234
Table 3 : Nodule number (m-21 at TPRI (37 DAP) comparing inoculation
(+ Inoc), application of nitrogen (+N = 30 kg N ha -11 and
Minjingu rock phosphate (MRP) or triple super phosphate (TSP).
MRP
TSP
Level of P-
Inoc
Kg P205 ha 1
- N
+ N
- N
+ N
-
-
0
689
300
689
300
60
1090
408
1631
633
120
1022
437
1858
812
180
1159
593
1873
740
0
+
604
226
604
226
60
+
828
259
1626
588
120
+
1013
367
1847
536
180
+
1234
533
2627
951
Standard Error
52.8
235
Table 4 : Seed yield (kg ha-11 at Erente (109 DAP) comparing inoculation
(2 Inoc), application of nitrogen (+N = 30 kg N ha-11 and
Minjingu rock phosphate (MRP) or triple super phosphate (TSP).
Level of P-
MRP
TSP
Kg P205 ha 1
Inoc
- N
+N
- N
+ N
0
421
528
421
528
60
552
596
530
552
120
546
500
668
698
180
398
540
797
866
0
t
367
563
367
563
6 0
t
500
476
622
546
120
+
599
494
586
632
180
t
449
608
790
962
Standard Error
24.6
-’
:‘.
:
‘.
236
Table 5 : Seed yield (kg ha-11 at TPRI (93 DAP) comparing inoculation
(2 Inoc), application of nitrogen (+N = 30 kg,N ha-l) and
Minjingu' rock phosphate (MRP) or triple super phosphate (TSP).
Level of P
MRP
TSP
kg P2O5 ha-l
Inoc
-N
tN
- N
tN
0
158
274
158
274
60
249
221
314
294
120
221
279
323
389
180
276
269
391
438
0
+
181
246
-181
246
60
+
237
264
286
342
120
+
215
279
319
332
180
t
262
408
338
388
Standard Error
10.1
237
Table 6 : Change with application (kg P205 ha-l) of Minjingu rock phosphate
(MRP) or triple super phosphate (TSP) in soil pH,and bicarbonate
soluble P (mg P kg-l) measured at Erente (109 DAP) and TPRI
(93 DAP).
Source and level of applied P
MRP
TSP
0 60 120 180
0 60
120
180
-
-
-
-
-
-
-
-
Erente
pH
5.3 5.3 5.2 5.4
5.3 5.3
5.3 5.3
P
2.2 2.8 4.7 6.2
2.2 4.2
4.7 6.5
TPRI
pH
6.6
6.8
6.6
6.7
6.6
6.6
6.7 6.7
P
1.4
3.9
10.2
11.8
1.4
5.8
14.3
19.0
238
References
ALLBN, DJ. 1986 Bean production systems in Africa CIAT, Cali, Colombia.
GRAHAM, P.H 1981 Some problems of nod&&ion and symbiotic nitrogen fixation in
Phaseolus vulgaris L. A review. Field Crops Res. 4, 93-112.
HOLLBY, J.D. and PETERSON, R.L. 1979 Development of a vesicular arbuscular mycor-
rhiza in bean roots. Can. J. Bot. 57, 1960-1978.
MAFF/ADAS 1986 The analysis of Agricultural materials. RB427. HMSO.
PHILLIPS, JM and HAYMAN, D.S. 1970 Improved procedures for clearing roots and
staining parasitic and vesicular-arbuscular mycorhizal fungi for rapid assessment of
infection. Trans. British Mycol. Sot. 55, 158-161.
SWIPE, A 1968 The effect of a cassava <<fallow* and various mammal treatments on
cotton at Ukiriguru, Tanzania. East A$ican Agric. and Forest. J. 33, 231-235.
SINGLETON, P.W. and TAVARES, J.W. 1986 Inoculation response of legumes in rela-
tion to the number and effectiveness of indigenous Rhizobium populations. Appl. En-
viron. Microbial. 51, 1013-1018.
SSALI, H and KEYA, S.O. 1983 The effects of phosphorus on nodulation, growth and
dinitrogen fixation of beans. Biol. Agric. Hort. 1, 135-144.
VINCENT, J.M. 1970 A manual for the practical study of root-nodule bacteria. IBP Hand-
book W 15, Blackwell Scientific Publications, London.
h4aximiser la FBA pour la production agricole et forest&e en Afiique
Variation in acid-Al tolerance -
of Bradyrhizobium japonicum
ASANUMA, S.(l) and A-ABA kc2)
(1) : Hokkdo National Agricultural Eqwiment Station,
1, Hitsujigaoka, Toyohira-ku, Sapporo. o@$ Japan.
(2) : De1 Monte Corporation, Walnut Creek Research Center.
P.O. Box 9004. Wabu~ Creek, California 94598. USA.
Abstract
Seventy-six stmins of Braa’yrhizobium japonicwn (54 African and 22 exotic)
were examined for tolerance of acidity @H 4.5). low P (5 @VI) and high Al (50
and 100 @f) by an agar plate method. Fortyfour strains were tolerant to acidity
regardless of P levels (1.000 or 5 @vi) of the medium and 22 were sensitive. The
remaining 10 strains differed in tolerance of acidity depending on level P of the
medium;nineweretolelantoflowP(5@vf)andoneofhighP(1.OOOcLM).All
of 21 strains which grew on the most severe stress used consisting of low P and
high Al (100 P.M) at pH 4.5 showed tolerance of acidity at both levels of P. Isolates
from a highly acidic soil @H 4.6, HaO) from Gnne, southeastern Nigeria, showed
different levels of acid-Al tolerance but tended to be more tolerant than those from
a slightly acidic soil (pH 6.5, HaO) from lbadan, southwestern Nigeria However.
20 % of the Omre isolates and 30 % of the Ibadan isolates were sensitive to acidity.
Thus, tolerance of stmins was not necessarily determined by the acidity of the soils
in which they reside. Gum production of certain strains occured at low P. but this
had no relation to acid-Al tolerance.
The great variation in acid-Al tolerance observed among strains of B. japonicwn
supports the strategy to select strains for use as inoculants in acid soils.
240
Introduction
Performance of inoculated strains of Bradyrhizobium japonicum in a highly acidic
soil (Ultisol, pH 4.6, %O) from Gnne, southeastern Nigeria, was good in terms of nodu-
lation of soybeans (Glycine mar (L.) Merr.) in the first year but was poor in the second
promfield and Ayanaba, 1980 ; IITA, 1981). It was suggested that some acid soil factors
are detrimental to the ecological behavior of introduced rhizobial strains and that use
of acid tolerant strains could lead to better utilization of the benefits of biological nitro-
gen fixation to grow soybeans under such acid conditions.
A great variation in tolerance of several acid soil factors has been shown among
strains of Rhizobium leguminosarum biovar tnifolii (Thornton and Davey, 1983 ; Thurman
et al., 1985). R. leguminosurum biovar phaseoli (Munns et al., 1979 ; Graham et al.,
1982 ; Lowendorf and Alexander, 1983), R. meliloti (Howieson et al., 1988) ; R. loti
and Bradyrhizobium sp. (lotus) (Cooper, 1982), Bradyrhizobium sp. (cowpea group) (Date
and Halliday, 1979 ; Keyser and Munns, 1979a and b ; Keyser et al., 1979 ; De Carvalho
v., 1981 ; Munns and Keyser, 1981 ; Ayanaba et al., 1983) and B. japonicum (Keyser
and Munns, 1979a and b ; Ayanaba et al., 1983). Little information, however, is available
on native African strains of B. juponiczun. Ayanaba ef al., (1983) reported that some
African rhizobia isolates nodulating both cow-pea (Vigna unguiculata (L.) Walp.) and soy-
bean cultivar TGm 344 showed wide variation in tolerance.
Tolerance of strains of rhizobii to acid soil factors is generally tested by growth
in defined liquid media (Date and Hall&y, 1979 ; Keyser and Munns, 1979a and b ;
Howieson, 1985). Keyser and MUMS (1979a) reported that in the test of tolerance of
acid soil factors with Bradyrhizobium sp. (cowpea group) and 8. japonicum, Al
(25-50 p.M) was probably a more severe stress than low Ca (50 @VI) and/or high Mn
(200 @VI). Thornton and Davey (1983) showed that in liquid culture media, growth of
R. trjfoliti was more limited by high Al (15 to 40 p.M) than by low P (1 to 6 pM).
Ayanaba et al., (1983) proved that screening of Rhizobium for tolerance to acid-Al stress
could be achieved by an agar plate method and that this method had an advantage in
handling large numbers of rhizobia.
In the present paper we report variation in tolerance of acid-Al stress among strains
of B. jczponicum isolated mostly from African soils. Acidity, low P and high Al were
factors examined by the agar plate method.
Materials and methods
Rhkobial strains
Seventy-six strains of B. juponicum from the culture collection of the Intematio-
nal Institute of Tropical Agriculture @TA), Ibadan Nigeria were used. Fifty-four of them
were isolated from African soils, 20 of which were from either soybean promiscuous
(Malayan, Orba or 4H/149/1) or non-promiscuous cultivars (TGm 294 or Bossier) grown
in a highly acidic soil of Gnne. Twenty were from the same cultivars grown in a slightly
acidic soil @H 6.5, %O) of Ibadan, southwestern Nigeria. Some of the African isolates
nodulate cowpeas (IITA, 1979). Three wild-type strains and their antibiotic resistant mu-
tants were included. One of them is resistant to spectinomycin and the other two to strepto-
mycin. All strains were cultured on slopes of yeast extract-mannitol agar medium (YJZM
agar) (Vincent 1970).
241
Test agar media
Six media prepared according to Keyser and Munns (1976b) were used : full -
defined ; acid ; low P ; low P, acid and two levels of high Al (50 and 100 pM). Each
medium was uspplemented with 15 g/l ionagar (Oxoid Co. LtrL, England). Each medium
contained the basal solution of Keyser and Munns (1979b). Five hundred @vI each of
KH$O, and K.$PO, and 1.1 g/l Na-glutamate were added to the two full-defined media,
whereas 5 pM KH$O,, 1.5 mM KC1 and 1.1 g/l Na-glutamate were added to the other
four media. Fulldefmed and low P media were adjusted to pH 6.8 with lN-NaOH and
the others were adjusted to pH 4.5 with IN-HCl before addition of AlK(SO,), and growth
factors (Geyser and Murms, 1979b). Furthermore, methyl red in ethanol was added as
a pH indicator to a final concentration of 0.002%. The media were autoclaved (121”C,
15 mm) before dispensing into plates (30 ml/plate). Autoclaving and the addition of these
substances slightly changed the pH values of the media but it did not affect the testing.
Testing of acid-Al tolerance
Loopfub of rhizobia grown on YEM agar slopes were suspended in sterilized
0.85% saline to give similar turbidity among strains. The six different agar media, pre-
dried for 3 h at 45°C were inoculated by transferring the bacterial suspensions with steri-
lized cotton swabs. Twenty strains were tested in one plate and the experiment was du-
plicated. Obervations were made daily during a lo-day-incubation at 28°C. Tolerance of
rhizobia to acidity or acid-Al stress was evaluated by comparing their growth on the
stress media with that on the full-defined medium. The tolerance of Onne and lbadan
isolates was as designated in Table 1, bases on the levels of severity of the stress.
Table 1
: Designation (I to VII) of the degree of response to acid-Al
stress by Bradyrhizobium isolates from Onne and Ibadan soils ;
media were at pH 4.5.a
Designation
P(uM)
Al
(uM) I II
III IV v VI
VII
-
-
-
-
-
-
-
1,000
0
S
'S
S
T
T T
T
5
0
S T
T
S
T T
T
5
50
s s
T
S
S T
T
5
100
s s
S
. s
s s
T
a:Growth on each medium was compared with that on control medium
(pH 6.8, 1,000 uM, P and OuM Al) and designated as T (tolerant) or S
(sensitive) accordingly.
242
Results and discussion
Most rhizobia grown on laboratory media are known to produce either acidic or
alkaline substances (Norris, 1965 ; Jones and Burrows, 1%9 ; Cooper, 1982, Hernandez
and Focht, 1984), depending on the preferential use of sugars or organic nitrogen com-
pounds, respectively, as the source of energy (Parker et al., 1977). In evaluating the growth
of Rhizobium on such media, therefore, changes in pH have to be monitored Methyl
red was employed in the present experiment for that purpose. About half of the strains
tested produced alkaEne substances when grown on the agar media (Table 2), but. it was
Table 2 : Bradyrhizobium japonicum strains tolerant to acid-Al (lOO@J, pH 4.5)
stress
IITA
Alkali
Colony type
accession
Origin or source
productiona
no.
High P
Low P
(1,OOO)lM)
(5)lM)
' IRj 2001
cv. Malayan. Onne
"db
nd
2002
cv. Halayan, Onne
+
white
watery
2004
cv. Halayan, Onne
+
white
.watery
2007
cv. Orba, Onne
+
white
watery
2013
cv. Halayan, Ibadan
+
white
watery
2019
cv. Orba, Ibadan
+
white
watery
2020
cv. Orba, Ibadan
--
nd
nd
2025
cv. TGmZ94, Ibadan
+
white
watery
2033
cv. Orba, Yandev
white
watery
2035
cv. Orba, Yandev
nd
nd
2038
cv. Malayan. Samaru
nd
nd
2039
cv. Halayan, Samaru
+
nd
nd
2048
cv. TGm294, Samaru
white
watery
2049
cv. TGm294, Samaru
nd
nd
2052
cv. Bossier, Samaru
nd
nd
nd
2109
Univ. of Wisconsin,
nd
nd
USA (SW 61A76)
2119
ENSA. Ivory Coast (63)
+
nd
nd
2120
ENSA, Ivory Coast (GB)
nd
nd
nd
2123
USDA (3llb138)
+
nd
nd
2128
Thailand (TWA21
+
white
not watery
2129
Thailand (THA5)
+
white
not watery
a
Alkali production : +, producing ; -, not producing ; nd, not determined.
b
nd, not determined.
243
observed that their growth preceded the shift in pH of the media. Ayanaba et al. (1983)
observed that strains of rhizobia tolerant of acid-Al stress on agar medium grew signifi-
cantly in liquid stress medium before a change in pH Thus, it seemed that tolerance
of rhizobia to acid-Al stress might not be directly related to their ability to produce alkali.
Rhizobial strains showed great variation in tolerance of acid-Al stress (Fig.1). Most
strains tolerated acidity stress (pH 4.5) similarly regardless of P levels (1.000 or 5 pM)
pH POJM) Al(pMI
4.5 1,000
Figure I : Effects of acidity and various levels of P and Al on growth of B. japonicum inoculated on
ag~mediaGrowthone~hmediumw~examinedby~mparisonwiththaton~ntrolmedium(pH
6.8.
P 1,000 @VI and Al 0 p.M) and designated T (tolerant) or S (sentitive) accordingly. Results are shown
with num& of strains and their proportions in percent in parentheses.
of the medium ; 44 strains were tolerant and 22 were sensitive. The remaining 10 strains,
however, differed in tolerance of acidity depending on the level of P. Nine required
low P 5 @vi) to be torerant and one required high P (1,000 pM) to be tolerant. Thus,
tolerance of rhizobia to acidity seemed to differ depending on the level of phosphorus
content of the medium. Table 3 shows the effect of P levels on the growth of rhizobia
under two pH conditions (6.8 and 4.5). Over 90 % of the strains grew similarly at pH
6.8 regardless of P levels. By contrast at pH 4.5, 25 % or 19 strains grew less at the
high P than at the low P. Nine strains which required low P to grow at pH 4.5 (Fig.1)
were also in this group. Uptake and utilization efficiencies of phosphorus were reported
to be greatly different among strains of rhizobia, which caused differences in low P tole-
rance of rhizobia (Cassman et al., 1981b ; Beck and Munns, 1984). The result shown
in Table 3 suggests that the efficiencies in phosphorus uptake and utilization might
change depending on the pH surrounding strains of rbizobia.
Growth of rhizobia at pH 6.8 and 4.5 was also compared at different P condi-
tions (1.000 and 5 pM) (Table 4). Most strains responded similarly to pH regardless
of P levels, suggesting that phosphorus is less influential on the growth of rbizobia than
pH. This supports observations with R. kifolii by Thornton and Davey (1983).
As shown in Fig. 1, tolerance of acidity did not always mean tolerance of acid-
Al, indicating that tolerance of acidity stress among rhizobia might be enhanced by the
presence of ahuninum. Thirty three of the 53 strains that tolerated acidity at the low
P level could grow under acid-Al (50 pM, pH 4.5) stress ; 21 of these or 27.6 % were
further tolerant of the most severe Al stress tested (100 J.LM). Similar results were obtained
,-. : :,i . .
..,..
244
Table 3 : Effet of P levels on growth of B. japonicum at pH
-
6.8 and 4.5a
PH
Growth in 1,000 )IMP compared to that in 5 PMP
Larger
Similar
Smaller
6.8
2 (2.6)
70 (92.1)
4 ( 5.3)
4.5
2 (2.6)
.
55 (72.4)
1 9 (25.0)
a Growth of strains was compared after lo-day-incubation. Visual
readings on a scale'& &5 were made of$colonies and the differences
of the minimum 2-point-readings were considered significant.
Results are shown with numbers of strains and their proportions
in percent in parentheses.
by Keyser and Munns (1979b) Who found that three of the eight B. jqwnicum strains
that were tolerant of pH 4.5 also tolerated acid-Al (50 p.M) stress. The most tolerant
strains (Table 2) were not restricted to one geographic region of the world
Does acidity of bulk soil determine the acid-Al tolerance of rhizobia isolated l3om
the soil ? To investigate it, acid-Al toll of isolates from a highly acidic Onne or
Table 4 : Effect of pH on growth of B. japonicum at two different
levels of pa
P (uM1
Growth at pH 6.8 compared to that at pH 4.5
Larger
Similar
Smaller
1,000
42 (55.3)
31 (40.8)
3 (3.9)
5
3 2 (42.1)
4 4 (57.9)
0 (0.0)
a See the footnote in Table 3.
245
slightly acidic Ibadan soil (pH 4.6 and 6.5 respectively) were compared (Fig. 2). The
Cnne isolates showed less varied levels of tolerance than those from Ibadan soil and
tended to be more tolerant of acid-AI than the Ibadan isolates, indicating that the acidity
of a bulk soil could affect the acid-Al tolerance of rhizobia residing in that soil. However,
20 % of the Onne isolates were sensitive to acidity (pH 4.5), and so were 30 % of the
Ihhn isolates. Thus, the tolerance of strains seemed not to be necessarily &tern&d
by the acidity of bulk soil from which they were isolated. Jones and Burrows (1969)
also reported no correlation between the final pH of the culture media in which isolates
of R. lrifolii grew and the pH of the soils from which they originated In soil, acidity
or acid-Al conditions of a microenvironment could affect the acid-Al tolerance of the
inhabitants, therefore, further research needs to be carried ..out in this connection.
Rhizobia nodulating non-promiscuous soybean cultivars are known to nodulate pro-
miscuous ones, too, but the reverse is not always true (IITA, 1979 ; Bromfield and Roughley,
1980). It means there is a difference among strains of rhizobia in terms of compatibility
with host cultivars. Promiscuity of soybean cultivars (Nangju, 1980 ; Pulver et al., 1985)
from which the Onne and Ibadan isolates were obtained and the acid-Al tolerance of
these isolates are shown in Fig. 3. Acid-Al tolerance of the isolates was not dependent
on the promiscuity of host plants from which they were isolated.
Certain strains formed watery and translucent colonies; regardless of P levels (1,000
or 5 pM) in the full-defined medium. However, some strains formed white colonies on
the same medium when P was high, and watery and translucent colonies when P was
low (Table 2). Watery colonies were formed presumably because production of copious
amounts of gummy substances occurred when the P level of the medium was low. Cassman
et al., (1981a) also observed abundant production of external gum by strains of B. juponicum
in defined liquid medium with low and moderate P, but not with the high P concentrations
routinely supplied in laboratory media. These results suggest that gum production of certain
rhizobia may be enhanced when the P level of the medium becomes low. In the study
of cowpea rhizobia isolated from three locations in West Africa, two types of colonies,
dry and pinpoint or large and gummy, were frequentIy observed when grown on YEM
agar (IITA, 1981 ; Ahmad et al., 1981) and Ayanaba et al., (1983) found that rhizobia
strains forming the former type of colonies were more sensitive to acid-Al stress than
those forming the latter ones. Cunningham and Munns (1984) showed that high production
of extracellular polysaccharide is positively related to the acid tolerance of rhizobia. In
this study, the relation between gum production and acid-Al tolerance of rhizobia was
not clear. In this connection, How&on et al., (1988) also reported that with R. trifolii,
there was no distinct~ link between polysaccaride production and acid tolerance.
Three wild type strains and their antibiotic r&st.ant mutants were included in this
study. Mutation induced by two antibiotics seemed not to affect the acid-Al tolerance
of rhizobia : all strains were as sensitive or tolerant as the wild types (data not shown).
This result supports the observation by Ayanaba and Wong (1982) that antibiotic resistant
mutants tolerated acidity to the same extent as their parents.
A consistent and stable property of tolerance of rhizobia to acid and aluminum
stresses (Munns and Keyser, 1981; Munns, 1986) and a great variation in acid-Al tolerance
among strains of B. juponicum observed in this study prove that strains can be selected
for possible use as soybean inoculants in acid soils. It is obvious, however, that there
are some other adverse, abiotic stresses like high temperature and/or desication (Osa-
Afiana and Alexander, 1982) in acid soils to which rhizobial inocula are probably expo-
246
0 Onne
Ibadan
I
I#
.I11 IV
V
V I
V I I
T o l e r a n c e o f a c i d - A l s t r e s s
Figure 2 : Variation in the tolerance of acid-Al stress shown by Bradyrhizobium isolates from
Onne and Ibadan soils. Twenty isolates each were tested. I - VII : See Table 1.
P r o p o r t i o n i n p e r c e n t
A,
0
5 0
1 0 0
I
I
1
I
I
I
I
I
I
I
Onne : Promiscuous
I
V
V I I
c u l t i v a r s
N o n - romiscuous
I v
V I
c u l t r v a r
Ibadan:
p r o m i s c u o u s
c u l t i v a r s
I
I
I’ I
I
I
I
I
I
I
I
Figure 3 : Comparison of acid-Al
stress tolerance between Bradyrhizobium isolates from promis-
cuous and non-promiscuous cultivars of soybean grown in Onne and lbadan soils. Eleven and 9
isolates were tested for the promiscuous and non-promiscuous cultivars, respectively. in
Onne
soil and each 12 and 8 isolates, respectively. in
Ihadan soil.
241
sed, particularly during fallow periods under tropical African conditions. Therefore, tests
are required on the effects of those stresses on the saprophytic competence of rhizobia
(Thornton and Davey, 1984) and their symbiotic effectiveness under field conditions. At
the same time, screening or breeding of soybean cuhivars adapted to acid soil conditions
should not be ignored (Mmms et al., 1981).
Acknowledgments
This study was supported by the United Nations Environment Programme.
References
AHMAD, M.N., EAGLESHAM, AR.S. and HASSOUNA, S. 1981 Examining serolo-
gical diversity of cccowpea>, rhizobia by the ELISA technique. Arch. Microbial. 130,
281-287.
AYANABA, A., ASANUMA, S. and MUNNS, D.N. 1983 An agar plate method for
rapide screening of Rhizobium for tolerance to acid-aluminum stress. Soil Sci. Sot.
Am. J. 47, 256-258.
AYANABA, A. and WONG, W.L. 1982 Antibiotic -. resistauce mutants identified from
nodules of uninoculated soybeans grown in a strongly acidic soil. Soil Biol. Biochem.
14, 139-143.
BECK, D.P. and MUNNS, D.N. 1984 Phosphate nutrition of Rhizobium spp. Appl. Environ.
Microbial. 47, 218 282.
BROMFIELD, E.S.P. and AYANABA, A. 1980 The efficacy of soybean inoculation on
acid soil in tropical Afkica Plant and soil. 54, 95-106.
BROMFIELD, E.S.P. and ROUGHLEY, RJ. 1980 Characterization of rhizobia isolated
from nodules on locally-adaptet Glycine max grown in Nigeria Ann. AppI. Biol. 95,
185190.
CASSMAN, KG., MUNNS, D.N. and BECK, D.P. 198laPhosphorus nutrition of Rhizobium
japonicum : strains differences in phosphate storage and utilization. Soil Sci. Sot.
Am. J. 45, 517-520.
CASSMAN, KG., MUNNS, DN. and BECK, DP. 1981b Grown of Rhiwbium s@ains
at low concentrations of phosphate. Soil Sci. Sot. AmJ 45, 520-223.
COOPER, JE. 1982 Acid production, acid tolerance and growth rate of Lotus rhizobia
in laboratory media Soil Biof. Biochem. 14, 127-131.
CUNNINGHAM, SD. and MUNNS, D.N. 1984 The correlation between extraceIIular
polysaccharide production and acid tolerance in Rhizobium: Soil Sci. Sot. Am. J. 48.
1273-1276.
DATE, R.A. and HALLIDAY, J. 1979 Selecting Rhizobium for acid, infertile soils of
the tropics. Nature. 277, 62-64.
DE CARVALHO, M-M., BUSHBY, H.V.A. and EDWARDS, D.G. 1981 Survival of Rhi-
zobium in nutrient solutions containing aluminum. Soil Biol. Biochem. 13, 541-542.
GRAHAM, P-H., VITERI, SE., MACKIE, F., VARGAS, A.T. and PALACIOS, A. 1982
Variation in acid soil tolerance among strains of Rhizobium phaseoli. Field Crops
Rex 5, 121-128.
248
HERNANDEZ, B.S. and FOCHT, D.D. 1984 Invalidity of the concept of slow growth
and alkali production in cowpea rhizobia. Appl. Environ. Microbial. 48, 206-210.
HOWIESON, J.G. 1985 Use of an organic buffer for the selection of acid to1erantRhiz~bk.m
melibti strains. Plant and Soil. 88, 367-376.
HOWIESON, J.G.,EWING,M.A. andD’ANTUON0,M.F. 1988 Selection for acid tolerance
in Rhizobium meliloti. Plant and Soil. 105. 179-188.
IINTERNATIONAL IINSTITUTE of TROPICAL AGRICULTURE. 1979 The BTA Col-
lection of Cuhures of Rhiwbium. Ibadan, Nigeria.
IINTERNATIONAL INSTITUTE of TROPICAL AGRICULTURE. 1981 Annual Report
for 1980. Ibadan, Nigeria pp. 155160.
JONES, D.G. and BURROWS, A.C. 1969 Acid production and symbiotic effectiveness
in Rhizobium trifolii. Soil Biol. Biochem. 1, 5761.
KEYSER, H.H. and MUNNS, D.N. 1979a. Effects of calcium, manganese, and aluminum
on growth of rhizobia in acid media. Soil Sci. Sot. Am. J. 43, 500-503.
KEYSER, H.H. and MUNNS, D.N. 1979b. Tolerance of rhizobia to acidity, aluminum,
and phosphate. Soil Sci. Sot. Am. J. 43, 519-523.
KETSER, H-H., MUNNS, D.N. and HOHENBERG, J.S. 1979 Acid tolerance of rhizobia
in culture and in symbiosis with cowpea. Soil Sci. Sot. AM. J. 43, 719-722.
LOWENDORF, H.S. and ALEXANDER, M. 1983 Identification of Rhizobium phuseoli
strains that are tolerant or sensitive to soil acidity. Appl. Environ. Microbial. 45, 737-
742.
MUNNS, D.N. 1986 Acid soil tolerance in legumes and rhizobia Adv. Plant Nun. 2,
63-91.
MUNNS, DN., HOHENBERG, J.S.,RIGHETTI, TL. and LAUTER, DJ. 1981 Soil acidity
tolerance of symbiotic and nitrogen-fertiked soybean. Agr. J. 73, 407410.
MUNNS, D.N. and KEYSER, H.H., 1981 Response of Rhizubium strains to acid and
aluminum stress. Soil Biol. Biochem. 13, 115-118.
MUNNS, DN., KEYSER, H.H., FOGLE, V-W., HOHENBERG, J.S., RIGHETII, T.L.,
LAUTER, D.L.. ZAROUG, M.G., CLARKIN, K.L. and WHITACRE, K.W. 1979
Tolerance of soil acidity in symbioses of mung bean with rhizobia Agr. J. 71, 256-
260.
NANGJU, D. 1980 Soybean response to indigenous rhizobia as influenced by cultivar
origin. Agr. J. 72, 403-406.
NORRIS, D.O., 1965 Acid production by Rhizobium a unifying concept. Plant and Soil.
22, 143-166.
OSA-AFIANA, L.O. and ALEXANDER, M. 1982 Differences among cowpea rhizobia
in tolerance to high temperature and desiccation in soil. Appl. Environ. Microbial.
43, 435-439.
PARKER, C.A., TRINICK, M.J. and CHATEL, D.L. 1977 Rhizobia as soil and rhizosphere
inhabitants. In : A Treatise on Dinitrogen Fixation Section IV. Agronomy and Ecology.
Eds. Hardy, R.W.F. and GIBSON, A.H. : A willey-Interscience Publication, USA,
PP 311-352.
249
PULVER, EL., KUENEMAN, E.A. and RANGA RAO, V. 1985 Identification of pro-
miscuous nodulating efficient in N, fixation. Crop Sci. 25, 660-663.
THORNTON, F.C. and DAVEY, C.B. 1983 Acid tolerance of Rhizobium trzfolii in culture
media. Soil Sci. Sot. Am. J. 47, 496-501.
THORNTON, F.C. and DAVEY, C.B. 1984 Saprophytic competence of acid tolerant strains
of Rhizobium tnyolii iu acid soil. Plant and Soil. 80, 337-344
THURMAN, NP., LEWIS, D.M. and JONES, D.G. 1985. The relationship of plasmid
number to growth, acid tolerance and symbiotic efficiency in isolates of Rhkobiurh
trifolii. J. Appl. Bacterial. 58, l-6.
VINCENT, J-M. 1970 A Manual for the Practical Study of the Root nodule Bacteria
IBP Handbook Np 15. Blackwell scientific Publications. oxford and Edingburgh. p.3.
Maximiser la FBA pour la production agricole et fond&e en Afiique
The influence of some parameters
of soil fertility on early growth
of Leucaena ZeucocephaZa and Cassia siama
LIYA, S.M$ MULONGOY, K.c2),
Odu, C.T.P, and AGBOOLA, AA.(l).
(I) : Department qf Agronomy, University of Ibadan, Ibaah, Nigeria.
(2) : International Institute of Tropical Agriculture. PMB 5320. Ibaahn, Nigeria.
Abstract
L~ume~ leucocephala and Cassia siamea were grown in concrete cylinders
containing 0.35 t of a sandy, high base status Entisol. The nutrient content of the
soils varied from one cylinder to another. Fertilizer was added to the rate
11 kg N, 88 kg P and 5 kg Zn per hectare. At 8 months after planting, Lxucm
was well no-&late-d and produced on average seven times as much biomass as the
na-nodulated Casriu. There were significant (P = 0.05) correlations with topsoil
N. With Ca&a, positive correlations were observed for plant height (r = 0.73) and
total stem length (r = 0.75). With Lfxc- there were significant negative correla-
tions for plant height (r = - 0.85) and estimated total biomass (r = - 0.90). L..euasta
plant height was also positively correlated with topsoil K (r = 0.81). Ninety
eight per cent of the variation in height could be accounted for by the equation
Y = 4.88 + 117.37X - 14.94X2 where Y = plant height and X = K/N ratio.
251
Introduction
Trials of alley cropping in farmers’ fields have run into problems on account of
the very uneven establishment of the trees due to poorly understood factors. The work
reported here provides indications of how certain parameters of soil fertility may influ-
ence early growth of Cassiu siamea and Leucaena leucocephala.
Materials and methods
Two month-old plants of LRucae~ leucocephala (K8) and Cassia siamea (bcal
variety) were transplanted into concrete cylinders (60 cm i.d. and 90 cm deep) sunk in
the ground and containing a sandy, high base status Entisol which had been previously
used for various experiments for over a decade at Intemational Institute of Tropical Agri-
culture (BTA), Ibadan. Before planting, composite topsoil (15 cm) samples were taken
by bulking five cores per cylinder, and their nutrient content determined at the university
of Ibadan. At planting time, the cylinders were fertilized with 88 kg P ha-’ as single
superphosphate and 11 kg N ha-’ as (NH&SO,. Two months later 5 kg ha-r Zn was
applied as ZnSO,.Plant height, total stem lenght (plant height plus the sum of branch
lenghts) and leaf number were recorded 6 months after transplanting and a rough estimate
of total aboveground biomass was made, using the formula of Anderson and Ingram (1987)
to estimate stem and branch biomass. Leaf biomass was estimated by multiplying leaf
number by the average weight per leaf of a sample of leaves.
Results
Soil analysis (Table 1) indicated marked differences in nutrient content between
the cylinders with coefficients of variation ranging fi-om 11 to 64%. Soil physical pro-
perties (not reported here) varied little. The soils could be described as having medium
to high organic matter, low available P. and medium to high exchangeable K. The plant
growth data show that Lmcaena produced on average about seven times more biomass
than Cassia. There was considerable variation from one plant to another. For Leucaena,
coefficients of variation for leaf number, total stem length, plant height and estimated
total biomass were 51,58, 17 and 48% respectively. For Cursiu the variation was lower :
17, 40, 18 and 36% respectively. The form of different trees of the same species also
varied appreciably. Leucaena had 3 to 24 branches, Cassia 1 to 7 branches. Leucaena
was well nodulated in all the cylinders, but Cu.&u did not nodulate.
Regression analyses indicated that Cassia and Leucaena were differently affected
by topsoil properties.
Cassia stem and leaf production were positively correlated with soil organic C,
total N, available P, exchangeable Ca and Mg but negatively correlated with exchan-
geable K. Except for total N, these correlations were not significant. Total N was sign&
tautly correlated at the 5% level with total stem length (r = 0.74) and there was a high
though not significant positive correlation with leaf number (r = 0.66). The correlation
coefficient for leaf number and available P was not significant (r = 0.41). Stem and leaf
production were positively correlated with exchangeable Ca (r = 0.58 and 0.19 respectively),
and stem production was positively correlated with exchageable Mg (r = 0.34). On the
other hand, plant height was negatively correlated with all these soil properties except
Table 1 : The range and variation of soil properties in cylinders planted with Leucaena leucocephala and Cassia
siamea and the range and variation of plant growth parameters of the two species.
Characteristic
Leucaena leucocephala
Cassia siamea
Mean
Range
cv (%I
Mean
Range
cv (%I
Soil properties
Organic C (4;
1.8
1.8- 210
1.5-2.1
11
Total N 1%)
0.13
0.07-0.24
2
O.lO-0.14.
Extractable P (ug g-1)
1.8-8.4
ki
K (meq/lOO g)
4.5
1.3- 8.4
ii
0.2-0.4
55
Ca(meq/lOO g)
F:i
0.2-
5.5-13.0 4.6
35
4.3-12.0
44
Mg(meq/lOO g)
K/N(meq/lOO g)
;:i
2.6-
1.5- 4.9 5.8
El
2.4-5.3
1.3-3.5
ii
Plant growth parameters
Leaf No. plant-'
113
23-190
51
39-62
17
Height (cm)
174
123-201
17
43-69
r18
Total stem length (cm)
517
123-935
58
137
69-251
40
Estimated total biomass (g)
737
79-1056
48
96
74-157
36
Table 2 : Correlation coefficients between soil properties and plant growth parameters
Tree
Plant
Soil property
parameters
C
N
P
K
Ca
MCI
K/N
~
-
-
-
-
-
Leucaena
Leaf No.
0.05
-0.76*
-0.22
0.05 -0.04
-0.53 0.30
leucocephala
Total stem
length
0.41
-0.69
-0.02
0.19
0.22
-0.41
0.60
Height
-0.58
-0.85"
-0.31
0.81*
0.01
-0.55
0.70
Estimated biomass
-0.13
-0.90*
-0.34
0.36
0.12
-0.49
0.53
Cassia siamea
Leaf No
0.02
0.66
0.41
-0.41
0.19
0.001
-0.57
Total stem
length
0.30
0.73*
0.32
-0.2
0.58
0.34
-0.18
Height
-0.45
0.75*
-0.56
-0.01
0.25
-0.43
-0.03
Cstimated biomass
0.09
0.47
0.03
-0.40
0.20
-0.09 -0.25
* Significant at P = 0.05
254
total N. There was no correlation between plant height and total biomass production
( r = 0.03).
With Lf~~aena, the results were in almost total contrast. There were negative cor-
relations between plant height, total stem length or leaf number and total N with the
following coeficients of correlation : -0.85 (significant at P = O-05), -0.69 and -0.76 respec-
tively. Also, correlations between exchangeable Mg and plant height, total stem length
or leaf number (r = -0.55, -0.41 and -0.53 respectively) were negative but not significantly.
There was however a significant correlation between plant height and exchangeable K
(r = 0.81). Plant height was also significantly correlated with total biomass production
(r = 0.87) and leaf number (r = 0.63). Plant height seemed to be governed by the soil
K/N ratio. The relationship was best described by a quadratic equation, and 98.9% of
the variation in Leucaena plant height could be dkcribed by the equation
Y = 4.88 + 117.37X - 14.94p, where Y = plant height in cm and X = K/N ratio in
the soil (Fig.l), with K expressed in meq/lOO g soil and N in percent.
Figure 1: Relationship betweenkucasla height and K/N ratio of the soil
Discussion
In spite of the deep rooting habit of Cassiu and Leucaena, it was interesting to
note some significant correlations with topsoil properties. The results highlight the very
different response to N of the two species. Growth of the non-nodulated Cassia was signifi-
cantly correlated with topsoil N and clearly limited by this element, so that plant biomass
production was very low compared with that of Leucaena. The results further suggest
that Cussia has a high requirement for P and exchangeable bases. This would explain
the negative correlation with K because of the antagonism between K and Mg. Our results
contour with observations by foresters that Cassiu siumea grows well only in fertile soils,
a fact which limits its use in re-aforestation programs (Anonymous, 1981).
254
total N. There was no correlation between plant height and total biomass production
( r = 0.03).
With L..eucuenu, the results were in almost total contrast. There were negative cor-
relations between plant height, total stem length or leaf number and total N with the
following coeficients of correlation : -0.85 (significant at P = O-OS>, -0.69 and -0.76 respec-
tively. Also, correlations between exchangeable Mg and plant height, total stem length
or leaf number (r = -0.55, -0.41 and -0.53 respectively) were negative but not signifkantly.
There was however a sign&ant correlation between plant height and exchangeable K
(r = 0.81). Plant height was also signikantly correlated with total biomass production
(r = 0.87) and leaf number (r = 0.63). Plant height seemed to be governed by the soil
K/N ratio. The relationship was best described by a quadratic equation, and 98.9% of
the variation in Leltcaena plant height could be decribed by the equation
Y = 4.88 + 117.37X - 14.94X, where Y = plant height in cm and X = K/N ratio in
the soil (Fig.l), with K expressed in meq/lOO g soil and N in percent.
Leucaenahdght (cm)
2 0 0
1 5 0
100
Y - 4.80+117.37X-l 4.942
0
1.0
2 . 0
3 . 0
4 . 0
5 . 0
6 . 0
W N
Discussion
In spite of the deep rooting habit of Cussiu and Leucaena, it was interesting to
note some signifkant correlations with topsoil properties. The results highlight the very
different response to N of the two species. Growth of the non-nodulated Cussiu was signifi-
cantly correlated with topsoil N and clearly limited by this elemenf so that plant biomass
production was very low compared with that of Leucue~~~. The results further suggest
that Cussia has a high requirement for P and exchangeable bases. This would explain
the negative correlation with K because of the antagonism between K and Mg. Our results
contour with observations by foresters that Cussia siamea grows well only in fertile soils,
a fact which limits its use in re-aforestation programs (Anonymous, 1981).
2 5 5
In contrast, all parameters ‘of Leucuenu growth showed negative correlations with
soil N, an indication that this woody species relied for its N supply on biological N,
fixation, a process inhibited by increasing levels of soil N. Apparently the N supplied
in the experimental soils enabled L.etia+nu the plant to make rapid and prolific growth
compared with cussia. ::.3:
L.eucaenu growth J&S also* shown to be influenced by available K in the soil
(r = 0.81 and 0.53 for plant height and total biomass respectively). These relatively high
correlations were likely due to the rather high soil Mg levels. There was a negative corm-.
lation though non-significant between plant height and exchangeable Mg (r = -0.55). As
K controls meristem growth, fast growing trees may be expected to have high K requi-
rcments. This has been demontrated for Leucuena which can respond to up to 800 ppm
K and also appears to be rather inefficient at extracting the element from the soil (Schmidt
and Saefuddin 1987). The K requirement of a plant is increased by an abundant N supply
(Mengel and Kirby 1987). Thus the K requirements of N,-fixing trees are likely more
pronounced than for non-N,-fixing trees. Indeed numerous studies have demonstrated that
K is required in greater quantities than any other soil mineral for total N accumulation
in legumes. It has also been establish that K-stimulated increases in legume dry matter
production are accompanied by increases in ,N, fiition when no fertilizer N is used.
This is believed to be caused by K-induced increases in photosynthesis, the transport
of photosynthates to the root nodules and increased root growth (Duke and Collins, 1985).
Fifty six percent of the variation in height of a related N,-fixing woody species AZbizia
fdcaturia over 28 different sites has been reported to be determined by soil K levels
(Delmacio 1987). Soil K levels of unfertilized soils were positively correlated with nodule
productivity of the N,-fixing Ingu jinicuil in Mexico (Van Kessel and Roskoski, 1981).
Relatively low levels of K fertilization stimulated N, fixation by this tree (Van Kessel
and Roskoski, 1983). The critical K level for maximum growth and N2 fixation of legu-
minous trees appears to be much higher than that required by cereals and needs to be
further investigated.
In this experiment, low soil N levels promote N, fixation and thus plant growth.
However, this condition increases plant K demand, and plant growth in turn becomes
limited by the soil supplied K. There appears to be an optimum K/N ratio (Fig-l).
It can be concluded that while Cussia is dependent on general soil fertility, particu-
larly N availability for adequate growth, nodulated Leucuena can grow well on poor soils,
if supplied with enough K and possibly P and Zn, which were adequately supplied in
this experiment. The relative growth of the two species at any one place will depend
on site factors. Cussie is likely to compete with interplanted food crops for N, but Leucuena
may not compete for K with cereals because legumes are usually less efficient than cereals
at extracting K from soils.
Aknowledgements
The authors thank the Federal Ministry of Education of Nigeria for the award of
a Commonwealth Scholoarship to S. LIYA.
256
References
ANDERSON, J.M. and INGRAM, J.S.I. 1987 Tropical soil biology and fertility methods
handbook. IUBS/UNESCO-MAB, Paris.
ANONYMOUS, 1981 Memento du Forestier. 2nd edition. Paris. Republique Fran&se,
Ministere de la Cooperation et du Developpement.
DELMACIO, M.I. 1987 Relationship between some site factors and growth of Albizia
falcataria. NFT Res. Rep. 5, 26-28.
DUKE, S.H. and COLLINS, M. 1985 Role of potassium in legume nitrogen fixation.
In : Munson R.D. (ed.) Potassium in Agriculture. pp 443446. American Society of
Agronomy, Madison, USA.
MENGEL, K. and KIRBY, E.A. 1987 Principles of plant nutrition. 4th edition. International
Potash Institue, Bern, Switzerland.
SCHMIDT, DR. and SAEFUDDIN, A. 1987 Response of Gliricidia and Leucaena to
potassium and lime. NFT Res. Rep. 5, 46-48.
VAN KESSEL, C. and ROSKOSKI, JP. 198 1 Nodulation and N,-fixation by Inga junicuil,
a woody legume in coffee plantation II. Effect of soil nutrients on nodulation and
N, fixation. Plant and soi1.59, 207-215.
VAN KESSEL, C. and ROSKOSKI, J.P. 1983 Nodulation and N, fixation by Inga junicuil,
a woody legume in coffee plantations III. Effect of fertilizers and soil shading on
nodulation and nitrogen fixation (acetylene reduction) of I. junicuil seedlings. Plant
and soi1.72, 95105.
MESURE DE LA FIXATION D’AZOTE
Maximisez la FBA pour la Production Agricole et Forest&e en Afrique
Evaluation of biological
nitrogen fixation in plants
DANSO, S&A.
’
Joint FAOJIAEA Division, P.O. Box 100, A-1400 Vienna, Austria
Abstract
Measuring the contriiution of atmospheric nitmgen (N2) to the growth of plants
is an kportmt step in understanding factors that influence the process, and for in-
creasing N2 fixation iu various systems. Although the uN techniques generally appear
to be superior to the other methods of measuring N, fixed, the facilities are not
available in many laboratories, and like all the others, it has its weaknesses as well
; undez some situations, or depending on the level of accuracy required. some of
the non-isotopic methods, many of which are less expensive, may be equally or
more suitable. The situations under which methods often regarded as inferior give
near accurate results have seldom been adequately highlighted. In this paper
thexfa the most fkequently used methods, total nitrogen difference ml9 acety-
lene reduction, xylem exudate analysis and the *W techniques have been examined
withthisinmind.Itisconcludedthatassimpleandinexpensiveasitis,theTND
approachcanundexmany cinxmsm give satisfactory estimates of N, fixed. The
acetylene reduction assay, on the other hand, has several drawbacks and cannot be
used to obtain quantitative measures of N2 fured. However. it should at least be
very useful in the initial scmening for wide diffexexes in N2 fuation among geno-
types and for distinguishing between absolute non-furers and potential N, furers.
259
Introduction
Increases in the contribution of nitrogen (N2) fixation to crop and tree growth,
and to high productivity of many agricultmaJ soils are attainable. It is however neces-
sary to lmow the current levels of N, fixed, so that any improvements and the conditions
necessary to achieve them can be ascertained. Several methods already exist, and have
been used to measure N, fixed in many crops and ecosystems. It is certain from all the
reviews on these methods that each has some disadvantages (Knowles, 1981 ; Chalk,
1985 ; Danso, 1985 ; Vase and Victoria, 1986). Judging from the diversity of N; fixing
crops and cropping systems, the prospects of an ideal method for ail situations are slim.
We are therefore compelled to make the best use of existing methodologies, insofar as
we are aware of their limitations, and to use the most appropriate for each particular
situation. This requires a thorough understanding of the strengths and weaknesses of each
methodology. Methods have unfortunately been on many occasions adjudged as the most
or least appropriate, without due regard for the situation. In this paper therefore, the
greatest strengths and weaknesses of the most commonly used methods for measuring
N, fixation will be examined, and as much as possible, the situation(s) under which each
is best suited will be assessed.
Total Nitrogen Difference (TND) Method
This is one of the oldest and simplest methods, and has provided us with many
estimates of N, fixed, upon which several useful management practices have been
devised. In its simplest form, the TND method is based on the difference between the
total N contents of plants or ecosystem in which N, is fmed and those that do not derive
N from N, fixation. The method in essence is based on an assumption that both the
N2 fixing and non-fming control plants absorb equal amounts of soil N for growth (Ren-
nie and Rennie, 1983). This assumption may not however hold under most situations,
as this would require that the different plants, in addition to being similar in root morpho-
logy and in several physiological attributes, should absorb their N from similar depths
and horizon. This assumption is therefore often the greatest limitation of the TND method
(Danso, 1985). In addition, where N, fixed in the whole ecosystem rather than only in
plants is being measured, the inability to recover completely all the different forms of
N using the usual Kjeldahl digestion procedures may result in erroneous estimates. For
measuring N, fnation in an ecosystem using the TND approach, it is also difficult to
quantify accu- rately all the many inputs and outputs of N into the ecosystem ; the errors
in these TND estimates therefore tend to be high, and can be very serious when N, fixa-
tion is low (because the analytical errors in the determination of total soil N alone may
sometimes be greater or close to the measured inputs from N, fixation). To overcome
this, especially in N-rich soils, N2 fixed is usually allowed to accumulate for a few to
several years before measurement. In this way, the TND method tends not to be only
tedious, but also requires long durations.
Despite the stated limitations, the TND method has often provided estimates that
are not significantly different from those obtained using more sophisticated and expen-
sive techniques (e.g., Ham, 1978 ; Phillips et al., 1983). These similarities therefore jus-
tify a close examination of the strengths of the TND method, and when it is most suitable.
As suggested by Danso (1988), and shown in the study of Paterson and LaRue (1983)
260
and Retmie (1984). the TND method reliably estimates of N, fixed in plants grown in
soils or systems in which the initial N content is low. Under such circumstances, almost
all the N would naturally be expected to be derived from N, fixation, with soil N contri-
buting only negligible amounts. Several such situations exist, e.g., in areas consisting
of sand dunes, systems undergoing primary colonization, and in many semi-arid areas.
It is therefore not surprising that the values of N, fixed in Casumina stands on the sand
dunes in Cape Verde (Dommergues, 1963) are still quoted universally, and Gauthier et
al. (1985) in their N, fixation studies on a sandy soil in Senegal frequently found no
sign&ant differences between TND estimates of N, fixed and those obtained using the
more expensive and more sophisticated (both mathematically and analytically) 15N me-
thodologies. The situation could, however, be different even where the soil N level is
low, depending on how much N, is fmed ; when this is low, the estimates made could
like when soil N is high, be highly erroneous. In this czule, the otherwise small errors
in the estimates of soil N uptake could be high, relative to N, fixed, and this would
account for many of the negative estimates of N, fixed that have been reported using
TND method (e.g. Patterson and LaRue, 1983 ; Rennie, 1984). All the same, it may
not be worth bothering much about the magnitude of such inaccurate estimates since in
my opinion, the real fact is, that the negative data only indicate that even if any N,
was fixed, it was so low that it could be regarded as negligible and not worth taking
note of.
The chances of very serious errors in TND estimates of N, fmed are certainly
high when inorganic N is high in the soil, unless the control non-N, fixing crop and
the N,-ftig crop absorb closely similar amounts of soil N. As reported by Rennie (MM),
this ideal situation is attained when the fertilizer utilization efficiencies (FUR) of the fixing
and control plants are equal, and this would have to be ascertained using ‘5N fertihzers.
However when the FUEs are unequal, fairly accurate estimates of N, fixed may
sometimes be obtained. As shown by Hatdamon et al. (1988), errors in N, fixation es-
timates become smaller and less important at higher N2 fmation rates, and although this
principle was demonstrated using only the 15N isotope dilution technique, the rationale
should hold for all the other methods. Thus, for some forage legumes, many of which
have been reported to derive around 80 to 90% of their N from Nz fixation (e.g., Herd]
et al.. 1968 ; Hardarson et al, 1988), it is likely that any erroneous estimate of N, fmed
asaresultoftheinaccurate
assessment of soil N uptake could be minimal, in contrast
to several grain legumes, such as many commonbean (Phusseotus vzdguris) varieties that
are known to be poor N, fixers.
Acetylene Reduction Technique Assay (ARA)
This methodology, used originally for N, fixation studies in pure microbial cul-
tures (Dilworth, 1966), and which was later developed for measuring N2 fixation in green-
house and field-grown crops (Hardy et al., 1968), is a paradoxical case of a method
which has, at one stage or another, been used in almost every laboratory for measuring
N, fixation, and which a little over a decade ago, stimulated an unprecedented interest
in N2 fixation studies, but which today, as shown in many reviews, is almost at the point
of becoming “the” unacceptable or obsolete way of measuring N, fixation (e.g. Witty
and h4inchin, 1988).
261
The early, fast and wide acceptance of the ARA technique for assessing N, fixa-
tion was due to its being a simple, rapid and highly sensitive method (Hardy et al., 1968),
and a reasonably lower instrumentational cost. Also, since ethylene is stable, the final
product could thus be stored (allowing large samples to be collected at a time) and analysed
when most convenient.
Although the early and popular acceptance of the ARA technique, for measuring
N, fixation was almost universal, many of its drawbacks soon became apparent, and various
attempts were made to overcome or minimize their effects. For example, several reports
called attention to the instantaneous nature’of ARA assays, and the doubtful validity in
extrapolating these short-term assays to cover N, fmed over whole growing seasons (in-
volving large diurnal, daily and sometimes seasonal variations in rates of N2 fixation)
(Rennie et al., 1978). Attempts to reduce this problem have- included making several se-
quential assays rather than a single or few measurements (e.g. Hardy et ol., 1968). Because
nitrogenase activity declines significantly in detached nodules, roots with attached nodules
more frequently replaced the use of detached nodules. Also, studies showed that the digging
of roots (especially in field-grown plants) resulted in unavoidable nodule losses, and the
subsequent preparation of root samples for ARA assays, such as washing of roots (e.g.
Carta and Sheaffer, 1983) or the drying of nodules in the field or after sampliig, adver-
sely affected the diffusion of gases, such as oxygen and acetylene into the nodules (Minchin
et al., 1983 ; Weisz et al., 1985 ; Durand et al., 1987). Also, it was noticed that the
levels of acetylene used sometimes inhibited the process being measured (Witty and
Minchin, 1988), while nodulated root segments exhibited a lower ARA than intact plants
(Wych and Rains, 1978). Efforts made to solve or minimize these problems included
conducting nondestructive in situ assays on intact plants in the field or greenhouse rather
than on decapitated roots (Wych and Rains, 1978 ; Pfeifer et al., 1983), and the use
of a flow-through gaz system (Minchin et al., 1983 ; Fellows et al., 1987) which allowed
a low and non-toxic concentration of acetylene to be continuously passed over the N,-
fixing system. None of these modifications was still able to overcome the need for artifi-
cial enclosures for the plants, and it was also often necessary to correct for the endo-
genous production of ethylene from soils.
Furthermore, it became known early that the validity of the factor of 3 originally
proposed for converting C&Id produced into values of N, fared (Hardy et al., 1968) was
not valid for all crops and systems (Rennie et al., 1978 ; Hudd et al., 1980). This discre-
pancy was largely attributed to all the electrons becoming available to nitrogenase in
the ARA assay, whereas Hudd et al. (1980), for example, reported that only about half
the electron supply was used to fix N, in field bean (Viciu f&u). resulting in a true
conversion ratio of 5.75 rather than 3. This value was also not universally acceptable,
since experimentally determined ratios have been reported to differ widely among diffe-
rent N, fixing plants and systems, with some of the reported ratios ranging from 1.5
to 8.4 (Bergersen, 1970 ; Hardy et al., 1973 ; Rermie et al., 1978), and perhaps greater.
Reunie et al. (1978) therefore recommended the experimental determination of the ratios
to use under each situation using lsN assays for calibration of the ARA technique.
However, lsN, gas is expensive, and th$ approach would require determining the correct
ratio to use for each study, as for the same crop, the electron flux (which the ARA tech-
nique essentially measures, rather than direct N, fmation) may differ under different experi-
mental conditions or assay periods (Witty and Minchin, 1988). It was therefore not
surprising that conversion of ARA assays into kg N/ha of N, fixed became less popular,
262
and was more often expressed as ethylene produced (either in nM or pM) per plant or
nodule, and per unit time. The shift allowed seemingly valid relative comparisons of treat-
ment effects to be made, by avoiding the need to use controversial conversion ratios.
This approach held much promise for comparing N, fixation abilities (where the actual
amounts of N, fmed were not essential or critical) until later evidence suggested that
insofar as electron fly, could be influenced differently by different treatments, such rela-
tive comparisons were of questionable validity (Witty and Minchin, 1988).
For those still dedicated to using the ARA technique, it seems that it may soon
become difficult for their results to be accepted, unless the method can somehow or the
other be imporved beyond its present capabilities. These adherents may however argue
that, despite all these stated problems, numerous AFCA estimates of N, fixation, statis-
tically similar or highly correlated to values obtained by other “acceptable” methods
abound in the literature (e-g. Hardy, et al., 1968 ; Paterson and LaRue, 1983). These
results lead us to question whether all the methods used in these cases gave wrong results
or that the limitation of ARA have in reality been exaggerated. These are valid ques-
tions for all scientists interested in N, fEation studies. It would be necessary to closely
examine how often ARA estimates have deviated significantly from those obtained by
other methods ; this would be particularly importaut in studies to rank N, fixation capa-
bilks. for which the ARA technique at one stage held much promise. If these deviations
are found to be only occasional, a case may have been established for Examining when
the ARA technique could be used to obtain valid results, and what precautions are neces-
sary. The method at least holds much promise for identifying potential N, fixing crops
in novel communities, and for grouping plants into broad categories in terms of N2 fixation
ability. It is also currently the only method that has consistently been used to establish
that a plant intended to be used as a reference crop (m the 15N methodology, and inci-
dentally often the method of choice now) is truly a non-N, fixer (e.g. Fried er al., 1983).
Relative Urelde Production
The xylem sap of N, fixing plants contains N compounds originating not only
from soil N uptake by roots, but in addition, contains N of nodule origin, through the
process of atmospheric N, fixation. In some N,-fming tropical legumes, differences in
the N compounds carried by the xylem sap have been found. This is influenced largely
by whether the N is of soil or nodule-origin (Pate et al., 1980 ; Herridge. 1982s) ; the
soil derived N is exported largely as free NO,- and amide amino acids (e.g. glutamine
and aspam&@, while N, fixation results largely in the formation and export of ureides
(e.g. allantoin and allautoic acid). The relative composition of ureides versus NO,- + amides
in these cases therefore serves as a useful index of the sources of the plant’s N. ‘Ibe
composition of the xyIem sap changes progressively from one dominated by ureides to
one composed largely of NO,- and amino compounds as N, fixation declines (Herridge,
1982a).The assay of ureides, contrary to what many expect, is not that sophisticated, and
involves only a simple calorimetric assay. However, like most calorimetric assays, it needs
to be calibrated, and both the ARA and ‘%I techniques have been used for this purpose
(Pate et al., 1980 ; Heniedge, 1982a).
The initial handicaps of the methodology included (i) obtaining a representative
exudate, since the ureide content of the plant was not. uniform in all organs (Herridge,
Maximiser la FBA pour la Production Agricole et Forestière en Afrique
Influence du déficit hydrique
sur la fixation symbiotique
de l’azote atmosphérique chez le soja
SALL, K., DREXON, J.J. et OBATON M.
Laboratoire de Recherches sur les Symbiotes des Racines,
lNRA. 1, place viala, 34060 Montpellier cedex 1, France
RQsumé
L’effet du déficit hydrique sur l’activité nitrogt?nase est étudié chez le soja par
application de mannitol. Ce travail a Bté Alis& en serre en culture hydroponique.
sur deux variktés de soja pnkentan~ dans les essais agronomiques. des sfmsibilit.&
diffkrentes à la sécheresse : Kingsoy n5sistante et Hodgson sensible. Sur ces plantes
de soja en symbiose avec une souche de Bradyrhizobium jqponicum, sont mesures
I’activitr? fïxahice d’azote et I’effet de pressions partielles d’oxygène croissantes,
sous diffhntes intensitks de contrainte osmotique : - 0.2, -0.4. -0.6 et -0.8 MPa
De fortes pressions partielles d’oxygene permettent de lever sen.sibkxnent,
mais pas
totalement, l’inhibition de la fixation due B la contrainte hydrique. L’infbrence de
la contrainte osmotique sur la nitrog&urse varie suivant la teneur en nitrate du milieu
de culture. Ces nkultats suggèrent que le déficit hydrique limite l’alimentation des
bactkroïdes en oxygi?ne mais egalement affecte la quantid de pouvoir r&hrcteur et
la quant% d’enzyme active @sente dans les nodositks pour la fixation de l’azote.
La fixation de l’azote est affect& diff&ennnent chez les deux cultîvars : elle est
moins dkpritnee chez Hodgson que chez Kinsoy. apparenunent parce que Hodgson
posskle une réserve énergétique plus importante dans ses nodosit&.
273
Introduction
Dans les conditions de la culture du soja en plein champ chez les plantes inocult5es
avec Brudyrhizobium juponicum, la fixation symbiotique de l’azote atmosphkrique fléchit
rapidement, pouvant même s’annuler lorsque le déficit hydrique est s&&e. (Obaton et
al., 1982 ; Bouniols et al., 1982 ; Wt5ry et al., 1986).
La baisse de l’activité nitrogénase par le déficit hydrique est observée par la méthode
de l’activité réductrice de l’ac&yEne (ARA) (Pankhurst et Sprent, 1975 ; Sheehy et ul.,
1983 ; Benett et Albrecht, 1984) et par la mCthode 15N (Sinclair et Goudriaan, 1981).
Alors que certains auteurs considknt que la diminution de l’activité nitrog&we est due
A l’inhibition de la photosynthbe (Huang er al., 1973 ; Sprent, 1976), Finn et Brun (1982)
concluent que la nitrogenase est plus sensible au &fkit hychique qu’à la photosynth&se.
Cette interprétation est soutenue par Bennet et Albrecht (1984) qui ont montré que la
fucation de l’azote atmosphkrique est étroitement corrt5Ee au-potentiel hydrique des nodules,
ce dernier &ant plus sensible au déficit hydrique que le potentiel hydrique foliaire et
la résistance stomatique.
Selon Weisz er al. (1983, l’inhibition de l’activid nitrogknase par le déficit hydrique
est détectable au champ trois jours seulement aprRs l’installation de la contrainte hydrique
et avant qu’il y ait une différence pour la surface foliaire, Ie poids des plantes ou des
nodules et le nombre de nodosités entre plantes stressks et plantes témoins.
Pankhurst et Sprent (1975) notent qu’un accroissement de la pression partielle d’oxy-
gène (pOJ lors d’un déficit hydrique, augmente la fixation de l’azote mesur& par l’activité
r&luctrice d’acétylène (ARA) et la respiration des nodules. Ce r&ultat obtenu sur nodules
excisks nous a conduit à réaliser des exp&iences similaires sur plantes entières pour essayer
de mieux comprendre ce phénomène, tout en perturbant au minimum la physiologie des
plantes.
Matbriels et méthodes
Matbrie biologique et culture des plantes
Deux vari&% de soja (GZycti max L. Merrill) sont cultiv&s : Hodgson du groupe
1 demi-prkoce, sensible à la skheresse, et Kingsoy du groupe II demi-tardif, retenue
pour sa bonne tol&ance à la skheresse (Bouniols, con-un. pers.).
L-es graines de soja, après dtifection A lhypochlorite de calcium (30 g 1-l pendant
30 mn), sont mises à germer a I’obscuriti durant 3 jours A 28°C sur vermiculite sttkile.
Les plantules de soja sont placées sur des plaques de polystyr?ne prkilablement pwc&s
pour laisser croître la partie racinaire. L’ensemble est ensuite placé sur des bacs de 40
litres (60 cm x 40 cm) contenant la solution nutritive a&& dont le pH est maintenu
à6J.Lesplantulesdesojasontinoc~~d~lamiseenplaceaveclasouchedeBradyrhizobium
japonicum G 49 (SB 16, IARI, Inde). Les cultures sont conduites en serre &Clair& pendant
16 h (klairement additionnel pendant l’hiver, 100 W mw2 par des lampes Philipps), à
une tempkrature variant entre 15 et 35°C. Pendant la phase obscure de 8 heures. la température
varie entre 15 et 20°C. L’humiditk relative, augmentée grâce A l’action d’un brumisateur,
est supérieure 3 70% et varie suivant les saisons.
274
La solution nutritive contient : qP0, : 025 p.M ; ,CaClr : 33 p.M ; K$O, :
125 plvl ; MgSO,. 7qO : 2.05 @I ; %BO, : 4pM ; MnSO, : 6,6 pM ; ZnSO,,7%0:
1,55 p.M ; CuSG4,5$0 : 1,56 p.M ; Na$foG,, 2qO : 0,12 p.M et du fer sous forme
d’EDTA (0,015 g 1-i de produit). Pendant les 2 premikes semaines, l’azote de complement
est apporte sous forme duree (1 mM). Lorsque la nodnlation est install& les plantes
ne reçoivent aucune autre source d’azote que la fixation sauf dans les exp&iences avec
nitrate où l’apport d’ut% est remplace par des doses variables de NO,-. Le milieu de culture
est retrouvele toutes les semaines, et r@likrement rkajusté avec de l’eau distillee.
Intensité! et mode d’application du dbficit hydrique
La contrainte hydrique est obtenue par abaissement du potentiel osmotique de la
solution par apport de mannitol pendant 24 h. Le mannitol est peu ou pas absorbe par
les cellules (Heath, 1977) et ne prksente pas de phytotoxicité. Pour eviter la multiplication
des microorganismes dans la solution nutritive, un antibiotique, la t&racycline, est apporte
(10 mg 1-l) avec le mannitol. Quatre traitements sont rt!.ali&-s : -0.2 MPa (0,074 M) :
deficit faible ; -0,4 MPa (0,153 M) : déficit moyen ; -0,6 MPa (0,233 M) : deficit fort
mais non destructif ; -08 MPa (0,312 4 M) : déficit tres fort et plus ou moins destructif
suivant les conditions climatiques.
Mesures des échanges gazeux
Les diffkentes mesures effectukes dans ce travail, sont faites à partir du qnarantieme
jour de culture ; les plantes sont alors en pleine floraison, stade le plus sensible a la
sécheresse (Bouniols ef ul. 1982). Les mesures de l’activité nitrogenase sont rt%is&s par
la technique de rkluction de l’ac&ylkne en dthylene (Hardy et cd., 1968) appliquée a des
racines nodulees en place SUT la plante. L’incubation est faite en flacon s&um de 500
ou 1000 ml, ce qui permet de r&liser des mesures non destructives de l’activité nitrog&ase.
LWanchei~ au niveau de la tige est assur& par un joint de coton imbibe de Rhodorsil
RTV 1502 (Rhône-Poulenc). Les incubations en presence de C,H, sont réalisees soit en
systkne ferme, soit en continu avec balayage du milieu racinaire par un mélange gazeux
contenant 10% CA.
Mesures en systbme fermé - Le protocole appliqué est celui dkrit par
Kalia et Drevon (1985). La partie racinak de la plante est herm&iquement enfermé
dans le flacon s&um, avec 10% (v/v) de solution nutrive pour maintenir la turgescence
et l’activid des organes. La nitrogtkw est saturee par une concentration en ac&yl$ne
de 10%. La quantité d’&hyEne accumulée dans ce compartiment mcinaire est &ahK?e
par des pr&vements d’&antillons de l’atrnosphere interne qui sont ensuite analy& B
l’aide d’un chromatographe Girdel30. muni d’un détecteur à ionisation de flamme (colonne
de Porapak T: 1,5 m de long et 3.2 mm de diam&re, tempkature du four : 80°C. gaz
vecteur azote à 30 ml mn-‘). Les mesures d’activité nitrogenase (ARA) obtenues dans
ces conditions sont plus élev&s que celles obtenues sur les racines dparees car l’excision
entraîne une baisse de 50 3 80% de l’activitt enzymatique @ilston et Imsande, 1982).
Toutefois, en systkme fermé, la rkluction d’ackylène ne reste constante que pendant environ
60 mn (Kalia et Drevon, 1985). Les mesures doivent donc être effectuées rapidement
275
Mesures en systtkne continu - Le principe de mesure de l’activité de
la nitrogénase est le même qu’en systeme ferme, mais cette technique a l’avantage de
permettre le contrôle du mélange gazeux qui balaie en permanence l’atmosphère du
système racinaire. Ainsi, grâce à des débitmètres Tylan, le flacon est balayé par un me-
lange gazeux a debit constant et contenant de l’adtylène (NS), de l’oxygène et de l’azote
avec la possibilité de moduler les concentrations respectives de ces deux gaz. L.e systkme
comprend le flacon s&um qui porte la plante, les débitmètres connectés aux différentes
bouteilles de gaz et un tuyau de sortie du milieu d’incubation qui est connecté au chroma-
tographe et au catharomètte à conductivité thermique. Le catharomètre permet de doser
les quantites d’oxygene et d’azote contenues dans les 6chantillons. Les deux appareils sont
munis chacun d’une vanne d’injection automatique qui fonctionne alternativement à la
fréquence programmk. Un autre débitmètre à bulles de savon, relié à la sortie du catha-
romètre, permet de vérifier le débit du mélange gazeux. Il est ainsi possible & suivre
l’evolution de l’activité nitrogénase en fonction de la pression d’oxygène (Drevon et al.,
1988). Le dispositif permet de réaliser quatre incubations simultanément Des prélève-
ments à la seringue d’échantiIlons de l’atmosphère racinaire permettent de mesurer le déga-
gement de CO, par des dosages avec un IRGA ADC. 225 MK3.
R&ultats
Influence de la contrainte osmotique sur I’ARA à 21% 0,
Les donnees présentées à la Fig.lA montrent que chez Kingsoy, l’ARA diminue
des les premieres heures qui suivent l’application du manuitol dans le compartiment ra-
cinaim. Cette inhibition de l’activite nitrogénase est d’autant plus forte que l’intensiti du
déficit osmotique est e1ewS.e. Apres 24 h en prksence de mannitol, l’ARA exprim& en
pourcent de celle du tkmoin n’ayant pas rqu de mannitol, n’est plus que de 72, 63, 52
et 29% respectivement pour une pression osmotique de -0,2, -0,4, -0,6 et -0.8 MPa. Après
avoir 1evC la contrainte osmotique, I’ARA continue a dkcliner pendant au moins 4 jours
chez les plantes ayant Cté soumises a la contrainte osmotique de -0,8 h4Pa. Bar contre,
chez les plantes soumises à -0,4 et -0,6 MPa, ce déclin d’activité cesse apparemment
après 2 jours, et chez celles soumises à une contrainte de -0,2 MPa la stabilisation de
SARA s’effectue aprks environ 24 h. Chez ces demieres, l’ARA augmente même signi-
ficativement durant le troisieme jour.
Chez la variété Hodgson (Fig.lB), l’ARA est apparemment moins affect& que chez
la varieté Kingsoy. En effet, apr& 24 h de contrainte osmotique, I’ARA exprimk en
pourcent du témoin n’est que de 82 a 62% pour les trois traitements. Il n’y aurait pas
de différence significative de l’inhibition de l’activité nitrogénase chez cette vari&& pour
des concentrations de mannitol correspondant a -0,4 et -0,6 M?a. Apres retrait du man-
nitol, la baisse de la fixation se prolonge encore pendant environ 3 jours pour les plantes
ayant subi les traitements Ct -0.4 et -0,6 MF’a alors que la stabilisation de I’ARA est plus
rapide a -0,2 MPa comme pour la variété Kingsoy. Ainsi, 4 jours après la fii de la con-
trainte osmotique, les plantes précédemment soumises a -0,2 MPa retrouvent 84 % de
leur activité fixatrice d’azote, contre 52 et 40 pour -0,4 et -0,6 MPa respectivement.
276
80
60
4 0
6 0
80
100
120
140
- 2 0
p
20)
4 0
6 0
8 0
1 0 0
1 2 0
140
mn!ralnte hydripue
TEMPS en HEURES
Fiiure 1 : Jnfluence du déficit hydrique sur la fixation symbiotique de l’azote mesur& par la méthode
ARA, chez les vari&tés de Soja KINGSOY (A) et HODGSON (B). Les résultats sont exprim& en
pourcentage du témoin. Chaque point est la moyenne de 5 rép&itions. Le traitement au mannitol dure
24 heures. Les mesures sont faites 1 heure avant le traitement, 5 h et 24 h après. Les mesures sui-
vantes sont effectuées au retour à un régime hydrique non limitant. 2 bars A ,4 bars A ,6 bars l
,
Sbarso
277
Influence du d6ficit hydrique sur la repense de I’ARA à la p0,
Les donnkes de la Fig.2 montrent qu’après environ 800 mn (environ 13 heures)
d’application d’une contrainte osmotique a l’aide de mannitol, 1’ARA devient à peu pr&s
stable. Cette stabilité relative permet donc d’étudier l’influence de p0, croissantes appliqukes
progressivement pendant plusieurs heures.
Chez Kingsoy, sur les plantes témoins n’ayant pas subi de contrainte osmotique,
l’ARA augmente jusqu’à 30% d’oxygkne puis baisse pour des p0, supkeures (Fig.3A),
En présence d’une contrainte osmotique de -0,2 et -0.4 MPa, l’AlU maximale (ARA
4
8BARS
6 0 0
1000
1 2 0 0
1400
1 6 0 0
1 8 0 0
-TEMPS EN MINUTES
Figure 2 : CinCtique de I’ARA chez le soja variéti HOLXXON. en fonction de l’intensité de la
contrainte osmotique (appliquée à l’aide de mannitol) : 0,4 , et 8 bars. Les mesures sont faites eu
continu à 20% d’oxygène. Chaque point est la moyenue de 5 à 8 n?p&tions.
ARA X du TE à 20% d’02
ARA X du TE à 20% d’02
w
0
0
B
:
0
0
0
279
correspondant a la p0, optimale) est atteinte pour une p0, plus élevée, voisine de 40%
et sup&ieure a 60% pour -0,6 IvlPa. La p0, optimale est donc damant plus élevée que
la contrainte hydrique est plus s&ère. Toutefois l’augmentation de la p0, ne permet pas
d’obtenir une ARA maximale comparable à celle du témoin.
Chez la variete Hodgson, en l’absence de contrainte osmotique, la p0, optimale observ&
dans fa Fig.3B est plus élevée que chez Kingsoy. En effet TARA est multipliée par 2
lorsque la p0, augmente de 20 à 40% d’oxygene ; elle baisse au delà de SO% d’oxygene.
De même que chez Kingsoy, l’ARA maximale en presence d’une contrainte osmotique
est observée pour des p0, d’autant plus élev&s que la concentration en mannitol est plus
forte. L’ARA a 60% 0, est alors d’environ 80 à 90% de celle du temoin n’ayant pas
subi une contrainte osmotique pour les traitements -0,2 et -0,4 MPa ; mais pour un deficit
hydrique sévère (-0,6 MPa) l’effet de la p0, est beaucoup plus faible (environ 25% de
1’ABA du témoin pour une p0, de 60% 03.
Influence du deficit hydrique sur la respiration nodulaire
En syst$me fermé chez Kingsoy, la respiration du système racinaim nodule est
influencée par l’intensif de la contrainte hydrique (Fig. 4A). Pendant le déficit hydrique,
la respiration diminue d’autant plus que la contrainte osmotique est plus intense. Au retour
a un régime hydrique non limitant, la respiration reprend legèrement, surtout pour la faible
intensité, mais ne retrouve pas le niveau du tkmoin. Hodgson se comporte de la même
façon que Kingsoy, mais est legèrement plus affecte que celui-ci, surtout pour le trai-
tement -0,6 h4Pa (FigAB). Après 24 h. de contrainte osmotique, la respiration chez
Kingsoy est de 95, 70 et 67% du témoin contre 88, 62 et 55% chez Hodgson respec-
tivement pour -0,2, -0,4 et -0,6 MPa.
Dans les expkïences de r@onse B l’oxygkne en continu, aussi bien chez Kingsoy
que chez Hodgson, le dégagement de CO, augmente avec la p0, au moins jusqu’à 60%
Or Contrairement à I’ARA, il n’y a pas Ctablissement d’un plateau et la respiration semble
être limitée par la p0, quelle que soit la variété ou l’intensité du déficit hydrique (Fig.5).
Ces mésures de l’émission de CO, réaMes simultanément a celles d’ARA permettent
d’etablir une relation linéaire entre ces deux activités (Witty er aI., 1983). La pente de
la droite de régression ainsi obtenue correspond à la quantité de CO, émise par molécule
de C,H, réduite. L’ordonnée a l’origine représente la respiration des tissus racinaire et
nodulaire qui n’est pas associée a la fixation. Les donnees présentées dans le
Tableau 1 ne rév&lent aucune difference significative entre les traitements pour ces deux
paran&tres, excepté pour la contrainte de -0,2 MPa. Les r&ultats obtenus chez Hogdson
(donndes non prkentkes) sont similaires. Ainsi la contrainte osmotique de -0,2 MPa aug-
menterait le coût Cnergetique de la fixation puisque le rapport vitesse d’krnission de
COJARA est voisin de ,lO p.mol CO,.p.mol-r C,H, pour -0,2 Mpa contre environ 5 chez
le temoin. Cette observation mériterait d’être vCrif& dans des travaux ultkieurs.
Influence de l’azote minkal sur I’lnhlbition par le deficit hydrique
L’action du nitrate est dépressive sur la nitrogenase : en effet la fixation est faible
chez les plantes avec nitrate et sans déficit hydrique comparees a celles soumises au même
traitement maïs sans nitrate (Tableau 2). Apres 24 h en présence de mannitol, les plantes
sans nitrate maintiennent une activité de plus 60% du témoin contre 12% chez les plantes
i., -i
~j.‘,.. ---
‘.
20
40
60
80
100
120
140
-20
f
27
40
60
80
T;;ps,,;~~,,;o
contrainte hydriaue
0 bar4 2 barsa 4 b a r s II
6 bars ü
8 bars x
Figure 4 : Influence de la contrainte osmotique sur la respiration des racines
nodulkes. Les mesures
sont effectuées simultakment
aux mesures d’ARA en système fermk, chez les vtidtis de soja
IUNGSOY (A) et HODGSON (T3). Le traitement au mannitol dure 24 heures. Les mesures sont
faites 1 heure avant le traitement, 5h et 24h après. Les mesures suivantes sont effectuées après retour
à un régime hydrique non limitant.
281
200
g 150
-ü-oc3d 100
E
3
pc
50
z
u
0
10
20
30
40
PO* m 02)
Obar A 2barsA 4 b a r s n 6barst-J
Figure 5 : Influence de la contrainte osmotique sur la respiration des racines noduM.es. Les
mesures sont effectuées simultan6ment aux mesures d’ARA en systhme famt, chez les vari6tis de
soja
KINGSOY (A) et HODGSON (T3). Les dsultats sont exprimks en pourcentage du tbrnoïn à
20% d’oxygène. Chaque point est la moyenne de 8 B 16 mesures sur une plante. L’expb-ience est
rép&?e deux fois.
282
Tableau 1 : Influence du déficit hydrique sur la respiration du système
racinaire nodulé et sur le coût énergétique de la fixation
de l'azote (ARAI chez le soja, variété Kingsoy. Les mesures
d'ARA sont faites en système continu. La respiration basale
correspond à l'ordonnée à l'origine de la droite de régres-
sion simple entre 1'ARA et la respiration, et le ooüt éner-
gétique est la pente de cette droite.
Intensité de la contrainte
Respiration basale
Coût énergétique de
osmotique en MPa
nitrogénase
'
pmol CO2 h-' pl-'
pmol CO2 umol -' C2H4
0
255 + 68
5,5 + 1,o
-
-
-0,Z
146 + 5 8
9,4 + 2,1
-
-
-0,4
224 + 89
4,3 + 2,6
-
-
-0,6
234 t 44
6,l + 2,l
-
:,
-.
.
.<
.
.
.
i
-1.
.
.
283
Tableau 2 : Effets cumulés du nitrate et du déficit hydrique sur la fixation
symbiotique de l'azote (ARAI du soja, variété Hodgson ; intensité
de la contrainte osmotique -0,4 MPa, pendant 24 h. Le point -1
correspond aux mesures faites une heure avant application du
mannitol, et le point 120h à celles réalisées 96h soit 4 jours
après suppression de la contrainte osmotique.
Heures
-1
ARA : umoles C2H4h -' Plante
après stress
Sans déficit hydrique
Avec déficit hydrique
-NO-3
+NO-3
-NO-3
+NO- 3
-1
52,0
11,o
52,0
11,o
+5
62,5
13,2
42,4
172
+24
59,7
25,0
37,1
330
+120
86,4
10,8
45,4
4,O
284
cultivkes avec 3 mM de nitrate. L’inhibition de la fixation par le nitrate et le déficit hy-
drique sont donc approximativement cumulatives.
A la suppression de la contrainte hydrique, les valeurs de l’ARA sont très faibles
chez les plantes avec nitrate, mais augmenteut assez rapidement. Chez les plantes sans
nitrate, l’inhibition de fa nitrog&rase n’est que partiellement levke.
Discussion et conclusion
D’après les travaux initiaux de Sprent (1976), la contrainte osmotique inhibe la
fition syrnbiotique de l’azote en limitant la quantiti d’oxygkne disponible dans les no-
dosit6s pour la synthkse de I’ATP qui est indispensable au fonctionnement de la nitro-
ghase. Les rthltats obtenus dans ce travail sur nodosites intactes en place snr la plante
sont en accord avec cette hypothèse. En effet l’inhibition de l’ARA par la comrainte os-
motique, due à l’apport du mannitol dans le milieu racinaire à 20% d’oxigeue, est lev6e
partiellement par l’augmentation de la p0, externe. Le d&placement de la p0, optimale
vers des valeurs d’autant plus Clevtks que la concentration de mannitol est plus forte,
ainsi que les observations sur la respiration nodulaire, renforcent cette interpretation et
suggerent que la diffusion de l’oxygene est perturbee par la contrainte osmotique.
Cependant, la manipulation de la p0, externe ne permet pas de lever totalement lkhi-
bition. En effet les valeurs dARA maximale en prksence de la contrainte osmotique sont
infkieures a celles observees snr les plantes tkrnoin ne recevant pas de mannitol. La
limitation de I’ARA par l’oxygène ne serait donc pas le seul mécanisme implique dans
l’inhibition de la fixation par une comrainte osmotique.
L’ARA maximale est vraisemblablement déterminée par le flux de pouvoir
réducteur aux bactkoïdes (Heckmann et Drevon, 1988) et par la quantité d’enzyme
active. L’irr&ersibilid de l’inhibition pour des pressions osmotiques inferieures à
-0.2 MPa suggère qu’tme fraction des protéines îmiispensables a fa fixation a &t5 dkruite
durant les 24 h d’incubation en presence du mannitol. Ceci est apparemment confkme
par l’observation de la Fig.2 : l’inhibition de l’AFUI durant les premières heures d’incu-
bation en prtkence de mannitol serait due à un ralentissement de la diffusion de 0, tandis
que la baisse plus lente d’activite nitrgenaso observée entre 14 et 20 h s’expliquerait par
une d&krioration du tissu nodulaire, due elle-même à un manque prolonge d’oxygene ou
encore a un manque d’énergie. Il est toutefois peu probable que œ manque d’énergie
au niveau nodulake provienne d’une diminution de la photosynth&se sous l’effet de la
contrairite osmotique car une diminution dela réduction foliaire de CO, par limitation
du flux de photons (Schweitzer et Harper, 1980) ou une stimulation par la pC0, ak-ienne
(I%n et Brun, 1982) n’affectent pas I’ARA avant au moins 24 h. Il est par contre pos-
sible que la contrantte osmotique ait ralenti le transfert des photosynthats aux nodositks,
œ qui d’après Sloger er al., (1975), diminue I’ARA à court terme, vraisemblablement
en limitant la quantid de pouvoir rkhtcteur disponible pour le fonctionnement des bac-
t&oides. Afin de vkritïer ces hypothkses, ii conviendrait de tester l’effet a plus court terme
du mannitol sur l’inhibition de la fixation et la reversibilid de cette inhibition, et de mesurer
l’activit6 ex-planta de bactkïdes prklev6.s à des moments successifs de la cidtique d’inhi-
bition. La présence de nitrate dans la solution nutritive pourrait, par son accumulation
vacuolaire, diminuer l’inhibition en permettant à la plante et aux nodositks de maintenir
leur turgescence. Les résultats pn%entés dans le Tableau 1 ne soutiennent pas cette hy-
285
pothbse puisque le pourcentage d’inhibition de I’ARA par une contrainte osmotique de
5,4 MPa est suphieure en présence (88%) qu’en absence (38%) de NO,-. L’inhibition
plus forte en prksence de cet ion serait due à son effet inhibiteur propre qui augmenterait
la rt!.sistance corticale a la diffusion de 0, dans les nodosités (Minchin et ai., 1986) et
diminuerait la quantité de pouvoir rkiucteur disponible pour les bactéroïdes (Heckmann
et Drevon, 1987). Ceci pourrait également être dû Zr une accumulation dans la nodosité
de nitrite issu de la réduction du nitrate (Rigaud, 1976 ; Streeter, 1982) d’autant plus
que cette accumulation de nitrite est plus forte en condition d’aliientation lim$ante de
0, (Hedmann et Drevon, 1987).
La comparaison des denx cultivars en l’absence de pertwbation osmotique, suggère
que Hogdson aurait d’avantage de rkserves glucidiques facilement mobilisables que
Kingsoy puisque sa p0, optimale pour l’ARA est plus élevée. Ceci pourrait expliquer
la meilleure r&wsibilite & l’inhibition chez Hogdson et soutiendrait lbypothkse d’un rôle
des réserves glucidiques dans le maintien de l’int&rité des nodositks en prksence d’une
contrainte osmotique de moins de 24 h.
Du point de vue agronomique, cette étude entraîne deux remarques :
(i) Il est reconnu que l’apport de nitrate à une culture permet B celle-ci
de mieux rkster à la sécheresse, car cet anion joue le rôle d’osmoticnm
dans la vacuole et permet à la plante de rester turgescente. Cependant,
cet anion diminue egalement la fixation et son rôle dépressif s’ajoute
ii cehri de la contrainte hydrique. Dans ce cas, le soja perd son intkêt
comme plante f-ce. Il serait utile de faire la même étude sur les
1Qumineuses tres résistantes à la skcheresse, telles celles des zones
dksertiques.
(ii) Le cuhivar de soja Kingsoy, dont le rendement est moins diminue que
celui d’Hodgson par une irrigation insuffisante (Paul, com. pers.), pn%kente
cependant une fixation d’azote très sensible au déficit hydrique. Cette
apparente contradiction est sans doute due au fait qu’un defîcit hydrique
de plusieurs jours ne provoque pas de perturbations physiologiques pro-
fondes chez Kinsoy du fait que la &ulahon stomatique est efficace
chez cette vari&.& (Salt, 1984). Par contre, œ déficit hydrique dkprime
plus vite et plus fortement son activite nitrog&tse, cette varieti pos-
skdant des nodositks ayant une plus faible rtkerve tkrgétique que celles
d’Hodgson.
Il y aurait donc sans doute intkrêt it rechercher des vari&& peu permrbees par
le delïcit hydrique, a la fois pour le rendement et la fixation de l’azote, ou additionner
œs deux crit&es par croisement.
Ce travail pr&minaire qui demande confiiation, notamment en poursuivant la
même recherche sur d’autres cultivars de soja ou même sur d’autres espèces de légumi-
neuses, montre l’int&t d’une étude physiologique sur les différentes étapes de Klabo-
ration du rendement pour optimiser la production d’une culture.
286
Remerciements
Les auteurs remercient Nathalie OLLAT pour son aide lors des premières exp&
riences sur la @onse de I’ARA a l’oxygkne. Ce travail a été partiellement fmcé par
la CEE-DG Xll (Contrat STD 1, No TS-A-180 F) et le Minist&re de llndustrie et de
la Recherche (Contrat 84 X 0881).
RQfbrences
BENETT, J.M. et ALBRECHT, S.L. 1984 Drought and flooding effects on N, fixation,
water relation, and diffusive resistance of soybean. Agron. J. 76, 735-740.
BOUNJOLS, A., PUECH, J. et MONDJES, M. 1982 Muence d’un déficit hydrique ap
plique durant la fructification sur la production du soja (Glyck mar L. Merrill).
Bull. Groupe français HumiaTimktre Neutronique. 11, 39-55.
DREVON, J J., KALJA, V.C., HECKMANN, M.O. et PEDELAHORE, P. 1988 In situ
open-flow assay of acetylene reduction activity by soybean nodules : influence of
acetylene and oxygen. Plant Physiol. Bbchem. 26, 73-78.
FJNN, A.G. and BRUN, A.W. 1980 Water stress effects on CO, assimilation, photo-
synthate partitioning, stomatal resistance and nodule activity in soybean. Crop Sci.
20, 431-434.
FJNN, A.G. and BRUN, A.W. 1982 Effect of atmospheric CO, enrichment on growth,
nonstructural carbohydrate content and root nodule activity in soybean. Plant Physiol.
69, 327-331.
HARDY, R.W.F., HOLSTElN, R.D., JACKSON, E.K. and BURNS, R.C. 1968 The ace-
tylene-ethylen assay for N2 fixation : laboratory and field evaluation. Plant P.hysioZ.
43, 911-914.
HEATH, R.L. 1977 Penetration of mannitol into the intracellular space of chlrorella soro-
kiniana. Plant Physiol. 59. 911-914.
HECKMAN, M.O. and DREVON, JJ. 1987 Nitrate metabolism in soybean root nodules.
Physiol. Plant. 69, 721-725.
HECKMAN, M.O. et DREVON, JJ. 1988 Effet du malate sur la @onse de l’activité
nitrogtkase (reduction de J’acétylene) des nodositks de soja à l’oxygene. CR. Acad.
Sci. Paris. 306, 559-564.
HUANG, J.C., CHEN, C.H. and BURRJS, R.H. 1973 Inhibition of nitrogenase-catalysed
reductions. Biophys. Biochim. Acta. 292, 256-270.
KALJA, V.C. et DREVON, JJ. 1985 Variation de l’activité nitrogenase (reduction de
l’acetylène) pendant l’incubation in situ des nodosités de Soja (Glycine max (L) Merr.).
CR. Ad. Sci. Paris 301, 591-596.
MJNCm, RF., INES MINGUEZ, M., SHEEHY, J.E., WJl-TY, J.F. and SKOT, L. 1986
Relationships between nitrate and oxygen in symbiotic nitrogen fixation by white clover.
J. Exp. Bot. 37, 1103-1113.
287
OBATON, M., MIQUEL,M.,ROBIN,P., CONEJERO, G., DOMENACH, A.M. et BARDIN,
R. 1982 Influence du déficit hydrique sur l’activité nitrate r&iuctase et nitrogénase
chez le soja (Glycine mi (L). Menill ; cv. Hodgson). CR. Acad. Sci. Paris. 294,
1007-1012.
PANKHURST, CE. and SPRENT, J-1. 1975 Effets of water stress on the respiratory
and nitrogen-ftig activity of soybean root nodules. J. Exp. Bot. 26, 287-304.
RALSTON, E.J. and IMSANDE, J. 1982 Entry of oxygen nitrogen into intact soybean
nodules. J. Exp. Bot. 33, 208-214.
RIGAUD, J. 1976 Effet du nitrate sur la fixation d’azote par les nodules de haricot (Phuseolus
vulgank L.). Physial. Vég. 14, 297-308.
SALL, K. 1984 Influence d’un déficit hydrique sur les développement végétatif et fruc-
tif& et sur l’alimentation azotee chez trois vari&& de Soja : Kinsoy, Jiangedou et
Quingmoushidou. Diplôme d’études approfondies, INP-Toulouse ; 94 pp.
SCHWEITZER, L.E. and HARPER, J.E. 1980 Effect of light, dark and temperature on
root nodule activity (acetylene reduction) of soybeans (Glycine max L. Mer-r.) Plant
Physiol. 65, 51-66.
SHEEHY. JE., MINCHIN, F.R. and WTITY, JF. 1983 Biological control of the resistance
to oxygen flux in nodules. Ann. Bot. 52, 565-571.
SINCLAIR, TR. and GOUDRIAAN, J. 1981 Physical and morphological constraints on
transport in nodules. Plant Physiol. 67, 143-145.
SLOGER, C., BEZDICED, D., MILBERG, R. and BOQNKERD, N. 1975 Seasonal and
diumal variation in N, (C~H.&fixing activity in field soybeans. Inter. Biol. Prog. 6,
271-284.
SPRENT, J.I. 1976 Nitrogen fixation by legumes subjected to water and light stress.
In : Nutman, P.S. (e&) Symbiotic nitrogen fixation in plants. pp. 405420 Cambridge
Univ. Pre.ss.
STREETER, J.G. 1982 Synthesis and accumulation of nitrite in soybean nodules sup-
plied with nitrate. Plant Physiol. 69, 1429-1434.
WEISZ, R.W., DENISON, R.F. and SINCLAIR, T.R. 1985 Response to drought stress
of nitrogen futation (acetylene reduction rates) by field-grown soybeans. Plant Physiol.
78, 525-530.
WERY, J., TURC, 0. and SALSAC, L. 1986 Relationship between growth, nitrogen frxa-
tion and assimilation in a legume (Medicugo sutivu (L.). Plant and Soil. 96, 17-29.
WITI’Y, JF., MINCHIN, FR. and SHEEHY, JE. 1983 Carbon costs of nitrogenase ac-
tivity in legume mot nodules determined by using acetylene and oxygen. J. Exp. Bot.
34, 951-%3.
Maximiser la FBA pour la Production Agricoie et Forest&e en Afiique
Effect of 16N-labelled mineral-N
and strains of Bradyrhizobium on biological
nitrogen fixation and N-partitioning
in cowpea (Vigna unguiculata) (L.) Walp.
LUYINDULA, N.(l) and WEAVER, R.W.t2)-
(1) Centre Rkgional &Etudes Nucldaires de Kimhasa.
BP. 868. Kinshasa Xl zaiie.
(2) Tear Agricdtmd Experiment Station ~FUI Texav A & M University,
College Station, Texas 77843, USA.
Abstract
‘SN-labelling technique was used to study the effects of different strains of
Bradyrhizobium and two doses of mineral-N on the fate of biologically fixed and
fertiker derived N on the partitioning of N in cowpea. Cowpea plants treated with
40 mg N plant-’ at planting had greater N-content and fixed more N2 than plants
receiving 240 mg N plant? in split applications. Plants treated with the lower dose
of mineral-N contained more N in all parts except the pods and roots as compared
to the. higher mineral N treatment. The high dose of mineral-N inhibited BNF of
cowpea, but N accumulation in the pods was not reduced For this treatment 52-
58% total N fixed and 32-58% of fetilizer-N absorbed were accumulated in the
pods. The nitrogen content of nodules was influenced by the strain of Bradyrhizobirun.
The strain of Bradyrhizobium also influenced the partitioning of N derived Erom
mineral fertilizer among plant parts.
289
Introduction
Cowpea is an important crop for tropical developing countries due to its high pro-
tein content. Scope for improving cowpea yield by applications of nitrogen fertilizer is
limited due to high costs of fertilizer nitrogen, especially in developing countries. Great
attention must, therefore, be paid to inoculation of cowpea with efficient strains of Bru-
dyrhizobium.
Many studies have been done concerning the effects of combined N on the phy-
siology of the Rhizobiumllegume symbiosis and the partitioning of f&r-N in .plants
during reproductive development (Summerfield et al., 1977 ; Eaglesham et al., 1983 ;
Douglas and Weaver, 1986 ; Weaver et al., 1987)
Many authors recommended the use of low doses of N to stimulate nodule growth
and N, fixation (Pate and Dart, 1961 ; Dart and Willon, 1970 ; Subba Rao et al.. 1974 ;
Eaglesham er al., 1983, Weavez et al., 1987). However, the determination of starter-N
level is complex and may depend on many factors, such as the form of the N-compound,
strain of Rhizobim, season, temperature, time of application, plant cultivar, etc...
Some experimental data indicated that applied N at a level lower than 360 mg
N/plant may have synergistic effects on N2 fixation by vigorously growing cowpea and
soybean (Broeshart, 1974, Eaglesham er al., 1983). Morever, Weaver et al. (1987), demons-
trated with glasshouse experiments that the addition of 60 mg plant-’ mineral-N (com-
pared with no addition) to cowpea changed the ranking of strains for effectiveness, in-
fluenced the total N content on shoots and whole plant and N concentration in leaves
and pods.
Partitioning of N during reproductive development of grain legume is largely a
genetic trait of the plant (Minchin et al., 1978) ; some authors found a Rhiwbium strain
effect on N partitioning in reproductive soybean plant parts (Morris and Weaver, 1983 ;
News et al., 1985).
In the previous research, rhizobial strains influenced N partitioning in plants, only
when they differed in effectiveness. The present work utilized four rhizobial strains of
equal effectiveness to determine their influence on N partitioning in cowpea in a long-
term greenhouse experiment under two doses of mineral-N.
Materials and Methods
Plant and bacteria
Cowpea (Vigna Imguiculuta) (L.) Walp. “Bush Purple Hull”) seeds were preger-
minated and planted in plastic pots of expanded vermiculite (Weaver, 1975). Four preger-
minated seeds were planted in each pot and were inoculated at’planting with 1 x lo7
cell I-’ per seed of Bradyrhizobium (strains BDC 22,6A, 24B from Texas A 8r M University
culture collection isolated from Barbados soils and strain 32lXl from Nitragin Co., Inc.).
One week after planting, the seedlings were thinned to one per pot. Plants were
watered with a N-free nutrient solution @vans et al., 1972). every 1 to 2 days and excess
nutrient solution drained from holes at the bottom into containers placed under the pots
(Eaglesham et al., 1983).
Mineral-N application
There were 10 treatments, five (four inoculated and one uninoculated) supplied
with 40 mg N/pot of 15N-labelled KNO, (1.2786 atom % “N excess) and five (four inocu-
lated and one uninoculated) supplied with 240 mg N/pot of lsN-labelled KNO, (0.4783
atom % 15N excess). The low dose was applied as a single application 1 week after planting,
and the higher dose was applied in split application 1.4, 5, 6 and 7 weeks after planting
at rate of 40, 40, 80, 40 and 40 mg N/pot respectively at each time. Also an unino-
culated treatment without N was included as a control to determine the quantity of N
coming from the plant support medium. These uninoculated plants without miner&N dis-
played early N-deficiency symptoms and yielded a maximum of 6.5 mg N/plant. Each
treatment was replicated four times and treatments were arranged in a completely rando-
mized design.
Harvesting and Nitrogen analysis
Plants were harvested at maturity (when 3-4 pods on each plant were completely
purple). Plants were separated into leaves, stems, peduncles, pods, roots and nodules and
oven dried (65-7O’C) for at least 72 h before dry weights were determined. Samples
of leaves, stems, peduncles, pods and roots were ground and N analysed by kjeldhal
method as described by Nelson and Sommers (1973) but was modified to include NO,
nitrogen.
l*N Determination
r5N content was determined by the method described by Hauck (1982) using a
Micromass 602-D mass spectrometer. From the atom % rsN excess in sample and atom
96 15N excess of mineral-N applied, the fraction of N in the plant sample derived from
fertilizer was determined by the formula :
Atom %15N excess in plant part
% N derived from fertilizer
=
X
100
Atom % r5N excess in fertilizer
Results and Discussion
Performance of rhizobial strains was not significantly different within each N
treatment with respect to total N and dry matter contained in various plant parts except
for nodules. The main effects are presented in Table 1. Starter N (40 mg plant -‘) resulted
in more dry matter being partitioned into nodules and less into roots as compared to
the split application of 240 mg N plant -I (Table 1). However, application of only starter
N resulted in more total N being contained in leaves, stems, peduncles and nodules
(Table 1). Whole plant dry matter was not significantly different (p = 0.05) for the two
N treatments (28.5 vs 29.8 g plant-‘) but total N content was higher for the starter N
treatment (624 vs. 596 mg plant “) as indicated by the F test for significance.
291
292
The dry weight of nodules was reduced by the continuous N treatment but within
N treatments rhizobial strains performed similarly (Table 2). Rhizobial strain bad a signi-
ficant influence on the N content of the nodules (Table 2). Nodules induced by strains
22 and 32Hl contained substantially more N than nodules induced by the other two strains.
The effect of strain was less pronounced when more mineral N was provided. The N
content of the nodules was a significant pcxtion of the whole plant N and for the starter
N treatment, the nodules contained more N than the roots (Table 1). The level of N
provided in the contirmous N treatment not only reduced nodulation (Tables 1 and 2)
but dinitrogen fixation (Table 3) even though the percentage of total N in the plant, derived
from the fertilizer was only 15%. Eaglesham et a2. (1983) reported a sigmficant decrease
in N, fxation when plant uptake of NO, provided more than 20% of the total N in the
plant The quantity of biologically fmed N contained in plant parts other than roots was
reduced by the continuous N treatment (Table 3) but the quantity of fertilizer N in all
plants was greatly increased.
Table 2 : Total dry weight and nitrogen content of nodules on cowpea as
influenced by strain of rhizobia and mineral nitrogen.
Dry weight
Total N
Strain
----___---__-_--_--________
------------_-------___________
Starter
Continuous
Starter
Continuous
----------g plant -1 ------__-
- -----__--_ mg plant -1 _-___-__-_
22
0.490
0.347
35.3
25.9
248
0.484
0.325
24.7
16.8
32Hl
0.535
0.340
40.2
24.7
6A
0.445
3
0.320
20.4
15.9
293
Table 3 : The influence of starter N and continuous N on partitioning
of biologically fixed N (BNF) and fertilizer N into various
plant parts.
BNF
Fertilizer N
Plant
part
Starter
Continuous
Starter
Continuous
- - -------______ -__-__-__ mg N plant-1 -____________--_-__-______
Leaf
154.0 (12.8)+
113.0
2.63
17.5
Stem
44.2 (5.27)
31.0
0.62
5.31
'Peduncle
39.2 (4.75)
23.7
0.32
4.55
Pod
326.0 (22.9)
291 .o
2.45
50.1
Root
23.5 (3.77)
23.0
1.15
6.32
Nodule
30.2 (5.83)
20.8
0.36
0.63
Total
617.1 (27.7)
502.5
7.53
84.41
+ The number in parenthesis is the LSD 0.05 for making comparisons on
the effect of mineral N treatments on the quantity of biologically
fixed N in the particular plant part. A test of significance was not
made for fertilizer N comparisons.
294
Fertilizer N treatments influenced the relative partitioning of fertilizer N into various
plant parts (Table 4). Distribution of starter N was largely into vegetative organs probably
because it was depleted from the plant growth medium relatively soon whereas for the
other N treatment the split applications made the N available throughout the growth period
and half of the featihzer N was contained in the pods (Table 4). The largest percentage
of the biologically fmed N was contained in the pods regardless of the fertilizer N treatment
In the case of the split application N treatment, the same proportion of the fertihzer N
and biologically fixed N in tbe whole plant partitioned into the pods. Eaglesham et al.
(1977) reported similar results for cowpea continuously provided with NO,. It appears
that the plant did not have a preference for the source of N for pods when both sources
of N were available throughout plant development. However, the data in Table 4 are
averages across four strains of rhizobia. The nodules only received 3% of their N from
mineral N in contrast to the roots which received 24% (calculated from Table 3). Eaglesham
et al. (197’7) also reported that nodules on plants fertilized with NO,- received 96% of
tbeii N from dinitrogen but roots received only 86% of their N from dinitrogen and concluded
that nodules were largely self-sufficient. Nineteen days after plants were pulsed with r5N
labelled NO,- the roots contained 19% of the NO, that was taken up and the nodules
only contained 2% (Douglas and Weaver, 1986).
Table 4 : Distribution of biologically fixed N (BNF) and fertilizer N
(FN) in plant parts of cowpea as a percentage of the total
plant N derived from each source.
Plant part
Leaves
Stems
Peduncles
Pods
Roots
Nodules
Nitrogen.
-
-
treatment
BNF
fN
BNF FN
BNF FN
BNF FN
BNF
FN BNF FN
-
-
-
-
-
-
-
-
m
m
-
-
-
Starter
25.1
35.2
7.3
8.3
6.3
4.3
5 3
3 2
3.8
1 5
4.8
4.9
Continuous
22.7
21.2
6.2
6.3
4.7
5.5
5 8
5 8
4.6
7.6
4.0
0.4
LSD.O5
2.3
'3.5
0.85
1.6
0.76
0.87
3.0
5.0
0.66 1.7 0.60 0.71
295
The strain of rhizobia did not influence uptake or distribution of starter N in the
plant (data not shown) but rhizobial strain influenced distribution of N from the continuous
N treatment (Table 5). The actual quantity of fertilizer N contained in various plant parts
was only signifkantly different for leaves, stems, and nodules. The coefficient of variation
was quite large for the reproductive plant parts and although large differences occurred
for pods they were not statist&ally significant. However, the data provide quite strong
evidence that the strain of rhizobia influenced the relative distribution of fertilizer N into
the pods (Table 5). For two of the strains approximately 52% of the fertilizer N was
Table 5 : The influence of rhizobial strain on the quantity and percentage
distribution of continuous N into plant parts of cowpea
Plant part
Strain
Leaves
Stems
Peduncles
Pods
Roots
Nodules
Total
-
-
-
-
-
-
---------------------------mg N plant-1 _____________--__--_------
6A
15.7
4.98
4.19
51.4 4.76 0.41 83.4
2 2
18.5
3.83
3.79
64.6 6.20 0.47 97.4
246
13.8
5.00
4.51
36.0 6.87 0.61 66.8
32Hl
22.1
7.42
5.72
48.3 7.46 1.04 92.0
LSD 0.05
4.4*
T.86*
2.21
23.9 2.96
0.24* 30.8
--_-___-____--__-__L--- ----i distribution------------------------
6A
19.7
6.32
5.15
'
62.5.. 5.85 0.53
2 2
20.1
3.90
3.95
65.2 6.45 0.48
-24B
20.8
7.50
6.62
53.7 9.88 0.93
32Hl
24.4
8.10
6.13
52.1 8.10 1.13
LSD 0.05
5.60
1.94*
1.76*
9.5*
2.18*
0.28*
* indicates significant difference by F-test (P = 0.05) between strains.
296
partitioned into the pods but for the other two strains approximately 63% of fertilizer
N was partitioned into the pods. These results indicate that the strain of rhizobia influenced
partitioning of fertihzr N in cowpea as has been reported for pigeon pea (Hernandez
et al., 1987). However, the data do not confii that cowpea strains influence uptake
of mineral N as report& by El Hassan and Focht (1986) but different strains of rhizobia
were utilized in our experiments and El Hassan and Focht (1986) grew their plants in
soil rather than a defined medium.
Evidence is accumulating that the strain of rhizobia may have an effect on nitrogen
partitioning in grain legumes. Our results indicated that four strains of rhizobia of equal
effectiveness based on total dinitrogen accumulated produced similar pod yields but the
relative distribution of fertibxer N partitioned into pods was different (fable 5). The strains
also influence the quantity of N contained in nodules frable 2). Depending on the mineral
N treatment the quantity of N in nodules was more or less than that contained in roots
(Table 1) which accounted for approximately 5% of the total plant N. Averaged across
strains of rhizobii the plant partitioned N from dinitrogen and continuously supplied mineral
N into pods in the same relative proportions (Table 4). This agrees with the data of
Eaglesham er ul (1977) in which 66% of the N from each N source was partitioned
into cowpea fruit. Thus, there was no evidence that the plant preferentially milked biologi-
cally fixed N over mineral N for pod development or that the rhizobial strain influen-
ced partitioning of biologically fixed N into pods.
Acknowledgements
This work was supported by funds from Regional Project S-170, the International
Atomic Energy Agency and US AID grant 84-CRSR-2-2521. The technical assistance of
Mrs. Heidi J. Mjelde is gratefully acknowledged.
References
BROESHART, H. 1974 Quantitative measurement of fertilizer uptake by crops. Netherland
J. Agric. Science. 22, 245.
DART, PJ. and WILDGW, D.C. 1970 Nodulation and nitrogen fixation by Vignu sineti
and Viciu atropurpuneu. The inlluence of concentration, form and site of application
of combined nitrogen- Australian J. Agric. Research. 21. 45%.
DOUGLAS, LA. and WEAVER, R.W. 1986Partitioning of nitrogen-15 labelled biologically
fvred and nitrogen-15 labelled nitrate in cowpea during pod development. Agronmy
J. 78, 499-502.
EAGLESHAM, ARJ., MINCHIN, F.R., SUMMERFIELD, RJ., DART, PJ., HULEY,
P.A. and DAY, JM. 1977 Nitrogen nutrition of cowpea Vigna unguicuha III. Distri-
bution of nitrogen within effectively nodulated plant Exp. Agric. 13, 369-380.
EAGLESHAM, ARJ., HASSOUNA, S. and SEEGERS, R. 1983 Fertilizer-N effect on
N, fixation by cowpea and soybean. Agronomy J. 75, 61-66.
EL HASSAN, G.A. and FGCHT, D.D. 1980 Comparison of inoculant and indigenous
rhizobial dinitrogen fmtion in Cowpeas by direct nitrogen-15 analysis. Soif Sci. Sm.
Am. J. 50, 923-927.
297
EVANS, HJ., KOCH, B., and KLUCAS, R. 1972 Preparation of nitrogenase from nodules
and separation into components. Methock Enzymology. 24, 470-476.
HAUCK, R.D. 1982 Nitrogen-isotope-radio-analysis. In : Page, A.L., Miller, R.H. and
Kenny, D.R. eds. Methods of Soil Analysis. Part 2. Ed ZAmericau Society of Agro-
nomy Soil Science Society of America Journal Madison. pp. 735-779.
HERNANDEZ, B.S., POTH. M. and FOCHT, D.D. 1987 Increased effectiveness of com-
petitive Rhizobium strains upon inoculation of Cajanus cajan. Appl. Environ. Microbial.
53, 2066-2068.
MINCHIN, FR., SUMMERFIE LD, R.J. and EAGLESHAM, ARJ. 1978 Plant genotype
Rhizobium strains interaction in cowpea Vigna unguiculatu (L.) Walp. Trop. Agric.
55, 107-11s.
MORRIS, DR and WEAVER, R.W. 1983 Mobilisation of 15N from Soybean leaves as
influenced by rhizobial strains. Crop Science. 23, 1111-l 114.
NELSON, D.W. and SOMMERS, LE. 1980 Total nitrogen analysis for soil and plant
tissues. Association Analytical Chem. 63, 770-778.
NEVES, M.C.P., DlDONET, A.D., DUQUE, F.F. and DOBEREINER, J. 1985 Rhizo-
bium strain effects on nitrogen transport and distribution on soybeans. Exp. Bot. 36,
1179-l 192.
PATE, J.S. and DART, PJ. 1981 Nodulation studies in legumes. IV. The influence of
inoculum strain and time of application of ammonioum nitrate on symbiotic response.
Plant and Soil. 15, 329-346.
SUBBA RAO, PAHWA, MR. and LAKESHMI KUMARI, M. 1974 Effects of combined
nitrogen on legume root nodulation. Acta Botunica Zndica. 2, 54.
SUMMERFIELD, R J., DART, PJ., HUXLEY, PA., EAGLESHAM, ARJ., MICHIN,
F.R. and DAY, J-M. 1977 Nitrogen nutrition of cowpea Vigna unguiculata L. effects
of applied nitrogen and symbiotic nitrogen fixation on growth and seed yield Exp.
Agric. 13, 129-142.
WEAVER, R.W. 1975 Growing plants for Rhizobium effectivenes tests. Soil Bid. Biochem.
7, 77-78.
WEAVER, R.W., DIPANKAR SEN, KYAN MOE and GRAHAM, RA. 1987 Effective-
ness of Bdyrhizobium sp. on cowpea provided with starter nitrogen. XI North Arne-
rican Rhiwbium Conference, August 9-15, Quebec, Canada.
Maximiser la FBA pour la Producrion Agricole et Forestibe en Afiique
Quantification of the contributkn of BNF
to field grown plants : the use of the 16N isotope
dilution technique - problems and some solutions
BODDEY, R M. and URQUIAGA, S.
EMBRAPA-UAPNPBS. Km 47. Seropkdica, 23851, Rio & Janeiro, Brazil
Summary
When the ‘?N isotope dilution technique is used to estimate the contribution of
biological N2 fixation (BNF) to plants, the assumption is made that both the “N2-
fixing” (test) and control crops remove N from the soil with the same “N enrichment.
The most practical method of labelling the soil with “N is to apply soluble labelled
featilkr but this @ces in the soil mineral N a rapid fall in lSN enrichment with
time. Hence plants with different uptake patterns can acquire soil-N with different
‘-9-l enrichments. While it is xzcornmended that control crops with the same uptake
patteznsasthetestcropsareused,thisrequirementisexcebdingly~~tttot
for, and in each study uptake patterns will vary with different edaphic and environ-
mental cm&ions. In these circumstances the use of several control crops in conjunc-
tion with slow release forms of lSN can be a successful strategy for the reliable quantifi-
cation of BNF. In a soil vjith a stable ‘?I label any control crop may be used or
even dispen& with altogether. These problems and possible solutions are discu-
ssed with examples of the quantification of BNF in both grain legumes and
grasses.
299
Introduction
In recent years the 15N isotope dilution technique has been widely used for the
quantification of the contribution plant-associated biological nitrogen fixation (BNF) to
both nodulated legumes and other crops. The acetylene reduction technique (Hardy et
al., 1968) has made a gmat contribution to knowledge in BNF research, but its use is
severely limited for the definitive quantification of contributions of N, fixation to higher
plants. The limitations of the application of this technique to the quantification of BNF
associated with nodulated legumes have been thoroughIy discussed in a very recent paper
by Witty and Minchin (1988), and the problems in its use to quantify possible N, fixation
associated with grasses and cereals have also been reviewed recently (Boddey, 1987).
While techniques using total N measurements can be very ‘valuable and should not be
discounted (Weber, 1966 ; Bell and Nutman, 1971 ; Nelson and Weaver, 1980 ; App
et al.. 1980, 1986 ; Lima et al., 1987), they are often of limited vahre in field expe-
riments unless very long term (Jenkinson, 1977, 1982).
It may be thought that the cost of using the stable isotope ‘5N is very high, but
a few hundred dollars (US) can buy sufficient of the Iabelled salts to perform one or
two field experiments. While the instruments to analyse ‘5N enrichment are very expensive
(a mass spectrometer is likely to cost upwards from US$ IOO,OOO, the less sensitive em-
mission spectrometer approximately 30 to 50% of this), commercial b&oratories in the
US and elsewhere will perform analyses for $10 to $20 per sample or even less if many
hundreds of analyses are required. However, if it is essential for the aims of the project
in question that the contribution of BNF to a crop be quantified, then the isotope dilution
technique can provide the answer if used with great care.
The principle of the 15N isotope dilution technique
The 15N technique requires that “N, fixing” crop under study (the test crop) is grown
in a soil enriched with ‘SN along with a control (“non-N,-fixing”) crop. The ‘5N enrichment
of the plant material of the control crop is regarded as representing the enrichment of
the plant-available (mineral) nitrogen in the soil. Hence if the test crop has a lower enrichment
than this it is concluded that unlabelled nitrogen from the air was incorporated into the
tissue of the test crop via N, fixation. It is apparent that the underlying assumption in
the use of this technique is that both the control and test crops remove nitrogen from
the soil with the same 15N enrichment.
Assuming this to be the case the calculation of the proportion of nitrogen in the
test crop derived from BNF (%NF) can be calculated from the equation :
Atom % 15N excess in test crop
. ..Eqn 1.
.’ Atom % *5N excess in control crop
>
It may be pointed out that the technique quantifies the contribution of N, Iixa-
tion specilically associated with the test crop, and that any input of unlabelled N derived
from rainfall, pollution or from BNF associated with cyanobacteria or other sources, if
equally available to both test and control crops, will not be detected by this technique.
For the acetylene reduction or total N balance techniques this may not be true.
300
Enrichment of the soil with 15N
It is inevitable that the test and control crops will remove N from the soil with
& me 15N enrichment if the soil is labelled uniformly in area and depth, and is stable
with t&e. III this case any crop which certainly has no associated BNF will we as
a valid control for any test crop. This approach has been used by several authors to
quantify BNF associated with both legumes and grasses (Kohl and Shearer, 1981; Broadbent
er al., 1982 ; Rennie et al., 1982 ; Butler and Ladd, 1985 ; Miranda and Boddey, 1987).
h fact the control crop in this case may be dispensed with altogether and the 15N enri-
chment of the N derived from the soil determined directly from measurements on the
soil mineral N, (Chalk ef al., 1983).
The most practical and usual method for enrichment -of a soil with 15N is by the
application of a solution of labelled nitrogen fertil&r (e.g. urea, ammonium sulphate
or potassium nitrate) to the surface of the soil. If the fertiKzer is uniformly distributed
over the area of soil to be planted then lateral variations in 15N enrichment should not
occur. Injection of labelled fert&er at points within the planted area such as that per-
formed by Rennie and Thomas (1987) are not to be recommanded as differences in lateral
root distribution between test and control crops may cause differences in the 15N en-
richment of the N derived from the soil.
The application of soluble fertilizer to the surf& of the soil does result in consi-
derable gradients with depth @add et al., 1981 ; Ledgard et al., 198.5). Boddey and Vic-
toria (1986) found that soil amended with repeated (small) applications of labelled am-
monium sulphate showed mean enrichments of mineralizable nitrogen of 0.140. 0.069,
0,048 and 0.030 atom % 15N excess in four successive 11 cm layers. They also found
evidence to suggest that this caused plants with different vertical root distributions to
obtain different “N enrichments in the soil-derived N.
However, the most important problem associated with the use of soluble labelled
fertiGzers is that immediately after application there is a sharp decline in lsN enrichment
in the mineral N in the soil, due to continuing mineralisation of unlabelled soil organic
N into a mineral pool continuously depleted by plant uptake, leaching and other losses
(Fried et al., 1983 ; Witty, 1983). This is illustrated by recent data from a study by
our group performed near Brasialia (Boddey, R.M., Urquiaga, S., Suhet, A., Peres, J.R.
and Neves M.C.P. in preparation). In one treatment of this experiment, inoculated
soybean was grown with four different control crops (a non-nodulating genotype of soy-
bean, sorghum, sunflower and rice) in soil amended with 10 kg N ha-l of lsN-labelled
potassium nitrate (10.0 atom 96 15N excess) at planting. The data show a continuous
decrease in “N enrichment in the plant aerial tissue due to the fall in the enrichment
of soil mineral N (Fig. I).
The reason for the initial incry in lsN enrichment of the soybean treatments
is that at the early harvests the unlabelled nitrogen derived fi-om the large N-rich soybean
seeds constitutes a large proportion of the N in the young plants. This contribution of
unlabelled N decreases as it becomes diluted with labelled N from the soil and the 15N
enrichment of the plants then increases to reflect that accumulated from the soil. This
effect is not significant for the sorghum and sunflower as the seeds are smaller and of
lower N content. It should be noted that if a small seeded cereal plant (e.g. wheat, barley,
sorghum etc) is used as a control for soybean or other large-seeded legumes, then at
early harvests the contribution of unlabelled seed N may be confused with that derived
from BNF.
I-
I-
I-
i--
0
ID
80
m
DAySAFIERGERhB(ATIoN
Figure 1 : lSN enrichment of nodulated soybean (
a ), non-nodulating soybean ( .+ ),
sorghum ( Q ) and sunflower ( d ) grown in a soil amended at planting with soluble lsN Iabelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % “N).
302
As the patterns of soil nitrogen uptake in this study were different (Fig. 2) this
resulted in the different control plants accumulating nitrogen from the soil with different
15N enrichments. This problem has been discussed by many authors and is particularly
well illustmted by the data reported by Witty (1983).
Most isotope dilution studies have been performed using single additions of soluble
labelled fettilizer in conjunction with the use of a single control crop (see reviews of
Chalk, 1985 and Danso, 1988). As the data displayed above, and that of Witty, illustrate,
in such studies it is dangerous to assume that different crops will accumulate N from
the soil with the same “N enrichment, and in many of these studies the different “N
enrichment observed in the nod&ted legume may have been partly due to differences
in N uptake patterns introducing an error of unknown magnitude to the resulting BNF
estimates.
It has been suggested that certain control crops are appropriate for specific legume
crops for use with this technique to quantify BNF (Rermie 1986). There is little evidence
from the sources cited by Rennie that these control crops were necessarily “appropriate”
(i.e. accumulated N with the same 15N enrichment as that derived from the soil by the
legume). Furthermore, as edaphic and environmental conditions will cause differential
changes in soil N uptake patterns, even if a control proved to be “appropriate” in one
experiment there is no guarantee that this would be true for other experiments on the
same legume at different sites, seasons or years.
Slow release forms of labelled nitrogen.
As Witty (1983) showed, if the decrease in r5N enrichment of soil mineral N is
slowed down by applying a slow release form labelled N fertilizer then the differences
in enrichment between different controls are attenuated. In our recent study near Bra-
silia a further treatment was included in the experiment where the same quantity of ‘5N
labelled potassium nitrate (10 kg N ha-r) was mixed with sugar to give a CzN ratio of
1O:l and added 7 days before planting. The patterns of uptake of soil N were not altered
significantly by the addition of sugar to immobilize the mineral nitrogen (Fig. 3) but
the rate of temporal decline in 15N enrichment was attenuated (Fig. 4) resulting in a narrower
range of enrichments of the four different control crops (Table 1).
Many different techniques have been utilized to slow down the decline in the 15N
enrichment of soil mineral N with time. Some workers have mixed various carbon substrates
(sucrose, glucose, cellulose, straw etc.) with labelled N fertilizer to immobilize the mineral
N (Legg and Sloger, 1975 ; Broadbent et al.,1982 ; Wagner and Zapata, 1982 ; Talbott
et al.,1982 ; Giller and Witty, 1987). Another similar technique adopted by some has
been to incorporate *sN-labelled plant material (especially prepared or resulting from other
studies) into the soil (Henzell et al.,1968 ; Boddey et al.,1984).
In a recent study on 10 cultivars of sugar cane, lsN-labelled ammonium sulphate
and potassium nitrate were composted with bagasse and filter cake from a sugar cane
mill to form a labelled organic matter which was mixed with soil to fill a concrete tank
(20 x 6 m) to a depth of 60 cm (Urquiaga et al.,1989). There was still a dramatic fall
in r5N enrichment of the soil mineral N with time (Fig. 5) although it should be emphasized
that the measurements were made over a period of almost 2 years. Data for the r5N enrichment
of the sugar cane cultivars and the control crop (Bruchiatiu rdcanr) at the first complete
harvest (250 days after emergence) suggested that there were considerable BNF contri-
303
0
1D
4l
60
80
m
DAYSAFllZCZ!RMNAllON
Figure 2 : Total N accumulation of nodulated soybean ( n ). non-nodulating soybean ( f ), sor-
ghum (Q ) and sunflower ( h ) grown in a soil amended at planting with soluble lSN labelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % 15N).
304
Figure 3 : Total N accumulation of nodulated soybean ( D ), non-nodulating soybean ( j- ), sor-
ghum (Q ) and sunflower (A ) grown in a soil amended 7 days before planting with 15N labelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % “N) mixed with sugar to give a C:N ratio of 10: 1.
3 0 5
22 T-
2
la
16
I4
12
1
a8
llfi
a4
02
0
Figure 4 : “N enrichment of nodulated soybean ( m ). non-nodulating soybean ( + ). sor-
ghum ( 0) and sunflower (h ) grown in a soil amended 7 days before planting with ‘W labelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % ‘?J) mixed with sugar to give a C:N ratio of 10:l.
306
Table 1. 15N enrichment of nodulated soybean and 4 control crops grown
in soil amended with 15N-labelled fertilizer, with and without
addition of sucrose. Means of 4 replicates.
ATOM % 15N EXCESS
CROPS
SOLUBLE FERTILIZER
FERTILIZER + SUGAR
Non-Nod Soybean
0.4039
0.3538
Sorghum
0.4367
0.3679
Sunflower
0.6209
0.4522
Rice
0.5044
0.4355
Nodulated soybean
0.3131
0.3208
Honest significant
Difference (P=O.OSl*
0.2083
0.1925
* Tukey test, Coefficient of variation 20.5%
L3
Figure S : ‘SN enrichment of soil mineral nitrogen extracted with 2M KC1 from Itaguai series soil
L2-
mixed with “N labelled compost. Data from Urguiaga et al (1988) and Urguiaga S., Cruz K.H.S.
and Boddey R.M. (in preparationj.
11 -
I-
0.9-
0.8 -
ti
CA7 -
?.i82
0.6-
8
E*
05 -
<
0.4 -
* 1.
03 -
0.2 -
0.1 -
u
El
0 (
I
I
I
I
I
1
I
1
I
0
loo
300
400
!3nJ
DAYS AFTER APPLICATlON
308
butions to several of the sugar cane cuhivars, notably Krakatau and SP 70-1143, but
because of the large fall in the 15N enrichment in the soil mineral N during this period
the estimates could not be regarded as accurate (Table 2) However, as subsequent to
this harvest there was very litde further drop in the enrichment of the soil mineral N,
the estimates of the BNF contribution to the ratoon crop can be regarded as being much
more reliable.
From the above disr;ussion it is apparent that in order to accurately quantify plant
associated BNF using the “N isotope dilution technique it is desirable that slow release
forms of labelled nitrogen are used. Nearly all of the slow release forms of N when
incorporated into the soil immobilize soil mineral N and hence change soil N availability
to the plants, and/or cause other changes in soil fertility. It is important therefore to examine
the objectives of the study in question to decide whether this will negate them. In our
work performed on grasses and sugar cane the interest has been in determining under
optimum conditions whether these plants are able to obtain significant contributions from
plant associated BNF and in the comparisons of cult&us with regard to this trait. In
this case the immobilization of nitrogen by the addition of slow release forms of labelled
N does not adversely affect the objectives of the experiment This may be true for many
studies with legumes, but there will be circumstances where it is necezzuy that the soil
N availability is not significantly affected by application of the isotope dilution technique.
One form of slow release labelled N which would probably only marginally affect
soil N availability is that based on the gypsum pellets described by Witty (1983). However,
more recent data suggests that release of N from such pellets is not much slower than
soluble N (Giller and Witty, 1987 ; Boddey et al.,1989). One other solution that may
be possible is the addition of small quantities of soluble labelled N at regular (e.g. weekly)
intervals. This thechnique has been adopted by many authors (Vallis et aI.,1%7 ; Goh
et al, 1978 ; Boddey et al., 1983, 1984 ; Ledgard et al., 1985) but as Rennie (1985)
has pointed out this technique has its dangers if sequential harvest of plants are to be
made and soil N availability is high. If in this case the test and control crops remove
different quantities of N from the soil then the residual mineral N in the soil at the sub-
sequent harvest will be different. Further addition of labelled N to the soil will then re-
sult in a different initial enrichment of mineral nitrogen for the two crops at the start
of the next growth period. Experience to date suggests that the use of small frequent
additions of labelled fertilizer may be the best technique of labelling the soil if soil N
availability is not to be radically altered (Boddey et af.,1983. 1984 ; Ledgard er al., 1985b)
but further, more detailed, studies on this subject are required.
A strategy for the field application of the isotope dilution technique
Unless the 15N enrichment of the soil mineral N is known to be reasonably stable
with time it is advisable that more than one control crop should be used within the same
experiment. This strategy was recently recommended following experiments performed
by our group at two different sites in Brazil (Boddey et u1.,1989). In both experiments
the objective was to quantify BNF contributions to inoculated and non-inoculated soybean,
cowpea and groundnut using three different control crops (a non-nodulating genotype of
soybean, sorghum and rice).
At the fiit site (Rio de Janeiro) the soil was enriched with 15N by addition of
labelled gypsum pellets (Witty, 1983). The data showed that there were considerable and
Table 2 : Estimates of the contribution of biological nitrogen fixation (BNF) to 10 sugar cane varieties
1st HARVEST 250 DAET
2nd HARVEST 380 DAE1
Varietv/
species
Total N
15N
Estlmate of the
Total N
T5N
Estimate of the
accumulation
enrichment
contribution. of BNF2
Accumulation
Enrichment
contribution of
(kg.ha-1)
(ATM % 15N ext.)
%
(kg.ha-1)
(kg.ha-1)
(ATM % 15N ext.) BNF2
%
(kg N.ha-1)
CB 47-89
265 ab
0.316 bc
41.9
111
209 bc
0.108 bc
19
41
CB 45-3
257 ab
0.305 bed
44.1
110
258 b
0.120 ab
10
27
NA 56-79
246 ab
0.313 bc
42.9
106
188 bed
0.126 a b
6
11
IAC 52-150
271 ab
0.300 bc
46.2
125
173 bed
0.115 abc
14
24
SP 70-1143
245 ab
0.261 bc
52.3
128
216 b
0.109 bc
19
40
SP 71-799
244 ab
0.298 bc
46.1
112
200 bc
0.109 bc
19
37
SP 79-2312
202 abc
0.344 bc
41.7
84
277 b
0.127 ab
5
14
Chunnee
152 c
0.349 b
37.2
56
119 bed
0,124 a b
7
9
Caiana
49 d
0.309 bc
42.2
21
26 d
0.123 ab
8
2
Krakatau
294 a
0.255 c
55.6
163
452 a
0.097 c
28
125
B. radicans
189 bc
0.546 a
43 cd
0.134 a
Coefficient
of variation (%I
27.0
11.4
35.6
7.6
'Days after emergence
2Values calculated using the isotope dilution technique using B. radicans as control.
310
significant (P < 0.05) differences between the 15N enrichments of the three different control
crops but all were far higher in i5N enrichment than any of the nodulated legumes which
indicated that the legumes were all obtaining considerable contributions from BNF
(Table 3). For each nodulated legume, three estimates of the BNF contribution could
be made both by using the total N difference method (total N in nodulated legume -
total N in control crop) and the isotope dilution technique and these estimates are dis-
played in Fig 6. It is apparent that the three different estimates of BNF for each nodulated
legume calculated from the total N difference are more diverse than the three estimates
derived from the isotope dilution technique. Furthermore, the estimates for cowpea are
negative and hence obviously erroneous. This was due to the fact that the cowpea grew
very poorly due to fungal disease and accumulated very little N from the soil. These
data show that, at least in this case, the isotope dilution technique yielded far better results
than the total difference N technique. However, the date show that despite the considerable
differences in 15N enrichment between the different control crops the estimates of BNF
contributions fall within a fairly narrow range. The probability that the true estimate of
the BNF contribution lies within this range increases with increasing numbers and diversity
of control crops.
At the second site (Brasilia) 15N-labelled compost (Urquiaga et al.,1989) was used
to label the soil. At this site there were smaller differences between the enrichments of
the different control crops suggesting that this form of slow release N was superior to
the gypsum pellets. This resulted in relatively small differences between the isotope dilution
estimates of BNF contributions (Fig. 7) and leads us to recommend this technique of
combining slow release fertilizer (or perhaps repeated small additions of soluble fertilizer)
with the use of several (3 or more) control crops. The advantage of using several control
crops of a diverse nature is that it gives the experimenter an idea of the range of possible
15N enrichments that crops of different growth habits are able to obtain from a soil with
.
a non-uniform 15N enrichment.
Aknowledgements
Financial support for many of our recent studies on the quantification of biological nitrogen
fixation has been provided by the Financiadora de Rstudos e Rrojetos (FINEP) of Brazil,
and by the U.S. National Academy of Sciences, National Research Council through a
grant from the U.S. Agency for International Development.
Table 3 : final grain yield, and dry matter nitrogen and 15N9)recovery at mid-pod fill stage of 3 cqntrol crops and
inoculated and non-inoculated soybeans, cowpea and groundnuts. Experiment 1, Riz de Janeiro. Means of
4 reolicates.
Dry matter
15N recovery
Grain(3)
Treatment
accumulation(2)
Total N(2)
% 15N1
g.m-2
mg 15N excess
yield
g.m-2
excess
m-2
kg ha-l
Soybean inoc'd(4)
29 w
621
20.6a.(5)
2.56bY5)
Soybean inoc'd(4)
638
20.3a
0.0104e
2.15bc
Soybean non ir#oc'd(4)
589
18.8ab
0.0125e
2.32b
Cowpea inoc'd
202
3.6de
0.0289cd
1.02cd
304
wz
Cowpea non-inoc'd
144
2.5e
0.0315cd
0.78d
334
Groundnut inoc'd
669
16.3ab
0.0143e
2.14bc
1710
Groundnut non-inoc'd
593
14.7b
0.0219de
3.29b
2322
Sorghum
1142
9.7c
0.0574a
5.08a
Soybean, non-nodu-
lating
408
7.7cd
0.0398bc
3.04b
Rice
443
5.2de
0.0484ab
2.67b
Coefficient of
Variation %
18.3
21.4
35.5
32.2
1. 15N added as slow release gypsum pellets (Witty, 1983) at 2.15 kg labelled N/ha
2. Mid-pod fill stage, 110 days after planting, Harvested area = 1 m2/plot.
3. Nature grain stage, 117 and 141 days after planting for cowpea and groundnut, respectively. Harvested area 2 mZ/plot.
4. Grain yield lost due to weevil attack.
5. Means followed by the same letter are not significantly different at P = 0.05 (Duncan).
Figure 6 : Total N difference ( El ) and isotope dilution ( n ) estimates of the contribution of N,
fixation to inoculated and non-inoculated soybean, cowpea and groundnut using 3 different control
crops ; sorghum (S), non-nodulating soybean (Y) and rice (R). Experiment 1, Rio de Janeiro. Means of
4 replicates. Data from Boddey et al (1989).
S O Y B E A N
C O W P E A
G R O U N D N U T
Inoculated
Inoculated
Non
Inoculated
Non
Inoculated
Non
2 9 w
C.8 1 8 0 9
Inoculated
BR 53
inoculated
I I A
inoculated ’
i
I I
I
I
I
I
I
I
11:
S Y R
SYR S Y R
S Y R S Y R O’?
In
313
z-uI
Contrlbutlon of
’
N6
0
UOllOX()
N,
flxatlon
‘N
40
Lo
UOl+l’lq(l(UO3
aN
.
rn-*
314
References
APP, A.A., WATANABE, I., ALEXANDER, M., VENTURA, W., DAES, C.,
SANTIAGO, T. and D DATTA, SK. 1980 Nonsymbiotic nitrogen fixation associa-
ted with the rice plant in flooded soils. Soil Science. 130, 283-289.
APP, A.A., WATANABE, l., VENTURA, T.S., BRAVO, M. and JUREY, C.D. 1986
The effect of cultivated and wild rice varieties on the nitrogen balance of flooded
soil. Soil Science. 141, 448-452.
BELL,F., and NUTMAN, P.S. 1971 Experiments on nitrogen fixation by nod&ted luceme.
Plant and Soil. Special Volume, 231-264.
BODDEY, RM., CHALK, PM., VICTORIA, R-L., MATSUI, E. and DOBEREINER.
J. 1983 The use of the 15N isotope dilution technique to estimate the conhibution
of associated biological nitrogen fixation to the nitrogen nutrition of Paspahm po-
tatum cv. batatais. Can. J. Microbial. 29, 1036-1045.
BODDEY, R.M., CHALK, P.M., VICTORIA, RL. and MATSUI, E. 1984 Nitrogen fma-
tion by nodulated soybean under tropical field conditions estimated by the ‘W iso-
tope dilution technique. Soil Biol. Biochem. 16, 583-588.
BODDEY, R.M and VICTORIA, R.L. 1986 Estimation of biological nitrogen fixation
associated with Bra&aria and Paspalum grasses using ‘SN labelled organic matter
and fertiIizer. Plant and Soil. 90, 265-292.
BODDEY, R.M. 1987 Methods for quantification of nitrogen fixation associated with
gramineae. CRC Critical Reviews in Plant Sciences. 6. 209-266.
BODDEY, R.M., URQUIAGA, S., SUHET, AR., PERES, JR. and NEVES, M.C.P. 1989
Quantification of the contribution of N2 fixation to field-grown grain legumes : A
strategy for the practical application of the ‘% isotope dilution technique. Soil Bid.
Biochem. (Submitted)
BROADBENT, F.E., NAKASHIMA, T. and CHANG, G.Y. 1982 Estimation of nitrogen
fixation by isotope dilution in field and greenhouse experiments. Agronomy J. 74,
625-628.
BUT&R, J.H.A. and LADD, JN. 1985 Symbiotically-fixed and soil-derived nitrogen
in legumes grown in pots in soils with different amounts of available nitrate. Soil
Biol. Biochem. 17, 47-55.
CHALK, P.M. DOUGLAS, L.A. and BUCHANAN, S.A. 1983 Use of 15N enrichment
of soil minezaIizable N as a reference for &ope dilution measurements of biolo-
gically fixed nitrogen. Can. J. Microbial. 29. 1046-1052.
CHALK, P.M. 1985 Estimation of N2 fixation by isotope dilution : an appraisal of tech-
niques involving lsN enrichment and their application. Soil Biol. Biochem. 17, 389-
410.
DANSO, S.K.A. 1988 The use of lsN enriched fertilizers for estimating nitrogen fixation
in grain and pasture legumes. In : Nitrogen fixation by legumes in Mediterranean
agriculture. Beck,D.P. andMateron,L.A.Eds. pp, 345-358. MartinusNijhoff, Dordrecht,
Netherlands.
315
FRIED, M.. DANSO, S.K.A. and ZAPATA, F. 1983 The methodology of measurement
of N, fixation by non-legumes as inferred from field experiments with legumes. Can.
J. Microbial. 29, 1053-1062.
GILLER, K.E. and WITTY, J.F. 1987 Immobilized 15N-fertilizer sources improve the
accuracy of field estimates of N,-fixation by isotope dilution. Soil Biol. Biochem.
19, 459463.
GOH, K.M., EDMEADES, D.C. and ROBINSON, B.W. 1978 Field measurements of
symbiotic notrogen fixation in an established pasture using acetylene reduction and
a 15N method. Soil Biol. Biuchem. 10, 13-20.
HARDY,R.WF.,HOLSTEN,R.D.,JACKSON,E.K.andBURNS,R.C.l968Theacetylene-
ethylene assay for N,-fixation : Laboratory and field evaluation. Plant Physiol. 43,
11851207.
HENZELL, E.F., MARTlN, A.E., ROSS, PJ. and HAYDOCK, K.P. 1968 Isotopic studies
on the uptake of nitrogen by pasture plants. IV. Uptake of nitrogen from labelled
plant material by rhodes grass and siratro. Ausr. J. Agric. Res. 19, 65-77.
JENKINSON, D.S. 1977 The nitrogen economy of the Broadbalk experiments. I. Nitrogen
balance in the experiments. In : Rothamsted Experimental Station Report for 1976,
part 2, Harpenden, Heats, U.K., pp; 103-109.
JENKINSON, D.S. 1982 The nitrogen cycle in long-term field experiments. PhiZsophicul
Transactions of the Royal Society (London) B. 296, 563-571.
KOHL, D.H. and SHEARER, G. 1981 The use of soils highly enriched in 15N to screen
for N,-fixing activity. Plant and Soil. 60, 487-489.
LADD, J.N., OADES, J.M and AMATO, M. 1981 Distribution and recovery of nitrogen
from legume residues decomposing in soils to wheat in the field. Soil Biol. Biochem.
13, 251-256.
LEDGARD, SF., SIMPSON, JR., FRENEY, JR. and BERGERSEN, F-J. 1985 Field
evaluation of lsN techniques for estimating nitrogen fixation in legume-grass associations.
Aust. J. Agric. Res. 36, 247-258.
LEGG, J.O. and SLOGER, C. 1975 A tracer method for determining symbiotic nitrogen
furation in field studies. In : Proceedings of the 2nd-International Conference on Stable
Isotopes. Klein, E.R. and Klein, P.D. Eds. pp. 661666. Oalbrook, Illinois, USA.
LIMA, E., BODDEY, R.M. and DOBEREINER, J. 1987 Quantification of biological ni-
trogen fmtion associated with sugar cane using a 15N aided nitrogen balance. Soil
Biol. Biochem. 19, 165-170.
MIRANDA, C.H.B. and BODDEY, R.M. 1987 Estimation of biological nitrogen fixation
associated with 11 ecotypes of Punicum maximum grown in isN labelled soil. Agronomy
J. 79, 558-563.
NELSON, AN. and WEAVER, R.W. 1980 Seasonal nitrogen accumulation and fmtion
by soybean grown at different densities. Agronomy J. 72, 613-616.
RENNlE, R.J. 1985 Theoretical errors in sequential Nrs labelling of soil to estimate N,
fmation by isotope dilution. Crop Science. 25, 891.
316
RENNTE, RJ. 1986 Advantages and disadvantages of nitrogen-15 isotope dilution to quau-
tity dinitrogen furation in field-grown legumes - A critique. In : Field Measurement
of Dinitrogen Fixation and Denitrification. Soil Science Society America Special Pu-
blication 18. pp. 43-58. American Society of Agronomy, Madison, Wisconsin, USA.
RENNIE, RJ. and THOMAS, J.B. 1987 15N-determined effect of inoculation with Nz fixing
bacteria on nitrogen assimilation in western Canadian wheats. Plant and Soil. 100.
213-223.
RENNIE, RJ., DUBETZ, S.. BOLE, J.B. and MUENDEL, HI-L 1982 Dinitrogen furation
measured by 15N isotope dilution in two Canadian soybean cultivars. Agronomy J.
74, 725.
TALBOTT, H., KENWORTHY, WJ. and LEGG, J-0. 1982 Field comparison of the ni-
trogen-15 and difference methods of measuring nitrogen fixation. Agronomy J. 74,
799-804.
URQUIAGA, S., BOTI’EON, P.B.L. and BODDEY, R.M. 1989 Selection of sugar cane
cultivars for associated biological nitrogen fixation using ‘sN-labelled soil. In : Nitro-
gen fixation with non-legumes. Skinner, F., Boddey, R.M and Fend& I. Eds. pp
311-319. KIuwex Academic Publishers, Dordrecht, Netherlands.
VALLIS, I., HAYDOCK, KS., ROSS, PJ. and HENZELL, E.F. 1967 Isotopic studies
on the uptake of nitrogen by pasture plants. III. The uptake of small additions of
15N labelled fertilizer by rhodes grass and Townsville luceme. Aust. J. Agric. Res.
18, 865-877.
WAGNER, G.H. and ZAPATA, F. 1982 Field evaluation of reference crops in the study
of nitrogen fixation by legumes using isotope techniques. Agronomy J. 74, 607612.
WEBER, CR. 1966 Nodulating and non-nodulatmg soybean isolines. II. Response to ap-
plied nitrogen and modified soil conditions. Agronomy J. 38, 46-49.
WITTY, JF. 1983 Estimating N2-fixation in the field using ‘5N-labekl fertilizer : some
problems and solutions. Soils Biol. Biochem. 15, 631-639.
WITTY, J.F. and MlNCHIN, F.R. 1988 Measurement of nitrogen fixation by the ace-
tylene reduction assay ; myths and mysteries. In : Nitrogen furation by legumes in
Mediterranean Agriculture. Beck, D-P. and Materon, L.A. Eds. pp. 331-343. ICARDA,
Aleppa Syria
Maximiser la FBA pour la Production Agricole et Forestibe en Afrique
Utilisation de la m&hode 16N pour estimer
la fixation symbiotique de l’azote
chez les plantes herbakes et ligneuses.
DOIMENACH, A.M., KURDALI, F. et BARDIN, R.
Universitk Lyon 1. laboratoire de Biologie des So&
VRA CNRS 71. bat. 741, 43 bid ak 11 novembre 1918,
69622 VILLElJRBAhWE Ceder, France.
La m&ode de mesure de la fixation symbiotique de l’aznte bask sur les varia-
tions isotopiques naturelles de l’azote est un outil privilkgiB pour des 6tudes sur le
terrain et en conditions naturelles. En effet elle ne n&e&e aucune intervention
de la part de rexpkrimcntateur
et n’entraine pas de perturbation du systkme ttudib.
Cet article veut mettre en 6vidence les exigences, d&e&nation de l’identiti isoto-
pique de l’azote fix6 et assinG. et les contraintes de cette mkthode, en particulier
au niveau de l’kchantillonnage,
tant pour les plantes annuelles que pour les arbres.
Les quelques exemples d’application don& illustrent les grandes possibiitis de cette
m&ode qui pennet, en particulier, d’aborder l’ktude des systkmes ligneux.
Introduction
L’estimation de la fixation symbiotique d’azote est we &ape essentielle dans toute
&de des systkmes fixateurs. En effet, aussi bien au niveau globaliste, pour connaitre
les rentrkes d’azote dans un tkzosyst~me et pour les optimiser, qu’au niveau rkduction-
nke pour tester des cultivars ou des souches microbiennes s&xXio~&~ ou transform&s,
318
la quantification s’impose. Aussi est-il nt?cessaire d’avoir des outils de mesures fables
a sa disposition.
Une nouvelle generation de spectrometres de masse a double introduction et double
collection dune tres graude sensibilite, a permis de penser qu’il Ctait possible d’utiliser
les differences isotopiques namrelles de l’azote entre fair et le sol pour estimer, chez
les plantes fixatrices, la part d’azote venant de l’une et l’autre source (Delwiche and Steyn,
1970 ; Rennie et al., 1976).
Mais si les variations isotopiques naturelles de l’azote ont Cte mises en evidence
d?s 1939 par Schoenheimer et Rittenberg, variations qui, d’apr& Hoering (1955), auraient
leurs explications dans les processus biologiques, leur utilisation darts la comprehension
et la quantification des processus biologiques a &5 plus difficilement admise. Edwards
(1975) trouve ces variations trap faibles pour Ctre exploitables. Les r&ultats de Kohl
et al., 1971 sur la quantitr5 de nitrate polluant la nappe phr&ique calculee a partir des
variations isotopiques naturelles ont souleve une vive polemique (Hauck ef al., 1972 darts
“a questionable approach”). Bremner (1977) s’etonne, que les critiques formulees n’aient
pas d&courage les recherches a ce sujet et pour lui, mCme si ces variations existent, elles
ne peuvent dormer que des informations qualitatives.
Pourtant, depuis 1977, la m&hode de mesure de la fixation symbiotique bas& sur
les variations nature&s des abondances isotopiques de l’azote s’est ClabonZe (Amarger
er al., 1977, Bardin et al., 1977, Kohl er al., 1980). Si cette m&rode prrjsente certaines
contraintes et limites, son champ &application reste t&s vaste et inclut m&ne les arlxes
fixateurs d’azote. Sa sp&ificite reside essentiellement darts le fait qu’elle n’implique au-
curie intervention de l’exp&imentateur et peut done Ctre appliqut5e sur les systi?mes na-
turels sans les modifier ni les perburber.
Principe
Rappelons que l’abondance isotopique dun compose azou5 est le pourcentage du
nombre d’atome 15N par rapport au nombre total d’atomes d’azote. L’azote atmospherique
mot-me une composition isotopique t&s stable qui est estimee a 0,3663 f 0,0002% r5N
(Mariotti 1984) et celle des composes azotes varie entre 0,364O et 0.37% 15N.
Ces variations sont l&s a l’existence de fractionnements isotopiques. On peut de-
fmir le fr-actionnement comme &ant la repartition diffr5rent.e des isotopes dun element
contenu darts le produit initial et final au tours dune reaction chimique ou biologique.
Les liaisons etablies par l’isotope leger seraient plus ais&nment bris&s que celles &&lies
par I’isotope lourd et les mol&tles contenant l’isotope leger r&$raient plus vite que celles
contenant l’isotope lourd Chaque processus biologique est une suite de phrsieurs reactions
et la teneur isotopique finale est une rrZ.sultante de l’ensemble des fractionnements.
Le principe de la m&rode de mesure de la fixation symbiotique bast5e sur les
variations isotopiques nature&s de l’azote est simple : une plante fixatrice est aliment&
par deux sources d’azote, l’azote mol&zulaire de l’air et l’azote min&-al du sol. Si ces
deux sources ont une abondance isotopique differente, iI peut ttre possible, comme on
le fait apms un marquage, de connaitre par un simple calcul de dilution isotopique, le
pourcentage d’azote venant de chacune de ces deux sources. Tout consiste done a determiner
la valeur isotopique de l’azote venant du sol et celle venant de lair.
319
Valeur isotopique de I’azote fix4
MCme si on sait que lair a une valeur isotopique constante (Mariotti, 1984). celle-
cipeutCtremodifi~lorsqueI’azotemoI~ulaireestm~bo~paruneplante.Cettemodification
t&s faible, negligeable lorsqu’on utilise le marquage, peut entraker, si on n’en tient pas
compte, une grossikre erreur darts l’estimation du taux de fixation pouvant aller jusqu’a
donner des valeurs aberrantes. Cette m&ode nkessite done la determination de l’abon-
dance isotopique de l’azote provenant de la fmation. Cette determination est faite a partir
dune plante fmatrice poussant sur un milieu depourvu d’azote et qui ne dispose que de
l’air comme source d’azote.
Dans la pratique, la mkthode utilise les valeurs isotopiques des parties akiennes
des planks fixatrices car souvent les racines sont difficilement accessibles. Darts ce cas,
il est n6cessaire de tenir compte egalement du fractionnement isotopique possible de l’azote
se produisant a I’intkrieur mCme de la plante. On a pu voir en effet, que les nodosites
du soja et du robinier s’emichissaient en 15N pendant que la partie aerienne s’appauvrissait
en 15N (Reinero et al., 1983, Domenach, 1985). Ce fractionnement lit a la fixation peut
Ctre different suivant l’espece de la plante h8te (Mariotti et al., 1980, Steele et al., 1983)
ou de la souche microbienne (Yoneyama et al., 1984). Si on a pu montrer que, si darts
le genre Aims, lksphx (A. incana et A. ghtinma) et le type de souches de Frankia
influencent peu la composition isotopique de l’azote fv;C (Dommenach et al., 1988), la
valeur trouv~ (615N = -2) est differente de celle du soja (PN = -3). 11 est done rkessaire
de determiner pour chaque syst&ne fixateur Ctudie, le fractionnement isotopique se produisant
a l’intirieur de la plante avant de pouvoir appliquer cette m&rode. Ce fractionnement
evoke au tours du temps au mains durant les premiers stades de developpement de la
plante (Shearer et al., 1980 ; Fig. 1) puis semble se stabiliser. En effet, les feuilles d’auhres
ages de 5-6 ans transplantf5s sur un milieu sans azote ont une valeur isotopique identique
B celle de jeunes planks de 8 mois (Domenach et al., 1988). Aussi il est nt5cessair-e
de kisser pousser les plantes sur milieu sans azote pendant phtsieurs mois avant de pou-
voir d&z-miner la valeur isotopique de l’azote fmt.
6-
4 ..
Figure 1: Evolution en
faction du tunpa du
3 .
6% des nod&t&, par-
tie3 ntriamcs ct plantes
2 ‘.
entiircs d e Robinin
psf?udoacacia palssant
sur milieu saris azote
(d’aprkr Domcnach.
1985).
320
Les valeurs trouvks de ce fractionnement varient d’un laboratoire a un awe et
semblent en park likes aux procedures de preparation des kchantillons et au type d’appareil
utilisk (Shearer and Kohl, 1986). Cette difficult& vient egalement du fait qu’il n’existe
pas de gaz standard dont on connait la valeur absolue. Ces variations d’un appareil a
un autre ne semblent pas gkuuu.es pour la mesure du taux de fixation si l’ensemble des
mesures se rapportant a une experience sont faites aver la mEme m&ode analytique
et sur le mCme appareil. Ceci impose a chaque laboratoire de d&erminer p&ciskment
la valeur de ses param&res et de ne pas utiliser pour ses propres piantes les valeurs
dorm&s par ia littkrature.
Devoir d&erminer les fractionnements isotopiques ii& a la fmtion de chaque syst&me
symbiotique est tme contrainte props a cette m&rode qui vient de l’exploitation de faibles
variations isotopiques qui sent de l’ordre de lo4 (une unit5 8). En respectant cette contrainte
la mkthode peut Ctre gCn&alisable ?I presque tous les systemes symbiotiques fixateurs.
Valeur isotopique de I’azote assimilb
Mesurer directement la valeur isotopique de l’azote du sol s’avkre difficile dans
le choix de la forme a mesurer (organique ou minkale). On sait en effet que le passage
dune forme ?I une autre provoque un fractionnement isotopique. De plus, la mesure des
diff&-entes formes d’azote du sol a un instant dorm6 ne permet pas d’intigrer la vakur
isotopique de l’azote assimilC par la plante durant tome sa vie. Enfin. un fiactionnement
isotopique pcut parfois se prod&e au moment de l’assimilation et modi!kr l’abondance
isotopique de l’azote provenant du sol (Mkiotti et al., 1980). Aussi comme darts la methode
de marquage du sol de Fried et Middleboe (19771, on apprehende la valeur isotopique
de l’azote assimild par le biais dune plante non furattice poussant dans lcs m&nes conditions
que l’on appelle plante de rkf&ence ou plante t&mom. Lorsque ies &&es sont faites sur
des systkmes annuels, cette plante timoin sera choisie en fonction de l’enracinement et
de la p&iode de croissance comparables aux plantes fixatrices &udi&s. Dans les kcosystkmes
naturels on chosira les planks temoins ayant un developpement equivalent a celui de
la plante fixatrice. Toutefois, le probEme de la plante temoin est moins crucial dans cette
m&ode que dans celle du marquage du sol (Rermie, 1982), car ii n’existe pas d’effets
lies au dkclin prononce de l’emichissement du sol avec le temps, aucun engrais marquC
n’ayant ktc5 rajoute. L’Cvolution de la valeur isotopique naturelle de l’azote dun sol est
lente car elle resulte de l’ensemble des processus lids a la formation de ce sol et la valeur
isotopique mesurke de l’azote assimile est peu differente dune arm& sur l’autre. Dans
les articles (Domenach and Corn-tan, 1984 ; Domenach and Kurdali, sous press& on peut
se rendre compte que les mesures faites sur tme aulnaie montrent la grande stabiliti de
la valeur isotopique de l’azote assimile et que le m5me champ de culture en donne la
mCme valeur sur deux ans d’exp&iences.
Cheng er al. (1964) et Karamanos and Rennie (1980) ont mis en evidence une
variation isotopique de l’azote en fonction de la profondeur du sol, aussi il est important
que la plante non fixatrice ait un systeme racinaire exploitant la m&me zone de sol. Ceci
est surtout vrai pour les kcosystkmes naturels install6 sur des sols t&s evoluCs qui peuvent
presenter ces variations. En systeme agricole, grace a la pratique des labours, la valeur
isotopique de l’azote assimile s’avere beaucoup plus homogene (Selles et al., 1986). Ainsi,
nous avons pu montrer, que si un soja gCnCtiquement non nodulant representait la plante
de reference la plus satisfaisante pour une culture de soja, le sorgho pouvait egalement
321
remplir ce role (Domenach and Corman, 1984), comme le ray-grass peut Etre pour le
m3le (Bergersen and Turner, 1983) ou la luzerne (Ledgard et al., 1985). Les plantes
fixatricesnepos~entpastoutesuneisolignegCn~quementnonnodulante,aussilapossibili~
d’utihser une graminke cornme plante de rkference permet d’appliquer la mt5thode a ces
dims plantcs f~trices et done globalement permet une grande utilisation de la m&ode.
Ceci donne Cgalement la possibiliti de choisir une plante non fixatrice bien adapt&e a
la region. Dans un systkme naturel, les planks de reference sont imposkes par le milieu.
Elles seront done des arbres si on veut estimer la fixation d’arbres symbiotiques. Ces
arbres non fixateurs peuvent avoir des valeurs relativement homogenes (Fig. 2) ou comme
S1TE 1
SITE 2
SITEJ
SITE 4
. . . . . . . . . . .
._._..__.._.... i
. . . . . . *-***
*“~...*............““““““‘i
1
1
1
I
1 I
Figure 2 : 6”N de feuilh d’A1nu.s incaM et de diffbxxes
plantes non fuauiccs rt?coltis dans quatrc stations d’une
sulnaie naturelle. * correspond P la vrdeur mesurke SUT un
aulne poussant sur milieu ssns aznte 250 jours zip&s
t
inoadation d’aprk Domenach et Ku&Ii, 1989).
322
le montrent Shearer et al. (1983), pour le desert de Sonoran, des valeurs assez h&Crogenes
(Fig. 3). 11 est alors n&ssaire de prelever le maximum de plantes non fixatrices afin
de rendre compte de l’ht%rogCneitk isotopique du milieu et de collecter les kchantillons
par pa&. Chaque paire sera form&z dune plante fixatrice et dune non-fixatrice les plus
proches afin d’optimiser la similaritt5 isotopique de l’azote que chacune de ces plantes
assimile (Shearer and Kohl, 1986).
J?ignre 3 : RQmition du 6-N
des plantes non fixatrices du
dCsert d e Sonomn (d’aprks
Shearer et al, 1983).
i23456789
SUNof leaf tissue
La m&ode bask sur les variations isotopiques naturelles de l’azote nkcessite l’exis-
tence dune difference entre la valeur isotopique de la plante ttmoin et celle de l’azote
fix& La precision depend directement de cette difference. Cette exigence marque une
limite dans l’application de la mkthode. La valeur de cette difference est fonction du sol
et kchappe a notre cornrole. Aussi, suivant les conditions du milieu et sans que nous
puissions le savoir a l’avance, le taux de fixation ne pourra pas ttre mesure ou au contraire,
aura une valeur tres precise. La prtkision des mesures peut se situer entre 5 et 25%.
Les sols agricoles foumissent en g&kal une valeur isotopique de l’azote assimile largement
plus haute que celle de l’azote fix& Par con&e, en sol de for-Et tempMe, l’azote assimile
aunevaleurisotopique~uventbasseetm~en~gativevenantdel’importancedel’immobilisation
comme le montre le modele de Shearer et al. (1974). Cette valeur est alors peu differente
de celle de l’azote fixe et entraine une estimation plus grossiere de la fmation et parfois
m&me entraine I’impossibilitC d’utiliser cette m&ode.
La valeur isotopique de la plante li &udier est une combinaison des valeurs isotopiques
des dew sources azotkes prkckdemment d&ermin&zs, l’azote assimilt et I’azote fix&
Le probleme se pose alors de savoir comment realiser l’kchantillonnage de ces
plantes. Les rksultats de Reinero ef al. (1983) et de Yoneyama er al. (1986) ont montre
une grande hCt&og&rtW dans les valeurs isotopiques trouvkes pour les differems organes
dun mCme soja au champ. Si on considere uniquement les feuilles de sojas nodules et
323
de sojas non nodules poussant dans les memes conditions (Fig. 4), on remarque une grande
h&krogtnCite IiQ a la position des feuilles (El correspondant aux feuilles les plus basses,
F15, aux feuilles les plus hautes) et li6es a leur stade phenologique (les valeurs isotopiques
des mCmes feuiks sont diff&entes a la formation des gousses et iI la maturation des
graines). &chant que des sojas poussant sur milieu sans azote ont un WN t&s homo-
gene (Mariotti, 1982), les variations observ&es dans la Fig. 4A sont likes ?r l’assimilation
(fractionnement isotopique lors de l’assimilation ou evolution dans le temps du 615N de
l’amte assimile ?). On peut peruser que la complete pertubation des valeurs mew&s a
maturation des graines (Fig. 4B) proviendrait de la remobilisation de l’azote au profit
des graines. Cette remobilisation affecterait plus spkialement l’azote fix& En effet, si
a15N 2 t
1
-%&-
F 2
F 3 4
FS6
Fll 12
-I
Figure 4 : 6”I-J de feuilles de
nivsau dc6 feuiir
sojn (Hodgson)iuocul~s et non
inoculCs en fonction de leur
position sur la plante (Fl MT-
A la maturation des gahes
respond BUX fcuiks lcs plus
basses), B dew stades phtno-
logiques.
B
a 15N
niveau des feuilles
324
on considkre la valeur isotopique de la partie aerienne en&e, la difference per&e entre
les sojas nodulQ et non-nodules alors que les valeurs isotopiques de leurs feuilles devien-
nent semblables. Cette remobilisation preferentielle de l’azote fme a d’ailleurs Cte misc
en Cvidence grace au r5Nz par Warembourg et al. (1984.)
Ces r&&ats montrent que, s’il existe une variation isotopique de l’azote assimilC,
celle-ci n’est pas g&rante darts la mesure oti elle est prise en compte aussi bien par la
plarite fixatrice que par la phmte non-fixatrice. De plus, ces r&&ats mettent en &kience
l’importance de lkhentillonge dans l’application de la mtthode bask sur les abondances
isotopiques naturelles. En effet, pour unmake le taux de fixation d’une plante fixatrice
herback, il est necessaire de p&lever la partie a&ienne en&z de ces planks : le fait
de mesurer exclusivement les graines entraine une surestimation du taux d’azote prove-
nant de la fixation et la mesure des feuilles pourrait entrainer, suivant la p&Me de p&E-
vement, une sous estimation de ce taux.
Pour estimer le taux de fixation d’un arbre, il parait difficile d’analyser tome la
partie aerienne de cet arbre. Les feuilles, production annuelle de ces arbres apparaissent
comme le mat&iel de choix qui peut Cue utilise pour estimer le taux de fixation dune
p&iode vegetative de cet arbre. Ainsi, si on choisit les feuilles, on se refire ZI la production
vCgCtale de km& alors que si on choisit le tmnc ou les tiges, le ph&tom&ne d’accumu-
lation interfere sur le rt5suhat. La valeur isotopique de l’azote mesurke sur le tronc devrait
reprksenter l’historique de la nutrition azot& de l’arbre alors que la vakur isotopique de
I’azote des feuilles reprksente celle de l’anntk Mais dans ces feuilles, une certaine quantiti
d’azote a Ctk remobilisk a partir des reserves situ&s darts les racines (Domenach et Kurdali,
1989) : pour des auks agb de 5-6 ans les r&serves reprksentent encore 10% de l’azote
des feuilles en fin de p&ode vCg&ative (Fig. 5). Cet azote remobilid constitue avec
l’air et le sol, une troisikme source de nutrition dont il faut tenir compte si on veut estimer
le taux d’azote fixC darts ces feuilles. Ce distingo perd de son importance si lXchantil-
lon est constitue des dernikres feuilles form&es, mais on risque ainsi de n’avoir qu’une
Flgure 5 : Pcurcentagc d’nzote provenant
dcs r&s-
ewes des ncines dans les feuilles d’Ahs gluhosa
e t d e PopullLF alha iges de 5 - 6 111s e t poussant e n
conditions naturelIes dumnt tome Ia pCriodc de
croissance. (d’apr2s Domenach et Kurd&i 1989).
--___.
325
vue partielle de ce qui s’est passk dans la saison. Le mEme probkme se pose dans la
mesure de l’activitt fixatrice des l~gumineuses herbacks pkrennes dont les parties akrien-
nes sont composks de 20% de I’azote venant des rkerves apri?s 40 jours de repousse
(Lescure et Chalamet, 1985).
Sur des aulnes en milieu naturel, les feuilles peuvent p&enter une valeur isoto-
pique relativement homog&ne ou au contraire, hCdrogtne comme le montre la Figd. L.es
variations peuvent etre le signe de l’existence de p&iodes de furation et d’assimilation
plus ou mains intenses. 11 est done important, si on veut avoir une estimation globale
de l’azote fix& de prklever, antant que possible, les fenilles snr l’ensemble de la hauteur.
Les arbres constituent un type de materiel biologique malaise? & &udier, mais quelles que
Gent les difficult& d’application de la m&ode bask sur les variations isotopiques na-
tuelles, il convient de noter que c’est la seule utilisable sur les arbres en conditions na-
turelles aujourd’hui.
M&hode de calcul du pourcentage de fixation (%Ndfa)
La base de la mkthode est simple et se rksume B une application du principe de
la dilution isotopique. Le calcul est le mCme pour les plantes herb&s ou ligneuses.
WN pl. non fixatrice - PN pl. fixatrice
%Ndfa =
x 100
ZPN pl. non fiiatrice - PN pl. fixatrice sur milieu sans N
Cette kquation est de la forme : %Ndfa = (X-Y) / (X-C). L’kcart-type peut Ctre
dond par la formule utiliske par Shearer and Kohl (1986).
o--C)
CAY)
6-w
A % N d f a = -
I I A x + - - -
I I AC
(C-xy
I(C-x>l + (C-X)2
Comme il n’est pas mk-ssaire de rajouter une source d’azote comme on le fait
dans la m&ode de marquage du sol, on n’utilise pas le concept de la valeur “A” ce
qui supprime les problkmes Ii& ii cette valeur. Toutefois lorsque l’on introduit des engrais,
la mkthode isotopique des abondances naturelles reste applicable mais peut poser quel-
ques probkmes. En effet, les engrais, &ant fabriqub a partir de l’azote de l’air, ont souvent
un 6”N t&s bas ce qui a pour cons@uence de diminuer la valenr isotopique de l’azote
assimilC. I1 en r&&e une moins bonne prkision dans la mesure de fixation. Cet incon-
vtkient peut Ctre palliC par le controle du rapport isotopique de l’engrais utili& et si n.5
cessaire, la modification de ce rapport par un infme apport de produit marquk.
Comparaison des m&hodes marquage du sol et abondance isotopique na-
turelle
La possibilitk de quantifier la furation par la seule connaissance des variations iso-
topiques naturelles entre une plante fixatrice et une plante non fixatrice est confirm&
par la comparaison des rksultats obtenus avec czux que donne la m&ode de marquage
du sol sur des plantes similaires.
326
Sit* 3
I IEv)E 6
----a
,
Figure 6 : 6”N de fcuilla d’Afnrrs
incnca (A) et de Acer psewiopkz-
L
fanw (B) r6colties au dkpan Ed 1
c
l’extr=SmirC dcs branches d’un
meme arbre sur les bois des
armies pr&ddentes (___ 1 et
I
327
Ces deux m&hodes sont en effet comparables car tomes deux sont indt$endantes
des quantit&s d’azote assimilkes, utilisent une plante de reference, le concept de la dilution
isotopique, et donnent une valeur globale de la fixation. Seul le principe de base change
: dans un marquage, on admet que les microorganismes et les plantes ne penvent pas
faire la difference entre 15N et 14N lorsqu’ils metaboiisent l’azote alors que c’est sur l’existence
de cette reconnaissance des deux isotopes par les microorganismes qu’est baske la m&hode
WN. LAX deux principes ne sont pas contradictoires et I’un ou l’autre s’applique suivant
l’kchelle de precision darts laquelle on se place, les variations isotopiques naturelles n’&ant
sensibles qu’a 1O4% 15N. Cette comparaison de-s deux m&hodes est essentielle pour &ablir
la validit de la mkhode bask sur les variations isotopiques naturelles meme si peu de
publications (Bergersen and Turner, 1983 ; Domenach et Chalamet, 1979; Ledgard et
al., 1985) en font &at.
Le Tableau 1 montre que les deux methodes donnent des rt5sultat.s 6quivalents aussi
bien sur les plantes her-backs que ligneuses justifmt l’emploi des variations isotopiques
naturelles pour mesurer la futation symbiotique de l’azote.
Si l'utilisation de la methode WN prksente quelques limites et contraintes que nous
avons soulign&s tout au long de cet article, elle offre egalement beaucoup de possibilit6s.
Comme la mCtkxie de marquage du sol, elle fait partie des m&odes globales intQrant
Tableau 1 - Comparaison de deux mGthodes d'estimation
du taux de fixation symbiotique : marquage du sol et
variations isotopiques naturelles de l'azote.
% FIXATION
TRAITEMENTS
plantes
marque
non marqw?
NH4+
Soja
6 0
68
5 6
60
Trefle
42
42
N03-
Soja
7 2
69
Aulne
98
100
N-organique
Trefle
45
4 7
Domenach et Chalamet 1979, Domenach and Kurdali 1989.
328
la fixation de mute la vie de la plante et permet ainsi d’avoir une estimation chiffrk
de la contribution de la fixation dans un systkme donne. Elle est Cgalement indepen-
dame de la production et n’utilise pas l’hypothke d’egalid des quantitks d’azote assimile
entre une plante fixatrice et une plante non fixattice. Mais conuairement a la m&hode
de Fried and Middleboe (1977), cette m&ode n’utilise pas de produits marqds. Cette
diffi&nce irnplique de nombrenx avantages :
(i)
il est possible de realiser des mesures sur l’ensemble du systkme 6tudiC
et non plus de se contenter de microparoelles comme on le fait avec
le marqnage, ce qui donne une meilleure image de la fixation rklle,
dun champ par exemple ;
(ii)
tous les problemes lies It l’hCt&ogedid du marquage due a la difti-
cult6 dune application homogene de l’engrais disparaissent ;
(iii) le choix de la plante temoin est plus large car le rapport isotopique
de l’azote du sol est homogene et ne varie que t&s lentement au tours
du temps, diffkemment du d&Am t&s rapide du %15N dun sol apr?s
un marqnage ;
(iv) enfin et surtout, elle n’implique aucune intervention de l’expkrimen-
tateur et peut done &e appliquke sur les kosystemes naturels saris
les perturber ni les modifier. Cette m&ode apparait ainsi la mieux
adapt& a la mesure d’activite fixatrice des arbres sur qui par le fait
mCme de leur taille, la m&ode a l’adtylene ou les mtthodes de mar-
quage semblent delicates a apphquer.
La grande difficult6 tenant a la methode vient des valeurs isotopiques de l’azote
assimil6 souvent basses en fotit dont la consequence est de n’avoir qu’une estimation
grossikre de la fixation.
Champs d’application
Si on peut disposer dun spectrometre de masse, l’emploi de cette m&rode in situ
est t&s aisk puisque la seule intervention de l’expkimentateur consiste a p&lever des
6chantillons au moment opportun. Son champ d’application est t&s vaste et quelques exemples
peuvent rendre compte des nombreuses possibilit6s qu’elle offre :
0 Cette m&ode pourrait permettre, a la mar&e de Virginia and Delwiche
(1982), de r&lker des criblages de plantes peu connues darts des milieux
vierges. En effet, ces auteurs, en r&disant des mesures isotopiques sys-
t&natiques sur les planks prksumkes non fixatrices, ont pu mettre en
evidence qu’une rosa&, Chamaebatia foliolosa, etait en fait fixatrice,
ce qui, par la suite, a pu Cue confirm&
8 Estimation des entr6e.s d’azote dans un systeme agronomique ou natu-
rel. Des estimations de taux de fixation de plusieurs legumineuses cul-
tivkes dans differentes regions de France (Tableau 2) montrent la plus
ou moins bonne adaptation du pouvoir fixateur de ces plantes au climat,
ce qui ne signifie pas que les dites plantes n’aient pas un bon rendement
m&me si leur taux de fixation est bas.
329
Tableau 2 - Estimation des rentrees d'azote dans differents systPmes
par le biais de la fixation symbiotique.
% d'azote fix&
Stations
Aulne*
Soja
Luzerne
TrPfle
Ornon
9 8
Montpellier
40-75
Dijon
10-20
58
56
* Apport de 1itiPre : 700kg N ha - 1
(d'aprPs Bardin et al 1977, Domenach and Corman.1984, Domenach and
Kurdali 1988).
On pourrait ainsi pr&ser, pour les differentes esp?ces de plantes, les
zones geographiques oti se situe I’optimum de leur capacite fixatrice.
Un exemple de la tres grande adaptation au milieu et au &mat est dorm6
par des aulnes, arbres a actinorhizes dont la repartition des espkes depend
naturellement des conditions de milieux. Par exemple en France, Alnus
incanu est localise en montage et Alms gbtimsu, au bard des rivieres.
Les mesures faites sur ces arbres croissant dans leur milieu naturel, soit
pour A. incanu, poussant en altitude sur un support constitd par des
moraines et pratiquement dQourvu d’azote, soit pour A. ghtinosu poussant
en plaine sur un sol riche en azote, montrent la t&s grande capaciti
de ces arbres a fner l’azote de fair (Tableau 3). On comprend mieux
ainsi le role pionnier que ces arbres fixateurs peuvent avoir dans la consti-
tution dun sol et l’inttk~t de boisements mixtes permettant le develop-
pement d’essences plus nobles.
8 Optimisation des entrees d’azote par la fixation symbiotique d’azote.
0)
par la meilleure connaissance de l’impact des facteurs de l’en-
vironnement sur cette activittk Un exemple est don& par Etude
de l’influence du stress hydrique sur la fixation du soja (Obaton
et al., 1982 ; Domenach and Corman, 1985). Cette estimation
est faite sur des cultures en champ ayant support6 des stress
hydriques durant l’optimum de leur cycle nitrate-r&luctase (stress
AN& arret de l’irrigation de mi-juin a mi-juillet) ou nitrogenase
(stress A.R.A., a&t de l’irrigation au mois d’aout) ou les deux
(stress A.N.R. + A.R.A., arret de l’irrigation de la mi-juin a
.
...*
330
Tableau 3 - Estimation du taux d'azote de fixation de deux espkes
d'aulnes dans leur milieu naturel, Alnus glutinosa (arbre de plaines,
de bard de riviere), Alnus incana (arbre de montagne, poussant sur
des moraines),
Espkes
PH
% C organique % N total % CaC03
% N fix@
A glutinosa
7.1
2.7
0.33
0
95-tl3
A incana
8.3
0.5
0.06
6 4
98+20
la fm aofit). Les mesures obtenues a partir de la m&ode des
abondances isotopiques naturelles montrent la grande sensibilite
de la fixation au manque d’eau alors que l’assimilation est moins
touch&z par ces stress (Pig.7, Tableau 4).
w par la recherche du couple symbiotique le plus efficace et le
mieux adapt6 aux conditions de milieu S&ction de la souche
microbienne : s’il est aid de tester des souches au laboratoire,
il est difficile de prejuger de leur capacitk en conditions naturelles
de culture. Pour connditre les modifications que ces souches
peuvent apporter dans la nutrition azotie de leur plante hbte,
la mesure de fixation s’impose (Tableau 5). La m&rode M’ac&ylene
per-met de classer les souches en fonction de leur activite
nitrog6nasique mais ne peut pas donner une estimation quan-
titative de l’azote fm6 et sumstime le pouvoir fixateur des souches
comme la G49 qui pr&ente une faible capacite hydrogenasique.
(iii) s&ction de la plante h6te : par la connaissance du rendement
et des capacitks futatrices de differems cultivars de sojas (Tableau
6), ou d’aulnes de difft?rentes origines (Fig.8), il est possible
de choisir, pour un sol donne ou une region don&e, la plante
h&e qui offrira le meilleur compromis qui tient compte a la
fois du rendement et des capacitks de fixation de I’azote.
331
Figure 7 : AcfivitCs nitrate rC-
ductase (A.N.R) u rcductrice dc
l’achyline (AKA) chcz le soja
(HOdgson) au cows du cycle
vCgitatif B 1’Cvapotranspiration
maximale @T&I) (d’aprks Obaton
cf AI., 1982).
100 iours spr;sscmis
V
V R
1
16 8 stadts phenol.
Tableau 4 - Influence du stress hydrique provoque lot-s de
l'optimum de l'activite nitrogenasique (stress A.R.A.) et de
l'activite de la nitrate-reductase (stress A.N.R.) sur la
fixation et l'assimilation de l'azote.
Conditions
% N fix@
Quantites assimilees
hydriques
(mg.pl-'1
M.E.T.
41
750
Stress A.N.R.
2 5
540
Stress A.R.A.
1 6
491
Stress A.N.R.+ A.R.A.
0
410
M.E.T. : Maximum d'evapo-transpiration. (Domenach and Corman 1985)
332
Tableau 5 - Soja (hodgson) poussant sur sol et inocules avec
differentes souches de Bradyrhizobium japonicum.
Souches
%Nfixe QNfixe
ARA.
i5N nodosites
mg/plante
nm./pl./h.
Gl
2 5
149
202
2.5
GA
3 6
293
1105
5.2
GMBl
81
1140
1614
8.7
649
6 9
1736
2374
9
G3S
87
1948
1777
9
GPsp
7 7
1990
2317
9.8
Avec la collaboration de J.C. Cleyet-Marrel, INRA, Montpellier,
laboratoire des Symbiotes des Racines.
Tableau 6 - % d'azote fixP par differents cultivars de sojas
inocules avec la m&w souche (63 Sp.)
I‘ 11 1 I he
quaIli 6 ir II i*t
E.T.M.
par plante (mg)
-tardives
106-17
93
1120
Amsoy
7 5
857
Al00
7 0
831
-precoces
Hodgson
5 7
579
Dobrudza
70
258
20-18-82
5 9
230
Avec la collaboration de M. Obaton, INRA, Montpellier,
Laboratoire des symbiotes des racines.
333
80-
60-
Figure 8 : Pounrntagc d’amtc fuC dam
les feuilles d’Alw g1dm.w de di6mtes
2
origines r&colt&es dans une plantation
s
d’arbres igts de 5 am.
* 40 .
20 -
Conclusion
Quoique spectaculaires, ces rbultats sont obtenus pour la plus grande part, B partir
d’essais rkalisCs en pays temphk Pour connaitre toutes les potentialit& de la n&&ode
d’estimation de la fixation symbiotique d’azote bask sur les variations isotopiques na-
tureks, il serait nkcessaire de la tester sous d’autres types de chats (tropical, semi-
dbertique...). Egalement, peu de travaux utilisant cette mkthode ont &15 r&tlis& sur les
tkosyst&mes naturels et il est difficile de se rendre compte des limites que le milieu peut
imposer (en particulier sur la valeur isotopique de l’azote assimilC 3 partir du sol).
Si les travaux qui seront effectuh dans le futur conferment et complhent les larges
possibilitks de cette mkthode, celle-ci pourrait pcrmettre une meilleure connaissance des
kosysthnes naturels, et Ctre particulikrement utiliske en foresterie, domaine dans lequel
les mCthodes classiques ne peuvent ou que difficilement Ctre appliqukes.
Remerciements
L,e CEA de Syrie a octroyk une bourse d’&nde B Monsieur Fawaz Kurd&.
334
Wfbences
AMARGER, N., MARIOTTI, A. et MARIOTIT, F. 1977 Essai d’estimation du taux
d’azote fix& symbiotiquement chez le lupin par le tracage isotopique nature1 (%).
CR. Acad Sci. Paris. 284, 2179-2182.
BARDIN. R. DOMENACH A.M. et CHALAMET A. 1977 Rapports isotopiques naturels
de l’azote. II - Application a la mesure de la fixation symbiotique de l’azote in situ.
Rev. Ecol. Biol. Sol. 14, 395402.
BERGERSEN, FJ. and TURNER, G.L. 1983 An evaluation of lsN methods for estimating
nitrogen fixation in a subterranean clover-perennial ryegrass sward. Aust. J. Agric.
Res. 34, 391-401.
BREMNER, J.M. 1977 Use of nitrogen tracer techniques for research on nitrogen fi-
xation. In : Biological nitrogen fmation in farming systems of tropics. Ayanaba, A.
and Dart, P.J. Eds. Pub., John Wiley and Sons, pp 335-352.
CHENG, H.H. BREMNER, J.M. and EDWARDS, A.P. 1964 Variations of is-N abun-
dance in soils. Science. 146, 1574-1575.
DELWICHE, CC. and STEYN, P.L. 1970 Nitrogen isotope fractionation in soils and
microbial reactions. Environ. Sci. Technol. 4, 929-935.
DOMENACH, A.M. 1985 Comparaison des abondances isotopiques naturelles de l’azote
chez deux legumineuses, l’une ligneuse, Robiniapsetuioacacia, l’autre hexlmct5e, Glycine
max. Acta Oecologica, Oecol. Applic. 6. 23-29.
DOMENACH, A.M. CHALAMET, A. et PACHIAUDI, C. 1979 Estimation de la fmation
d’azote par le soja a l’aide de deux m&odes isotopiques. CR. Acad Sci. Paris 289,
29 l-293.
DOMENACH, A.M. and CORMAN, A. 1984 Dinitrogen fixation by field grown soy-
beans : statistical analysis of variations in 15N and proposed sampling pmedure. Plant
Soil. 78, 301-313.
DOMENACH, A.M. and CORMAN, A. 1985 Use of “N natural abundance method for
LIIC: btuuJ UI by~ti~~uuz mauon 01 rrckr-grown soyneans. Iniluence 01 tixauon on assi-
miiation and effect of water conditions on these two functions. Soil Sci. Plant NW.
31, 311-321.
DOMENACH, A.M. et KURDALI, F. 1989 Influence des reserves azo&s sur la for-
mation des feuilles &Alms glutinosa (L) Gaerm. et ses consequences dans l’estimation
de la fixation d’azote. Can. J. Bot. sous presse.
DOMENACH, A.M. and KURDALI, F. 1989 Estimation of symbiotic dinitrogen fixation
in alder forest by the method based on natural 15N abundance. 7* IntemationaI Meeting
on “Frankia and Actionorhizal plants” Stmrs, Conr~~ticut. Plant Soil, sous presse.
DOMENACH, A.M. KURDALI, F. DANNIERE, C. et BARDIN, R. 1988 Determination
de I’identite isotopique de l’azote fix6 par Frank51 associe au genre Alms. Can. J.
Bot. 66, 1241-1247.
EDWARDS, A.P. 1975 Isotope effects in relation to the interpretation of 15Np4N ratios
in tracer studies. In : Symp. on isotope ratios as pollutant source and behaviour in-
dicators. IAEA SM. - 191/32.
FRIED, M. and MIDDLEBOE, V. 1977 Measurement of amount of nitrogen fixed by
a legume crop. Plant Soil. 47, 713-715.
335
HAUCK, R.D. BARTHOLEMEW, V.W. BREMNER,J.M. BROADBENT, F.E. CHENG,
H.M. EDWARDS, AS. KEENEY, D.R. LEGG, J.O. OLSEN, SX. and PWORTER
L.K. 1972 Use of variations in the natural nitrogen abundance for environmental stu-
dies : a questionable approach. Science. 177, 453456.
HOERING, T. 1975 Variations of i5N abundance in naturally occuring substances. Science.
122, 1233-1234.
KARAMANOS, R.E. and RENNIE, D. A. 1980 Changes in natural i5N abundance asso-
ciated with pedogenic processes in soil. II Changes on different slope positions. Can.
J. Soil Sci. 60, 365-372.
KOHL, D.H.. SHEARER, G.B. and COMMONER, B. 1971 Fertilizer nitrogen conhiiution
to nitrate in surface water in a corn belt watershed. Science. 174, 1331-1334.
KOHL, D.H. SHEARER, G. and HARPER, J.E. 1980 Estimates of N, fixation based
on differences in the natural abundance of 15N in nodulating and normodulating iso-
lines of soybeans. Plant Physiol. 66, 61-65.
LEDGARD, SF. SIMPSON, J.R. FREENEY, J.R. BERGERSEN, F.J. and MORTON,
R. 1985 Assessment of the relative uptake of added and indigenous soil nitrogen
by nodulated legumes and reference plants in the i5N dilution measurement of N,
fixation : glasshouse application of the method. Soil Biol. Biochem. 17, 323-328.
LESCURE, C. et CHALAMET, A. 1985 Etude de l’influence des nkerves azot&e.s dans
l’estimation par 15N de la fEation symbiotique dune plante pkenne, Trrfolium prutense.
Can J. Bot. 63, 991-994.
MARIOTI’I, A. 1982 Apports de la gkochimie isotopique a la connaissance du cycle
de l’azote. These d&at, Universitk Paris VI, 476~.
MARIOTTI, A. 1984 Natural r5N abundance measurements and atmospheric nitrogen stan-
dar calibration. Nature. 311, 251-252.
MARIO-F-I-I, A. MARIOTTI, F. AMARGER, N. PIZELLE, G. NGAMBI, J.M. CHAM-
PIGNY, M.L. and MOYSE, A. 1980 Fractionnements isotopiques de l’azote lors des
processus d’absorption des nitrates et de la fmation de l’azote atmospherique par les
plantes. Physiol. Veg. 18, 163-181.
OBATON, M. MIQUEL, M. ROBIN, P. CONEJERO, G. DOMENACH, AM. et BARDIN,
R. 1982 Influence du deficit hydrique sur l’activite nitrate rkluctase et nitrogthase
chez le soja (Glycine max L. Merr. cv. Hodgson). CJI. Acad. SC. Paris 294, 1007-
1012.
RElNERO, A., SHEARER, G., BRYAN, B.A., SKEETERS, J.L. and KOHL, D.H. 1983
Site of namral 14N enrichment of soybeans nodules. Piant Physiol. 72, 256-258.
RENNIE, RJ. 1982 Quantifying dinitrogen (N,> fixation in soybeans by i5N isotope dilu-
tion : the question of the non fixing control plant. Can. J. Bot. 60, 856861.
RENNIE. D.A., PAUL, E.A. and JONES, LE. 1976 Natural nitrogen-15 abundance in
soil and plant samples. Can. J. Siol Sci. 56, 43-50.
SCHOENHEIMER, R. and RlTlENBERG, D. 1939 Studies in protein metabolism : I-
general considerations in the application of isotopes to the study of protein meta-
bolism. The natural abundance of nitrogen isotopes in aminoacids. J. Biol. Chem.
127, 285-290.
336
SELLES, F., KARAMANOS, R.E. and KACHANOSKI, R.G. 1986 The spatial varia-
bility of nitrogen-15 and its relation to the variability of other soil properties. Soil
Sci. Sot. Am. J. 50, 105110.
SHEARER, G., DUF’FY, J., KOHL, D.H. COMMONER, B. 1974 A steady-state model
of isotopic fractionation accompanying nitrogen transformations in soil. Soil Sci. Sot.
Amer. Proc. 38, 315-321.
SHEARER, G., KOHL, D.H. and HARPER, JE. 1980 Distribution of r5N among plant
parts of nodulating and nonnodulating isolines of soybeans. Plant Physiol. 66. 57-
60.
SHEARER, G. and KOHL, D.H. 1986 N,-fixation in field settings : estimations based
on natural “N abundance. Aust. J. Plant Physiol. 13, 699-756.
SHEARER, G., KOHL, D.H., VIRGINIA, R.A., BRYAN B.A., SKEBTBRS, J.L., NJLSEN,
E.T., SHARIFI, M.R. and RUNDEL, P.W. 1983 Estimates of N, fixation from varia-
tion in the natural abundance of lsN in Sonoran desert ecosystems. Oecologia (Berlin)
56, 365-373.
STEELE, K.W., BONISH, P.M., DANIEL, R.M. and OHARA, G.W. 1983 Effect of
rhizobial and host plant on nitrogen isotopic fractionation in legumes. Plant Physiol.
72, 1001-1004.
VIRGINIA, RA. and DBLWICHE, C.C. 1982 Natural 15N abundance of presumed N2-
fling and non-fixing plants from selected ecosystems. OecoZogiu (Berlin) 54, 317-
325.
WAREMBOURG, FR., HOEGEL, B., FERNANDEZ, M.P.et MONTANGE, D. 1984 Dis-
tribution et utilisation des assimilats carbon& en relation avec la fixation symbio-
tique de I’azotc chez le soja (Glycine max. L. Mcrr.). Plant Soil.082, 163-178.
YONEYAMA, T., YAMADA, N., KOJIMA, H. and YAZAKI, J. 1984 Variation of natural
lsN abundance in leguminous plants and nodule fi-actions. Pkznt Cell. Physiol. 25,
1561-1565.
YONEYAMA, T., FUJLTA, K., YOSHIDA, T., MATSUMOTO, T., KAMBAYASHI, I.
and YAZAKI, J. 1986 Variation in natural abundance of “N among plant parts and
in “Np’N fractioation during N, fmation in the legume-Rhizobia symbiotic system.
Plant Cell. Physiol. 25 no 8 : 1561-1565.
Maximisez la FBA pow la Production Agricole et Forestihe era Afrique
Nitrogen fixation in tropical trees :
Estimations based on 15N techniques
SANGINGA, N.(l), ZAPATA, (l) and DANSO, S.K.A@)
(I) : FAOIIAEA Programme, IAEA Seibersdorf Laboratory, A-2444 Seibersdor$ Autria ;
(2) : Joint FAOIIAEA Division, IAEA Wapmerstrarse 5, A-1400 Vii, Austria.
Abstract
Because of the importance of trees in agroforestry. there is a crucial need to
assess the magnitude of N2 fixed by them, directly in the field The long-term evalu-
ation of the N,-fixing potential and actual amounts of N, fixed in a given tree raise
“rr\\h,,~mr :_ q”v,.-.?:l, n,.,...ti ^r_.. .: .,I. 1, .-. .I..-“.:-_. .r .- 1
I
.ij -...-.. .,.. .,_ D’
I.. .I .-7 ‘.,-P 7. ;,,
. . . . -. ‘,
obtaining reference crops over several seasons. For young or small trees,
‘sN proce-
dures similar to those adopted in gram or pastnre legumes would be expected to
give equally satisfactory results. The greatest problems with
N2 fixation estimations
using “N will occur in mature trees, due to their perennial nature and massive s&s.
leading to logistic and sampling difficulties or differences in tsN/?N ratio of soil
due to diierences in nitrogen turnover processes that occur under the fixing and
reference crops with time. The influence of these effects may differ depending on
how the 15N is applied. An examination of which of the existing ‘sN procedures,
e.g. isotope dilution, A-value or the natural “N abundance and what method of ‘?J
application should be adopted under different situations is therefore urgently needed.
In the paper we have considered the strategies for obtaining representative
samphng
(as against the whole destructive plant sampling) that would closely reflect the overall
r5N enrichment in the whole plant, and the selection of the appropriate reference
tree. The advantage in using the
lSN labelled plant samples for additional agronomic
studies besides N, fixation, e.g. contribution of tree litter to plant N uptake and
soil organic matter are briefly discussed.
338
Introduction
In recent years the potential of leguminous and actinorhizal trees has aroused con-
siderable interest., especially as regards to their socioeconomic importance and increased
use as firewood, paper pulp, forage and timber (NAS, 1979 ; Dommergues, 1982 ; Torrey,
1982). These nitrogen fming trees (NFTs) contribute to restoration and/or maintenance
of soil fertility (Kang, 1985) and are very important in sound tropical soil management
for sustained agricultural productivity, and controlling soil erosion and degradation.
Until recently however, few studies have been conducted on the examination of
root nodules (Allen and Allen, 1981) and the magnitude of atmospheric N, fixed by dif-
ferent Rhizobium or Fmnkia strains, key factors in the successful growth of NFTs in
nitrogen deficient soils. To maximize the contribution of this natural and inexpensive
source of N in agroforestry systems, reliable methods are needed for quantifying biologi-
cally N, fixed in trees.
Various methods, such as the total nitrogen difference (Comet et al., 1985 ; Gauthier
et al., 1985 ; Sanginga et al., 1985,1986 ; Ndoye and Dreyfus, 1988), acetylene reduction
assay (Roskoski, 1981; Hogberg and Kvarnstrom, 1982 ; Lulandala and Hall, 1988) nodule
number and mass (Roskoski, 1981 ; Hogberg and Kvamstrom, 1982) and ureides pro-
duction (Kessel et al., 1987 ; Sanginga et al., unpublished data) have been used to assess
N, fixed by NFTs. Most of these techniques, analyzed in several critical reviews (Fried
et al., 1983 ; Rennie and Rennie, 1983 ; Danso, 1985, 1986 ; Dommergues, 1987),
are based on indirect criteria, and are either qualitative or cannot distinguish between
the various sources of nitrogen in a faring plant.
Nitrogen-15 techniques, that involve the application of “N labelled materials to
soil have been suggested in many studies to be the most reliable method for quantifying
N,-fmed (Legg and Sloger, 1975 ; Rennie, 1982 ; Vose et al., 1982 ; Rennie and Kemp,
1983 ; Hardarson et al., 1984 ; Danso, 1985 ; Danso et al., 1986). These 15N techniques
are particularly useful as unlike many others, they provide an integrated measurement
of the atmospheric N, fixed by field-grown crops during the experimental period and
can discriminate among the various N sources available to the crop.
T-1.- 15%. 1 . .
I . -1.
_
.
.,.. 1 . ...-1 U..i .tiLA...:.; ~~L~~~L~~u~.J +iiL III>, Y3LU L” L>LIIII&LL lh2 ilXc;cl III
pasture and gram legumes. They have however not been used much in tree N, fixation
studies. Experience gamed from grain and pasture legumes would therefore be valuable
for extending this technique to trees.
This paper examines critically the advantages and difficulties reported for the various
techniques involving the application of lW-labelled fertilizers to soil, and the reliability
of the method based on the differences in the natural abundance of lsN between soil and
atmospheric N, to measure N, fixed and how these could be used in devising suitable
procedures for tree N2 fixation assessment Some estimates of N, fixed in various NFTs
under greenhouse and field conditions using these methods will be also reported.
Basis of the lSN methods in NFT studies
There are two main approaches of the r5N techniques for assessing N,-fixation :
the natural abundance of 15N and the W-labelled fertilizers. The “N natural abundance
is based on the small differences in 15N abundance which frequently occur in nature between
atmospheric N, and soil N, in contrast to r5N substratum labelling techniques where the
339
larger differences in 15N between the soil N and atmospheric N, are induced by applying
15N labelled materials.
Both approaches are however in essence, based on the same principle, i.e. the
higher 15Np4N ratio absorbed from soil N by the N,-fixing plant decreases during fma-
tion by incorporation of the lower 15N enrichment of atmospheric N, ; the more N, that
is fixed, the greater this dilution. Since it is not possible to assess the integrated
i5Np4N ratio of the soil N over a growing season by chemical extraction a non-fixing
reference plant is used for this purpose.
Several articles have reviewed the concepts, equations and procedures used in ap-
plying the lsN methods for measuring nitrogen fixation of grain and pasture legumes (Shearer
et al., 1978 ; Knowles, 1981 ; Fried et al., 1983 ; Rennie and Rennie, 1983 ; Danso,
1985, 1986 ; Dommergues, 1987 ; Hardarson et al. 1987). The researcher must be aware
that there are however as yet no generally accepted procedures which can be claimed
to be valid under all conditions. NITS present specific problems which deserve special
attention.
Factors affecting the validity and precision of N, fixation estimation by trees
using 15N-labelled materials
Reference tree
The accurate determination of the actual amounts of N, fixed in the field is crucial
only in some instances, such as in N-balance studies in agro-ecological systems i.e. agrofo-
restry, or in comparing N, fixed in different seasons, years and environments. In these
cases the reference crop constitutes the main potential source of error in the 15N tech-
nique. The criteria used in the selection of suitable reference crops have been discussed
by Fried et aL, (1983) and Danso (1985, 1986). However, according to Danso (1986)
and Danso et al. (1986) the need for the precise quantification of N, fixed may not be
that compelling in studies to simply compare treatment effects or rank plant genotypes
or RhizobiumlFrankia strains for N, fixing abilities.
Uninoculated N,-fixing legume or actinorhizal trees have often been used as ref-
erence crops (Table 1). Since no indigenous rhizobia or Frunkia were present in the experi-
mental soil the uninoculated nitrogen fixing trees were found to be suitable reference
crops (Comet et al., 1985 ; Gauthier et al., 1985
; Sanginga et al., 1986 ; Sougoufara
et al., 1988). However, care must be taken to thoroughly examine roots of such unino-
culated trees to ensure that they are not nodulated or run an acetylene reduction test
to confirm the absence of nitrogenase activity. Cross-contamination of uninoculated control
from inoculated tmatments has heen observed in our pot experiment involving NITS (unpu-
blished data) and has also been reported by others (Sougoufara et al., 1988). Using such
controls, N2 fixation measured by the isotope dilution and the difference methods were
underestimated, (Sanginga et al., unpublished data). It may be difficult to avoid cross-
contamination in the field and even in the greenhouse unless special precautions are taken
(Bromfield and Ayanaba, 1981). If possible, uninoculated control plants should be com-
pared with other potential reference crops such as known non nitrogen-fixing trees. The
non-fixing trees, however, have to fulfill the conditions outlined by Fried et al. (1983)
and Danso (1986) to be an appropriate reference crop. The validity of such selections
can be established by comparing the isotopic composition of nitrogen with that of non-
nodulating isolines (Fried et al., 1983). Since non-nodulating tree genotypes have not
340
yet been identified, further search for non-nodulating fixing trees in natural populations
or their development (e.g., through mutation induction) are most valuable for the rsN metbo-
dology and attempts have commenced. A simple procedure can be planting of several
hundreds of Camarina and Allocasuarina photosynthetic branches in a sandbath irrigated
with heavy suspensions of Frunkiu. After a period of growth, the roots are examined
for the absence of nodules. The identified potential non-nodulating isolines can be vegeta-
tively propagated and again re-inoculated with Frunkiu to ‘ensure that this trait is stable.
Screening fning trees for host and strain specificity could also provide suitable controls
in some situations.
Table 1 : Reference plants for estimating N2 fixation using 15N techniques
Authors
Type of
Fixing
Reference
experiment
plant
plant
GAUTHIER et al.
-
-
Field
Casuarina
Casuarina
(19851
=-F-i?-
v
CORNET et al.
Field
Acacia
Acacia
(19851 - -
holosericea
lX?%%icea
(II
(Ul
SANGINGA (19851
Field
Leucaena
Leucaena
leucocephala
leucocephala
ZAHARAH et al.
-
-
Field
Leucaena
Setaria ancepts
(19861
'FT?-
(U)
NDOYE and
Pot
Sesbania
Sesbania
DREYFUS (19881
lm3
rostrata (11
rostrata (UI
5
.
n
(I)
-
-
S.n (U1
-
-
SOUGOUFARA et al.
-
-
Pot
C. equisetifolia
C. equisetifolia
(19881
(15,50,1000 11
--Tr----
-0
SANGINGA et al.
Pot
C. equisetifolia
Allocasuarina
IunpublisriS;d~
(5 kg)
c. cunmnghamiana
Allocasuarina
SANGINGA et al.
Pot
Leucaena,
(1987. 193X3-
(5 kg)
(;llncldia
Mmea
-
-
Acacia albida
I: Inoculated
U: Uninoculated
341
Genetic variation in 15N fixation
Large plant-to-plant variation in nodulation and N, fixation have been reported
for NFIs (Gauthier el al., 1985 ; Sanginga et al., 1988b; Sougoufara et al., 1987). The
genetic heterogeneity of NFTs has hardly received the attention it deserves. Danso er
al. (in preparation) indicated that this genetic variation may contribute to the poor preci-
sion in the estimates of N, fixation using the 15N methodology. Duhoux and Dommergues
(1985) observed that average percentage N, fixed based on tree sire, varied from 16 (in
the smallest trees) to 70 (in the biggest-sized trees). In such cases, it is essential to select
well-matched reference crops for each plant to lessen errors. Where plants can be grown
by vegetative propagation, this variation could certainly be reduced, and this could be
practical with many NJ!Ts.
Labelling techniques
There are several practical questions on the techniques of applying the labelled
15N materials which need to be addressed before designing “N experiments to measure
N, fixed. These include the N rate of application and lsN enrichment, chemical and phy-
sical form and time of application. There is evidence that labelling techniques can have
a significant effect on estimates of N, fixed. These techniques, described in several re-
views (Chalk, 1985 ; Danso. 1985, 1986), have mainly been used for grain and pastures
legumes. The limited data available for NITS indicate that much more work still needs
to be carried out to develop adequate techniques for measuring N, fixation by NFTs.
(i) Type and rate of “N labelled fertilizer
The chemical form of the labelled fertilizer is important because of its influence
on N, fixation. This is due to the difference in solubility, uptake by the reference and
fming crops, relative movement and biological transformations in the soil. The isN fertilizers
used so far for NFfs have been in the form of NH,, i.e. ammonium sulphate (Comet
et al., 1985 ; Gauthier et al., 1985 ; Zaharah et al. 1986 ; Ndoye and Dreyfus, 1988 ;
Sanginga et al., 1988a) although no specific reasons can yet be adduced for this preference.
Lower cost of the ammonium sulphate, less interference with fmations and lower potential
losses of N may be one of the reasons. Only recently, Sougoufara et al. (unpublished
data) used urea but there is a high risk for significant N losses.
The amount of “N applied for labelling a tree will depend upon the N rate and
the ‘W enrichment of the labelled material utilized. The N rates and i5N enrichment used
for grain and forage legumes can be readily adopted for small tree plants grown in the
greenhouse or in field conditions.
If isotope-aided experiments are performed with large-sized trees, the amount of
N already present in tree may constitute an extra-dilution factor of the 15N applied. Under
such conditions, it is highly advisable to conduct a preliminary experiment with few trees,
to ascermin several questions, e.g. total N and its partitioning among tree organs, labelling
techniques, as well as sampling procedures.
Single application rates of 2g N/m2, or 0.5 g N/tree enriched with 10 atom %
N have been used to NETS (Comet et al., 1985 ; Gauthier et al., 1985 ; Ndoye and
Dreyfus, 1987). Sanginga et al. (1988) in a single application used twice this rate. Appli-
cation rates will however vary for different NITS, type of studies, locations and soil charac-
teristics, and 15N measuring equipment. As a general guide, the 15N rate must be enough
not to significantly interfere with N, fixation in NFTs (Danso, 1986), and also capable
342
of being detected within the sensitivity range of the isN measuring equipment. For an
emission spectrometer assay, a higher 15N level is needed for detection than by a mass
spectrometer.
In most studies, the isotope dilution method, which involves the application of
the some amount of N-15 labelled material to both the reference and fixing NETS has
been adopted (Table 2). Comet et al. (1985) and Gamhier er al. (1985) compared the
“A value” with the isotope dilution method for estimating N, fixed by Acacia holosericeu
and Cusuarinu equisetifoliu grown in a N-deficient sandy soil. To prevent N-deficiency
in the reference crop so as to ensure comparable satisfactory growth as the N,-fixing
plant, a higher rate of N. 100 kg N/ha was added to the plots of the non-fixing reference
crops, while a lower rate, 20 kg N/ha that would not significantly inhibit N, fixation
was applied to the fixing crops. These authors found that the “A value” technique gave
slightly lower estimates of N, fmation than the isotope dilution and the difference method.
Altbough the “A value” expression is not dependent on fertilizer N addition rates and
fertilizer recovery by both fixing and non-fixing plants, errors can be introduced with
factors affecting differentially the applied fertilizer, e.g. losses due to leaching, deninifi-
cation, N turnover, etc.
Table 2 : Labelling techniques for N2 fixation studies using 15R methods
Authors
Type of
Form
N Addition Atom % 15N Comparison
experiment
rate
excess
other methods
GAUTHIER et al
- -*
Field
(NHqk?S04
;;OUby)
10.5(I.U)
Difference N
11985)
1.91 U ) A-Value
CORNET et al.
(1985) - -
Field
lNH412SO4
2O(I,UI
10 1I.U)
Difference
SANGINGA
Field
lNH412SO4
401I.U)
10 lI,U)
Difference
t 1985)
ZAHARAH et al.
-
-
Field
lNH4)ZS04
2011,R)
10
lI,R)
-
119861
NDOYE and
Pot
lNH4)2SO4
2011,U). 1 0 lI,U)
Difference
DREYFUS
11987)
SOUGOUFARA
Pot
Urea
201I.U)
9.19(I.U) Difference N
et al.
115.50.
100 (U)
1.89 1lJ)
A-Value
7l9aE)
1000 1)
SANGINGA et
Pot
KNOX
2OlI,UI
Difference N
al. 1198r
(5 kg)
100(u)
'! IY'
A-Value
T!%8)
SANGINGA et
Pot
KN03
20 (I)
10 1I.U)
Difference N
~18\\1987~
15 kg)
20,100(uI
2 (u)
A-Value
343
(ii) Methods, frequency and time of “N application
Several workers have reported that the decline in the soil 15NF4N ratio with
time influences the accuracy of Njfixation estimates ; the lower the rate of decline, the
less serious the errors associated with any mismatch between the N uptake by reference
and fixing crops (Fried et al., 1983 ; Witty, 1983). This could be very important in field
experiments involving NITIs because of the long period of growth. In this case a single
initial application of labelled fertilizer may prove to be inadequate for measuring N,-
flation over a long period. Multiple additions of small concentrations of highly labelled
fertilizer to the soil, use of slow-release N fertilizers and ‘SN-labelled organic matter offer
great promise in N,-fixah‘on with NFfs, since the i5NP4N ratio would remain relatively
stable due to the rather slow release of a small but fairly constant amount of N with
time. The greatest problem with these approaches however is that when applied N rates
are lower than needed to support a non-fixing crop over several seasons, the reference
crop may grow poorly or even die.
The lsN fertilizer has been applied to NITS at seeding (Comet et al., 1985) or
soon after germination (Sanginga et al., 1988) or at transplantation of the 3.5 month-
old plants (Gauthier et al., 1985). However, it has been proved by several workers that
early growth of many NITS is slow and that nodulation and N,-fmation is delayed some-
times for several months (Reddell et al., 1986 ; Sanginga et al., 1988a). It is therefore
advisable to delay 15N application until both fixing and reference crops are well-matched
in growth and probably in N uptake.
i5N labelled fertilizer has generally been applied in solution, although a few in-
vestigators (Sanginga et al., 1988a) have broadcasted (NHJ,SO, on field plots and then
incorporated it into the soil with water additions. Several other methods of 15N appli-
cation have been tried for grain legumes (Fried et aI., 1983 ; Chalk, 1985). They have
been reported to give similar results and are thus all satisfactory ways of adding i5N to
soil for N,-fixation studies. Methods of application and their influence on N,-fixation esti-
mate by NITS have not been investigated yet. In some cases the soil injection may provide
a satisfactory approach to apply the “N labelled fertilizer to similar root depths of both
fixing and non-f=ing trees. The choice of one over the other for NITS will be dictated
largely by the homogeneity of application, necessity of diluting the labelled materials
and the objectives of the experimental work.
Cropping systems and nitrogen cycling
The integration of trees, especially NFTs, into stable agroforestry systems can con-
tribute significantly towards restoring and maintaining soil fertility, combating erosion
and desertification and in providing fuelwood. The potential of such trees has already
been shown in alley farming systems in the humid tropics (Kang et al., 1981, 1985).
In this system in which rows of NFTs alternate with several rows of food crops,the NFTS
are periodically lopped, the foliage is used as green manure for food crops or as forage
for animals and the stems provide fuel. However much of scientific background required
for the development of agroforestry technology is still in its initial stage and almost
nothing is known about the magnitude of factors that influence N, fixation and N cycling
processes in this system.
344
The cropping systems could also affect N, fixation as shown in grain and pasture
legumes (Danso et al., 1986) by influencing the suitability of the reference crop and therefore
the validity of the estimates made.
Variations in N, feting activity with the age of trees, or interference by different
processes such as litter fall and its decomposition, and the redistribution of nitrogen in
the different compartments of the tree/soil system may present some difficulties for measuring
N,furation using lsN isotopic methods. Significant amounts of N, fixed by NITS are undoubtedly
finally incorporated into the soil, but no attempts have been made so far to quantify
their magnitude and significance on soil N pool. This could be very important in agroforestry
systems such as in alley cropping where NFTs and non-NFTs are grown in alternate
rows. The soil nutrients under these contrasting plants have been reported to mineralize
at different rates due to differences in litter quality (Swift, 1984, 1985). NITS, because
they contain more N than non-NFTs tend to contribute more N to the soil. This may
lead to conditions where the fixing and non-fixing crops are not absorbing soil N from
the same isotopic composition (15NP4N ratio) during the growing period and thus invalidate
one of the most important assumptions of N, fixation estimates (Chalk, 1985). These
factors can significantly affect the application of the r5N methodology. The periodic collection
and an estimation of the N in fallen leaves in a given area could be used to assess the
error of measurement caused by N in falling and decomposing litter.
Transfer of N, fared by NFfs to associated reference non-NFTs could be another
source of error for N,-fixation estimates (Fried et al., 1983 ; Chalk, 1985 ; Danso et
cd., 1986).
Sampling of plant material
Witty (1983) emphasized that N, fixation estimates based on isotope dilution are
not estimates of the amount of N, fixed by a crop, but rather estimates of the amount
of fixed N contained in the harvested portion of the crop. The r5N enrichment in different
plant parts of crops grown on “N enriched soils frequently differ (Fried et al., 1983 ;Rennie
et al., 1978). Measurements of N,-fixation, based on only one plant part are therefore
not adequate because the “N enrichment of one plant part is not representative for the
whole plant. This difference in enrichment among plant parts has been cited as a problem
in the “N-isotope technique for estimating N,-fmed (Heichel et al., 1984 ; Chalk, 1985).
According to Danso et &(1986) and Fried et &(1983), errors due to differences in
r5N enrichment can be minimixed by sampling each of these plant parts with fairly uniform
N-isotopic composition separately for N and 15N determinations and using theweighted
atom % lsN as a basis for the N, fixation estimate for the whole plant (Danso et aI.,
1986). This seems to be a standard procedure adopted in most N2 fixation studies that
use 15N and is feasible for tree seedlings and young NFTs. Problems may occur when
sampling big trees, because of their high above and below-ground biomass. In contrast
to grain legumes, NFIs have significant proportions of their total biomass and N content
below ground (Young 1985). In this case, quantification of N, fixed that disregards roots,
nodules and crowns would result in serious errors and the proportion of N, fixed may
be largely underestimated. In addition, for many perenniel legumes the N stored in these
underground structures and unharvested portions of foliage may, after decomposition exert
a significant effect on the dilution of the soil 15Nr4N ratio. This is a practical problem
345
which may not be easy to resolve, given the difficulty in recovering all roots and litter.
As suggested by Danso (1986), it is advisable to estimate this stored N and make
a correction for its effect on the isotopic composition of N at harvest, An examination
of procedures for obtaining a representative sampling (as against whole plant sampling)
that would closely reflect the overall r5N enrichment of the whole plant is urgently
needed.
In alley cropping and most the agroforestry systems however ‘W assay in leaf sam-
ples appears to be the practical method of sampling. The validity of the technique would
be based on the assumption that any defined type of leaf sample would probably provide
the same answer with regard to N, fixation of the whole plant. Leaves of similar age
and morphological position should be sampled from the tree at various intervals after
application of lsN. The method of sampling however, should be established after a series
of preliminary experiments. Another approach may involve the forester methods of using
formula to estimate the standing biomass of trees. In this case, it should be possible sample
a few trees to estimate small portions of different organs of the tree for lsN enrichment
and using the formula to estimate total N, fixed.
Sampling of NITS constitutes the greatest problem in N,-fixation estimation and
ways to overcome this needs to be seriously addressed. These sampling procedure problems
are further confounded by the practicality to use single trees usually as experimental units
to cut down cost, a practice that introduces high variability in the estimates.
N, fixation estimates by some NFT using 15N isotope techniques
The isotope dilution method has predominantly been used to evaluate N, fixation
by a few trees in the field. Estimates of N,-fixed of some NFI’s by different methods
used are summarized in Table 3.
The natural 15N abundance based on the study of small differences between the
natural abundance of 15N in non-fixing and fixing trees has been used for tree by few
workers (Khol and Shearer, 1981 ; Domenach, 1987). One of the first studies using this
method was carried out on Prosopis in the Sonoran desert. The natural 15N abundance
in the tree was significantly lower than in the soil, indicating that it had fixed about
40 to 60% of its nitrogen though no nodules were found. Prosopis was presumed to
develop nodules on deep roots which are not normally harvested (Virginia et al., 1981).
More recently, Hogberg (Pers. comm.) reported in a pilot study of N,-fixing potential
of legume trees in a mixed tropical woodland in Tanzania that the N:P ratio and the
15N natural abundance appeared to be more reliable indicators on N,-fixation than the
N content alone. However, its application is limited by the availability of a high precision
massspecm>meterandspecialsamplepreparationprocedures(Amargeretal.,
1979; Domenach
et al., 1979). Other problems include the natural variations in atom 96 15N in soil with
depth (Steele et al., 1981) the spatial variability in the field (Shearer et al., 1978), and
the soil enrichment in 15N during litter decomposition and subsequent plant i5N uptake
(Turner et al., 1983 ; Domenach, 1987). This could be important in agroforestry systems
such as alley cropping where NFTs are sequentially cut and used as N source after decom-
position (Kang et al., 1985).
The feasibility of using 15N-depleted material for assessing N, fixation of legu-
minous treeshasonlybeeninvestigatedonce(KesselandNakao, 1986). Significantdifferences
in atom % 15N and percentage N derived from N2 fixation were measured in Albizziu
Table 3 : N2 fixation estimates (Proportion <rtd actual amounts of N2 fixed) by NFT's
N2 fixation estimates
Authors
Fixing plant
% Nitrogen derived from Atmosph.
N2 fixed from Atmosphere (g/pL)
Difference
Isotope
dilution
A-value
Difference
Isotope
dilutton
A-Value
-
~:;UM;ER E 2.
C. equisetifolla
49
55
3
9
3.07
3.27
2.31
CORNET et al. (1985)
-
-
A. holosericea
29
33
32
1.12
1.35
1.26
SANGINGA et al.
C. leucocephala
39
4Q
ND
98*
134*
ND
(1985) - -
ZAHARAH et al.
L. leucocephala
-
ND
78
ND
ND
230*
ND
(1986) - -
,-
NDOYE and DREYFUS
S. rostrata
35
ND
0.59
0.62
ND
(1987)
S. sesban
18
:i
ND
0.12
0.13
ND
-
-
SOUGOUFARA et al.
C. equisetifolia
62
45
46
3.8
2.5
2.8
(1989) - -
SANGINGA et al.
Casuarina,
NO
30-80
30-80
ND
NO
ND
(unpublisliZdT-
Allocasuarina
SANGINGA et al.
;: law;z;ephala
ND
6-52
37-72
4-37
21-47
(unpublisl%dr
ND
14-44
14-37
;;
3-9
3-11
GliYiXKa sepium
ND
19-78
28-79
ND
13-40
28-52
ND * not determined
l
= kg/ha/yr
347
l&beck and leucaena indicating the possibility of 15N depleted ammonium sulphate for
measuring N,-fixation and N-allocation in NITS. However. difficulties such as those des-
cribed for the natural abundance methods may limit the usefulness of this method for
measuring N,-fixed by trees.
Acknowledgements
We thank G. Hardarson for his careful reading of the manuscript and his helpful
criticisms and suggestions. We also acknowledge the assistance of M. Tadjbakhsh for
secretarkd services.
References
ALLEN, O.N. and ALLEN, E.K. 1981 The leguminosae. A source book of characteristics,
uses, and nodulation. Madison, University of Wisconsin Press.
AMARGER, N., MARIO-I-II, A., MARIOTTI, F., DURR, J.C., BOURGUIGNON, V and
LAGACI-IERIE, B. 1979 Estimate of symbiotically fixed nitrogen in field-grown soy-
beans using variations in 15N natural abundance. Plant and Soil, 52, 269-280.
BROMFIELD, E.S.P. and AYANABA, A. 1980 The efficacy of soybean inoculation on
acid soil in Tropical Africa. Plant and Soil, 54, 95-106.
CHALK, P.M. 1985 Review. Estimation of N, fixation by isotope dilution : An appraisal
of techniques involving 15N enrichment and their application. Soil Biol. Biochem.
17, 389410.
CORNET F., OTTO, C.. RINAUDO, G., DIEM, H.G. and DOMMERGUES, Y. 1985
Nitrogen fixation by Acacia holosericea grown in field simulating conditions. Acta
Oecologica, 6 (20), 211-218.
DANSO, S.K.A. 1985 Methods of estimating biological nitrogen fmation. In: Biological
Nitrogen Fixation in Africa. PI. Ssali and S.O. Kew. eds.) nn. ??4-?44 Rhivhium
MLRCEN, Nairobi, Kenya.
DANSO, S.K.A. 1986 Estimation of N24ixation by isotope dilution: An appraisal of tech-
niques involving 15N enrichment and their application-Comments. Soil Biol. and Bio-
them. 18, 243-244.
DANSO, S.K.A., HARDARSON. G. and ZAPATA, F 1986 Assessment of dinitrogen
fixation potentials of forage legumes with 15N technique. In : proceedings of a Workshop
on Potentials of Forage Legumes in FarmingSystems of Sub-Saharan Africa, (Haque,
I., Jutzi, S. and Neate, PJ.H. eds.) p. 26-57. ILCA, Addis Ababa.
DOMENACH, A.M. 1987. Estimation de la fixation symbiotique chez des plantes he&a&es
et ligneuses ; utilisation et validite de la m&bode ba& sur la mesure des abondances
isotopiques naturelles de l’azote. These de Doctorat - Lyon I. 189 p.
DOMENACH, A.M., CHALAMET, A. and PACHIANDI, C. 1979 Estimation de la futation
d’azote par le soja a l’aide de deux methcxles d’analyses isotopiques. CR. Acad. Sci.
Ser. D 289, 291-293.
348
DOMMERGUES, Y. 1982 Ensuring effective symbiosis in nitrogen fixing trees. In : Bio-
logical nitrogen fixation technology for tropical agriculture. (Graham, P.H. and Harris,
SC. eds.) p. 395413. CIAT. Cab, Colombia
DOMMERGUES Y. 1987 The role of biological nitrogen fixation in agroforestry. In :
Agroforestry a decade of development. (Steppler, H.A. and Nair, P.K.R. eds.) p. 245-
271 ICRAF - Nairobi, Kenya.
FRIED, M., DANSO, S.K.A. and ZAPATA, F. 1983 The methodology of measurement
of N, fixation by non-legumes as inferred from field experiments with legumes. Can.
J. Microbial. 29, 1053-1062.
GAUTHIER, D.L., DIEM, H.G., DOMMERGUES, Y. and GANRY, F. 1985 Assessment
of N, fixation by Casuurina equisetifolia inoculated with Frankia 0RS021001 using
“N methods. Soil Biol. Biochzem. 17, 375-379.
HARDARSON, G., ZAPATA, F. and DANSO, S.K.A. 1984 Field evaluation of symbiotic
nitrogen fixation by rhizobial strains using 15N methodology. Plant and Soil. 82,369-
375.
HARDARSON, G., DANSO, S.K.A. and ZAPATA 1987. Biological nitrogen fixation
in field crops. In : Handbook of Plant Sience in Agriculture (Christie, B.R. ed.) p.
165-191 CRC Press.
HEICHEL, G.H., HARDARSON, G., BARNES, D.K. and VANCE, C.P. 1984 Dinitrogen
fixation. herbage yield, and rhizobial preference of selected alfalfa clones. Crop Sci.
24, 1093-1097.
H&BERG P. and KVARNSTROM, M. 1982 Nitrogen fixation by the woody legume
Leucuena leucocephalu. Plant Soil 66, 21-28.
KANG, B-T.. WILSON, GF. and SIPKENS, L. 1981 Alley cropping maize (Zea muys
L.) and leucaena (Z.eucuenu leucocephalu (Lam.) de Wit in Southern Nigeria. Plant
and Soil, 63, 165-179.
KANG, B.T., GRIMME, H. and LAWSON, T.T. 1985 Alley cropping sequentially cropped
maize and cowpea with leucaena on sandy soil in Southern Nigeria. Plant and Soil
_,
_
o-)> iv i .-.L / , .
KESSEL, C. VAN and NAKAO, P. 1986 The use of nitrogen-15depleted ammonium
sulfate for estimating nitrogen fixation by leguminous trees. Agronomy J. 78, 549-
51.
KESSEL, C., VAN, NAKAO, P., ROSKOSKI JP. and KEVIN, K. 1987 Ureide production
by N,fixing leguminous trees. Soil Biol. Biochem (in press).
KOHL, D.H. and SHEARER, G.B. 1981 The use of soils lightly enriched 15N to screen
for N2 fixing activity in plants. Plant and Soil 60,, 487-489.
KNOWLES, R. 1981 The measurement of nitrogen fixation. In : Current Perspectives
in Nitrogen Fixation (Gibson, A.H. and Newton, W.E. eds.). pp. 327-333 Australian
Academy of Sciences Canberra, Australia
LEGG, J.O. and SLOGER, C. 1975 A tracer method for determining nitrogen fixation
in field studies. Proc. Int. Conf. Stable Isot. 2nd 1975. pp. 661666.
349
LIJLANDALA, L.L.L. and HALL, J.B., 1988. Leucaena leucocephala’s biological nitro-
gen fixation : a promising substitute for inorganic nitrogen fertilization in agrofo-
restry systems. In : Biotechnology of nitrogen fixation in the tropics (BIGtufT), Pro-
ceedings of UNE!XO Regional Symposium and Workshop, UPM, Malaysia, 25-29
August 1986, p. 259-265.
NAS. 1979. Tropical legumes : Ressorces for the future. National Academy hess Washington
D.C. 331 p.
NDOYE, I. and DREYIFUS, B. 1988 N2 fvtation by Sesbania rosfrutu and Sesbunia sesbun
estimated using 15N and total N difference methods. Soil Biol. und Biochem. 20,209-
213.
RENNIE, R.J. 1982 Quantifying dinitrogen fuation in soybean by 15N isotope dilution:
The question of the non-fixing control plant. Can J. Bot. 60, 856861.
RENNIE, RJ. and RENNIE, A.D. 1983 Techniques for quantifying N2 fixation in asso-
ciation with non-lehumes under field and greenhouse conditions. Can. J. Microbial.
29, 1022-1035.
RENNJE, RJ. and KEMP, G.A. 1983 Dinitrogen fixation in field beans quantified by
15N isotope dilution. I. Effect of strains of Rhizobium phuseoii. Agron. J. 75, 640-
644.
RENNIE, R.J.; RBNNIB, D.A. and FRIED, M. 1978 Concepts of 15N usage in dinitrogen
fixation studies. In : Isotopes in Biological dinitrogen fixation. IAEA, Vienna, Austria,
STI/PUEV478 pp. 107-130.
REDDELL, P.W., BOWEN, G.D. and ROBSON, A.D. 1986 Nodulation of Casuarinaceae
in relation to host species and soil properties. Aust. J. Bot 34, 435-44.
ROSKOSKI JP. 1981 Nodulation and N, fixation by Zngu jinicuil, a woody legume in
coffee plantations. I. Measurements of nodule biomass and field C!& reduction rates.
Plant Soil 59, 201-m.
SANGINGA, N., MULONGOY K. and AYANABA, A. 1985 Effect of inoculation and
mineral nutrients on nodulation and growth of Leucuenu leucocephulu. In : Biological
*,;u”giiL AlLLUII ill AiIILd. \\3kl**, Il. ;IllLI I\\cya,s.u.txs.), p. 4lY-;t30. iwkCM4,P4aIrobI.
SANGINGA, N. 1986 Inoculation of Leucaenu leucocephulu (Lam.) de Wit with Rhizobium
and its nitrogen contribution to a subsequent maize crop. Biol. Agric. Hort. 3, 347-
352.
SANGINGA. N., MULONGOY, K. and AYANABA, A. 1988a Nitrogen of field ino-
culated Leucuena Zeucocephulu (Lam.) de Wit estimated by the nitrogen-15 and the
difference method. Plant and Soil (In Press).
SANGINGA N., DANSO, S.K.A. and BOWEN, G.D. 1988b Nodulation and growth response
of Cusuurinu and Allocusuurinu to P fertilization in an acid soil. Plant und Soil.(Ac-
cepted).
SHEARER, G., KOHL, D.H. and CHIEN, S.H. 1978 The nitrogen-15 abundance in a
wide variety of soils. Soil Sci. Sot. Am. J. 42, 899-902.
SOUGOUFARA B., DUHOUX, E. and DOMMERGUBS, Y.R. 1987 Improvement of
nitrogen fixation by Cusuurinu equisetifoliu through clonal selection. Arid Soil Research
and Rehabilitation 1, 129-132.
350
SOUGOUFARA B., DIEM, H.G. and DOMMERGUES, Y.R. 1988 Response of tield-
grown Cusuarina equisetifoliu to inoculation with Frunkia strain ORS021001 entrap-
ped in alginate beads. Plant and Soil (Accepted).
STEELE, K.W. 1983 Quantitative measurements of nitrogen turnover in pasture systems
with particular reference to the role of lsN. In : Nuclear Techniques in Improving
Pasture Management. IAEA, Vienna, Austria. pp. 17-35.
SWIFT, MJ. 1984 Soil biological processes and tropical soil fertility: a proposal for col-
laborative programme of research. Biol. Internat. Spec. Issue 5, 38 pp.
SWIFT, M.J. 1985 Tropical soil biology and fertility (TSBF). Planning for research. Biol.
Internat. Spec. Issue 9, 24 pp.
TORREY, J.G. 1982 Camurina - Actinorhizal nitrogen-fixing tree of the tropics. In :
Biological Nitrogen Fixation Technology for Tropical-Agriculture (Graham, P.H. and
Harris, S.C. eds.) p. 427-441. CIAT. Cali, Colombia.
TURNER, G.L. and BERGERSEN, F.J. 1983 Natural abundance of N in root nodules
of soybean, Iupin, subterranean clover and luceme. Soil Biol. Biochem. 15, 525-530.
VOSE, P-B., RUSCHEL, A.P. VICTORIA, R.L., SAITO, S.M.T. and MATSUI, E. 1982
15N as a tool in biological nitrogen fEation in soil. In : Biological Nitrogen Fixation
Technology for Tropical Agriculture (Graham, P.H. and HARRIS, S.C. eds.) p. 575
593.
VIRGINIA, R.A., JARRELL, W.M. KOHL, D.H. and SHEARERE, G.B. 1981 Symbiotic
nitrogen fixation in Prosopis (Leguminosae) dominated desert ecosystems. In : Current
perspectives in nitrogen fixation. (Gibson, A.H. and Newton, W.E. eds.) Canberra:
Australian Academy of Science.
WITTY, J.F. 1983 Estimating N-fixation in the field using lSN-labelled fertilizer: some
problems and solutions. Soil Biol. Biochem. 15, 631-639.
YOUNG, A. 1985 The potential of agroforestry as a practical means of sustaining soil
fertility. ICRAF Working Paper 34, 28pp.
ZAHARA, A.R., SHARIFUDDING, H.A.H. and STIBRAMANIAM. R. 19%. Nitrrmn
lmauon by Leucaena leucocephala as measured by N-15 dilution technique. Pertanika
91, 17-22.
CONCLUSIONS ET RECOMMANDATIONS
MaCmiser la FRA pour la production agricole et forestière en Afrique
Les recherches sur la fixation biologique de N,
et leur impact présent et futur sur la production
végétale et forestière en Afkique
DOMMERGUES, Y.,(‘)
BORDELEAU, L.MJ2) et GUEYE, ML3)
(1) : BSSFTKTFT, 4.5 bis Av. de la Belle Gabrielle, Nogent sur Marne, France ;
(2) : Canada Agriculture, Station de Recherche Sainte Foy,
2540 Blvd Hochelaga. GIV 2J3 Canada ;
(3) : MIRCEN-CNRA, BP. 53 Bambey, Sénégal
La troisième Conference de I’AABNF qui s’est tenue a Dakar du 7 au 12 no-
vembre 1988 revêtait, au moins autant que les prt!c&lentes, une importance considérable
pour le développement de l’agriculture africaine. Les experts sont en effet unanimes à
reconnaître que l’Afrique n’a que très faiblement bénéficié de la Révolution Verte en raison
notamment des caractkistiques particulièrement defavorables de beaucoup de ses sols et
de certains de ses climats. Cest seulement en recourant a un matériel veggétal robuste
et frugal, notamment en ce qui concerne l’azote, que l’on pourra rattraper le retard accu-
mulé par rapport a d’autres continents. Il est &ident que l’optimisation de la fixation
de l’azote constitue une voie prometteuse mais encore insuffkmment exploitée pour at-
teindre cet objectif.
Dans cette note finale, on tentera de souligner les domaines dans lesquels les efforts
entrepris devront être poursuivis et intensifiés. Ces domaines concernent plus particulie-
rement la maîtrise de l’inoculation avec les microorganismes fixateurs d’azote, l’kude qua-
litative des effets de l’inoculation et l’&&ration de la fixation de l’azote, les stratégies
d’amelioration des partenaires des symbioses f=atrices d’azote et lIntt5gration des systemes
fixateurs de N, dans les systèmes culturaux.
Outre ces recommandations qui portent sur les programmes de recherche fonda-
mentale ou appliquee, on indiquera les grandes lignes de la politique à suivre pour mener
a bien ces programmes de recherche et assurer le transfert des rbu1tat.s à l’application.
353
Maîtrise de l’inoculation
Identification des caractbristiques des sols
Il est absolument necessaire que, contrairement a ce que l’on observe trop
souvent, l’on fasse pr&%der les essais d’inoculation d’une étude systématique des sols.
Cette étude doit consister essentiellement a déterminer les niveaux de populations de
Rhizobium (ou F~&&I) prksents dans les sols que l’on utilisera. Compte tenu de l’état
actuel de nos connaissances et, sauf cas particulier, si le nombre de Rhizobium com-
pétents dt$asse le seuil de 102 2 IV par g de sol, il ne faut pas s’attendre a une r@onse
significative. & l’inoculation. Cette règle relative aux Rhizobizun ou Frankia s’applique
Cgalement aux champignons mycorhiziens et Cventuellement aux bact&ies fixatrices
d’azote en association. Il importe de mettre au point des méthodes relativement simples
et ne nécessitant aucun appareillage particulier pour determiner approximativement ce
potentiel infectieux des sols.
L’analyse chimique des sols effectuée conjointement devra consister à déter-
miner le pH et si possible en analyser le complexe absorbant, évaluer les teneurs en
C, N, P (assimilable), K et Cventuellement permettre l’identification de carences en
certains f%ments.
S6lectlon des souches
Cette sélection sera effectuée à partir des collections de souches, notamment celles
des h4IRCENs.
Parmi les critères dont il sera nécessaire de tenir compte, le critère de sp6cificité
devra être considéré avec attention. Il est désormais admis par beaucoup d’experts qu’il
vaut mieux developper des souches spkcifiques (c’est-Mire à spectre d’hôte étroit) que
des souches a large spectre.
Conditionnement des inoculums
La tourbe est encoure le support le plus souvent utilise, mais d’autres supports
méritent encore d’être essayés notamment dans tous les pays où il est difficile de se procurer
la tourbe. L’usage des inoculums matriciels - résultant de l’inclusion des microorganismes
dans une matrice de polymère - devrait être développé à l’avenir.
Survie des inoculums
Les inoculants doivent conserver leur pouvoir infectieux le plus longtemps possible
au cours du stockage et même après l’application dans le sol ou sur les graines. Il est
souhaitable qu’une méthode simple puisse permettre, dans un proche avenir, de vkifier
la survie des bactkrks dans l’inocuhun immkdiatement avant l’application de celui-ci au
champ.
Pour répondre à la question de savoir si un inoculum doit être appliqué plusieurs
années de suite, il est nécessaire que l’on mesure, dans différents sols et sous diffkrentes
conditions climatiques, le taux de survie des bactéries apres la ou les premieres appli-
cations d’inoculum.
354
Etude qualitative des effets de l’inoculation et haluation de la fixation
de N,
Effets spkifiques autre que la fixation de N,
Il est certain que l’inoculation d’une plante avec Rhizobium (ou Fmnkiu) peut mo-
difier la distribution de l’azote dans la plante, cet effet étant mesuré par l’indice de récolte ;
il peut y avoir aussi une modification du développement des parties aériennes par rapport
aux racines, cet effet étant mesure par le rapport parties akriennes (shoots) / racines. On
sait aussi que les Rhizobium peuvent, dans certains cas particuliers, réduire l’incidence
de l’attaque des pathoghes.
Il apparaît n6cessaire de developper les recherches de nature phénoménologique
sur ces effets afin de pouvoir en entreprendre ensuite une étude mécanistique, ce qui
implique le developpement de recherches sur la physiologie des symbioses consider6es.
Evaluation de l’azote fixe! par voie biologique dans les hosyst&mes
Il convient de souligner ici qu’avant de choisir la méthode d’estimation à utiliser
(notamment s’il s’agit dune méthode isotopique qui peut être relativement coûteuse) il
est indispensable de définir les objectifs prkis de l’expkimentation. C’est ainsi que la
sélection de clones d’une plante donnée en fonction de son pouvoir fixateur de N, ne
requiert pas obligatoirement le recours a la méthode isotopique.
Il faut être toujours conscient (i) du fait qu’aucune méthode n’est parfaite, (ii) de
la rkcessité de mettre en place une ou plusieurs plantes témoins et (iii) de la nécessité
d’apporter le plus grand soin au marquage du sol (lorsque l’on utilise la methode de di-
lution isotopique).
Les problemes spkifiques aux arbres ont Cd &oqués à plusieurs reprises au cours
de la conference. On ne les &mm&era pas ici, mais il apparaît nkessake d’attirer l’at-
tention sur ces problemes dont certains risquent de passer inaperçus.
Il est bien connu que l’extrapolation au champ d’essais effectues en pot est très
risquée. Toutefois, les n%ultats rapportes suggerent que la culture de plantes dans des
bacs de très grandes dimensions (tanks) pourrait, dans une certaine mesure, se rapprocher
de la culture en plein champ. Il est donc recommandé de faire appel autant que possible
à ces dispositifs expkimentaux.
Stratégies d’am&lioration des partenaires des symbioses fixatrices de N,
Ces stratkgies sont bien COMIKS dans leur principe ; elles portent a la foi sur
la plante hôte et sur le microsymbiote. Jusqu’à pnkent., elles ont Cté fondées essentiel-
lement sur l’emploi de méthodes conventionnelles qui consistent à exploiter la variabilité
naturelle des populations microbiennes ou vegetales. Mais grâce au développement
rkcent de la biologie mol&Aaire, on peut envisager avec optimisme la mise au point
de nouvelles techniques d’amelioration des deux partenaires de la symbiose.
Dans le cas des arbres fixateurs de N,, la stratégie fondke l’expIoitation de la varia-
bilité intraspkcifique des plantes hôtes et la multiplication végétative des individus les
plus performants devrait permettre, dans un tres proche avenir, de mettre à la disposi-
tion des agriculteurs et forestiers un matkriel vegétal à haut rendement et doté d’un
potentiel fixateur de N, considerablement accru.
IrMgration des plantes flxatrices de N, dans les systèmes culturaux
Jusqu’à présent, il est certain que, sauf rares exceptions, peu d’efforts ont éte con-
Sac&s à l’étude de l’int&ation des plantes f=atrices de N, dans les systemes culturaux
qu’il s’agisse de cultures mixtes (multiple cropping), cultures alternées (intercropping),
ou cultures séquentielles (sequential cropping), ou de systèmes agroforestiers. Il apparaît
indispensable : (i) de chercher à identifier et à quantifier les interactions entre les cultures
composant les systèmes et les microorganismes du sol notamment ceux qui intervien-
nent dans le cycle de l’azote ; (ii) d’évaluer avec précision les économies d’azote que
l’intégration des plantes f=atrices de N, permet d’obtenir dans les différents systkmes cul-
turaux ; (iii) de mettre au point les méthodes d’aménagement des differents systèmes de
façon à tirer le profit maximum des apports d’azote provenant de la fixation et de façon
à assurer au mieux la conservation et la rCgén&ation des sols.
Ces différentes investigations se heurteront à des difficultés qu’il faudra tenter de
surmonter, ces difficultks Ctant liées notamment (i) à l’évaluation de la fixation biolo-
gique de N,, (ii) à l’évaluation des transferts d’azote des plantes fixatrices aux plantes
non fixatrices de N,, (iii) aux problemes inhérents aux recherches sur le terrain, re-
cherches qui doivent Ztre conduites dans des conditions simulant aussi parfaitement que
possible celles de la ferme.
Politique de formation, de collaboration et de vulgarisation
A l’unanimité, les s&minaristes sont arrivés a la conclusion que l’effort de formation
entrepris devait être amplifie, étant entendu que cette formation devrait être continue.
Une attention particulière a été accordée à la conception des programmes établis
en collaboration dans le sens Nord-Sud et Sud-Sud, ces programmes devant idéalement
presenter l’un et/ou l’autre des caractires suivants, c’est-à-dire être : (i) focalises sur des
thèmes bien précis ; (ii) liiit6s à des zones géographiques pn5sentant de grandes simi-
laritis sur le plan écologique ; (iii) fondés sur l’exploitation des découvertes récentes
ouvrant la voie a des applications pratiques importantes (ex. découvertes des susbtances
élicites des gènes nod excrétées par la plante hôte) ; (iv) considérés comme prioritaires
dans une région donnée.
La vulgarisation des résultats obtenus au niveau du laboratoire ou de la station
de recherche devrait faire l’objet d’un effort tout particulier. Il s’agit Iii dune activité dif-
férente de celle du chercheur de laboratoire ou de station et à laquelle des chercheurs,
ingenieurs et techniciens sp6cialist% devraient être affectes. Dans le cadre de cet effort
de vulgarisation, il est souhaité que l’on quantifie, partout où ce sera possible, les bénéfices
qui pourront découler, pour le paysan, des nouvelles technologies mises a sa disposition.
Liste des participants
ANGLETERRE
AMIJEE, FEROZ
Wye College London University WYE, KENT
TN25 5AH United Kingdau
ANGOLA
PASSOS TEIXEIRA G. DEIP Biologica Faculdade de Ciêucias Da Uni-
LIDIA
versidade A. Neto Av. 4 de FevereironT3 Luanda
R.P. ANGOLA
C.P. No425 Faculdade de Ciencias Agrarias Uni-
vasidade A. Neto Luanda RP. ANGOLA
MORAS CORDEIRO,
C.P. Np 425 - Faculdade de Ciencias
J. MANUEL
Agrarias Uuiversidade Agostinho Neto
Luanda - RP. ANGOLA
BELGIQUE
DESMADRYL,, DIRK
Laboratoire de Physiologie Végétale Institut
Camoy, Place Croix du Sud, 4, Université
C8thdique&eLwvain,B-l348Lowain-La-Newe,
BELGIQUE
Laboratoire de Physiologie VegCtale Institut
VAN HOVE, CHARJ&S Camoy. Place Croix du Sud, 4, UniversitC
CathdiquedeLouvain,B-1348Lowain-La-Nwve.
BELGIQUE
BRESIL
BODDEY, ROBERT
EMBRAPA - UAPNPBS, Km 47. S&opédica,
MICHAEL
23851 Rio de Janeiro, BRESIL
BURKINA FASO
DIANDA, kfAHAM,%DI lRB~~B.P.7047Guagadougcu, BURKINA FASO
Station de Recherché 2560 Hochelaga, Saiite Foy
CANADA
BORDELEAU, LUCIEN ~utbec GIV 2j3. CANADA
CONGO
PANDZOU, JOSEPH
Centre &Recherches Agronauiques de Loudima,
B.P.28 Lmdima, CONGO
COTE DWOIRE
IUMOU, ANKOMIAN
D6ptemeut d’Agratomie Laboratoire de Micro-
biologie et des Iudustries Agro-alimentaires
ENSAA. B.P.35 ABIDJAN 08
EGYPTE
BADR EL DIN, SAID
NationalResearchCmterDokki.Cairo,EGYPT
KHALAFALLAH, MOHA- National Research CmterDokki, Cairo,EGYPT
MED A.
MOAWAD, HASSAN
National Research Cmter Dokki. Caim. EGYPT
ETATS UNIS
ANGLE, FAY SCOTT
Department of Agronomy University of
D’AMERIQUE
Maryland College Park, Maryland 20742 USA
DODO, HORTENSE
Tuskege University Alabama 626 South PUGH
St # 4 College PA 16801. USA
WEAVER, RICHARD
Dept. Soil and Crop Sciences Texas A CI M
University College Station, Texas USA 77843
FRANCE
BJZJq =FA. mIJA Socitté CALIOPE, 17 Rue SEBASTOPOL
34500 - Béziers - FRANCE
BEUNARD, PIERRE
IRATKIRAD Avenue du Val de Montferrand
B.P.5035 34032 MontpeBier Cedex FRANCE
DOMENACH, ANNE
Université Lyon f, Laboratoire de Biologie des
MARIE
Sols, Bat741, 43 Bld du 11 Novembre 1918,
69622 Vi!.leurbanne Cedex FRANCE
DOMMJZRGUES, YVON
BSSFf/Cl’pT, 45 Bis. Avenue de la Belle
Gabrielle 94736 Nogent/Mame FRANCE
DUHOUX, EMILE
BSSFftCTpT, 45 Bis, Avmue de la Belle
Gabrielle 94736 NogmtlMame FRANCE
FERRET, ROGER
Société CALIOPE, 17 Rue SEBASTOPOL
34500 - BCziers - FRANCE
GALIANA
BSSFT/CFI’. 45 Bis. Avenue de la Belle
Gabrielle 94736 Nogent/Mame FRANCE
MAROC
AURAG, JAMAL
Université Mohamed V. Facultt des Sciences, De-
partement de Biologie, Labo.de Microbiologie -
B.P.1014. Rabat, MAROC
BERRAHO, EL BEICKAY
Université Mohamed V, Faculté des Sciences. Dé-
paxtement de Biologie, Labo.dc Microbiologie -
B.P.1014. Rabat, MAROC
HKLALI, ABDELALI
D+rtemmt des Sci- du Sol, Institut Agm-
nomique et V&kinaire Hassan IIE .P.6202. Rabat,
MAROC
NIGERIA
~ONGOY, WEMANJ
LLT.A., O y o R o a d - P M B 5 3 2 0 Ibadan, NIGERIA
ISICHEI, AUGUSTINE
Depanment of Botany. Obafemi Awolowo
University Ile-Ile, NIGERIA
ODEYEMI, OLO
Department of Microbiology - Obafemi,
Awolowo Univeniry, Ile-Ile. NIGERIA
OUGANDA
San ME (‘MS-
Departmmtof Soil Scienœ.Facultyof Agriculmre
andFomstry.MakerereUniversiry,P.O.BOX 1002,
TINE’
Kampala, OUGANDA
PAYS-BAS
v&q BRU~~, JO~
Uuivaity of Dar Es Splam. Bot=~ Departmfflt.
P.O.BOX 35060, Dar Es Salam. TANZANIA
RWAh’DA
RUGABA, SILAS
Projd Rizicole Butare B.P.76. Butare, RWANDA
HAKIZIMANA, A.
LS.A.R. - RUBONA - B.P. 138 - Butare
RWANDA
SIERRA LEONE
A.MARA, DENIS
soascimœ~mSNyllaUnivnsityCdlege,
PMB. Freetown. SIFRRA LEONE
TAA’ZANIE
MuRuKE, MASOUD
UniversityofDuEaSalam,DepartmmtofBotany,
SALUM
Microbiology Iabarrtory, P.O.BOX 35060. Dar
Es Salam, TANZANIA
FRANCE
OBATON, MICHEL
Laboratoire de Recherches sur les Symbiontes des
Racines INRA, 1 Place Viala 34060 Montpellier Cedex
FRANCE
RINAUDO, GIRARD
Universiti Claude Bernard Lyon 1. Ecologie Micro-
bienne (Bat741) 6%22 Villembme Cedex FRANCE
SALL, KEBA
Laboratoire de Recherches sur les Symbiontes des
Racines INRA, 1 Place Vi& 34060 Montpellier Cedex
FRANCE
GHANA
KALEEM, FEZRHAT
CropsResearchInstituteP.O.BOX55-Educati<mRidge
TamaIe - GHANA
HAITI
FELIX, JEAN FENEL
2290 Rue Ste Foy G, W, A8 Québec, CANADA
JAPON
ASANUMA, SHUICHJ
Hokkaido National Agricultural Experiment Station 1,
Hitsujigaoka, Toyohira-Ko Sappzm, 004 JAPAN
KENYA
Botany Department University of Nairobi, P.O.BOX
mAm’ m*cols
30197 Nairobi . KENYA
ODEE, DAVID W.
Kenya Foresty Research Institue P.O.BOX 20412,
Nairobi. KENYA
ODUOL, PETER AL-
ICRAF. BOX 30677. Nairobi, KENYA
LAN
LJBERIA
KAGABO, WILSON
CentralAgriculmalResearchInstitute.P.O.BOX3929,
EMMANZI
Ma-mvia, LIBE!RIA
MAW
LAHE3IB MESSAOTJD
Ecole Normale Supérieure B.P.241. Bamako, MALI
TRAORE TAHIROU
Bcole Normale Sup&ieure B.P.241, Bamako, MALI
MAROC
ABOUROUH, MOHA-
DivisionRecherchesForestièresB.P.763.Agdel.Rabat,
MED
MAROC
TUNISIE
BAGHDADI, RIVIERE
Ministi?redel’Agticultuulture@irectiatG&tCrale
KHAOULA
de la Production VCgbale). 30, Rue Alain
Savary, Tunis, TUNISIE
ZAIRE
ALrMASI, MWALIBANTU pro ramme National Engrais, B.P.3325.
g
KinshasaKioinbe, ZAIRE
LUYINDULA, NDIKU
Commissariat Gbtral à 1’Energie Atomique,
BP.868. Kinshasa XI, ZAIRE
ZIMBABWE
NYIKA, MICHAEL AN-
Department of Research and Specialist
DREW
Services. Chemistry and Soil Research Ins-
titute,P/BAB37S7,M~~den,Z1MBABWE
AGENCES INTER-
DIARA, HENRI
W.A.R.D.A. Azolla Pmject, ADRAO Project
NATIONALES
Azolla, BP.%, Saint-Lxis. SENEGAL
RIVEROS, FERNAND0
Groupe de Paturage (Grassland Group), FAO,
Via Delle Terme di Caracalla, 00100, Rome,
lTALIA
SCAGLIA, J. ALEXAN-
F.A.O. AGLF. Via Dclle Terme di CaracalIa,
DRE
00100. Rome, lTALIA
WANGARI, ELISABETH
UNESCOMAB, BP.33 11 .Dakar, SENEGAL
BA, AMADOU TIDIANJZ
UNES33 ABN. BP.33 11. Dakar. SENEGAL
DANSO, SETH KOFI
Point FAO/IAEA Division Soil Science
Section, 5, Wagramerstrasse, 100-A 1400.
Vienne, AUSTRIA
SANGINGA, NTERANYA
Soil Science Unit, IAEA Seibersdorf Labo-
ratoty A 2444, Seibersdorf, Vienne, AUS-
TRIA
HAQUE, 1.
ILCA. P.O.5690, Addis Abeba, ETIBOPIA
SENEGAL.
SIAKA SADIO
ISRA/CNRF BP.53 Dakar. SENEGAL
OUSMANE DIAGNE
Mme AMlNATA NIANE
Is-A/BAbIBEY BP.53 Ehbey,
SENEGAL
BABACAR NDAO
”
DEMBA FARBA MBAYE
n
FRANCIS GANRY
”
Mme FATOU GUEYE
h4AMADOU GUEYE
BERNARD DREYFUS
ORSTOM BEL-AIR BP.1386.Dakar.
SENEGAL
BASSJROU SOUGOUFARA
<*
JEAN PAUL COLONNA
MARC DUCOUSSO
MAGGIA LAURENT
PHILsIE’PE DE LAJUDIE
DANIEL THOEN
n
KODJO TOMEKPE
”
PHILIPPE TRAN
I,
Mme MARIE MADELEINE
Université Cheikh Anta DIOP. Dakar,
SPENCER BARRETO
SENEGAL
Mme YAYE KANE GASSA-
MA SEYE
LES PUBLICATIONS DE L’ISRA
Publications non périodiques
“Cahiers d’Information”
Monographies destikes plus particulièrement
aux producteurs agricoles et à leur encadrement
technico-tkonomique immédiat.
“Etudes et Documents”
Monographies éditées a l’intention des cadres du
développement agricole et des chercheurs.
“Actes”
Proposant des comptes rendus des rkunions,
,
colloques et séminaires dans lesquels ont été
impliques des agents de 1’ISRA.
“Réflexions et perspectives”
Monographies sur la politique de la recherche, du
développement, de la liaison entre la recherche et
le développement.. _
Publications périodiques
“Revue Sénégalaise des Recherches Agricoles et Halieutiques” Publication trimestrielle.
“La Lettre d’Information”
Trimestrielle, donnant des renseignements sur la
vie de l’Institution, B l’usage de ses agents et des
utilisateurs situes en amont et en aval de ITSRA.
“Bulletin analytique Documentaire”
Trimestriel, contenant les réferenccs des travaux
publiés par les chercheurs de l’lSRA, qu’il s’agisse
de rapports, de comptes rendus, de mémoires ou
d’articles...
“Bulletin Agroclimatologique”
Paraissant au moment de l’hivernage.
Pour tous renseignements concernant ces diverses publications et en particulier
leur prix de vente, s’adresser a la Direction Gémkale, et à l’attention de 1’ UNIVAL.
S’agissant de la revue, le tableau suivant donne les conditions de cession
4 En FCFA
Abonnement annuel
Numéro séparé
/
Etats d’Afïique
et Madagascar
6000
2 000
Autres uavs
10000
3000
ue n’a que très faiblement bédficie de la
en raison notamment des caractéristiques
défavorables de ses sols et de certains de ses
e de 1’Azot.e (FBA) est une des voies prometteuses
foresterie en Afrique dans le contexte général de l’éco-
n Africaine 5-m la Fixation Biologique de l’Azote
du 7 au l%@ovembre 1988 a particulièrement
dans lesquels les efforts entrepris doivent être poursuivis et
nference comprenait les sessions suivantes :
l
Amélioration de la FBA par la sélection des plantes ;
l
Amélioration de la FBA pour les pratiques culturales ;
*
Inoculation : production d?noculum et essais d’inoculation ;
*
Mesure de la FBA ;
l
FEIA chez les legumineuses à graines ;
l
FBA chez les légumineuses fourragères ;
l
F%A chez les arbres ;
l
Fixation d’Azote associative ;
l
Association AzollaJAnabaena ;
l
Economie de 1’Azote dans les systemes de culture en association et en
l r-._ _. /
l
/ \\ :
i
/
,
I
I
!
--... -,
1
I-L-=,-, -4
Il est reconnu que l’Afrique n’a que très faiblement bénéficié
de la Revolution Verte en raison notamment des caractéris-
tiques particulièrement défavorables de ses sols et de certains
de ses climats.
Maximiser la furation Biologique de l’Azote (FBA) est une des
voies prometteuses pour développer l’agriculture et la foresterie
en Afrique dans le contexte génkal de l’économie de l’azote.
La 3ème conferennce de l’Association Africaine pour la Fixa-
tion Biologique de 1’Azote (AABNF) qui s’est tenue à Dakar
du 7 au 12 novembre 1988 a particulièrement souligné les do-
maines dans lesquels les efforts entrepris doivent être
poursuivis et intensifies. Cette 3ème conférence comprenait les
sessions suivantes :
l Amélioration de la FRA par la sélection des
plantes ;
l Amélioration de la FBA pour les pratiques
culturales ;
l Inoculation : production d’inoculum et essais
d’inoculation ;
l Mesure de la FBA ;
. * u.,
TTn 4 Chi 12s l~gumincuscs à graines ;
l FBA chez les légumineuses fourragères ;
l FBA chez les arbres ;
l Fixation d’Azote associative ;
l Association AzollaJAnabaena ;
l
Economie de 1’Azote dans les systkmes de culture
en association et en agroforesterie.
African
INSTITUT SENEGALAIS
Association for
AABN
DE
Biological
q
I!l RECHERCHES AGRICOLES
Nitrogen
B
Fixation
A C T E S
MAXIMISER LA FIXATION
BIOLOGIQUE DE L’AZOTE
POUR LA PRODUCTION
A G R I C O L E
ET FORESTIÈRE
EN AFRIQUE
IIIème Conférence de I’AARNF
7-12 Novembre 1988, Dakay Sénégal
Mamadou GUEYE
Kalemani MULONGOY
Yvon DOMMERGUES
ISSN 0850-072 X
VOL 2
N 02
1988
.
.
.,.
.
I
ISRA
Institut Si-n+lais de Recherches Agricoles
Document réalis? par
La Direction des Recherches sur les Productions \\‘égi-tbies
Route du Front de Terre
B.P 2057 DAKAR-HANN
SENEGAL
= 32.53.03
.<
...->
I
Fixation Hioiogiqui: .iv I’A;. -1: pur ia l’rod~~lion Agricole el Fo~eslièrc cn
Airique : Papicrs prtSscn:cs a la 3e confkence de I’AAEINF à Dakar,
7-12 Novembre 1988 Collcc:on Actes de I’ISRA. Vol. 2, nQ 2.
0c ISRA, 1 9 9 0
Concrpmn et RCalisaticm LJNIVALISRA
--
. .
--
SOMMAIRE
_- - _
.L
.F,..<
. . .
.,
-.* L‘ :
.G’
‘i;;~k!i
;:,
>
‘
>,’
HI Pages
2..
0_’
Les persptxtives de la fmtion ‘biologique de
GUEYE,M.
1
l’azote en Afrique
FIXATION BIOLOGIQUE DE L’AZOTE
I. .
$,
-j,
.a
:
.“’
CHEZ LES LEGUMINEUSES A GRAINES
ferrallitique~forteme désature en bas;~, C&e
d’ivoire
r
. .
,.._ 1.n
3.,:&.” .a L‘:.‘:: ..;.
‘ - _,’
_ I _,,
,x,. ,. i ‘1 ‘i i, ^ ,-/
‘23
,:.
Influence of rhitibii ir&ula~on’on so$ean
AWONAIKB, K.O., OLUFA-
15
(Glycine max L.) performance mNige% ’
: JQ, 0.0. md mu, 1.~.
i
Nodulahon and growth of bambam groundnut : MULONGOY, K and,
21
(Vigna subterranea) in tree soils.
GOLI, A.E. .
.
Effet de l’inoculation avec des souches delihizo-
GUEYE, M.
34
bium et de la fertilisation azotée sur le rende-
, <
,.: II..> ,.
mentengrainsduVoandzou(Vignasubt&?ànea
(L.) Tbouars) au S&t@al.
’ .’
HXAtiON BId&)GIQ* ~gp&~*&hcA ZI.: *i%iv j($
;
f-t.3 *~~!‘~~:‘~!ii~:~’
fnfji&itffrjf$,
,
CHEZ LES LEGUMINEUSES!:F&JR-
RAGERES
Nitrncrn fïu;llinn nf lhc fornm lcc:llrncs 2nd
r!:\\Qr’E, 1
.;,i*
therr resrdual el’kcts on whcat growth and yield.
FLXATION BIOLOGIQUE DE L’AZOTE
CHEZ LES ARBRES
. . . I
Effect of the addition of phosphoks and&o-
BADJI, S ; THOEN, D. ;
54
bium inoculum on thenodulation and thegrowth DU,CGUSSO, M; and ,
of two species of gum arabic trwAcacia senegal
COLONNA, J.P. i
L. Willd and Acacia luetu R. Brl cx Benth.
:.
Nodulation survey of leguminous and non kgU-'
&J&, p-A. ; AKUNDA,
6 3
minous trees growing at ICRAF’s field station.
E.M. and WAMBUGU, P.N.
Effect of inoculation withe Rhizobium, P appli-
GICHURU, M. and
.
12
cation and liming on early growtb of leucaena
MULONGOY, K. ‘r-
(Lewuena leucocephulu Lam. de wit).
:i..:;
,.
, -+
:
.,:.
A preliminary study on the compatibility bet-
ODEE, D.W.
8 1
ween indigenous rhizobia and some tree legu-
mes in Kenya
FIXATION D’AZOTE ASSOCIATIVE
Effect of inoculation on the growth and yield of
MWAIJRA, F.B. and
92
three maize cultivars.
WIDDOWSON, D.
Population of nitrogen fixing bacteria in sweet
HORTENSE, W.D. ; WAL-
99
potato fibrous roots.
TER,A.H.;MULONGOY.
K. ; ADEYEYE, S.O. and
HAHN, S.K.
ASSOCIATION AZOLLA / ANABAENA
La recherche sur Azolla et ses applications :
VAN HOVE, C. et DIARA,
1 0 6
évolution et tendances actuelles.
H.F.
Phosphorus needs and accumulation potential
DESMADRYL, D. ; GO-
1 1 8
of various Azoh species and strains.
DARD, P. ; WAUTHELET,
M. and VAN HOVE, C.
Sensitivity to Aluminium of various Azolfu _ ,WAUTHELET,,M. ; GO-
1 2 6
species and strains.
‘DARD, P. ; DESMADRYL,
D. and VAN HOVE, C.
INOCULATION :
PRODUCTION D’INOCULUM
.-. L_ .,.>. ,;:, ii’i:\\\\,cLL.;; ;\\,:\\
Production and uses of rhizobium inoculants.
BORDELEAU, L.
1 3 2
Caractkkation et choix de supports pour ino-
BEUNARD, P. et
1 4 4
culum.
SAINT MACARY, H.
Progrès récents dans la technologie des inocu-
DIEM, H.G. ; BEN KHALI-
1 5 3
lums utilises en agriculture et en foresterie.
FA, K. ; NEYRA, M. and
DOMMERGUES, Y.R.
ECONOMIE DE L’AZOTE DANS LES
SYTEMES DE CULTURE EN ASSOCIA-
TION ET EN AGROFORESTERIE
Biological nitrogen fixation and nitrogen trans-
MULONGOY, K. and
1 7 4
fer in multiple cropping in tropical Afiica.
EHUI, S.K.
AMELIORATION DE LA FIXATION
SYMBIOTIQUE DE L’AZOTE PAR LA
SELECTION DES PLANTES
.:
Importance de la plante-hôte et amélioration
DUHOUX, E.
193
génétique de la fixation biologique de N,,chez
les arbres tropicaux.
kralysis of the plant genes involved in the De LAJUDIE, P.
.
ti
207
nitrogen fting symbiosis. Sesbania rostrata as
a model.
.
Sesbania rostrata :
TOMEKPE, K. ; .,
213
un modèle pour la biologie mol6culaire de la
HOLSTERS, M. ;
nodulation
GOETHALS, K. ; VAN
DEN EEDE, G. ;
De LAJUDIE, P. ; TRAN, P.
and DREYFUS, B.
AMELIORATION DE LA FIXATION
SYMBIOTIQUE DE L’AZOTE PAR LES
PRATIQUES CULTURALES
Amélioration de la Fixation Biologique de
GANRY, F.
221
SAiste (NJ par les pratiques culturales.
:
Response of Phaseolus vulgaris L. to inocula-
AMIJEE, F. ; EDJE, 0-T. ;
228
tion with Rhizobium and fertilization with
KOINANGE, E.K. ; BlTA-
nitrogen and phosphorus in Northem Tanzania
NYL, HF. ; BRODRICK,
S.J. and GlLLER, K.E.
The influence of some parameters of soi1 fertili-
LIYA, S.M. ; MULONGOY,
250
ty on early growth of Leucaena leucocephala
K. ; ODU, C.T.I. and
and Cassia siamea
AGBOOLA, A.A.
MESURE DE LA FIXATION D’AZOTE
Evaluation of biological nitrogen fixation in
DANSO, S.K.A.
258
plants.
Influence du déficit hydrique sur la fixation
SALL, K. ; DREVON, J.J. et
212
symbiotique de l’azote atmospherique chez le
OBATON, M.
soja
Effect of ‘5N-labekd mineral-N and strains of
LUYINDULA,
N. and
2 8 8
Bradyrhizobiumon biological nitrogen and
N- WEAVER, R.W.
partitioning in cowpea (Vigo wguiculazu) (L..)
Walp.
Quantifïcation of the contribution of BNF to fïeld
BODDEY, R.M. and
298
grown plants : the use of the rsN isotope dilution
URQUIAGA, S.
technique - problems and some solutions.
Utilisation de la méthode 15N pour estimer la
DOMENACH, A.M. ;
317
fixation symbiotique de l’azote chez les plantes
KURDALI, F. et
herbacées et ligneuses.
BARDIN, R.
Nitrogen fixation in tropical trees : estimations
SAGINGA, N. ; ZAPATA
3 3 7
based on lsN techniques.
and DANSO, S.K.A.
CONCLUSIONS
ET RECOMMANDATIONS
Les recherches sur la fixation biologique de N,
DOMMERGUES, Y. ;
352
et leur impact présent et futur sur la production
BORDELEAU, L.M. et
végétale et forestière en Afrique.
GUEYE, M.
PREFACE
Aprèslesréunions~Nairobien 1984etauCaireen 1986,l’InstitutS~nnégalaisdeRecherches
Agricoles (ISRA) a organisé la troisième confërencede l’Association Afïicaine pour laFixation
Biologique de l’Azote (AABNFQ. Le thème g&kral de cette troisième confkrence était «
Maximiser la Fixation Biologique de l’Azote pour la production agricole et forestikre en
Afrique ».
Maximiser la « Fixation Biologique de 1’Azote » pour la production agricole et foresti&re
est sans nul doute une des voies que les pays en voie de développement, particulièrement
ceux d’Afrique, doivent adopter pour relancer leur production agricole dont la tendance est
2 la décroissance et sauvegarder leur patrimoine forestier exposé dangereusement a la
dégradation.
En effet, dans un environnement à forte propension aux bouleversements kologiques
et aux déséquilibres d’&osyst&mes naturellement fragiles, les pays africains ont beaucoup
de défis à relever parmi lesquels :
0 la sécurisation de la production agricole et la couverture de la demande alimentaire,
l la sauvegarde du milieu naturel.
11 s’agit la d’enjeux de toute importance exigeant la mise en œuvre d’actions correctrices
permanentes liées aux opkations agricoles. Deux actions paraissent prioritaires à l’heure
actuelle pour la sauvegarde des agrosystèmes : le maintien et l’amélioration de la fertilité
des sols (pour la plupart potentiellement pauvres et fragiles) et la lutte contre la désertification
par un effort soutenu de reboisement.
Dans ce cadre, la biotechnologie en général et la fixation biologique de l’azote en particulier
offrent d’excellentes perspectives pour les pays africains. La mise en œuvre effective de cette
technologie nécessite :
* l’isolement et la culture en quantités industrielles de souches de microorganismes
fixateurs d’azote;
l la mise au point de méthodes pratiques et efficaces d’inoculation à la portée du
développement;
fl
‘,:
,’
.:,lj’i/ >Jjb;.:..:‘. ,’ i. ,,. i ::
,hL ;:, ;.:lL’.% L’:
:.i!ii L:< icur aptitude à l
a
nodulation et à la fixation de l’azote;
l
lkolement et l’identification de souches de champignons micorhiziens et la mise en
œuvre de méthodes pratiques d’inoculation facilement utilisables en pépini&e.
L’Association Africaine pour la Fixation Biologique de l’Azote devra aida les pays
africains à atteindre ces objectifs pour l’accroissement du niveau de vie des populations rurales
par la promotion d’opkrations agricoles intégrant l’utilisation des microorganismes fixateurs
d’azote.
Je remercie toutes les institutions nationales et internationales ainsi que la communauté
scientifiquemondialequiontaimablementassisté1’ISRAdansl’organisationdecetteconf~ence
de I’AABNF.
Mouhamudou El Habib LY
Directeur Général de 1’I.S.R.A.
AVANT PROPOS
troisième conférence de l’Association Africaine pour la Fixation
Biologique de l’Azote (AABNF) a et6 organisée ti Dakar, Senégal,
par l’Institut Sénégalais de Recherches Agricoles (ISRA) du 7 au
12 Novembre 1988.
L’ISRA, établissement public chargé de concevoir, orga-
niser et mener à bien tout= la recherches relatives au secteur mal
au Sénégal, bénéficie de la collaboration très étroite de l’Université Cheickh Anta Diop
de DakartJJCAD) et de celle de l’Institut Français de Recherche Scientifique pour le Dé-
veloppement en Coopkration (ORSTOM) au Sénégal.
Plusieurs agences internationales ont contribué financièrement à l’organisation de
la 3éme conférence AABNF :
* l’Organisation des Nations Unies pour I’Education, la Science et la Culture
(UNESCO) à Paris par ses divisions : SER, ABN, ROSTA et MAB.
LUNESCO centralise le réseau mondial des MIRCENs ;
l le Programme des Nations Unies pour l’Environnement à Nairobi ;
l l’Organisation des Nations Unies pour 1’Alimentation et l’Agriculture
(FAO) a Rome ;
l le Programme des Nations pour le Développement (UNDP) à Dakar ;
l le Centre Technique de Cooptkation Agricole (CTA), installé depuis 1983
à Ede/Wageningen au titre de la convention de Lame entre les états du
groupe ACP. Le CTA est à la disposition des états ACP pour leur permettre
un meilleur accés à l’information, à la recherche, à la formation ainsi
qn’aux innovatinns dans Icc w-trur9 de r-b~rln~~cmcnt ?&r-nl- nt nlrqt rf
dc ia vuigmsdtwr ,
* le Centre de Recherche pour le Développement International (CRDI) à
Ottawa ;
l la Fondation Internationale pour la Science (FIS) à Stockholm ;
l YUnion Internationale des Sociétés de Microbiologie (IUMS).
La troisième conférence AABNF a réuni 80 participants venant de : Angleterre,
Angola, Autriche, Belgique, Brésil, Burkina Faso, Canada, Congo, Côte d’ivoire, Egypte,
Etats Unis d’Amérique, Ethiopie, France, Ghana, Haïti, Italie, Kenya, Liberia, Mali,
Maroc, Nigéria, Ouganda, Pays Bas, Rwanda, Sénégal, Sierra Leone, Tanzanie, Tunisie,
Zaïre et Zimbabwé.
Comité d’organisation
Le comité d’organisation de la 3è conférence AABNF comprenait :
BA A.
ORSTOM, Dakar, Sénégal
BA A.T.
UCAD, Dakar, Sénégal
BADJI S.
Eaux et Forêts, Dakar, ShZgal
BARRETO M.M.S. (Mme)
UCAD, Dakar, Sénégal
BORDELEAU L.M.
Canada Agriculture, Quebec, Canada
DIAGNE 0.
ISRA, Dakar, Sénégal
DIARA H.
ADRAO, St-Louis, Sénégal
DOMMERGUES B.
BSSFWCTFI’, Nogent sur Marne
DREYFUS B.
ORSTOM, Dakar, Sénégal
DUHOUX E.
BSSFI’/CTFT, Nogent sur Marne
GANRY F.
ISRA/ClRAD, Bambey, Sénégal
GUEYE M.
ISRA, Bambey, Sénégal
MBAYE D.F.
ISRA, Bambey, Sénégal
NDOYE 1.
ORSTOM, Dakar, Sénégal
SADIO s.
ISRA, Dakar, Sénkgal
SEYE Y.K.G.(Mme)
UCAD, Dakar, Sénégal
SOUGOUbRA B.
Eaux et Forêts, Dakar, Sénégal
MAMADOU NGUER
ISRA/CNRA/Bamtxy BP.53, Bambey,
SENEGAL
ABDOURAHMANE DIOM
.>
KHOUDIA NDIAYE
n
Mme DIOR FALL MBODJ
<I
ROBERT DIOKH
”
MBAYE DIOUF
91
.
r
-,
*
‘,
:...
,
..,
.- .
b!
<
.<.
:+,y;
-3i.P
!X,‘,
‘J’a:#i
0’
4,. ., . ..*
GYYE,
M.
i.1
,,‘;
,“.,
I;,
:;’
‘.
.’
,..
:.
.‘.
i
:,
.’
‘i
‘:
:.
..
,
;:>;,i
~.
;t;Qk
i:;;.
!&
MIRCEhKNRA
BP. 53 Bambey., SknkgaI
:
..,
‘.
-.,
~‘~~:Yy~:,“. I!t:’
!‘<
:”
;j?{.‘: . ..‘ ;\\;II?: !“‘!
,
.;
.\\
‘ ) 2.“‘!:, i ‘. i,
. .
ri’. ,
:.
I
,
,‘$:J . .:y*. >i(>- ,.;,
,.’ ,.. .- “.1 .:.
: : ; .. \\.>
.’
:
..?. :, .<_. . . . . . . .:_ ,..- .~Xfi&Fi[i: m ~XiSI
-.z ,:<‘i
:.-
5 ; i i .i
i,:rf;!“i,,;: of rl: *‘(;:j
1
.
&y$j+;&~*,c
‘&&&j UsJrt i.& .3$x3 d &a’:J : LTm+l$&
.$i$:
)“!.
h &c&&-jm ,&&&.x&&nnfnfl~;& fiútitihhe+j- &?jj&@&$$&’ afri-
cainesidQend .en premier, lieu et en ti2.s @-ande Pa&ie ++@$&@~~~a&
culture, qui sont eux-mêmes handicapés par la pa&reti d& &ls peb I&les,
carencés en CEments majeurs (N, P, K) et en oligdléments (zn, CU). POU~
obtenir des rendements
nrricoler suftkmts dans de tclc SOIS, il est nfccq-
sue de recourir à un emploi massif d’engrais chimiques. Or, les cultivateurs
afïicains, dans leur grande majorité. n’utilisent pas ces engrais parce qu’ils
sont importés, et donc trop coûteux. Ainsi, pour am6liorer la production
agricole en. Afrique, il faudra obligatoirement faire appel~à”d$tio&el&
I.
t&nol@& &,Ja fo~&ficaces @. mat&&&. sp~~~&$&&*b~?
I’introduction-sy~~matique
,des plantes’ fixatnces; ~~t~:~i.~~’ -~-’ . ~
d
agricoles devrak~permettre le maintien et l’an&tior&on dé ~.‘f&&&&$
sols. L’ag&&.ure.itin&ante~en &t un ezémj$e. fis’p~tes&@&$@&~
jouent un..rôle prépqnd&ant dans Ile cadk ‘de. syst&mes de&& &.@&-&$
ou à~.~v&r& la’ n-&&e orgaiq&, ‘&n$, & ’ &&O~S &*~?&Q~$
les cultures en -association ou en couloir.
j:, o”2*z;,’ ” !**
‘S
:
,._
. . .
;
a-
‘.
.‘-
.; (’
t ,
.’
; 2 . . É.‘..:i. +::J:
9,.
,.
I
-I
‘:.
‘s .>.. .,!! ..<1,
2
Importance des plantes fixatrices d’azote dans l’agriculture et la foresterie
,)
,..,.
en Afrique
Les l@wGneuw a graines ont fait l’objet de nombreuses études. Elles occupent
une place pr@ond&ante dans l’agriculture et y jouent un rôle V&S important aussi bien
dans la satisfaction des besoins alimentaires des hommes et des animaux que dans le
maintien de la fertiliti des sols. En effet, grâce à leur capaciti de fmer l’azote atmosphé-
rique en association avec les Rhizubizmz, les légumineuses à graines permettent d’&cono-
miser l’azote du sol. L’utilisation des variétés a haut potentiel fïxatew d’azote dans les
sys&mes culturaux accroit la teneur en azote des sols et en assure la fertilité. D’autre
part, dans le cadre d’une rotation de culture l&umineuse-cMale, au cours des processus
de décomposition, l’azote des racines et des nodules peut être transf&6 a la &Sale suivante.
En Afi-ique~l~e dizaine de/-légumineuses $ gra$es.soFt cultiv&s~~,.$s wf6rentes
zones 6colog1q&%%&&!r&an-i~‘&~ le .&$ (G&ne max) le haric$ &aseol& vuigaris),
.
le ni&5 (Vigna zuzgzziczdata), l’arachide (Arachis hyp@$z&],&&s d&e ‘&%&i&zz&z)
et le pois bambara ou voandzou (vigne subterranea).
Les espèces ligneuses fixatrices d’azote sont appelées elles aussi 2 jouer un rôle
de plus en plus important en Afrique pour plusieurs raisons. D’abord, elles peuvent pousser
sur des sols tr&s pauvres ou totalement dépourvus d’azote fournissant ainsi du bois ou
du fourrage pour l’alimentation du bétail : c’est le cas de.Acacid &da et d’a&es acacias
introduits en Afrique sahélienne ainsi que de Medicago arborea et Acacia cyanophylla
en Afrique du Nord. D’autre part, beaucoup d’espèces ligneuses contribuent à la n5g6nnémtion
des sols en reconstimant le stock de matière organique dans le cadre des cultures intercal-
laires en sys&~es agroforestier (Leucaena leucocephala, Glirictia sepizmz) ou dans le
cas de la rizicula (Sesbania rostrata). Les esp&es ligneuses contribuent également &
la fixation d&dunes : c’est le cas du ftio (Caszzarina eqzzisetifolia).
., ,,I~;~Jw&I<&~c; &J%tape des “ii%herche?actiellésd& t&&lê fnhimum de profit
‘&& ‘~nnaiss&&~,~quises pour, V&iser au inieux la fiXarion$biologique de l’azote
dans le but d’améliorer la production agricole et forestière en Afrique.
Amélioratiori de la fixation biologique de l’azote par action sur les micro-
organismes fixateurs d’azote
<
..-., ‘.
La pren$&déanarche pour augmenter la fixation biologique de l’azote par action
.
sur les microorg+n@es ,COIXXW.~ .s&.ctionner .judicieusement pour chaque espèce de
légumin&$ ~@-e~L~~c~aque~vari&5 .de l@mineu& ftitrice ‘d’&t&‘u6e‘Ou des souches
“’ ‘i’
de Rhzrobizm c.apal$ewde@~.@& effectivement -l’azote~atmosph&ique4 la suite de leur
in?aUation &r%. p&&-hôte, considérée. Le schéma de s&ction a ét6 4r&s bien décrit
k b@&$$%.&& ‘@&Cent, 1970). Il repose sûr ‘la connaissance de critères bien
précis : (ij infect&&& :&iffecGvit6 dans la fixation biologique de l’azote ; (ii).comp&itivité
pourlafo&&~&&&&&.set(i&) adaptationauxconditionsdel’enviroimement:temp&ature,
.) ,;: .J
pH, concentration en ‘~1, &pe de sol, taux d’azote ~$/OU de matière organique du sol,
antagonistes (certains actinomycètes...).
L’inoculation des plantes fnatrices d’azote avec des souches sélection&s de mi-
croorganismessymbiodquesagénéralementdonnédebonsrésultatsdanslecasdesl~gumineu~s
suivantes : soja @mon, 1976). haricot (Keya et aL, 1982), luzerne (Medicago sativa),
(Bordeleau et al., 1977). stylosanthes (Stylosanthes guineansis) leucaena (Skerman, 1982)
4
‘;‘. : : Ii; cL;$ii “ii’I
..<#Y? ‘;-.:lt x ..‘ial ‘.f ,“,‘.-.
Actuellement, on commence & &r&ppel-‘aux techmques de .g&ie ‘g&néhqtie fi&
cn5e.r des souches ~pluQerformhntes parles ~proceksus de transformation, de traduction
et.de conjugair8n-b~ac~~~~éntiLiue l’information g&&ique qui déter-
mine la fixation biologiquei~W~ le!~%k&ies”& cO&5iue dans une séquence
de. genes~d6nornmés N&dorlt la~ta;‘strucmre:et le ~fonc%ionne&nt kont décrits avec précision
(Sasson, 1984). hs gènes Nif de Klebsiellaqneumoniae
ont tu5 transf6res avec succès
très peu effective et n
Am6lioration de la fixation biologique de l’azote par la sblectlon des plantes
RL!i‘I?i‘!OI!~ ijlii: IA i)‘JllbliISC:
2SL L)X Ci:I’,:a:Li<,J, “’
u18~ d!ovCid~JOi;l Liiilb L1CU.i ~);LLi~lI;liIi~
pour un bénéfice reciproque. Il est donc logique de chercher a maximiser la fixation biolo-
gique de l’azote en agikanf’non ‘s&lement sur le-partenaire bactérien, mais aussi en agis-
a-: Q3
A h fm de h ‘d&$&&‘~~@i), .uùe.nouvell~~~~no~ogie s’est &,&pp& :‘i;l~~git
de la multiplication vég&a&ve in vitro qui permet une exploitation raison& de la’!&&+
bilit6 intraspkifique chez les ‘phuites. Dans le cas des esp&es fixat.rices d-te; cette
technologie permet d’accroître substantiellement la quantité d’tite fixé ‘en permettant
la multiplication desr clones iden$fï~ comme les plus, performants.
‘, ‘; ,,&) J1::~,
Cette démarche’peut’facikment s~a@iqi.$r’&x:ri.rb~~, fiiateu&d’&ote’&r~i@ pt-&
sentent une variabilité très importante et leur densite a l’hectare est comprise entre 2ooi)
4
et 5000 plants. En revanche, on ne peut envisager de recourir à cette méthodologie dans
le cas des plantes furatrices d’azote annuelles dont la densité est de l’ordre de 100.000
plants à l’hectare (Dommergues, 1987).
Les centres de ressources microbiologiques (MIRCENs)
L’amélioration de la fixation biologique de l’azote ktant une priorité pour les pays
africains, puisqu’elle contribue à l’autosuffisance alimentaire avec un minimum d’engrais
azotes, il y a lieu de collecter et de préserver les microorganismes fixateurs d’azote capables
de s’associer de faç~~~esfective
avec les plantes amtliorées. C’est la un des rôles majeurs
des h4IRCENs. Le rôle de ces MlRCENs est de :
fournir l’infrastructure d’un réseau mondial de laboratoires travaillant en
l
collaboration au niveau régional et interr6gional à la distribution et ci l’uti-
lisation des pools de gènes des microorganismes ;
renforcer les efforts relatifs à la préservation des microorganismes d’intérêt
l
économique ;
promouvoir l’application de la microbiologie a l’agriculture ;
l
servir dé, centre de formation pour la diifusion des connaissances sur la
l
1 microbiologie.
Actuellement, il y a trois MlRCENs en Afrique. Le MIRCEN de Nairobi avec
une collection de 500 souches de Rhizobium, produit de l’inoculum sur un support de
boue fïltréc pour la culture du haricot. Le MlRCEN du Caire renferme une collection
de 200 souches de Rhizobjum et conduit plusieurs experiences sur l’inoculation de la
fève (Viciafubu). La‘collection du MIRCEN de Bambey renferme 300 souches de Rhiwbium
pour les exp&i~,~~~s@ !a fi@ion biologique de l’azote chez l’arachide, le niébé, le
;)‘:: .,;<
.’
<
V~&&U, ‘ie soja et lis &a&&
L’association africaine pour la fixation biologique de l’azote (AABNF), les
institutions internationales
Dans la plupart des pays ahcains, existe une institution nationale, privée ou semi-
privée, travaillant sur un aspect de la fixation biologique de l’azote. Cependant, ces institutions
nationales ont des moyens très limites qu’il conviendrait de rassembler sous la forme
d’un réseau. C’est encore là un des objectifs des MIRCENs et par delà un objectif de
I’AABNF. II n’est.donc pas étonnant que les MIFKENs d’Afrique aient soutenu et supporté
les premières activk% de I’AABNF. A côté des institutions nationales, il y a les institu-
tions internationales.
L’impact dè l’International Institut of Tropical Agriculture @TA) de Ibadan, Nigeria
dans les programmes de recherches sous régionaux n’est plus à démontrer : depuis les
recherches sur le ni&6 en particulier jusqu’aux r6cents travaux sur l’exploitation des arbres
fixateurs d’azote dans les systemes de culture en couloir. L’International Council of
Research in AgroForestry (ICRAF) a démontre, dans les régions au Sud du Sahara, l’impor-
tance des arbres fixateurs d’azote et leur intégration dans les systèmes agroforestiers.
Il y a tgalemnt les nombreuses agences internationales oeuvrant pour le développe-
ment de l’agriculture en Afrique a travers des programmes axes sur la fixation biologique
de l’azote : CRDI, CTA, FAO, FIS, NAS-BOSTID, UNDP, UNEP, UNESCO.
5
11 apparait donc indispensable d’&.aIuer dès a présent, le potentiel humain dont
dispose actuellement l’Afrique pour mener à bien les recherches envisagées. C’est pourquoi,
PAABNF, par l’intermediaire de I’IITA à Ibadan, dresse un annuaire des microbiolo-
gistes en Afrique travaillant dans le domaine de la fixation biologique de l’azote. Il faut
maintenant raisonnablement mesurer ce qui reste a faire pour que tous les pays africains
puissent utiliser avec le maximum de profit toute la technologie de la fixation biologique
de l’azote. Cet objectif requiert au prt%able la formation d’un personnel qualifié et une
plus large diffusion des connaissances acquises sur la fixation biologique de l’azote en
Afrique.
Rf$f&ences
BORDELEAU, L.M., ANTOUN, H. et LACHANCHE, R.A. 1977 Effets des souches
de Rhizobim meliloti et des coupes successives de la luzerne (Medicago sativa) sur
la fixation symbiotique de l’azote. Can. .J. Plant Sci. 57, 433-439.
BURTON, J. 1976 Problems in obtaining adequate inoculation of soybcans. In: World
soybean research. Pmceedmgs of tlte world soybean research conference. Ed. by Lowell
D. Hill. The Interstate’Printers and Publishers, Inc. Danville, Illinois p. 170-179.DIEM,
H.G., GAUTHIER, D. et DGMMERGUES, Y. 1983 An effective strain of Frankia
from Casuarina sp. Can. J. Bot. 61, 2815-2821.
DOMMERGUES, Y. 1987 Comment accroître la fixation symbiotique de l’azote par les
arbres en milieu tropical. In : Actes des semina& sur les arbres fixateurs d’azote
et l’amélioration biologique de la fertilité du sol. Editions de l’ORSTOM, Paris p.
18-27.
GUEYE, M. and BORDELEAU, L.M 1988 Nitrogen fixation in bambara groundnut,
Voandzeia subterranea (L.) Thouars. Mircen J. 4, 365-375.
KEYA, SO., BALASUNDARAM, V.R., SSALI, H. et MUGANE, C. 1982 Multiloca-
tional field responses of Phase&s vulgaris to inoculation in eastem Africa. In :
BNF Technology for tropical agriculture. Papers presented at a workshop held at
fT,AT \\fnrch 4.12 JO‘?j C~>~~ITY PM, y!,.1 TI,.....' c p '- ' r, 1. .r' ' : ,!
,
23 l-234.
PGSTAGATE, J. 1987 Nitrogen fixation, 2nd edition. Edward Amold (Publishers) Lui,
London 73~.
SASSON, A. 1984 Biotechnologies : challenges and promises. Eds. UNESCO, Paris. P.
109-155.
SKERMAN, PJ. 1982 Les légumineuses fourrageres tropicales. Collection FAO produc-
tions veggétales et protection des plantes. Rome. 666 p.
VINCENT, J.M. 1970 A manual for the practical study of root nodule bacteria, IBP
n* 15. Blackwell Scient& Publications, Oxford.
FIXATION BIOLOGIQUE DE L’AZOTE
CHEZ LES LEGUMINEUSES A GRAINES
Maximiser la FBA pour la Production Agricole et Forestière en Afrique
x-.
..; I 1
,r;i.‘, t’ :,<y ,,.. g*;&; .::;.
_,
b
‘.
:‘,‘T
.,
.f,6..Lu1$e-.h’ d;*$*ja Gly&i& m& ‘&) fie,*i , ” j_<
et du niéb& Vigna unguiculata (L.) Walp.)
sur sol feryallitique fortement désaturé
..a .’
.,..<.:..
.en .basse&~te &Ivoire.
. . .
L$xmztoire de Microbiologie et des Industries Agro-alimentaires. EhTA, OS BP 35 Abidjan
08,Côtcd’Ivoùe.
,.‘..
Des essais d’inoculation du soja (Glycine max) et du ni6bk (Vigna wzguiculata) ont
td conduits en serre et en plein champ sur un sol ferrallitique fortement désature
en .Mai, .Juin et Juillet 1987. Les résultats obtenus indiquent d’une part qu’en sol
.I
.;!
;, 1 - &’ &&:(PH 4,8),‘3 ~t+po&iblë $a&&&& &&e &&&&a~~*&&~s
&&&a-
tions symbiotiques (L&rmineuses-Rhizobium)
bien adaptées aux conditions extrêmes
de culture. D’autre part, l’inoculation du soja et du nitW peut être bénéfique en
Côte d’ivoire. du moins dans la @on où l’expérience a Cte conduite.
La nutrition arot& des légumineuses est un important facteur limitant du rendement
I
cn graiilzj du SOJ;I ((J/JC& mm) CL du nlcbk (l-iztid bryucuidld) CI dc leur I”;L.
en tant que précédent cultural. or, cette nutrition azotée est assurée par deux voies :
l’assimilation de l’azote combiné du sol ou des engrais azotes et la fixation biolo-
gique de l’azote atmosphérique. Ces deux voies sont influenckes différemment par
les facteurs de l’environnement si bien qu’elles peuvent être concurrentes ou
compl6mentaires suivant les stades physiologiques de la plante ou les conditions du
milieu (Decau et al.. 1975 ; Obaton et Kimou. 1985).
La modification & l’dquilibre entre les contributions de ces deux sources peut avoir
d’imporumtes implications agronomiques et économiques. En effet, des gains sub-
stentielsdefixatonetderendementsontpossibleschezcertainese~del~gumineuses
par inoculation avec des souches trés efficientes. Mais l’eflïcacité de cette inoculation
peut être compromise par des conditions pkloclimatiques
inadéquates. Dans le présent
travail, nous avons cherché à estimer l’influence des
modalitts de fourniture d’azote
par le sol ou par la fumure sur la participation relative de l’une ou l’autre source
et à en évaluer les cons6quences agronomiques en sol fcrrallitique fortement désa-
tur6 très typique des zones forestiks de la basse Côte d’ivoire.
,.
.,,..,
~)
.._
_>.
I
,.,-<. ..-.,.
.
.
.
-,..-
c.
,I
MatWels et mbthodes
Le niébé cv. V, et le soja cv. Tropical ont été cultivds parallèlement en champ
et en serre en pr&ence de 0 et 90 kg N/ha. L’azote est appport& sous forme d’ur&.
Pour chaque type de IQnrninenses. les ,@nes ont ét.6 s@il&+ parJrempage pendant
30 mn dans une solution d’hypochlofité- dè calcium (66 g/l) püi ,$@@mrnent-rinu&
à l’eau distillk stirile et ensuite répar& eti deux lots. Un lot ,#$noc& avec une suspen-
sion de Rhizobium japonicum G49 @WA-Franc@ paw le ,yJa et,(e:R@zobium tropical
N7 (ENSA-Côte d’ivoire) pour le &~‘~~‘i&culation ,a &té ‘rt?aWë8 une dose forte
(lOg Rhizobium par graine) juste avant le semis; L’ati&e’lot SI~ @nec est Seme sans
inoculation.
L’expérience au champ a été conduite sur la parcelle exp&imentale de 1’Ecole
Nationale Agronomique d’Abidjan en une seule saison culturale de Mai a Juillet 1987.
Cette période correspond ZI la grande saison detipluies&&ti
&?&hüre en Côte’d’Ivoire.
Les sols sont du type sablo-argileux contc+mt en moyenne .2;84JWde matière organique
et 1080 ppm d’azote total ; 17,s ppm de phosphore assimilable ; complexe absorbant
25,ll % ; Mg : 0,23 meq/lOO g ; Ca : 0,84 me@00 g ; K : 0,05 me@00 g et pH
eau 4,8. Le site chosi pour l’exp&ience est relativement homog&ne et n’a jamais porté
de culture de soja et de ni&.
En serre, les plantes ont été cuItivées en pots de 5 1 sur des échantillons de sol
provenant de la parcelle d’expérience au champ. La serre est un abri avec un toit en
verre transparent qui n’arrête que les eaus de pluies. I-es murs sont constitués de grillage
de 0,5 cm de maille. Ainsi les plantes en serre bénéficient des mêmes conditions d’&zlaire-
ment naturel (durée d’ensoleillement var&nt entre 10 et 12 heures), de’temp&ature (tempéra-
ture ,dii$;,: .g,O. à, ,30.?&; ,~~ra&n&wcmme S?4-&26930)&~&fiuhUmiditi (humidité
relative pendant la p&iode de.culIturec95 WXl~:%): quec&ItWSpletichtip.
Les plantes
sont arros&s au fur et à mesure des -besoins, en solution non coulante à l’eau distill&.
En pleine floraison, 45 jours après le semis pour le niébt! et 60 jours après le
semis pour le soja, quelques plantes sont déterrées pour le dénombrement des nodosités
formées. La partie aérienne de chaque plante est sectionnée au noeud cotylédonnaire et
i I i .;; Il i>L!:~i ;i . L...,<’ .i e;
.
I
L jLiicid:ii -3~1 11. LC +ds dz. lidikc: sAc: JC: ia p.ut~c
aérienne est déterminé par pesée. Les rendements en graines sont mesurés après la récolte
?+ la maturité et après séchage au soleil jusqu’à 15 % d’humidité.
Le dispositif expérimental utilisé en champ et en serre est en blocs de Fisher entière-
ment randomisés. Il comporte 4 traitements .(t&noin absol& témoin! avec inoculation,
90 kg N/ha et 90 kg N/ba avec inoculation.
_.
;.
-,
Rkwltats et discussion
Les &ultats de l’inoculation sont indiqués dans les Tableaux 1 et 2 respective-
ment pour le soja et pour le Niét&
Inoculation du soja - Les effets de l’inoculation du soja, cv. Tropical avec
la souche G49 sont très nets. Au champ, en l’absence de fumure azotée, l’inoculation
permet de passer de.O, t/ha (témoin absolu) à 2.62 t/ba soit un accroissement du rendement
en graines de 23 % par rapport au témoin non fumC et non inocul& (Tableau 1A). En
l’absence d’inoculation, l’apport de 90 kg N/ha augmente le rendement en graines, d’environ
9
67 % mais ne provoque aucune différence significative sur la production de matière séche
des parties aériennes. En présence d’inoculation, l’apport de 90 kg N/ha augmente
significativement la production de matière sèche des parties aériennes de 13.9 % et provoquent
dans le même temps une diminution de 21 96 du rendement en. gra$tes~ Cette baisse
du rendement est due à un avortement elevé:des fleurs snite à un developpement vég&atif
excessif des plantes suivi de leur verse. Ces difft%ntes observations faites au champ sont
confirmées par les resultats en serre (Tableau 1B).
Tableau 1A : Essai d'inoculation de Glycine max (cv. Tropical) avec
l'inoculum 649 (INRA-France) : résultats au champ.
Nombre de no-
Poids de ma-
Rendement en
Traitements
dosités par
tiére sèche
graines
plante
t/ha.
t/ha
Témoin non
inoculé
0 kg N/ha
O b
5,69 c
0,78 c
90 kg N/ha
O b
5,92 c
1.30 b
Inoculé
0 kg/ha
72 a
7,97 b
2,62 a
90 kg/ha
69 a
9,08 a
2,16 a
Dans chaque colonne les valeurs suivies de la m@me lettre ne dif-
fèrent pas au seuil de 5 % (méthode Newman-Keuls)
10
Tropical)avec‘
z.
T'inoculum G49.SINRd-France) : résultats en serre
P,
. . . .
Nombre de no-
Poids de ma-
Rendement en
Traitements
dosités par
tiëre sèche
graines
plante
gfplante
g/plante
Témoin non
inoculé
0 kgfha
O b
18,28 d
3,57 d
9 0 kg/ha
O b
24,32 c
8,90 c
Inoculé
0 kg/ha
5 0 a
31,14
b
15,04 a
9 0 kg/ha
5 6 a
41,68 a
Il,83 b
- 90 kg N/ha correspond à 33 mg N/lOO g du sol utilisé en serre
- Chaque valeur est une moyenne sur 16 plantes (4 plantes x 4 répëti-
tionsl
Inoculation du niébé - Les observations faites sur l’inoculation du soja, cv.
Tropical sont identiques à celles faites sur l’inoculation du niébé cv. V,. Cependant, l’effet
de cette inoculation semble moins spectaculaire sur le niébé comparé aux résultats obtenus
sur le soja. Au champ, en l’absence de fumure azotée, l’inoculation augmente le rendement
en graines de 32,75 % par rapport au témoin non inoculé (Tableau 2A). Pour les autres
traitements, aucune différence significative n’est observée. La différence de résultats observée
entre le soja et le niébé pourrait être justifiée par le fait que le niébé non inoculé porte
des nodosités qui possèdent une certaine aptitude à furer l’azote atmosphére (section rouge
des nodosités). Un traitement témoin non nodulé aurait été intéressant pour infirmer ou
confirmer cette hypothèse.
_
*
,
..,
;.~,.ti’;,’ -..
.
.
.
..:,
‘.‘.
:
_.
Tableau 2A : Essai d'inoculaiion de Vignatunguiculata (Vl) avec
.,.<'i,'.. 1 ../-
l'inoculum'local N 7 (ENSA-Côte-d'Ivoire) : résultats au
champ.
-Nombre de nodo-
Poids de ma-
Rendement en
Traitements
sités par
tière sèche
graines
plante
t/ha
t/ha
Témoin non
!: :>.:
<. .'
i n o c u l é
..-.;,F'.
0 kg N/ha
8 b
6,OO b
0,76 b
9 0 k g N/ha
7 b
6,65 a b
0,85 a b
.nnr,:lé
0 k g N/ha
20 a
5,86 b
1,OOla
9 0 k g N/ha
1 5 ab
7,08 a
0,86 ab
Dans chaque colonne les valeurs suivies de la mëme lettre ne diffèrent
pas au seuil de 5 4 (Méthode Newman-Keuls).
I
,
12
Tableau 28 : Essai d'inoculation de Vigna unguiculata (Vl) avec l'ino-
culum local N7 (ENSA-Côte-d'Ivoire) : Résultats en serre
Nombre de nodo-
Poids de ma-
Rendement en
Traitements
sités par
tiëre g/par
graines
plante
plante
g/pl ante
Témoin non
inoculé
0 kg N/ha
8 b
23,08 b
3,75 c
90 kg N/ha
8 b
3,95 c
Inoculé
90 kg N/ha
27 a
24,90 a
4,99 b
- 90 N/ha correspond â 3 mg N/lOO g du sol utilisé en serre
- Chaque valeur est une moyenne sur 16 plantes (4 plantes x 4 répéti-
tions.
- Dans chaque colonne, les valeurs suivies de la mëme lettre ne dif-
fèrent pas au seuil de 5 49 (Méthode Newman-Keulsl.
13
Conclusion
De ce travail deux points essentiels semblent retenir particulièrement notre atten-
tion :
en sol tr?s acide, il est possible d’avoir une bonne nodulation avec certaines
l
associations symbiotiques «légumineuse-Rhizobium» bien adaptées aux
conditions extrêmes.
9 l’expérience conduite sur le domaine exp&imental de 1’Ecole Nationale
Sup&ieure Agronomique d’Abidjan montre que l’inoculation du Soja cv.
Tropical et du Ni&6 cv. V, peut être bénéfique. Bien que l’inoculation
du niébé soit rarement pratiquée dans le monde, cette augmentation de
rendement d’environ 33 % dans le sol présentant de telles caractkis-
tiques nécessiterait d’effectuer des expérimentations analogues en d’autres
points du pays. La confirmation de tel effet bénéfique pourrait conduire
à une application pratique (inoculation du soja et du niébé) en milieu
Pw=n.
Remerciements
Ce travail n’aurait certainement pas eu lieu sans l’aide financière de la Communauté
Economique Européenne (Contrat CEE 02.087/TSD-A-180) et la franche collaboration
des chercheurs de 1’INRA de Montpellier en particulier M. Obaton et J.J. Drevon. Qu’ils
rqoivent ainsi que tous les membres de la CEE nos sincères remerciements.
RQft$rences
DECAU, M-J., BOUNIOLS, A., LENCREROT, P. et PUECH, J. 1975 : Concurrence
ou complémentarité de l’alimentation azotée non symbiotique et de la fixation bactérien-
ne chez le Soja (Glycine max. L. Merr.). CR. Acad. Sci.. Paris 281, 535-538.
<.,_
..\\.....~.. z< LL .i...IDC
nase au cours du cycle végétatif chez le Soja. In : Nutrition azotée des légumineuses,
Versailles, 19-21. Ed. INRA, Paris, 1987 (les colloques de I’INRA, N* 37).
Maximiser la FBA pour la Production Agricole et ForestZre en Afrique
I,
t.;.:::;’
‘.
.;.
f.j,d
.
I.
.i,.
. . ;
I, ,,d
.1,:
;
:
,.
‘.,I.
,
.
.:
.
,.
-(
.
.
1
.
:.
In.flu&nce of rhiz6bia inoculation
on soybean -(GZ~cine .max L.)
perfbrmtitice’ ‘iti : NigeGa
AWONAIKE, tiO.,“’ .~LUFAJO, 0.0.‘2’
and. Ah,, J.K.(2)
: htitutc of Agricdhual Research and Training Obrgemi Awolowo University
PMB. 5029, Ibadan. Nigeria;
(2):. Institute for Agricultural Research Ahmadu Belle University Samaru, tiia, Nigeria
’
The response of soybean (Glycine max L.) to inoculation with exotic and indigenous
rhizobia (no inoculation) and inorganic nitrogen fertilization was studied at Ilorin
and Samaru, Nigeria, during the 1985 cropping season In the two locations. nodula-
tion was enhanced by all inoculants. while inorganic nitrogen fertilkation significan-
tly enhanced vegetative dry mattex production. Location effect was observed for the
respmse of soybean to inoculation with regard to seed yield. While in Ilorin, grain
yield was not enhanced by inoculation, in Samaru, only strain IRj 2133 significan-
tly’ influenced the seed yield of some soybean cuhivars.
j
A generaliiation 04 the need or otherwise to inoculate soybean for enhanced perfor-
mm% in Nigeria is not possible from this study. However, a more justifiable conclu-
sion may be reached if other parameters, such as amount of nitrogen fixed and the
proportion of it which is available to a subsequent grown cmp, are assayed.
16
lntroductlon
The potential contribution of soybean to the improved nutrition of most Nigerians
that can ill-afford animal proteins cannot be overemphasized. Although it has been culti-
vated for sometime in traditionally growing areas of Nigeria, its wide national acceptance
and cultivation is relatively recent.
Soybeans derive N directly from the soil as inorganic residuals or as mineral&d
organic soil fractions and indirectly by the symbiotic nodule relationship where the bacte-
rium Bradyrhizubim juponicwn fix N present in the atmosphere. Harper (1974). confirmed
that both symbiotically fixed and inorganic N are required for highest yields of soybean,
while estimates of the fared N to the total N range from 25 to 60% (Vest et al., 1973).
Thus, yield increases from N fertilization would be expected to be relatively common.
Yet only a few reports of measured yield increases from N fert&ation are available
in the literature (Bhangoo and Albritton, 1972 ; johnson and Hume, 1972) and these
increases were often very small (Sorrensen and Penas, 1978). Somme other investigations
however found no effect of N fertilization on soybean field (Wagher, 1962, Beard and
Hoover, 1971).
Ayanaba (1977) reported that the tropical soils are devoid of indigenous B. japo+un,
although the cowpea type Rhizobizm strain (which is prevalent in these soils) readily
form effective nodules on the roots of soybean varieties of Asian origin (Nangju. 1980).
It is generally believed that when soybean is cultivated in a field with no previous history
of soybean cultivation, artificial inoculation with B.japonicum
is essential to ensure high
seed and protein production (Graham, 1985). Although several investigators have reported
beneficial effect of inoculation on soybean (Rang, 1975 ; Pal and Saxena, 1975 ; Chowdhury,
1977 ; Singh and Tilak, 1977), others have reported inoculation failnres (Salema and
Chowdhtuy, 1980 ; Awai, 1981 ; Awonaike, 1988).
Addition of combined nitrogen and inoculation have been shown to respectively
enhance vegetative matter yield and nodulation in some grain legumes (Ayanaba and Nangju,
1973). Their combined effect on seed yield of some tropical grain legumes under field
conditions is inconclusive (Ofori, 1973 ; Luse er al., 1975 ; Pal et al., 1985).
The objectives of this study were therefore to ascertain the need or otherwise to
inoculate soybean for enhanced performance in Nigeria and if necessary, to identify strains
of B. juponicum that are adapted to the environmental conditions.
Materials and methods
,
Field experiments were conducted on newly cleared land (without a history of
soybean cultivation) in 1985 at the Ballah substation of the Institute of Agricultural Research
and Training in Ilorin (8”3O”N, 4O35’E) and on the Institute for Agricnltural Research’s
experimental farm at Samatu, Nigeria (ll”ll”NN, 7”38’E). The soil at Ilorin belongs to
the Apomu series and that of Samaru is a well-drained, leached ferrogenous tropical sandy
loam.
(i) : Treatments of the llorin trial - Eleven soybean cuhivars were subjected
to three nitrogen nutrition regimes (viz., N, : uninoculated control treatment without nitrogen
fertilization ; N2 : 150 kg urea per ha provided in three split applications at 25 days
interval commencing at planting ; N3 : inoculation with a mixture of B. juponicum strains
IRj 2114 and IRj 2133). A randomized complete block design incorporating four replicates
17
of the 33 treatments was used. A basal application of 400 kg single superphosphate
pe.r hectare was applied before planting because of the low P status of the soil (3.2 ppm
p>-
(ii) : Treatments of the Samaru trial - Treatments consisted of five soybean
cultivars-grown with four single IRj strain peat inoctits (2114,2123. 2133 and 2144)
and a mixture containing equal proportions of the four strains. An uninoculated control
treatment without nitrogen fertilizer and another uninoculated treatment with 100 kg
N/ha applied as calcium ammonium nitrate (CAN) (26 96 N) were also included. A randomi-
ml complete block design incorporating three nql@es of each of the 35 treatments
I
i.‘.
was used.
Each plot size in the both trials was 6 m by 2.25 m made up of four rows plants,
75 cm between the rows and ‘5 cm within the row.
Symbiqtic effectiveness was evaluated at 9 weeks after planting (WAP) by asses-
sing fresh weight of nodules produced and dry weight of the shoots (Ilorin trial) and
tie nodules number and dry weight (Samaru tial) of five competitive plants per plot,
The remaining plants were left to mature when the seeds were harvested.
AI1 the data collected were subjected to statistical analysis of variance.
Results
Nodulation and vegetative matter production. The nodulation data as expres-
sed by the fresh weight of nodules per plant (Ilorin trial) and number and dry weight
of nodules per plant (Samaru trial) at 9 WAP arepresented in Tables 1.2 and 3, respectively.
Averaged over cultivars, all inoculants significantly enhanced nod&ion when compared
to the uninoculated control in both locations. Also, nitrogen f-on decreased nodula-
tion, but not significantly. Among the inoculants at Samaru, strain IRj 2133 gave the
hig&t nodules number and weight while strain IRj 2123 performed poorest. The cultivars
also differed significantly in their ability to nodulate. As expected at Ilorin, Bossier cultivar
nodulated very poorly with the indigenous rhizobia but nodulated highly with the mix-
ture of inoculant strains, while cultivars TGx 539-!%, Samsoy 2 and M-90 nodulated
well with both the indigenous rhizobia and the inoculant mixture. At Samaru however,
cultivar TGx 814-26D produced the highest number and weight of nodules while the
Samsoy cultivars produced the least number and weight of nodules. Significant interaction
between cultivar and inoculant strain was observed at Samaru in that while cultivar TGx
88849C responded to all inoculants except strain IRj 2144, cultivars TGx 536-02D, Sam-
soy 1 and Samsoy 2 responded td all inoculants ex&pt strain kj 2123. TGx 814-26D
responded to all inoculants except the mixture of strains.
At Ilorin (Table 4) and Samaru (result not shown) inorganic nitrogen fertilisation
significantly enhanced vegetative matter production. There was however no significant
difference in the mean vegetative matter production of the inoculated and uninoculated
control plants. Varietal differences were also observed.
Seed yield. The seed yield data are presented in Tables 5 and 6.
Despite the significant effects of source of nitrogen on nodulation and vegetative
matter production at Ilorin, these were not reflected in the seed yield (Table 5). Significant
varietal effects were however observed. TGx 814-26D and TGx 539-5F significantly out
yielded the rest of cultivars. The very poor yield of TGx 573-209D was due to poor
18
Table 1. Effect of inoculation wi'th B. japonicum strains on nodulation
-
of soybean cultivars (Ilorin location).
Fresh weight nodules (ng per plant)
Cultivars
Uninoculated
Uninoculated
Mixture of
(indigenous
Mean
+ 150 kg N/ha strains
rhizobia)
1Rj 2114
and 2133
_-
-
Samsoy 1
0.06
0.05
0.24
0.12
TGx 297-10F
0.05
0.03
0.14
0.08
TGx 539-5E
0.10
0.03
0.38
0 . 1 7
M-90
0.05
0.03
0.11
0.06
TGx 814-26D
0.04
0.04
0.16
0.08
TGx 136-0211
0.02
0.01
0.15
0.06
TGx 813-6D
0.01
0.05
0.08
0.05
Samsoy
2
0.08,
0.01
0.16
0.08 ;.
TGx 573-209,
':
0.02
0.02
,0.13
0:oi )-
TGx 888-49C
0.05
0.05
0.19
0.10 '
Bossier
0.01
0.01
0.46
0.16
Mean
0.04
0.03
0.20
C.V. % 32.5
LSD (P = 0.01) N = 0.01
LSD !P = 0.01) V = 0.02
LSD (P. = 0105) N x V = 0.07
~ -
Table 2 : Effett'of inoculation with B_,,;japonicum strains on the noduIation of
" sdy'#an cultivars (Samaru location),
,:. :I
Number of nodules per plant
TGx
TGx
TGx
Inoculant
88-49C
536-02D
Samsoy 1
Samsoy 2
814-26D
Mean
1Rj 2114
70.2 (b-h)
43.5 (c-i) 67.9 (b-h)
14.5(hi)
98.9 (a-c) 59.0 (tu)
1Rj 2123
51.3 (b-i)
22.3 (f-i) 29.1 (e-i)
16.7
g-i)
93.0 (a-d) 42.5 (uv)
1Rj 2133
79.7 (b c)
87.3 (b-d) 72.5 (b-g)
88.4
a-d)
140.3 (3)
93.6 (s)
1Rj 2144
23.7 (f-l)
102.7 (a b) 69.7 (b-h)
75.0
b-f)
104.1 iab)
75.1 (st)
Mixture
59.8 (b-h)
63.7 (b-h) 43.3 (c-i)
59.2
b-i)
82.3 (0-e) 62.3 (tu)
Uninoculated
23.3 (f-i)
6.8 (i)
17.'3 (g-i)
23.1
f-i)
41.2 (d-i) 22.3 (VW)
Uninoculated+
.
100 kg N/ha
9.4 (i)
7.3 (i)
20.2 (f-i)
18.3 (g-i)
49.1 !I)-i) 20.9 (w)
Mean
45.4
48.2
45.7
42.2
87.0
SE inoculant means = 7.26; SE cultivar means = 6.13 ; SE inoculant x (.liltivar mean =
16.23. Mean followed by the same letter or letters are not significant at the 5 per
cent level, according to Duncan's multiple range test. Means followed lly 2 letters
connected with a hyphen are included in all groupings between the two letters in
the alphabet.
Table 3 : Effect of inoculation with B japonicum strains on the nod .!ation of
soybean cultivars (Samaru location).
Nodule dry weight (mg per plant)
TGx
TGx
TGx
Inoculant
88-49C
536-020
Samsoy 1
Samsoy 2
814-26D
14 c a n
-
-
-
--
1Rj 2114
382.9
222.4
353.9
55.6
254.7
253.9 (bc)
1Rj 2123
267.6
100.5
99.7
83.3
344.1
179.0
cd)
1Rj 2133
429.8
472.7
304.7
524.5
486.5
4ti3.6
a)
1Rj 2144
139.2
364.0
327.3
336.3
460.3
3i',.4
b)
Mixture
472.4
315.5
159.4
320.3
299.1
31 j-3
b)
Uninoculated
153.4
43.8
64.4
88.1,
173.7
104.7
d)
Uninoculated+
100 kg N/ha
52.2
45.1
91.3
68.8
218.9
94.7 (d)
Mean
271.1
223.4
200.1
210.5
319.6
----
__-
S.E.D. of inoculant means = 44.4 ; S.E.D. of variety means = 37.5 ; ' .F:.D. of
inoculant x variety means = 99.25.
Values in column followed by the same letter are not significantly di' ferent at
the 5 percent level, according to Duncan's multiple range test.
21
establishment of the plants. At Samaru, the relatively earlier maturing cultivars
TGx 888-49C and TGx 536-02D yielded significantly higher than the other three cultivars
which did not differ from each other. Also, the interactive effect of cultivar and strain
on seed yield was significant (F’ = 0.05) at Samaru. Inoculants did not increase the yields
of cultivars TGx 88849C. TGx 53602D. Samsoy 1 and TGx 814-26D, while the yield
of Samsoy 2 was improved by strain IRj 2133.
Discussion
Nangju (1980) postulated that the local cult&us can form effective nodules with
.indigenous rhizobia and thkrefore can be grown without seed inoculation with R. juponicum.
All the cultivars used in this study (except cultivar Bossier) were selected under the
local conditions and were found to respond to inoculation. The fact that the strains differed
in their ability to form effective nodules with the different cultivars indicated that the
most important
is the use of cfllcicnt strains of R. jupunicum that can compete
with the indigenous rhizobia. The comparative advantage of higher number of organisms
in the itioculank than in the soil may also account for the response. Other researchers
have reported differential performance of soybean inoculants (Kang, 1975 ; Chowdhury,
1977 ; Chowdhury er al., 1983).
The positive re&nses of all the cultivars to inorganic N fertilization (with regard
to vegetative matter production) and to inoculation (with regard to nodulation) suggest
that the indigenous rhizobia do not adequately supply the N needs of the plants. It is
however not clear why these (increased matter production and nodulation) do not enhance
seed yield. Further studies on the distribution of absorbed and fixed N to the parts of
an actively fixing soyv plant especially at its reproductive phase should help to explain
‘.’ ‘the situation.
.
Until this is done however, the inoculant requirement situation in Nigeria for enhan-
ced soybean production is uncertain.
Acknowledgement
The authors are grateful to Dr. K. Mulongoy (IITA, Ibadan) for supplying the
peat inoculants. Also a grant from the Federal Ministry of Science and Technology of
the Federal Government of Nigeria, for the National Co-ordinated Research Projects on
Soybean under which auspices this study was conducted, is gratefully acknowledged.
22
Table 4 : Effect of inoculation with B. japonicum strains on vegetative dry
matter yield hg per plant) of soybean cultivars (Ilorin location).
Uninoculated
Uninoculated+
Mixture of strains
Cultivars
(indigenous)
150 kg N/ha
1Rj 2114 and 2133
Mean
Samsoy 1
10.5
13.2
9.3
11.0
TGx 297-1OF
12.3
14.1
11.4
12.6
TGx 539-5E
10.6
15.8
7.6
11.4
M - 90
8.5
12.8
7.0
9.4
TGx 814-26D
10.8
15.9
7.9
11.5
TGx.536020
12.6
16.9
9.6
.13.0
:,
TGx 813-6D
13.6
14.9
11.1.
13.0
Samsoy 2
9.5
18.2
10.7
12.8
TGx 573-208D
12.2
15.6
11.0
13.0
TGx 888-49C
10.7
16.8
11.5
13.0
Bossier
7.2
22.4
9.2
12.9
Mean
10.8
16.1
9.7
C.V. (“i, ) = .19.5
LSD (P = O-01) N = 3.6
LSD (P = 0.05) v = 1.0
Table 5 : Effect of inoculation with 8. japonicum strains on tl., seed yield
'a:+(kg per ha) of so$bean~~ultivar^s (Ilorin location).
'
! ! i',..
I
: LL:I:
;;'*
-*,(>q
;>E
I
\\I
j c: j
i
. :
., ,.
!:rq,
!I
Uninoculated
Uninoculated+
Mixture of c.Irains
Mean
Cultivars
(indigenous
+ 150 kg
1Rj 2114 ant:
Mean
rhizobia)
N/ha
2133
--
--
-
Samsoy 1
1941
2499
2335
2285.3
TGx 297:1OF
2260
1665
2173
2032.7
TGx 539-5E
3648
3203
3242
3364.7
M - 9 0
2508
2320
1930
2286.0
TGx 814-260
3608
3143
3170
3307.0
TGx 536-021)
1578
2008
2903
2163.0
Samsoy 2
1917
1685
1814
TGx 573-209D
860
218
337
471.7
TGx 888-491:
2812
2588
2773
2724.3
Bossier
1615
2092
1908
1871.7
Mean
2208.1
2118.0
2201.8
C.V. (%I 15.8
q
LSO (p = 0.01) v = 268.3
-v..--I
=
e---n..-.--
-A
-I----
A
.-
Y A--
I8
Y
&.A
a
‘3
Y
a,
Y
IX---
‘3
--......
‘3
6
24
bb
Y
Y
‘3
A
Y
25
References
I
AWAI, J. 1981 Inoculation of soybean (Glycine mar L. Merr.) in trinidad. Trap. Agric.
;‘”
(Trin.).
~58,?,3~“;~18.
::
‘_,
“&&&&,. K;ij’. .i(j8i Respon;iof $y&, ~&~~ & ,L*) u, indigenous *izobia
and other sources of nitrogen in south-western Nigeria. Nig. J. Agronomy (in press).
AYANABA, k. 1977 Towards better use of inoculants in the humid tropics. In : Biological
nitrogen fixation in farming systems of the tropics. (Ayanaba, A-and Dart, P.J. eds.),
pp. 18f-187. .Jdhn Wiley and Sons, New York.
AYANABA, A. and FANGJU, D. 1973 Nodulation and nitrogen fixation in six grain
’ iegdipe&‘& 1 Pro&dings of the first IITA grain legume improvement workshop.
pp. 198-204. IFA, Ibadan publ.
BEARD, B.H. akd HOOVER, R.M. 1971 Effect of notrogen on nodulation and yield
o f irrig,qt& soybeans. Apron. J. 6 3 , 815-816.
BHANGHOO, MS. and ALBRITTON, DJ. 1972 Effect of fertilizer nitrogen, phosphorus
and potassium on yield and nutrient content of Lee soybeans. Agron. J. 64, 743-
746.
CHOWDHURY, M.S. 1977 Response of soybean to Rhizobium inoculation at Momgoro,
Tanzania. In : Biological nitrogen fmation in farming systems of the tropics. (Ayanaba,
A. and Dart, PJ. eds.) pp. 245-253. John Wiley and Sons. New York.
CHOWDHURY, M-S, MSUMALI, G-B. and MALEKELA. G.P. 1983 Need for seasonal
inoculation of soybeans with rhizobia at Morogoro, Tanzania Biol. Agric. Hart. 1,
219-228.
GRAHAM, P.H. 1985 Problems of soybean inoculation in the tropics. In : World soybean
research conference III. mgs (Shibles, R. ed.) pp. 951-959. Western Press,
Inc.
HARPER, J.E. 1974 Soiland symbiotic nitrogenrequirementforoptimum soybean production.
Crop Sci. 14, 255-260.
JOHNSON, H.S. and HUME, D.J. 1972 Effects of nitrogen sources and organic matter
on nitrogen fixation and yield of soybean. Can. J. Plant. Sci. 52, 991-996.
KANG, B.T. 1975 Effect of inoculation and nitrogen fertilizer on soybean in Western
Nigeria. Expl. Agric. 11, 23-31.
LUSE. RL., KANG, B.T., FOX, R.L. and NANGJU. D. 1975 Protein quality in grain
legumes grown in the lowland humid tropics with special reference to West Africa.
In : Fertilizer use and protein production. XI. Collegium International Potash Institute,
Denmark.
NANGJU, D. 1980. Soybean response to indigenous rhizobia as influenced by cultivar
origin. Agron. J. 72, 403406.
OFORI, C.S. 1973 The importance of fertilizcr N in grain legume production on soils
of granitic origin in the upper region of Ghana. In : Proceedings of the Fit IITA
grain legume improvement workshop. IITA Ibadan publ.
PAL, U.R., OLUBAJO, 0.0. and ADU, J-K. 1985 Recent findings on the cultural and
nutritional requirements of soybean in the Guinea savanna zone of Nigeria. In : Pro-
ceedings of the 5th annual workshop of the nigerian soybean scientists. pp. 52-64.
26
PAL, U.R. and SAXENA, MC. 1975 Response of soybean to symbiosis and nitrogen
fertilization under humid sub-tropical conditions. Expl. Agric. 11, 221-226.
SALEMA, M.P. and CHOWDHURY, M.S. 1980 Rhizobial inoculation and fertilizer effect
on soybean in the presence of native rhizobia at Morogoro, Tanzania. Beitz Tropiqh.
a
Landwirt. Veterinamed. 18, 245-250.
SINGH. H.P. and TlLAK, K.Y.B.R. 1977 Response of soybean to inoculation with various
commercial inoculants of Rhizobium. Indian J. Agron. 22, 57-59.
SORRENSEN, R.C. and PENAS, EJ. 1978 Nitrogen fertilization of soybeans. Agron.
J. 70, 213-216.
VEST, G., WEBER, D.F. and SLOGER, C. 1973 Nodulation and nitrogen fixation.
In : Soybeans : improvement, production and uses (Caldwell, B.F. ; Howell, R.W.and
Johnson, H.W. eds.) Amer. Sot. of Agron., Madison, Wisconsin.
WAGNER, G.H. 1962 Nitrogen fertilization of soybeans. Missouri Agrir. EXP. Starim
ices. b’uu. IY I.
I.‘__
Maximiser fa FBA
la Production Agricole et Forestitre en Afrique
Nodulation and growth
o f baididra &dunddut (Vigrtu su7jterrLd,l,eu>
in th&e soils
.,:
:.
._
MULONCiOY, K ‘and’ GOLI; A.E.
Inter~ional htitute of Tropical Agriculture (II’EA)
PiUB. 5320 Oyo Road, Ibaakn. Nigeria
-1~
.
Ab’stract
L
:.r
:,.;.A:
.:
.,I.
. .
Nodulation and dry matter production of ten cultivars of bambara groundnut differing
in growth habit, maturity period and yield potential were assessed in potted soils
from three locations in southern Nigeria. Nodulation occurred in all three soils with
the least number of nodules plant-’ in the acid soil (Ultisol) from Onne. near Port
Harcourt
Urea did not affect nodulation in general. Differences were observed between cultivars
. .‘bS I.* on the:basis “6f ntkhrk number
:>f-> (1 ,a.
but not of nodule’ dry..weight Relative symbiotic
(_
. -:::,. $3 ..
.,.
~~‘~effectiven&~vahresbased
c-k phmt dry ‘tieight av&rage$W % for all cultivars grown
in the three soils. This indicated that bambara groundnut
$xablishes effective symbioses
with iitdig&ot&‘iG&a in thkse ‘soils. Only cultivar TVsu 1231 responded to urea
i
N. signifrc~tly (at P g 0.05) in soils from the International Institute of Tropical
.*’ Agriculture (ETA) and Pashola (Alfiiols). There &s’no correlation between growth
habit or yield p&x&l and plant dry matter’pmduction.
28
Introduction
A few years back, bambara groundnut (Vigna subterranea
(L.) Verdcourt) was
one of the most widely cultivated leguminous crops in Africa (Doku and Karikari, 1971).
Nowadays, with the expansion of groundnut production for oil. bambara groundnut is
only sparsely grown in Africa (Goli, 1987) in spite of its pest and disease tolerance both
in fields and during storage, its nutritional value and its adaptation to a wide range of
soils (Doku and Karikari, 1971).
Few studies have been conducted to document the ability of bambara groundnut
to fix nitrogen and grow on low-N soils. Duke et al. (1977) reported in their review
that high soil N promotes vegetative growth of bambara groundnut at the expense of
seed production. Stanton (1968) found that inoculation of new fields of bambara ground-
nut with soil from plots where the crop grew well improved crop yields. But Johnson
(1966) and Denarit et al. (1968) showed that inoculation with rhizobia did not improve
yici& signil’ictiltiy. Marc inucuiauurr studits are being carried in pars oi Senegal where
bambara groundnut does not nodulate freely (Gueye, M. pers. comm.).
The germplasm unit of the International Institute of Tropical Agriculture @TA),
Ibadan, Nigeria, possesses an important collection of bambara groundnut with 2027 ac-
cessions. In view of the potential of this legume in tropical agriculture, we assessed its
growth and nitrogen-fming ability in three low-N soils comprising an Ultisol from the
humid region, and two Alfisols from the subhumid forest-savanna transition zone. Cultivars
used in this study represented a range of growth habits, maturity periods and yield poten-
tials of bambara groundnut
Materials and methods
Ten cultivars of bambara groundnut were compared on the basis of their nodu-
lation, dry matter production and response to urea to assess their nitrogen-fixing ability.
The experiment was conducted in pots at IITA. Growth characteristics and yield potential
of the cultivars used are presented in Table 1. The completely randomized design was
used with five replications and a factorial combination of 10 bambara groundnut cultivars,
three soils low in N from IITA, Fashola (at about 70 km North of Ibadan) and Onne
near Port Harcourt, all in Nigeria, and two rates of urea applied at 0 and 100 mg
N kg-’ soil. Some properties of the experimental soils are presented in Table 2. Urea
.y,as applied as a water,, solution at one and two weeks after plant emergence. All pots
” received a basal application of single superphosphate (25 mg P kg-’ soil) and muriate
of potash (15 mg K kg’ soil).
Five seeds of each cultivar were sown in each pot containing 3 kg of soil previ-
ously passed through ‘a 10 mm sieve to remove stones and large plant parts. Seedlings
were thinned to three at one week after emergence. Plants were harvested at 8 weeks
after planting for nodule count and mass, and plant dry weight.
Results and Discussion
Nodulation occurred in all the soils with an average of 30 nodules plant-’ in Fashola
soil (Table 3). Nodule number was 21 % and 44 % less in IITA and Onne soils respec-
tively ; but plant nodule dry weight was the same in the three soils indicating that nodules
29
Table 1. Growth characteristics and yield potential of cultivars of
Bambara groundnut studied.
Cultivars
Yield jqtential
'Growth hab4t'
(TVsuI
Days to maturity
(g plant-')
5
spread
123
4
9a
bunch
113
27
9b
spread
115
1 5
132
spread
131
1 8
409
intermediate
105
5
465
b u n c h
105
1 5
2"
790
b u n c h '
127
094
1 055
bunch
150
5
1 061
intermediate
157
5
1 231
intermediate
123
87
30
Table 2. Some properties of the experimental soils collected from
IITA, Fashola and Onne.
-
Parameter
IITA
Fashola
Onne
P H
5.8
6.2
4.6
Organic C, St
0.53
0.36
1 .Ol
Total N, %
0.11
0.04
0.13
Extractable P (Bray-l),
mg kg-'
21.0
7.8
54.8
CECa,
meq 1OOg -1
1.89
1.21
2.32
Total acidity (Al + H), meq 1OOg -'
0.02
0.01
1.72
aCation e x c h a n g e c a p a c i t y
were relatively smaller in IITA and Fashola soils. Thompson and Dennis (1977) reported
numbers ranging between 92 and 318 nodules plant-’ in Ghana. Free modulation was also
reported in Madagascar (DenariC et al.. 1968) and in Zaire (Rassel, 1960).
Nodulation in this experiment was not affected by fertilization with urea except
in Fashola soil (results not presented) where nodule number of cultivars TVsu 9b and
TVsu 409 decreased and that of cultivar TVsu 9a increased in the N treatment These
differences were also observed when the average nodule number for all three soils was
calculated (Table 3). In general, there were in both N treatments differences between
cultivars of bambara groundnut on the basis of nodule number but not of nodule dry
weight (Table 4). Correlations between nodule. numbers or weights and shoot, root or
plant weights were poor at P = 0.05 (r < 0.253) in the no-fertilizer N treatement. For
instance, plant dry matter production was highest in Onne soil where nodulation was
the poorest (Table 3). Also, there was apparently no relationship between growth habit,
maturity period or yield potential and nodulation, relative effectiveness or dry matter pro-
. :
. . ..__ :;...
Table 3. Nodulation and growth of Bambara groundnut in three
different soils (1)
.:
Plant dry'
fiodule
Nodule
Single nodule
Soil
weight '.
number
dry weight
dry weight
(g plant-')
plant -1
(mg plant-')
(mg)
Ibadan
4.2
2 3
8 7
3.8
Fashola
3.2
41
100
2.4
Onne
5.3
_
16
80
5.0
LSD (0.05)
0.89
4.3
NS
1.6
(1) Means of 10 cultivars in the no-fertilizer N treatment.
32
Table 4.
Response of ten cultivars of bambara groundnut to fertilizer Nl
and their relative effectiveness (R.E.)2 in nitrogen fixation
based on plant dry weight.
..,>i,“l’
r>O,:L
“‘y
~*liuii
1 t ItUlilbcr
I.UliU I c til y
>lhglc
II”UVI’
:..:
weight
weight
dry weight
(TVsu)
N O
Nl NO
N1 NO
N1
NO
N1
__- -
-
-
-
_
g plant -1
No. plant -1
mg plant -1
-----w-----
?z
5
4.3 5.3
3 0
27 168 98
5.6
3.6
8 2
9a
3.5 4.0
1 8
31 54
71
3.0
2.3
8 8
9b
3.7 4.4
2 2
16 80
126
3.6
7.9
8 4
132
5.3 4.0
2 3
27 88
59
3.‘8
2.2
133
409
3.3 4.0
3 4
21 111
146
3.3
7.0
a 3
465
5.1 4.7
28
32 92 97
3.3
3.0
109
790
5.5 6.3
2 5
32 83
7 3
3.3
2.3
8 7
1055
4.5 3.8
2 2
33 88
132
4.0
4.0
118
1061
5.7 6.0
2 0
13 69
5 6
3.5
4.3
9 5
1231
4.3 6.7
11
16 60 71
5.5
4.4
6 4
LSD(S%)
1.48
11.1
96
4.32
ND3
1NO = without fertilizer N; N1 = with 100 mg urea N kg-l soil
2R.E. : Relative effectiveness = Plant dry weight at NO * lOO/Plant dry
weight at NT
3ND
: not determined as R.E. values were obtained from computed values.
3 3
Dry matter production was only slightly increased by N fertilization. Except for
cultivar TVsu 1231. relative effectiveness based on plant dry weight ranged between 82
and 132 %. This was an indication of effective symbiosis between bambara groundnut
and indigenous rhizobii present in the three soils studied. Cultivar TVsu 1231 had relative
effectiveness values of 421.64% and 87% in Fashola, IITA and Onne soil respectively,
with an average relative effectiveness of 64%. Thus cultivar TVsu 1231 had the least
effective symbiosis with native populations of rhizobia in BTA and Fashola soils and
would require inoculation with the appropriate rhizobia to fix nitrogen adequately in these
soils.
Aknowledgement
The autors thank KE. Dashiell for his suggestions and W.O. Oyekanmi and E.I.
Jegede for technical assistance.
References
DENARIE, J., ANDRIAMANATENA, S. and RAMONJY, J. 1968 L’inoculation des 16-
gumineuses a Madagascar. R&nltats de I’expkimentation. Agr.Trop. 23, 925-966.
DOKU, E.V. and KARIKARI, S.K. 1971 Bambara groundnut. EconBot. 25, 255-262.
DUKE, J.A., OKIGBO, BN., REED, CF. and WEDER, J.K.P. 1977 Voandzeiu subferranea
Q Thouars. Tropical Grain Legume Bulletin. 10, 8-l 1.
GOLI, A.E. 1987 Plant exploration in Senegal. G.R.U. Exploration 1986. BTA, Ihadan,
Nigeria.
JOHNSON, D.F. 1966 Bambara groundnut. A review. Rhod. Agric. .I. 65, 1-4.
RASSEL, A. 1960 Le voandzou (Voandzeiu subterranea L. Thouars) et sa culture au
Kwango. Bulletin Agricole du Congo Belge. 51, l-26.
STANTON, WR. 1968 Grain Legumes in Africa. Rome : FAO.
TOMPSON, EJ. and DENNIS, E.A. 1977 Studies on nodulation and nitrogen fixation
by selected legumes. Proceedings of the University of Ghana Council for Scientific
and Industrial Research : symposium on grain legumes in Ghana. University of Ghana
p. 85-102.
35
Très rtkemment, les essais d’inoculation du voandzou (Vigna subferranea (L.)
Thouars) ont débuté en station expérimentale. Le voandzou est une lt5gumineuse à graines
de haute qualiti nutritive (les graines contiennent 17% de protkiies), r&&ante aux
maladies et aux insecks, et parfaitement adaptée aux conditions semi arides des rkgions
tropicales (NA& 1979). Le voandzou est cultivé au swkst du Séneg+ dans les petites
exploitations agricoles, seul ou en association avec une c&%le.
Ce prkent rapport dkrit les premiers rkwltats obtenus sur l’effet de l’inoculation
avec des souches de Rhiwbium préalablement s6kctio~6es (Guéye et Bordeleau, 1988)
sur le rendement en grains de deux variCt&s de voandzou.
Matbriels et mbthodes
Un .&sai d’inoculation du voandzou au champ a &5 conduit en station e.xp&-@e&e
au sudest .du %n&al & Sinthiou h4aRme en 1986 sur lé %e A (pluviom&rie : 794 km)
et en 1988 sur le site B @luviom&rie : 873 mm). Les caractkistiques physicochimiques
dti sol des deux sites sont indiquées au Tableau 1.
. c
,. . .
Tableau 1
: Caractéristiques physico chimiques du sol de Sinthiou Ma]$c
-~
wm
. .
Comppsifion (5;)
,.
Passimilabk '
Sites
C total
N total
P total- fOlsen*) Argile Limon Sable pH (H20
---y
-
-
-
-
A
4380
450
557
156,5
4,8
?,5
97.7
6.40
E?
4540
470
402
10995
4,3
2,3
93,4
6,64
*Olsen e t .a?< (19541
,_,_,
‘,
.>.
’
”
;
hxirniser la FBA pour la production agricole et forestière en Afrique
.
:1.
<. .’
i
2~. . . : 2.,
.’ St S:*I!, ~ “,’
_;,
‘“:x
‘2.
Effet de l’inoculation avec des souches
de Rhizobium et de 15, fyC$sation azotée
sw;. i&‘A,2dé&&ti *é;;;dgrains
” “,
du Voandzou (Vignci subterranea (L.) ‘Thouars)
.._
au Sénégal
GUEYE, M
.
MIRCEhKNRA, BP. 53. Bamhq. S&dgal
:
_.._I’
I-..
.
_
_.._.
-
_____
.-
.--,,-,.I1._
_-.
I
..
--
:,;y,
.-
.,‘.
RBsunlé
Desttssaisauchampontétécondllitc PU S&+~~RI m !Wf;ct 10PR c-T~T~~:P- r..~L-:z:.~~:z’.:
pour étudier la réponse du voandzou (Vigo subterranea (L.) Thouars) à l’inoculation
avec deux souches de Rhizobiwn : MAO 113 et TAL 2.2. L’inoculation des deux
vari&% étudiées a augmenté leur rendement en grains de 27 à 58%. L’amélioration
de. rendement la plus significative (58%) a kté obtenue en 1986 avec la variété
83-131 inocul6e avec la souche TAL 22 En 1988, l’ion avkc la-rsouche
MAO 113 a amklioi le rendement de la vari&? 79-l de 52%.
Introduction
L’inoculation des légumineuses A graines avec des souches de Rhizobium n’est
pas une pratique courante au SMgal bien que ses effets btkéfiques y aient été mis en
évidence sur le soja en particulier. L’inoculation du soja au champ avec la souche de
Rhizobium USDA 138 conditionnée dans de la tourbe a augmenti significativement le
rendement en grain du soja @archer et al., 1984) ; au laboratoire, Dommergues er al.
(1979) avaient egalement obtenu une augmentation tr&s significative du poids sec des
nodules et de l’azote total des parties aériennes du soja inocul6 avec la souche de Rhizobim
USDA 138 incluse dans un polymkre de polyacrylamide.
Deux va&&-de voandzou (79-l et 83-131) ont 6t6 inocuEes avec deux souches
de Rhizobiwn : MAO 113 isol6e du voandzou au S&&gal (Guèye et Bordeleau, 1988)
et TAL 22 fournie par le projet Niftal à HawaiL De plus, l’essai comportait deux traite-
ments supplémentaires sans inoculation avec une souche de Rhiwbium : M traitement
consstantenrmeapplicationd’~~azo~a~de50kg~etuntraitement
sans azote. L’essai a 6t.6 conduit en blocs compl&tement randomisés avec quatre @&itions.
Chaque panAle Unenta& (2,4 m x 15 m) avait reçu une fertil&tion de fond de
60 kg P,o/ha sous forme de supertriple et de 120 kg KCJIha. L’inoculum a 6t6 apport.&
au sol sous forme de tourbe (5g/poquet) contenant approximativement 109 Rhizo-
bium/g. Le voandmu a été sem6 à raison d*,F”gr@e par poquet selon un Ment
de 30 cm entre les lignes et 15 cm F .la ligne.
Le rendement en grai& du voandiou a & Cvalué à la &olte. Les anaIyses statis-
tiques ont &fZ effectu&s en utilisant le test de .Duncan (1955). .
.I i .
.
:.
’
;. . .
.i.,
_,
Rb~Rats
e
t
dls&sslon : -
:
:
:’
En comparaison avec les parcelles t6moins (Tableau 2). le rendement en grains
du voandzou n’a pas été significativement affect& par un ap~~~rt de 50 kg ur6e/ha bien
que sur le site B en 1988, il.y.aït eu 30% d’augmentation avec la vari&6 83-131. Par
contre, l’inoculation du voandzm avec les souches de Rhizobium MAO 113 et TAL 22
a augmenté le rendement en grains des vari&?s 79-1 et 83-131:En 1986, l’inoculation
avec lasoucheh4AO 113 aaugmenté IerendementdelavarH 83-131 de51%; l’inoculation
aveclasoucbeTAL22aaugmentélerendementdesdeuxvaritWs79-1
et83-131 respectivement
de 50 et 58%. En.1988, l’inoculation avec la souche MAO 113 a am6lionZ le rendément
des vari&+ 79-l” et 83-131 ~tivement de 53% $ 33% ; avec la. souche TAL 22,
les m~@at$yq@ * .+$%&r la varit% .:ZJ-llet 27% :F ~puZt6 X3-131:
Ces &ultats. montrent que ,la difE.rence g&&ique entre les deux va&& uti&s
influence fortement la r@onse ‘au champ du voandzou a l’inoculation avec les souches
de Rhizobium : la Aponse de la variété 83-131 est plus stable que celle de la vari&&
79-l. Cela était pr&isible car une exmence précédente avait mont& que l’indice de
nodulation de ces deux variétés révClait une symbiose plus efficiente de la variété
83-131 en association avec les souches MAO 113 et TAL 22 (Guèye et Bordeleau, 1988).
Ces résultats confhment donc les travaux de Dadson er al. (1988) qui avaient mis en
évidence l’influence du g&otype des variétés de voandzou à l’inoculation avec des souches
de Rhizobium. L’effet bénéfique de !‘,noc@tion~,+~~ 1~. ,so+~ de &$yrhfzobium
sur
le rendement en grains.du &andzou était @lemenbpr&i@ble cardes &des antieunzs
(résultats non publ&) avaient montré que le sol de sinthiou MalémeÏ&ferme t.r& peu
de Rhiwbium (c 2OO/g) et que le voandzou a souvent besoin d’&re inoculé (Johnson,
1968). On pourra3 donc recommander l’inoculation du voandzou dans les sols du StMgal
renfermant t&s peu de Brdyrhizobizm. Cependant, on doit garda pr&ent A l’esprit le
fait que l’efficience d’une association Rhizobim x plante-hôte varie considérablement
d’un site 2 un autre.
3 7
:
f
‘..;i..
.:
:
&fF-.?,, -:>.:L2<- ,*..a. -’ J,.’ 1.. . . - .
.; .:. i.-~~~~~~~~~~~~~~~~
;<,, : .-
;---
.<
__c-
-----
----..
-
-.__-----
- .
-.
.-----
---
I
.Tableau 2 : Effet de-l'azote et-de-l'inoculation avec les souches de
BradyrhizobiÙm:yD:-113 et TAL 22 sur le rendement en grain
du.vo?uizou ~varj.~t~-79-1 et 83-131.) cultivé au champ, à
]
-
la station ~e&&'iÏn&ale de Sinthiou MaliMe.
l
:
_-
I
.-
.: .:
:
,
:
:.-:/:
:.
~ Y.
.,
.1-p.. - 3.
Rendement en.grain- {kg/ha)
/..,.
.'.
*
Varietés: Traitements
Site A (1986) Site B (1988)
. . ;.
_-
.3062,50 b‘
Témoin
'
3080,30 Ii
3555.40 b
Azote (50 kg urée&)
3832;90 ab
4624,lO a
r';r?"
_I
."_
,.~
-->
TAL 22
4863,90 a
4524,40 a
'... Daqs.:chaque colonne et?poÛr.chaque variété les valeur suivies d'une:
'
.
:: &jnê+:&'ttr~ ne ront.jign~~iCaMveiaent- diffbentes 'd'après. le test de
ii.- ~Duncan.:31955) '
a
u
seuil:d$~,5 ?h:- ',: i
..r_
A;
_ I
-- L. U-L I_ ___ --._ ..1.- -.a--
.y,
_ --~i...,.-eT
._ ::.-:
_-
3 8
Remerciements
Ces travaux ont étk réalises d’une part grâce au support financier & 1’U.S. BOSTID-
NAS par la bourse no BNF-SN-2-84-21 et d’autre part grâce au financement du projet
UNEP-FAO CP/FP 6106-84-02. L’auteur remercie Alioune Gning, Cheikh Samb et Omar
Tour6 pour leur assistance technique.
Bibliographie
DADSON, R.B., BROOK, C.B. and WUT.OH, J.G. 1988 Evaluation of sekcted rhizobial
strains in association with bambara groundnut (Voandzeiu subferreneu (L.) Thouars).
Trop. Agric. flrinidad). 65, 254-256.
DOMMERGUES, Y.R., DIEM, H.G. and DIVIES, C. 1979 Polyacrylamide-entrapped
Rhizobizm as an inoculant for legumes. Appl. Environ. Microbiol. 37, 779-781.
DUNCAN, D.B. 1955 Multiple range and multiple test. Biomefricu 11, l-42.
GUEYE, M. and BORDELEAU, L.M 1988 Nitrogen fixation in bambara groundnut, Vound-
zeiu subterruneu (L.) T~OUTS. Mircen J. 4, 365-375.
JOHNSON, D.T. 1968 The bambara groundnut : a review. Rhodesiu Agric. J. 65,
l-4.
LARCHER, J., WEY, J. et GANRY, F. 1984 Recherches sur le soja de 1978 a 1983.
Rapport ISRA, CNRA Bambey, 86 p.
NATIONAL ACADEMY OF SCIENCES. 1979 Puises. In : Tropical Legumes : ressources
for the future. pp. 47-53. National Academy of Sciences, Washington D.C.
OLSEN, S.R., COLE, L.V., WATANABE, F.S. and DEAN; L.A. 1954 Estimation of
availablephosphate in soils by extraction with sodium carbonate. Circzdur of US DepurtmeM
of Agriculture No 939.
FIXATION BIOLOGIQUE DE L’AZOTE
CHEZ LES LEGUMINEUSES FOURRAGERES
Uaximiser la FBA pour la Production Agricole et ForestiZre en Afiique
Nitrogen fixation of the forage legumes
and their -residual effects on wheat growth
and yield
Soils rmd Plant Nutrition Section, h&national Livestock Centre for Af&a (ILCA),
P.O. Box 5689. AaZis Ababa. Ethiopia
Abstract
The symbiotically fixed nitrogen of various forage legumes was evaluated for two
seasons un% upland soil conditions in Ethiopian highlands using ‘5N technique and
Avena sativa (oats) as a non nodulating reference crop. Forage oats gave higher
dry matter yield than Vi& darycxzrpu (vetch), however, the latter gave feeding stuff
of highex N concentration and yield (169 kg N/ha&ear for vetch vs. 73 kg N/ha/
year for oats). In 1985. 78, 78 and 73% of the total N in vicia, clover (Trifoliwn
steudneri)andmedics(Mcdicagopdymorpha)p;iginatedfnrm~~.Vi~medics
and trifolium fixed 131, 77 ‘and 61 kg N/ha nz&uively. In 1986. the dry matter
and grain yield of wheat following vi&, medics and trifolium was significantly higher
than that following oars. Between 12 and 31 kg of additional N was taken up by
wheat straw and grain following legumes and fallow as CompBTed with those following
Oats.
Introduction
Of the many possible nutrient deficiencies, lack of N imposes the most widespread
and strongest restrictions on plant and animal production in sub-Saharan Africa (Nnadi
and Haque, 1988a). While the N status of soils can be improved by the addition of N
1:
.
.
:
^- ; : ,-
41
fertilizer, it is expensive input and is reflected in its low consumption in the region @‘DC,
1986).
In Ethiopia, the average yields of most of the cereals are below 1000 kg/ha. In
order to improve the productivity of feed and food in various production systems and
to sustain it at reasonable level, there is an urgent need for alternate sources to increase
the influx of N in the system. Use of legumes in various production systems to benefit
from biological nitrogen fmation @W) contribution to following cereal crops may attract
quick adoption by resource poor farmers. Legumes also contribute to soil productivity
by improving soil structure (Andrew, 1965 ; Stoveman, 1973) and can control pests and
diseases of cereals in legumecereal rotations (Moore et al. 1982).
Owing to the multiple facets of the contribution of legumes to cereals and the
difficulty of isolating these factors, the contribution of legumes is often studied by measu-
ring the yield of following crops (Muss and Burhan, 1974 ; Papastylianou et al.
1981 ; Singh et al. 1985 ; Ayoub, 1986 ; Osman and Nersoyan, 1986 ; Mohamed-Saleem
et al. 1986 ; Papastylianou, 1987 ; Papastylianou and Samios, 1987 ; Keatinge et al.
1988).
This communication reports BNF by annual forage legumes and their residual effect
on subsequent wheat crop on an upland soil in Ethiopian highlands.
Materials and methods
Climate
Rainfall, potential evapottmspirattion (PE) and water balance.
The rainfall exhibits strong scasonality and of the total annual rainfall of 866 mm, 84%
is recorded in the rainy season extending from June to September (Fig. 1). Short season
is from March to May, when a total of about 140 mm rainfall is received. PE is 1320
mm (Fig. 1) and values are lower (2-3 mm/day) during the rainy season as compared
with short rainy season (34 mm/day).
The water budget (Fig. 1) shows that in the months of July and August there
is a surplus of about 300 mm of water. Daniel (1977) reported that potential runoff at
Debre Zeit, in an average year, is likely to be 100 mm which could be collected and
reuse for crop and livestock production in the post rainy season.
Ten?perature and radiation. Temperature and radiation are two factors
that limit crop and livestock production in Ethiopian highlands. Mean maximum, minimum
and average monthly and an&al temperatures. are shovvn in Fig. 2 Relative sun&ine
hours are about 5O%‘d&ng the rainy months and are 70 to 80% during the rest of the
year. Total radiation is around 500 cal/cm2/day and somewhat lower during the rainy
months (300450 cal/cm2/day) and higher in the dry period from February to May (500-
550 cal/cm2/day).
Soils
Debre Zeit is situated 8” 44’W, 39” 02”E at an altitude of about 1850 m above
sea level, about 50 km south-east of Addis Ababa in Ethiopian highlands. The soils are
mainly vertisols and some alfisols. The experiment was conducted on an alfiiL
Along the toposequence, alfisol, soil with vertic properties, and vertisol stored 113,
187 and 278 mm water per meter depth respectively. During the maturity stages of forages
:
,.
.l‘i..‘;
;.::..t,
:
‘.
._ ,.... ‘> .,., i,;:
,’
:
.
.
._.
./
42
Debre Zeit, Ethiopia
8044'N. 3g002'E, 1850 as1
e Rainfall
m m
- - - - pE
240
Annual roinfoli - 866mm
A n n u a l P E
- 1319mm
1 6 0
A
--A-,
/
/
C-
I
i’
-8C
Figure 1:
Rainfall,
pan
evapotranspiration
(PE) and water
budget at Debre Zeit for five years.
. .
43
Debre Zeit. Ethiopia
as1
30
8O44'N, 39O02'E, 1850
Maxlmum
26.0%
;3
0
-g 20
A-g=. - _ - ,
-.
/.A’
---2.
.
-.-. -._.
18.7%
3
I-
-
-
-
0
I
I
I
I
I
I
I
JFMAMJJAS () i D Annual
Month
Figure 2:
Maximum, minimum and average temperatures at Debre
Zeit.
“.
‘/‘...
and crops, xrtisol stored 57% more moisture than alfisol and remained moist throughout
the year below 50 cm depth (Kamara and Haque, 1988). Some of the physico-chemical
properties of experimental soil are given in Table 1.
Agronomy
A total of five treatments were tested along with a &llom and the non-nodulating
reference (oats) treatments (fable 2). A randomized complete block design with four re-
plications was adopted. The plot size was 6 x 4.50 m and there were 2 and 1 m paths
between replications and plots respectively. AU the legume treatments as well as non-
nodnlating stada-d crop recdd 20 kg W F+ _~r~~zq%nting~ except in- ~mplots.
Within each of these plots, a n&oplot of 2 x%!5 rir rcce$d 20 kg N/ha of 5 % 15N
a.e. in the form of urea solution.
All these treatments received a uniform application of 30 kg P/ha as triple su-
perphosphate. Medics and clover were planted at 25 ,cm apart at a seed rate of 10 kg/
ha.Vetchwasplantedat5Ocmapartat.25kg/ha.whileoats,25cmapartat1OOkg/
ha. The experiment was plantcd on July- 11, -1985 and 2Okg N/ha of 5 96 ‘%I ae. urea
was applied on July 15, 1985.
Plants were harvested 98 days after planting (DAP) for dry matter estimation and
1 mz was sampled from ea$.of ?-%zlabelled microplots
,.:;.: ‘Z.. :..<. .,-,\\t-T:., .j
.‘..L,< & :: (1 ‘r
for total N and %I analysis. The
praporpon of ~~t.N .&.g?&..m ,~bi~tic,N.~~~~.~:;~~~‘to
the
procedure given by Hardarson (1984). The legume residues were removed from the plots.
The following year, wheat cv. Buhe at 25 cm apart was planted on 18 June, 1986
atseedingmteof1OOKg/ha.TwoNrates(Oand6OkgN/ha)intheformof~ureawere
superimposed on each plot before planting. Wheat was barvested from 2 x 2.5 m at
113 ,DAP for dry matter and grain e&nation. Composite samples of straw and grain
I.
w&eanalysedfortotalNtoestimateNuptake.-
I*.
‘{ 2. I 7. p ;f ,7 f *:: I-:.. ;:‘T i.
: . .
:
.;.
.-
,_:..
.:
:
45
’
. Table'1 : Physico-chemical properties of experimental soil
Properties
Values
‘_I
.’
.,...
“
.:
‘-
5,
..
s.
Sand
(%),
49
Silt 1%)
19
j 'Clay 1%)
32
pH (H$I)
.
-: b;iganic matter (%)
‘;_ 6.!
.i' :I :
.;
,. 2 . 2
Nitrogen (%)
0.05 :
Available P (Bray II - ppm)
7
Exch. cations (me/100 g)
K
0.65
Ca
13.20
k.l
9.76
N
a
0.41
I
.’
rs
.I
:’ : . . : .‘.
.. 46
Results and discussion
Dry matter and nitrogen yield of legumes or oats
There were no signifi--t diffpk am&ig,.pats, yetch and medics with respect
to dry matter yield ‘and thi: l@~&%*ced $0&x .dry: mattex than oats (Table 2).
Si~tlyhigherNyieldwasob~~frdin~~asw’nparedwithothertreatments
and there were no difference$$$J:jyields among the others (Table 2). Papastylianou
and Samios (1987) reported that wooly p&l vetch gave higher N concentration and yield
(86 kg N/ha&ear for vetch vs 55 G N/ha per ye&) ‘than barley.
:,
:,
6lologlcal nirogen fix&n ! : _
Table 2 shows the 15N a&?&derived from &xation and N flied by various
legumes l%J enri&ment of oats w& si&ificantl~higher~as compared with that of legumes.
However, there was no significant differences among the legunies though lowest ‘W enrichment
was noticed in vetdh (Table 2).
Nitrogen derived from ftion varied fi-om 73 to 78% in three legumes and no
significant differences among various.trealments were observed (Table 2). Total N fixed
was 131,77 and 61 kg N/ha in vet&me$ics and clove resptively. Vetch significantly
fixed hi@@,, as ‘Compared ivit6:&ve&(Tabl~Table; ?). .&&~ti&e,& al :.(1988) reported that
.;,&m*:;:-$ i‘..
vetch’-m .#). to 4g tig Nb at ,dy md Td &ii .g$-v~+&y & No*m syria
They further observed that vetch fixed do$Ie the amount of N in kiproved management
as Compared with traditional .qp, Higher N fixation by vetch in Ethiopian highland
conditions might be due io mor( reJiab1~ rainfa& prkence of indigenous rhizobii and
higher fertility of upland soils as timpared with no&&n Syrian’conditions.
-;.. *
- :
i
. Residual &f&t bf &g&i& ,$ $#$$
i
ic
i’
.py “-..’ j.
“u ::.
Dry matter atid gn$e y@!~.The+nxid~.effem oQ985 season cropping
treatments on dry matter and grain yield of wheht at two N rate~-ire given in Tables
3 and 4. The dry matter yield of whe~~~following v&h was significantly higher than
that following oats or natural faIIow. However, no significant difference in dry matter
was ohserved -following v&h, medics and clover.
The grain yield of wheat following ‘vetch was sigriificantly higher than that following
oats. No significant difference in wheat grain yield was observed in;plots previously cropped
to, the legumes. Yields in fallow. * oats plots w? jthe lowest Fable 4).
s. r:
-2.
e s
., .)
..%.s ; 1
.*
::
“ .
i
Table 2 : Dry matter yield, total N yield, percent T5N atomic excess, percent N derived from
fertilizer, N fertilizer yield,
Ndf fixation and N fixed by forage legumes,
Debre Zeit, 1985.
Dry matter
N yield
15N
Ndf
N fert
Ndf N
Treatments
(kg/ha)
a.e
fert.
yield
.'fix
fixed
non
Isotopic
%
%
(kg/ha)
%
.'
(kg/ha)
' :
Isotopic
;
----
Vicia dasycarpa
4531b*
49DOa* 169a*
O.O99b* 1.852b* 3.015b*
77.93a* 13la*
Medicago polymorpha
4163b
39D;Oab 106b
0.123b 2.306b 2.425b
72.73a
77ab
Trifolium'steudneri
2150~
2625b
78b
0.102b 1.903b 1.475b
77.93a
61b
Avena sativa (oats)
5813a
5325a
73b
0.471a
8.827a'
6.300a
-
-
*Values withincolumns followed by the same letter are not significantly different at 5% level according
to Duncan's Multiple Range Test.
I
UJ 0
Y
\\ -c
(7,
ca
I
I
u ccr 2: 2: L h 7-u
48
L aJ
- IA
!
_
--
-
.--
.._
Table 4 : Effect of previous cropping and nitrogen application
on grain yield of wheat, Debre Zeit, 1986. .
Previous cropping
kgN/ha
1985
-
Vicia dasycarpa
2612a*
2810a*
Medicago polymorpha
2531a
2280bc
Trifolium steudneri
2600a
2661ab
Fallow
2090ab
2046~
Avena sativa
1753b
2119c
*Valueswithin columns by"the same letter are not significantly
different at 5% level according to Duncan's Multiple Range Test.
;.,
,,
.
,l_
.-
.,
I
( /
I,
t
.:\\
:..
I,.
:
:
:
j,
t 1 I ,,.. ;,rz+>-
. . ‘,. . . .
.
.
., i..;<.. : .:l ,. i :
.,,
:
,.
.
:
.,
50
Nitrogen yield of wheat straw and graln. In this investigation, we
wish to quantify the residual benefit in N derived from legume crops. It can be seen
from Fig. 3 that between 1 and 8 kg/ha of additional N are taken up by wheat straw
following legumes aud fallow compared with those following oats. Pmceding vetch crop
resulted iu 53% higher N uptake by wheat as compared with amtroL
ItisclearfromPig.3thatbetween
11 and23kglhaofadditionalNaretaken
up by wheat grain following legumesand fallow compared with following oats.
Vetch
produced 58% higher N uptake as compared with control. Nnadi and Haque (1988b) reported
that vetch species could corm&me .a considerable amount of N to following crops even
when tops are not~retumed to the:‘&iL Minemkation of root N occured in vetch species
where root N content exceeded 296.
-..
Conclusion
.-
The results of the present study could be of importance wherefarmers cannot afford
to apply fertihzers which may often be unavailable, The choice ,+ to which is the most
beneficial forage to be grown should depend on feed production, nuuitive value for animals,
N economy and rotational value for following crops.
.:, : ‘z
Recommendation for cereals in Ethiopian highlaudsjs ;2l&l$j;.j (EPID, 1975a).
OFF, @Js y$ ~XXEJS on +mf=’ fields W
show np t.0 80%
increak in yield over current recommendations with applicatioqof q.kg N/ha (Christensen,
1973 ; EPlD, 1975b). The amount of N fixed and Contributed by vetch used in this study
could raise substantially the yield of cereal crops and livestockproduction on upland
soils ill Ethiopian highlands.
::
.-:
Acknpwledgements
.
.-I
-..
‘Ihe author express his appreciation to Mr,,Peter Ibara, Ato Yirsaw Wubete, Ato
Muhommed Yasin, Wzt. Abeba G&tom. and WzC Tsehay~Aberra and other staff of Soil
ScieIlceandPlantNuhritionSectionf~technicalassistance,andtotheSel’beEsdorfL,
International Atomic Energy Agency, Vienna for the ‘?N analysis.
:
_
.
.
._
.:
.,.
.
_’
..
,.
;.i,..
.!
”
.:
Straw
Grain ',
I
I
fzzrzm-
N yield (kg ha')
100-r
V.D = Vicia dasvcarcta
M.P = Medi,cago boly'morpha
,.
“..
.
. 1
1 j
= Avena sativa
60
40
20
0
Pb.Fceding crops
Figure 3:
Effect of preceding Crops on nitrogen yield of wheat straw end grain
in unfertilized plots, Debre Zeit, 1 9 8 6 .
51
References
ANDREW, E.D.1965 The effect of some pastme species on soil structure. Aust.‘.T. Expl.
Agric. Anim. Hush. 5. 133-136.
.:
AYOUB, A.T. 1986 The potential contriiution of some f&age legum&oth~~nitrogen
budget and animal feed in the Sudan Gezira farming system. -In : Potential~of forage
legumes in farming systems of sub-Saharan Africa.. Eds. Hague, L, Jutzi, .S.C. and
Neate, PJ.H. p. 5868. Pmczdings of a conference held at ILCA, Addis Ababa,
Ethiopia, 16-19 September, 1985.
.:
G&, . ,/.1 I ., I .,,i;
CHRISTENSEN, II. 1973 Fertihxr and variety triaIs~&idfdemonstmtions:in~Ethiopia,
’
197241973. Extension and Project Implementation deprtment (EPID) Pd. No. 10.
Minstry of Agriculture, Addis Ababa, Ethiopia.
DANIEL, G. 1977 Aspect of climate and water budget in Ethiopa. Addis Ababa University
Press. 71 p.
EXTENSION and PROJECT IMPLEMENTATION DEPARTMENT @PID). 1975a Hand-
book for Agronomy of Crops. EPID Pd. No. 10. hdinistry of Agriculture, Addis
Ababa, Ethiopia.
:
^
/
EXTENSION and PROJECT IMPLEMENTATIONDEPARTMENT@PID). 1975bResuhs
,. .; ‘?“.<’ ‘I, j:.7i. i *’
of EPID PubL No. 31. Minis&y, of Agriculture:~ ~~<~Abjb$~~Ethiopia$$?~~~-
HAQUE, I. and TOTHILL, J.C. 1988 Forages and pastmes-in mixed &mingsystems
of sub&haran Africa. In I Land management and managementofacidsoilsinAf.iica.
Eds. Latham, M. and Ahn, P. p. 107-13 1. Pnxxdings ofthe Second Region&Workshop
on Land Development and Management of Acid Soils in Africa held in Lusalm and
Kasama, Zambii 9-16 April, 1987.
HARDARSON, G. 1984 Biological nitrogen fixation-of grain-legumes : The use of =N
methodology to assess nitrogen fixation and guidelines’ for improvement of nitrogen
fmtion in grain legumes. FAO/IAEA Agricultural Biotechnology Laboratory, Sei-
bersdorf, Austria.
INTERNATIONAL FERTILIZER DEVELOPMENT CENTER (IFDC).1986, Fertilizer si-
tuation : Africa. IFDC Special Publication, Alabama, USA.
KAMARA, C.S. and HAQIJE, I. 1988 Soil moisture storage along a toposequence in
the Ethiopian highlands. Iti : Management of vertisols in sub-Saharan Africa Eds.
Jutzi,S.C.,Haque,I.,McIntire,J.andStares,JE.p.183-20O&oceedmgsofaconference
held at ILCA, Addis Ababa, Ethiopia. 31 August to 4 September, 1987.
KEATINGE, J.D.H.. CHAPANIAN, ‘N. and SAXENA, MC. 1988 Effect of improved
management of legumes in legume-cereal rotation on field e&mates of crop nitrogen
uptake and symbiotic nitrogen fixation in Northern Syria. J. Agric. Sci. (Cumb.). 110,
651-659.
MOHAMED-SALEEM, MA, SULEIMAN, IL and OTSYINA, RM. 1986 Fodder banks
for pastorahsts or farmers. Zn : Management of vertisols in sub-Saharan Africa. Eds,
Jutzi. SC., Haque, I., McIntire. J. and Stares, JE. p. 420-437. Proceedings of a
conference held at ILCA., Addis Ababa, Ethiopia. 31 August to 4 September, 1987.
MOORE, KJ., HERRIDGE, D.F., STARR, G. and DOYLE, AD. 1982 Evidence for
disease control as a factor in improved wheat yield in a 1upincereaI rotation. Proceedings
of the 2nd Australian Agronomy Conference, Wagga, Working paper. p. 327.
.:
:
z.
,:
.I_
:.
..),.’
;
....-..
j_
..,
,: ._: . . . .
:
52
h4USA. UM. and BURHAN, H.O. 1974 The relative performance of forage legumes
as rotational crops in the Gezira ExpZ. Agric. 10, 131-140.
NNADI, LA. and HAQUE. I. 1988a Forage legumes in African croplivestock systems.
.,;%r JL(X B~u&.$$: IO+: ._ :.
NNADI, LA. and HAQUE, I. 1988b Root nitrogen transformation and mineral composition
in selected forage legumes. J. Agric. Sci. (Camb.). 111, 513-518
OSMAN, AE: and NERSOYAN, N. 1986 Effect of the proportion of species on the
yield and quality of forage mixtures and on the yield of barley in the follckng year.
:Expl..tyl@;~22;z:r;5-351;,
PAPASTYLtiOU, I. 1987 Effect of preceding legume or cereal on barley grain and
nitrogen yield. J. Agric. Sci. (Camb.). 108, 623-626.
PAPASTYLIAN~~, I. and SAMIOS, Th. 1987 Comparison of rotations in which barley
for grain folloqrs wollypod vetch or forage barley. J. Agric. Sci. (Cumb.). 108, 609-
6 1 5 .
PAPASTYLIANOU, I.,‘lkXRIDGE, D.W. and CARTER, E.D 1981 Nitrogen nutrition
of cereals in short term rotation. I. Single SeaSOn treatments as a source of nitrogen
‘.
for sul@eq~$~$&xeal ,xrops. Au&. J.’ Agric. Res. 32, 703-712.
<.
S~G&C&$&;f&$& RB: 1985 Contr&tion of legume & he fern nitrogen
economy in a maize based cropping system. J. Agric.Sci. (Camb.). 105, 491-493.
STOVlZ+AN, T.C. 1973 Soil structure changes under wheat belt farming systems. J. Dept.
Agric. K A&. 94, 209-214.
:
1,
FIXATION. BIOLOGIQUE DE L’AZOTE
CHEZ LES ARBR+ES
.
hkhiser la FBA pour la production agricole et foresti&e en Afiique.
Effect of the addition of phosphorus
and Rhizobium inoculum on the nodulation
and the growth of two species of gum arabic
trees Acacia Senegal L. Willd
and Acacia la&a R. Br. ex Benth
BADJI, SP, THOEN, DP,
DUCOUSSO, M.0’ and COLONNA, J.P.a’
mLabo&oire & Microbiologic des Sols, ORSTOiU B.P. 1386, Dakm, Sknegal ;
m Fondion Luembourgwise ;
w ISRA:DRPF, BP 2312. Dakar, Senkgal.
Abstract
Four trea$~ were applied to seedlings of Acacia senegal and Acacia laeta
grown in containers filled with sandy soil poor in available phosphorus (11 ppm.).
Compared .to the con@ Rhizobium inocdum treatmmt induced rm incrwxq not
always statistically s&&ant, of the measured parameters of growth and nodulation.
Thcfoliarandnodulebiomasses
of Acacia Senegal were increased upto 132% and
283% nsptctively. and upto 21% and 66% for Acacia &eta
‘;i.
Rhizobium~andaddirionof30or60ppmPenhaficedvays~~.
antheparrrme~ofgrowthandnodulationoftfietwogumarabicspacies.Inthc
ease of Acacia senegal, the addition of phosphorus W the growth 7
~5~to2oo4b.tbefoliarbiomassabout4oo46.thenodulatianptaametcrJfrom
1100% to 1400%. In ampmkm, forAcacia kzeru, the addition of phosphmns increa-
sed the growth parameters from 40% to 131%. the folk biomass about 15?% and
the nodulation parameters i?om 250% to 350%. However, there was no difference
between these both treatments. Addition of 30 ppm P per plant seems to be suffi-
cient to assure optimal response.
Further research is planned to determine the complete response curve of both
trees to phosphorus in the presence or absence of Rhizobium.
.’
._..
.‘:
.-‘.
55
Introduction
Among the roughly one thousand acacia species (New, 1984), Acacia Senegal
was used, like many other woody leguminous trees, to reforestation in the Sahelian zone.
(Badji et aI., 1989).
The Sahalian ecosystem has been severely degraded due to the combined effects
of an extended period of drought and the excessive exploitation of natural resources by
man and his livesto&. The nahd population of Acacia Senegal has decreased dramat-
ically. This species is an ecologically beneficial tree because of its important role in the
reconstitution of the ‘vegetative cover of the zone.
Acacia senegul is able to fix atmospheric nitrogen through rhizobial nodulation
(Basak and Goyal, 1980 ; Dreyfus and Dommergues, 1981), and hence may contribute
to the improvement-of the soil. Found natumlly throughout the semi-arid Sahel, this species
seems to be well adapted to the edaphic and climatic conditions of the region (Giffard,
1975). The tree has many useful products for local populations, such as aerial fodder,
firewood, wood for tools and construction. Moreover, the tree exudes gum arabic, a product
of considerable economic value.
It may lx possiile to introdue the Acacia laeta in Senegal, onother gum arabic
producing species commonly found m t$ Sudan ; the major gum arabicexporting countries.
Because a pdsitive effect of p;hbSjhoruson Acacia holoseticeaand &zcia ra&&na
has been shown earlier in Senegal (Comet and Diem, 1982 ; Cornet er al. 1982). the
aim of this study is to determine the influence of the addition of phosphorus with R&o-
bium inoculum on the nodulation and the growth of these two acacia species grown on
sandy phosphorus poor soils found in the SaheL
Materials and methods
The seeds of Acacia Senegal belonged to the Mlight grey>, phenotype (Dione, M.
pers. comm). As with the seeds of Acacia &ta, they were furnished by the acDirection
des Recherches sur les Productions Forestieres~ (DRPF) of the &stitut Senegalais de
Recherches Agricoles~ (ERA). The seeds were putted in concentrated sulphuric acid for
14 mn, rinsed thoroughly in sterile water, pregerminated in sterile vermiculite, and after-
wards transferred into individual containers filled with’soil sampled from sand dunes Ioca-
ted near Kebemer. This soil has a sandy texture (94,3% sand) and a low available phos-
phorus level (11 ppm,). Each container was filled with a 1.5 kg mixy of soil and polys-
t y r e n e m (2..,‘v/v)s
:: .3F& :,“i..’
..,
- The factorial experiment, with randomimtion within the blocks~ had three replicates
.and three contmlled factors. The first factor is species with two levels : the Acacia senegal
and the Acacia ha. The second is soil sterihzation, also divided into two levels : non-
sterilised soil and autoclaved soil (120°C 60 mn). The third control factor is fertilization
with four levels : 1) no inoculation .; 2) rhizobii inoculation ; 3) rhizobial inoculation
plus 30 ppm of phosphorus ; 4) rhizobial inoculation plus 60 ppm of phosphorus. Phos-
phorus is supplied as KHQO, in single application. Each elementary plot contained fours
plants. The mean values for each plot was used in the statistical analysis.
At the transfer to the greenhouse, the seedlings wexe inoculated with 2 ml ofRhizobium
suspension (109 cells/ml). A fast growing strain (ORS 1007) selected previously Q3adji
er al., 1988) was employed. It was isolated from nodules of Acacia Zaeta and it is effective
on the two acacia species. The cross nodulation was demonstrated (Badji et al. 1988).
56
Twenty five weeks after planting (WAP) nodulation and growth were recorded
using the following parameters : diamqr at the collar (A); height of the main stem (B);
total length of branches (C!); number of shoot nodes (D); dq biomass of the leaves Q;
number (Fj and dry weight (G) of the nodules.
;-
:
I
.,
R e s u l t s ahd’,‘d&ss& ’ ”
I
Species effect ‘.
Table, 1 shows that .dming :the fkst months of growth and development, Acacia
ZaetuandAcaciasenegalpresentedsomesignificantdifferenceswith~~totheparameters
A, B, C, E, and F. One non significant increase occm-ed for number of shoot nodes (D)
and dry weight of nodules (G).
Table 1 : Differences in growth and nodulation between 25 weeek old seedlings
of Acacia Senegal and Acacia Laeta
i”
Measured Parameters*
Growth
Nodulation
SPECIES
‘,. A.
6
ci
:c
D
E.
F
G
A'!- - - - .~
Acacia
4.85
24.50 40.90 -35.38 1558
14.06
307
Senegal
b
b.
b
-
b
b
-
(AS)
Acacia
5.61
35.85
63.25 38.17 2356
17.09
309
Senegal
a
a
a
-
a
a
(AS-AL) 100
e----------_
--t16
t46
t55
+ 8 +51
+22
+l
AS
. ..
* per plant, A = collar diameter in mm ; B = height of plant in cm ;
C = total length of stem and branches in cm ; D = number of shoot nodes ;
E = dry weight of leaves in mg ; F = number of nodules ; G = dry weight of
nodules in mg. For each parameter data in each column without letter or
followed by the same letter are not significantly different (P f 0.05 -
Newman Keuls test).
57
Soil sterilization effect
For both species and for all growth parameters (A to E), the values were higher
on non-sterilized soil than on sterilized soil (Table 2). However, these differences are
not significant for the collar diameter (A) and the height (B). They are significant for
total length of stem and branches (C) and, in the case of Acacia Zueta, for the number
of shoot nodes (D) and the dry folk biomass Q. The differences were not significant
for the number of nodules (F) and for the nodule dry weight (G). To conclude, the effect
of soil sterilization was rather depressive but not significant (Newman and Keuls test,
p = 0.05). Soil sklixation lolled the native rhizobium and the endomycorhizal propagules;
this might explain the ‘depressive effect.
Fertilization effect
The statistid analysis showed that generally, the effects of the fertilizing treatments
were very significant (Table 3) on the growth parameters (A to E) and the noduIation
parameters (F and G.). Acacia senegal responded better than Acacia Zaetu. There was
no interaction between soil sterilization and fertilizing treatments.
Inoculation with rbirobium, in comparison with the control induced an increase
of both growth and nodulation parameters with the exception of parameter B for Acacia
laeta. However, these increases were not always significant. The significance threshold
was reached only for the pammetexs A, F and G by Acacia Zaetu and for B, C, D, and
E by Acacia Senegal. Bhizobial inoculation plus 30 ppm or 60 ppm of phosphorus induced
important increases for all the parameters. The values reached were significantly different
from those of the other treatments. No significant differences occurred between the phosphorus
treatments for both acacia species. One may suppose that in this case, the economically
optimum level for the application of phosphorus is 30 ppm.
Table 4 shows the percentage of increase of all parameters for the treatments 2,
3, and 4 in comparison with the control treatment. They were low for treatment 2, they
ranged from 20% to 132% for the growth parameter for Acacia senegd and no more
than 21% for Acacia Zuetu. For the number and dry weight of nodules, the increase ranged
respectively from 190% to 283% and 66% to 142%.
For treatments 3 and 4, with phosphorus supplements to the rhizobial inoculation,
these increases were higher ; for the growth parameters (A to D) of Acacia senegul they
ranged from 50% to 200% and 35% to 131% for Acacia Zueta. The gain for the dry
foliar biomass was 506% for Acacia senegal and 177% for Acacia Zueta. For the number
(F) and dry weight (G) of nodules, the mean increase was approximately 1250% for Acacia
Senegal and 300% for the others species. Given what is lmown about the role of phosphorus
in the structure of essential molecules, in the energy process, and in the metabolism of
the plant and recent new information (Bielesky, 1973 ; Heldt et aZ., 1977; Preiss, 1982 ;
Mengel, 1984), this effect of phosphorus is not surprising. On the phosphorus deficient
soil from Kebemer, P fertikation combined with by Rhiwbium inoculation had a positive
effect on the growth of both gum arabic trees and probably on N, fixation. Indeed the
phosphorus treatments largely increased the nodulation. As nodule formation and nitrogen
fixation require energy and thus ATP, one might assume that the increase of the available
phosphorus level in the deficient soils induces a better functioning of the rhizobial symbiosis.
Table 2 : Effect of sol1 sterilization on growth and nodulation parameters of Acacia Senegal and Acacia Taeta
Measured Parameters *
Growth :
Nodulation
Species
Soil sterilization
(factor 1)
(factor 2)
A
B
C
D
E
F
G
'--
a. sterilized soil
4.69
24.10
3'7.86b
33.50
1462
14.32
281
3
Acacia
b. non sterilized soil
5.01
24.99.
43.93a
37 .'27
1654
13.80
328
Senegal
( a -b). 100 / a
tll
to4
+16
tll
+13
- 0 4t17
a. sterilized soil
5.56
35.13
59.13b
34,66b
1958b
18.64
304
Acacia
b. non sterilized soil
5.66
36.56
67.36a
41.6ga
2755a
15.53
307
laeta
(a-b).lOO/a
+02
to4
t14
t20
t41
-17
to1
* As in Table 1
; each data for each species and in each column without letter or with the same letter are not
significantly different (P = 0,05 - Newman Keuls test).
,,
_.
.
.
.
.
.
...:.
,-
. . __ .
.
.)
:
~.
_
59
Table 3 : Effect of fertilizing treatments on the growth and nodulation
of Acacia-Senegal and Acacia Laeta
Measured Parameters*
Species
Fertilizing
treatments
A B
C
.D
E
F G
-
-
-
-
-
-
-
A. Senegal
No inoculation
3.63b 17.00~ 19.5Oc 16.83~ 423~
1.67b 39.6b
Rhizobial ino-
4.35b 22.97b 30.93b 25.87b 981b
4.85b 151.5b
culation
Rhizobial ino-
5.61a 28.55a 56.93a 50.90a 2232a 25.85a 475.5a
culation + 30
PpmP
Rhizobial ino-
5.73a 29.07a 55.28a 47.27a 2566a 23.43a 552.4a
culation + 60
Ppm p.
A. laeta
No inoculation
4.52~ 30.17b 40.17b 23.17~ 1310~
5.5oc 117.4c
Rhizobial ino-
5.14b 29.35b 47.13b 28.17~ 1580~ 13.32b 195.4b
culation
Rhizobial ino-
6.34a 42.75a 79.40a 47.38b 2840b 26.82a 462.6a
culation + 30
tvm P
Rhizobial ino-
6.35a 40.85a 85.42a 53.43a 3631a 22.37a 444.9a
culation + 60
Ppm p
* Asin Table 1 ; for each species, in each column, data follow by the same
.
letter are not significantly different (P = 0,05 - Newman Keuls test).
i‘“‘.
1 :
~
-
__
. , I . ,
_,,
,
_
, .
,?.
”
- .
,.
::-
.;;:..
.(,
.:
:
.
.:
60
-
--
.,_- ---- :. ._. ~.
_
; Table 4 : Variations in percentage for each parameter due to fertilizing
treatments (2, 3, 4) versus control (11
Measured Parameters*
j
::
j' Species
Fertilizing
treatments**
A
8
C
D
E
F
G
-
-
-
-
-
-
' A. Senegal
2
20
35 59 54 132 190
283
3..
55
68 192 202 428 1447 1100
4
58
71 183 181 506 1302 1294
A. laeta
2
14
-3
17
22
21
142
66
3
40
42
98
‘1 04
117
388
294
4
40
35
113
131
177
307
279
* Parameters A.to G as in iable 1
;
**Fertiiizing treatments as in Table 3.
)
_.
I
.:.
.
..’
”
y”
..,-
.-*
61
Further investigations are planned to determine the influence of lower phosphorus
levels on the growth of acacia in the presence or absence of rhizobii inoculation. Such
experiments should help to solve the problem of costs and benefits of phosphorus appli-
cation during afforestation programmes.
References
BADJI, S., DUCOUSSO, M., GUEYE, M. et COLONNA, JP. 1988 Fixation biologique.
de l’azote et possibii de nodulation croisk chez les deux espkes d’acacias producteurs
de gomme dure : Acacia senegal L. Willd. et Acacia laeta R. Br. ex Benth. CR.
Acud. Sci., S&. IJI. 307 (1 l), 663668.
BADJI, S., DUCOUSSO, M., GUEYE M, et COLONNA, JP. 1989 Effets de l’inoculation
par diverses souches de Rhizobizun du S&r&al sur les deux principaux Acacias gom-
miers en culture semi-aseptique. In : &I&w Symposium SUF le Go&r et la Gomme
Arubique (SYGGA III), 25-2-8 Oct. 1988, St. Louis du Senegal. Collections Actes de
I’ISRA, p. 171-179.
BASAK, UK. et GOYAL, SK 1980 Studies on tree legumes. II. Further additions to
the list of nodulating tree legumes. Plant and soil. 56, 33-37.
BIBLESKI, R.L. 1973 Phosphate pools phosphate transport and phosphate availability.
Ann. Rev. Plant Physiol. 24,225252.cENTRE DU COMMERCE IN’IFRNATIONAL
CNUCED. 1978 Le marche de la gomme arabique, CNUCED Ed. Geneve, 181 p.
CORNET, F. et DIEM, H.G. 1982 Etude comparative & l’effkaciu5 des souches de Rhizobium
d’ Acacia isolees de sols du Senegal et effet de la double symbiose Rhizobium - Glomus
mossae SUT la croissance d’Acacia holosericea et A. raaWana. Bois et For&s &s
Tropiques. 198, 3-15.
CORNET, F., DIEM, H.G. et DOhIMBRGUES, Y.R. 1982 Effet de l’inoculation avec
Glomus mossae sur la croissance d’Acacia holosericea en p&pin&e et apr&s trans-
plantation sur le terrain. In I Les mycorhizes : biologie et utilisatior~ Colloque de
I’INRA n”13, 5-6 Mai 1982, Dijon, Ed. INRA Publ. pp. 287-293.
DREYFUS, B.L. et DOMMERGUES, Y.R. 1981 Nodulation of Acacia species by fast
and slow growing tropical strains of Rhiibium. AppIEnvironBioZ. 41, 97-99.
GIFFARD, P.L. 1975 Les gommiers, essences de reboisement pour les regions sah&ennes.
Bois et Fon%s des Tropiques. 161, 3-21.
-HELDT, RW., JA CHONG, C., MARONDE, D., HEROLD, A., STANKOVIC, Z.S.,
WALKER, D.A., KRAMlNGER, A., KIRK, MR. et HEBER, U. 1977 Role of or-
thophosphate and other factors in the regulation of starch formation in leaves and
isolated chloroplasts. Plant physiol. 59, 1146-1155.
MENGEL, K. 1984 Uptake of mineral nutrients and their biological functions. In. VIeme
Collocpe International pour l’Opt.imisation de la Nutrition des Plantes. 2-8 Sept 1984,
Montpellier, public par AIONP/CIRAD et ACCT, pp. 1495-1524.
NEW, TR. 1984 A biology of Acacia. Oxford University Press, Melbourne, 153 p.
6 2
PREISS, J. 1982 Regulation of the biosynthesis and degradation of starch. Ann.
Rev. Pfmr
Physid. 33. 431454.
WALKER, D.A. 1980 Regulation of starch synthesis in leaves : @e role of orthophosphak
In. Physiological aspects of crop productivity. Intern. Potash Institute, Beane, pp.
195-207.
&icole et Foredre en Afrique
I
_:
i.’
.
Nodulation survey ~~6~:~~&i&1$nous
and non legum&&i trees’ growing
at ICRAF’s field station
ODUOL, PA, AKUNDA, E.M.
and WAMBUGU, PX.
ICRAF P-0. Box 30677 Nairobi, Kenya
Introduction
Substained productivity, of &d is very much dependent on nitrogen input to the
soil from both biologiqal and non-biological sources. Legiunes have been used in building
and c3mmving soil fertilitj since thk beginning of agriculbnz because of their ability
to NIX nitrogen from the atmosphere. It is estimated that this symbiotic process between
bacteria and host plant accounts for approximately 20% of the total nitrogen fmed an-
nually in the world (Dazzo and Hubbel 1974).
Most legumes grow as wee+ (shrubs and trees), thus very little knowledge in
the extent of nodulation is known. The presence of nodules is the prelinkury and visual
confirmation of nitrogen fixation in legumes. Nodulation among legumes is affected by
64
several factors among which include soil temperature (Santo and Dobereiner, 1970 ; Habiih,
1970 ; Gibson, 1977), soil moisture (Habish. 1970), absence or presence of excess minerals
(Jutse and Haque, 1984). All these factors and the presence of the right Rhizobium affect
the rate and position of nodulation.
About 49 trees (Table 1) grow at ICRAF’s f=ld station (some indigenous and
introduced). Most of these species are likely to be used in agroforestry interventions which
focus on fertility restoration. Up to date, no quantitative or qualitative information on
the nodulating ability of these species has been given under the sub-humid, moderately
acidic condition of Machakos field station, the center of ICRAF’s field trials.
This paper reports he findings of the noduktiug survey of some leguminous (47)
and non-leguminous (2) trees being gr&n ‘at the field station in a sub-humid and serni-
arid environment.
Materials and methods
The study site lies between l”33”S latitude and 37”lS’E longitude at an altitude
of 156Om in the sub-humid to semi-arid zone with bimodal annual rainfall of about 700mm.
Mean annual air temperature of 22”C, with deep soil profile, well drained, moderately
leached, weakly acid soils with pH rate 6.0-6.5 and medium nutrient levels. Soils gene-
rally classified as al&ols.
Forty nine plants were examined for the presence of nodules. These plants were
leguminous and non-leguminous woody shrubs and trees Some of these 49 species have
been reported to nodulate in various reports (Table 1) (Corby, 1974 ; Allen and Allen
1981). The species examined were both indigenous as well as introduced exotics
(Table 2).
Periodic sampling,,for nodules was carried,out during the wet active growing sea-
son (and’dry season) of the ‘kees/sbrubs. This was done on diverse dates during April
and May in 1986, 1987 and 1988 after digging up a hole near the tree trunk to expose
the root system. The excavated roots were washed under runuing tap water and the root
systems were floated in trays of water to facilitate detection of nodules. The presence
or absence and degree of nodulation on the root systems were noted. The morphology
of mature nodules was described and their sizes (diameter) measured in mm. Special
care was taken to distinguish root nodules from root malformations, such as those caused
by nematodes, insects or other root inhabiting pathogenic microorganisms, by disecting
the nodules to establish whether they are red, brown or green iu internal colour.
Results
Observations on nod&ion are presented in Table 2, which also lists whether the
plants are indigenous or introduced species The mcorded presence of root nodules on
the examined species was checked against published lists of Allen and Allen (1961,1981),
Corby (1974), Lim (1979) and Mohammad and Mahmoud (1985). Table 2 also presents
species which nod&ted during the three seasons of sampling from May 1986 to May
1988. Nodulation octm-ed in 24 species, represented by : Mimosoideae, Caesalpinoideae,
Papilionoideae, and non-legumes representing Simaroubaceae and Casuarinaceae. Majority
of the species were moderately to heavily nodulated, some were very heavily nodulakd,
specially Gliricidia sepium, Leucaena leucocephala, Samanea sanuq Sesbania se&an,
65
- - .
-I__--_
_...._
-__-
_
&=.a-.--
__---.-.~
.._
.,.
-.-
-_-
Table 1.; .-
List of Nitrogen fixing trees growing at ICRAF's Field Station
Machakos - Kenya
Nodulation
Species
Present
Previous Record
Study
(in literature1
Acacia albida
Acacia deamii
y'
Acacia farnesiana
Y
Acacia holosericea
NI
Acacia mangium
;
Y
Acacia mellifera
Y
Acacia paniculata
N I
Acacia pennatula
Y
Acacia stuhlmanii
t
Y
Acacia saligna
t -
Y
Acacia tortilis
+
Y
Acacia victoriae
Y
Acacia xanthoplea
Y
Albizzia longepedata
**Alvarodoa amorphoides
t
!
Ateleia herbert-smithii
NI
Caesalpinia velutina
+
NI
Caesalpinia coriaria
-
NI
Caesalpinia eriostachys
t
\\’
Crescentia olata
;:-. ..-;::, ;:- ,
Cajanus cajan
-I-
Y
Calliandra calothyrsus
t
Y
Cassia alata
B
Cassia siamea
Y
*jCasuarina equisetifolia
Y
i
Cordeauxia edulis
Y
Enterolobium cyclocarpmum
Erythrina abyssinica
Gliricidia sepium
Haematoxylon brasilleto
Leucaena diversifolia
Leucaena shannoni
Leucaena leucocephala
(Cunningham).
Leucaena leucocephala (Peru)
+
Leucaena leucocephala
t
(Hawaiian giant - K81
Mimosa tenuiflora
Myrospermum frutescens
Parkinsonia aculeata
Pithecellobiun? dulce
Prosopis juliflora
Prosopis nigra
Table 1 : (continued)
Nodulation
Species
Present .i
Previous Record
Study
(in literature)
Prosopis pallida
Y
Prosopis alba
Y
Prosopis chilensis
Samanea saman
+
::
Sesbaniamacrantha
+
Y
Sesbania grandiflora
+
Y
I
Sesbania sesban
+
i
Senna atomeria
NI
r* iWi~ieguminous
+ Presence of nodules
- absence of nodules
Y Nodulation previously observed in other studies
B Species previously investigated but nodulation never observed
NI No information available.
68
Sesbania macrantha and Sesbania grandiflora. Heavy nodulation ocurred in species Alvarodoa
amorphides, Casuarina equisetifolia, Calliandra Calothrysus, Erythrina abyssinica,
Enterolobium cyclocarpomum and Pithecelobium dulce. Sparse to moderate nodulation
oaxm-4 in species of Acacia saligna, Acacia tortilis, Acacia stuhlmanii, Prosopis jul#lora.
&e.&pinia velutina, h?Sa@da eriostachys, ittxcae~ diversifo~ia. kucaena shamoni,
provenanees of Lacae~ leucocephala (Hawaiian giant K.8) and Cunningham and Mimosa I
tenufflora and Myrospemum frutescem
Most of the nodules were located on the upper parts of central tap root systems
in clusters as well as on lateml roots mg.1). Most of the young nodules were rcknd,
white and smooth-surf&xl whi@ .older nodules were of various shapes, mostly brown
in internal colour and rough-surfaced. Nodule shapes were classiied into different types
as described by the drawings of Lim and Ng (1977). The nodde shapes in Mimosoideae
were f&n shaped, elongated, lobed, globose and bifurcate. cardlloid types were found
in Sbnaroubaceae and Caskeae. In Caesalpinoideae, the nodules were globose. and
in Papi~Yonoideae the nodule shapes were globose and fan-&p&d (Table 2). Fig. 2 shows
globose shaped nodules for &sbania sesban occuring in clusters.
Table 2 also shows average diameters for mature nodules. The nodule diameters
are way variable, ranging from 1.7 mm in Lmcaem leucocephala Peru) to large diameters
of 4 mm in Mikosa ten@orti .and Sesbania granal.@ora.
.:
,~
:
, ~
;,;..,
:,.:::::
,,
.
.
. . .
.
.
‘,’
I
.,.:
68
Sesbania macrantha and Sesbania grandifora. Heavy nodulation ocurred in species Alvar&a
amorphides, Casuarina equisetifolia, Calliandra Calothrysus, Erythrina abyssinica,
Enterolobium cyclocarpomum and Pithecelobium aklce. Sparse to moderate nodulation
occurred in species of Acacia saligna, Acacia tortilis, Acacia stuhlmank Prosopis julipora,
Caesalpinia veluthuz, Caesalpinia eriostachys, L+eucaena diversifolia. Leucaena shannoni,
provenances ofLeucaena leucocephala (Hawaiian giant K.8) and Cunningham andMimosa
tenuiflora a n d Myrospermum jktescens.
Most of the nodules wexe located on the upper parts of central tap root systems
in clusters as well as on lateral roots (Fig.l). Most of the young nodules were rotid,
white and smooth-smfaced while older nodules were of various shapes, mostly brown
in internal colour and rough-surfaced. Nodule shapes were classified into different types
as described by the drawings of Lim and Ng (1977). The nodule shapes in Mimosoideae
were fan shaped, elongated, lobed, globose and bifin-cate. Carolloid types were found
in Simaroubaceae and Casuarinaceae. In CaeSalpinoiakae, the rqdules were globose, and
in Papilionoideae the nodule shapes were globose and fan-shaped (Table 2). Fig. 2 shows
globose shaped nodules for Sesbania sesban occuring in clusters.
Table 2 also shows average diameters for mature nodules. The nodule diameters
are very variable, ranging from 1.7 mm in Leucaena leucocephala (Peru) to large diameters
of 4 mm in Mimosa tenuijlora and Sesbania grandiJora.
Figure 1 : Exposed root system of Sesbania bispirwsa showing nodules confined to the upper parts
of the tap root and in late&l roots.
69
Figure 2 : Sesbaniu sesban globose shaped nodules in clusters
69
;,.
i.
i,
i
;
,.
,’
‘,
.-.
. . -,
1.
‘-.
‘.
;
,$.
i
I
. .._
*
.;
*
;;
.,.
.”
.
.
”
- .
:.
70
Discussion
In this survey, 24 out of the 49 species were recorded for presence of nodules.
However, the absence of nodules on other species is not a conclusive evidence for lack
of nodulating ability. The findings of this survey are in broad-agreement with the surmnary
of the world-index of nodulation by Allen and Allen (1961). However, some of these
species which were not observed to nodulate in this study have been reported elsewhere
to be nodulating (Table 1). Successful nodulation in these plants indicates the presence
of potential infective autochthonous rhizobial strains in the soil. The degree of nodulation
was sparse in some species ; thii can be attributed to the fact that most of these species
were not inoculated at the time of planting (Table 2). Those nodulating species somehow
formed a symbiotic relationship with the indigenous rhizobia in the soil. Conclusive evidence
cannot be drawn on whether there is correlation between those introduced and indigenous
species in nodulation. There was a general absence of nodules on these trees/shrubs
during the dry season. This confirms the inability of species to nodulate under extreme
or adverse environmental conditions. Nodules were generally active during the wet seasons
when there was a general active vegetative and reproductive growth in these species.
However, the nodules during the dry season were generally inactive. Nodulated legumes
and non-legumes can contribute significantly to the nitrogen status of the soil. In order
to exploit the potential of fixed nitrogen, it is necessary to successfully identify nitrogen
fming nodulated legumes and non-legumes.
In this study, the non-legumes Alvaroaba amorphoides and Casuarina equisetifolia
nodulated prolifically at this site, thus indicating a great potential of non-leguminous plants
with nodules possessing nitrogen fixing capacity. In future, more and more nitrogen-fixing
tree species with multiple benefits will be used in farming systems . So, more information
is required on the agronomic and agricultural aspects of nodulated plants to enable the
symbiotic system to be .exploited to its maximum. What is presented in this paper is
an account of the potential contribution of these nodulated plants to the nitrogen cycle
in the biosphere.
The study shows prolific nodulation which varies among species. The next stage
is to select the most prolific nodulating species for quantitative nitrogen fixation studies
to be carried out at different periods of the year.
References
ALLEN, E.K. and ALLEN, 0-N. 1961 The scope of nodulation in theLegminosae. Recent
Advances in Botany. Vol. 1, University of Toronto Press, Canada.
CORBY, H.D.L. 1974 Systematic implications of nodulation among Rhodesian 1egume.s.
Kirlia. 9. 301-329.
DAZZO, R.B. and HUBBEL, D.H. 1974 Biological nitrogen fixation. Soil and Crop Science
Sot. Florida Proc. 34, 71-74.
GIBSON, A.H. 1977 In : A Treatise on dinitrogen fixation. IV. Agronomy and Ecology
(Hardy, R.W.F. and Gibon, A.H. eds.). pp. 350-393. Wiley Interscience, New York,
U.S.A.
JUTSE, S. and HAQUE, I. 1984 Soil plant nutrition and forage agronomy research (F’ro-
ject F’rotocols). Highland Programme, ILCA, Addis Ababa, Ethiopia
7 1
HABISH, D.A. 1970 Effect of certain soil conditions on nodulation of Acacia spp. Plant
and Soil. 33, l-6.
LIM, G. and NG, HL. 1979 Root nodules of some tropical legumes in Singapore. Plant
und SoiI. 46, 317-327.
MOHAMMAD, A. and MAHMOOD, A. 1985 Qualitative study of the nodulating abiity
of legumes of Pakistan. List 3. Trap. Agric. 62 (IQ 49-51.
SANTO, D.M. and DOBBRBINER, J. 1970 Perquisa Agrqxcuaria. Brmileira. 5,
365-371.
Maxim&r la FBA pour la Protfuction Agricok et Foresti&e en Aj?ique
Effect of inoculation- with Rhizobium,
P application and liming on early growth
of Ieucaena (Leucaena Zeucocephalci Lam. de wit).
. .
GICHURU, M. and.MULONGOY, K.
InternatM Ikstitti of Tropical Agricuhre. PMB 5320. lbaahn, Nigtfria.
Abstract
The .effects of P applioation (0 and 50 mg kg’ aoil), inooul&on with soil from
a field with well nodulated plants of LeucaeM Ieu~~~ephaIa, and with a water sus-
pension of a Rhizobiwn strain IRC 1045 specific for leueaena, on establishment of
leueaena were studied in four soils. two Psammentic Ustorthenta, one Oxic Paleustalf
and one Typic Paleudult. Application of P resulted in imProved leueaena growth
and nodulation in all four soils. The sharpest response was obtained in the Psammentic
Ustorthent from a farm where leucaena established Poorly. Soil inoculation with rhi-
zobia resulted in inqxoved leueaena growth increased dry matter yield and nodulation
except ou the Oxic Paleuatalf from the International Institute of Tropical Agriculture
(IlTA) on which leueaena normally paforms welL The aoil inoarkmt performed better
than the water auqension of atrain IRC 1045. Dicaleium phosphate affected the early
growth of leueaena on the TyPic Paleudult because of its liming properties.
Introduction
Alley cropping is an agroforcshy system in which food crops are grown in alleys
formed by hedgerows of suitable trees or shrubs which can be pruned periodically to
prevent excessive shading and which provide mulch and green manure for companion
food crops. Leucacna (Leucaena feztcocephdu var. K 28) appears to bc a suitable hedge
row tree for alley cropping on Alfiils and Inccptisols
of southwestern Nigeria (Kang
and Wilson, 1987). The problems of establishment and of slow initial growth of leucacna
7 3
remain however a major limitation for adaptation alley cropping by farmers in some areas,
particularly when the direct seeding method is used. The poor establishment experienced
in farmers’ fields has been attributed to competition with companion crops or weeds (Jones
and Bray, 1983 ; Atta-Krah and Kolawole, 1987) to lack of effective strains of Rhitobium
for nitrogen fixation (Ruaysoongnem,et al., 1984 ; Sanginga et al., 1988) to P deficiency
(Sanginga et al., 1988) and to acidity in highly weathered soils (Hutton and Andrew,
I
1978 ; Olverra et al., 1982).
The purpose of this study was to evaluate some of the soil factors that are respon:
sible for the poor establishment of leucaena in southern Nigeria. The effect of lime and
dicalcium phosphate on leucaena early growth in an acid soil was also compared
Materials and methods
A greenhouse pot experiment was conducted at the InternationaI Institute of Ti-o-
pical Agriculture (ETA), Ibadan, Nigeria. The experimental design was a split-plot with
four soils, two from on-farm sites near Alabata village, approximately 15 km Northwest
of Ibadan (Psammeutic Ustorthent), one from the IITA main station (Oxic Paleustalf)
Ibadan, southwestern Nigeria, and one from the IITA high rainfall station (Typic Paleudult)
at Onne, southeastern Nigeria, as main plot treatments. Leucaena establishment was rela-
tively good at one on-farm site (farm-l) and poor at the other site @m-2). The soil
fi-om IlTA station usually supports good leucaena growth whereas leucaena performance
in the acid soil is normally very poor. Chemical properties of these soils are shown iu
Table 1. The subplot treatments were two rates of dicalcium phosphate (0 and 50 mg
P kg’ soil) and three inoculation treatments (uninoculated control, inoculation with a water
Table 1 : Selected chemical properties of the soils used for the pot experiment
Soil
PH
Organic Total Bray-l Ca Mg K
Total
C
N
P
acidity
--v-P
- -
%
%
kg -1 ------
w
meq/loog-------------
IITA
6.3 1.49 0.121 11.2
0.74 0.32 0.29 0.16
(Oiic Paleustalf)
Farm-l
5.8 0.65 0.052 7.2
0.53
0.17 0.30 0.05
(Psammentic Ustorthent)
.
Farm-2
5.6 0.51 0.052
2.2 0.52 0.15 0.24 0.00
(Psammentic Ustorthent)
Onne
4.9 1.14 0.089 62.2. 0.73 0.25 0.30 0.36
(Typic Paleudult)
,.-
,f.”
:‘:
.<.’
74
suspension of Rhizobium strain IRC 1045 and inoculation with a soil from a field with
well n&dated leum) in a factorial arrangement, Rhiwbium strain IK 1045 is an
elite strain isolated from .no+les of leucaena grown on an Alfisol in southwestern
Nigeria (S.anginga ef al, 1988). Ten mihiliters of the water suspension of this strain (109
cells ml-l)’ ixp‘g: of the’ ,++l :inoculano, (10’ leucaena rhizobia g-l soil) were mixed
,withtbesurfaci:soil~inthepots,Thefinaweigfitofsoilineachpotwas3kg.
A second pot experiment was carried out to study the effect of lime (CaCOJ and
dicalcium phosphate fextiker on early growth of leucaena in the acid soiL The treatments
wereOand2000mgCaCO,kg’soil,andOand50mgPkg1soilinafactorialarran-
gemIlL
I ”
Six seeds of leucaena (Leucaena leucocephala var. K-28) previously scarified in sul-
furic acid we sown in each pot Seedlings were thinned tb three after one week. The
pots were miformly watered with 100 ml of deionised water daily for the first 6 weeks
and twice daily af&wa&. Bi-weekly height measurements were made until 14 weeks
after planting (WAP). At harvest (14 WAP), above-ground dry matter was determined.
Roots wea-e washed in tap water over a sieve to miuimi~ losses. Nodules were separated
from the roots, counted and dried.
The Statistical Analysis Systems (SAS) package was used for statistical evaluation
of variances. Least significant differences (LSD) are given only when the overall F-test
is signiticant @ = -0.05).
Results and discusslon
Height- measurements
Th~slxqz@al effect of.,P appbcation on leucaena growth was obseaved from 4 WAP
and the effe&w sign&c&t for; the rest of the experiment (Table 2). Application
of 50 mg -P kg’ as dicalcium phosphate increased leucaena height by 18 - 42%. The
largest effect was observed at 6 and 8 WAP.
Table 2 : Effect of P application on leucaena height at different times
(Means of four soils, 3 inoculation treatments and four replications)
Weeks after planting
. .
,. ‘;
J
.
Phosphorus‘,
rate
2
4
6
a
10
12
14
--P---P
-1
“‘g kg
-----------------_-_ (cm) --------------------------- :
0
7
9
13
17
22
27 34
j'
50
7
11
17
24
28
33
40
Ls0(0.051
NS
0.4
0.4
0.8
1.2
1.6
2.6
75
Response of leucaena to inoculation is shown in Table 3. Leucaena growth on the
IITA station soil showed no reponse to inoculation except at 4 WAP. This indicates that
effective rhiiia capable of infecting leucaena were not limiting at this site. In conrast,
growth on other soils showed variable response to iuoculation. ‘I$ two on$$n showed
a delayed response to inoculation compared to ‘the Omie soil but the in$ial gt6vth of
leucaena in the latter soil which is acidic was much slower than in the other soils. The
soil inoculant perform better than Rhizobizun strah IRc 1045 on all soil types. Soil ino-
culauts are also known to supply other useful microorganisms like endomycorrhi# fungi
and some soil nutrients.
,.
The better response to soil inoculant can explain the improved establishment of’tians-
planted leucaena compared with directly seeded plants in areas where leucaeua is imro-
duced for the first time. Soils used in nurseries are likely to be from sites with favorable
nutrieut aud microbiological properties.
Dry matter yield and nodulation
Above ground and root dry matter were significantly enhanced by P application with
sign&am interactions between soil type and P iu root dry matter yield (Table 4). Without
phosphorus application, leucaena growing on Farm-2 soil produced significantly less bio-
mass compared with the other three soils. The sharp response to P by leucaena on Farm-
2 soil indicates that available P was limiting leucaena growth on this soil. This was confir-
mtxl by the soil analytical results (Table 1).
Nodulation was not significantly enhanced by P application except on the acid soil.
Sign&am interactions between soil type and inoculation were noted iu dry matter yield
and nodulation with the acid soil producing much sharper responses compared to the
other soils (Table 5). The llTA soil did not show response to iuoculation There was
a general tendency for increased nodulation with inoculation although the effects were
significant only in the acid soil. Nodule counts were highly variable particularly in the
sandy soils due to the difficulty of collecting the small nodules.
Effect of lime and dicalcium phosphate
In the aid soil, leucaena dry matter yields and nodulation showed significant inter-
actions between lime and P (Table 6). Growth and nodulation were increased by the
application of either lime or dicalcium phosphate. Lime produced greater increases. Extrac-
table P is high in this soil (Table 1). With the application of lime, phosphorus produced
no additional increase in native leucaena growth and nodulation. The explanation for this
is that application of lime to this soil makes the native soil phosphorus more available.
76
Table 3 : Effect of inoculation with a soil inoculant and a water-
suspension of Rhizobium strain IRc 1045 on height of leucaena
on four soils at different times.
Inoculation
Weeks after planting
treatment
2
4
6
8
10
12
14
-----~
- ----_- --------------(cm)---------
---_-_---------_____
I ITA
Control
7
12
16 22
27
33
3 9
Soil inoculant
7
9
15 21
27
34
39
IRc 1045
7
12
16 23
30
35
42
LSD(O.05)
NS
1.8
NS NS
NS
NS
NS
Farm 1
Control
7
11
15 21
25
31
39
Soil inoculant
7
11
15 23
28
33
40
IRc 1045
7
10
14 20
24
27
34
LSD(O.051
NS
NS NS NS
2.6
3.3
NS
Farm 2
Control
7
11
14 17
20
23
28
Soil inoculant
7
10
14 19
24
28
33
IRc 1045
7
11
14 18
22
26
30
LSD(0.05)
NS
NS
NS 1.6
2.3
3.6
NS
Onne
Control
5
8
13 18
22
27
35
Soil inoculant
6
9
15 23
29
34
42
IRc 1045
5
9
14 21
25
31
40
LSD(O.051
0.5
0.8
1.4 2.1
3.1
4.5
NS
77
Table 4 : Effect of P application on dry matter production and
nodulation of leucaena on four soils at 14 weeks after
planting.
Phosphorus
applied
Top
Root
Nodules
Nodule
(mg kg-l)
weight
weight
number
weight
g plant -1
g plant -'
No. plant -1
q plant -1
IITA
0
5
5
34
107
50
7
7
42
100
LSD(O.05)
0.5
0.6
NS
NS
Farm 1
0
5
5
36
_
140
50
6
6
39
110
LSD(O.05)
0.6
0.7
NS
NS
Farm 2
0
3
3
28
60
50
5
5
39
a3
LSD(0.05)
0.3
0.5
NS
NS
Onne
0
5
5
42
a3
50
7
9
74
157
Lso(o.o5)
0.8
1.0
25.0
60.0
78
Table 5 : Effect of inoculation with a soil inoculant and a water-suspension:
of Rhizobium strain IRc 1045 on leucaena dry matter yield and
nodulation on four soils, at 14 weeks#after planting.
Inoculation
Top
Root
Nodule
Nodule
treatment
weight
weight
number
weight
g plant -1
g plant -1
No: plant-'
mg plant -1
,,
IITA
Control
6
6
32
120
Soil inoculant
6
6
42
90
IRc 1045
6
6
40
97
I
LSD(0.05)
NS
NS
NS
NS
Farm 1
Control
6
5
28
87
;
Soil inoculant
6
6
43
157
I
IRc 1045
5 .
5 .
42
130
I
LSD(O.05)
0.7
NS
NS
5
3
Farm 2
Control
3
3
19
53
Soil inoculant
4
4
39
87
IRc 1045
4
4
42
73
LSD(O.05)
0.3
NS
NS
NS
Onne
Control
5
5
30
167
Soil inoculant
7
7
97
153
IRc 1045
5
5
50
97
LSD(O.05)
1.0
1.2
30.6
NS
79
.
I/
I
Table 6 : Effects of lime and dicalcium phosphate on the dry matter yield
,
and nodulation of Leucaena leucocephala in a pot experiment with
i
an acid soil from Onne, at 14 weeks after planting.
P applied
CaC03 applied, mg kg-l
(mg kg-')
0
2 0 0 0
Means
TOPS, (g plant-')
0
4
10
7.0
50
6
10
8.0 SE _f 0.54
Means
5.0
10.0
SE + 0.54
SE (1 cell) = k 0.76 :
CY 21%
t
Roots, (g plant-')
0
4
7
5.5
50
6
6
6.0 SE + 5.53
SE + 0.55
SE (1 cell) = 0.78
CV 27 %
Nodule number (No.plant-')
0
18
112
65.0 SE + 10.0.
50
62
97
79.5
Means
40.0
104.5
+
10.0
E
SE (1 cell) = 14.2
39 %
80
Acknowledgements
The authors thank J. Pleysier and AN. Atta-Krah for their suggestions and E.I. Jegede
for the technical assistance.
References
A’ITA-KRAH, AN. and KOLAWOLE, G.O. 1987 Establishment and growth of leucaena
and gliricidia alley cropped with pepper and sorghum. Leucaena Res. Reports. 8,46-
47.
HUTTON, E.M. and ANDRBW C.S. 1978 Comparative effects of calcium carbonate
on growth, nodulation and chemical composition of four Leucaena leucocephala lines,
Mucroptilium Iathyroides and L&on&s buinesii. Aust. J. Exp. Agric. and An. Hush.
18, 81-88.
JONES, RJ. and BRAY, R.A. 1983 Agronomic research in the development of leucaena
as a pasture legume in Australia. In : Leucaena Research in Asian Pacific Region,
pp. 4148, IDRC, Ottawa, Canada.
KANG, B.T. and WILSON, G.F. 1987 The development of alley cropping as a promising
agroforestry technology. In : Steppler, H.A. and Nair, P.K.R. (eds.) Agroforestry
a Decade in Development, pp. 227-243. ICRAF, Nairobi, Kenya.
OLVERRA, E., WEST, S.H. and BRUE, W.G. 1982Establishment ofkucaena leucocephula
in acid soils. Leucaena Res. Rep. 3, 84-85.
RUAYSOONGNERN, S., SHELTON, H&L and EDWARDS, D.G. 1984 ‘Ihe influence
of lime, combined nitrogen supply and Rhizobixun sbxin on growth, nodulation and
nitrogen status of I!.+?UCU~?M kucocephakz cv. Cunningham. In : I.R. Kennedy and
L. Copeland (eds) The Seventh Australian Legume Nodulation Conference, Aust. Inst.
of Agric. Sci. Occusionul Pub. W 12, 21-22
SANGINGA, N., MULONGOY, K. and AYANABA, A. 1988 Response of L.eucueunu/
Rhizobim symbiosis to mineral nutrients in southwestern Nigeria. Plant and Soil.
112, 121-127.
Maxbnisa la FBA pour la Pmduction Agricok et Fore&&e tn Afrique
_ I
’
.,.
A prelimkxry study on the compatjbility
between indi&nou.s rhizsbia
and some tree? Ieguixiqs in Kenya
ODEE, D. W. r
Kenya Fomstry Rescaxh Indtute. Nairobi, P.OBox 20412, Kenya.
..’
.,.
.
‘,..(
: -
..i
cInm’- stldy with indigenous rhizobia was carried out on seval host s&X-
tics of isolation : Acacia a&&z, Acacia meamrii, Callkmdra caldhyrsuq Lencaena
leucocephala. Prosopis jul$lora. Sesbania graruiifrora ad Sesbania se&an. The Rhi-
zobium strains were highly infective and effectiveness was variable across the host-
strain combinations. There occured four cross-inoculation groups within the range
of Rhizobium strains and host species studied. The two members of the genus Se.sbania
were more specific in their rhimbial partna-s in terms of beneficial ass&a&n Nodule
shape and clieutio? ,was kmktent’onthcrootsystemofcachhostspeciesin
.I. _.
i 1
: effectivc : symtnosis
:*
Introduction
Acacia albida, Acacia meanki. Calliandra calothyrsus. L.eucaena leucocephala,
Prosopis julijlora, Scsbania grandifIora and Sesbania se&an are grown in Kenya to provide
green manure and mulch in intercropping systems, browse or fodder for livestock and
fuelwood. These nitrogen fming trees (NJ?lJ nodulate sporadically with indigenous rhi-
mbia in the forest nurse&s and field conditions (Odee, unpub$shed data).
Despite the inaeasin g use of these NET in the country, little is known about their
symbiosis with the naturally occurring rhizobia. Nodulation in both exotic and indigenous
82
legumes by natural rhizobii has been described as erratic (Keya and Van Eijnatten,
1975 ; Keya, 1977). but most of the work reported on natural nodulation have been of
grain and pasture legumes (McDonald, 1935 ; Bumpus, 1957 : Morrison, 1966 ; Souza,
1969).
This papet reports on the results of cross-inoculation study of selected Rhiwbium
strains on the given host spectrum and dkusses the salient features.
Materials and methods
Eight strains of Rhiwbiwn isolated from geographically different areas were se-
lected for the study. Strain NUM 777 was acquired from the Nairobi Rhizobium h4tRCEN.
Each strain was charactetized by streaking on plates of yeast extract mannitol (YEM)
agar. The composition of the medium was as described by Ber&nsen (1980) as follows
(g 1-l) : manttitol, 10 ; yeast eXtraCl, 0.4 ; K.$IPGd, 0.5 -i MgSOi; 7H.+I, 0.2 ; NaCl,
0.1 and distilled water. 0.5% alcoholic solution of bromothymolblue(BTB) was incorporated
into the medium at the rate of 5 ml per liter for. pH reaction of the Rhiwbium strains.
Streaked plates were incubated at 26°C for 10 days. Growth rates were described
by cultural and colony size (Vincent, 1970). Fast growers achieved moderate to abundant
growth with colony sizes of equal or greater than 2 mm within 5 days. Slow growers
achieved slight to moderate growth with colony s&es of less than 2 mm after 5 days.
The characteristics of the strains are shown in Table 1.
Leonard jar assembks (Vincent, 1970) with vermiculite as the growth medium
were used to grow seedlings. The composition of the nutrient solution for the assemblies
was as described by Somasegaran and Hoben (1985) as follows (pM) : CaC4,2%0 :
1000 ; K.$ZPG, : 500 ; CJXsOrFe, %O : 10 ; MgSO,, 740 : 250 ; K$04 : 250 ;
M&O,, II-&O : 1 ; qB0, : 2.; Z&O,, WO : 0.5 ; CuSO,, 5H.&#~ ; .CoSO,, TO:
0.1 ; Na@G, 2qO : 0.1 and 0.05% KNO, (w/v) was added to the nutrient solution
in the plus nitrogen (+N) control assemblies. The pH was adjusted to 6.8 with 1N NaOH.
The entire assemblies with nutrient solution and vermiculite medium were autoclaved at
121°C for 1 h.
Seed of the tree legumes obtained from the Kenya Forestry Seed Center were scari-
fied with concentrated sulphuric acid for the following times (mn) : Acucica albida and
Acacia mearnsii, 20 ; L.eucaeim leucocephala, Calliandra calolhyrsus &d Sesbania sesban,
15 ; Prosopis julijlora and Sesbania grandjlora, 10. After treatment, the seeds were washed
in several changes of sterile distilled water until all %;.$f the, acid were removed.
The seeds were then aseptically placed onto 0.75% (w/v) wather agar plates and incubated
at 28°C for 72 h.
Two pm-germinated seedlings of similar size and radicle length were grown per
assembly. The seedlings were inoculated by dispensing 1 ml of a fully grown test culture
around theii roots using a fresh pipette for each strain of Rhiwbizm. Sterilized dry gravel
was spread over the surface of the vermiculite. Plants were thinned to one per assembly
by excising the shoot of the unwanted plant with sterilized scissors.
Each treatment of host-strain combination and uninoculated plus nitrogen (+N) and
minus nitrogen (-N) controls were replicated four times and grouped according to host
species. The plants were then grown in the glasshouse for 10 weeks. Replenishment of
water and nutrient solution was done once every fortnight.
.-
Table 1 : Growth characteristics of eight. I<hizobium straiis isolated from sev'en tree legumes
Rhlzdbium
Locality
Host [II' Isolation
Growth
Reactton on
Colonial morphology
strain
Bromothymol
Blue (BTB)
1Bb
Magarini
Sesbil,lla randiflora
Slow
Alkaline
Flat, dull, cream coloursd'with
-
-
(Coast Province)
translucent margin differentiating
into greater opacity at the center
of the colonies
lla
Turbo
Sesballl a sesban
Fast
-
-
-
Acidic
Flat, shiny, white coloured and
(Rift Valley)
evenly opaque colontes
2c
Mbits point
Leucaena leuceupheld.,; Fast
Acidic
domed,; shiny, cream coloured with
(Nyanza Province)
tran$lucent margin differentiating
,..
into greater opacity at the center
,I
of the colonies
6b
Kisumu
Leucaf:na leuceu hala!'
-
-
Fast
Acidic
Domed, shiny, cream toured with
'(Nyanza Province)
', z.,
translucent margln dtfferentlatlnQ
fnto greater opacity at thecente'r
of the colonies.
15b
Katangi
Proso;)is
jultflord
- .-
Fast
Acidtc
Flat, shiny, white coloured and
(EasternProvince)
evenly,opaque colonies.
16~
W u n d a n y i
Calli,lridra calothyrs&
~ . ..-
Fast
Acidic
Domed, shiny: white coloured wit);
(Coast Province)
t,ranslucerit margin differentiating
'i.
into greater,opacity at the center?.
of the colonies
e%r
,
;- :
22
hbita Point
Acacl albida
'
-
-
Fast
Acidic
(Nyanza Province)
.?--
5-L
/..
NM 7?7*
Eldoret wattle
Acacl,!
mernsii
'[;.
-.. -
Fast
Neutral
Flat, dull, white Colau&d and evhnly
e;;;;ctin f ctory
T..?
opaque colonies ::
al ev
e
* Source : The Nairobi Rhizdbitxn MIRCEN.
,! !
..: .:
.,i
, -.?
.’
:.
.
.
.
: :
84
Infectiveness and effectiveness of each host-strain combination was described by
visual growth ratings. An effective association (B) produced nodulated dark green healthy
plants showing excellent growth ; partially effective association (F’E) produced nodulated
plants with light green colour and restricted growth ; ineffective association (e) produced
nodmated plants which were pale green and stunted ; and plants which were not nodulated
were designated non-infective (ni). Nodule shape and distribution were described for each
host-strain association according to Somasegaran and Hoben (1985). The dry weight of
the plant shoots were determined after oven-drying at 70°C for 48 h.
Results and Qiscu,+@n
The results of the cross-inoculation tests are given in Table 2. Uninoculated control
treatments remained free of nodules in all host species tested.
The Rhiwbium strahs exhibited 8 high degree of infectiveness with the test plants.
All strains were infectiv& iin their own host of isolation. Strains isolated from CaNziudra
calothyrsus, Leucaena leucocephala and Prosopis jul$i’ora nodulated all the hosts whereas
those from Sesbania grandi$lora, Sesbania sedan and Acacia mearnsii variably nodulated
five host species each.
There were four cross-inoculation or infective groups within the scope of this study.
These were as follows : (i) first group : Acacia mearnsii and Prosopis jzdij7ora ;
(ii) second group : Leucaena leucocephala. Callzkdra calothyrsus and Prosopis Julifzora ;
(iii) third group : Acacia albida. Acacia mearnsii, Sesbania grandiflora and Sesbunia
sesban ; (iv) fourth group : Acacia Albida, Leucaena leucocephala, Sesbania grand@lora
and Sesbania sesban. I-,
Rhizobii in each infective group were mutually interchangeable between or among
the plant host members as defined by Graham (1976). Given the narrow range of Rhiwhm
strains and host species, ‘and the fact that strains +olated from a given tree can also
be different, the apparent infective groups may be regarded as atypical because of the
evident overlap. Howevertheconcept of ctiective group>> is not to be emphasized particularly
in the tropics because even in the recognized groups, infection across the boundaries have
been shown to be prevalent (Lange, 1961 ; Habish and Khairi, 1968 ; Trinick, 1968).
Symbiotic effectiveness as rated by visual appearance of test plants gave the following
own host associations : effective symbioses in Sesbania sedan, Sesbania grandiflora,
Calliandra calothyrsus and Acacia albida ; partially effective symbioses in Leucaena leu-
cocephala and Acacia mearnsii ; ineffective asskation in Prosopis julijlora. Effective
cross-inoculation symbioses were obtained by Prosopis julij7ora strain 131 on Lt~aena
leuwcephala and Calliandra calothyrsus ; Calliandra calothyrsus strain 16~ on Leucaena
leucocephala, and Acacia mearnsii stxGn NUM 777 on Acacia albida.
On the basis of beneficial symbiosis, Sesbania sesban and Sesbania grandflora
were more specific in their rhizobial requirement. Others were less specific because in
addition to forming effective or partially effective own host symbioses, they also exhibited
varying degrees of effectiveness in cross-inoculation with the exception of Prosopis jul$ora.
The ineffective own host nodulation in Prosopis jdifora is not an isolated case. Mettinen
et al (1968) working in Bum (Eastern Kenya) reported a similar occurrence and attributed
it to cross-inoculation with indigenous soil rhizobia compatible with the indigenous Acacia
Senegal. This would seem possible because Prosopis jdflora is not endemic or as natur&zed
as the other species in Kenya and thus may merely be acting as trap-host for indigenous
fable 2 : Ineffectiveness and effectiveness ratlqgs of host-straln combinations
Rhizobtum
Host of
Sesbania
Sesbania
Leucaena
Calllandra
Acacia
Acacia
strain
Isolation
grandlflora
SeJblln
-hala
calothyrsus m
-
liizarnst I
-
18b
Sesbanla
PE(2) ':
E(1)
:r”
e
mora
ni.(?:)
,,(JI
:
lla
Sesbanla
5eSbln
e
e
ni
e
2C
Leucaena
leucbcephala
e
e
PE
e
. ,.
P E
6b
Leucaena
-hala
e
.) e
PE
e
e
E,
15b
e
‘- e
E
E
PE '
PE
ii!iEs?
:,.
16c
Calllandra
calothyrsus
e
e
E
E
e
e
22
Acacia
m
e
.'
e
e
nf
IE
NUM 777
Acacia
E
PE
,e
,’,:‘.
(1) E = Effective, (2) PE = Partially effective, t31e = Ineffective, (4) nl = non infective.
,:
.
86
rhizobii Although the Rhiwbium strains were isolated from effective (pink) nodules, the
partiahy effective or ineffective own host associations reported in this study emphasizes
the need to re-authenticate strain effectiveness on host of isolation prior to use in cross-
inoculation studies. The response to inocuIation in terms of shoot dry weight on own
host and cross-inoculation combinations gave similar trend as in visual ratings. Own host
association of strains lla, 18b, 16c and 22 (Fig. la, b, c, e) gave shoot dry weights
that were signihcantly higher (p = 0.05) than the tminoculated -N control. Shoot dry
weights of cross-inocuIation association of strains 15b on Leucaena Zeucocephafa and
Calliandra calothyrsus, 16c on Lmcaena leucocephala and NUM 777 on Acacia nlbida
(Fig. lc, f. e) were aIso significantIy higher than the uninoculated -N control other beneficial
symbioses of shin 16c on Prosopis juliflora (Fii. Id) and strains 6b, 18b, NUM 777,
2c, 15b, 22 on Acacia meamsii (Fig. lg) were not significant. The insignificant response
of the Iatter two host species could be attributed to their growth rates which were observed
to be relatively slower than the other host species. There also occurred a marked increase
of shoot dry weight by the following symbioses and magnitudes relative to +N control :
strains lla on Sehmia sesban, x 6.0 ; strains 16c and 15b on Leucaena leucocephala,
x 4.7 and 3.8 respectively ; strain NUM 777 on Acacia alIda, x 1.6 ; and strain 15b
on Calhndra calothyms, x 3.0. Despite the fact that the amount of inorganic N applied
in +N control may not have been optimum, the magnitudes of the comparisons indicate
some potential in biological nitrogen fucation of these IWT with the indigenous rhizobia.
The plant host determined the shape and distribution of nodules on the root system
in each host species irrespective of host of isolation of the Rhizobium strain forming
the association (Table 3). However, consistency in nodule shape and distribution was main-
tained only in effective symbioses.
The resuhs presented in this paper have shown varying degrees of infectiveness
and compatibihty of indigenous rhizobii with the NFT. These have in effect posed a
great challeng& The chaIlenge to assess the need of inoculation, select effective symbioses
under simuIated environmentaI conditions in the laboratory/gIasshouse and evaluate the
selected symbioses in the field sites where they are to be grown. Kenya Forestry Research
Institute is addressing itself to these studies.
.”
‘i
Table 3 : Nodule shape and distribution of effective host-strain associations 10 weeks after planting
.'
Test plant
Rhizoblum rtraln and host
of isolation (In brackets)
Nodule shape
Nodule $strlbutlon
Sesbania grandlflora
lab
(2. grandlflora)
Se&globase
Prolfflc t&-root and bcca-
,, -
;.
sionpl lateral root
._.,.
ribdulatlon
'f
.
Sesbanla sesban 3;
-
-
Globose':
5.:z? Occassional tap-root and large
humber of root nodulation
.
.
e
:
,(
Leucaena leucecophala
16~ (C. cdloth rsus)
U;$ate and
"^
"'
Occasstonpl'thp-root and
15b (F T+
lateral root nodulation
I
Prosopts juliflora
16c (C. calothyrsus)
Semi-globose and
Occasslonal tap-root and
lobed
lateral root nodulation
Calliandra calothyrsus
lbc IC. calothyrsus)
Semi-globose
Large number of tap-root and
15b (p. juliflor.aT
lateral root nodulation
Acacia alblda
-
-
22 (A. albfda)
Semt-glabose
Prolific tap-root and occas-.
NUM 777 ~arns~l)
-
-
stonal lateral root nodulatlon
88
.‘
(al sesbd;jia &
(b) Sesbania grandifkxx
(Cl Callii~ cabthyrsus
-
-
T r e a t m e n t s
Fig.1. a b c 6 d. Effect of inoculation
with strains of m and application of
inorgrjc N on shoot dry weights d various tree species in
Leonards jars at 10 weeks after p4ant-q.
/..,
.:,.t,
3;.
. ..c.:
:
-
‘:
I
II
‘..I
;?
L
i. ‘.:.. .
__
9 0
Acknowledgements
I am very graMid to Professor S.O. Keya and Nairobi Rhiwbiwn MIRCEN staff
for technical assistance, and IDRC for financial support to attend the meeting.
References
BERGENSEN, FJ. 1980 Methods for evaluating biological nitrogen fixation. John Wiley
and sons, Ltd.
BUMPUS, ED. 1957 Legume nodulation iu Kenya. East Afi. Agric. For. J. 23, 91-99.
GRAHAM, P.H. 1976 Identification and classification of root nodule bacteria-In : Symbiotic
nitrogen fixation in plants. (Nutman, P.S. ed.). pp. 99-l 12. Cambridge University Press,
Cambridge.
HABISH, HA. and KHAJRI. S.M 1968 Nodulation of legumes in the Sudan : cross-
inoculation groups and the associated Rhiwbium s&aim. Exp. Agric. 4, 227-234.
KEYA, LO.1977 Nod&ion and nitrogen fucaton in legumes in East Africa. In : Biological
nitrogen fixation in farming systems of the Tropics. (Ayan&a, A. and Dart, PJ. eds.).
pp 233-243. John Wiley and Sons, Chichester.
KEYA, N.C.O. and Van EIJNA’ITEN, C.L.M. 1975 Studies on oversowing of natural
grassland in Kenya. 2. Effects of seed inoculation and pelleting on the nodulation
of Desmodirun uncinatum (JACQ.) DC. East Afk. Agric. For. J, 40, 351-358.
LANGE, R-T. 1961 Nodule bacteria associated with indigenous Leguminosae of South-
Western Australia. J. Gen. Microbial. 26, 351-359.
MCDONALD, J. 1935 The inoculation of leguminous crops. &sf Afr. Agric. For. J. 1,
8-13.
MORRISSON, J. 1966 Productivity of grass and grass/legume swards in Kenya Highlands.
East Ajr. Agric. For. J. 32, 19-24.
SOMASEGARAN, P. and HOBEN, H.J. 1985 Methods in Legume - Rhizobium technology.
University of Havaii, NiiAL
Project and MIRCEN.
SOUZA, D.I.A. de 1969. Legume noduhuion and notrogen fixation studies in Kenya.
East Aj-. Agric. For. J. 34, 299-305.
TRINICK, MJ. 1%8 Nodulation of tropical legumes. 1. Specificity in the Rhizobium
symbhsis of Leucaena leucpcephala.
Eap. Agric. 4, 243-253.
VINCENT, J.M. 1970 Amanual for the practical study of root nodule bacteria. IBP Handbook
n* 15. Oxford, Blackwell Scientific Publications.
FIXATION D’AZOTE ASSOCIATIVE
Maximiser la FBA pout la Production Agricole et Forestih en Aftiqve
Effect of inoculation on the growth
and yield of three maize cultivars
IMWAURA, F.B. and WIDDOWSON, D.
Dqwtment of Botany, University of Nairobi, Nairobi, Kkya
Summary
Illocallation exp+mexlra were con&ted using three lnaizc cultivars (z?u &yiL)
and three N,-fixing plant-growth-lvomotinging-rhizobacteria
The plants WQC grown to
maturity and various vegetative and reproductive m were determined. The
results ohtaincd indicated significanr increases in the shoot dry mat&s yield (342%)
of one maize arltivar with only a small inaease in seed dry weight Significant
increases in seed dry weight (38.4%-495%) were mcorded for inoculated plants of
two cultivars. No si&ficaat increase in the cnrde protein of the grain was observed
in the. inoculaled tTeatments.
lntroductlon
It has been suggested that increasing the proportion of rhizosphere w in
grass-bacteria associations, through inoculation, may result in &eased N, fixation and
plant yield (Pan-iquin, 1982 ; Okon, 1984). Such increases in the numbers of N,-hxing
rhizobacteria may be achieved by application of highly competitive bacterial strains, heavy
inoculation or by the introduction of N,-fixing genes to other root associated bacteria
(Kleeberger and Khngmuller, 1980 ; Postgate and Kent, 1987). However results obtained
in inoculation stuclies of cereal crops and forage grasses are very varied ranging from
no response (Schank et al. 1980 ; Wright and Weaver, 1982) to highly significant responses
(Kapulnik et al., 1981 ; Mertens and Hess, 1984 ; Smith et al., 1984). Differences in
inoculation responses attributable to plant genotype effects have previously been demon-
.
.
I
93
trated in wheat and maize among other cereal crops (Kapulnik et al., 1981 ; Reynders
and Vlassak, 1982 ; Millet et al., 1984).
In this study we report the response of three maize cultivars to inoculation with
N,-fixing plant-growth-promoting-rhizobacteria in a tropical soil.
Materials and Methods
Three facultatively anaerobic diazotmtrophs C, J and L previously isolated from
maize plant roots were grown under static conditions at 24°C in a liquid nitrogen deficient
glucose medium (L&Berg and Granhall, 1984 ; Mwaura and Granhall, 1986). The cultures
were grown to stationary phase, centrifuged and washed twice in 0.05 M sterile phosphate
buffer pH 7.0 @erg et al., 1980). The cells were finally resuspended in buffer and trans-
ferred into sterile petri dishes.
Seeds of three maize cultivars LG 11, Makueni and Katumani were washed under
running water and in sterile distilled water to remove the fungicides. The seeds were
germinated on moistened filter paper in the dark at room temperature. The seedling lot
of each cultivar was divided into four batches. For each cult&r, one batch of seeds
was used as a control while the other three batches were inoculated with the three bacterial
strains C, J and L respectively. The seedlings were inoculated by placing them in petri
dishes containing the bacterial suspensions for 30 mn. Control seeds were immersed in
sterile buffer.
Four maize seedlings were sown in 30 1 plastic troughs containing well moistened
soil from a weeded plot. Great care was taken to avoid cross contamination between
the containers. Each treatment was replicated four times and all containers were randomly
arranged in an open space but under a roof of clear plastic sheeting to prevent flooding
of the containers by rain water. The soil in each container was covered with washed
vermiculite lo millhisalgal growth:All plants were watered every two or three days
as necessary. The plants were thinned to three in each container 10 days after planting.
After growing for 110 days, the plants were harvested and their shoot lengths measured.
The shoot, ear and lOOO-seed dry weights were determined in each treatment.
The plant materials were dried to constant weight in a forced-air oven at 80°C.
The materials were then milled and the nitrogen content was determined by the Kjeldahl
method (Hesse, 1971). The crude protein content of the maize grain was estimated by
multiplying the nitrogen content with a factor of 6.25 (AOAC, 1960).
Results
The results are summarised in Tables 1 and 2. Improved vegetative growth was
particularly evident in maize plants of cultivars LG 11 and Katumani inoculated with
the three bacterial strains Substantial increases in plant height (11.6 - 43%) and shoot
dry weights (3-l- 342%) were recorded over the uninoculated plants (Table 1). However,
little or no improvement in vegetative growth was observed in inoculated maize plants
of the Makueni cultivar. Among the three inoculants tested, bacterial strain C was the
most effective in increasing the shoot dry matter yield. Inoculation however did not increase
the nitrogen content of maize plant shoots except for plants of cultivar LG 11 inoculated
with bacterial strain C (Table 1).
Table 1. Effect of jnoculation on the vegetative growth and N-yield of three maize cult'ivars.
Cultivar
Inoculant
Plant height
Shoot d,wt./plant
% N
Ndylel d/plant
(cm)
(g)
(mg)
LG 11
control
130.2 +
12.1
39.1 +’
4.9
0.35
136.9
i_
152.1 7
5.4
51.1 7
2.8*
0.34
J'
186.0 7
5.6
":'
47.7 T
2.5
0.37
;:;*;
,C
180.3 z
9.1
:
52.5 z
4.4*
0.54
283:5.
.:
P
Katumani
control
184.7 t
14.3
:,
77.6 2
7.0
0.65
504.4
206.2 7
6.6
8.1
0.64
512.0
!i
225.5 7
15.4
,.'
0.61
C
244.9 7
9.2
9712
;i.;';2' ii::*
0.57
Makueni
control
175.1 +
8.9
75.4 t
6.6
0.41
309.1
J'
196.8 187.7 7 11.9 10
80.0 76.1 7
7.3
oi.39
296.8
L
8.1
0.38
304.0
C
162.7 z 11.5
1.
74.2 z
6.0
q.39
289.4
* significant increases (P = 0.05').
Results are mean of six replicates plants t S.E.
_,
-
.
._ _
.
.
I_-. --
. ..-.........
_ ..^. ..-..--
_.
..--&-
--..
-.
_-_-
__
.
_
95
With regard to the reproductive parameters, ear dry weight and the mean seed
dry weight increased in inoculated plants compared to those of uninocuhued treatments.
‘Ihe most significant responses were observed in Kammani and Makueni cultivars for
which the lOOO-seed dry weight increased by up to 49.5% and 38.4% respectively frable
2). The results indicate that bacterial strains J and L were the most effective in promoting
reproductive growth of the maize plants. Higher ear and seed dry weights were observed
in maize plants inoculated with the two strains compared to those of plants inoculated
with bacterial strain C. The crude protein content in the grain was however unaffected
by inoculation except for the Makueni cuhivar in which higher crude protein levels were
recorded in inoculated than in the uninocuJated treatment (Table 2). It was also noted
that the inoculated maize plants flowered and matured about 10 days earlier than the
control plants.
Discussion
In addition to the inoculated plants maturing slightly earlier than the controls, ino-
culation increased plant height and dry matter yield Earlier heading and flowering has
similarly been observed in wheat, Setaria iralicu, sorghum and Punicum sp. in the green-
house and in the field after inoculation with N,-fiig bacteria (Kapuhtik et al., 1981
; Kapulnik et al., 1983 ; Lin et al., 1983 ; Yaholon et al., 1984 ; Sarig er al., 1984).
Jnocnlation of the Makueni cultivar did not promote vegetative growth of the maize plants
but the seed dry weight increased by up to 38.4%. Large increases in ear dry weight
(31.9%) and seed dry weight (49.5%) were also recorded for the Katumani maize cuhivar
inoculated with the test bacteria. The observed differences in make cultivar response to
inoculation with the N,-fixing rhizobacteria were rennuI&le in that the third cultivar LG
11 responded best in terms of plant dry matter yield increases but recorded the lowest
increase in seed dry weight among the three cultivars. Similar genotype effects have been
established for maize (Kapuhuk et al., 1981 ; He@ and Monib, 1983) and other cereals
(.YMillet et al., 1984).
In Israel, Kapuhtik et &(1981) reported seed and cob dry weight increases of
28.2% and 34.4% respectively in inoculated maize grown under field conditions. Few
published reports on inoculation effects in maize and other cereals in Africa are presently
available. In Egypt, Hegazi and Monib (1983) reported increases of 150-170%
180-270%, 120-130% for straw yield, grain yield and total N yield respectively in maize
inoculated with AzospiriUum
sp. under field conditions. More concerted efforts need to
be made to study the potential for increasin g grain and forage yield of cereal crops
through inoculation with growth-promoting-rhizobacteria.
‘.
.: ..*
. .
-
.-
:_
Table 2.
Effect of inoculation on yield parameters of three maize cultjvars.
Cultivar
Inoculant
ear d.wt,
lOOO-seed
mean seed
X crude
/plant 19)
d.wt. (g)
d. wt. (mg)
protein
-
I
LG 11
control
57.3 f 5.2
*I
129.8
129.8
13.1
J'
56.7 54.0 7 0.5 4.8
._
130.8 135.2
130.8
12.8
-..
',
135.2
C
67.7 2.13.2z .:',
141.7
',
141.7
1;::
Katumani
control.
32.9 ,+ 4.3
74.1
74.1
-
13.2
:
43.4 38.4 7 7 7.2 3.7,
i
110.8
9660
96.0
:
10.5
11.6
c
"
39.2 2 4.1
:
76.5
':E
.
14.1
Makueni
control
4442 t 1.9 i"
81;6
81.6
46.4 7 10.7
112.9
112.9*
109*:
:
49.8 7 6.6 +.
c
.._
42.8? 2,.5 :"'.-
'07K
'%
.'
10:5
.
.F.
.
11.4
",e #
: ,. *
* signlflcant'increases (P = 0.05).
, ,..I.'
,, .,
Results are means of SIX replicate plants 2 S.E.:,Ears of replicate plants were pooled, shelled and the
grain dried to constant welght.
'
.
.
.
+.
.’
:.
.-
.+
97
Acknowledgements
We wish to thank the International Foundation for Science and the Kenya National
Council for Science and Technology for financial support to one of us (Mwaura, B.)
References
AOAC. 1960 Methods of Analysis. Washington D.C.
BERG, R.H., TYLER, ME., NOVICK, NJ., VASIL, V. and VASIL, Ix. 1980 Biology
of Azospbillum - sugarcane association : Enhancement of nitrogenase activity. Appl.
Enrivon. Microbial. 39, 642-649.
HEGAZI, N.A. and MONIB, M. 1983 Response of maize plants to inoculation with azos-
pirilla and (or) straw amendment in Egypt. Can. J. Microbial. 29, 888-894
HESSE, PP. 1971 -A textbook of soil chemical analysis. John Murray Publishers.
KAPULNIK, Y., SARIG, S., NUR, L and OKON, Y. 1983 Effect of Azospirillum inocu-
lation on yield of field-grown wheat Can. J. Microbial. 29, 895-944.
KAPULNIK, Y., SARIG, S., NUR, I., OKON, Y., KIGEL, J. and HENIS, Y. 1981 Yield
immases in summer cexeal crops in Israeli fields inocdakd with Azospirillm Expl
Agric. 17, 179-187.
KLEEBERGER, A. and KLINGMULLER, W. 1980 Plasmid mediated transfer of nitrogen-
fixing capability to bacteria from the rhizosphere of grasses. Mol. Gen. Gem. 180,
621-627.
LlN, W., OKON, Y. and HARDY, R.W.F. 1983 Enhanced mineral uptake by Zea mays
ad Sorghum bicolor mm inoculated with Azosptillwn bradense. Appl. Environ.
Microbial. 45, 1775-1779.
LINDBERG, T. and GRANHAIL,, U. 1984 Isolation and characterisation of dinitrogen-
fixing bacterial from the rhizosphere of temperate cereals and forage grasses. Appl.
Environ. Microbial. 48, 683-689.
MERTENS, T. and HESS, D. 1984 Yield increases in spring wheat (Triticum aestivum
L.) inoculated with Azospirillm lipoferum under greenhouse and field conditions of
a temperate region. Plant ana’ Soil. 82, 87-99.
MILLET, E.. AVM, Y. and FELDMAN, M. 1984 Yield response of various wheat geno-
types to inoculation with Azospirillum bradense. Plant and Soil. 80, 261-266.
MWAURA, F. and GRANHALL, U. 1986 Nitrogen fixation (CA reduction) associated
with Maize (Za mays L.) in a Swedish soil. Swedish J. agric. Res. 16, 49-56.
OKON, Y. 1984 Response of cereal and forage grasses to inoculation with N2-fixing
bacteria. In : Veeger, C. and Newton, W. E. (eds.). Advances in nitrogen fixation
research, pp. 303-309. Martinus Nijhoff/Dr. W. Junk Publishers, The Hague.
PATRIQUIN, D.G. 1982 New developments in grass-bacteria associations. In : Subba
Rae. N.S. (ed.). Advances in agricultural microbiology. pp. 139-190. Butterworth Scien-
tific, London.
POSTGATE, J.R. and KENT, H.M. 1987 Qualitative evidence for expression of Klebsiella
pnewnoniae Nif in Pseudomonas put&. J. Gen. Microbial. 133, 2563-2566.
98
REYNDERS, L. and VLASSAK, K. 1982 Use of Azospirillwn brasilense as biofertilizer
in intensive wheat cropping. Plant and Soil. 66. 217-223.
SARIG, S., KAPULNIK, Y., NUR, I. and OKON, Y. 1984 Response of non irrigated
Sorghum bicolor to Azospirilium inoculation. Expl. Agric. 20, 59-66.
SWANK, S.C., SMlTH, R.L. and WEISER, G.C. 1980 Responses of two Pearl Millets
grown in vitro after inoculation with Azospirillum brasilense. Soil Crop Sci. Sot. Fla.
Proc. 39, 112-115.
SMITH,R.L., SCHANK, S.C.,MILAN, JR. andBALDENSPERGER, A.A. 1984Responses
of Sorghum and Pennisetwn species to the N,-fixing bacterium Azospirillwn bradense.
Appl. Environ. Microbial. 47, 1331-1336.WRIGHT, S.F. and WEAVER, R.W. 1982
Inoculation of forage grasses with N,-fixing Enterobacteriaceae. Plant and Soil. 65,
415-419.
YAHALON, E., KAPULNIK, Y. and OKON, Y. 1984 Response of Setaria italica to
inoculation with Azospidlum brasilense as compared to Azotobacter chrooccocum.
Plant Soil. 82, 77-85.
MW la FBA pour la P&don Agricofe cl For&e en Ajkiquc
Population of-~i&ro~e&ixhg bacteria
in sweet potato fibrou& ri3ots
HORTENSE, W.DP’, WALTER, A.HF,
IMULONGOY, K.t2’, ADEYEYE, S.0.‘2’,
and HAHN, SX.“‘. ”
(1) Tiuk-egee University. Alabama,VSA;
_,
{2). I- Znstit+ .~..T~~~.:~~~~=,,~.Niguin.
:
. .
Abstract
Fibrous e’s of six sweet ~q~~<ul$~ars w~,ev@ated
for rhizosphese popula-
ti. of N$.&y&.;~
JuWAikc.,~~k‘~aamplingdates.PopllatiomOf
the bacteria were de&mix& using culture tubes (pellicle formation and acetylene
reduction activity), and colony plate counts. Plant growth parameters evahuted
were : fresh and dry weights of storage roots. foliage, fibrous roots and total
biomass. The bacterial population ranged from 10’ to 106 g per dry root weight
depending on the counting method used- The N$xing bacterial populations were
positively correlated (r = 0.67 to 0.88) with foliage arid total biomass dry weight
for cultivars TIS 70357 and TIS 8441 ; for the same parameters, negative corre-
lations (r = -0.68 to -0.80) were found for cultivars TIS 9265 and Tlb 4 and corm-
lations were not significant for aikivars TIS 2498 and TLS 8504.
100
Introduction
Sweet potato cultivars from the breeding program at the International Institute of
Tropical Agriculture @TA) produce fi-om 40 to 50 mt/ha of storage roots and up to
31 mt/ha foliage when grown in soils with low soil N indexes (IIahn et al.. 1984 ; Hill
et al., 1985). Azospidum-like bacteria am widely distributed in tropical and temperate
region soils and plants, and have been suggested as important associative N,-fming bac-
teria under the direct influence of plants. Several strains of A. brudense and A. Zipoferum
associated with the roots of different sweet potato varieties have been isolated and characte-
rized (Hill ef al., 1983 ; Hill and Bacon, 1984). Reviews of Azospirillum inoculation
studies by Boddey and Dobereiner (1982) and Dart (1986) show inconsistent responses
of cereals and grasses for several parameters including mtrogenase activity, nutrient
uptake and top dry matter. When sweet potato was inocul& with Azospidhm bra-
dense, the root yield .and N ‘content were increased (Crossman and Hill, 1986).
The purpose of this study was to quantify naturally occuring nitrogen fixing ba-
cteria in sweet potato rhizosphere and determine the effect of fertilizer nitrogen addition
on growth and microbial parameters, and to evaluate relationships between sweet potato
growth parameters and parameters used to measure associative N, fixation.
Material and Methods
Six sweet potato cultivars were selected from the IITA breeding program : TIb4,
TIS 2498, TIS 70357, TIS 8441, TIS 8504 and TIS 9265 were planted in loamy sand
at the IlTA Ihadan, Nigeria. The experimental design was a split plot with two rates
of fertilizer N (0 and 50 kg of calcium ammonium nitrate N ha-‘) as main plots and
cultivars as subplot treatrmzirts. Each treatment con&ted of four rows of a sweet potato
cultivar, 13.2 m long and Im wide each, with plants spaced 0.5 m within rows. At 6,
10 and 14 weeks after planting (WAR), four plants were collected at random f?om the
left center row, and storage roots, foliage and fibrous root weights and associative N,
fmation parameters were evaluated. At each sampling date, sub-samples of 2 g of fibrous
roots plus adhering soil were analysed for number of Azospirillum-like bacteria, using
the serial dilution method described by Wollum (1982) for colony, plate count and the
most probable number method described by Knowles (1984). Malate N free media was
used for both solid plates and semi-solid tubes. Acetylene reduction assay (ARA) was
carried out on all tubes -showing pellicle formation, using a Carle Age 211 Gas Chro-
matograph’ equiped with a flame ionisation detector.
Results and Discussions ..
Storage root weight was higher for TIS 8504 and 8441 than for the other cultivars
at 6, 10 and 14 WAP irrespective of fertilizer treatment (Table 1). Differences in total
dry biomass and storage root weights were most apparent at 14 weeks after transplanting
for both 0 and 50 kg N ha-‘. Fertilizer N addition tended to depress storage root yield
and foliage weight of most cultivars. Acetylene reduction activity was detected in culture
tubes at 10’ dilution for all six cultivars at all three sample dates suggesting the presence
of N,-fixing bacteria in the rhizosphere of all the cultivars studied. The most probable
numbers (MPN) of N,-fixing bacteria were low, and significant differences were not apparent.
101
Table 1 : Effect of fertilizer N (50 kg N ha-') on sweet potato storage
roots and foliage growth.
Tuber wt
Foliage wt
Cultivars
(kg ha-l)
(kg hi-11
N = 50
N = O
N = 50
N = O
TIS 8504
824 a
635 a
1822 b
2162 a
TIS 8441
492 b
548 b
2291 ab
2075 b
TIS 2498
319 cd
330 c
1547 bc
1439 bc
TIb 4
390 c
374 c
1322 c
1172 c
TIS 9265
156 d
147 d
1967 b
1705 bc
TIS 70357
188 d
151 d
2502 a
2227 a
Means with the same letter in the same column are not significantly
different at the 5 X level by the Duncan's multiple range test.
The lowest yielding cultivar TlS 70357 had a significantly higher MPN and percentage
of N,-fixing bacteria (Table 2). Significant correlations (r = 0.67 to 0.88) were found
for foliage and total dry biomass weight, with MPN of AzospitiZZIun-like bacteria in the
rhizosphere (Table 3).
Correlations between Nz-fixing bacterial population and foliage weight or total biomass
were positive for TlS 70357 and 8441 suggesting that Azospirilhm-like bacteria were
responsible for the increase in foliage production. Correlation for the same parameters
were negative for TIb 4 and TlS 9265 (r= -0.68 to -0.80) and not significant for cultivar
TIS 2498 and 8504.
These results suggest that associative N2 fixation in sweet potato is cuhivar dependent,
and that foliage and total biomass production of some sweet potato cultivars on soils
with low N indexes may be related to the presence of N,-furing bacteria However the
association between Azospidlum-like bacteria and sweet potato fibrous roots seems to
be regulated by complex mechanisms that need to be more investigated.
102
Table 2 : Number of.Azospirillum bacteria in sweet potato fibrous roots
by the acetylene'reduction assay (ARAB and the most probable
number (MPN)
Cultivars
ARA
Log ARA
MPN
Log MPN
105 /g roots
105 /g roots
TIS 8504
8.20 b
3.25 abc
0.090 a
3.75 a
TIS 8441
9.86 b
3.17 bc
0.067 a
3.62 a
TIS 2498
11.79 b
2.63 c
0.042 a
3.50 a
TIb 4
5.77 b
3.34 abc
0.055 a
3.67 a
TIS 9265
7.07 b
4.09 ab
0.089 a
3.72 a
TIS 70357
88.31
a
4.57 a
0.010 a
3.72 a
Means with the same letter in the same column are not significantly
different at the 5 % level by Duncan's multiple range test.
103
Table 3. Correlation coefficients between selected growth.para&ters and -
rhitosphere Np-fixing bacterial count for six sweet potato -'
I,
cultivars. -
Log
Log
Log -. ;
CULTIVAR
MPN-A'
HPN-N2
CPC3
HPN-A
WPN-N
CPC
TIS 8504
FDW4
N S
NS
NS
NS NS
NS
TD85
N S
NS
NS
NS
NS
NS
TIS 8441
FDW
NS
0.75*
0.84***
NS
NS
NS
-
TDB
NS
0.77**
0.88***
NS
NS
NS
TIS 2498
FDW
NS
N S
NS
NS
NS
NS
I
FDB
NS
N S
NS
NS
NS
NS !
Jib 4
* FDW
-0.80**
N S
NS
-0.7&
-0.76*
. I
-0.70*
TDB
-0.72*
N S
NS
-0.78** N S
NS
TIS 9265
FDW
N S
N S
NS
NS
NS
NS
TDB
NS
N S
NS
-0.75**
-0.68**
NS
TIS 70357
FDW
0.71*
N S
0.75*
0.72*
0.67*
NS
TD8
N S
N S
NS
NS
NS
NS
1. Most probable numbers of Azospirillum
2..Most probable numbers of Ricronitropliilic
bacteria
3. Colony plate counts
4. Foliage dry weight
5. Total dry biomass
*Significant at the 5% level
l +Significant at the 1% -level
***Significant at the 0.1% level
‘..‘.
.’
104
Acknowledgements
This paper is contribution No. PSOO8 of the George Washington Carver AgricuI-
turaI Experiment station, Tuskegee Univexsity. The research was supported by funds fi-om
the U.S. Agency for International Development (project No. DAN-5053-G-55-6062-00)
and the I$SDA/CSRS @ant No. ALX-SP-1).
References
BODDEY, RIkL and DOBEREINER,J. 1982 Association of AzospirifZum and other dia-
zotrophs with tropical gmminae. In : Non symbiotic nitrogen fixation and organic
matter in the tropics. pp. 28-47. Symposium Papers. I. Transactions
12th Int. Cong.
Soil Sci New Delhi, India.
CROSSMAN, S.M. and HILL, WA.. 1986 Inoculation of sweet potato with Azospidlum
brasilense. Hart. Sci. 18, 169.
DART, PJ. 1986 Nitrogen fixation ass&a&d with non-legumes in agriculture. Plant and
Soil. 90, 303-334.
HAHN, SK, ALVAREZ, UN., CAVENESS. F.E. and NG, S.Y. 1984 Sweet potato
improvement at the International Institute of Tropical Agriculture. Proc. 1st Ckibean
Regional Workshop in Tropical Root Crops. pp. 93-97. University of the West Indies,
Kingston. Jamaica.
HILL, W.A. and BACON, P. 1984 Fertilizer N use efficiency and associative N, fixation
with non legumes. Proc 6th Int’l. Symp. Tropical Root Crops. CUP. Lima, Peru, Feb.
21-26, 1983.
HILL, WA, BACON, P., CROSSMAN, S.M and STEVENS, C. 1983 Characterkation
of N2-fixing bacteria associated with sweet potato roots. Can. J. Microbial. 29,
860-862.
HILL, W.A., HAHN, S.K. and MIJLONGOY K. 1985 Most probable numbers of nitrogen-
fming bacteria associated with sweet potato roots. Proc. 3rd Symposium Int. Sot.
Trop. Root Crops, African branch, Owerri, Nigeria, 16-23, August 1986. IDRC, Ottawa
Canada.
KNOWLES, R. 1984 Free living dinitrogen bacteria. In : Page, AL., Miller, RH. and
Keeney, D.L. (E&z.) Methods of soil analysis, part 2, Chemical and Microbiological
properties.p. 1070-1090 Agronomy Monograph Np 9, American society of Agronomy,
Madison, WI.
WOLLUM, A.G. 1982 Cultural methods for soil microorganisms. In : Page, A.L., Miller,
R.H. and Keeney, D.L. (&is.) Methods of soil analysis, Part 2, Chemical and Microbio-
logical properties. p. 781-802. Agronomy Monograph Np 9, American society of Agro-
nomy, Madison, WI.
ASSOCIATION AZOLLA / ANABAENA
MaruKrer la FBA pour la production Agricole et Forestière en Afrique
La recherche sur AzoZZa et ses applications :
évolution et tendances actuelles
VAN HOVE, C.(l) et DIARA, H.F.@)
(1) Ldzboruto~ de physiologie v&&ale, hiversit~ Catholique de haùh
Place croix ak sud, 4. B-I348 LmvahJa2deuvc. Belgïum.
(2) Association pour le Dtfveloppemmt
de la Rizicuhre
en Afrique de I’OM (ADRAO), BS. 96, Saint-buis, Sknpgal
i
Summaty
AwIIa uses as biofertilizer and food, traditional in some par& of China and Vietnam,
havebeenignoredarneglectedinothcrcounhies(andeveninwideareasoftbese
cmntrks) uutil the si~ties, when research programmes were launched with a view
to inueasing areas under Awlla cukivation and impmving cultural pra&e.s adap
ted to various mmmnes. Sincc thcn resxch, pure and applied, has blossomed, and
effurtstojntroduceAwZiammanykriancountriesbutalsomotha~ofthe
“
world~veinrreasod,withvarying~sts.ThepuIposeofthispapaistooutline
the major steps of ouf lcnowledge an AwZ& to bring out the main prt&nt re-
search trends, to descrîble thc various possible AzoZZu uses and theii m,
withspecialemphasisOn~SUgg~~~finanypnsenttd,~atpromoting
Azolkrinoroductioninrhiscontmen5~yf~søanfarmers,wbo’arethemain
potaltid beneficiaries.
Introduction
L’objectif du prbent document est de dresser un bref historique des principales
connaissances acquises sur la fougkre aquatique Azollu, de degager les grands axes de
recherche en cours, de dkrire les diverses possi%ilitt% de valorisation d’AzolZu en agriculture
:
107
et les problemes poses par son utilisation, l’accent étant mis sur le cas particulier de
l’Afrique. Des suggestions sont enfim émises en vue de promouvoir l’azolliculture dans
ce continent, spécialement aup&s des petits paysans qui en sont les principaux b&&-
ciaires potentiels.
L%volutior) de la Recherche
Des origines ii la crise p&roll&e
La valeur fertilisante et ahme&re de la fouragere aquatique AzoZZu est connue
et mise à profit depuis de nombreux siècles dans certaines figions tri2 lirnitks de Chine
et du Vietnam ; situ6es le long de la côte entre le 18ème et le 3Oi?me degre de latitude
Nord, ces r&gions se caractkisent ‘par un climat subtropical humide, à variations saison-
nières de temp&ature souvent consiMles. Jusqu’en 1960, la maîtrise d’Azollu y &ait
essentiellement empirique.
Dans le monde occidental par ailleurs, dès le début du dix neuvième siècle appa-
raissent quelques études fondamentales sur AzoZ2u ; ce n’est cependant qu’au début du
vingtième siècle, en grande partie gr& B l’intervention de chercheurs travaillant en Jndo-
chine, que l’int&& pratique d’AzolZu est signal& C’est ainsi qu’il est fait mention, sans
d’ailleurs que des arguments probants soient. toujours pr&ent&, & l’effet d@essif d’une
couverture de plans d’eau par Azolla, sur le &veloppement des moustiques (Smith, 1910),
de la capaciti de rkluction de l’azote (Ces, 1913), .du file fertïkant d’AzolZu en riziculture
et de sa valeur alimentaire pour des animaux d’blevage (Chevalier, 1926), de l’action
inhibitrice d’AzoZZu sur le développement des adventices (Braemer, 1927) ou enfin de
son influence negative sur les pertes d’eau par évapotranspiration (Nguyen, 1930). On
avait donc, des cette époque, relevé l’essentiel des proprit& pratiques d’AzoZlu conuues
aujourd’hui, mais il faut bien recormaît~~ que ces informaticms ne suscit&rent pas grand
int&êc
En Crient par contre, dks la fin des annees 50, les gouvernements vietnamien et
chinois décidèrent de promouvoir la recherche scientifique ayant trait a Azollu. J.l en re-
sulta un foisonnement & travaux, en g&u?ral a orientation appliquée. En Chine, Lumpkin
et Plucknett (1982),par la mise au point & techniques diverses de conservation en saison
froide, permit d’étendre consid&ablement l’aire de culture vers le Nord, jusqu’au delà
de Beijing ; dans le Sud par ailleurs des techniques & protection contre les tempkatures
élevées de l’été permirent~i’utilisation d’Azollu pour les cultures automnales.
L’étude des principaux .mseUes..@@tetns ,d’Aw&x (qui constituent un autre pro-
blktne majeur en Asie) et des modes &,lutte contre ceux-ci, ainsi que le développement
de techniques d’enfo uissement d’AzoZlu entre les Lignes de riz am&ior&ent quant à eux
I’effkacit6 de l’azolliculture.
Du Côt6 vietnamien (Dao ami Thuyet, 1979). les efforts ‘Jxntkent sur la s6lection
d’écotypes (indig&nes) d’A. pinnuru var. irnbkuta et sur l’am&oration des pratiques cul-
turales en association avec la culture du riz d’hiver, avec pour r&uhat un accroissement
considéxable des surfaces couvertes. lXs 1965, se développa la monoculture hivernale
d’AzolZu, dans divers systkmes de rotation. En même temps, de nombreuses recherches
portant sur les prkiateurs, les exigences en phosphore, la conservation d’Azolfu en saison
chaude, etc..., ttaient entreprises.
108
L’explosion des années 1970
c’gst durant les annees ‘1970, en particulier grâce aux publications vietnamïemres
plus ais&nent accessibles que celles provenant de Chine, que les rnilkux scientifïques
extkaiy a ces deux.~~ys pr$ent rkllement conscience de l’inter& potentiel d’Adla
eti igri~m :; @~Ymonographi~,de Moore (1969) joua certakment un r& knportant
acet~~.La,aupointen1%7&latechniquede mesure de sactiti nitrog&
aisue QJJ pstchromatographie gazeuse suscita par ailleurs un regain g&.&al d’intfZzi?t
pour l’&ude de la fixation biologique de l’azote. La crise pétroliere de 1973 enfin fut
un stimulant puissant a la recherche sur les engrais biologiques A cette @que com-
mencent dès @rs B se multipliex les programmes de, recherche sur ,&oZZu de par le monde.
Parmi les th&nes’de~recherche initiés et les acquis majeurs des ann&s 1970, on
retiendra particuli&ement :
‘k &f&n&o~ de la localisation de la nitrogtkase au sein d’Anubuenu
l
(Peters and .Mayne, 1974) ;
les preniières applications de la méthode isotopique au lsN pour l’étude
l
des relations entre les partekes de la symbiose (Pe&rs et al., 1977) ;
la description d&aillee du cycle de ~développemént d’AzoZZa (Konar and
l
lbpoc=, 1974) ;
l’am&xation des connaissance sur les relations morphologiques et phy-
l
siologiques entre Azolla et Anabama @Sers, 1984) ;
les tentative de culture in vitro d’Anabaena ~azollae (Bai et al., 1978 ;
l
Newton and Herman, 1979) ;
la confirmation de l’importance du phosphore (Subudhi and Watanahe,
l
1979).; . : ‘,
,
‘.d -“:9 la’conf&n&ti~~~~de l’effet «herbicide» d’Aha .(N$, 1973 ; Talley et
az., 1977) ;
. le développement des recherches sur la composition chimique d’AzolZa
et sur son utilisation comme aliment (Buckingham et al., 1977, 1978;
Subudhi and Singh, 1978, 1979) ;
l’étude de la d6composition d’AzoZla dans le sol (Brotonegoro and Ah-
l
,
$dkadir, 1978) ;
les essais de compostage d’AzolZa (Tran and Dao. 1973) ;
l
l%mod&on et la diffusion d’Azolla jiliculglaes:en Chine en 1977, et
l
. .
,.! des autres .esp&=.+xr 1979 &un#n ,a& mucloleta, 1982)
l’organisation des premiers essais coordonnb de l’International Network
l
on Soil Fertility and Fert&er Evaluation for Rice (INSEEER) sur AzolIa
(Watanahe, 1987) ;
l’@iation des pmmi&res grandes collections de &erence d’AzolZa 31 l’in-
l
temational Riche Research Institute, aux Philippines et a I’Unviversiti
Catholique de Louvain, en Belgique.
Les ann6es 1990 : I’Age d’Or 3
La stimulation de la recherche sur AzoZla pendant les anntks 1930 et les espoirs
qu’elle suscite sur le plan des applications donnent lieu à l’organisation de conf6rences
109
internationales CI Porto Rico (Silver and Schroder, 1984) et en Chine @IRI, 1987) et
de cours de formation sur Azollu. au P&ou et au S&I&~ en 1983, en Thailande en 1984.
en Chine en 1987.
En 1982, la FAO et l’Agence Internationale de I’Energie Atomique développent
un progmnme de recherche coordonm? sur Azolla, tandis que paraît une excellente et
tr&s compl&e monographie (Lumpkin and Pluknett, 1982). qui fait encore autoriti à ce
jour. En Chine, le CentreNational de Recherches sur Azolla (NARC) est inaugun5 2
Furhou, en 1985. Le rythme des publications continue a s’acc&%er et des prog&s décisifs
emq$str&s ,:
‘* <
le prob&ne~de l’identification d’AzoZZu au niveau spécifique et subspe-
cifique, crucial pour les milisateurs, s’il n’est pas rt%olu, fait l’objet d’études
approfondies basées sur les méthodes classiques (Pe&ins et al., 1985 ;
D&am and fowler, 1987), et de chimiotaxonomie (Mai Kodomi et al.,
1984 ; Zimmerman et al., 1984 ; Zimmerman et al., 1988) ;
la cara&risation des Anabaena pr&ents dans les diverses e~pkes d’Azolla
l
progresse grike à la g&r&ique mol&rlaire (Cohen - Bazire and Franche,
1985 ; Franche and Cohen - Bazire, 1985), et aux techniques des anticorps
monoclonaux (Gates et al., 1981 ; Ladha and Watanabe. 1982 ; Tang
et az, 1987) ;
l’intervention d’un troisième partenaire, probablement obligé, de la sym-
l
biose, a savoir une bactérie (du genre Arrhrobacter ?) semble se confirmer
(Gg et al., 1980 ; Wallace and Gates, 1986 ; Formi and Grilli Caiola,
.
- ,
l’insensibïïtt! relative de la nitrog@ase du complexe Azolla - Anabaena
l
&IX ,compo& azptés est mise en évidence, avec de fortes diff&ences
sp&ifiques’ et’ subsp6cifiques (Okoronkwo et aZ., 1989) ;
des différence considdrables quant a la valeur alimentaire des diverses
l
espèces d’AzoIZu sont démontrrees (Antoine et al., 1986 ; Sanginga and
Van Hove, 1989) ;
les $remih hybridations entre esp&es d’AzolZa sont r6alis4e.s (Wei et
l
af., 1986) ; dans le même ordre d’idée, de nouvelles combinaisons sont
obtenues par transfert d’Anabuenu provenant d’une espèce d’AzolZa 21 une
autre- esp&% pn%lablement dt%aras& de son propre symbionte (Lin et
al.;388);:; L
;
&.-&t&.h G!@j&&on de biogàz a p& #holIa sont entrepris
-
l
(charlier et az.. 1983) ;
un systhe de cultm-e intiti Riz-Azolla-poisson est testé, apparemment
l
avec des r7%uhats t& positifs (L.iu, 1987) ;
le pouvoir ~accumulateur de potassium d’AzoZZu est mis en évidence
l
ainsi que divers avantages de la culture d’AzoZZa sur sol humide (Liu,
E%n ;
enfin, les Uanges de souches entre centres de recherche et les essais
l
d’adaptation dans diverses 6coxones s’intensifîent.
‘. . .
:
:
’
:
.,.i
.._
:..
.
.
‘.’
_
.’
Y
.”
..;
.:
,Y.,._,
.,
.,.
.
.
.
-....
110
Les principaux modes actuels d’utilisation d’Azo!la
Azolla comme engrais vert
C’est principalement en culture h-r&& qu’Azo&z est utilis6 comme engraisvert.
nyestc~tiv~enassociatianaveclerizetenfolrigImeou&~reprisesau~du
&veloppement de ce dernier ; lorsque le calendrier cultnral et la dïspom%iW en eau
le permettent, une mokcuhure d’tlzolla et un enfotiat avant repiqkge du riz peuvent
~JII%&X la dam associée.
ces pratiques foumisent de I’amte au riz mais a outre inhibent les adve&ces,
am&oreIlt la structure du sol et diminuent les pertes d’ea&l par kipotmnspi&œl.
D’autres cultures inigukes, telles que celles du Lotus et de la Colocase (ta@,
sont Cgalement fert&&s de cette manike.
Azollu peut -enfin servir d’engrais vert en cultures maraîchères ou horticoles de
pleine terre (par exemple jardin potager jouxtaut les p&ni&res de maintenance) ; il-peut
y être incorpo~ au sol & l’&at frais ou sec (en vue de facilitez le transport ou de différer
son emploi), ou apri3s compostage, seul ou en m&nge avec d’autres composants tels
que la paille de riz.
Azolla comme aliment
:,..
L’amBioration des cormaissances sur la valeur alimentaire d’Azont.l, G lltkmsité
de valorisez la biomasse produite dans les pépiniikes en dehors des saisons,d’utili&ion
en r%re, l’extention de l’azolliculture dans le nord de la Chine, r&ion à traditiond’élevage,
ont eu pour effet de stimuler l’emploi d’AzoZlu comme aliment pour divers animaux,
principalement porcs et volaille mais aussi ruminants et plusieurs esp&Xs & poissons
(Su~hudhi and Singh, 1978 ; FAO, 1985, 1985b, ; Que@ii et al., 1986)2n &gle gt%-
rale, il ,semble qu’AzoZZa puisse se su~tituer à en@o~..2@% des raisons classiqks, mais
il faut reconnaître que des 6tudes compl~mentaiks sont indispensables pou+ &ner ces
estimations. Quoi.qu’il en soit cette situation d’AwZZa est en forte expansion actuellement.
Azolla en systbmes d’exploitation complexes
Comme signal6 plus haut une technique promeueuse consistant a associer riz, pois-
sons et Azolla a 6t6 rtkemment mise au point, chacun des partenaks b&&iciant de la
prt5sence des autres. Cette--technique se développe actuellement dans plusieurs pmvinces
de Chine (L~U Chung Chu, Comm. Pers.).
+
Un systkme Mgr6 impliquaut culture d’Azo&z,, &wage de volailles en voli&res
1 ti piloti su$om&nt des étangs a Tilapia et &eva&‘de porcs est &alernent B l’essai
en Côte dIvok
La production de biogaz enfin, par fermentation anakobie d’un m6lange d’AzoZZu
et de paille de riz, et la rkup&ation de l’effluent du digesteur pour fertiker les cultures
voisines semblent donner des &suk& positifs aux Philippines (Johnson, 1988).
La situation en Afrique
Les premiem rtisultats portant sur le potentiel d’AzoZZu en agriculture africaine (Roger
et Reynaud, 1979) suggkaient qu’Atolla &icanu (A. pimuta var. pùmata) pourrait être
valoris6 dans la val& du fleuve S&@al pour autant que le probEme du manque d’eau
111
en saison sMe soit rksolu. Des essais d’adaptation de cette souche en riziere mont&ent
cependant sa sensibilite aux basses temperamms de décembre a f&ier (Ton That, 1980).
En 1980, I’ADRAO lance, en collaboration avec Wniversid Catholique de Louvain
(Belgique), un programme derecherche ayant pour objectif de tester le potentiel de valori-
sation d’Azdla dans:diverses conditious kologiques d’Afrique. de I’Ouest. Les premiers
essais sont rtMis6s a Richard Toll ‘(Sénégal) et B Rokupr (Sierra Leone) ; ils portent
sur l’adaptation d’une-vingtaine d’kotypes reprkentant les sept espkes et deux vari&%
d’AmIla. 11 apparaît rapidement que certains écotypes s’adaptent bien dans chacune des
stations, et que l’introduction d’espkes &rangeres est essentielle, les souches d’origine
loeale~&ant..en.-gMral p e u performanks..:
‘_
Suite ii ces r6sultats encourageants, la collection de refere.nce est progmssivement
enrichie en vue d’identifier des souches a haute productivité ; des m6thodologies adaptées
à l’étude compark de ces souches en conditions contrôhks et en champs sont mises
au point
“, . .
Les premiers essais de fertilisation -des rizieres sont entrepris en 1981 ; les effets
escomptes (apport d’N, inhibition des adventices) sont manifestes
Un cours et des stages de formation sont organises a l’intention des chercheurs
des pays membres de I’ADRAO et une plaquette sur Azolla est difkst% dans les milieux
intétessés (Van Hove er al., 1983).
Ces actions, ainsi que di- iutementions de la FAO, sont suivies de la mise
en place de programmes de reche,rche dans plusieurs pays d’Afrique de l’Ouest, ainsi
qu’au Burundi, au Rwanda et en Tanzanie.
Plus rkemment enfin des acti&s de vulgarisation et d’encadrement en milieu
paysan ‘ont d6buté.
De l’expérience ahsi acquise (Diara and Vand Hove, 1983 ; Diara et al., 1984a,
1984b. 1987 ‘i Van ‘Hove,: 1987, ;:;..Vji iuove and Diara, 1987a, 1987b ;. ,Van Hove er
al., 1983, ‘1987)“Se dégagent. les kon&isions suivantes :
’
les conditions climatiques de nombreuses r&ions d’Afrique sont fâvo-
l
rables a une bonne, ou même à une excellente productivité d’AzoZZa, et
ce pendant toute l’année, pour autant qu’un choix approprié d’kotypes
soit effectue ;
les prédateurs d’Azq& qui empêchent ou contrarient souvent l’utili-
l
sation â’Azolla en Asie, ne provoquent pas de dégât sérieux en Afrique.
Cet& situation n’est cependant sans doute que momentan&z et il faut s’at-
I
tendre, s~.yt~$t+y$y .y!? tr@=Ja hqi$y. B =: que. œmh~ de
ces pt-édatéurs’ tiiiiumyu, &ja pr&entS en A~&U< posent pkbli?me au
fur et à me.s&$re I’azolliculture se d&eIopptG ;
le manque de maîb+e de l’eau. dans une proporGon élevée des zones
l
agricoles africaines. constitue probablement le kteur physique limitant
le plus habituel ;. on doit esp6rer cependant une am&ration progres-
sive de la situation dans ce domaine, amt%oration d’ailleurs requise pour
une meilleure production & riz, ind@endamment d’AzoZla.
pour autant que la culture d’Azolla soit bien pratiquée, ses principaux
l
effets en riziculture sont manifestes : augmentation sensrbledelaproductivin5
(chaque enfouissement d’AzoZZa &ant 6quivalent a l’application de 30 a
40 unités d’azot&a), et diminution du développement des adventices (et
112
donc des travaux de d6sherbage) ; il faut y ajouter l’konomie en eau
-
i
et l’amt%oration de la structure du sol ;
les tentatives d’introduction de l’azolliculture en milieu paysan sont ac-
l
a-.
:.,:. ic: :; i.’ ,~c’&$,@ diversement selon la structure des exploitations agricoles. Dans
-.; ~ ; .,,. y.:, * ! - 5i, &$ iii@ à foa.,e++raUent technologique (distribmion des intrants ,as-
, ,~“.s$e? mr+mwtion pousA...) elles semblent avoir peu de chance d’aboutir,
la +otivation des agriculteurs &ant jnwfkinte pour soutenir un effort
pr$+ig6 de .maîhise des pratiques culturales reqkes. Ailleurs par contre,
.
_ ,:,,,ief x-t%ctio~ sont souvent positives, et dans quelques ca& tles lime
-
,.
I., ,.,jl,~~~l$+, actuelle,iLest vrai-,les paysans prakpmt dt5jZt effectivement
-.
‘.
et $e leur propre inihative I’azoUiculture.
Dkcusslorr .ti &nc!usions
. . ,,a. ,
‘.::’ <’ 7 .
:.
?
&es ipz.heFhes sur Azollu ont considkablement progre& ces 20 dernieres années
en m&& temps que s’am6lioraient et se diversifîaient les pratiques ~ulturales destinées
ii ‘tirer profit de cet+ plaqte et que s’accroissaient les superficies cultivk.
Depui.4 quelques temps cependant, un courant d’opinion peu favorable à la poursuite
: des !+F!qkF: œ ,@n+ne ig dessine dans œrtains milieux, les arguments invoqué
plus ou moms .exp@ement &ant l’effondrement du prix des engrais de synthèse sur
.,._
le marche mondial, la diminution de l’emploi d’AzoIla comme engrais vert en Chine,
ainsi que le faible pourcentage de Assite dans les tentatives d’introduction de l’azolli-
culture dans d’autres pays.
Si l’on consid&re le cas particulier de l’Afrique, il faut cependant reconnaître que
ces arguments~,sont~peu convaincauts. Même au prix actuel (qu’il serait sans doute im-
pN*;%@*a.q mm?: immgble)~4J~qqpis,commexcialxesteraxans doute encore
longtemps imxcesible au petit pays& 6loignC des centres de distribution et daus Ce cas
AzoZZa peut repn%enter une aknative,,.i
‘~‘Q&tu’lar&ession
de l’&ohkuhure enkine, il faut signaler qu’elle effecte
principalement 1’utiUation d’AzoZZa en monoculture, et dans une moindre mesure en culture
assoc& avec le riz; tandis que la production pour lWmentation animale et en systkme
intkgre est au contra& en expansion. Les raisons justifiant cette r6gression partielle sont
l’am&ration de la situation konomique qui facilite l’ac&s aux engrais commerciaux,
la diminution de la main d’oeuvre disponible dans les campagnes et la nt!cessité de consacrer
la total& ducalendrier 8 la production alimemare, compte tenu de la forte densit6 de
population ; c’est ainsi que la motroc&m5hiveanale d’Azo2ki’y’êst de plus en plus souvent
_ remplack par des Vc#ures maraîchkes ou & ble d’hiver. De teks raisons ne peuvent
pas’être invoq+ dans les conditions g&k%ales actuelles de lkgriculture atricaine.
Ilestvraienfînqueles
pmgrammes de recherche t+lant Zt introduire l’azolliculture
dans plusieurs pays‘dYAsie n’ont pas souvent (mais parfois) 6tt5 couro~ de suc&.
Les causes en sont sans doute diverses~Dans cer&s cas, des conditions similaires a
celles prdvalant actuellement en Chine peuvent être invoqu&s ; dans d’autres, les condi-
tions du milieu sont effectivement défavorables au dkloppement d’AzoZla (abondance
de prklateurs, temp&ature et humidit6 excessives, alcabuite du milieu favorable à la prolife-
ration des algues...), dans d’autres, enfim c’est l’insuffisance des moyens mis en œuvre
pour la recherche...), la formation et/ou l’encadrement qui est en cause. Pour de nom-
breuses r&ions d’Afrique les deux premières raisons ne peuvent certes être invoquées,
; .,_. :
113
et les n%uhats de la recherche sont en outre dés & prksent suffisamment positifs pour
y justifm l’introductiou de l’azolliculture.
La question se pose d&s lors de savoir quels sont les moyens a mettre en œuvre
pour que le paysan africain tire effectivement parti & ces acquis. L’expkence montre
qu’un encadrement suivi est essentiel; au moins dans un premier temps. Un personnel
compkent doit donc &re forme et des conditions de travail lui permettant & se tenir
en permanence a la disposition des paysans doivent lui &re fournies. La prise en charge
de ce personnel sera assur& selon les cas par des n%eaux nationaux de vul~on,
des coop&atives, des organisations non gouvernementales. Si on se re&re aux pays où
Azollu est effdvetit valc+rist! ii i’heure actuelle; ‘force est de constater que le dkvelop
pement de l’azolliculture y a toujours &e promu de marriere tri?s directe par les pouvoirs
publics qui y ont vu un investksem em, non négligeable, sans doute, mais rentable à terme
; en cas de rkssite de l’opkation les d@enses consenties prkentent en effet le triple
avantage de ne pas requérir de devises &rangères, de n’être pas récurrentes (contrairement
aux dépenses de subsidiation des engrais) et de toucher une tranche de paysannat particu-
lièrement d&vorisee. ll faut noter enfii qu’une telle stratégie ne comporte pas de risques
élevés, compte tenu de ce qu’elle peut fort bien être testée. à titre probatoire, dans des
régions limitées. Son succks dépendra du choix des zones d’introduction ; certaines sont
en effet plus prometteuses que d’autres (maîtrise de l’eau, conditions climatiques, stmc&ms
du paysannat) ; il d@endra peut-&-e surtout de la formation du personnel’d’encadmment
et de sa disponibilité sur le terrain.
Rbfbrences
ANTOINE, TS., CARRARO, S., MJCHA, JC. and VAN HOVB, C. 1986 Comparative
,‘, :- appetency for Azolla of Cîchksoma and Or-eochrornis (Tiipia). Aqua&ure ‘53,
95-99.
’
’
BAI, K.Z., YU, S.L., CHBN, WL. and YANG, S.Y., 1978 Cultivation of Alga-free
Azolla, isolation of Anabaena azollae and preliminary attempt at their association.
In : Proc. Symp. on Plant Tiie ~Culture, Peking, 23-27 May 1978, Sciences Press
PubL Peking. pp. 455458.
BRABMER, P., 1927 La culhm des Azolla au Tonkin. Rev. ‘Bot. App. Agric. Col. 7,
816819.
_
BROTONEGORO, ST and ~ABDULKADIR, S. 1978 The decomposition of Azolfu pinnata
.e in moist and$ooded.,,it Ann.. Bogor., 6. 169-175.
: i
~BtiCKJNG~~ Kti., ELA, S.W., MORRIS, J.G. and GOLDMAN, C.R. 1977Protein
quality evaluation sud proximate analysis of AzoZZa fiZicnZoiifes Fed. Proc. 36, 1179.
BUCKJNGHAM, KW., ELA, S-W., MORRIS, J.G. and GOLDMAN, CR. 1978 Nutri-
tive vatue of the nitrogen-fixulg aquatic fern Azolla jïlîculoïdes. J. jlgnc. Food Chem.
26, 1230-1234.
CHARLIER,S.,LEGROS,A.ASJNARJDXSANMARZAN O,C.M,NAVEAU,H.,NYNS,
EJ., VAN HOVB, C. and GODARD, P. 1983 Biomethanation potentialities of Azolla.
In : 3d Intem. Symposium on Anaerobic Digestion, 14-20 Aug 1983 Boston, USA
PP- 7.
114
CHEVALIER, A. 1926 La culture d’Azollu pour la nourriture des animanx de basse-
cour et comme engrais vert pour les rizi&res. Rev. Bot. App. Agrk Col. 6.350-360.
COHEN-BAZIRE, G. and FRANCHE, C. 1985 Restriction sites in the nif H, D, K gene
region qf habaena uwllae endosymbionts from nine different Atolla qxcies.:In :
fitrogen fixation research progress. Evans, H.J., Bottomley PJ. and. NemqW.E.
Eds. M&inus Nijhof Publ. Dordrecht, p. 146.
DAO ‘Ihe Tuan and TRAN QUANG THUYET 1979 Use of AzolZa in Vietnam. pp.
395405. In : Nitrogen and rice, IRRI, Ed-, Los Banos.
DIARA, H.F. and VAN HOVE, C. 1983 ,The influence of AzolZa on rice productivity
. .
Pdmmary results obtained at WARDA Richard Toll project - Senegal. WAkDA
tecticaI NewsZetter 4, 7-8.
DIARA, HF., DD[oN, C.A. and VAN HOVE, C. 1984a Adla in West Africa : first
results IÏom the WARDA program. In : Developments in plant and soil sciences :
practical application of Azolla for rice production. Silver, W.S. and Schroder, E.C..
Eds., Martinus Nijhof Publ. Dordrecht, pp 202-207.
DIARA, H.F., DIXON, CA and VAN HOVE, C. 1984b Influence d’Azolla sur la produc-
tivité du riz à Rokupr, en Sierra Leone. Bulletin technique de PADRAO 1, 3-4.
DIA&& m., VAN BRANDT, &, DIOP, A.M. and VAN HOVE, C. 1987 Azolla and
its mie in riœ culture in West Africa. In: AzolZu utilizatï~ IRRI, Ed., I.oS.Banos
pp. 147-152.
DUNHAM, D.G. and FOWLER, K. 1987 Taxonomy and species recognition in AzoZlu
Lam. In: holIa utilisation, IRRI, Ed, Los Banos pp. 7-16.
FAO 1985a Propagation and agricultural use of Azollu in Viehxun, ZRCN. 34,257~263.
FAO l!$b Utilizat@ of Azollu in China. IRCN. 34, i65-273.
.)1
FORNI,’ C’and GRILL1 CAIOLA, M. 1988 Physiological cbaractekation of the bacteria
isohaed f.kom the aquatic fem Azollu Lam. p. 235. In : Nitrogen fixation : Himdred
years afta. Bombe, De Bruyn, Newton, Eds., Gustav Fi&er, Stuttgart.
FRANCHE, C. and COHEN-BAZIRE, G. 1985 The structural nif genes of four symbio-
tic Anubuena uzollue show a highly conserved physical arrangement Plant Sci. 39,
125-131.
GATES, JB., FISHER, R-W, and CANDLER, R.A. 1980 The occurence of coryneform
bxteria in the, leaf cavity of Azolldrch. Microbiol. 127, 163-165.
GAT@. 3.E.. FISHER, R-W.. GOGGIN, T.W. and AZROLAN, N.T.. 1981 A ukparïson
: Of @é kfkx antigedicy of the N-fïxing cyanobacterial symbionts of Ada pinnuta
A carolinianu and A. microphilla. Plant Physiol. 67, 134.
IRRI 1987 Azolla utilization. Proœedings of the Workshop on Azolkà Use, Fuzhou, Fujian,
chine. 31 Mar - 21 Apr 1985, IRRI, Ed. Los Banos, 2% pp.
JOHNSON, P. 1988 A smala farmer reaps honor and succes frorn Green Gold, Agribusiness
Weekly. 5, 4-5.
KONAR, R.N. and KAPOOR, RK 1974 Embryology of Azollu pinnutu. Phytwnorphology.
24, 228-261.
LADHA, J.K. and WATANABE, 1.1982 Antigenic similarity among Azollu uwllae separaH
from different species of Azolla. Biochem. Biophys. Res. Commun 109, 675-682.
: .. ., .:..
115
LIN, C.,LIU, C.C.,ZHENG, D-Y., TANG, LF: and WATANABE, 1.1988 Re-establishment
of symbiosis to Anabama - Ce& Azolla (in chinese). Acta Scientia sinica (B). 31,
700.708.
LIU, C.C. 1987 Reevaluation of Azolla utiliz&k In : Azolla utikation, IRRI, Ed., Los
Banos, pp- 67-76 ‘...,‘( ii;
:: *., i ,’ .;‘i:“’ ‘28 :
*
LUMPKIN, TA and PLUCKNE’IT, I&I,. 19~2-Az&& a green manure : use and manag-
ement in trop pmduction. Westviey Press, Boulder, 230 pp.
MAI KODOMI, Y., GODARD, P. and VAN JjIOk, C. 1984 F%vonoid analysis in the
genus awlla: Arck: In& Physùik Bfkx~~~2: 34-35.
MOORE, kW. 1969 %ok i && and @$&ok&&nikance.
Bot. Rev. 35, 17-34.
NEWTON, J.W. and HERMAN, AL .197?, 1solat.i~ of cyanobacteria from the aquatic
fcrn Azolla. Ar~h. Microbiol. 120, :161-165.
1
NGO-GIA-DINH ‘i973 The effect of k Azolla caver ‘on the germination of Barnyard
gras (Echinochloa crus-galli (L.) Beau~.). In ~‘Proc. 2nd Indonesian Weed Sci. COI&,
Yogyakarta, 2-5 Apr 1973. p. 201.
NGUYEN, CT. 1930 L’Azolla cultiv6e comme e&ra& vext Bull. Econ. Indochin. (Hanoi).
33, 335-350.
OES,A.l913Uebexdi&ssim&iondesfrei~s&&ffsdurchAzoZ~BotnnischesCentralbl
Il, 164-165,
OKORONKWO, N. VAN HOVE, C. and ESKEW, DL. 1989 Evaluation of Nitrogen
fixation by different stmins of adla &abaena symbiosis in the presence of a high
level of ammonium. Biot. Fert. Soik 7, 275-278.
PERKINS, S., PETERS! 94, LUMP~,~X,awi C+&VERT, H. 1985 Scanning elec@n
microscopy of perine ar&t.e&ure~ “aS a%&&&~ &l k the genus Azolla Gmark.
Scanning Electron MicrOs. 40, 1719-1734.
PETER!$ GA. 1984 AzoZZa ‘&nubaenu symbiksës: basic biology, use and prospects for
the fumure. In : Developments in p@nt and soil Sciences: practical appkation of Azolla
for ricc production. Silver, W.S. and Schr&ier, E-C., Fxls., Martinus Nijhof PubL Dor-
drecht, 1-14.
PETER$ GA. and MAYNE, B.C. 1974 The Azolla-Anabaena azollae relationship. II:
Lmalmtion of nitrogenase activaty as assayed by acetylene reduction. Plant Physiol.
53, 820-824.
ship. V : lsN, fixation, acetylene reducti&‘ kd H, production Plant Physiol. 59,
1021-1025.
QUERUBIN, LJ. ALCANTARA. P.F. and PRINCESA, AO. 1986 Chen&al composition
of three Azolla spekes (A. caroliniana, A. rnicrophilla and A. pinnata) and feeding
value of Azolla meal (A. microphilla) in broilex ration Phil. Agric. 69, 479490.
ROGER, P. and RBYNAUD, PA. 1979 Premiks données sur l’kologie d’Azolla africana
en zone sah&ienne (SCn6gal). Uecol. Plant. 14, 75-84.
SANGINGA, N. and VAN HOVB, C. 1989 Amino-acid composition of Azolla as af-
fected by strains and population density. Plant Soil. 117, 263-267.
:
‘.
,.
-.:
.:_.
.i
,..,,_
116
SILVER, W.S. and SeOIDER, E.C. 1984 Practical application of Azolla for rice pro-
duction. Pmceedhgs of an International Workshop, Mayaguez, Puerto Rico, Nov. 17-
19, 1982. 227 pp.
SMITH, JB. 1910 Azollu vs. mosquitoes. Enromol. New. 21, 437-440.
SUBUDHI. B.P.R. ad SINGH, S. 1978 Nutritive value of the water fem Azolla for chicks.
Poult. sci. 57, 318-m.
SUBUDHI, BSR. and SINGH, P.K. 1979 Effect of macronutients and pH on the growth,
nitrogen fixation and soluble sugar content of water fem Azolla pinnata. Biol. Plant.
21, 66-70.
SUBUDHI, B.P.R. and WATANABE, 1. 1979 Minimum level of phosphate in water for
growth of Azolla determined by continuous flow culture. CUIT. Sci. 48, 1065-1066.
TALLBY, S.N., TALLEY, BJ. and mS, WD. 1977 Nitrogen fixation by Azolla in
rice fields. In : Gene& engineering for nitrogen fuation, Alexander Hollaender, Bd.,
Plenurn Press, New York, pp. 259-281.
TANG, LP., ZHBN, Q. ZHEN, D.Y. and LIU, C.C. 1987 A fluorescent antibody assay
for detecting antibodies to surface antigen on Anabaena-azollae : a prehmiuary report
p. 290. In : Azolla utilization, IRRI, Bd. Los Banos.
TON THAT TRINH 1980. Intmductory notes on Azolla &ricana, an atmospheric nitrogen-
fixing fem in irrigated rice culture; importance and prospects of its use in OMVS
(organisation pour la mise en valeur du fleuve Sénegal) irrigated amas. IRCN 29,23-
27.
TRAN QUANG THUEY and DAO THE TUAN. 1973 Azolla : a green compost. Vief.
StKd. 38, 119-127.
VAN HOVE, C. 1987 Potentiel de valorisation d’Azolla en agriculture afrkxiue. Cahors
dkthologie appliquée. 7, 31-42.
VAN HOVE, C. de WAHA BAILLONVILLE, T. DIABA, H-F. GODARD, P.. MAI KO-
DOMI, Y. and SANGINGA, N. 1987 Azolla collection and selection pp. 77-87.
In I Azolla utilimtion. lRRL, Bd., Los Banos.
VAN HOVE, C. and DIARA, HF. 1987a Azolla introduction in afiican agriculture :
progress and pmblems. IRCN 36, l-4.
VAN HOVE, C. and-DI@A, B.F. 1987b Bxplanatory gui& of a series of slide pictures
illustrating the methods of Azolla utilization in agriculture. 9 pp. 40 slides, C. Van
Hove, Ed Lonvain-La-Neuve, Belgium.
-VAN HOVB, C DIARA, H.F. and GODARD, P. 1983 Azolla en Afrique de l’Ouest
- in West Africa, 56 pp. 11 illmons. Van Hove, C. Ed. Louvain-La-Neuve, Belgium.
WALLACE, W.H. and GATES, J.E. 1986 Identification of Eubacteria isolated fiom leaf
cavïties of four species of the N-Mg Azolla fem as Arthrobacter Conn and Dinmick.
Appl. Environ. Microbiol. 52, 425-429.
WATANABE, 1. 1987 Summary report of the Azolla program of the intematioual net-
work on soil fertility and fertilizer evaluation for rice. pp. 197-205. In : AzoZZa utili-
zation, IRRI, Bd. Los Banos.
WBI, W.X., JlN, G.Y. and ZHUAN, N. 1986. Primat-y report of Azollu hy&idi&on
study (in chinese). Bull. Fujian Ad. Asrie. Sci. 1, 73-79.
;
a>.‘,,.
:
:->
.,,..;.
:
.a :
117
ZIMMERMANN, W. LUMPKIN, T. and WATANABE, 1. 1988 Biochemical taxonomy
of Azolla Anabaena, section Azollu. p. 240. In : Nitrogen fixation: Hundred years after.
Bothe, De Bruyn, Newton, Eds., Gustav Fisher, Stuttgart.
:
_i_’
.:’
.
.
.
.
.
.,.
h4mimi.uz la FBA pour la Production Agricole et ForestiSre en Afrique
Phosphorous needs and accumulation
potential of various AzoZZa species and strains
DESMADRYL, D., GODARD, P.,
WAUTHELET, IMa and VAN HOVE, C.
Laboratory sf Phmt Physiology, Catholic Universiry of Lmvuin.
B-1348 Louvain-la-Neuve,
Belgium.
It is know-n that Apda are able to accumulate phosphorus when grown on
P-rich media, and it has been suggested to take advantage of this luxury consump-
tion, loading up Awlla with P in the nurseries before mating them into the rice
field, ww they could continue to multiply using their excess P. The purpose of
the present styly was to identify AwZla ecotypes with high accumulation power
,ad/or eveutually low P needs, comparing 20 high m tolerant strains. Pour
A. caroliniana (ACA). 4 A. microphylla (Ah@, 6 A. pimata var. imbrica& (API)
and 6 A. pinnafa var. pinnafa (APP) have been grown first on N-free Hoagland
solutions (dihlted 2J5) containing respectively 15.30.45 and 60 ppm P for 18 days.
None. or only dight effects of the treatments are observed on productivity, but Awl[a
P content increases line&y with external P concenhation. strongly in ACA and AhJI,
only slightly in API and APP, and varies widely from one strain to another, even
from the same species. When such variously P-loaded Azolla are transferred
(60 g/n?) on a P-free medium, the biomass they math at the plateau phase is stmn-
gly dependent on the pre%matrnent for ACA and AhJI, not for API and APP ; at
this stage Awl& P content varies fi-om 0.023 to 0.106% (DM.). but is not correlated
withfmalbiomassnorwithspecies.
i
*
.:..,
:
,
I
.-..
119
Introduction
Phosphorous is frequently the limiting factor of AzoZZu productivity (Tung and
Watanabe, 1983 ; Lumpkin, 1987) ; on the other hand Azollu is able to make some luxu-
rious P consumption (Lumpkin, 1987 ; Reddy and Debusk, 1984 ; Subudhi and Singh,
1979 ; Sub&i and Watanabe, 1981), and it has been suggested to make advantage from
this property, enriching AzoZZu in the nurseries before field inoculation (Jumpkin, 1987).
Few data are nevertheless available on the growth potential of P-preloaded Azollu
when cultivated on P deficient media
The purpose of the present research is to identify AzoZZu ecotypes with high ac-
cumulation power and/or low P needs ; the investigation is restricted to a group of high
temperature tolerant strains.
Material and Methods
Twenty Azollu straius from the ADUL collection (Van Hove el al. 1987). previ-
ously selected among 150 for their tolerance to high temperatures (unpublished results)
are tested in this experiment : four A. cmolitiunu (ACA) : ADUL 8, 61, 109, 113 ;
Four A. microphyZZu (AMI) : ADUL 65, $6, 104,175 ; six A. pinnatu var. pinnufu (APP) :
ADUL 98, 99, 111, 129, 136, 144 and six A; pikufu var. imbricufu (API) : ADUL 6,
26, 87, 92, 100, 137.
Experiment 1 : four culture baths containing 40 plastic boxes with the bottom
replaced by a mosquito net (Van Hove et al. 1987) are filled with 120 1 of diluted
(4/10) Hoagland solution in which nitrates are replaced by chlorides and phosphorus.is
added as KH.$O, at respective concentrations of 15, 30,45 and 60 ppm P (initial pH :
6.00, 5.29, 5.19,4.89). Each box is inoculakd with lg (F.W.) AzoZZu, every strain being
duplicated, at random, in the four culture baths. Culture media are replaced after 9 days.
The experiment is realized in a greenhouse in which conditions are : light intensity :
natural + 65 pE.m-2S.-1; air to : 35 f 4°C ; water to : 33 f 4°C ; relative humidity :
47 Ik 5%.
On day 19, AzolZu samples are harvested and their fresh weight measured after
1 mn centrifugation (16 g} of the AzoZZu containing boxes, followed by 20 s drainage
on absorbent paper ;- 1 g of each sample is set apart as inoculum for the second experi-
ment ; the remaining biomass is used for dry weight measnrement and phosphorous ana-
lysis.
r.F.1 : \\ :
..’
j
Experiment 2 : The device is similar to the pmceding one except that the culture
medium (initial pH : 6.05) is P-free, and is renewed every 7 days ; air t!’ : 38 f 5°C ;
water to 37 zk 4°C ; relative humidity : 51 f 5%. Each box is inoculated with 1 g of
variously P-enriched AzoZZu from the first experiment (60 gm-9.
Fresh weight is measnred as above, every 2, 3 or 7 days according to the growth
stage until the plateau phase is reached (two successive not mcreasing biomass values).
Azollu P content is measured at the end of the experiment.
Chemical analysls : Total phosphorous is determined calorimetrically as its re-
duced phosphomolybdate complex, after miueralization in diluted nitric acid.
;.
I
120
Results
Experiment 1 : Effect of P concentration in the medium on Azollu growth and
P cQntent.
Probucfiv~fy - Under the prevailing experime&l conditions A. carolinium
and A. microphylla are the most productive species, followed by the A. pinnata varieties
(Table 1). Increasing P concentration Erom 15 to 60 ppm has no or only slight effects
on productivity.
Table 1 : Mean productivity and standard deviation (gFWm2.day) of 4 Azolla
species or varieties as a function of the external P concentration
(Ppm PI.
Species
15 ppmP
30 ppmP
45 PpmP
60 ppmP
ACA
122.2 2 4.8a
120.8.: 3.7a
118.8 f 5.0a
116.8 + 5.3a
I.,
AM1
' 116;g +, G;.;;'
118.0 2 7.6a
11955 f 4.4a
111.0 _+ 9.0a'
API
75.9 i 4.lb
'70.1 _+ 5.5b
72.7 _+ 8.8b
70.4 _+ 7.6b
APP
62.7 +14.9c
63.9 f 3.0b -
68.7 +10.3b
66.3 + 9.5b
In a column, values with the same letter are not significantly different at
P = 0.05 (Duncan test).
Phosphoms confent - Internal P concentration at the end of the experiment
(Fig. 1) varies considerably Corn one strain to another in a given species or variety. At
low (15 ppm) external P concentration, there is nevertheless no specific or varietal diffe-
rence (Table 2) but P content increases linearly with increasing external concentration,
slightly in the two A. pinnata varieties. for which angular coefficients of the regression
line (internal P content (%)/medium P content (ppm)) vary fmm 0.003 to 0.009, strongly
in A. caroliniana and A. microphylla (angular coefticients : 0.01 to 0.02).
.
.
.
:,‘-:
121
Figure 1 : Three typical behavior of 20 strains of Azolla species or varieties growing on a P deficient medium after a pretreatment on 4 P concentrations : ACA
& AMI (a), best APP (b) and API or other APP (c).
Growth curves as biomass in kgFM/m2 related to time in day after pretreatment at 15 (I) ), 30 ( + ), 45 ( s ) and 60 ppm P (a ).
Maximum biomass in kgFM/m’ ( 0 ) and final P content in %DM (% ) related to P concentration in ppm P of the pretreatment medium.
122
Table 2 : Final P content and standard dev!ation (XDM) of 4 Azolla species
or varieties as a function of the external P concentration
(ppm P).
Species
1 5 ppm P
30 ppm P
45 Ppm P
60 ppm P
ACA
0.94 + 0.14a
1.05 + 0.07b
1.32 + 0.12a
1.53 + 0.16a
AM1
1.00 + 0.19a
1.25 -r 0.17a
1.39 + 0.18a
1.59 + 0.15a
API
0.93 + 0.16a
1.01 + 0.18b
1.11 + 0.20b
1.20 -I
16b
APP
0.88 2 0.16a
0.96 f 0.176
0.99 f 0.13b
1.11 f O.llb
In a column, values with the same letter are not significantly different
at p = 0.05 (Duncan test)
Experiment 2 : AzoZla growth in P-free media as influenced by initial P content.
Maximum biomass - Maximum biomass for a given pretreatment (Table
3) is highly strain dependent in A. pinnutu var. pinnata (high standard deviation), not
in A. pinnata var. imbricata which always shows low plateau values, or in the two other
species. As expected from results obtained in the first experiment, A. carolinianu and
A. microphyllu final biomasses are highly influenced by the pretreatment, whereas A. pinnatu
are practically not. Fig. la and lb illustrate these two typical growth patterns. Also for
A. caroliniana and A. microphylla there is a good correlation, at the species level, between
initial P content and maximum biomass. For A. pinnatu on the other hand such a correlation
only exists at the strain level as illustrated in Fig. lb and c.
Final Azolla P content - Azolla P content at the plateau phase (Table
4) is generally higher in A. microphylla and A. caroliniana than in A. pinmta, which
presents very high intraspecific variations. It must be mentioned that the observed values
are higher than the calculated values (initial P content. initial biomass. final biomass’) ;
this has to be attributed to traces of P brought in the system by demineralized water
compensating evapotranspiration.
123
Table 4 : Final internal P content and standard deviation (% DM)
of 4 Azolla species or varieties cultured in a P deficient
medium after a pretreatment at different P concentration
(Ppm P)
Species
Pf(%DM)
ACA
0.076 + 0.009 b
AMI
0.087 + 0.013 a
API
0.056 + 0.022 c
APP
0.041 + 0.018 d
In a column, values with the same letter are not significantly
different at p = 0.05 (Duncan test).
Table 3 : Maximum biomass yield and standard deviation (kgFM/mZ) of 4 Azolla
species or varieties cultured in a P deficient medium after a
pretreatment at different P concentration (ppm P).
Species
15 ppm P
30 ppm P
45 ppm P
60 ppm P
ACA
2.4 + 0.4ab
2.7 + 0.3a
3.4 + 0.3a
4.1 + 0.4a
AMI
2.0 + 0.3bc
2.4 + 0.3a
2.9 k 0.4b
3.6 + 0.5b
API
1.6 + 0.3~
1.8 + 0.2b
2.1 + 0.3~
2.0 + 0.3d
APP
2.5 5 0.8a
2.6 2 0.6a
2.7 + 0.6b
2.9 _+ 0.6~
In each column, values with the same letter are not significantly different
at p = 0.05 (Duncan test)
124
Dlscusslon
Subudhi and Watanabe (1981) have shown that A. pinnuta growth does not improve
when external P concentration increases from 5 to 10 ppm, whereas Singh ef al (1984)
obtained growth increases up to 15 ppm P (the highest concentration tested) forA.fiZicuZoiaks
and A. mexicana, to 10 ppm for me A. pinnata strain, and to 5 ppm for another. Cary
and Weerts (1982,1984) observed better A. jXcu2oide.s growth at 20 ppm than at 5 ppm ;
for A. pinnata optimal growth was reached in the 5 to 20 ppm range, with a significant
decrease at 40 ppm.
In the present study, the P concentration range was extended to 60 ppm, which
is very high according to agricultural practice standards, with the view of testing phos-
phorus accumulation potential of various AzoZla species and strains. Results are generally
in agreement with previous information, except for the fact that no depressive effects
of high P concentration were generally noticed
Subudhi and Watanabe (1981) also found that P content in A. pinnata increases
linearly with increasing external P concentration between 0 and 10 ppm. We observed
the same behaviour for the 20 strains tested between 15 and 60 ppm P ; the fact that
productivity is not affected in this range of concentration confirms the luxury consumption
capacity of Azollu (Lumpkin and Pluchnett, 1982). Significant interspecific differences
nevertheless exist, A. caroliniana and A. microphylla having a higher accumulation power
than A. pinnutu ; considering that they are also characterized by a good correlation bet-
ween P content and maximal biomass, over fertilization of these two species at the nur-
sery level seems justified. If P fertihzer availability is restricted on the other hand, some
A. pinnata var. pinnata should be preferred.
Looking at the Azolla P content at the plateau phase in P deficient culture media,
it appears once more that A. caroliniana and A. microphylla differ from A. pinnata, with
a higher P content and a lower intraspecific variability. The extreme values are 0.023%
and 0.106% ; assuming a 1.2% P content in the inocuhtm, which is easily obtained with
a good P fertilization in the nursery, this means biomass multiplication potentials, in a
P-free medium, of respectively 52 and 11. Azolla strains with low minimal P content
L::y..c ‘l7t.p
. r
r
,.
mmpq:-1.. :,T b,z ;a,nbiJ.l’- ’ ;..
L . . . . . . . . ..I. a.,
_ cu ALA ,&c;iol~ j;iUt;itiII.
I\\i~l’i; iikuiuduu* MUUIU IK,VU
theless be necessary concerning N content of Azolla in these extreme conditions since
it is known (Watanabe et al., 1980) that P and N contents are correlated.
Furthermore, the capacity of a given strain to reach a high biomass in P deficient
culture media is not attributable to lower P needs, final biomass and final P content being
poorly correlated (r = 0.35); this means that strains having similar initial and final P
contents can differ by their fmal biomass, due probably to differential capacities of their
root system to extract traces of phosphorus present in the culture medium, as suggested
by Subudhi and Watanabe (1981).
Conclusion
The aim of the present research was to estimate the range of phosphorus needs
and phosphorus accumulation power of various Azollu species and strains, with the view
of introducing, eventually, such criteria in Azolla selection program.
125
Based on a limited number of strains, the conclusion is that the two criteria have
to be considered, somme strains from the A. pinnutu group being characterized by their
low treshold P level, some others, from the A. caroliniana and A. microphylla group
having a high phosphorus accumulation power.
References
CARY, P.R. and WEERTS, P.G.J. 1982 Nutritional and water temperature factors affec-
ting growth of aquatic plants. Rex Rep., CSIRO Div. Irrig. Res. 1981-1982, 37-40.
CARRY, P.R. and WEERTS P.G.J. 1984 Factors affecting growth of Azollu spp. Res.
Rep., CSIRO Div. Irrig. Res. 1983-1984, 43-45.
LUMPKIN, T.A. and PLUCKNEIT, D.L. 1982 Azolla as a green manure : use and ma-
nagment in crop production. West View Press, Boulder, Colorado. 230 pp.
LUMPKIN, T.A. 1987 Environmental requirements for successful Azolla growth, p.
89-97. In : Azollu Utilization. IntemationaI Rice Research Institute, Manila.
REDDY, KR. and DEBUSK, W.F. 1984 Phosphorus removal by Azollu curoliniana
cuhured in nutrient enriched waters. p. 151-162. In : Practical application of Azollu
for rice production. Silver, W.S. and Schroder, E.C. (eds). Martinus Nijhom W.
Junk, Dordrecht.
SINGH, P.K., PATRA, R.N. and NAYAK, S.K. 1984 Sporocarp germination, cytology
and mineral nutrition of AzolIa, p.5572. In : Practical application of Azolla for rice
production. Silver W.S. and SchrUder E.C. teds). Martinus Nijhofm W. Junk, Dor-
drecht.
SUBUDHI, B.P.R. and SINGH, P.K. 1979 Effect of phosphrus and nitrogen on growth,
chlorophyll, amino nitrogen, soluble sugar contents and algal heterocysts of water fern
Azolla pinnata. Biol. Plantarum 21, 401-406.
SUBUDHI, B.P.R. and WATANABE, I. 1981 Differential phosphorus requirements of
Azolla species and strains in phosphorus-limited continuous culture. Soil Sci. Plant
Nutr. 27, 2??-347.
TUNG, H.F. and WATANABE, I. 1983 Differential response of Azollu -Anubaena as-
sociations to high temperature and minus phosphorus treatment. New Phytol. 93,
423-431.
VAN HOVE, C., DE WAHA BAILLONVILLE, T., DIARA, HF., GODARD, P. MAI
KODOMI, Y. and SANGINGA, N. 1987 Azollu collection and selection, p.77-87.
In : Azolla utilization. International Rice Research Institute, Manila.
WATANABE, I., BERJA, N.S. and DEL ROSARIO, D. 1980 Growth of Azolla in paddy
field as affected by phosphorus fertilizer. Soil Sci. Plant Nub-. 26, 301-307.
Maximisez la FBA pour la Production Agricok et Forectit?re en Afrique
Sensitivity to Aluminium of various AzoZZa species
and strains
WAUTHELET, M., GODARD, P.,
DESMADRXL, D. and VAN HOVE, C.
L&oratory of Plant Physiology, Catholic University of Louvain,
B-1348 L.ouvain-la-Neuve,
Belgium.
Abstract
High alumivillm rnnwntTrt:nn iq 2 +r?j” Fnhl-- ir: mz~)r ~ci~.Ic tr-;i;::! 2 -i!:,
where it interferes among others with phosphorus availability, an element especially
important for Azolla. Information on Azolla sensitivity to Al is nevertheless mis-
sing, and it was the purpose of this preliminary research to estimate the range of
Al concentrations allowing proper Azollo growth and to select, eventually, Al tole-
rant straim. Twenty high temperature tolerant strains, four A. caroIiniona (AC&.
” four A. microphylla (Ah@, six A. pinnata var. imbrkata (API) and six A. pinnata
var. pinnuta (APP), were grown on N-free 2/5 Hoagland solutions with respectively
0.2.5 and 8 ppm Al, and their biomass (fresh matter) was measured after 11 days.
Results show that, according to the strain productivities decrease with increasing
Al concentration either +_ linearly or exponentially. At the species level API is the
most sensitive, followed by ACA and AMI and then by APP. Intraspecific diffe-
rences arenevertheless high, except for API strains. According to the strain, productivities
at 2 ppm vary from 100 to 50% and at 8 ppm from 73 to 16% as compared to
the control.
127
Introduction
Aluminium toxicity is an important growth limiting factor for plants in many aci-
dic soils, in which Al is easily solubilized. Al toxicity can induce mineral nutrition distur-
bances, such as P (Randall and Vosc, 1963) and Ca (Awad et al., 1976), uptake decrease ;
it also inhibits cell division (Horst et al., 1983).
As far as the aquatic fern Azolla and its endosymbiont Anabaena are concerned,
only scarce information is available (Lumpkin and Plucknett, 1982). The purpose of the
present research is to estimate the range of Al concentrations allowing satisfactory Azolla
growth, and to select eventually Al tolerant ecotypes (among a group of high tempe-
rature tolerant strains).
Material and Methods
AZ&Z strains as well as the experimental device are identical to those described
by Desmadryl et al., (page 118 in this volume). Culture media contain 0, 2, 5 and
8 ppm Al as Al, (S0,),.18H,O), their respective pH being 4.80 f 0.41, 4.14 + 0.11,
3.72 f 0.04 and 3.59 f 0.06 during all the experiment. Each box is inoculated with
5 g (FW) Azda (& 300 g/m”) and each strain is duplicated. The conditions in the green-
house are : light intensity : natural + 65 pE.m-2.s~1; photoperiod 12/12 h; air to :
31 * 2T; water to : 30 rt 2°C; relative humidity : 55 + 5%.
After 11 days AzoZZa samples are harvested and their fresh weight measured after
1 mn centrifugation (16 g) and 20 s drainage on absorbent paper.
Results
In the conditions of the experiment, the productivity of the controls (0 ppm Al)
varies strongly between strains, but also between species and varieties (Pig. 1) A. caroliniana
and A. microphylla are very similar and the most productive, followed by A. pinnata
var. pinnatu and finally by A. pinnata var. imbricata. Productivity of all the strains de-
creases with increasing Alummium concentration, eilher exponentially, for the most sensi-
tive strains, or more or less linearly for the others. Concentrations of 2 and 5 ppm Al
only slightly affect growth of tolerant strains, while for the sensitive ones, growth is
already strongly depressed. At 8 ppm Al only one strain (ADUL 136 PP) grows very
poorly. A. pinnutu var. pinnata is the most tolerant variety (Table l), A. pinnuta var.
imbricatu the most sensitive, the two other species being intermediate. Intraspecific and
intravarietal differences (except for A. pinnatu var imbricata) are nevertheless high.
Discussion and Conclusion
Sensitivity to Al varies widely among plants. Gossypium hirsutum root growth
for example stops above 0.5 ppm Al (Rios and Pearson, 1964), other plants are affected
at 1.5 ppm Al (Pay and Brown, 1964) whereas Vigna unguiculuta sustains 2 - 4 ppm
(Horst et al., 1983), Sorghum bicolor 8 ppm (Purlani and Clark, 1981). Pennisetum typhoides,
16 ppm (Long et al., 1973) and Cumelia Sinensis up to 27 ppm at pH 6 (Matsumoto
128
et al., 1976). Or-pa sativa which is of special concern here, is not depressed at 10 ppm
at the adult stage (Howeler and Cadavid, 1976), whereas seedlings are quite sensitive
(Thawomwong and Van Die&, 1974).
Compared to these plants, Azolla in general thus seems moderately tolerant to Al.
It must be mentioned that growth inhibition observed in this experiment at high Al con-
centrations could be partly attributed to pH effects ; nevertheless it is well known that
Azolla is able to grow properly in a very wide range of pH (Nickell, 1961 ; Lumpkin
and Plucknett, 1982 ; Van Hove et al., 1983).
In view of the wide Azolla species and strains Aluminium sensitivity differences,
selection based on this characteristic has to be considered for Azolfa use on Al rich soils.
Table 1 : Productivity as compared to the control and standard deviation
(%I of Azolla species or varieties as influenced by the external
Al concentration (ppm Al).
Species
2 ppm A l
5 ppm A l
8 ppm A l
APP
90.51 + 11.81a
63.46 !' 16.60a
46.52 & 15.27a
ACA
80.42 + 13.52ab
45.07 i 16.16b
33.76 _f 14.19b
AM1
76.20 + 13.50b
3:.33 + ;J.Sib
23.01 :
/.6/bc
API
61.74 + 8.70~
33.06.+ 7.77b 21.70 _+ 5.09~
In a column, values whith the same lettre are not significantly
different at P = 0.05 (Duncan test)
129
\\
1OC
\\
\\
a
b
p IC
r
t
u
c a0
:v
1
t 40
J
a0
0
100
d
P 80
I-
:
u
c oo
t
I
I
i 40
t
J
a0
0’
I
t
t
0
a
4
0
8 0
a
4
a
I
PPm Al
PP= Al
129
Figure 1 : Productivity (g/m’. day) of 20 Azolla strains as influenced by the external Al concen-
tration @pm Al). Strains are characterised by their entry number in the ADUL colle&on
(a) ACA
:
6851
6 1
109
113.
(b) Ah4I
:
66:
104:
175.
(c) +PI
:
6,
26,
87.
92,
100,
137.
(d) APP
:
98,
99.
111.
129,
136.
144.
130
References
AWAD, AS., EDWARDS, D.G. and MILHAM, P.J. 1975 Effect of pH and phosphate
on soluble soil alumimium and on growth and composition of Kikuyu grass. Plant
Soil 45, 531-542.
FOY, C.D. and BROWN, J.C. 1964 Toxic factors in acid soils : II. Differential alumi-
nium tolerance of plant species. Soil Sci. Sot. Am. Proc. 28, 27-32.
FURLANI, P.R. and CLARK, R.B. 1981 Screening sorghum for aluminnium tolerance
in nutrient solutions. Agron. J. 73, 587-594.
HORST, W.J., WAGNER, A. and MARSCHNER, H. 1983 Effect of aluminium on root
growth, cell-division rate and mineral element contents in roots of Vigna unguiculatu
Genotypes. Pflanzenphysiol. 109, 95-103.
HOWELER, R.H. and CADAVID, L.F. 1976 Screening of rice cultivars for tolerance
to Al-toxicity in nutrient solutions as compared with a field screening method. Agron.
J. 68, 551-555.
LONG, F.L., LANGDALE, G.W. and MYHRE, D.L. 1973 Response of an Al-tolerant
and Al-sensitive genotype to lime, P, and K on three atlantic coast flatwoods soils.
Agron. J. 65, 30-34.
LUMPKIN, T.A. and PLUCKNE’IT, D.L. 1982 AzoIZa as a green manure : Use and ma-
nagement in crop production, Westview Press, Boulder, 230 pp.
MATSUMOTO, H., HIRASAWA, E., MORIMURA, S. and TAKAHASHI, E. 1976 Lo-
calization of aluminium in tea leaves. Plant and Cell Physiol. 17, 627-631.
NICKELL, L.G. 1961 Physiological studies with AzolIa under aseptic conditions. II. Nu-
tritional studies and the effects of the chemicals on growth. Phyton. 17, 49-54.
RANDALL, P.J. and VOSE, P.B. 1963 Effect of aluminum on uptake and translocation
of phosphorus 32 by perennial Ryegrass. Plant Physiol. 38, 403-409.
RIOS, M.A. and PEARSON, R.W. 1964 The effect of some chemical environmental
factors on cotton root behavior. Soil Sci. Sot. Am. Proc. 28, 232-235.
THAWORNWONG, N. and VAN DIEST, A. 1974 Influences of high acidity and alumi-
nium on the growth of lowland rice. Plant Soil 41, 141-159.
VAN HOVE, C., DIARA, H.F. and GODARD, P. 1983 Azolla en Afrique de 1’Ouest
- in West Africa. 56 pp. 11 illustrations. C. Van Hove, Ed., Louvain-la-Neuve, Belgium.
INOCULATION :
PRODUCTION D’INOCULUM
ET ESSAIS D’INOCULATION
.
. . (
I
. “ . ,
.
.
.
. : . .
Umimisa la FBA pour la Production Agriwle et Foresti& en Afiiqu~
.I.
:
Prbd&tioti x&i uses of rhizobium inoculants
BORDELEAU, L.
Research Sation, A@ult~ Canada,
2.560 Hochelaga Bhd. Sbinte-Fey,
Qdbec. Canada GIV 213.
Abstract
‘,
‘.
The inoculation of legumes with Rhizobiwn inoadants is one of the few cases
of man’s successful exploitation of microorganisms fqr avp action. This bio-
technology is outstanding in its wide applicability in diffmt agricultural envimn-
merits. We outline here background information on the various aspects of legume
inoculant technology, namely : strain selection and characcetistics, preparation and
properties of carriers, growth and survival of rhizobirg quality control program from
strain selection tbmugh preparation and distribution of final products. Criteria used
in the selection of Rhizobium strains for inoculation include prompt effective nodu-
lath over a wide range of field conditions and root tempaatures, ability to grow
wellinculture.togrowandiveincarrierandto‘~~onthe~Other
characters are competi-tiveness in nodule formation and survival and multiplication
inthefoilpH~l~pesticidetoleranceand~inthepesenceofcom-
bined nitrogen The sxxxss of legume inoculation~depauis on both the quality of
inoculantp-eparationQldthe~~~usedto~~~rhizobiaintosailatsowing.
lnoculant carriers must be selected accord+ to availability of material and t&no-
log&. and must permit the introduction of the highest possible number of infective
cells in the soil where the host plant will first accept an infection Finely ground
peat is the rhizobial carrier most extensively used. New technologies such as polymer-
entrapped microorganisms, and concentrated freezedried cells powder inoculants are
arriving on the market A quality control program is an essential part of inoculation
technology.
133
Introduction
Food and fibers production is often limited by low soil fertility and nitrogen
appears to be one of the most limiting factors. Industrially manufactured nitrogenous ferti-
lizeds are not afordable for most of developing countries that want to intensify their agri-
cultural production. Therefore, the alternative of using some microorganisms to convert
atmospheric nitrogen to a form available for the nutrition of higher plants was extensively
exploited
The nitrogen-fixing symbiosis between leguminous hosts and their rhizobia is an
outstanding biotechnology because of its wide applicability in diverse environments and
agricultural systmns. The inoculation of legumes with Rhiwbium inoculants is one of
the few cases of man’s successful exploitation of microorganisms for agricultural pur-
poses. ibis biological system has attracted considerable attention of scientists having
diverse backgrounds because of the complexity of the phenomenon which demands the
application of many skills for its understanding and more successful utilization. However,
there is a real risk when the enthusiastic non-microbiologist moves in to invest&ate a
system in which a bacterium, which may be more difficult to control, plays its own major
role in an intimate association with the more tractable host. Whether he is plant physiologist,
biochemist, geneticist, soil scientist, or even a microbiologist unfamiliar with rhizobium,
a worker coming into this field cannot afford to take this microorganism too lightly. Diffi-
culties can be encountered by the relatively slow development of even the faster-growing
rhiiobia, leaving them open to overgrowth by a contaminant, by the gummy nature of
its growth, increasing the likelihood of an associated organism remaining undetected even
with single colony isolation. Furthermore, the difficulty one has to definitively distinguish
rhizobia from some other genera, apart from demonstration of nodulating capacity, adds
up to the confusion due to contamination or wrong identification and may lead to a whole
body of developmental work which is bias to the practice of legume inoculation. In such
a case, either the newcomer takes the trouble to fanGnize himself sufficiently with the
nature of the microorganism he is dealing with, and the technique needed for its purifi-
cation and maintenance, or he works cooperatively with someone who has this knowledge
and technical capacity.
In this presentation, we will outline background information on the various aspects
of legume inoculant technology, including factors affecting quality and technique for
quality control
Selection of rhkobk for inoculants
Many of the failures to establish rhizobia in soil successfully may be attriiuted
either to insufficient or unsuitable rhizobia in the inoculant. As agriculture extends further
and further into regions which are marginal because of low and/or erratic rainfall and
poor fertility, rhizobia will be called upon to perform with a wider range of host plants
and environmental and edaphic conditions. The selection of suitable strains of rhiiobia
is the first step in the produciton of legumes inoculants. At one time, nitrogen fvtaton
efficiency was the only criterion seriously considered for a good inoculant strain, but
now there are many that should be considered as well : (i) competitive ability with other
strains, including indigenous population, for prompt effective nodulation of the host legumes ;
(ii) nitrogen fixing ability over a wide range of environmental conditions ; (iii) nodule-
134
forming and nitrogen-fixing abilities in the presence of soil nitrogen ; (iv) ability to multiply
in broth and to survive in carriers ; (v) ability to survive on the seed ; (vi) ability to
survive in adverse physical and physiological conditions such as desiccation, heat, free-
zing, low soil fertily ; (vii) tolerance to pH changes, pesticides ; (viii) genetic stability
during-w~growth;
In the context of inoculant production, strain competitiveness and specifkity refer
to effectiveness in forming rapidly an effective association with a particular host (Antoun
et al., 1979). The host legume exerts a strong influence on the effectiveness of the sym-
biotic association in fixing N,, and strain specificity exists for particular hosts (Gueye
and 3ordeleau, 1988). while intemctions between the symbiotic association and the environ-
ment have been also described @rockwell, 1980 ; McLoughIin et al., 1984). l&se create
a dilemma in deciding wether to produce numerous inoculants with highly effective
strainsforindividualhostorU,usea~~~strainsthatvaryfromgoodto
excellent with a range of legumes. The use of multiple strain iuoculants should be avoided
due to possible antagonistic and competitive effects witbin the culture and the likelihood
of competition in nodule formation from less effective strains in the inoculant (Bordeleau
and Antoun, 1977). This problem poses a delicate situation for microbiologists who must
decide between practical aspects of inoculant production and marketing, and scientific
excellence in providing the best available mater2 for inoculant production.
As a rule, foreach legume that we want to improve yields through nitrogen fixa-
tion, it is preferable to select strains of rhixobia directly from the cultivar we intend to
use and then check the efficiency. The main criterion used in selection is the amount
of N, fixed, either assayed directly, or indirectly by determining plant dry weight @or-
deleau et al., 1977). The latter method integrates all the factors that influence the symbiotic
system. With grain legumes, pediodical m easurements of biomass and nitrogen content
of plant parts during the development and at maturity will pezmit to calculate t&harvest
index of a strain, since it is hown that plant inoculation with Rhiwbim can modify
the nitrogen distribution in the plant. Tests in controlled environment provide reliable
comparative information on the ability of a number of strains to fix N,, but they do
not evaluate ability to colonize the rhizosphere, to compete for nodule sites, or to persist
in the soil. In the other hand, field ~CS~.S cm CV~USZ only a Llii-ALcd liu~~~k~ of strains
because of demands on time, labor and equipment. Both systems, therefore, are commonly
used in sequence, the former to select a set of strains with above average N,-fixing ca-
pacity and the latter for an overall evaluation of the strains’ ability to fm N2 under a
variety of field situations. Specific separate tests can be conducted if a particular cha-
racter is bemg evaluated. However, field trials provide the ultimate test of a strain’s perfor-
mance, since N2 fixation and the proportion of nodules formed by the inoculum are the
end result to the interactions of all the factors involved.
Large-scale production of rhizobla
The first step in the production of legume inoculant is massive production of a
selected Rhiwbium species in liquid medium. Rhizobia are heterotrophic and generally
not highly demanding in their nutrient requirements.
Different media have been proposed, they are composed of a carbon source, a
source of nitrogen and growth factors, and different minerals such as potassium phos-
phate, magnesium sulfate, and sodium chloride. Mannitol and sucrose are the conventional
135
carbon sources used in routine media, although growth of slow growers on mtitol is
variable. The most satisfactory carbohydrate is glycerol, on which slow growers have
shorter generation tunes than on glucose, mannitol, galactose, or sucrose (Stowens, 1985).
Yeast extract can be used as the sole carbon and nitrogen source ; however, high concen-
trations cause cell d&or&m and even growth inhibition of certain Rhizobium; which can
be co~tmcted by calcium. The economy of largescale production is largely governed
by the price and availability of a suitable carbon source. Corn-steep liquor, proteolysed
pea husks, malt sprouts, whey and industrial-grade yeast extract have been pmposed as
media for Rhiwbium production at the industrial level (Meade et al., 1985 ; B&OM&&
et al., 1986). All of these materials ate industrial by-products which contain growth factors,
nitrogen and carbon. Efficient uulization of a carbon source often depends on aeration,
method of ster%zation and equipment available. For the production of inoculants, a high-
count broth is required to ensm a highquality inoculant ; increased numbers of viable
cells may be achieved by varying the source and concentration of materials acconhng
to the individual strain requirements. A- short fermentation time for the broth culture is
desirable in order to reduce contamination risks. A large inocuhun, providing 106-10’
rhizobia/ml culture medium at the beginning of fermentation, reduces the time required
to reach maximum viable numbers, and, therefore, reduces the chance and effects of
contamination. Fast-growing strains reach maximum viable number iu 30-40 hours, com-
pared with 80-120 hours for slow-growing strains. Temperature of incubation for rhii
bia is generally kept at 2628”C, although few strains can be produced at 32-35°C
without affecting their nod&ion ability.
Facilities for multiplication of rhizobia in liquid culture vary from elaborate
large-capacity (10002000 liters) industrial fermentors. to simple flasks or drums
(10-100 liters) (Date and Roughley, 1977). Good fermenters can be fabricated from
steel drums fitted with brass inlet, outlet, inoculation and sampling ports Tbe air inlet
extends to the bottom to provide stirring of the broth. Proper filtration systems st&lize
the air from an external supply. The vessel with filters and outlet tubes attached and
two-thirds filled with medium is sterilized by autoclaving. Maximum infective rhizobia
after incubation should be in the order of 109-KY0 cells/ml.
Carriers of rhizobia and their preparation
High-quality legume inoculants is &pendent on selection of suitable carrier ma-
terials and their pre-treatment (Kremer and Peterson. 1983). An assessment of the literature
on carriers of rhizobia leaves the net impression that peat is still unchallenged as a carrier
and that it is relatively easy to devise a substrate from a variety of matea%& that would
support satisfactory growth and survival of rhizobia (Burton, 1976 ; Brockwell, 1980).
As a rule, carriers must have a high water-holding capacity, provide a nutritive medium
for growth of rhizobii and enhance survival during storage, distribution and when ino-
culated onto seed Various carriers of organic origin have been investigated These include
: soil and peat mixtures, soil plus charcoal, Nile silt plus nutrients, soil plus coir dust,
coal with and without nutrients, composted maize cobs, bagasse, filter mud from sugar
cane, cellulose powder, rice husk, bentonite, ground common talc.
The most important factors to consider before selecting a particular carrier are
the availability and characteristic of the material. Whereas the choice may be aided by
a chemical analysis, actual multiplication and survival studies of Rhizobium strains in
136
it are essential. Pinely ground peat is the rhizobial carrier used most extensively in conven-
tional leguine inoculant pxodllctial. Afm pmper neutmnsation with taco,, and milling
to as fine a particle size as possiile, broth culture of rhizobial strain containing at least
109rhizobidmlaremixedwithdryunsterilizcdpeatinbulkandthenmaauedinshallow
trays for 7296 hours at temperatures near WC. Inoculants pmpared with unsterilized
peat usualIy contain up to lOO-fold fewer rhizobia than skriked peat Packaging is done
in polytbene bags, moisture content cormsponds to 3540 % on a wet weight basis. Rhi-
zobia g,row and survive better in sterile than in tuwfxile peat. The extent of this dif-
ference is depeknt on the suitability of the peat and the temperature and duration of
storage. ‘Ihe choice of a method for steriWng peat depends on the type of container
inwhichtheculMeistobe~~,thenumberofculturestobeprepared,andthe
availability of steriking facilities, Autoclaving, ethylene oxide treatment, and gamma irra-
diation are used. Stklized peat in sealed polythene bags is inoculated by injecting enough
broth culture to raise the moisture content to 50-60 %. The puncture is sealed with
adhesive tape, the peat and broth are blended by manipulation and then incubated at
26°C for 2 weeks followed by storage at 4°C. High quality inoculants containing up to
lOlo rhizobia/g can be prepared by this method (Roughley and P&ford, 1982).
Alternative can&s to prepare Rhizobium inoculant are polymer gels ma& of po-
lyacrylamide, alginate, or xauthan. They were successfully tested 10 years ago by Dom-
mergues et al. (1979). Alginatebased polymer-entrapped Rhizobizun inoculant contain 109
to 10” ceils/g dry material of high survival ability Rhizobium (Diem et al., 1988).
However, this technology requires specialized equipment and a certain level of expertise
that are not yet adapted for use in developing countries.
Improved technologies for dried microbial cells preparation through free;tedrying
processus, in which various protectants are added, were recently applied for Rhizobium
inoculauts. Thus;high number of rhizobia can be added to seed because of the very high
count of the inoculant so prepared. Rhizobium can sukve for a very long time in these
concentrated inocuWs (E. Brochu, Rosell Institute, Montreal, Canada, pers. comm.).
Survival of rhkobla
Survival ability of rhizobia throughout the manufacturation and uses of inoculants
am very important, since the inoculum must contain very high number of rhizobia if
success of nodulation is to occur in field situation. Survival of Rhiwbium has been
studied in culture media (Bordekau and McLoughling, 1982), in carriers &renter and
Peterson, 1983), in the seed environment (Burton, 1986) and in soils (Bushby and Mar-
chall, 1977). Cultural factors and conditions may affect survival ability ; it is generally
recagnized that yotmgex cells of Rhiwbiwn survive better when subjected to stress con-
dition because they have more endogenous energy available to them to overcome the
adverse conditions (Bordeleau and McLaughlin. 1982). In liquid medium, survival
abiity of fast growing Rhizobium is not significantly affected by limited amounts of
carbon, phosphorus and potassium in growing medium. However, under desiccation, sur-
vival is poor unless cells are protected by agents such as lactose. It appears that cell
leakages, and therefore cell damages, are limited by lactose and the exact nature of this
protection is unknown. Commercial production of fast-growing rhizobia on lactose rich
by-product such as whey is now becoming a practice (Eissonnette et al., 1986). not
only because this medium permits the production of high concentrated broth (lOlo rhi-
137
zobia/ml), but also acts as a good protectant and contributes to high survival. Strains
of tibia varied in their susceptibility to desiccation, and they have to be selected ~CCOT-
ding1 y.
Survival of organisms plays an important part in microbial ecology, especially
when the environmental stress may adversely affect the cell. These environmental
stresses are generally the lack of nutrients (energy) in the environment. la& of moisture,
presence of other organisms, changes in pH, Presence of chemicals, etc. When applied
to the rhizobii, the ability of legumes to be properly infected with desirable strain depeuds
on the rhizobia to survive in the carrier, on the seed, and in the soil under adverse conditions
until it is ready to infect the host. With the current practice of. inoaihnmg the seeds,
ability of the rhizobia to survive throughout the process is of the first importance. As
mentioned earlier, carriers can be manipulated to sustain high number of cells. However,
exposure of inoculants to hot and dry conditions during shipping, storage, and planting
often results in decreased numbers of N,-fixing effectiveness of the rbizobia. Variable
inoculant quality is a common problem in the tropics and subtropics where imxulants
are often subjected to unpredictable handling and storage during distribution and use.
Therefore, inoculant formulation and packaging cannot be overlooked from a practical
point of view.
The survival of rhizobia on seed is of the greatest impormnce since seed pelleting
as a method of legume inoculation has received widespread attention and it is now a
standard agricultural practice. Rhizobia die quickly on freshly inoculated seed particu-
larly with broth inocula (Burton, 1976 ; Brockwell, 1980). Therefore, protective agent
assuring survival of rhizobia and a durable coating the seeds without affecting their genni-
nation must be used Lost of humidity is an important factor for the death of rhizobia
on the seeds (Salema et al., 1982) and, on the other hand, the coating has to be dried
enough to stay stable and to equilibrate with the moisture level of the seed. Other l&or-s
must be taken into consideration such as toxicity of the coating, pH and nutrient value.
Pelleting of leguminous seeds with organic based inocula and pulverized limestone and
nutritive sticking agents such as lactose, gum arabic and cheesewhey provides both a
good survival of rhizobia and uniformity of coating. Pelleted seeds are often treated with
fungicides to protect them against fungal diseases. Problems arise when legumes ino-
culants are used in conjunction with pesticide treatments (Tu, 1981). Such treatments may
affect the survival and/or the infectivity of Rhizobium imculants, unless resistant strahs
and compatible doses of pesticides are used (Odeyemi and Alexander, 1977).
Quality control program
The quality of legume inocula depends on both the number of rhizobia they
contain and their effectiveness in fling nitrogen with the intended host. Quality control
tests measure the number (quantitative) and identify the type (qualitative) of rhizobia in
the incculant during its recommended useful life (Roughley and Pulsford, 1982). Quality,
therefore, is determined by the various strain attributes mentioned earlier. The standards
of quality adopted need to reflect both the number of rhizobia required for successful
nodulation and the ability of manufacturers to produce such inoculants. The principal
reason for quality control is consumer protection, a function normally performed by go-
vernment agencies. However, such control applies only to the fmal products. Quality control
during preparation and before sale is usually the responsibility of manufacturers.
138
It can be safely stated that the highex the number of infective rhizobia contained
in the inodant, provided the strain is efficient, the better the inoculant. All investiga-
tors recogn& the importance of inoculant having at least a lo-fold more rhizobia when
produced than required at the time of usage to compensate for die-off. Ideally the standard
for an inocuh@ is based on the mm&r of infective rhizobia/seed necessary to initiate
prompt and effective nod&Con. This number will vary considerably from the favcrable
conditions of aseptic culture in greenhouse to the highly competitive situation in the field.
In practice, standards are strongly infhrenced by the level of rhizobii that manufacturers
can consistently attain and maintain in their inoculants with available technology -Many
investigators recommend an inoculum which would provide approximately l(Y to 1V rhi-
zobia per seed for a good nodulation. Even at this level of inoculation and where indi-
genous soil rhixobial population might be as high as lWcells/g soil, the introduced
Rhiwbium is so heavily outnumM that it needs special advantages in order to become
established. In Cauada, quality standards of inoculants were established at levels that
provide at least103 infective rh&bia/seed for small seeded legumes, and up to 10s infec-
tive rhizobia/seed for larger seeds such as soybean and peanuts (Bordeleau and
Prevost, 1981).
Basically any program should involve both qualitative aud quantitative tests of
the broth culture before impregnation of the carrier, fmal carrier culture immediately after
manuf&tu&on and before release, and inoculants obtained from the retail outlets. Tests
on broths and carriers before release am essential for all batches of culture. Broth cultures
should be examined for authenticity by appropriate immunological tests to ensure the
identity of the inoculaut stmin, and for freedom from contamination by absence of growth
on glucose-peptone agard and gram-stained smear free of positive cells. Results of these
tests~aMilablewithin24hours.Thestrainintbeb~thshouldalsobecheckedfor
infectiv’~ess and effectivemss by inoculating plants growing on N-free medium in an
aseptic environment.
After maturation of the inoculant in the carrier and before distribution, plate counts
on sterile carriers should be made to determine the number of rhiibii present and to
ensure that contamination is low. Where unsterile peats are used, plate counts am not
feasible because no specific media exist to differentiate rhizobial cells, plant-infection
tests should be made (Bordeleau and Prevost, 1981). Although, labor tune, space and
material consuming, the plant-infection test provides a useful tool to assess identity and
number of infective rhizobia in non-sterile carriers. Only after completion of these tests
should the inocuhnt be released for use and an expiry date assigned. The length of the
useful life of the inoculant will depend on the chamcteristics of the packaging and carrier
materials, storage conditions, methods of distribution and sale. Therefore, it is impossible
to set general expiry date.
The quality control program should include testing of inoculants sampled from
retails outlets. Such a program was established in Canada and it contributed greatly to
improve the quality of legume inoculant products offered to the farmers. This program
provided information on the effect of storage conditions, on the performance of particular
inoculant and a check on tire labelling and claims by manufacturers. It also provided
maximum consumer protection aud more direct assistance to manufacturers by early de-
tection of problems such as poor survival in carrier culture and on pm-inoculated seeds,
or t%hxre to produce nodulation on the recommended host-plant.
139
Field applications of legume inoculants
The act of inoculation concentrates the inoculum close to the position where tbe
host plant will first accept an infection whereas competing soil organisms, sometimes
much more numerous, are distributed randomly. A major advantage enjoyed by .inocu-
lant rhizobii is proximity to a developing legume rhizosphe.re where it can multiply.
Moreover, the time between sowing and ger&nation is a critical period. If the inoculant
mortality rate is high due to a soil environm ent that is physically, chemically, or microbio-
logically unfavourable, delayed gemGnation can lead to a situation where the &z&ii
popdation is too low to colonize the rhizosphere. The likelihood of nod&ion can be
increased by reducing the mortality rate of the inoculant, by raising the rate of inoculation,
or by decreasing the time to seedling germination. Increased oi; massive rates of inoculation,
may lead to the establishment of the inoculant in the soil micro-habitat by weight of
numbers. Alternatively, seed pelkting procedures protect the rhizobia against deleterious
environmental factors. We will &sent here a few methods in usage for inoculant appli-
cation in the field, with a relative apprakal of each method.
Seed Inoculation.
Dusting. At one time some manufacturers recommended that ino-
culant be merely mixed with the seed immediately before sowing or sprinkled on to the
seed in the seed box. Some culture adheres to the seed by lodging in the micmpyle and
in scratches and irregularities on the seed coat, and by electrostatic attraction. However,
much of it is shed particularly during passage of the seed through machinery. Dusting
is certainly the simplest method of inoculation but it is also the least effkient with conven-
tional products, and unless break-through in this method proves a better efficiency. by
;.
the
use of new products, it is not a recOmmended method.
Slumy inoculation. In order to increase the amount of inoculant
adbering to the seed, the culture is applied as a water suspension ; alternatively the ino-
culant is mixed with moistened seed. The seed must .be dried before sowing (not in direct
sunlight), but as it dries a proportion of the inoculant is lost. More inoculant can be
attached to seed by using an adhesive in the slurry ; e.g. a 10 96 solution of household
sugar (but sugar lacks tenacity). Adhesives such as gum arabic and substituted cellulose
compounds are more tenacious. Caution must be exercised to avoid samples of gum arabic
which contain preservatives k&al to rhizobia. The celluloses tend to be rathe~ cheaper
and more readiIy available than gum ambic.
Seed pelkting. Coating legume seeds with lime to promote no-
dulation apparently had its origin in Maine, U.S.A., more than 40 years ago. The development
of the process as a technique for establishing pasture species on acid soils was begun
in Australia in the 1950”s (Roughley and Pulsford, 1982). Its use is now widespread
extending beyond more protection against soil acidity. Advantages include protection of
the inoculant rhizobii against rhizobiotoxic substances contained in some legume seed
coats, unfavourable physical and chemical conditions in soil, competition from the soil
microflora, the effects of acid fertilii, and against seed-harvesting ants. Pelleting makes
aerial sowing of inoculated seed a practical proposition and ensures better survival of
the rhizobia when delays between inoculation and sowing are unavoidable.
140
Materials used for pellet coatings include calcium carbonate in many forms, dolo-
mite, various grades of gypsum, bentonite and other clay minerals, rock phosphate and
other phosphorus compounds, inert materials such as titanium dioxide and talc, soil and
humus, and activated &u-coal. The main requirements for a good coating mate&l are
tbatitsbouldbe~v~yc~toneutralityiapHand~lygrwnd(9o%atleast
passing 300 mesh). Adhesives used for attaching the coating material to seed include
synthetic glues, glues of vegetable or animal origin, gelatin, various sugars, and honey.
The adhesive should have sufficient tenacity so that the coating material does not slough
off, but should not be so tenacious that the cotyledons are damaged during geimina-
tion.. Adhesives must be free of preservatives.
Seed bed inoculation
There are several situations in which seed application of rhizobia may be an inef-
ficient means of inoculation : (i) preemergence diseases- or insect attack may make it
necessary to use seed dressings of fungicides or insecticides all of which have some effect
on the viability of rhizobii and some of which are very toxic indeed (Odeyemi and Alexan-
der, 1977 ; Tu, 1981) ; (ii) inoculation for broad-acre sowings of crop legumes with
high seeding rates is a major task, limiting the speed of the sowing operation ; (iii) some
seeds are too fragile to be inoculated, e.g. shelled peanuts (huch hypoguea L.) ; (iv)
some legumes such as soybean (GZycti mur (L.) Merr.) and subterranean clover (nijb-
lium subterranewn~
lift the seed coat out of the soil during emergence so that the rhizobia
on the seed coat are not deposited in the soil ; (v) the dimensions of the seed surface
place a test&ion upon the number of rhizobia which may be applied ; sometimes, .espe-
dally when small-seeded IegMles are used or where naturally occurring competitive rbi-
zobia are present, it is necessq that rates of inoculant application should exceed this
Limit ; (vi) the seed coats of some~species contain substances toxic to rhizobia. In these
circumstances, almtive means of inoculation are needed.
Solide inoculant - Solid inoculant is made by coating solid, granu-
lated material with peat inoculant in an adhesive. Suitable adhesives include a 25%
aqueous solution of gum arabic (no preservatives) or a 5% solution of methyl cellulose.
Tenacity of the adhesive solution can be improved by chilling overnight Peat inocu-
lant is thoroughly stirred into the adhesive and this suspension poured on to the beads
and mixed together until all beads appear evenly coated. The beads should be dried by
~ginathinlayer.Whendryandanylumpshavebeenbrokenup,thematerial
is ready for use. other inert particulate, freely flowing material such as marble chips,
coarsegraded sand, polyethylene beads, clay gram& can be used for preparing solid
inoculant. Availability and economic factors are the main points to consider in choosing
material. For large-scale application, solid inoculant can be applied through the fertilizer
box or insecticide attzzhment of a seed drill.
Liquid (spray) inoculation - Excellent nodulation can be obtained
by spraying inoculum into the row beside or beneath the seed. A peat culture of rhizobia
or fi-ozen, concentrated broth cultures, is mixed into a paste with water, diluted to a slurry,
then added to a water-filled tank prior to spray application. Any spray equipment is satis-
factory provided it has not been used previously for toxic chemicals. This method provides
very good results because high number of rhizobia can bc applied. However, marketing
141
problems exist especially with frozen concentrated cultures which deteriorate during
handling if proper conditions are not given to avoid thawing of the preparation
Pre-inoculation before sale
pre-inocnlationistfieinoculationofseedpriartosaleor~utionforfarmers;
it implies a ,peaiod of storage of theinoculamd seed before it is sown. The main objec-
tiveis~allowseedmetchants~buildupas~~~ofinoculatedseedino~to
meet predictable heavy demand by farmers at time when sowing deadlines are imposed
by climate. Bacteriologically, the objectives should be to use culuues containing very
high populations of rhizobia and to preserve their viability so that, at sowing, the level
of inoculum on pre-inoculated seed is equal to that on freshly inoculated seed. New tech-
nology exists to produce pelleted, pre-inoculated seed with excellent mechanical proper-
ties and large population of viable rhizobia. Pm-inoculamd seeds should meet the same
standard in the quautitative terms of infective rhizobii by seed as that provided by ino-
culation at sowing time. Seed pelleting represents the major aud probably the only real
advance in the technology of inoculant application, and appears to offer the best prospects
for the development of a reliable inoculatior~
Concluding remarks
The introduction of a legume seed inoculant strain into soil aud its establishment as a
permanent member of the soil community is an exercice in applied microbii ecology.
Quantitatively, the number of rhizobii associated with iuoculamd seed mt only a
way small fraction of the total soil microflora. The choice of a strain, the quality of
the inoculant and its method of application are impor%+ to insure that the mtmduced
Rhiwbium strain survives and establishes an effective symbiosis with its designatkd ho%
Unless proper methods of inoculaut manufacturation and application are used, inoculant
will fail to meet the objectives of its practice, regardless of the strain. Quality control
programm is an essential part of inoculation technology.
References
ANTOUN, R, BORDELBAU, L.M. and LACHANCE, R.A. 1979 Rendement de la hrzerne
(cultivar Saranac) inocul6e avec une souche t&s effkace de Rhiwbizun meldoti en
presence d’autres eqhes de Rhiwbium. Can. J. Plant Sci. 59, 521-523.
BISSONNETIE, N., LALANDE, R and BORDELBAU, L.U 1986 Large-scale produc-
tion of Rhiwbium meliloti on whey. Appl. Environ. Microbial. 52, 838-841.
BORDELEAU, L.M. and ANTOUN, H. 1977 Effets de l’inoculation mixte avec des
souches de R. meliloti sur le rendement de la luzeme. Can. J. Plant Sci. 57,
1071-1075.
BORDELEAU, L.M. aud McLOUGHLIN,‘TJ. 1982 Survival studies with Rhizobium
species. Can. Sot. Microbial. Ann. Meeting Proc. EM-I.
142
BORDEIXAU, L.M. and PREVOST, D. 1981 Quality control program by federal go-
vernment agencies and method of testing legume inoculant and preinoculated seed
pmducts in Canada. In : Clark, KW. and Stephens, J.H.G. (eds.), Pmuxdings of
the 8th North Amexican Rhiwbium Confea~~ce, Winnipeg, Canada, pp. 566-579. The
university of - F+iirhg stices, wilmipeg.
BORDELEAU, LM. ANTOUN, II. and LACHANCE, RA. 1977 Effets &s souches
de Rhiwbium melihi et des coupes successives de la hrzerne (Medicago sclrivp) sur
la fvration symbiotique d’azote. Can J. Plant Sci. 57, 433-439.
BROCKWELL. J. 1980 Experiments with crop and pasture legumes - Principles and Prac-
tice. In : Bergersen, FJ. (ed). Methods for evaluating biological nitrogen fixation,
pp. 417488. John Wiley and Sons, N.Y.
BURTON, J.C. 1976 Methods of inoculating seeds and the effect on survival of Rhizobia.
In : Nutman P.S. (ed). Symbiotic nitrogen fixation in plants, IBP 7, pp. 175190.
Cambridge University Press, Cambridge.
BUSHBY, lXV.A and MARCHALL,, KC. 1977 Some factors affecting the survival of
root-nodule bacteria on desiccation. Soil Biol. Biochem. 9, 144-147.
DATE, R.A. and ROUGHLEY, RJ. 1977 Preparation of legume seed inoculants.
In : Hardy, R.WF and Gibson, AH. (e&z.). A tmatise on dinitrogen fucaton, section
IV, pp. 243-275. John Wiley and Sons, N.Y.
DIEM, H.G., BEN KHALIFA, K., NEYRA, M. and DOMMERGUES, Y.R. 1988 Recent
advances in the inoculant technology with special emphasii on plant symbiotic mi-
croorganisms. Workshop on advanced technologies for increased agricultuml produc-
tion, Santa Margherita Ligure, Italia, 25-29 Sept. 1988.
DOMMEKC;tiES, Y.K., DIEM, H.G. and DIVLES, C. 1979 Polyacrylamide-entrapped
Rhizobium as an inoculant for legumes. Appl. Environ. Microbial. 37, 779-781.
GUEYE, M. and BORDELEAU, L.M. 1988 Nitrogen fixation in bambara groundnut, Voand-
z&a subterranea (L.) l%ouars. Mircen J. 4, 365-375.
KREMER, RJ. and PETERSON, ILL. 1983 Effects of carrier and temperature on sur-
vival of Rhiwbium spp. in legume inocula : Development of an improved type of
iddant. Appl. Environ. Microbial. 45, 1790-1794.
McLOUGHLIN, TJ., BORDELEAU, L.M. and DUNICAN, LX. 1984 Competition stu-
dies with Rhizobium mfolii in a field experiment. J. Appl. Bacterial. 56, 131-135.
WE, J., HIGGINS, P. and GGARA, F. 1985 Production and storage of Rhizobium
leguminosarum cell concentrates for use as inoculants. J. Appl. Bacterial. 58,
5 17-524.
‘_.
.:
.’
1,
‘!...
..--.-
:
143
ODEYEMI, 0. and ALEXANDER, M. 1977 Resistance of Rhiwbium strains to Phy-
gon, Spexgon, and Thiram. Appl. Environ. Microbial. 33, 784-790.
ROUGHLEY, RJ. and PULSFORD, DJ. 1982 Production and control of legume
ino-
culants. In : Vincent, J&i. (ed.), Nitrogen Fixation in Legumes, pp. 193-209. Act+
demic Press, Sydney.
SALEMA, M.P., PARKER, CA. and KEDBY, DJL 1982 Death of Rhizobia on inocu-
lated seeds. Soil Biol. Biochem. 14, 13-14.
STOWENS, M.D. 1985 Carbon metabolism in Rhiwbium species. Ann. Rev. Microbial.
39, 89-108.
. _
TU, CM. 1981 Effect of fungicidal seed treatments on alfalfa growth and nodulation.
Chemosphere 10, 127-134.
Mmimisez la FBA pour la Production Agricole et Forestière en Afrique
Caractérisation et choix de supports
pour inoculum
BEUNARD, P. et &UN!I’ MAURY, H.
IRAT~CIRALI - BP. 5035. 34032 MONTPELLIER cedex France.
L’étude réalis& avait pour but de tenter de relier quelques unes des carackk-
tiques Dhvsicn-chimiques de SU~JXX?.Q possihlw d’innmlllm WCC leur cayGt6 r;rllr
à permettre la survie des rhimbiums.
Les modalitis de conditionnement des inocuhuns ont tgakment fait l’objet de mises
au point visant B l’obtention de produits de qualité Llev&
Les supports utilisks proviennent de pays où des unit& de production d’inoculum
pour l@mineuses ont &5 installh dans le cadre de programme de coop&tion
technique de la FAO : Madagascar, Tunisie, Zaïre, Bhutan, Nicaragua, Turquie., France.
Les rhultats obtenus ont permis de montrer : (i) L’importance du pH initial du
mpport, l’optimum se situant entre 6 et 6.5. Lorsque œ n’est pas le cas, une neutra-
lisation peut êh~ effecde avant l’utilisation ; (ii) les conditions & sttkilisation des
apports ont des cons&wnces primordiales sur la survie des rhk9knns ; (iii) I’exis-
tente dans la majoritk des cas d’une Ikg&re multiplication des rhizobiums au cours
des premiks semaines suivant l’ensachage puis une dkroissanœ
plus ou moins
rapide selon les conditions de conservation (te.mp&at.ure de stockage).
Ces r&ultats ont conduit & des recommandations sur les mkhodes de production
des irmxulums dans des unités de petite taille. Leur mise en œuvre fait tgalement
robjef d’une jxckerltation.
145
Introduction
Le nombre de rhizobiums contenus dans un inoculum conditionne en grande partie
son efficacitk. Lors de la production 2 l’tkhelle pilote, pour les besoins de l’expkimen-
tation en milieu paysan, la qualiti du produit utilis5 joue un rôle primordial et doit être
aussi élevée que possible. La multiplication des rhizobiums dans des petits fer-menteurs
simplifiés mis au point par I’IRAT et baptises UPIL (unité de Production d’Inoculums
pour Légumineuses) est maintenant bien maitriske et adoptée dans plusieurs pays (Saint
Macary et al, 1986). L’étape suivante de la production, dans laquelle la culture liquide
est mélangée a un support nkessite la recherche, au niveau local, de composés permettant
une bonne survie des rhizobiums. Dans le cadre de projets de coopération technique de
la FAO, de tels supports ont été recherches et leur aptitude & être utilises dans les unités
a et& évaluée.
Matbriel et m&hodes
Les principales caractéristiques physiques et chimiques des supports utilises sont
prksentks dans le Tableau 1.
L’ensemble des supports a &.k séché a l’air puis à l’étude a 50°C pour obtenir
une humidité finale de 12 à 15%. Ils ont ensuite été broyés. passes au tamis de 0,5 mm
puis conditionnés dans des sachets en plastique autoclavables (PolytWrylène moyenne den-
site ; épaisseur 150 p.) a raison de 80 g par sachet. L’ensemble des sachets contenant
les supports a été stérilise par trois passages B l’autoclave (120°C, 20 mn), les autoclaves
étant espacds de 24 h
Les cultures de rhizobium utilisées pour injection dans les sachets ont ktk rkalisks
en fioles d’erlenmeyers, agitées & 30°C contenant du milieu yeast extract mannitol
(YEhî), (Vincent, 1970). Les inoculums ont été préparés en injectant en conditions skiles
dans les sachets une quantité de culture liquide correspondant ?t une valeur situke entre
40% et 60% de la capacité maximale de rétention en eau du support (Date et Roughley,
1977).
Api2 JiiLeLdib L~I~~S, CL, szion ics ui~~kmmth ;ippqucs ~~LCIIII~~UUII
uu non,
neutralisation ou non, température de stokage), le nombre de rhizobiums contenus dans
les sachets a été déterminé selon les méthodes de suspension dilution puis étalement en
boite de p&ri (FAO, 1984).
RBsultats et Discussions
Neutralitb des supports
Les résultats obtenus sur la tourbe de Betafo (Madagascar), représentés dans la
Fig. 1 montrent clairement que la neutralisation par du carbonate de calcium d’un support
tri?s acide peut considerablement améliorer sa capacité a maintenir une population micro-
bienne en vie. La tourbe neutralis&, actuellement utilisée pour la fabrication d’inoculums
permet la survie de Brudyrhizobium japonicum pendant pres de deux mois, ce qui permet
d’envisager un délai entre la fabrication des produits et leur utilisation.
Tableau 1 : Caracteristiques chimiques des supports utilises
Lieu de
Identification
Conductivi-
Pays d'origine
Nature
prélèvement
sur Fig. 4
pH (eau)
Carbone (%)
C/N
té è 25'C
(micromhos)
Madagascar
Tourbe
Betafo
3,78
15,86
16
1180
Turquie
Tourbe
Bol~
5,60
41,06
23
1650
Trabzon
4,60
24,65
2s
600
Kayseri
7,60
9,90
17
1700
B h u t a n
Tourbe
Par0
4,60
21,80
13
-
Nicaragua
Tourbe
Las Playitas
6,35
14,70
64
176
6,68
6,38
36
106
6,90
7,56
34
116
San Tarlos
5,90
19,02
20
104
Balle de riz
5,14
27,70
69
397
Ecume de canne
5,92
37,62
51
2750
Bagasse de canne
5,54
14,74
79
350
Bagacillo de canne
6,18
15,60
80
335
Pulpe de caf@
7,37
46,94
42
2330
0
0
147
i
.
148
SMlIté des supports
Le traitement par la chaleur humide de tourbes de Turquie (Fig. 2) amkliore la
conservation des rhizobiums. Il est naturel, en effet, qu’en l’absence d’autres microorga-
nismes, les rhizobiums ne rencontrent aucune comp&ition pour les sources d’éikments
nutritifs contenus dans le support. L’hypothèse souvent exprimk d’un risque de dégra-
dation du support pendant la stkilisation et d’une libération de composb toxiques pour
les rhizobiums ne se trouve cependant v&if%e pour aucun des deux supports comparés
dans cette &ude. Entre ces deux supports cependant, des difftkences notables existent
et la tourbe extraite de Bolu fait aujourd’hui l’objet d’une utilisation par l’industriel de
fabrication d’inoculum turque.
Tempbature de conservation des supports
L’étude r&lis& sur une tourbe de Paro (Bhutan) comparant l’effet du stockage
d’inocnlnms à 4’C et 25OC a permis de montrer (Fig. 3) qu’il n’existait pas, pour la
p&iode consid&& de diffkrence importante entre les traitements. D’autres r6sultat.s (non
présentés ici) ont montré que la conservation des inoculums & tempkrature ambiante ou
& 30°C pouvait rkwlter. dans un premier temps en une multiplication des rhizobiums
et l’obtention de meilleurs inoculums que ceux conservés dès leur conditionnement à
basse tempkature.
Relations pH, matibre organique, survie
La survie de la souche de Bradyrhizobium japonicum IRAT FA3, apr6s 50 jours
de conservation dans différents supports est repr&ent& dans la Fig. 4, où sont égale-
ment indiquées deux caractkristiques chimiques des supports : le pH et la teneur en
carbone. On constate qu’il n’existe pas de règle gtkkale, au moins pour les deux facteurs
pH et C, permettant de pr6dire la valeur d’un support, Les meilleurs rkwltats ont Cd
obtenus avec des supports pmches de la neutralité (pH = 5 à 7) et pour des teneurs
en C moyennement élevks. Ces conditions peuvent semb1e.r nécessaires mais ne sont
pas suffites puisque certains supports, comme la bagasse de canne (2) ou la tourbe
de Las Playitas (C) ne donnent que des r&ultats mtiocres. On peut remarquer cependant
cjti: kS xillcurs rCstiki1;s 51;iii SOUVCiii &&;;ds a\\&. &s cornpusk i>rg:ili~;qu~;b LUIIIIIIC
la tourbe (Uantillons D, B, K, N), les matières organiques fraîches (Z, R, S) étant
souvent de mauvais supports d’inoculums.
Conclusion
L’ensemble des rkwltats obtenus dans ces &udes et dans d’autres essais non pré-
sentés ici, montrent la difficulté qu’il y a à prévoir, par une simple analyse chimique,
le comportement d’un support vis-à-vis des rhizobiums. Parmi les rkgles pouvant être
tkonckes pour la fabrication d’inoculums dans lesquels les rhizobiums pos&ient une pro-
babifti de survie tlevtk, celles de la neutralité et de la st&ilid du support semblent
primordiales. Par la suite, une &Valuation bas& sur des tests biologiques de survie des
microorganismes semble obligatoire pour dCterminer les meilleures conditions de prc$a-
ration et de conservation des produits. En la matière, une bonne connaissance de l’&o-
lution du nombre de rhizobiums dans le support permet de déterminer la dur& maximale
d’utilisation des inoculums, de prkvoir leur dur& de production en conskquence, et de
prkiser les conditions de leur distribution, la tifrigération ne constituant pas toujours
une obligation.
149
Nombre de rhizobiums
,
vivants par gramme d’inoculum
(Log)
Sauche employ6e :
R. lquminowum biovar phnsd PIXC 40?4
8 i--.
>
0
8
15
2 2
3 6
50
Temps (jours)
Figure 3 : Effet dc! la tempkature de conservation des inoculums sur la survie des rhtzobiums
151
152
Cette étude n’a pris en compte, comme critère de qualité des inoculums, que le
nombre de Rhizobium vivants. Il n’est pas certain que dans tous les cas des modifica-
tions de l’efficacite des Rhizobium ne se soient pas produites. L’obtention et l’applica-
tion d’un nombre aussi élevé que possible de Rhizobium viables constitue cependant un
prt%lable indispensable a une bonne qualité des inoculums. C’est pourquoi le crX?re de
nombre a été seul consid6re dans l’etude.
Remerciements
Le auteurs remercient T. DORJI, S. MARTIN DU PONT, M. NEYRA, M. OCHOA,
D. PINOIT, P. RAKOTONDRAMASY et J.A. SCAGLIA pour leur participation aux ex-
p&imentations décrites. Ils remercient également le Service des engrais et de la nutrition
des plantes (AGLF) de la FAO pour la fourniture des différents supports.
Rbfbrences
DATE, R.A. and ROUGHLEY, R.J. 1977 Préparation of legume seed inoculants.
In : A treatise on dinitrogen fixation. Section IV, Edited by Hardy, R.W.F. and Gibson,
A.H., John Willey and Sons. p. 243-275.
FAO 1984 Fichier technique de la fixation symbiotique de l’azote, Organisation des
Nations Unies pour I’Alimentation et l’Agriculture. Rome.
SAINT MACARY, H., BEUNARD, P., MNTANGE, D., TRANCHANT, J.P. et
VERNIAU, S. 1986 Setting and diffusion of a production system for legume Rhizo-
bium inoculants. Symbiosis. 2, 363-366.
VINCENT, J.M. 1970 A manual for the practical study of root nodule bacteria.
IBP Handbook N*15. Blackwell Scient& Publications, Oxford and Edinburgh.
Maumiser
lo FBA pour la production agricole et fortxtière en Afrique
Progrès récents dans la technologie
des inocülums utilisés en agrictiture
et en foresterie
DIEMI, H.GP’, BEN KHALIF’A, KP’,
NEYRA, M.@) and DOMMERGUES, Y.R.“‘.
(1) : BSSlT (ORSTOMICTTFTICNRS)
45 bis, avenue de la Belle Gabrielle 94736 Nogent-sur Marne cedex (France) ;
(2) : SocU Calliope, 17, rue de Shastopol,
34500 Bhziers (France) ;
(3) : ORSTOM, BP. I386, Dakar (Sénégal).
Summaty
The preparation of inoculants requires (i) mass production of the target micro-
organism to be introduced in the agrosystem, and (ii) an appropriate formulation
of the microbial culture ensuring long survival during storage and case of appli-
cation
The pxinciples of mass culture methods for gmwing rhizobia, ectomycorrhizal
fungi, Frankiae, and VAM fungi are briefly presented. Emphasis is specially piaced
on the di!Sulties encountenzd when gmwing Frmkiue and on the solutions adopted
to increase its bîomass production, which includes biphasic culture, two stage culture,
ancl gmwth in a polymeric matrix.
The modem technique of processing the inoculants is based on the inclusion of
bacteria in a polyrneric matrix (generally alginate). Two types of polymeric it~~~hmts
cari be prepared : PEM inoculants made of microbial cells entrapped in the polymeric
matrix ; PEC inoculants made of miaobîal colonies entrapped in the matrix (the
colonks resulting from the growth of entrapped isolated cells). PEM and PEC ino-
culants are usually dried, then applied as a powder onto seeds or soil. or rehydrated
before beiig used. Polymeric ir~~hmts were shown to be quite effective and reliable
in field expeziments with rhizabia, VAM mycorrhîzae and frankiae.
154
Introduction
On peut classer en deux catkgories les microorganismes rhizosphériques et
phyllosph6rique.s qui améliorent la production végétale.
Le premier groupe est constitué par les microorganismes qui agissent directement
sur le veg&al en contribuant ZI sa nutrition azotée ; par ex. Rhiwbium ou Frankia, en
am&nanf l’absorption des min&aux et de l’eau, par ex. Awspidzun (Okon, 1984 ; Nayak
et al., 1986 ; Boddey et Dobereiner, 1988), champignons mycorhiziens (Menge, 1983;
Gianinazzi-Pearson e t G i i , 1986) ; en solubilisant certains 616ments. par ex. bac-
tkries solubiit les phosphates (Banza et al., 1%7 ; Azcon et af.. 1976) ; en synthe-
tisant des substances de croissance,* par ex. Awtobacter (Brown, 1982), Awspirillum
(Okon, 1987 ; Boddey et Do~re&r, 1988).
L.e deuxième groupe affecte indirectement la croissance du végétal en luttant contre
les microorganismes pathogènes, par ex. Trichoak-ma SP., Agrobacterizun radiobacter ;
en luttant contre les arthopodes, par ex. Beauveria bassiana ; en luttant contre les mauvaises
herbes (microorganismes pathogknes de ces plantes notamment mycoherbicides), par ex.
Phytophthora palmivora ; en pr6servant des effets du gel, par ex. Pseudomonas
syringae (Okon et Hadar, 1987).
En plus des microorganismes rhizophénques et phyllosph&iques, des microorga-
nismes endophytiques (c’est-à-dire vivant dans le xylème des plantes) pourraient être utili&s
à l’avenir, apAs manipulation génétique, comme agent de lutte biologique contre dif-
férents ennemis des plantes, spécialement des insectes (Carlson, 1988).
On pense que l’inoculation des plantes effectuée dans le but d’améliorer les r&oltes
est un type de biotechnoiogie &Zmentaire ; en fait il s’agit d’une op&ation relativement
compliquée (Bashan, 1986). Les techniques conventionnelles donnent des r&ultats in&
guliers et les m&hodes d’inoculation sont souvent primitives. C’est ainsi que dans certains
pays, on inocule encore les plantes actinorhiziennes avec des bmyats & nodules. On-cIoit
interdire une telle pratique en raison du risque d’introduction de souches ineffectives (non
furatrices d’azote) et surtout du danger de contamination des semis ou des boutures avec
des agents pathogènes, tels que Rhizoctonia solani ou Pseudomonas solanacearum dans
Ic cas dc Cuslrarina cqtisefifolia (Liang Zizh, l986) ou dcb nhu~des lorsque l’on
a affaire à des plantes sensibles à ces derniers parasites, telles que les acacias australiens
introduits en Afrique de I’Ouest..
Cette communication fait le point des progr&s les plus marquants réalisés r&em-
ment (i) dans la culture des quatres types les plus importants de microorganismes symbio-
tiques, Rhizobium, Frankiu, champignons ectomycorhiziens (EMC) et vésiculo-
arbosculaires (VAM), et (ii) dans le conditionnement du microorganisme utilisé pour
l’inoculation, Op&tion fond& sur l’inclusion du micmorganisme dans une matrice de
polymére, par exemple alginate. On donnera enfin les r6sultats d’essais d’inoculation au
champ portant sur ce nouveau type d’inoculum.
Culture en masse du microorganisme
La production d’inoculum repose sur la culture en masse du microorganisme que
l’on souhaite introduire dans l’agrosystème. Alors que l’on peut cultiver facilement certains
microorganismes in vitro (rhizobiums, champignons ectomycorhiziens), il est beaucoup
plus difficile de produire des quantit&s importantes d’autres microorganismes (par ex. Fran-
155
kia). En outre on n’est pas encore parvenu à cultiver in vitro un certain nombre de
microorganismes, en particulier les champignons VAh4.
Rhizobiums
On peut cultiver facilement les rhizobiums dans de nombreux types de fermen-
teurs, du type &atch» ou semi-continu (!Vii, 1984).
On peut aussi cultiver ces bactiries B l’état inclus dans une matrice polyémrique
suivant le pro&% pour les frankias (cf: infï-a). Contrairement aux lka&ias dont les CO--
lonies sont régulièrement distribuées à l’intkrieur des billes, les rhizobinm& comme les
autres microorganismes aérobies tels que Buciilus subtiZis (Baudet et al., 1983), poussent
seulement dans la couche externe des billes, dont le diam&re.ne devrait donc pas dépas-
ser 1 mm.
On sait maintenant que, dans la symbiose rhizobienne, nombre de composés appa-
rentés aux flavones stimulent l’expression des gi?nes nod, y compris les génes nod ABCEF
(seul le g&ne nod D est constitutif chez les rhizobiums ZI l’ttat libre), et acAlèrent la
nodulation, ce qui rend plus compétitifs les rhizobiums incubés en prksence de ces subs-
tances. Ces composés se trouvent dans les exudats des plantes-hôtes, mais la quantiti
exsudée peut ne pas être suffisante pour GAer la nodulation (Halverson et Stacey, 1986 ;
Redmond et al., 1986 ; Phillips et al., 1988). L’addition d’exsudats ou de substances
inductrices réduit le temps nécessaire à la nodulation (Rolfe et al., 1987 ; Wijffelman
et al., 1987). Ce traitement sera probablement commercialii dans un proche avenir.
Champignons ectomycorhiziens
On sait cultiver les champignons ectomycorhiiens (par ex. Stillus granr&tz~ ,
Rhizopogon luteus, Thelephora terrestris. Pisoüthus tinctorius) soit en milieu liquide,
dans difftkents types de fermenteurs y compris de grands fermenteurs industriels (Marx
et Kenney, 1982), soit sur substratum solide imprQné de milieu nutritif. Dans le second
cas, on met 2 incuber le substratum inoculk avec le champignon dans des sacs de plas-
tique de 5 à 10 litres dot& sur une face d’une membrane permettant les échanges gazeux
tout en empêchant des conramirwtinn?
Pwvtuelles (Le Taccr! c! Uxbr;:, 1936 ; Gsg-,~n
et ai., 1987).
Frankias
La culture in vitro de Frankia soulkve plusieurs problèmes @km et al., 1988)
dûs aux caractkistiques de cet actinomyc&e : (i) une croissance tri% lente, le temps de
doublement &ant de 50 A 60 h, (ii) l’inhibition de la croissance en batch lorsque la bio-
masse de Frankia dans le milieu atteint le seuil anormalement bas de 10 pg de prot&nes
cytoplasmiques par ml c’est-Mire 70-100 1.18 de protéines totales par ml de milieu,
(iii) une diminution II& rapide de la biomasse physiologiquement active à la fin de la
phase active de croissante (Fig.l), cette phase de dUin étant caract&is& par une lyse
marquée des diffkrentes structures de Frunkia, (iv) une sporulation souvent irr&uliére
et instable. Ce dernier point mérite d’être ttudib car on peut penser qu’en obtenant une
sporulatîon plus abondante, on améliorera la viabilité de Frankia dans l’inoculum
pendant le stockage.
Des expériences rikentes effectuks dans notre laboratoire suggèrent que l’on peut
stimuler la sporulatîon de Frunkiu en utilisant un milieu de culture limité en N ou en
apportant certains acides aminés (leucine, alanine, phénylalanine, arginine et valine).
L . E
h
156
Figure 1 : Courbe de croissance de la souche de fkankia ORS 020607. La courbe en fonction du temps de la teneur de la culture en protéines
comporte trois phases : (i) phase de latente (jusqu’au Sème jour), (ii) phase de croissance active, que l’on suppose de type linbaire, du 5kme
au 15ème jour et (iii) une phase plus ou moins stationnaire jusqu’au 962me jour.
157
Jusqu’à p&ent, on a utilise surtout les cultures en batch, avec ou sans agitation,
avec ou sans a&ation. Comme on ne peut utiliser ces méthodes pour produire les quantitks
importantes de Fr& nkessaires à la pr@aration des inoculums, nous cherchons ac-
tuellement a mettre au point trois nouvelles techniques de production en masse.
Culture blphaslque
Etant doMe que Frunkia pousse bien lorsqu’il est attaché à une surface solide,
Diem et Dommergues (1985) ont mis au point une m&hode de culture biphasique dans
laquelle on inocule le milieu liquide avec des fragments de gélose provenant de l’krasement
d’une culture de Fr& ayant pousse sur un milieu nutritif gt%&, les fragments de
gélose constituant des matrices où sont incluses de nombreuses structures de Frunkiu.
Cette méthode permet d’obtenir une croissance satistàisante de FrÜnkiu 21 l’interface entre
la courbe de gelose 6cras& (phase solide) qui tombe au fond du flacon de culture con-
tenant le milieu liquide (phase liquide). On obtient ainsi la formation, à l’interface solide-
liquide, d’un enchevêtrement homogène d’hyphes de Fr& ayant 2 mm d’6$akeur.
Une telle am&ioration de ht croissance au niveau d’une interface liquidesolide a aussi
et15 observée dans le cas d’autres bactt%ies (Bauder et al., 1983). Ce syst.&me & culture
biphasique permet en outre (i) de rktiver la croissance de vieillesoul~ de FrankG
qui ne pouvaient redm A la suite du simple transfert sur les milieux solides ou liquides
habituels, (iii d’assurer le développement des microcolonies obtenues au cours dé l’iso-
lement de Frmkiu iI partir des nodules, la biomasse initiale de ces microcolonies étant
trop faible pour en assurer leur croissance uMrieure dans les milieux solides ou lîq&les
habituels.
Culture ii deux atapes
Neyra a mis au point dans notre laboratoire une methode de culture a deux &apes
(resultats non publiés). Dans la premikre Ctape, on cultive Frunkia dans un chkmostat
conçu par Belaich, J.P. (comm. pers.) pour la culture des microorganismes a croissance
lente. Dans la deuxième ttape, la culture de Fr& qui sort du chemostat est utilisée
pour inoculer des batchs ancillaires contenant du milieu frais.
Culture en billes d’alginate
Le principe de la mtthode est le suivant : les celhrles de Frankîa sont incluses
aseptïquement dans des billes d’a@nate (qui constituent lamatrice polym&ique) obtenues
conform6ment au protocole dkrit au paragraphe consacre au conditionnement des îno-
culums et leg&ement modifie. Cette modification consiste a remplacer Yeau distilltk utilis&
pour faire la suspension d’alginate a 5 5% par un milieu de cuhure renfermant moins
de phosphate que le milieu usuel. Ces billes sont ensuite lavtks dans l’eau stkik et incubees
pendant 10-15 jours dans le milieu de culture. Les colonies de Frmkia qui se d&elop-
pentu à l’inteneur des billes (Fig. 5) renferment de nombreux sporanges au milieu de
chacune de ces colonies. Lorsque les colonies form6e.s a l’intérieur des billes sont suffkm-
ment développées, les billes sont &ch&s et constituent ainsi l’inoculation PEC en poudre
(cf : infra).
Dans le cas des plantes actinorhiziennes, l’expression des gks twd est proba-
blement induite par des substances du même type que ceBes qui interviennent dans la
symbiose Rhizobium-16gumineuse. En effet, Prin et Rougier (1987) ont montré que les
exsudat.s de 1’Alnus sont impliqub dans le processus d’infection.
-
158
Champignons mycorhiziens v&ico-arbosculaires (VAM)
Ainsi qu’on l’a signalé anterieurement, les champignons VAM ne peuvent être
cultivb in vitro, c’est-a-dire en l’absence de la plante-hôte. Contrairement aux rhizo-
biums et fknkias, ces champignons sont si peu spécifiques que presque toutes les espèces
végétales dotées d’un système racinaire abondant peuvent être utiliskes comme
plantes-hôtes.
La méthode habituelle, encore largement utilisée, consiste à cultiver la plante-hôte
dans un sol ou un substratmn stérile place dans un rtkipient ou directement sur le plancher
de la serre. On inocule le sol ou le substratum avec des spores ou des racines deja in-
fectks avec le champignon VAM à multiplier, provenant de cultures en pot effectuées
anterieurement. Linoculum ainsi prépare est volumineux et.. difficile 21 manipuler-
Récemment, Mosse et Thompson (1984) ont mis au point une méthode appelée
NET (Nutrient Film Technique). Cette methode consiste a faire pousser la plante-hôte
dont les racines ont et6 préalablement infectées avec le champignon VAM à multiplier
dans un canal étroit où coule la solution nutritive pour le vegétal. Le réseau dense de
racines formé dans ces conditions constitue un milieu idéal pour la multiplication du
champignon VAM. L’avantage de l’inoculum ainsi obtenu est qu’il suffit de 60 kg de
racines infectées pour traiter 1 ha, poids bien infkrieur à celui de l’inoculum classique
dkrit ci-dessus. En admettant que le rendement en racines soit de 300 g (poids frais)
par m de canal, on peut considérer qu’avec 200 m de canal on peut inoculer 1 ha @mes
et al., 1983). Il est 6vident que le matiriel ainsi obtenu n’est pas exempt de contami-
riants, les plus fi@uents ktant des protozoaires ou des nkmatodes saprophytes (Masse
et Thompson, 1984). La méthode NET a éte expkimentke avec succès sur haricot (Phu-
seolus vulgaris), maïs (Zea mays) et d’autres hôtes infestés avec GIomus mosseae,
G. fasciculatus, G. caleahium et Acadospora luevis @mes et Mosse, 1984).
Une autre. méthode est fond& sur l’emploi de plantes-hôtes cultivtks aeroponi-
quement. Hung et Sylvia (1988) ont confirmé des études antérieures montrant que cette
methode peut être utilisk avec SUC~S pour multiplier les champignons VAM. car une
bonne aération de la rhizosphère stimule la sporulation.
Enfin, Mugnier et Mosse (1987) ont proposé une approche originale pour amé-
liorer la culture des champignons VAM : au lieu d’utiliser la plante entière, on utilise
des racines transforrkes par le T-DNA cSAgrobacferium rhizogenes. Les racines trans-
formées de Convolvufw sepium, que l’on peut cultiver dans des fermenteurs usuels, sont
infectées par des spores de Glomus mosseae et assurent une bonne croissance du cham-
pignon. Ce pro&&, qui donne satisfaction au niveau du laboratoire, n’a pas encore été
commercialise.
Inoculants constitu& par des mtcroorganismes inclus dans une matrice
polym&ique
Le conditionnement des inoculums microbiens utilisés en agriculture est très im-
portant. Le succès de ces inoculums dépend en effet non seulement de la survie du plus
grand nombre possible de cellules microbiennes pendant le laps de temps s’écoulant entre
la fabrication et l’application, mais aussi de la formulation qui doit être simple mais per-
mettant un transport facile et une application aisée. Le conditionnement classique est fondé
sur l’adsorption de la culture microbienne sur un support protecteur, en g&&ral la tourbe
(Date 1977 ; Burton, 1979 ; Williams, 1984). Des supports de remplacement ont et& pro-
159
posés et utilisés tels que lignite, charbon, paille, cellulose, bagasse, résidus de récolte
broyés en poudre et sol (Chao et Alexanda, 1984). Dans certains cas, la culture et le
conditionnement du microorganisme sont effectues en même temps. C’est ainsi que les
inocuhrms d’ectomycorhizes sont faits d’un melange d’hyphes fongiques et du substrat
de vermiculite irnpr&ne du milieu spécifique sur lequel le champignon a pousse (Max
et Kenney, 1982).
Di+s 1977 et 1979, Dommergues et aL, ont suggeti que l’on pouvait utiliser des
gels de polym&es pour inclure les microorganismes b&%ques pour @parer des ino-
cuh.uns a usage agricole. Les premiers inocuhnns faits de microorganismes inchrs’ dans
des matrices polym&iques ont et6 tes& sur des rhizobiums il y a 10 ans. Ce concept
est maintenant appliqué avec succès, sur une &helle expt?rimentale, à d’autres groupes
de microorganismes, par ex. champignons VAM (Ganry ët 42.; 1982), champignons
ectomycorhiiiens (Jung et al., 1981 ; le Tacon et d., 1985), AzOspirilIum (Bashan, 1986)
et aussi divers chamignons utilises en lutte biologique (Lewis et. Papavizas, 1985). Nombre
de substances utili&s pour I’inunobilisation des cellules, enzymes, anticorps et autres
protéines (Scott, 1987) ont éte test&s pour preparer des inoculums polymériques :
polyacrylamide, alginate, xanthane (Jung et al., 1982) et de nombreux autres gels na-
turels tels que le carrageenan.
Cellules mlcroblennes
incluses dans une matrice polymerique
(PEM)
On peut utiliser divers polymeres pour inclure les cellules microbiennes dans une
matrice polyémerique, le produit obtenu étant désigné sous le terme d’inoculum poly-
merique PEM ou plus brièvement inoculum PEM (Polymer-entrapped microbial ceps).
Dans le cas de I’alginate de sodium (par ex. Alginate S 170 fabrique par SATIA, 15
avenue d’Eylau, 75116 Paris, France), qui est un polymère t&s largement employé, on
prepare une solution B 5% (poids/volume) de ce compose dans l’eau distillée et on mélange
cette solution avec le même vohnne de la culture microbienne à la fm de sa phase active
de croissance. Gn laisse tomber goutte Li goutte le melange constitue par la solution d’al-
ginate et la culture microbienne dans une solution O,l-0,2 M de CaCl, sous agitation
magnétique. En entrant en contact avec la solution de CaCI-. chaque gmtte forme une
bille sphérique indépendante (diamètre 14 mm) résultant du pontage de I’afginate par
les ions Ca2+ (Fig. 2). On maintient les billes dans la solution de CaCl, pendant 45 mn,
puis on les récolte et on les lave plusieurs fois dans l’eau st&ile ; finalement on les
sèche pendant 24 h sous hotte a flux laminaire (Fig. 3). On peut broyer les billes s&ches
pour en f%re une poudre de 50-200 prn. L’inoculum ainsi fabriqué est stocke dans des
récipients fermes. Dans le cas du rhizobium, chaque g d’inocuhrm-PEM séché renferme
10g a 101r unités formant des colonies (CFU).
i
160
160
Figure 2 : Inoculum PEh4 : billes d’alginate fraîches (non séchées) où sont inclus des rhizobiums
(diamètre : envinm 3 mm)
Figure 3 : Inoculm PEh4 : billes d’alginate séchées où sont inclus des frankias (longueur : environ
1.5mm)
161
Dans le cas des champignons VAM, le materie microbien inclus dans la matrice
polymérique n’est pas seulement constitué de structures fongiques (spores, hyphes, vé-
sicules) mais aussi de fragments de racines infect& par le champignon (Fig. 4). Toutes
ces structures doivent être soigneusement lavees avant d’être mélangées avec la solu-
tion d’alginate. Les demiks phases de la préparation de ces inoculums sont les mêmes
que celles concernant la préparation des inoculums bactériens (Jung er al., 1981 ; Gamy
et al., 1982).
Figure 4 : Inoculum PEN : billes d’alginate fraîches (non séchées) où sont inclues des structures
de champignon VAh4 et des fragments de racine (diamètre : environ 3 mm)
L’iuoculum PEM est appliqué aux plants (généralement sur les graines ou dans
le sol de la raie de semis) soit sous forme skchke, soit sous forme d’une pseudo-solution
obtenue en mettant l’inoculnm seché dans une solution tampon (0,03 M KH,PO, ;
0,17 M K.$PO, ; pH 7,4) pendant 4-8 h.
L’inoculum sec peut être faxé sur les graines avec une substance adhtsive (Elegba
et Rennie, 1984), telle que la gomme arabique (40% poids/volume). On peut ainsi fixer
entre 1 et 2 mg d’inoculum PEM sec sur une graine de soja, soit plus de 106 à
2 x 10” CFU par graine.
Les principaux avantages des inoculums PEM à base d’alginate (avantages pro-
bablement partages avec les autres inoculums PEM) sont les suivants : (i) leur faible
poids qui les rend faciles à manipuler et à transporter de l’unité de fermentation au site
d’application ; (ii) ils peuvent facilement être melanges les uns aux autres ou avec dif-
ferents produits chimiques tels que des pesticides (Dreyfus et QI., 1983) ou des subs-
tances protectrices ; (iii) apres un skchage convenable de l’inoculum, de nombreuses
espkes de microorganismes peuvent survivre longtemps : On peut conserver Frankia et
les champignons VAM pendant 2-3 ans à 20-25 C sans qu’ils perdent leur pouvoir infectif.
161
Dans le cas des champignons VAM, le matkiel microbien inclus dans la matrice
polymérique n’est pas seulement constitué de structures fongiques (spores, hyphes, ve-
sicules) mais aussi de fragments de racines infectés par le champignon (Fig. 4). Toutes
ces structures doivent être soigneusement lavées avant d’être mélang&es avec la solu-
tion d’alginate. Les dernières phases de la ptiparation de ces inoculums sont les mêmes
que celles concernant la préparation des inoculums bactériens (Jung et al., 1981 ; Ganry
et al., 1982).
L’inoculum PEM est appliqué aux plants (généralement sur les graines ou dans
le sol de la raie de semis) soit sous forme séchée, soit sous forme d’une pseudo-solution
obtenue en mettant l’inoculum séché dans une solution tampon (0,03 M KH,PO, ;
0,17 M K$PO, ; pH 7,4) pendant 4-8 h.
L’inoculum sec peut être fixé sur les graines avec une substance adhésive (Elegba
et Rennie, 1984), telle que la gomme arabique (40% poids/volume). On peut ainsi fixer
entre 1 et 2 mg d’inoculum PEM sec sur une graine de soja, soit plus de 106 à
2 x 108 CFU par graine.
Les principaux avantages des inoculums PEM à base d’alginate (avantages pro-
bablement partagés avec les autres inoculums PEM) sont les suivants : (i) leur faible
poids qui les rend faciles à manipuler et à transporter de l’unité de fermentation au site
d’application ; (ii) ils peuvent facilement être mClangés les uns aux autres ou avec dif-
férents produits chimiques tels que des pesticides (Dreyfus et al., 1983) ou des subs-
tances protectrices ; (iii) après un séchage convenable de l’inoculum, de nombreuses
espèces de microorganismes peuvent survivre longtemps : On peut conserver Frankia et
les champignons VAM pendant 2-3 ans à 20-25°C sans qu’ils perdent leur pouvoir infectif.
162
De même il est clair que certains champignons ectomycorhixiens béneficient gran-
dement de leur inclusion dans l’alginate (Jung et al., 1981 ; le Tacon et al., 1985).
La survie des rhizobiums (Jung ef al., 1982). et plus gén&akment des bactkies
ne formant pas de spores, depend de la uSrance des souches au processus d’inclusion
(g&fkation du polyr&re), de la nature des carbohydrates pr6sents dans la matrice et
de la teneur f&e en eau (en fait l’activité de l’eau : a> de l’inoculum seché. Mugnier
et Jung (1985) ont montre que les carbohydmtes à faible poids mohZculaire (C, a C$
avaient un effet défavorable sur la survie, alors que les composes de poids molkculaire
plus klev6 (C, à C,d protégaient les microorganismes. Avec les souches test& par ces
auteurs, le nombre de cellules vivantes (10’0 par g d’inwulum .sec) reste constant pendant
un stockage d’une dur& sup&ieur.e B 3 ans a 28°C quand a, de l’inoculum est maintenue
au dessous de 0,069. Dans I’iutervalle de aw 0,069-0,830, le taux de survie diminue d’autant
plus vite que a, est élevée. La tokkance des rhizobiums a la dessiccation varie en fonction
de la souche considérée et d’autres facteurs tels que la phase de croissance (Mary et
al., 1985). Lorsque a” de I’inoculum tombe au-dessous d’un certain seuil, la perte d’eau
des cellules peut devenir ir&versible de sorte que les bactkies incluses meurent
Jl est ensentiel que le nombre de bactéries viables dans l’inoculum se maintienne
à un niveau Hevt! entre le moment de la production et celui de l’application. Une bonne
conservation de la viabiité pendant le stockage permet au fabricant de pnZparer les mo-
culums bien avant qu’ils ne soient r&lam& par l’utilisateur, ce qui est tout particulie-
rement important lorsqu’on doit disposer d’un large éventail de souches.
Des recherches sont encore nt%ssaiw pour améliorer la survie des microorga-
nismes dans les inoculums PEM et PEC. Differentes approches peuvent être employées
pour assurer la survie des microorgauïsmes pendant l’inclusion, la dessiccation et le
stockage :
(i) utilisation d’un agent g&fiant approprie, par ex., dans le ‘cas de l’al-
ginate, remplacement de CaC& par le gluconate de Ca (C,$J&IaO,S
(Fravel et ai., 1985) ;
(ii) incorporation dans la matrice de substances protectrices telles que les
argiles (Jung et ul., 1982 ; Fravel et ul., 1985) ou des carbohydrates
de poids moléculaire élevé comme on l’a signalé plus haut ;
(iii) stimulation de la production in vitro de structures de rf%istance teks
que les sporanges (Fr&), ascospores et conidies (champignons) ou
sclerotes (par ex. dans le cas de Phlebopus soudanensis, (Ihoen, D.
Comm. pers.) ;
(iv) stockage à basse tempt%ature.
Colonies incluses dans une matrice polymbrlque (PEC)
Ce type nouveau d’inoculum est obtenu à partir de billes d’alginate dans lesquelles
on a laissé croître les microorganismes, qui forment donc des colonies à l’intérieur des
billes @iem et Dommergues, 1985). On désigne cet inoculum sous le terme d’inoculum
polymétrique PEC ou plus brièvement inoculum PEC (Polymer-entrapped colonies)
(Fig. 5). La préparation de ces inoculums a Cté dkrite ci-dessus. D’après les recherches
pr&ninaires en cours, les inoculums PEC renferment un plus grand nombre de CFU
que les inoculums PEM, chez lesquels il n’y a pas eu de croissance des microorganismes
*-’ f.. :. - - _... . ,._ ,.
-
__
-_
__
._._-_ _ _, _ ___ .____ .-e^.i.i-i.~:=
,.
*-. . .
.-
-.-
--.
_,
_____
163
après l’inclusion dans la matrice polymérique. Des expériences au champ sont actuel-
lement en cours pour comparer la réponse des legumineuses et des plantes actinorhizien-
nes à l’application des inoculums PEC et PEh4.
Figure 5 : Inoculum PEC : colonies de fkmkia poussant dans des billes d’alginate (diamètre : environ
3mm)
Expériences au champ
Inoculum polymérique PEM de rhizobium
En 1979, on a commencé des expériences portant sur l’utilisarion d’inoculums
polymériques PEM renfermant des rhizobiums, aussi bien pour l’application au sol dans
la raie de semis que pour l’enrobage des graines. Une de ces expériences a été mise
en place à Guérina au Sud du Sénégal, dans un sol sableux déficient en N (0,026% N)
pendant la saison des pluies, c’est-à-dire de Juillet à Octobre, 1980. L’objectif de cette
expérience était de comparer la réponse du soja A l’inoculation avec trois types d’ino-
culums, deux inoculums polymériques (PEM-alginate et PEM-xanthane) et un inoculum
liquide (culture liquide de rhizobium). Le sol utilisé n’avait pas été cultivé en soja aupa-
ravant de sorte qu’il ne renfermait pas le Bradyrhizobium spécifique du soja. La fertili-
sation avait consiste dans l’apport de 30 kg P (superphosphate) et de 100 kg de K (KCl)
par ha.
On a comparé cinq traitements : pas d’inoculation (témoin) ; inoculation des graines
enrobees avec la poudre d’inoculum PEM-alginate ; inoculation des graines enrobées avec
la poudre d’inoculum PEM-xanthane ; inoculation du sol avec une culture liquide de Rhizo-
bium appliquée à une dose très élevée ; application d’engrais azoté à la dose de
300 kg N par ha. Le dispositif expérimental comprenait quatre blocs randomisés constitués
de parcelles élémentaires de 8 m x 8 m. Le Tableau 1 montre que les inoculums PEM
163
après l’inclusion dans la matrice polymerique. Des expériences au champ sont actuel-
lement en cours pour comparer la r@onse des légumineuses et des plantes actinorhiiien-
nes à l’application des inoculums PEC et PEM.
Exptkiences au champ
Inoculum polym&ique PEM de rhizobium
En 1979, on a commencé des expériences portant sur l’utilisation d’inoculums
polyrnériques PEM renfermant des rhizobiums, aussi bien pour l’application au sol dans
la raie de semis que pour l’enrobage des graines. Une de ces exp&iences a Cté mise
en place a Gu&i~ au Sud du Sénégal, dans un sol sableux déficient en N (0,026% N)
pendant la saison des pluies, c’est-a-dire de Juillet à Octobre, 1980. L’objectif de cette
exp&ience etait de comparer la @onse du soja à I’inocuhtion avec trois types d’ino-
culums, deux inoculums polymériques (PEM-alginate et PEM-xanthane) et un inoculum
- liquide (culture liquide de rhizobium). Le sol utilise n’avait pas eté cultivé en soja aupa-
ravant de sorte qu’il ne renfermait pas le Brudyrhizobium spécifique du soja. La fertili-
sation avait consiste dans l’apport de 30 kg P (superphosphate) et de 100 kg de K (KCI)
par ha.
On a compare cinq traitements : pas d’inoculation (témoin) ; inoculation des graines
enrobees avec la poudre d’inoculum PEM-alginate ; inoculation des graines enrobees avec
la poudre d’inoculum PEM-xanthane ; inoculation du sol avec une culture liquide de Rhizo-
bium appliquee à une dose très élevée ; application d’engrais azoté à la dose de
300 kg N par ha Le dispositif exp&imentaJ comprenait quatre blocs randomisés constitues
de parcelles élémentaires de 8 m x 8 m. Le Tableau 1 montre que les inoculums PEM
164
sont aussi efficaces que l’inoculum liquide appliqué directement 3 trks forte dose. En
ce qui concerne la nodulation, I’inoculum PEM-xanthane est apparu supbrieur & I’ino-
culum PEM-alginate ; mais pour les rendements en graines, il n’y a pas eu de diffhence
significative entre ces deux inoculums (Jung et aI., 1982).
On a utilisé ultkrieurement avec succès les inoculums PEM dans le cas des Iégu-
mineuses à nodules de tige, plus particuli&ment celui de Sesbuniu rosfrafa (Dreyfus
et al., 1983).
Tableau 1 : Réponse du soja à l'inoculation des graines avec l'inoculum PEM-
alginate et l'inoculum PEM-xanthane et à l'inoculation du sol avec
une culture liquide. Expérience effectuée à Guérina, Sénégal
(Jung et al., 1982).
Traitements
Rendement en grain (kg/ha)
Poids sec des
nodules
(g/plantel
Poids sec
N total
Pas d'inoculation
1919 c
107 b
0,08 c
'
Inoculation des graines
r i
PEM-alginate
2752 ab
190-a
1,6b
"'
PEM-xanthane
2773 a
176 a
5,0 a .
~
Inoculation du sol
culture liquide
2445 ab
16G a
2,2 b
Engrais azotë
2366 bc
130 ab
0,02 d
(300 kg N/hal
Les chiffres suivisdes mémes lettres ne diffërent pas significativement pour
P = 0,Ol.
165
Inoculum polym&ique PEM de Frankia
On a mis en place au Sénégal une experimentation sur une échelle relativement
grande pour évaluer la réponse de Casuarina equisetifolia 2 l’inoculation avec un ino-
culum PEM de Fr&.
Le 15 mai 1984,‘on a plante des semis âgés d’un mois dans des gaines de polyé-
tbylene remplies d’un sol tr& pauvre (OP% C ; 0,026% N) st&ilisé au pliable au bromure
de méthyle. La moiti6 des semis a été inoculee ; l’autre moitié @moins) ne l’a pas IX
L’inoculum PEM de Frankiu a été fabriqué avec la souche de Frunkzb ORS 021001 (Diem
et al., 1983) incluse dans des billes d’alginate prepardes comme il est indiqué ci-dessus,
ces billes ayant été séchées a l’air et conservées pendant 2 ans a la temp&ature du labo-
ratoire (20-25°C) avant d’être utilisées. On a procédé à I’inoculation en introduisant
dans chaque gaine une suspension de Frankia dans un tampon phosphate pendant
4-8 h (15 g d’inoculum PEM-alginate sec dans 1,5 1 de tampon) et en homogén&ant
la gelée liquide ainsi obtenue. Pendant tout leur séjour en serre, chacun des semis de
Cusuarinu equisetifolia a rqu une fois par semaine 100 ml d’une solution Hoagland
sans azote diluée au 1/4. Le 23 juillet 1984, lorsque les plantes avaient environ 20 cm
de haut, elles ont ti mises en place a Notto, à 80 km au Nord de Dakar et a 5 km
de la Côte. Le sol qui était un sol dunaire sableux très pauvre (0,047-0,121% N), n’avait
jamais eté plante auparavant en Casuarina et ne contenait pas de Fmnkia. La planta-
tion a été faite apres que les pluies eussent mouille le sol jusqu’à une profondeur de
30-40 cm. Le dispositif exp&imental comportait huit parcelles elémentaires de 416 m2
(quatre temoins, quatre inoculées) avec 84 plants par parcelle, ce qui correspond à environ
2 000 plants par ha. Une fertilisation phosphatde a ti appliquée aux doses suivantes ;
premiere année, 5 g K.$PO, ,* deuxieme année 10 g ; troisieme année 15 g. Les pluies
ont et& respectivement de 215, 375 et 473 mm en 1985, 1986 et 1987. Au cours de
la Premiere année, on a apporte 200 1 d’eau .a chaque plant, mais cette irrigation n’a
pas Cté répétée au cours des deux années suivantes.
Pendant trois ann&s consécutives, soit en Juillet 1985, Juillet 1986 et Juillet 1987,
dans chacune des huit parcelles, on a arraché cinq arbres repérés au hasard. Apres les
avoir mesures et pesés, on a divise chacun des arbres en quatre fractions : rameaux assimi-
lateurs, branches,’ racines et nodules. Chaque fraction a éte séchée et broyee. Après
echantillonnage, on a déterminb N total. Le Tableau 2 montre que la reportse de Cusuarina
equisetifoliu à l’application d’inoculum PEM de Frankia est significative (P < 0,05) ou
très significative (P c 0,Ol) sauf en ce qui concerne le paramètre hauteur. L’effet favorable
de l’inoculum a porte sur la biomasse des différentes fractions des arbres exprimée en
poids sec ou N total. Le rapport parues a&iennes/racines calculé sur la base du poids
sec ou de N total a tte également favorablement modifié par l’inoculation (Sougoufara
et al., 1989).
Les résultats indiquent clairement que l’inoculum PEM de Fnmkia est un ino-
culum efficace et qu’il peut être stocke pendant 2 ans sans perdre ni son inefectivité
ni son effectivité.
Inoculum polymérique (PEM) de champignon VAM
En 1982, Ganry et al. ont mis en place au Sénegal une exp&ience destinée a
evaluer la rt$onse du soja à l’application d’un inoculum PEM de Glomus mosseue fait
de spores, mycelium et racines mycorhiennes homog&&&es, ces différentes structures
166
Tableau 2 : Réponse de Casuarina equisetifolia à l'inoculation avec l'inoculum
PEH-alginate de frankia 1, 2 et 3 ans après la mise en place au
champ. Expérience effectuee à Notto. Sénégal (Sougoufara et al., 1989).
Poids sec tg/arbre)
N total (g/arbre)
Rapport(l)
(3)
P.A./Rac.
basé sur
br.
ras.
rat.
nod.
br.
ras.
roc.
nod.
PS N
-~ -
-
-
-
-
-
-
-
-
Juillet 1985 (1 an après la transplantation au champ)
0
151
163
89
266
0
0.63
1.46
1.24
0
0.95
1,69
+
182
324
119
309
2.55
1.99
2.17
1.60
0.04
1.42
2,54
*
*t
**
*
**
**
*
*
**
Juillet
986 (2 ans aprës la transplantation au champ)
0
291
890
628
507
0
5,64 -9,03
2,52
0
2,99'
5,82
+
313
1349
871
626
12.18
9.39
16.24
3.21
0.20
3,48
7.5;
NS
*t
**
**
**
**
**
* *
Juillet
987 (3 ans après la transplantation au champ) (5)
0
500
1451
1000
840
0
7,32
15,57
4.74
0
2,91
4,85
+
575
2220
1356
1031
22.75
12,20
23.35
5,12
0.40
3,39
6.47
NS
**
**
**
**
**
NS
**
*
Seuil de signification : ** P = 0, 01 ; * P = 0, 05 ; NS = non significatif
(1) 1. = inoculation. 0 : pas d'inoculation (témoins) ; + : inoculation avec l'ino-
culum PEM-alginate de la souche de frankia ORS021001.
(2) Hauteur moyenne des arbres en cm.
(3) br. = branches y compris les troncs ; ras = rameaux assimilateurs ; rat. = racines;
nod. = nodules.
(4) Rapport "Parties aëriennes/Racines" exprimé en poids sec (PS) ou N total (NI.
(5) Au cours de la 3ëme année des attaques de borers ont très sensiblement ralenti
la croissance des arbres.
167
ayant été incluses dans des billes d’alginate. Glomus mosseae avait été au préalable mul-
tiplie sur ni&& (Vigne unguicdafu). Le champ d’expérience, situe à la station ISRA (IIE-
titut Sertégalais de Recherches Agricoles) de Séfa, au Sud du Sénégal, avait été choisi
en raison du faible potentiel infectieux de son sol en ce qui concerne les champignons
VAM. Ce sol deficient en N (environ 0.04% N) n’avait jamais éd cultivé auparavant
en soja Le dispositif exp&imental utilise était du type split-plot avec huit répétitions.
Les trois traitements principaux (chaque parcelle principale avait 40 rnq ont été les sui-
vants : T (témoin), inoculation avec une souche ineffective de Bradyrhiwbium sur support
tourbe ; B, inoculation avec une souche effective de Bradyrihiwbium sur support tourbe ;
BG, inoculation avec la souche effective de Bradyrhizobiwn sur support tourbe (comme
dans le traitement B) et inoculation avec inoculum PEM-alginate de Glomus mosseae
(conserve à la chambre froide pendant plus d’un mois) appliqué a la dose de 15-20 billes
fraîches (non s&h&s) à 34 cm de profondeur autour de chaque semis de soja, lorsque
ceux-ci avaient 15 jours. Les deux sous-traitements ont été les suivants : OP, pas d’appli-
cation de fertilisation phosphatée ; P, application de superphosphate à la dose de 20 kg
par ha. L’azote marqué (‘-?Y) a été appliqué sur des microparcelles de 1.65m2 pour
mesurer la fixation de N, par la methode de la valeur A. On a analyse séparément
feuilles, tiges, gousses et graines. Par souci de simplification, les donnees concernant les
diff&eme.s parties a&iennes des plantes, à l’exception des graines ont été regroupées sous
le terme de «parties aeriennew. Les chutes de pluie ont Cté de 170 mm avant le semis
(ler Mai - 17 Juillet) et de 692 mm pendant le cycle cultural (17 Juillet - 10 Octobre).
La distribution des pluies a été assez regulière, sans période sèche marquée.
Le Tableau 3 montre clairement que l’inoculum PEM-alginate de Glomus mosseae
peut être utilise avec suceb au champ, en tant que traitement complementaire de l’ino-
culum de Bradyrhizobium, ce qui confirme les experiences publiees ant&ieurement par
Ganry et 42.. (1982).
Quand on a appliqué une fertilisation phosphatée legère (20 kg P/ha) I’inoculum
PEM-alginate de Glomus mosseae a augmenté la fixation de N, de 109 (inoculation avec
Bradyrhizobium seul) B 139 kg N$a, ce qui constitue une amélioration substantielle.
Dans les parcelles ayant reçu la fertilisation phosphatée, l’inoculum PEM-alginate de
Glomus mosseae a augmente significativement le rendement en grain exprime en poids
sec, et N total, et en outre améliore l’indice de récolte. Ces r&ultat.s positifs revêtent
une grande importance pour le fermier.
Conclusion
II est maintenant clair que l’on peut ameliorer considerablement la réponse des
plantes à l’inoculation en faisant appel à la technologie fondée sur l’utilisation des
microorganismes inclus dans une matrice polym&ique.
Etant donne que la production de ces inoculums perfectionnes exige un certain
niveau d’expertise, les usines de fabrication d’inoculums polymeriques devraient étre ins-
tall&s clans un nombre limite de centres d’excellence. En outre, les exp&iences au
champ, qui sont capitales pour toutes les souches, devraient être mises en place dans
les diff&-entes regions écologiques où l’inoculation est reconnue indispensable. Ceci est
particulièrement vrai pour les arbres fixateurs de N,
On peut penser que le developpement de l’utilisation d’inoculum de haute qualité
n’accroîtra pas seulement la fixation de N2 et la croissance des plantes, mais aussi ren-
168
Tableau 3 : Réponse du soja à l'inoculation avec une souche non effective
(non fixatrice de N2) de Bradyrhizobium (traitement T), d'une
souche effective de Bradyrhizobium seule (traitement BI OU
associée à Glomus mosseae, ce dernier inoculum étant un ino-
culum PEM-alginate (traitement BG). Dans la moitié des parcelles
on n'a pas apporté de P (traitement PO), dans l'autre moitié
on n‘a apporté P sous forme de superph0sphate.à la dose de
22 kg P/ha (traitement Super). Expérience effectuée à Séfa,
Sénégal (Ganry et al., 1985)
'Traitement
Rendement en grain
Grains et parties
Indice de
N2 fixé
(kg/ha)
aériennes (kg/ha)
récolte
sur la base (kg'ha'
PS
N
N.
P
de N total
T OP
Super
1093
1725
65,7
101,6
84,2
127,2
1;::
3,55
3,98
0
B OP
1423
90,3
4,18
73,1
B Super
2017
133,8
4,71
109,o
.
BG OP
1431
98,2
120,o
699
4,42
80,2
BG Super
2290
154,7
183,7
11,8
5,33
139.3
PPDS entre les traitements'
T,
B et BG
192
12,8
OP et Super
197
13,5
169
forcera leur rt%iitance aux parasites et pathogenes, ce qui permettra une diminution des
besoins en engrais et pesticides, d’où une r&htction des sources de pollution.
Actuellement, on met au point en France la fabrication à l’echelle industrielle
d’inoculums PEM et PEC pour les rhizobiums, Frunkziu et differents microorganismes
impliques dans la lutte biologique. Cette technologie devra être adaptée aux pays en voie
de développement dans un proche avenir.
Remerciements
Ce travail a &6 financé en partie par le contrat ANVAR no A 87 03 133 Q
AL 1340. Nous sommes reconnaissants au Professeur J.P. Belaich des conseils qu’il
nous a don& -pour le lancemem de la culture continue de Frankia.
AZCON, R., BAREA, J.M. and HAYMAN, D.S. 1976 Utilization of rock phosphate in
alcaline soils by plants inoculated with mycorhizal fungi and phosphatesolubiliziig
bacteria. Soi1 Biol. Biochem. 8, 135-138.
BAREA, J.M. NAVARRO, E. and MONTAYA,iE; t1976 Production of plant growth
regulators by rhizosphere phosphate-solubilizing bacteria. J. Appl. Bacteriol. 40,
129-134.
BASHAN, Y. 1986 Alginate beads as synthetic inoculant carriers for slow release of
bacteria that affect plant growth. Appl. Environ. Microbiol. 51, 1089-1098.
BAUDET, C., BARBOTIN, JN. and GUESPIN-MICHE!L, J. 1983 Growth and sporu-
lation of entrapped Bacilles subtilis cdis. Appl. Environ. Microbiol. 45, 297-301.
BODDEY, R.M. and DGBEREJNER, J. 1988 Nitrogen fmtion associated with grasses
and cemals : Recent results and perspectives for future research. Plant and Soil. 108,
53-65.
BROWN, M.E. 1982 Nitrogen fixation by free-living bacteria associated with plants -
fact or fiction ? In : Rhodes-Roberts M. and Skinners F.A. (eds.) Bacteria and plants,
pp. 2541. Academic Press, London.
BURTON, J. 1979 New developments in inoculating legumes. In : Subba Rao NS (ed.)
Recent advances in Biological nitrogen furation, pp. 380-405. Oxford and IBH Pu-
blishers, N e w D e h l i .
CARLSON, P.S. 1988 Microorganismes endophytiques : un syst&me r6volutio~ de
protection biologique des plantes. In : Abstracts «Advanced technology and plant bree-
ding strategy. Potential applications to woody species» meeting ,organized by the
Mo& Hennesy. Louis Vuittorm group, 19-21 Sept. 1988, Versailles.
CHAO, WL. and ALEXANDER, M. 1984 Mineral soils as carriers for Rhîzobium ino-
culant.~. Appl. Environ. Microbiol. 47, 94-97.
DATE, R.A. 1977 The development and use of legume incculants. In : Ayanaba, A.,
Dart, P.S. and Gibson, A.H., (cds.) Biological nitrogen fixation in farming systems
of the tropics, pp. 243-276 John Wiley, New York
170
DIEM, H.G. and DOMMERGUES, YR., 1985 Procede de culture de microorganismes,
notamment du groupe des Frankia et pr@ration d’inocnlums bacteriens. Demande
de brevet no 85-17933 du 04-12-85 (Method for cultivating microorganisms parti-
cularly of the Frankia group and preparation of bacterial inoculums. Patent appli-
cation PCT/FR86/00417 December 4, 1986).
DIEM, H.G., GAUTHIER, D. and DOMMERGUES, YR. 1983 An effective strain of
Frankia from Casuarina sp. Can.J. Bot. 61, 28152821.
DIEM, H.G., DUHOUX, E., SIMONET, P. and DOMMERGUES, YR. 1988 Actinorhizal
symbiosis biotechnology : the present aud the future. In : Durand G., Bobichon L.,
and Florent J. (eds.) 8th International Biotechnology Symposium Paris, 1988. Procee-
dings, Vol. II, pp. 984-995. Soc%% Française de Microbiologie, Paris.
DOMMERGUES, YR., DIEM, H.G. and DIVIES, C. 1977 procédé microbiologique
destine à maîtriser la productivit& des plantes cultivées. Demande de brevet no 77-
10254 du 05-04-77 (Microbiological process for comrolling the productivity of culti-
vated plants, US patent no 4 155-737, May 22, 1979).
DOMMERGUES, Y.R., DIEM, H.G. and DIVIES, C. 1979 Polyacrylamide-entrapped Rhi-
zobium as an inoculant for legumes. Appl. Environ. Mcrobiol. 37, 779-781.
DREYFUS, B., RINAUDO, G. and DOMMERGUES, YR. 1983 Pro&& d’inoculation
des plantes, notamment des I&umineuses à nodules caulinaires. Demande de brevet
no 83 02847 du 22.02.83.
ELEGBA, M.S. and RENNIE, R.J. 1984 Effect of different inoculant adhesive agents
on rhizobial survival, nodulation, ,and nitrogenase (acetylene reducing) activity of
soybeau (Glycine max (L.) (h!terill.). Can. J. Soi1 Sci. 64, 631-636.
ELMES, RP. and MOSSE, B. 1984 Vesicular-aruscular endomycorrhizal inoculum pro-
duction. II. Experiments with maize and other hosts in nutrient flow culture. Can.
J. Bot. 62, 1531-1536.
ELMES, R.P., HEPPER, C.M., HAYMAN, D.S. and O’SHEA, J. 1983 The use of vesi-
cuhrr-arbuscular mycorrhizal roots grown by the nutrient film technique as inoculum
for field sites. Ann. Appl. Biol. 104, 437-441.
FRAVEL, D.R., MAROIS, JJ., LUMSDEN, RD. and CONNICK, Jr. WJ. 1985 En-
capsulation of potential biocontrol agents in a alginak-clay mauix. Phytopathology.
75, 774-777.
GAGNON, J., LANGLOIS, C.G. and FORTIN, J.A 1987 Growth of containerized jack
pine seedhngs inoculated with different ectomycorrhizal fungi under a controlled fer-
tiliiation schedule. Cari. J. For. Research. 17, 840-845.
GANRY, F., DIEM, H.G. and DOMMERGUES, YR. 1982 Effect of inoculation with
Glomus mosseae on nitrogen fixation by field-grown soybean. Plant and Soil. 69,
321-329.
GANRY, F., DIEM, H-G., WEY, J. and DOMMERGUES, Y.R. 1985 Inoculation with
Glomus mosseae improves N, fixation by field-grown soybean. Biol. Fert. Soifs. 1,
15-23.
GIANINAZZI-PEARSON. V. and GIANINAZZI, S. (eds) 1986 Physiological and gene-
tical aspects of mycorhizae. INRA, Paris.
.
..i
171
HALVERSON, LJ. and STACEY, G. 1986 Signal exchange in plant-microbe interactions.
Microbiol. Reviews, June, 193-225.
HUNG, L.L. and SYLVIA, D.M. 1988 Production of vesicuhu-arbuscular mycorrhizal
fungus inoculurn in aeroponic culture. AppJ. Environ. Microkiol. 54, 353-357.
JUNG. G., MUGNIER, J., DOMMERGUES, Y.R. and DIEM, H.G. 1981 ProcedC d’inclusion
de microorganismes du groupe des mycorhizes et des actinorhizes. Demande de brevet
no 81-04474 du 06-03-81 (Process for inclusion of mycorrhizae and actinorhizae in
a matrix, US patent pending).
JUNG, G., MUGNIER, J., DIEM. H.G. and DOMMERGUES, Y .R. 1982 Polymer-entrapped
Rhiwbium as an inoculant for Iegumes. Plant and Soi1 65, 219-231.
LE TACON, F. and GARBAYE, J. 1986 La maîtrise des associations mycorhiziennes
en pépinière forestière. Rev. For. Fr. 38, 249-257.
LE TACON, F., JUNG, G.. MUGNIER, J., MICHELOT, P. and MAUPERlN, C. 1985
Efficiency in a forest nursery of an ectomycorhizal fungus inoculum produced in a
fermentor and entrapped in polyméric gels. Cun. J. Bot. 63, 1664-1668.
LEWIS, J.A. and PAPAVIZAS, G.C., 1985 Characteristics of alginate pellets formulated
with Trichoderma and Gliocladium on their effect on the proliferation of the fungi
in the soiL Plant pathology, 34, 571-577.
LIANG. ZICHAO 1986 Vegetative propagation and selection on Casuarina for resistance
to bacterial wilt. Tropical Forestry (Science and Technology) Guangzhou. 2, l-6 (In
chinesse, summary in english).
MARY, P., OCHIN, D. and TAILLEZ, R. 1985 Rates of drying and survival of Rhizobium
meliloti strains during storage at different relative humidities. App. Environ. Micro-
biol. 50, 207-211.
MARX, D.H. and KENNEY, D.S. 1982 Production of ectomycorhizal fungi inoculants.
In : Schenck N.C. (ed.) Methods and principles of mycorrhizal research, pp. 131-
146. The American phytopathological Society, Saint-Paul, Min.
MENGE, J.A. 1983 Utihzation of vesicular-arbuscular mycorhizal fungi in agriculture.
Can. J. Bot. 61, 1015-1024.
MOSSE, B. and THOMPSON, JP. 1984 Vesicular-arbusctdar endomycorrhizal inoculum
production. 1. Exploratory experiments with bean (Phaseolzu vulgaris) in nutrient flow
culture. Can. J. Bot. 62, 1523-1530.
MUGNIER, J. and MOSSE, B. 1987 Vesicular-arbuscular mycorrhizal infection in trans-
formed mot-inducing T-DNA roots grown axenically. Phytopathology. 77, 1045-1050.
MUGNIER, J. and JUNG, G. 1985 Survival of bacteria and fungi in relation to water
activity and the solvant properties of water in biopolymer gels. Appf. Environ. Microbiol.
5 0 , 1 0 8 - 1 1 4 .
NAYAK. D.N., LADHA, J.K. and WATANABE, 1. 1986 The fate of marker Azospirillum
lipoferum inoculated into rice and its effect on growth, yield and N, fixation of plants
studied by acetylene reduction, “N feeding and “N dilution techniques. Biol. Fert.
Soils. 2, 7-14.
OKON, Y. 1984 Response of cereal and forage grasses to inoculation witb N,-fixing
bacteira. In : Veeger C. and Newton W.E. (eds.) Advances in nitrogen fixation
research, pp. 303-309. NijhoffBung, the Hague.
172
OKON, Y. and HADAR, Y.A. 1987 Microbial inoculants as trop-yield enhancers. CRC
Critical Reviews in Biotechnology. 6, 61-85.
PHILLIPS, D.A., JOSEPH, C.M. and MAXWELL, C.A. 1988 Flavonoid nodulation signais
in alfalfa. In : Abstracts 7th Congress on nitrogen fixation, 13-20 March 1988, Cologne.
PRIN, Y. and ROUGlER, M. 1987 Preinfection events in the establishment of Alnus-
Frankt% symbiosis : study of the root hair deformation step. Plant Physiol (Life Sci.
Adv.). 6, 99-106.
PRIN, Y., NEYRA, M. and DIEM, H.G. 1989 Evaluation of Frankia growth and viavility
using Bradford protein and INT reduction tests. Plant and Soi1 (in press).
REDMOND, J.W., BATLEY, M., DJORDJEVIC, M.A., lNNES, R.W., KUEMPEL, P.L.
and ROLFE, B.G. 1986 Flavones induce expression of nodulation genes in Rhizobium.
Nature. 323, 632-635.
ROLFE, B.R., REDMOND, J.W., BATLEY, M. and DJORDEJEVIC, M.A. 1987 Nodu-
lation inducing factors. European patent application no 87302621.5, date of
filing 26.03.87.
SCO’IT, C.D. 1987 Immobihzed cells : a review literature. Enzyme Microb. Technol. 9,
66-73.
SOUGOUFARA, B., DIEM, H.G. and DOMMERGUES, Y.R. 1989 Response of field-
grown Casuarina equisetifolia to inoculation with Frankia str’ain ORS 021001 entrap
ped in alginate beads. Plant and Soil (in press).
WIJFFELMAN, C.A., SPAINK, H.P., OKKER, R.J.K., VAN BRUSSEL, A.A.N., ZAAT,
S.A.T. and LUGTENBERG, EJ.J. 1987 Process for activating rhizobium nodulation
promoters European patent application no 87200348.8. date of filing 26.02.87.
WILLIAMS, P.M. 1984 Current use of legume inoculant technology. In : Alexander,
M. (ed.) BiologicaI nitrogen fixation, pp. 173-200. Plenum Press, New York.
ECONOMIE DE L’AZOTE
DANS LES SYSTEMES
DE CULTURE EN ASSOCIXTION
ET EN AGROFORESTERIE
h4aximisez Ia FBA pour la Production Agricole et Forest&e en Afrique
Biological nitrogen fixation
and nitrogen transfer in multiple
cropping in tropical Africa
MIJLONGOY, K. and EHUI, S.K.
International Institute of Tropical Agriculture,
PMB 5320, Ibadan, Nigeria
Abstract
Traditional farming systems throughout the humid and subhumid tropics are cha-
racterized by multiple cropping, the association of crops .in time and space on a
piece of land during one calendar year. These systems usually involve leguminous
crops and trees. Legumes can harness atmospheric nitrogen (NJ into the systems
and benefrt non-leguminous crops through the transfer of nitrogen (N) in the root
systems anddecomposition of their residues. Biologicalmechanisms such as competition
for light, water and nutrients that determine the yield levels for the individual crops
can affect the proee~~es involved in N, fiiation. Legumes can improve soil N status
with resulting N and grain yield benefits to the succeeding nonleguminous crops.
Even in the absence of net N benefit, intercropped nonlegumes have been shown
to derive important fractions of their N from N, fmed by leguminous component(s).
Only few studies have been published on economic profits of using legumes in cropping
systems in the tropical Africa. Although they show interesting returns relative to
the use of chemical fertilizers, they usually do not consider risk factors. A quadratic
programming approach is thus proposed as a more appropriate method to appraise
the economic potentials of legumes in multiple cropping in sub-saharan Africa.
175
Introduction
Traditional agricultnre in tropical Africa is based on shifting cultivation and related
bush fallow systems. The restorative success of these systems depends largely on the
litter cover from the fallow vegetation and on the nutrient recycling function of the deep-
rooted trees and shrubs (Sanchez,, 1976). In recent years, however, due to increasing po-
pulations and/or the scarcity of labor in areas with low population densities, fallow periods
have been systematically reduced or suppressed, leading to more intensive cultivation of
marginal lands. Continuous cultivation of the same fields has resulted in soil erosion,
increased weed populations, and decreasing soil fertility and crop yields (Rang et al.,
1985 ; Lal, 1987).
Earlier research approaches towards alleviating the food deficit problem in tropical
Africa were based on the Green Revolution which was associated with high-yielding va-
rieties, sole cropping, monocuhure and enormous inputs of non-renewable resources to
extend the cropping season, increase soil productivity and control pests. The failure of
the Green Revolution, some fifteen years ago, led udevelopers, to revise their approach
for the promotion of food production in the tropics. They began to emphasize the needs
for crop production stability, resource conservation, reduction of external inputs and more
complete use of natural resources. This tendency has stimulated a lot of research effort
in the area of multiple cropping aiming at defining plant populations of the component
crops, planting patterns, plant structure, relative duration of the life cycles of the component
crops, and fertihzer managements that would result in combined yield advantages (Okigbo
and Greenland, 1976).
Legumes are usually included in multiple cropping systems for three basic reasons.
Low-growing legumes can be integrated into the architecture of tall cereals without exces-
sive reduction in yields of component crops, but with significant increases in protein and
crop yields per hectare (Norman, 1982). Legume-cereal mixtures are advantageous because
legumes can fm atmospheric N, symbiotically and transfer this additional N to other crops
in the system. Finally, farmers are risk averters who base their farming decisions not
on profit maximization but on the judicious choice and arrangement in time and space
of various crops to suit their microecologies and conditions and ensure crop production
with minimum risk.
There is however a dearth of information about how much N is fmed ; how this
process is affected by the various plant interactions such as shading in the cropping sys-
tem ; how leguminous plants and, in general, N,-furing systems contribute to the yield
advantages and the maintenance of soil fertility in multiple cropping. Economic evalua-
tion of Nz fixation and N transfer is also lacking partly because of the poor knowledge
of complex N transfer pmcesses in multiple cropping systems. Although, a great deal
of research work on legumes in multiple cropping has been conducted in tropical Asia,
comparatively little work has been conducted on this topic in tropical Africa.
In this paper, we briefly review the major ecological factors determining yield
levels in multiple cropping systems and their influence on biological N, fixation. We
then discuss N contribution of legumes in selected types of multiple cropping. Our dis-
cussion does not include associative N, fixation and pasture systems to simplify the presen-
tation of the paper. However, principles described for other systems are also valid for
these systems. Finally, we briefly review the economics of using legumes in multiple
cropping and consider the quadratic programming approach that includes the risk mini-
176
mixing attitude of farmers as a possible tool for economic assessment of the potential
of legumes in sub-saharau Africa.
Some definitions
Before addressing these issues, it is useful to define certain concepts. Multiple
cropping is the general term which is used to describe systems where two or more crops
are grown on the same field in a year. The system is said sequential when the crops
am grown in sequence. In this case, the intensification of crop production is only in the
time dimension. Sequential cropping is rare in traditional afiican farming systems. It charac-
terizes areas where on-farm agricultural technology is geared towards commercial pro-
duction (Andrews and Kassam, 1976).
Intercropping, the growing of two or more crops simultaneously on the same field,
is widespread in tropical Africa with the highest complexity in compound gardens of
the rainforest region where are intercropped annual staples, vegetables and perennial fruit
trees (Okigbo and Greenland, 1976). Mixed intercropping, the simultaneous growing of
two or more crops without distinct TOW arrangement, and relay intercropping, the growing
of two or more crops simultaneously during part of the life cycle of each component,
are the predominant practices in the tropics. In row intercropping, two or more crops
an3 grown simultaneously with one or more crops planted in rows. Strip intercropping
is the growing of two or more crops simultaneously in strips. Both row and strips intercrop-
ping are rare in Africa except where animals or tractors are used in cultivation (Okigbo
and Greenland, 1976 ; Steiner, 1982). In the live-mulch system, food crops are grown
in an establish field of a low-growing perennial leguminous cover crop (Akobundu, 1980).
The live mulch prevents soil degradation and erosion, suppresses weeds and provides
biologically fixed Nz
Trees such as oilpalm, coconut, cocoa are also often intercropped to a range of
annual crops in the early years of plantation (Okigbo and Greenland 1976). This mul-
tistorey cropping is considered as the ideal farmiug system in the humid tropics as it
resembles the natural vegetation. Related to the multistorey cropping system is alley crop
ping, the growing of food crops between hedgerows of woody species that are perio-
dically pruned to prevent shading of the associated crops and to make available mulch
materials and the nutrients locked up in the prunings (Rang et uL, 1981).
Multiple cropping is generally more complex in the forest regions than in the sa-
vanna areas. In the forest region, crops are more diverse, farmers’ holdings arc relatively
small and the growing season is longer (Okigbo and Greenland, 1976). Within the same
ecological zone, the diversity in component crops and their spatial arrangement increases
with soil fertility which in turn improves with the proximity to the homestead (e.g. corn-
pound gardens).
Factors determining yield levels in multiple cropping systems
The major factors determining yield levels in multiple cropping systems are the
interactions of component crops during the use of growth factors, the N,-fxing ability
of intercrops and stresses interfering with biological N,-fixation. We discuss each of these
processes subsequently.
177
Interactions of component crops
When plants are intercropped or grown in sequence, they interact through various
forms of modifications imposed on the environment. These interactions and their effects
on crop yields were reviewed by Trenbath (1976). Willey (1979a) and Palaniappan (198.3.
The interactions are said non-competitive when the associated crops share growth factors
that are present in adequate quantities, such that some positive level of output from one
crop is possrble without any reduction in the output of another crop. Competition occurs
when component crops share a limited pool of growth resources. In this case, one output
must be forgone in order to produce more of the other output.
In an intercropping situation, ‘when the component species are able to exploit the
supply of growth factors in different ways in time or space, or if one species is able
to help the other in the supply of a factor, the interaction is said to be complementary.
The pmcess by which growth factors are used is referred to as annidation. Other synonyms
found in the literature include facilitation, cooperation aud compensation. Another form
of interaction observed among plants grown in intercropping is the influence of one com-
ponent crop on pests that may affect the performance of other crops in the system.
Plant growth factors that are likely to play important role in multiple cropping
are light, water and mineral nutrients. Light is absorbed by the leaves, and water and
nutrients mainly by the roots. In healthy young plants, growth is responsive to the rate
of absorption of any one of the growth factors when it is in relatively short supply (Tren-
bath, 1976). Competition for light occurs when the photosynthetic canopy of one com-
ponent is set higher than that of another. Thus taller components compete more effectively
for light. But leaf area and leaf inclination are likely to determine the magnitude of com-
petition. If shading is partial, the shaded plants can adapt to low light conditions by re-
ducing their rate of dark respiration, lowering their root/shoot ratio, or increasing the
leaf area/leaf weight ratio. These adaptative changes can be part of selection traits in
breeding genotypes suitable for multiple cropping. Intense shading will reduce growth
considerably and induce early senescence of leaves. Light is different from other growth
factors as it cannot be much influenced by man, Therefore, by maximixing light inter-
ception, multiple cropping can suppress weeds and reduce evapotranspiration losses from
the soil surface effectively.
The uptake of water, dissolved nutrients and oxygen (03 by a root surface esta-
blishes a concentration gradient in the vicinity of the root The zone of depletion of water
and mobile ions such as nitrates will expand rapidly as compared to that of P and cations
like ammonium, calcium and potassium that are strongly adsorbed onto the surfaces of
soil particles. Thus in multiple cropping particularly in intercropping, a component crop
can induce water and nutrient stress for the other crops. Experiments on competition for
0, have not clearly demonstrated the magnitude and consequences of this interaction.
Competition for carbon dioxide (CC2) has been demonstrated in sealed containers, but
it is unlikely to occur in the fields owing to the turbulence of air within canopies (l&n-
bath, 1976).
Yield advanfages of nitrogen-fixing intercrops
Nitrogen-fixing intercrops affect crop yields through annidation in space, in time
and with respect to nitrogen. Spatial differences between crops are related to shoot height,
plant architecture, and root depth and distribution. Willey and Rao (1981) demonstrated
178
the importance of spatial interactions in intercropping by comparing the performance of
17 genotypes of pigeon pea in association with sorghum. None of the genotypes affected
sorghum yields adversely but their own yields ranged between 36 and 73% of sole crop
yields. Genotypes that were compact in the early growth stage and that spread later to
fully utilize the resources after sorghum harvest minimized competition with sorghum
significantly. They were found to be most suitable for high combined yields.
Another example of annidation in space is the multistorey cropping involving
planting leguminous shade trees or low-growing cover crops such as Pueraria or Centrcy
sema in cocoa and tea plantations. Trenbath (1976) noted that where the taller intercrop-
ped component has steeply inclined leaves and the shorter component has prostrate leaves,
there is a possibility of a land equivalent ratio (LER) > 1. In alley cropping, FIemingia
congesta is one of the few nitrogen-fixing shrubs that perform well under the shade of
plantain. Its N-rich pruning can improve plantain (mtemational Institute of Tropical Agri-
culture, 1985) and maize gram (Table 1) yields . The principle of exploiting different
vertical space layers is also found in live-mulch system. However, cover crop species
currently used in the system such as Psophocarpus palumis and Centrosema pubescent
tend to climb and strangle the associated crop if they are not sprayed with growth retardants.
-
Table 1.
Effect of prunings from Cassia siamea, Flemingia congesta,
Gliricidia sepium and Leucaena leucocephala on maize' N
content and grain yield (Mulongoy, unpublished data).
Pruning N
Maize grain
Pruning
applied*
yield
Maize N
-------------------kg ha-' ------------______-__________
Cassia
168
16.7
2430
Flemingia
90
13.9
T540
Gliricidia
180
21.8
-
2870
Control
0
8.0
-
1100
1 On the basis of 40000 plants ha-'. Nitrogen content was determined at 8
weeks after planting.
2 Prunings were applied only at maize planting
179
Some degree of root stratification must be common in intercrops of dissimilar com-
ponents on deep soils (‘frenbath, 1976). But only few reports have been published on
this topic. In alley cropping, Kang et al. (1985) showed that both maize and leucaena
had distinct root feeding zones with leucaena extracting moisture and nutrients from deeper
layers. However, hedgerow trees established from cuttings may have shallow nx)t systems
(AttaXrah et al. 1985) and thus compete with the associated crops.
Two intercrops of widely different maturities will likely reach their peak demands
for growth factors at different times. This dissimilarity will reduce interference between
them. A minimum maturity difference of 30-40 days for yield advantage has been sug-
gested (Palaniappan, 1985). After the early crop is harvested, the other crop will fully
exploit the growth factors. Yield advantages in relay cropping are partly based on this
principle. Another temporal effect occurs in situations where senescent plant materials,
litter or prunings could make nutrients available to another crop. Norman (1982) notes
for instance that up to 70% of the aboveground materials of guar (Cyamopsis tetrago-
doba) returns naturally to soil as leaf fall. Charreau and Vidal (1965) also noted the
leaf fall of Acacia ulbiuir and its contribution to soil N and N uptake of cereals planted
around it.
In an intercropping system involving Njfiing and non-N,-fixing systems, the
N,-freer will obtain part of its N from biological fixation of atmospheric N,, while the
other component(s) will utiIize exclusively soil mineral N. If the non-N,f=ing component
is very competitive for soil N, a depleted soil in N will favour nodulation and nitrogen
fixation of the N,-fixer intercrop. Pigeon pea seems to nodulate better where the roots
intermingle with those of intercropped sorghum (Willey, 1979a). Ennik cited by Trenbath
(1976) reports that white clover intercropped with grass on N-poor soil gives a LER up
to 6.7 due to the interactions between the component crops. Some investigators explain
the positive effect of added legumes on LER by some transfer of the N, fixed to the
non-leguminous component (Trenbath, 1976). But more data are needed to corroborate
this explanation.
interference stresses on biological nitrogen fixation
Various stresses can impair biological N, fixation in multiple cropping systems.
Biological N, fmation is markedly dependent on current photosynthesis regulated by light
This is shown by diurnal variations in nitrogenase activity (Sloger et al., 1975) and by
shading effect on NZ fixation. Gibson (1976) reported that nitrogenase of Trifolium sub-
terraneum last 40% of its activity within five hours when the plants were transferred
from an ilhnninance of 620 ttBrn-%-i to one of 165 l.t.Brn-W. Nutman (1976) reported
also that 50% shading of soybean plants at the end of flowering decreased fixation from
125 to 91 kg N ha-‘. He concluded in his review that there is a direct relationship between
light intensity and N, fixation through the amount of photosynthate produced. Similarly,
growth and CO, uptake of intercropped azolla declines in response to a developing rice
canopy (Lumpkin and Plucknett, 1982). Photosynthesis by leaves of legumes seems satu-
rated at 30-50% of maximum sunlight (Ludlow cited by Henzell, 1981). As stated earlier,
plants exhibit various phenotypical adaptations to low light to ensure maximum light inter-
ception.
The detrimental effect of water stress on the host-plant, the rhizobia and symbiotic
effectiveness is well documented (Sprent, 1971, 1972 ; Eaglesham and Ayanaba.. 1984).
180
Fthizobial populations are reduced in size when soils become desiccated. Slow-growing
rhizobia tend to survive desiccation better than fast-growing rhizobia Soil type influ-
ences the desiccation tolerance of rhizobia. Nodulation tends to fail at low soil moisture
levels not only as a result of death of the rhizobia, but essentially because root hairs
assume a configuration incompatible with the infection process (Eagle&m and Ayanaba,
1984). The water potential levels allowing nodulation of legumes should be defined.
Sprent (1971) noted that soil moisture loss below 80% of the maximum resulted in irre-
versible damage of soybean nodules. Some legumes such as clovers can recover more
rapidly from drought than others like soybeans that shed nodules under dry conditions.
Water stress can also reduce photosynthesis by reducing leaf area and inducing leaf
fall ; this in turn will reduce symbiotic N, fixation.
It is accepted that N,-fixing plants require more P than plants supplied with mineral
N. Phosphorus deficiency will limit nodule development, nitrogenase acrivity and host-
plant growth. A number of experiments have demonstrated the important role of vesicular-
arbuscular mycorrhiza in increasing P uptake of infected plants. Also, dual inoculation
with effective strains of rhizobia and mycorrhizal fungi shows synergistic effects on no-
dulation and reduction of atmospheric N, in Pdefrcient soils. Recent studies reveal ge-
netic diversities within plant species in tolerance to low soil P (Sanginga et al.. 1990).
The effect of K on the Rhizobium - lugume symbiosis seems to be indirect and
to act through the physiology of the plant (Jardim Freire, 1984). There are however reports
showing nodulation response to K under field conditions (Ayanaba and Dart, 1977).
Danso et al. (1987) observed a reduction in dry matter yield, total N and N2 fixed
by fababeans intercropped with barley. Mulongoy (1986a) found a similar reduction in
dry matter of intercropped cowpea and maize that resulted in reduced total plant N. With
increasing plant population densities of either fababeans in the sole crop or barley in
the intercrop, Danso et al. (1987) obeserved increases in the proportion of N derived
by fababeans from fixation. Greater Nz fixation may either increase the total N yield
in the legume or just the proportion of N from fixation as compared to N derived from
mineral sources by the legume. Danso et aZ. (1987) found in their study that the presence
of barley decreased soil and fertilizer N uptake by fababeans and enhanced symbiotic
N, fixation of the legume component. The mechanisms governing these processes were
not described. However, since the reduced soil N uptake by the legume was not reflec-
ted in increased soil or total N in barley, there was therefore no evidence of N-sparing
effect by the legume for the associated non legume. Data presented by Danso et al. (1987)
do not support the hypothesis of N transfer by the legume to maize.
Mulongoy and Nkwiine (1987) observed that populations of rhizobia decreased
when maize was included in a sequence of soybean monoculture (Table 2). Whether the
reduction in rhizobia populations was due to allelopathic effects was not investigated.
Reduction in the size of rhizobial population can result in poor nodulation and nitrogen
fixation. It is important to note that poorly or not nodulated legumes can deplete soil
N just like cereals. Mulongoy and Nkwiine showed in the same experiments that crop
ping cowpea or soybean after maize increased the population of Bdyrhizobizm jupo-
nicum. It is important to note that poorly or not nodulated legumes can deplete soil N
just like cereals.
From this discussion, it is evident that various types of interferences occurring
in multiple cropping can reduce the number of rhizobia and affect nodulation and nitro-
gen fixation adversely. Few investigations have been carried out to monitor changes in
181
these factors and how they affect biological nitrogen fixation of leguminous components
in multiple cropping.
Table 2.
Effect of cropping sequences of soybean(s) and maize on the
lo910 of the numbers of soybean rhizobia per gram of soil
recorded after the last cropping season, on a loamy sand.
(Mulongoy and Nkwiine, 1986a)
Cropping sequence
Bradyrhizobium japonicum
(log10 number)
Soybean - Soybean - Soybean
3.710 a
Soybean - Soybean - Maize
3.492 b
Soybean - Maize
- Soybean
3.710 a
Soybean - Maize
- Maize
3.118 c
Nitrogen contribution of legumes in multiple cropping
Nitrogen contribution of legumes and, in general, of N,-tixing systems in multiple
cropping can occur through (1) exudation or excretion of nitrogenous substances by active
nodules and roots, (2) leaching of soluble N from living leaves and plant parts during
rainfalls, (3) litter fall and decomposition, (4) root and nodule decay, (5) decomposition
of crop residues and prunings, and (6) animal manure. All these modes of N transfer
are possible in an intercropping situation, but only residual effects will be involved in
sequential cropping. In this section, we examine successively N contribution in legume-
non-legume intercropping, livemulch systems, alley cropping and sequential farming.
Nitrogen contribution in legume-nonlegume intercropping
Various reports give evidence that the legumes benefit the associated crops in
intercropping systems (Evans, 1960 ; Ckiru and Willey, 1972). But Trenbath (1976) com-
piled data (607 values) from various sources and showed that the effect of legumes on
LER was not necessarily better than adding nonleguminous intercrops. He concluded that
the main advantage from the legume~components was in the saving of N-fertilizer. He
also noted that while competition for N would be miuimized in mixtures involving legumes,
competition for other growth factors could occur.
Early scientific reports on active N tranfer from the root nodules to associated
crops were published in the 193Os, in pot experiments (Willey, 1979a). It was noted that
shading favours N excretion from the legumes. This may however be significant only
when the legumes have already fmed N,, as shading will reduce N, fixation. Defoliation
can also promote N release in the root system. Whitney and Kanehim (1967) found N
182
contribution of 0.1 to 0.6 kg ha-’ from the root system of defoliated Desmodilun inrortum
to pangola grass.
Although direct evidence of active N excretion from the root nodules of-legumes
is lacking, other data suggesting N transfer from the legumes include the lack of response
of cereals to fertiliser N when they are intercropped with legumes (Agboola and Fayemi,
1972 ; Remison, 1978), higher N uptake of the cereals in intercropping than when they
are grown as sole crops (De, 1980 ; Eaglesham et al., 1981) and ‘%I studies (Eaglesham,
1982 ; Patra et al., 1986). Short cycle legumes seem to contribute more nitrogen to the
associate4l crops than late maturing species. Maize derived more benefits from Vigmz i&a
than it did from cowpea (Agboola and Fayemi, 1972), and from L.uthyrus sp. and early
maturing cowpea or groundnut than it did from Phuseoh vulgaris (Waghmare and Singh,
1984). This suggests that N contributions might occur mostly through senescence and
decomposition of the short cycle legumes.
Nitrogen transfer from legumes to nonlegumes in intercropping is not always trans-
lated in N (Fig. 1) or grain’yield (Danso et al., 1987) advantages. Nitrogen-15 studies
provide the most reliable data on the occurence and magnitude of N transfer in intercrop-
ping involving legumes. Eaglesham (1982) found that intercropped maize had access to
a source of N not available to sole maize crop. He presented 15N data consistent with
N transfer from cowpea to the intercropped maize. In experiments conducted by Patra
ef al. (1986), 28% of maize N representing 21.2 kg N ha’ was obtained from the atmos-
phere through transfer of fmed N by cowpea grown in association with maize. They con-
firmed their data in pots with either maize or wheat intercropped with cop
N content. kg ha-l
Cbwpoa (LSD 5% - 6.4)
I-
Maize (LSD 5% - 4.2)
Figure 1: Response of wwpea and ma&N
content to intezcmpping. Cowpea v&&s
are average for 6 wwpea cultivi3rs ; for
maize, average of maize intezcropped with
I-
6 wwpea cultivars (Adapted fium Ma&m-
gay. 1986a).
sole cropping
lnb3rcropping
183
Nitrogen contribution in livemulch systems
Live-mulch is a crop production system in which food crops are planted into a
i’
low-growing cover crop with minimum soil disturbance (Akobundu, 1980). The cover
crop smothers weeds and protects the soil. Akobundu (1980) observed that after five sea-
sons of continuous cropping, maize did not respond to fertihzer N in live-mulch plots
of Psophocarpus pahris and Centrosema pubescens whereas maize yields in bare plots
under minimum or conventional tillage increased when N fertilixer (60 kg ha-‘) was given.
Mulongoy (1986b) compared N contributions to maize of newly established and 3:years
old live-mulches of Psophocarpus, Centrosema and Arachis repens. In the newly esta-
blished field, cover crops and maize competed for N to the disadvantage of maize. In
the longteim plot, P. p&.mis was the best live-mulch with N contributions averaging
30.7 kg ha-‘.
In another experiment, nitrogen uptake by maize was monitored for five cropping
seasons in a live-mulch of P. palustris. During the first four cropping seasons, P. paLlcstrii
and maize competed for soil N in the live-mulch, and maize grown in the live-mulch
contained less N than maize grown by itself. In the fifth season, the live-mulch contributed
15 kg N ha-l to maize in the plots that were cropped continuously for five seasons. Maise
grown in a psopho livemulch that had aheady been established for four cropping seasons
contained 36 kg ha-’ more N than maize grown by itself. This positive N contribution
came, at least in par& from increases in microbii biomass and in organic matter derived
from the cover crop in the form of litter and earthworm casts (Mulongoy, 1986b).
The live-mulch system, however, needs refinement because, at present, a net contri-
bution of N and an increase in food crop yield occur only after the cover crop has been
established for some cropping seasons. This lag in benefits has to be shortened if the
live-mulch system is to be accepted by farmers. One way to shorten the lag is to minimize
competition for soil N between the cover crop and the food crop during live-mulch esta-
blishement. Also, if the pathways of N transfer from the live-mulch to the food crop
were understood, it might be possible to manage the system so as to increase the amount
of N contributed by the legume.
Simpson (1976) and Ladd er al. (1981) also found little N transfer from low-growing
legumes to associated crops, in pastures. The main N value from the legume seems to
be the building-up of soil organic N for adequate N supply to subsequent crops. Milller
and Sundman (1988) who monitored the fate of iw released during decomposition of
leguminous plant materials buried in the field found that 2746% of input N was re-
tained in the soil. A succeeding crop of barley took up 6-254s of input.
Nitrogen contribution in alley cropping
In the alley cropping, leguminous trees are preferred to monleguminous trees for
the hedgerows. They can bring atmospheric N, into the soil system. Well-nodulated trees
or shrubs can yield more than 500 kg N ha-l in a year (Guevarra, 1976) owing to their
high N,-fixing ability. Estimation of N, fixation of trees is difficult. But current data
indicate that nodulated trees can fix symbiotically more than 50% of their N (Roskoski,
1981 ; Hogberg and Kvarnstrom, 1982 ;~Rinaudo ef al., 1982 ; Sanginga et al., 1986;
MuIongoy et al., 1988 ; Ndoye and Dreyfus, 1988). The mojor proportion of hedgerow
tree N is found in the prunings consisting of leaves, twigs and succulent plant parts inclu-
ding immature pods. Application of legume prunings increases N and grain yields in the
184
associated crops uable 1 ; Kang, 1987). However, N contribution of the prunings to
the associated crop may not always be very efficient. Kang (1987) and Mulongoy and
Van der Meersch (1988) reported it in the range of 10 and 40% of the input N, and
Kang et al. (1985) found that up to 80 kg fertilizer N was needed for optimum maize
grain yields. Some results suggest that part of the N from prunings is retained in the
soil organic-N pool (Mulongoy and Van der Meersch, 1988). In a hydromorphic soil crop
ped to rice, N contribution of Sesbmiu rostrutu was estimated equivalent to 40 kg N
ha-’ representing about 30% of the N applied in the prunings (Mulongoy, 1986c). Mana-
gement practices such as incorporation of prunings in the soil can increase the efficiency
with which N is transferred from prunings to the associated crops (Read et al., 1985 ;
Wilson et d., 1986).
Nitrogen contribution Corn belowground parts can be considerable. Sangiga et al.
(1986, 1988) estimated it at about 30% of total N contribution of leucaena to maize
in a rotation experiment. Nitrogen contribution from roots of hedgerow trees needs to
be investigated.
Nitrogen contribution in sequential cropping
The residual effect of leguminous crops in a sequence will greatly depend on
several fixton including the legume N at harvest time, N harvest index, management
of the residues (Whether retained on the soil surface, incorporated in the soil or partly
removed for domestic use or for feeding animals) and on the time interval between crops
in the sequence. Kang (1987) discussed some examples of N returned to soil in crop
residues of pigeon pea, cowpea and soybean. The following generalizations can be
drawn : (1) legume N yield i.e. the sum of atmospheric nitrogen fixed and N absorbed
from soil, is high, varying in these examples between 73 and 205 kg N ha-* ;
(2) N in fallen leaves and root system is usually low ; but it can be as high as
40 kg N ha-’ in well nodulated shrubs at harvest ; (3) N removed in crop residues, pods
and seeds can be relatively low ; it ranged in these examples between 26 and
85 kg N ha-’ ; (4) N balance can be very low even negative indicating that the legume
can deplete soil N particularly when N, fixation is low and/or N harvest index is high.
The N retmned to the soil in the residues will then be mineralized before it is
used by the succeeding crops. Plowing the residues under may reduce the probability
of N loss by volatibzation and erosion, and may increase N contribution to the succee-
ding crops.
Kang (1987) reports a number of experimental data showing that the legumes im-
proved soil N status, with resulting N and grain yield benefits to the succeeding non
legtmrinous crops. There were differences among plant species in the level of N contri-
bution. Pnzeding non-leguminous crops with a legume increased their yields more than
intercropping. Some investigators however reported no or negative N contribution from
legumes in sequential cropping. A good knowledge of the N cycle in the system will
allow to devise practices that can maximize the resources for low-input agriculture
suitable for the smallholder farmers.
185
Methods for evaluating the economic potential of legumes In multiple cropping
systems
The purpose of this section is to examine methods by which to appraise the eco-
nomic potential of legumes in multiple cropping systems. The basic question is what
are the best systems (with or without legumes) capable of raising or maintaining the level
of soil fertility while satisfying the farmer’s goals and resource limitations. The need
to answer this question is pressing as results from past agronomic studies show that
multiple cropping systems (especially when legumes are associated with non
leguminous food crops) are much more efficient than sole cropping (Willey, 1979a and
1979b ; Steiner, 1982 ; Horwith, 1985 ; Spencer, 1985). Leguminous crops have been
shown to improve soil fertility (through nitrogen fmation and transfer) thus contributing
to yield stability (risk reduction) over different seasons and higher yields in a given season.
Few economic studies are available which evaluate the role of leguminous crops
in multiple cropping systems. Mittal er aI. (1985) conducted a study to determine yield,
total production and net returns by intercropping legumes (black gram, green gram, cowpea
and groundnut) with maize. Their results show that maize in- with groundnuts
is the most profitable and suitable system for rainfed conditions in Shiwalik region, India.
Singh and Chand (1980) also studied the income effect of intercropping grain legumes
with maize. Their results show that maize intercropped with soybean is the most remu-
nerative system with 120 kg N ha-l.
Although these studies show interesting results, they are limited for a number of
reasons. First, they were conducted essentially in tropical Asia and therefore their results
may not be applicable to sub-saharan Africa where the rate of demographic change is
higher and the physical conditions which determine technical potential are more difficult
(Matlon and Spencer, 1984). Most of these studies use the partial budgeting approach
in which the market value of land and labor are unknown or included in an ad-hoc manner.
Yet the real economic prices 0.e. shadow prices) of these resources are essential for eva-
luating the profitability of a cropping system and how for example labor and different
land types are allocated in the cropping seasons and among crops. Partial budgeting ap
preach ignores the substitutability of inputs and how they are allocated based on fixed
endowments and the implicit prices of the resources. Finally, these studies do not consider
risk factors. Farmers in sub-sahamn Africa are risk-averse. They hedge against risk by
diversifying into as many crops as possible through multiple cropping. Crop yields and
prices vary from one growing season to another thus affecting the expected income of
the farmer. If this is true, then an economic evaluation of multiple cropping systems using
riskless models is inappropriate.
For appraising the economic potential of legumes in multiple cropping, a moderne
theory based on utility maxim&ration and risk programming is proposed. This is a much
more appropriate model to describe farmers behaviour in sub-sahamn Africa than the standard
profit maximization models. Farmers practice diversified farming not only to take ad-
vantage of the complementary nature of the various crops but also to minimize risk in
aggregate farm income given a level of expected income. Quadratic programming which
is based on the expected-income variance criterion is a practical way of selecting income-
risk efficient plans. It assumes that the farmer’s preference among alternative farm plans
is based solely on expected net farm income, E and variance, V (Freund, 1956). Quadratic
programming model was first applied to agriculture by Freund (1956) and has since been
186
recommended for analysis of uncertainties in gross margins, gross returns and net variable
costs in farm planning (Heraedy and Candler, 1958).
The problem of resource allocation under risk can be mathematically formu-
ItiaS:
h&x EV(X) = RX - (aLI)XQX
Subject to AX-3, where X is an (n x 1) vector of activities (e.g. hectares planted to
maiz&mpea ; sole maize, maiz4cassava/leucaena etc.)
R
=
(n x 1) vector of expected net returns associated with the vector of acti-
vities
t
=
trampos of R and X
A =
(n x n) matrix of technical coefficients (e.g. labor, rotational or land
requirement etc.)
B
=
(n x 1) vector of resource constraint level (e.g. available labor or capital,
or land etc.)
=
is the risk averse coefficient
;
=
(n x n) matrix of tbe variance-covariance matrix of gross returns.
The special feature of the quadratic programming approach is the variauce cova-
riance component of the objective function, EV. When the risk aversion coefficient cc
is zero, the problem collapses to a standard linear programming problem indicating that
farmers are risk-neutral. A positive (negative) value of a indicates the farmer is risk averse
(lover). Although there is no clear method by which to estimate a the latter is usually
obtained by eliciting farmers’ preferences. Application of the quadratic programming ap-
proach also mquires that expected returns and variance and covariauce of returns for all
enterprises be known. ~Covariances am fundamental for efficient diversification among
farm enterprises as a means of hedging against risk (Heady, 1952). For example combi-
nation of activities that have negative covariate gross margins will usually have a more
stable aggregate return that the return from more speciabzed strategies. A case in print
in sub-saharan Africa is the intercropping versus the monocropping system. Returns of
crops in an intercropping may covariate negatively thus providing in the aggregate a stable
return, compared to monocropping systems.
The quadratic programming approach assumes that the farmer is rational and thus
restricts his choice to those farm plans for which the associated income variances are
minimum for given expected income levels. This assumption is based on the premkes
that individuals like increases iu expected returns and dislike increases in variance of
returns. The variance and covariance of return can be obtained using standard statistical
packages or a culcu@ion using the variance formula
V = cj(Rj - r)2/n-l
where Rj denotes the gross margin of activity j, and r is the average return on activity
j over n cropping seasons. Other data needed depend on the problem specification and
are usually obtained by means of survey techniques, direct measurements and secondary
sources of information.
187
Conclusions
Multiple cropping is typical of agricultural practices in the tropics. Depending on
the level of agricultural technology, multiple cropping involves the growing of sole crops
in sequence or various mixtures of rain-fed crops. Intercropping is the predominant prac-
tice by small-holder farmers producing relatively low crop yields. It does however make
better use of the resources than sole cropping, and help minimize variability in returns.
Legumes are usually included in multiple cropping as sources of protein. This is
because they fit the temporal and spatial requirements of the system. It is indeed believed
that legumes competeonly little with other crops for N, and that they can reduce the
need for fertilizer N in the system since they can fix atmospheric N,
There is a dearth of information in multiple cropping’.on the dynamics of rhizobia
and other nitrogen-fixing microorganisms when they are living in the rhizosphere of non-
host plants, on how nodulation is affected by the various interactions observed between
component crops, and on how much N, legumes and other N,-fixing systems fn in mixtures
or in sequence with non-leguminous crops. The response of legumes and other crops
such as maize to crop mixture and high population densities have been widely studied
utilizing the aeplacemenb series (De Wit, 1965) or, most commonly, the u additive >
technique (Chang, 1981). The same techniques can be used to monitor the changes in
biological N, fijration parameters. Also, only little quantitative data is available on the
process of N cycling in multiple cropping. Nitrogen-15 studies will provide the most
accurate information about N, fixation and N transfer. They allow a distinction on one
hand between N derived from atmosphere, soil and fertilizer and, on the other hand, bet-
ween N transfer and competition (Pa&a et al., 1986).
Although, research results have demonstrated the N and grain yield advantages
of including legumes in multiple cropping, analyses are laking which determine the eco-
nomic potential of nitrogen fixation and transfer of legumes in multiple cropping systems
in Africa. In this paper, a method of economic analysis is highlighted which takes into
account the risk minimizing behaviour of farmers in sub-saharan Africa.
References
AGBGGLA, A.A. and FAYEMI, A.A. 1972 Fixation and excretion of nitrogen by tropical
legumes. Agron. J. 44, 409412.
AKOBUNDU, Lo.1980 Live mulch : a new approach to weed control and crop pro-
duction in the tropics. Proc. British Crop Protection CoyT - Wee&. pp. 377-382.
ANDREWS, D.T. and KASSAM, A.H. 1976 The importance of multiple cropping in
increasing world food supplies. In : Papendick RI., Sanchez, PA. and Triplett G.B.
(eds). Multiple Cropping, pp. l-10 ASA Spec. Publ. 27 Madison.
A-I-TA-KRAH, AN., SUMBERG, J.E. and REYNOLDS, L.1985 Leguminous fodder trees
in the farming system. An overview of research at the Humid Zone Programme of
lLCA in southwestern Nigeria. In : Haque, L., Jutzi, S. and Neate P.S.H.(eds.). Potentials
of Forage Legumes in Farming Systems of Sub-Saharan Africa, pp. 307-329 ILCA,
Addis Ababa, Ethiopia.
AYANABA, A. and DART, P.J. 1977 Biological Nitrogen Fixation in Farming Systems
of the Tropics. John Willey and Sons, Chichester.
188
CHANG, JE. 1981 An analysis of competition between intercropped cowpea and maize.
Ph.D. Thesis, lowa State University.
CHARREAU, C. and VIDAL. P. 1965 Influence de I’Acacia albida De1 sur le sol, nutrition
minhale et rendements des miIs Pennisetum au Senegal. Agronomie Tropicale. 20,
600-625.
DANSO, SKA, ZAPATA, F., HARDARSON, G., and FRIED, M. 1987 Nitrogen fixation
in fababeans as affected by plant population density in sole or intercropped systems
with barley. Soil Biol. Biochem. 19, 411415.
DE, R. 1980 Role of legumes in intercropping systems. In : IAEA. Nuclear Techniques
in the Development of Management Practices for Multiple Cropping Systems. pp.
73-84. IAEA, Vienna.
DE WIT, CT. 1%5 Competition among herbage plants. Netherlands JOIUM~ of Agricul-
tural Science. 13, 212221.
EAGLESHAM, ARJ., AYANABA A., RANGA, RAO, V. and ESKEW. D.L 1981 Impro-
ving the nitrogen nutrition of maize by intercropping with cowpea. Soil Biol. Biochem.
W, 169-171.
EAGLESHAM, A&J. 1982 Assessing the nitrogen contribution of cowpea (vigna ZUI-
guhlatu) in monoculture and intercropping. In : Graham, PH. and Harris, S.C. (eds).
Biological Nitrogen Fixation Technology for Tropical Agriculture. pp. 641-646. CIAT,
Cali; Colombia.
EAGLESHAM, ARJ. and AYANABA, A. 1984 Tropical stress ecology of rhizobii, root
nodulation and legume fixation. In : Subba Rao N.S. (ed.). Current developments
in biological nitrogen fixation pp. l-35. Oxford and IBH Publishing Co. New Delhi.
EVANS, AC. 1960 Studies of intercropping. I. Maize or sorghum with groundnuts. East
African Agricultural and Forestry Journal. -26, I-10.
FREUND, RJ. 1956 The introduction of risk into a programming model. Econometrica.
34, 253-263.
GIBSON, AH. 1976 Recovery and compensation by nodulated legumes to envrionmental
stress. In : Nutman, P.S. (ed.). Symbiotic nitrogen fixation in plants, pp. 385-403.
Cambridge University Press, Cambridge.
GUEVARRA, A.B. 1976 Management of L.eucaena leucocephala (Lam.) de Wit for maxi-
mum yield and nitrogen contribution to intercropped corn. PhD. Thesis, University
of Hawaii.
HEADY, E.O. 1952 Economics of agriculturaI production and ressource use. Englewood
cliffs, NJ : Prentice-Hall.
HENZELL, E.F. 1981 Forage legumes. In : Broughton WJ. (ed). Nitrogen fixation. Volume
1 : Ecology, pp. 264-289 Clarendon Press, Oxford.
HOGBERG, P. and KVARNSTROM, M. 1982 Nitrogen fu;ation by the woody legume
Leucaena leucocephala in Tanzania. Plant and Soil. 66. 21-28.
HORWlTI-I, B. 1985 A role for intercropping modem agriculture. Bioscience. 35, 286-
291.
INTERNATIONAL INSTlTUTE of TROPICAL AGRICULTURE. 1985 Annual Report
for 1984, pp. 182-183. Ibadan, Nigeria.
189
JARDIM F’REIRFL JR. 1984 Important limiting factors in soil for the RHZZOBZUM -
legume symbiosis. In : Alexander, M.. (ed.). Biological nitrogen fixation. Ecology,
technology, and physiology, pp. 51-74. Plenum Press. New York and London.
KANG, B.T., WILSON, GF. and SIPKENS, L. 1981 Alley cropping maize (Z&z mays
L.) and Leucaena leucocephulu &am) in southern Nigeria. Plant and Soil. 63, 165
1 7 9 .
KANG, B.T., GRIMMB, H. and LAWSON, T.L. 1985 Alley cropping sequentially crop-
ped maize and cowpea with leucaena on a sandy soil in southern Nigeria. Pht arid
Soil. 63, 165-179.
KANG, B.T. 1987 Nitrogen cycling in multiple cropping systems. Paper presented at the
International symposium on advances in nitrogen cycling in agricultural ecosystems.
Brisbane, Australia.
LADD, J.N., OADBS, J.M. and AMATO, M. 1981 Distribution and recovery of nitrogen
from legume residues decomposing in soils sown to wheat in the field. Soil Biol.
Biochem. 13, 251-256.
LAL, R. 1987 Managing the soils of sub-sahara Africa Science. 236, 1069-1076.
LUMPKIN, T.A. andPLUKNB’IT, D.L. 1982AzoZZu as a green manure : use and management
in crop production. Westview Press, Boulder,. Colorado.
MATLON, PJ. and SPENCER, D.S.C. 1984 Increasing food production in Sub-saharan
African : Environmental problems and inadequate technological solution. Amer. .Z.
of Agric. Economics. 66, 671-676.
MITTAL, SP., GREWAL, S.S., AGNIHOTRI, Y. and PRATAP SING 1985 Economics
of intercropping legumes in maize under rainfed condition. Intern. J. Trop. Agric.
3, 187-191.
MULLER, M.M. and SUNDMAN, V. 1988 The fate of nitrogen (%) released from dif-
ferent plant materials during decomposition under field conditions. Plant and Soil.
105, 133-139.
MULONGOY, K. 1986a Nitrogen fixation and N contribution of cowpeas in a maize-
cowpea intercropping system. Paper presented at the Second AABNP Conference,
Cairo, Egypt Dec. 15-19.
MULONGOY* K. 1986b Microbial biomass and maize nitrogen uptake under a Psopho-
carpus palustris live-mulch grown on a tropical Alfkol. Soil Biol. Biochem. 18,
395-398.
MULONGOY, K. 1986c Potential of Sesbuniu rostrutu (Brem.) as nitrogen source in
alley cropping systems. Biol. Agric. and Hortic. 3, 341-346.
MULONGOY, K. and NKWIINB, C. 1987 Effect of cropping on rhizobial pbpulation
and nodulation of soybean. In : BTA resource and crop management Program. Annual
Report for 1986, pp. 77-78. Ibadan, Nigeria.
MIJLONGOY, K. and VAN DER MBBRSCH, M.K. 1988 Nitrogen contribution by leu-
caena (Leucaena leucocephulu) prunings to maize in an alley cropping system. Biof.
Fert. Soils. 6, 282-285.
190
MlJTXlNGOY, K., OWOAJE, B.T., IBEWIRO, E.B. and OKOGUN, J.A. 1988 Nitro-
gen fixation and N contribution of selected trees in alley cropping system. In : Bothe,
de Bruijn and Newton @is.). Nitrogen fixation : hundred years after Gustav Fisher,
p. 824. Stuttgart, Germany.
NDOYE, I. and DREXFUS, B. 1988 N2 fmtion by Sesbania rosfruta and Sestmniu sesban
estimated using 15N and total N difference methods. Soil Biol. Biochem. 20, 2091
213.
NORMAN, MJ.T. 1982 A role for legumes in tropical agriculture. In : Graham, P.H.
and Hank, S.C. (eds.). Biological nitrogen fixation technology for tropical agriculture,
pp. 9-26, CIAT, Cali, Colombia.
NUTMAN, P.S. 1976 Symbiotic Nitrogen fnation in plants. Cambridge University Press,
Cambridge.
OKIGBO, BN. and GREENLAND, DJ. 1976 Intercropping systems in tropical Africa
In : Paper&k, RI., Sanchez, P.A. and Triple& BB. (eds.). Multiple cropping, pp.
63-101. ASA Spec. Publ. 27. Madison.
OSlRU, D.S.O. and WILLEY, R.V. 1972 Studies on mixtures of dwarf sorghum and
beans (Phuseolus vulgaris) with particular reference. to plant population. Journal of
Agricultural Science, Cambridge. 79, 531-540.
PALANIAPPAN, S.P. 1985 Cropping systems in the tropics : principles and management,
Wiley Eastern Ltd., New Delhi.
PATRA, D.C. SACHDER, M.S. and SUBBIAH, B.V. 1986 lsN studies on the transfer
of legume fmed nitrogen to associated cereals in intercropping system. Biol. Fertil.
Soils. 2, 165171.
READ, M.D., KANG, B.T. and WILSON, GF. 1985 Use of kucaena Zeucocephala (Lam.)
de Wit leaves as a nitrogen source for crop production. Fert. Res. 8, 107-116.
REMISON, S.U. 1978 Neighbour effects between maize and cowpea at various levels
of N and P. Erpl. Agric. 14, 205-211.
RlNAUDO, G., DREYFUS, B. and DOMMERGUES, Y.R. 1982 Intkence of Sesbania
rostrata green manure on the nitrogen content of rice crop and soil. Soil Biol. Biochem.
15, 111-113.
ROSKOSKI, JP. 1981 Nodulation and N,-fixation by Inga junicuil, a woody legume
in coffee plantations. Measurements of nodule biomass and field CA reduction rates.
Plant and Soil. 59, 201-206.
SANCHEZ, P. 1976 Properties and management of soils in the tropics. John Wiley &
Sons, New York, USA.
SANGINGA, N., MULGNGOY, K. and AYANABA, A. 1986 Inoculation of Leucuena
Zeucocephula (Lam.) de wit with Rhizobium and its nitrogen contriburion to a subsequent
maize crop. Biol. Agric. Hortic. 3, 347-352.
SANGINGA, N., MULONGOY, K. and AYANABA, A. 1988 Nitrogen contribution of
L.eucaenalRhizobium symbiosis to soil and subsequent maize crop. Plant and Soil.
112, 137-141.
SANGINGA, N., BOWEN, G.D. and DANSO, S.K.A. 1990 Irma-specific variation in
growth and N,-futation of L.eucaena leucocephala and Giiricidia sepium at low levels
of soil phosphorus. Plant and Soil. (In press).
191
SIMPSON, JR. 1976 Transfer of nitrogen from three pasture legumes under periodic
defoliation in a field environment. Aust. J. Exp. Agric. Anim. Hub. 16, 863-870.
SINGH, C.M. and CHAND, P. 1980 Note on economics of grain legume intercropping
and N fertilization in maize. Indian J. Agric. Res. 14, 62-64.
SLOGER, C., BEZDICEK, D., MILBERG, R. and BOONKERD. N. 1975 Seasonal and
diurnal variations in N, (C,H@ixi.ng activity in field soybeans. In : Stewart W.D.P.
(ed.). Nitrogen fixation in free-living micxorganisms. pp. 271-284. CambridgeUniversity
Press, London.
SPENCER, D.S.C. 1985 Aresearch strategy to develop appropriate agricuhural technologies
for small farm developement in sub-saharan Africa. In :--Ohm H. and Nagy. J. @is.).
Appropriate technologies for farmers in semi-arid West Africa;pp. 308-325. J. Intema-
tional Programs in Agriculture, Purdue University, West Lafayette, Indiana, USA.
SPRENT, J-1. 1971 The effects of water stress on nitrogen-fixing root nodules. I. Effects
on the physiology of detached soybean nodules. New Phytol. 70, 9-17.
SPRENT, J-1. 1972 The effects of water stress on nitrogen-fixing root nodules. IV. Effects
on whole plants of Vi& f&a and Glycine max. New Phytol. 71, 603611.
STEINER, KG. 1982 Intercropping in tropical smallholder agriculture with special reference
to West Africa. German Agency for Technical Cooperation. Germany.
TRENBATH, BR. 1976 Plant interactions in mixed crop communities. In : Papendick,
R.I., Sanchez, P.A. and Triple& G.B. (eds.). Multiple Cropping, pp. 129-169. ASA
Spec. Pub. Np 27. ASA, Madison, Wisconsin.
WAGHMARE, and SINGH. 1984 Sorghum-legume intercropping and the effect of nitrogen
fertilization. I. Yield and nitrogen uptake by crops. ExpZ. Agric. 20. 251-259.
WHITNEY, A.S. and KANEHIRO, Y. 1967 Pathways of nitrogen transfer in some tropical
legume - grass associations. Agron. J. 59, 585-588.
WILLEY,R.W. 1979aIntercropping - its importanceandreseamh needs. Part 1. Competition
and yield advantages. FieZd Crop Abstracts. 32, l-10.
WILLEY, R.W. 1979b Intercropping - its importance and research needs. Part 2. Agronomic
and research approaches. Field Crop Abstracts. 32, 73-85.
WILLEY, R. W. and RAO, M.R. 198 1 Genotype studies at ICRISAT. In : Proc. Int. Workshop
on Intercropping. 10-13 Jan. 1979. ICRISAT, Hyde&ad. pp. 117-127.
WILSON, G.F., KANG, B.T. and MULONGGY, K. 1986 Alley cropping: Trees as sources
of green-manure and mulch in the tropics. BioZ. Agric. Hort. 3. 251-267.
Maximiser la FBA pour la Production Agricole et Forestière en Afrique
Importance de la plante-hôte et amélioration
génétique de la fixation biologique de NS
chez les arbres tropicaux
DUHOUX E.
Université Paris VII et BSSmlCTFT (CIRADIORSTOM),
4.5 bis, Av. de la Belle Gabrielle. 94736. Nogent sur Marne France.
Summaty.
The host plant plays a most important role in the pmce~s of symbiotic N2 fixa-
tion Some of host gene products involved in the development and maintenance of
the symbiosis have already been identified. But further studies are necessary before
attempting the transfer of the symbiotic genes to novel plants.
Improvement of the host plant performances to increase its N,-fixing potential
and tolerance to environmental stresses could be achkved by : exploiting natm=al
genetic variability, expanding genetic variabiity through somaclonal variation and
introducing new genes in the plant genome by genetic transformation.
Theresultsdiscussedinthepapersupportthe recomlnwdation that selectian of host
plant should take into account various traits, specially N2-fixing potential, nitrate
assimilation, and tolerance to combined nitmgen inhibition.
Introduction
Les stratigies d’amt%oration des arbres forestiers reposent, comme chez les plantes
cultiv&s annuelles, sur une amélioration génétique des individus et également sur I’utili-
sation de mCthodes culturales adapteeS. Ce qui diff&re chez les ligneux, c’est tout d’abord,
la dur& de la vie de la plante qui peut être parfois consid&able et ensuite, la grande
variabilité g&&ique naturelle qui existe dans les peuplements, par suite de l’absence de
194
travaux de g&tique’sur ces végétaux a cycle de gMration tr&s long. De surcroît, dans
les associations symbiotiques. l’etude du partenaire veggétal a été longtemps d&tiss& au
profit de celle du microorganisme. Il n’est donc pas étonnant que l’arbre, en tant que
plante-hôte des associations symbiotiques forestières, ait été peu utilise en tant que syst&me
fixateur de N2
Les récents progr&s obtenus ces dernieres années dans les technologies avancees
de transfert de gènes aussi bien chez les microorganismes que chez les plantes herbacees
ont permis d’obtenir des r&sultats tout à fait remarquables (Shah et al., 1986 ; De. Blok
et al., 1987). Qu’en est-il des systbmes forestiers fixateurs de N2 ?
Avant de faire l’état actuel de nos coxuxGmnces, nous rappellerons l’importance
du partenaire végétal en dégageant quelques-uns des rôles connus de la plante-hôte dans
le fonctionnement de la symbiose furatrice de N2. Nous envisagerons ensuite les approches
possibles que l’on peut mettre en œuvre pour ameliorer la fixation de N2 en faisant appel
A la plante-hôte, et les premiers r&wltats concernan t cette stratégie seront présentés. Enfin,
l’amélioration gennétique de la plante-h& sera discutée en fonction des autres approches
Utilisées.
Importance du rble de la plante-hôte dans la fixation de N,
Qn sait que la planteh& fournit 1’6nergie snffiite au métabolisme de la fixation
de N, et à la survie du microorganisme symbiotique. Elle contrôle l’activité de I’hydrog&tase
et participe aussi à la protection de la nitrogénase. Depuis un demi-siècle, de nombreux
travaux ont demontre abondamment que chacune des etapes du processus de l’infection
et de la nodulation est contrôlt5e par la plante, en relation étroite avec le microorganisme
symbiotique. Le contrôle porte en particulier sur le nombre, la taille, la morphog&se
du nodule et la potentialit6 de forer N2’ On peut mettre en évidence ce rôle, non seulement
dans l’expression phénotypique, mais aussi, de plus en plus, au niveau génomique.
L’infection
L’existence des groupes d’inoculation croisée entre l’hôte et la bactérie symbiotique
est une manifestation évidente du rôle de la plante-hôte dans la symbiose. Une espèce
donnée de Rhizobium ou de Fnmliiu ne nodule pas toutes les espèces de lt5gurnineuses
ou de plantes actinorhiziennes mais seulement un ensemble d’espèces vegetales appartenant
a un groupe dit wl’inoculation crois&». Ce groupe peut comprendre un grand nombre
d’especes (qui sont dites non spt?cifïques) ou un petit nombre d’espèces ou même de
varie (groupe spécifique).
Sion s’en tientau syst&meL&umineuse-Rhizobiumquiconstituelemodelesymbiotique
le mieux connu actuellement, on peut souligner les rôles du genome de la plante qui
interviennent lors de l’infection de la plante par le Rhizobium :
émission de substances chimiques excr&es par la racine.
Ces substances jouent le r61e de signaux pour l’activation ou non des g&nes nod de la
bact&ie. Certaines de ces substances, hydroxyflavones et isoflavones (Fiiin et aL. 1986 ;
Peters et al., 1986) seraient stimuIatrices, tandis que d’autres, coumarines et isoflavones,
reprimeraient l’expression de ces mêmes genes (Djordjevic et al., 1987a). Les substances
stimulatrices seraient secretées dans la zone d’émergence des poiis absorbants, les substances
inhibitrices proviendraient des autres zones de la racine (Djorvdjevic et aZ., 1987b).
195
actlvatlon des syst&mes de reconnaissance. Sous l’influence
de polymères polysaccharidiques varies, exopolysaccharides, lipopolysaccharides, polysac-
charides capsulaires (Leigh et aL, 1985) synth&ises dans la paroi bacterienne, le génôme
de la plante neconnait ou non le Rhizobium spkcifique. En cas de reconnaissance il participe
a la courbure du poil absorbant et a la formation du cordon d’infection. Dans le cas
contraire il induit une réaction d’incompatibilité.
mlse en place de f’autorégulation de la nodulation. Ce phé-
nomène déclenché par le g6nome de la plante permet de limiter le nombre des nodules
et définit leur r@artition sur la racine (Nutman, 1952 ; Pierce et Bauer, 1983). La zone
de la racine susceptible d’&re infect& est Emit& à la région d’émergence des poils absor-
bants. Par ailleurs, une premiere infection inhibe un deuxième point d’inoculation situe
à proximité (Rolfe et Gresshoff, 1988).
Etablissement de la symbiose
La variabilité dans les réponses obtenues concernant la nodulation est le rbultat
de l’interaction du génome des deux partenaires. Par l’interm&liaire des mutants obtenus
à Ia fois chez la batterie et Ia plante-hôte on a mis en évidence une serie d’anomalies
dans la nodulation qui ont comme support le g&tôme des deux parknaires. Nous nous
attacherons ici a pr6sente.r celui de la plante-hôte.
On résumera dans le Tableau 1 quelques exemples de caractkstiques contrôlées
par la plante en indiquant, lorsqu’ils sont connus, les gènes responsables.
Fixation de N,
Chez de nombreuses associations L&umineuses-Rhizobium, Pisum (Cousin et al.,
1985), Ttifolium (Nutman, 1971). Meukugo (Viands et al., 1981 ; Barr~es et ai., 1986)
on a pu mettre en Cvidence une large variabilid génétique dans l’activité fvratrice. Une
part importante de cette variabilité revient a la plan-hôte.
Râble de I’hBte dans la quantité de N2 ~IX&. On choisira comme
exemple, une méthode d’amélioration de la fixation biologique de N, développée chez
la luzerne (Phillips et al., 1985) bas& sur la sélection des génotypes de la plante. Deux
génotypes HP et HP32, sont inocules avec quatre souches de Rhizobium. Il apparait (Fig.
1) que la quantité de N2 fixé est de 22 à 53% plus elev6e dans le génotype HP32. Les
performances symbiotiques sont donc nettement améliorees si on choisit le génotype de
Ia plante-hôte.
Influencede l’hôtedanslavariabilitédepotentlalité
de fixation
de A/, En utilisant le modele expkimental dtkrit dans la Fig.1. Phillips et ol., (1985)
ont montre que plus de 70 96 de la variabilité relative au poids sec des tiges, au poids
sec total, a à l’azote totzd peut être attribuée aux génotypes de la plante. Les souches
de Rhiwbium n’interviennent que pour 10 à 15 % dans la variabilite.
Au cours de la fmtion de N2, on sait que le g6ndme de la plante intervient &alement
Des gènes vegétaux sont exprimes uniquement dans le nodule conduisant à la formation
de substances spécifiques appeks nodulines. Dans les symbioses L&umineuses-Rhizobium
on en connait au moins une vingtaine (Govers et al. 1985) qui interviennent dans le
fonctionnement de la symbiose.
Tableau 1 : Rôles de l'hôte dans la nodulation
Caracteris-
hôte
g&ie de
Phénotype
Références
tique de la
l'hôte
observe
symbiose
responsable
pas de nodule
Glycine max
rjl
pas de cordon
CALDWELL, 1966
d'infection
pas de nodule
Pisum sativum
syml
pas de cordon
LIE, 1971
taille et
Trifolium sp. -
complexe
in NUTMAN, 1981
nombre des
nodules
nombre des
Glycine max
nombre x 10 NOS-
CARROLL et al.,
nodules
tolérant = super-
1985
nodulants
structure et
Lupinus sp.(L.) -
t R. Lupini no-
KIDBY et
forme des
Ornithopus SP.(O)
dules en anneau
GOODCHILD, 1966
nodules
(L) 1 bacte-
roTde/mb nodules
cylind. (0) pl.
bactérofdes/mb
mode d'in-
Eleagnus
Frankia
MILLER et
fection
Myrica
infection
BAKER, 1986
interc. (E)
infect. poil
abs. (Ml
1 9 7
102F28
lOîf65
414
445
.sescbes k RM2wbram mermtr
Fiiure 1 : N2 total fïxC par deux génotypes HP 32 et HP de luzerne inocul& avec quatre souches de
Rhizobium mdiloti. (Phillips et coll.. 1985).
,-
Les approches pour I’amMoratlon g&Gtique de la plante-hbte chez les arbres
Exploitation de la variabilit6 g&Mique naturelle
Approches conventionnelles. Une technique simple et efficace
consiste à utiliser les variations g&r&iques spontan@ obtenues dans un lot de graines
d’une même provenance (c’est à dire, d’une même aire geographique) ou de provenances
différentes. L’identification des individus ayant un potentiel fixateur de N, élevé doit être
bas& sur des tests non destructeurs, comme de la num&ation des nodules ou des tests
de réduction d’ac&yEne. Ensuite, les individus s&ctionnés sont propages par des techniques
de multiplication végétative classique (Sougoufara et al., 1987) et testes au champ pour
leur aptitude a fixer l’azote. Les auteurs ont pu ainsi identifier un clone 6 qui fixait 1.8
fois plus d’azote que le clone a inoculC avec la même souche de Frunkiu.
Approches modernes. Dans ce cas, l’ensemble des op&ations est
effectut5e au laboratoire. L’inoculation et la nodulation des plantes est obtenue in vitro
dans des systèmes expérimentaux adaptes aux arbres (Dhawan et Bhojwani, 1987 ; Alabarce,
1987). L’inoculation experirnentale en conditions aseptiques permet d’assurer le contrôle
de nombreux paramètres et d’optimiser la nodulation.
La s&ction des génotypes repose sur les comparaisons des tests d’effectivitt5 et
les mesures de biomasse des nodules et des plantes. Les’conditions d’asepsie permettent
de réaliser une micropropagation efficace des plantes s&ction&es. La multiplication vCg&
tative est assurée par les techniques de culture in vitro, en .proc&ant au developpement
des bourgeons axillaires vCg&atifs chez AZnus (Tremblay et Lalonde, 1984 ; Simon et
al., 1987), Eleugnus (Bertrand et Lalonde. 1986), Casuurin~ (Sougoufara et al., 1986).
Acacia ulbidu (Gassama et Duhoux, 1986), Acuciu nmngium @aha, com. pers.), ou
au developpement de bourgeons axiBaires d’inflorescencesimmaturesdecmumina@uhoux
et ul., 1986).
Induction par culture de tissus d’une nouvelle variabiltt6 g&étique par
modification globale du gBn6me
La variation somaclonale. La culture in vitro des cellules veggétales
s’accompagne tres souvent d’anomalies gén&.iques. Plus la phase de désorganisation tis-
sulaire (cal) est longue, plus fnQuents sont observes les variants lors de la I-ég&u%ation.
Ces anomalies relevent de plusieurs catégories : chromosomique (aneuploïdes, polyploïdes),
Cpigertique (d’origine nucléaire et/ou cytoplasmique) (Mestre et Benbadis, 1985).
La variation somaclonale repr&ente ainsi une methode pour produire une variation
gennétique en Cvitant la voie de la reproduction sexuel. Ces dernières annees, on a pu
appliquer ce proc&le pour l’obtention de divers variants chez les plantes herbacees (.Y&
lection de plantes tolerant aux toxines de pathogènes, sélection de plantes r&Aames aux
herbicides, Larkin. 1987).
Chez les arbres fixateurs de N, cette technique ne peut-être mise en œuvre que
chez les plantes capables de regtnérer a partir de cals : Alnus (Hutinen et ul., 1982 ;
Tremblay et ul.. 1985) ; Acacia kou (Skolmen et Mapes, 1976) ; Dulbergiu lutifoliu (Rao,
1986) ; Allocusuurina, (Duhoux, n%ultats non publiés).
199
Hybridation sumafique. Avec la fusion de protoplastes, deux
partem+s (es+ ou genres diff&ems) peuvent proceder à un échange d’information
génétique (nucleaire et/ou cytoplasmique) en contournant aussi une partie des obstacles
de la reproduction sexu6e.
Cette technique pourmit conduire 3 la cr&tion d’hybrides somatiques pos&dant
de nouvelles cam&&iques quant Zr leur nodulation, ou pr6sentant mie tol&ance accrue
à certaines contraintes de l’environnement. Actuellement, l’obtention de protoplastes d’arbres
fixateurs de NZ demeure rare ; Alnus ghtinosa. A. incana (Hutinen et al., 1982) et A.
incana (Tremblay et al., 1985) et I’absence fréquente de leur pouvoir de division et de
régtn&ation demeure encore une limitation de l’emploi de ces techniques.
introduction dans le gbnome de la plante d’un ou de quelques caract&es
nouveaux.
Les r6cents succès en biotechnologie ont montre des r&tssites remarquables pour
des plantes modeles comme le tabac avec laquelle Shah et al., (1986), De Block et al.,
(1987) ont obtenu des plantes transg4niques n5sisumte.s B un herbicide, le glyphosate. L’ob-
tention d’arbres transg&tiques rt%istants B watains herbicides est envisagee. Des telles
plantes possederaient ainsi un avantage sélectif certain, vis à vis de plantes herbacées
qui entrent en compétition au moment de la mise en place des plantations.
Les techniques de transfert de génes utilisent soit, des vecteurs biologiques
(Agrobacterium fum&acien.s, A. rhizogenes, le virus de la mosaïque du chou fleur), soit
le transfert direct (&ctroporation, liposomes, microinjection, microprojectiles enrobes de
DNA...). Le nombre de plantes transgeniques obtenues chez les arbres est encore faible
; un gène de tol&ance à un herbicide a été transf6r6 et exprimé chez un peuplier (Fillati
er al., 1987). Chez les arbres fixateurs de N,, des travaux ont débuté chez 1’Aulne
(Mackay et d., 1987a) et les Caswriw (Phelep, 1988).
Il est a noter que si les travaux..de gt%%que moEculaire concernan t la fixation
de N, sont maintenant bien avancés pour Ie microorganisme (Rhizobiwn), il n’en est pas
de même pour la plante-hôte (16gumineuses et plantes actinorhiziennes). Il en r6sulte que
nos co~aissances actuelles de la biologie mol6culaire de la plante-h&e ne sont pas encore
suff=antes pour envisager l’utilisation du génie g6nt%que dans l’amélioration de la poten-
tialité de fixer Nz.
D’autres strategies, comme celle de transf&er les g&nes de la nodulation zi des
arbres (le bouleau) non fixateurs de Nz, mais proches d’un point de vue systematique
des plantes actinorhiziermes (aulnes), sont envisagées (Tremblay et al., 1985).
Appllcatlon de I’exploltatlon de la varlabillt6 naturelle a l’am6lioratlon de la
flxatlon de N, chez les arbres
Les r&ultats expérimentaux
Les travaux concernant les arbres sont peu nombreux. Ils seront illustres ici par
un exemple d’am&ioration basée sur l’exploitation de la variabilité naturelle chez une
plante actinorhizienne (Sougoufara et al., 1987, 1989).
Un criblage a et.6 effectué sur de jeunes plants issus de semis de C. equisetifolia
et inocules avec la souche de Fmnkia ORS 021001 selon la methode déjà dt?crite (Sou-
goufara et ol., 1987). Trois individus a. 6 et S qui pn%entaient des differences signifi-
200
catives en ce qui concerne la biomasse de leurs parties aériennes, leur nodulation, et leur
aptitude à fixer N, ont été retenus. Des boutures âgées de 3 mois de ces individus ont
et.6 plantées dans un sol stérile déficient en N, marqué avec de l’urée 15N afin de comparer
les quantités d’azote de la plante provenant de la fixation de Nz (Ndfa) et du sol @kifs).
Une moitié des plantes a Cté inoculée avec la même souche de Fmnkàa, l’autre n’a pas
1505 inocul&. Les clones inocules de 12 mois diffktient entre eux, non seulement par
le poids de leurs nodules mais aussi par leur aptitude à fucer N, (Ndfa) et a absorber
N du sol (Ndfs), (Sougoufara et al., 1989). Les clones non inocules differaient Cgalement
entre eux par leur absorption spkfique de N du sol (Fig. 2).
10
8
6
4
2
Figure 3 : Biomasse exprim6.e en g/N/plante de trois clones de Casucwinu equisefifoia (S. U, $) 12 mois
apr&s leur transplantation (SOUGOUFARA et al., 1989).
1: plants inocul6-s avec la souche de
ORS 021001. La contribution du sol est repkentie par les
rectangles vides et celle de la fixation de N2 par les rectangles avec l’indication “Nz”.
201
On a pu ainsi mesurer, dans le cadre d’une variabilite intraspkifique de plantes-
hôtes de C. equisetzfolia, difft%ents critikes, Ndfa, Ndfs, biomasse des plantes (exprimée
en N total), qui sont indépendants des expkimentations. Les differences observées sont
uniquement le reflet de caractikes phenotypiques de la plante. Autrement dit, chaque clone
posskde des caracteristiques propres en ce qui concerne la potentialite & fmer NZ, d’absorber
Ndusoloudepmduiresabiomasse.
Dans le cadre de cette experience, les augmen-
tations de rendement consécutives à l’inoculation sont spectaculaires : +200% pour le
clone S, +158% pour le clone a, et +157% pour le clone B. La micropropagation des
clones s&c-tionn6s et leur transfert au champ peut donc amener des gains de produc-
tivité consid&able.
Choix des crit&es de s&ection
L’arrklioration de la plante-hôte doit tenir compte & la potentialitk de fixer N,
et Cgalement des crit&es d’adaptation a l’environnement (conditions climatiques, &a-
phiques, biotiques).
Pf~prï&&? symbiotiques. Les. mesures de biomasse de nodules
constituent M premier Grnent d’apprkmtion. Les techniques de mesure directe de la
fixation de N, étant difficiles à mettre en oeuvre chez les arbres âges, on se contentera
d’individus de quelques années, v@re même de jeunes plants. Les mesures sont réalisées
avec le test de réduction de l’ac&yli?ne (ARA) ou mieux, par les méthodes isotopiques
(Dommergutx et a/., 1988).
Adaptation $ l’environnement. Un des problemes importants
pourlamiseenplaced’un
pmgramme de sklection est de d&erminer, afin de les éviter,
les facteurs Limitant la croissance de la plante tek que la prknce d’azote combine du
sol, les contraintes hydriques, salines, et pathogenes. La t&rance aux herbicides peut
constituer pour ces vegetaux une caractkistique gknétique à introduire quand on sait, qu’en
zone aride, la compktition de la couverture herbacee est teks forte pendant les premières
ann&s de la vie de l’arbre. Parmi les facteurs biotiques, il faut inclure l’influence de
la comp&ition avec les souches sauvages de Rhizobium et de Frankia du sol. Considkant
ce dernier point, il est recommandé (Dommergues et ~2.. 1988) de s&ctionner des plantes-
hôtes capables de noduler seulement avec des souches trks effectives de microorganismes
spécifiques, pour eviter I’&abhssement de la symbiose par des souches sauvages pkentes
dans le sol.
Critéres de sélectlon propres aux arbres
(i) Utilisation de la multiplication v&&ative comme technique &
propagation des plantes sklectionnkes : il apparait peu realiste d’envisager un programme
d’amélioration bas& sur des expkiences d’hybridation sexuée chez les arbres tropicaux
fixateurs de N, qui prtkntent une trks forte hkkozygotie. La multiplication végttative
par les techniques conventionnelles et les techniques de culture in vitro s’avexent être
pour le moment, les seules mttbodes pratiques de propagation des individus s&ctio~&.
On connait la perte progressive de l’aptitude au developpement orthotrope et de la capacitt!
rhizogene des arbres âges, il s’avère donc preferable d’utiliser les techniques de culture
in vitro dont la puissance de multiplication est par ailleurs remarquable.
202
(ii) Sélection précoce des individus : la micropropagation est beaucoup
plus aisée a partir de jeunes individus (meilleure rkactivité, infections réduites...) que sur
des sujets matures. Si on peut détecter. par ailleurs, les carackkistiques physiologiques
des jeunes plants comme dans l’exemple du Cusuurinu (Sougoufara et al., 1989) la s&ec-
tion prkoce des individus réduira la dur& du processus d’amélioration de la plante.
(iii) Di@xltk du Eh& s’es critéres aé sélection : compte tenu des
résultats pr&minaires que nous posskdons, il est n6cessair-e de d&erminer des parambtres
simples comme la biomasse des plantes (exprirntk en N total et tenant compte a la fois
du potentiel fixatem de N, et de l’aptitude de l’arbre a utiliser N du sol) et de la biomasse
des nodules (exprimke en N total) qui semble assez bien corr&.s avec la fixation de
N, (Sougoufara et al., 1989).
Pour identifier les individus caract&ist% par un bon potentiel fixateur de N,, il
est indispensable de faire appel à une méthode kotopique qui permet en outre, de mesurer
l’aptitude des plantes a utiliser l’azote du sol. La s6lection des clones selon leur aptitude
à absorver N du sol, mais aussi selon leur susceptibilité a ce facteur extkieur semble
être un paramere important. Par ailleurs, Mackay et-d., (1987b) ont montre que la perfor-
mance des génotypes d’Alnw glz&wsu inoculés avec des souches de Frunkia &aient dépen-
dants de l’apport en N combiné du substrat, certains clones ttant plus d6pendants que
d’autres.
(iv) L’ éiectrophort?se 6 isozymes pmrmit-elle permettre d’ identjfier
des ghotypes bons jùateurs de IV, ? L’Ctude des profüs électrophorétiques (protéines,
DNA) constitue un outil puissant pour identifier et reconnaître a un stade precoce des
clones (Feldmann et al., 1983), pour contiler l’efficaciu5 d’un brassage génétique et en
déduire le type de fécondation, auto ou allogamie, pour estimer la variabilid génétique
(Vigneron, 1984). Cette technique peut-elle être utilis& pour l’identification de gknotypes
bons furateurs de N, ? Nous sommes tentés d’en douter, car on ne connait pas encore
le support gén6tique de la potentialiti plus ou moins grande de fuer NZ qui est complexe
et vmissemblablement polyg6nique. Par ailleurs, la variation quantitative d’une fonction
génique traduit la variabilité des gènes de r&ulation affectant les quantités de prot&nes
et la variation qualitative correspond en majorité & la variabilité des genes de structure
(Damerval et af., 19%). Une étude &ctrophor&ique pourrait donc, sans doute, identifier
et repérer des génotypes non fkateurs par comparaison avec des individus futant NZ- Par
contre, il semble plus improbable de mettre en evidence une liaison entre une variation
dans le degre de fixation de N, chez un génotype et une variation de la nature des marqueurs
enzymatiques ; les techniques &ctrophor&iques n’appr&ant que des variations quali-
tatives des molkcules (protkii, ADN) analyseeS. C’est d’ailleurs la raison pour laquelle
les caractkres enzymatiques ne sont que tres rarement lies a la valeur agronomique des
individus.
Conclusions
L’amélioration de la fixation biologique de N, à travers la sélection et la gennttique
est un probleme complexe car ce phenomène relève des deux partenaires. Les travaux
de recherche dCvelopp& ces dernières annt%s ont largement contribué à l’amelioration
de nos connaissances sur la complexité des mtkanismes mis en jeu. Les premières recherhes
ont surtout été orientées vers le microorganisme et une première moisson de résultats
t&s importants concernant le Rhizobium a dejà Cté récoltée.
203
Nous avons prkenté ici l’etat de nos connaissances concernant la plante-hôte dont
le rôle joué dans la symbiose commence a être de mieux en mieux connu, même s’il
reste encore beaucoup à faire en gédtique et en biologie mol~ulaire. Le s&ctionneur
peut améliorer efficacement la biomasse produite en utilisant la grande variabilite géné-
tique naturelle des arbres. Selon les rksultats dont nous disposons actuellement, il apparait
que l’objectif des travaux de s&ction à venir serait d’identifier des gknotypes p&entant
les trois caract.&istiques suivantes : grand pouvoir de fixation de N,. faible susceptibilité
de la symbiose a l’egard de N combiné du sol, et accroissement de l’assimilation de
N du SOL Ces trois caractkristiques sont des caractères indQe&am.s sur le plan g&rt%iqué ;
elles peuvent vraissemblablement agir en synergie, et sont pour quelques cas connus, le
Cas~ (Sougoufara et al., 1989) et le soja (Imsande, 1985) corr&es avec une biomasse
importante de l’appareil vég&atif de la plante. Il serait Cgalement utile de sélectionner
des arbres capables de continuer a fixer N, en présence de N combine du sol en particulier
dans les peuplements denses (forêts) où N du sol s’accumule au cours du temps.
Rfifbre nces
ALABARCE, J., 1987 Localisation et developpement des poils absorbants de Casuarina
stictu et Acacia mangium : mise au point d’un dispositif expkimental et r&ultats
prr%minaires. Mém. D.E.S.U. Université Paris VII, 62~.
BARNEiS, DK. JESSEN, D.L., HEICHEL, GI-I. et VANCE, CP. 1986 Selection for
multiple traits increases alfalfa N, fixation and yield. In : Advances in nitrogen fixa-
tion research, 7-9.
BEDMAR, EJ., EDIE, S.A. et PHILLIPS, D.A. 1983
Host plant cultivar effects on
hydrogen evolution by Rhizobim legumhwsarum. Plant Physiol. 72, 1011-1015.
BERTRAND, LJ. et LALONDE, M. 1985 In vitro propagation and nodulation by Frun-
kiu of actinorhizal Russian Olive (Eleugnus ungustijiiliu L.). Plant und Sd. 87, 143-
152
CALDWELL, B.E. 1966 Inheritance of a strain-specific ineffective nodulation in soybeans.
Crop. Sci. 6, 427-428.
CARROLL, B.J., McNEIL, DL. et GRESSHOFF, PM. 1985 Isolation and properties
of soybean (Glycine max) mutants that nodulate in the presence of high nitrate concen-
trations. Pro~. Nat. Acad. Sci. (USA). 82, 41624166.
COUSIN, R., MESSAGER, A., PICARD, J., De LAREMBERGUE, H., et MARGUET,
P. 1987 Variabiliti g&tique pour la fixation de l’azote chez le pois. In : Nutrition
azotée des @umineuses, Versailles, 19-21 Novembre 1985, &l., INRA, PARIS, 333-
342.
DAMMERVAL, C., De VIENNE, D. et HEBERT, Y. 1986 Divergence morphologique
et divergence mesur& par électrophorese bidimensionelle des protemes d&tatur&.s.
Sbminaire sur les distance.s génétiques. M&ibel, 9-16 Mars 1986, lp.
De BLOCK, M., BOTTERMAN, J., VANDEWIEIE, M., DOCKY, J., CTHOEN, V.G.,
RAO MOWA, N., THOMPSON, C., VAN MONTAGU, M. et LEEMANS, J. 1987
Engineering herbicide resistance in plants by expression of a detoxifying enzyme.
Embo .l. 6, 2513-2518.
204
DHAWAN, V. and BHOJWANI, S.S., 1987 In vitro nodulation of micropropagated plants
of Leucaena leucocephala by Rhizobium. Plant ami Soil. 103, 274-276.
DJORDJEVlC, M-A., REDMOND, J.W., BATLEY, M. et ROLFE, B.G. 1987a Clovers
secrete specific phenolic compounds which stimulate or repress nod gene expression
in Rhizobium tri&olii. Embo J. 6, 1173-1179.
DJORDJEVlC, M.A., GABRIEL, D.W. et ROLFE, B.G. 1987b Rhiwbium : t.k refmed
parasite of legume. Ann. Rev. Phytopathol. 25, 145168.
DOMMERGUES, YR., DUHOUX, E. et DIEM, RG. 1988 Les arbres fixateurs de N,
en foresterie et agroforesterie tropicales. Bull. Sc. Bot. Actual. Bot., (sous presse).
DUHOUX, E., SOUGOUFARA, B. et DOMMERGUES, Y. 1986 Propagation of Canrmina
equisetifolia through axillary buds of immature female inflorescences cultured in vitro.
Plant Ce11 Rep. 3, 161-164.
FELDMANN, P., LACROIX, J.L. et GLASZMANN, JC. 1983 L’électrophorèse d’iso-
zymes, outil du dlectionneur et du généticien. Bulletin d’information et de liaison
de I’ZRAT. 1-2, 33-38.
FlLLATI, J., SELLMER, J., McCOWN, B., HAISSIG, B. et COMAI. L. 1987 Agrobacte-
rium mediated transformation and regeneration of Populus. Mol. Gen. Genet. 206,
192-199.
FIRMIN,J.L., WESON, KE.,ROSSEN, L., JOHNSTON, A.W.B. 1986 Flavonoidactivation
of nodulation genes in Rhizobium reversed by other compotmds present in plants.
Nature. 324, 90-92.
GASSAMA, Y.K. et DUHOUX, E. 1986 Micropropagation d’arbres adultes d’Acacia
albida @.eguminosae). In : Les arbres fucateurs d’azote. L’amelioration biologique
de la fertilité du SOL Acte des S&ninaires 17-25 Mars 1986, Dakar, Bd. ORSTOM
PARIS, 216224.
GOVERS, F., GLOUDEMANS, T., MOERMAN, M, VAN KAMMEN, A. et BISSELING,
T. 1985 Expression of plant genes during the development of pea root nodules. Embo
J 4, 861-867.
HUTlNEN, O., HONKANEN, J. et SIMOLA, L.K. 1982 Ornithine and putrescine supported
divisions aud cell colony formation in leaf protoplasts of alders (Ahus glutinosa and
A. incana). Plant Sci. Lett. 28, 3-9.
IMSANDE, J. 1985 Plant genotype and the control of nitrogen fixation on soybean.
In : Nitrogen fixation reseamh progmss. Evans, HJ., Bottomley, P.J. et Newton,
W.E. &l., Martinus Nijhoff Publisbers, 34.
KIDBY, D.K. and GOODCHILD, DJ. 1966 Host influence in tbe ultrastructure of root
nodules of Lupinus luteus and Ornithopus sativus. J. Gen. Microbiol. 45, 147-152.
LAFKIN, PJ., 1987 Somaclonal variation : history, method, and meaning. Zona State
J. Res. 61. 393-434.
LEIGH, J.A., SIGNER, E.R., WALKER, G.C. 1985 Exopolysaccharide deficient mutants
of Rhizobium meliloti that form ineffective nodules. Proc. Natl. AIX& Sci. USA. 82
62316235.
LIE, T.A. 1971 Symbiotic nitrogen fixation under stress conditions. Plant and Soi1 Spec.
vol. 117-127.
h4aGmiser la FBA pour la Production Agricok et Forestihe en Afrique
Analysis of the plant genes involved
in the nitrogen fixing symbiosis.
Sesbania rostrata as a model
De hWUDIE l?
Laboratoire & Biologic ah Sok, ORSi”OM, B.P. I386. Dakm. SMgal.
Summary
During nitrogen ftig nodule formation on ieguminous plants, several plant
genes. so-called nodulin gene& iue specifically expressed. In this paper the current
knowledge on the expression, induction and regulation of noduh genes in the diffe-
rent symbiosises studied so far is summarized, together with main results obtained
in the case of Sesbania rostra&z on nodule-specific plant gene expression.
Introduction
Each stage in the Rhizobium-legume symbiosis is characterized by a series of
developmental events concerning both bacteria and plans resulting in a complex, well
organizedand wellcoordinatedplantorgan.Thishasbeen welldescriifromamorphological
point of view (Bauer, 1981).
By classical genetic experiments, several plant genes involved in nod&ion and
symbiotic nitrogen fixation have been identified in pea, soybean, clover and alfalfa (Nutman,
1981 ; la Rue et al., 1985). Mutations in the plant genome can result in disturbed nodule
development, varying tiom the absence of nodules to the developement of wild-type-
like but ineffective nodules (Vincent, 1980).
Theknowledge of nodule formation has made rapid progress since the rise of molecular
biological techniques. It has been found that a number of genes in both plant and bacterium
are only expressed in nodules. An effective symbiosis is accomplished by different&ion
of bacteria into bactero&ls on the one hand, and differentiation of plant cells into a
root nodule on the other hand. The major part of research activity has concentrated on
the nitrogen fixing bacteria, which are more easily accessible to gehetic manipulation
in comparison with legume plants. Most of the bacterial symbiotic genes (genes for nodu-
lation and nitrogen fixation) have been identified and are well documented. The plant
is equally important, providing the right environment, the energy and using the fixed
nitrogen for its growth and development. Over the last few years, interest in the role
of the plant in the symbiosis has considerably intensified, with the evidence of plant
genome-encoded, nodulespecific proteins, called nodulins &ego&i and Vern& 1980)
::,:
differentially expressed during nodule development.
Compared with other plant differemiation processes, root nodule development is
unique in the involvement of a prokaryote in the induction and control of plant development
_
Developmental program leading to root nodules, as an organ, is as complex as other deve-
lopmental programs in plants and involves numerous genes. Nodulin gene expression is
one of the most specific aspects of nodule differentiation and its study may be a useful
approach to understand root nodule development.
Nodulin definition. By definition, nodulins are plant gene-encoded proteins, which
are found only in root nodules and not in uninfected roots nor in other part of the host
plant (Van Kammen, 1984). Nodulin genes are, by consequence, plant genes exclusively
expressed during the development of the symbiosis. Around 10 to 40 different nodulins
have been found in the various systems studied so far, like soybean (Legocki and Verma
1980). pea (Bisseling et al., 1983), Medicago sativa (Lullien et al., 1987), Phaseohs
vulgaris (Campus et al., 1987), Sesbania rostrata (De Lajudie and Huguet, 1988). Nodulin
genes are differentially expressed during nodule development (Bisseling et al., 1983 ;
Gloudemans et al., 1987). The majority of nodulin genes is expressed around the onset
of nitrogen fixation, like leghemoglobin, and are called nodulin genes. They most p&a-
bly function in establishing and maintaining a proper environment within the nodule that
allows nitrogen fixation and ammonium assimilation to occur. Few nodulins are detec-
table at earlier stages of development when the nodule structure is being formed, which
are called early nodulin genes.
Nodulin functions. In all systems studied so far, the use of mutated or engi-
neered bacterial strains that arrest nodule development at different stages has enabled
to establish parallels events at the levels of nodule structure and nodulin gene.expression
by coupling histological and molecular biological data. Very few nodulins have been iden-
tified Moreover, nodulins must play roles at every stage of nodule developement and
function : (i) for plant-microbe recognition : lectins, enzymes of host origin capable of
degrading the capsular polysaccharides of the bacteria (Batter, 1981) ; (ii) during infection
process : root hair curling, infection thread formation, development of a meristem from
cortex cells, release of bacteria in the meristematic cells and differentiation of host cells
and bacteria into a nitrogen fixing nodule ; (iii) for the transport of metabolites including
organic acids, amino acids and sugars, as well as N, and 0, ; (iv) for the assimilation
of fixed nitrogen produced by bacteroyds ; (v) roles in all the morphological, cytological
and physiological changes that take place in the infected cell, like modification of endo-
membrane systems, increased abundance of free ribosomes. polyribosomes. proplastids
and mitochondria ; carbon, nitrogen and oxygen metabolism changes dramatically during
the differentiation of the me&tern into a root nodule, as indicated by the increase in
activities of several enzymes such as PEP carboxylase, malate deshydrogenase, glutamine
synthetase and uricase (for a review see Verma and Long, 1983).
A few early nodulin genes have been identified. Among them only one has been
well character&d : ENODZ, a nodulin of soybean (Glycine mar.), which has been also
found in other legumes like pea, vetch, alfalfa and clover (Govers et al., 1987). The
strucW features of this nodulin, like its hydroxyproline-rich sequence, suggest that it
is a cell wall protein closely related to the extensin (Nap, 1988).
Late nodulins are more documented The most famous is leghemoglobin, which
constitutes 30.? of the total soluble proteins. Leghemoglobin is a hemoprotein with high
oxygen affinity, which provides high yield of oxygen to bacteroids at a low partial tension,
compatiile with nitrogen fixation. It is generally detectable just before nitrogenase activity
can be measured. Leghemoglobin is a true symbiotic protein since the heme is a product
of the bacterold and protein is plant-encoded. In alI legumes studied so far, there are
several leghemoglobins, encoded by several genes (Marcker et al., 1984). Other late nodulius
have been identifield. One is a soybean nodule-specific uricase, a key enzyme in the
ureide biosynthetic pathway for ammonia assimilation (Bergman ef al., 1983). Two enzymes
of glutamine synthetase. which catalyses the first reaction in the assim&tion of ammonia
into organic nitrogen, are present in nodules of Phaseoh vulgaris, one of which is nodule-
specific (Cullimore et al., 1983). A sucrose synthase, an enzyme which cataIyzes the
degradation of sucrose into fructose and glucose, has been found specific to nodules of
soybean (Thummler and Verma, 1987). Other nodule-specific forms of enzymes that differ
in physical, kinetic and immunochemical properties from the corresponding enzymes in
roots have been found : phosphoenolpyruvate carboxylase, choline kinase, xanthine
deshydrogenase, purine nucleosidase, malate deshydrogenase (Nap, 1988). But there is
still no evidence that they derive from nodule-specific gene expression or from nodule-
specific modifications of root enzymes.
Several other nodulins, whose functions are yet unknown, have been found as-
sociated with peribactero~d membrane, the site of closest interaction and exchanges between
the bacteroRls and the host cell.
Molecular studies on nodulin genes have progressed a lot since the last few years :
gene cloning, organization in the genome, regulation, sequence analysis (for a review,
see Bissehng et al., 1986).
Sesbanla rostmta
Sesbaniu rosrrata is a tropical legume which forms nitrogeu’ fixing uodules on
stem and roots. This makes Sesbania a very efficient system for nitrogen fixation and
a real potential as green manure in tropical agriculture. Two bacterial genera are capable
to induce nodules on S. rostruta : Azorhizobium caulinodans and Rhizobium sp. (Dreyfus
et al., 1988). Due to its particular properties (in vitro nitrogen fixation), A.cudinodam
has been extensively studied. Opposite to this, few data are available on the plant itself
(E3ogmz er al., 1987 ; De Lajudie and Huguet, 1988, 1989). We focussed on S. rosrrutu
to investigate plant genes activated during nodule development and function. S. rostrata
constitutes a good model in such a study since, opposite to that of root nodules, age
of stem nodules can precisely be known : stem nodules are homogenous in age, opposite
to root nodules that form continuously.
210
Our approach consisted in measurin g the level of translatable mRNA present in
each tissue by comparing one-and-two dimensional polyacrylamide gel electrophoretic
patterns of in vitro translation products of poly(A)+ RNA from uninfected stems and stem
nodules, uninfected roots and root nodules.
Plant gene expression in stem nodules appears very similar to that in root nodules,
except some quantitative differences in the intensity of certain spots which could be ex-
plained by the heterogenous age of root nodules. We found 16 nodule-specific polypep-
tides, in addition to around 20 other present at different relative intensities in nodules
and in uninfected tissues. During stem nodule development, majority of nodule-specific
genes are expressed 12 days after inoculation, concommitantly with leghemogt’cibin and
initiation of nitrogen fixation. These could be correlated with transport of substrates
towards bacteroyds and assimilation of fmed nitrogen. Some other genes are expressed
either during the early stages of infection, or transiently, suggesting a role in plant-microbe
recognition, infection process, and nodule morphogenesis. Some others, expressed later
(day 18), could have a role in maintaining nodule structure and function, or in nodule
senescence. All these observations indicate that there is a sequence in the activation of
nodule-specific genes during nodule life, like in other descrii systems.
Among the nodule-specific polypeptides, we identified six in vitro translation
products that cross reacted with a serum anti-leghemoglobin of SesbunLz rostrata. During
nodule development, these leghemoglobin components are differently expressed and their
relative proportions vary : some of them are fmt major to become minor during the
late stages of nodule development.
Root nodules induced either by Rhizobium sp. (ORS 51) or A. caulinodans
(ORS 571) are very similar except for the expression of one more specific gene in the
case of ORS 51. One of the great differences pointed out between stem and root nodules
is the presence of chloroplasts and photosynthesis in stem nodules. We found interes-
ting to compare plant gene expression in stem nodules developed either in the dark or
under light : they show only minor differences ; moreover, stem nodules in the dark
and root nodules show some striking similarities. This, together with other arguments
like the same seven leghemoglobin isomeres (Bogusz er aI., 1987), the same infection
process in stem and root nodules (Ndoye and Truchet, in preparation), or the fact that
root nodules developed under light are photosynthetic, suggests that stem and root nodules
are not of completely different nature.
We used different nitrogen fixation deficient mutant strains of A. caz&odans
ORS 571 and observed that induction of nodule-specific gene expression is not depen-
dant on nitrogen fixation. However the level of expression of leghemoglobin and several
other nodulin genes is lowered when there is no nitrogen fming activity in nodule. This
is in agreement with all systems described so far.
Conclusion
Besides the agronomic aspect, nitrogen fming symbiosises are a tools towards elu-
cidating several fundamental biological problems. They constitute a model for the study
of plant microbe interactions, which are diverse and numerous in nature, especially of
parasitism and pathogenicity.
Opposite to Agrobacterium tumefaciens, the other genus of the Rhizobiaceae
family, which induces tumor formation, a set of undifferentiated cells, Rhizobium is able
211
to bypass the plant defense mechanisms and to induce a de-differentiation of plant cells
and the development of a nodule, a true organ specialized for nitrogen fixation. Therefore,
Rhizobium-Legume symbiosis with Frankia actinorhii plants are an interesting and rare
example for plant molecular biology of the induction by a prokaryote of an organoge-
nesis in a eulmryote.
In less than a decade, our knowledge on plant gene expression during symbiosis
has progressed a lot, and we begin to know more about the complex mechanisms of
symbiosis, bacterial genes implied in the induction of nodulin genes, which is a first
step in identifying bacterial signals to the plant that induces expression of the successive
classes ofnodulin,genes. A better understanding of alI the mechanisms implied in sym-
biosis will help in a better domestication for a practical use in the field
R&&ewes
APPLEBY. CA. 1984 Leghemoglobin and Rhizobium respiration. In : Briggs WR. (ed).
Ann. Rev. Plant Physiol. 35. 443-478.
BAUER, WD. 1981 Infection of legumes by rhizobia. Ann. Rev.. Plant Physiol. 32,407-
449.
BERGMANN, IX, PREDDIE, E. and, VERMA, D.P.S. 1983 Nodulin 35 : a subunit
micase (uricase II) induced and locahzed in the uninfected cells of soybean nodules.
Embo J. 2 (12), 2333-2339
BISSELING, T., BEEN, C., KLUGKIST, J., VAN KIMMEN, A., NADLER, K. 1983
Nodule-specific host proteins in effective and ineffective root nodules of Pisum sa-
tivwn. Embo J. 2, 961-966.
BISSELING, T., VAN DEN BOS. R.C. and VAN KAMMEN, A. 1986 Host specific
gene expression in legume root noduks. In : Nitrogen fixation. Vol 4 : Molecular
Biology. Broughton WJ. and Piihler A. cds. Clarendon Press, Oxford, 280-312.
BOGUS& D., KORTI’, kk, APPLEBY, CA. 1987 Sesbania roslrata stem and root
nodule leghemoglobiu : purification and relationships amongst the seven major com-
ponents. Arch. Biochem. Biophys. 254 (l), 263-271.
CAMPOS,F.,PADILLA,J.,VAZQUEZ,M.,OR-IEGA,J.L.,ENRIQUEZ,C.,andSANCHEZ,
F. 1987 Expression of nodule-specific getles in Phareolus vulgaris L. Plant Mol.
Biol. 2 , 521-532.
CULLIMORE, J.V., LARA, M, LEA, PJ., and MIFLIN, BJ. 1983 Purification and
properties of two forms of glutamine synthetase from the plant fraction of Phaseolus
root nodules. Planta.157, 245-253.
DREYFUS, B.L., GARCIA, JL. and GILLIS, M. 1988 Characterization of Atorhizo&un
caulin0dan.s gene nav. sp. nov., a stem-nodulating nitrogen fixing bacterium isolated
from Sesbania rostrata. Int. J. Syst. Bacterial. 38, 89-98.
GLOUDEMANS, T. De VRIES, S., BUSSINK, HJ., MALIK, N.S.A., FRANSSEN, H.J.,
LOUWERSE. J. and BISSELING, T. 1987 Nodulin gene expression during soybean
(Glycine mux) nodule development. Plant Mol. Biol. 8, 395-403.
GOVERS, F., NAP, J.P., MOERMAN. M. FRANSSEN, HJ., VAN KAMMEN, A., and
BISSELING T. 1987 cDNA cloning and developmental expression of pea nodulin
genes. Plant Mol. Biol. 8, 425-435.
212
LAJUDIE (De), P. and HUGUET, T. 1988 Plant gene expression during effective and
ineffective nodule development of the tropical stem nodulated legume Sesbania ros-
trata. Plant Mol. Bioi. 10, 537-548.
LAJUDIE (De), P. and HUGUET, T. 1989 Mise en evidence de nodulines chez Sesbania
rostra& In : Sesbania rostrata : fundamental research and application to agriculture.
Inpress-
LA RUE, T.A., KNEEN, BE. and GARTSIDE, E. 1985 Plant mutants defective in sym-
biotic nitrogen fixation. In : Analysis of the plant genes involved in the legume-
Rhizobium symbiosis. OECD, Paris, pp. 39-48.
LEGOCKI, R.P. and VERMA, D.P.S. 1980 I&ntification of nodulespecific host pro-
teins (nodulins) involved in the development of Rhizo.bium-legume symbiosis. Cell.
20, 153-163.
LULLIEN, V., BARKER, D.G., DE LAJUDIE, P. and HUGUET, T. 1987 Plant gene
expression in effective and ineffective root nodules of alfalfa (Medicago sativa). Plant
Mol. Biol. 9, 469-478.
MARCKER, A., LUND, M., JENSEN, E.O. and MARCKER, K.A., 1984 Transcrip-
tion of the soybean leghemoglobin genes during nodule development. Embo J. 3 (8),
1691-1695.
NAP, J.P. 1988 Nodulins in root nodule development. Ph. D. Thesis. Wageningen University
(Nederland).
NUTMAN, P.S. 1981 Hereditary host factors affecting nodulation and nitrogen fixation.
In : Gibson A.H. and Newton WE. (eds). Current perspectives in nitrogen fixation.
Elsevier, North Holland, pp. 194-204.
THUMMLER, F. and VERMA, DP.S. 1987 Nodulin 100 of soybean is the subunit of
sucrose synthase regulated by the availability of free heme in nodules. J. Biol. Chem.
262 (30), 14730-14736.
VAN KAMMEN, A 1984 Suggested nomenclature for plant genes involved in nodulation
and symbiosis. Plant Mol. Biol. Rep. 2, 43-45.
VERMA, D.P.S. and LONG, S. 1983 The molecular biology of Rhizobium - legume
symbiosis. Int. Rev. Cytol. Suppl. 14, 211-245.
VINCENT, J.M. 1980 Factors controlling the legume - Rhizobium symbiosis. In : Nitrogen
Fixation II, Newton W.E. and Grme Johnson W.H. eds, University Park Press, Bal-
timore, 103-129.
la FBA pour la production agricole et forestière en Afrique
Sesbania rostrata :
un modèle pour la biologie moléculaire
de la nodulation
TOMEKPE, Kil), HOLSTERS, M.*,
GOETHALS, a(2), VAN DEN EEDE, ~9,
De LAJUDIE, P.(l), TRAN, P.(l)
and DREYFUS, B.0)
(1) : aratoire de Microbiologie des Sols ORSTOM.
B.P. 1386, Dakar, Senegal ;
(2) : Luboratorium $znetikz, Uniwsiti de Gent 9000, Belgique.
Sesbunia rostrata forme des nodules aussi bien sur les tiges que sur les racines.
Sesmicro-symbiotesappam~~8dwxgenresf~ts:AzorhuobiumetRhizobiwn.
Les souches telle que ORS 571 qui induiint des nodules de tige et de racine et
qui fixent l’azote atmosphkrique en culture pure appartiument a la nouvelle espèce
Azwhizobium cauIino&ms~ Les .s~~&e.s isolkes des racines de S. rartrata et qui
forment des nodules effectifs sur d’autres espèces de Se.&mia appartiennent au genre
Rhùobium.
La mutagenèse de la souche d’Azorhimbim caulinodanr ORS 571 a permis d’iden-
tifier deux loti Nodl et Nod2 qui sont impliqu& dans la double nodulation racikre
et caulinaire. L’hybridation molkulaire et l’analyse des s6quences du locus Nodl
ont n%lé l’existence de trois gknes ayant une forte homologie avec les gènes nod
ABC de Rhizobiwn meliloti. L’analyse génktique a mont& que l’expression de ces
g&nes est induite en pr6sence des racines et des primordia racinaires situ6.s sur la
tige de S. rostrata ; elle a par ailleurs permis de d6couvrir. dans le gknome de
ORS 571, un troisAne locus dont la pr6sence est essentielle pour l’induction des
gènes du locus Nodl. Ce troisAne locus semble réguler l’expression des gènes de
nodulation racinaire et caulinaire.
214
Chez S. rostrufu, un mutant incapable de former des nodules de tige a été isolé.
Ce mutant a une croissance atkknne normale mais sa racineprincipale est plus vigou-
reuse et ses racines secondaires moins abondantes que celles de la plante de type
sauvage. Des diffkrences dans la composition des protiines ont été mises en hidence
entre le mutant et la plante normale. Une caracttkisation mol6culaire approfondie
devrait permettre de rechercher chez S. rostrufa les gi?nes impliquks
dans la nodu-
lation caulinaire.
Introduction
Sesbaniu rostrutu est une ltgumineuse que l’on rencontre sur les sols hydro-
morphes de l’Ouest-africain. Son aptitude à former des nodules caulinaires lui confkre
un potentiel de fixation d’azote exceptionnel et lui permet de noduler et de fixer l’azote
sur des sols inondks ou renfermant des quantités elevées d’azote combine (Dreyfus et
id., 1984).
Les nodules caulinaircs se développent à partir de sites pr6d&ermin~ (Duhoux
et Dreyfus, 1982 ; Tsien et ul., 1983). Ces sites sur les tiges sont une caractdristique
particuliière de Sesbuniu rostruta, de certaines espkces d’Aeschynomene et & Neptuniu
(Dreyfus et ul., 1984) ; ils sont constitues d’ebauches racinaires adventices qui se dévelop-
pent en nodules lorsqu’elles sont infectées par des symbiotes spkcifïques (Duhoux et
Dreyfus, 1982).
Les bactéries qui sont associées à Sesbunia rostrutu appartiennent .à deux genres
différents : Azorhizobium et Rhizobiwn.
Les souches d’Azorhizobium cuulinodans du type ORS 571 (Dreyfus et ul., 1988)
induisent des nodules fixateurs d’azote aussi bien sur les tiges que sur les racines. Ces
souches sont capables de fixer l’azote en culture pure. Bien que les souches d’Azorhizobium
soient capables de former des nodules sur un grand nombre d’espèces du genre Sesbania,
elles ne fixent l’azote qu’avec Sesbuniu rostruta.
Les souches de Rhizobium nodulent effectivement les racines de Sesbania rostrata
et des autres espèces du genre Sesbuniu ; la plupart d’entre elles sont capables de former
des nodules fixateurs d’azote sur les tiges de Sesbania romura, certaines y induisent
des pseudonodules ou des nodules ineffectifs et quelques unes ne forment pas du tout
de nodules sur les tiges.
Chez Azorhizobium cuulinoduns, la mutagen&se par transposon Tn5 de la souche
ORS 571 a permis d’identifier et de cloner deux loti Noël et Nota essentiels pour la
nodulation racinaire et caulinaire (Van Den Eede et al., 1987). La technique d’hybrida-
tion moléculaire a permis de mettre en Cvidence une homologie de kquence entre le
locus Noël et les genes de nodulation nodABC de Rhizobium meliloti (Kondorosi et al.,
1984). Les gènes homologues no&IBC de la souche ORS 571 ont ensuite étk isoles et
caracti& par le skquençage du locus NodI.
Le pr6sent article est principalement consacré a l’étude de l’expression des gènes
nod de la souche ORS 571 ; nous nous intkresserons par ailleurs aux résultats prt%mi-
naires concernant la caractérisation d’un mutant de Sesbunia rosrrutu &pourvu de sites
de nodulation de tige.
215
Induction et txigulatlon des genes de nodulatlon de la souche ORS 571
Chez certains Rhizobium à croissance rapide tel que R. meliloti et R. Legumino-
sarum, les gènes nod ne s’expriment qu’en pr&ence d’exsudats racinaires issus de la
plante hôte (John et al., 1985). Différents flavonoïdes isoles a partir de ces exsudats notam-
ment la lutMine chez la luzerne, sont capables d’induire l’expression de ces g&res. Par
ailleurs, il a & mis en 6vidence dans le gtkome de ces Rhizobium, des gènes régulateurs
indispensables à l’induction des genes nod par des flavonoïdes de la plante hôte (John
et al., 1985).
Pour étudier l’induction de l’expression des gènes nod de la souche ORS 571,
nous avons construit plusieurs fusions entre le locus Nodl et le gène LucZ de la
B-galactosidase de E. coli à l’aide d’un transposon (Stachel et al., 1985). Le principe
de ces fusions consiste à insérer dans un fragment d’ADN le gene LacZ sans son pro-
moteur afin que son expression soit sous la seule dépendance de l’expression du gène
dans lequel il s’est inske ; il faut pour cela que la fusion soit orientée dans la direction
uanscriptionnelle et qu’elle soit en phase de lecture. Les fusions construites dans le locus
Nodl étaient ensuite transf&ees dans une population de ORS 571 et testées pour l’ac-
_ tivité de la &galactosidase en présence et en absence de Sesbania rostrata.
Quelques unes des nombreuses fusions testées étaient induites par les fragments
de tige et de racine de Sesbania iostrata. Ces fusions inductibles sont toutes localisées
dans les genes nod et transcrites dans la même direction ; de plus, ces gènes semblent
dépendre de la même unité transcriptionnelle, ce qui indique une organisation sous forme
d’opéron comme pour les gknes noaXBC de R. meliloti.
Pour isoler la substance qui induit l’expression des g&es de la souche ORS 571,
des exudats racinaires de Sesbania rostrata ont été analysés par chromatographie liquide.
Le test d’activite de la l3-galactosidase à partir des diff&entes fractions issues de ces
.
exudats a revele l’existence d’un inducteur de type flavonoïde. Nous caract&isons actuel-
lement cette substance. ParalElement, nous p&parons des extraits de tige afin d’isoler
et de comparer la substance inductrice de tige à celle des racines. Nous avons par ailleurs
teste les fusions inductibles en prksence d’une s&ie de flavonoïdes de provenance com-
merciale ; des niveaux elevés d’induction ont été observés avec la naring&tine (Nar).
Quelques mesures de l’activité de la &galactosidase sont prksentkes dans le Tableau 1.
Etude de la rbgulation de l’induction des génes Nod de la souche ORS 571
Comme l’induction des genes de R. meüloti par la lutt!oline est sous la d@endance
d’un gkne r&ulateur no&, nous avons pense à une situation similaire chez ORS 571.
Deux approches ont été choisies pour rechercher le gène régulateur homologue du g&ne
nodD dans le g&rome de ORS 571.
(i) Après avoir introduit une fusion inductible dam Agrobacterium tumefaciens,
nous avons recherche dans la banque d’ADN de ORS 571 les clones qui permettent son
induction par la naringtnine. Un clone ayant un fragment de 16 kilobases (clone pRG910-
16) a ainsi Cté sklectionné.
(ii) Dans une deuxième approche, nous avons isolé un mutant spoman6 (ORS 571-
IN-) qui ne permet plus l’induction des fusions ZacZ par les fragments de Sesbania ros-
trata. Inocule aux racines et aux tiges de Sesbania rostrata, ce mutant induit des nodules
anormaux et ir&uliers. Lorsqu’on introduit dans ce mutant le clone pRG910-16 qui per-
216
Table 1
: Activités de la
-galactosidase avec et sans induction des
souches ORS 571 portant des fusions 1acZ dans le locus Nodl
ORS 571 avec différentes
Activités de la pgalactosidase
fusions lac Z
(unité!-gai)
-Nar
+Nar
ORS571-Mu63
22
650
ORS571-Mu3
11
205
ORS571-Mu22
1 7
240
ORS571-Mu21
1 8
470
ORS571-T20
1680
Les unités de b-galactosidase ont été mesurêes (Miller 1972) 12 heures
après l'addition de 10pM de naringénine (Nar) à des cultures de ORS571
en phase logarithmique.
met l’induction des fusions lu& dans A. twnefaciem, il est complémenté pour tous ses
phénotypes mutants.
Ces deux approches indiquent la présence sur le fragment de 16 kilobases d’une
fonction r6guMrice ntkessaire à l’expression des gènes de nodulation du locus Nodl.
Pour isoler cette fonction, une matagenèse par transposon a été effectuée dans le clone
pRG910-16 ; une population de clones mutés a été introduite dans le mutant ORS 571-
IN- et les colonies qui n’étaient plus induites ont été sélectionnées. Les fragments mutes
ont kti extraits de ces colonies et sont actuellement cartographiks afii d’isoler les g&nes
impliques dans la rt5gulation de la nodulation.
Caract&isation du mutant de Sesbanh rostrata sans site de nodulation de
tige
Chez Sesbania rosrrafa, les nodules de tige apparaissent sur des sites pl-édéter-
mines qui correspondent a des primordia racinaires. Nous avons rkemment obtenu par
mutagen& chimique un mutant de Sesbania rosfraru ne possédant plus de sites de nodu-
lation sur sa tige.
Ce mutant ne prksente aucune anomalie dans sa croissance a&ienne ni dans la
nodulation racinaire. Il a par contre une racine principale plus vigoureuse et des racines
secondaires moins abondantes que colles de la plante normale. Dans les sols inondes,
217
la quantité d’azote fixé par le mutant décroît alors que celle de la plante normale est
augmentke suite a la prtknce de nodules de tige.
Nous avons compare par &ctrophor&se en une dimension les protéines totales du
mutant et celles de la plante normale. D’après les profds &ctrophor&iques, il existerait
une ou deux prot&nes supplementaires chez le mutant Nous proc&ons actuellement a
une analyse approfondie qui devrait nous permettre de rechercher et d’identifier les fonc-
tions qui gouvernent l’initiation et le développement des sites de nodulation de tige chez
Sesbania rostrata.
Discussion et conclusion
L’analyse génétique et moléculaire de la souche d’Azh%izobim ORS 571 a permis
d’identifier et de caract&iser des gènes impliques dans la nodulation des racines et des
tiges de Sesbania rostrata. Il a été montré que ces gènes ne sont induits qu’en présence
de fragments racinaires ou caulinaires de Sesbania rostruta et que cette induction est
sous la dépendance d’un gbne régulateur. Comme chez Rhizobium meliloti, l’interaction
entre ce r&ulateur et la substance inductrice pourrait determiner l’amorce de la nodulation.
Nous pensons que la capacité de ORS 571 à exprimer pleinement ses genes md dépend
de l’affinité entre la molécule inductrice et ce gène r&ulateur~ Il se pourrait que les g&nes
nod des Rhizobium de racine soient induits par une substaricc pr6sente dans les racines
et pas dans les sites de nodulation de tige.
La découverte de quelques souches de Rhizobium qui n’induisent que des nodules
non fixateurs d’azote sur les tiges de S. rostrata suggère que Azorhizobium ORS 571,
contrairement a ces Rhizobium, possederait des gènes ou des mkanismes spécifiques de
la fixation ou de la transformation de l’azote atrnosphkrique dans les nodules de tige
(g&nes I@, ntr et&). La mise en évidence et l’étude de l’expression de nombreux gknes
fut et nif d’une part (Danèfle et al., 1987 ; Pawlowski et al., 1987 ; Kaminski et al.,
1988) et 1’6tude des mutants de la r@lation de ces gènes d’autre part (de Bruijn et
al., 1988) devraient permettre de rechercher et de comprendre les mkcanismes spécifiques
de la fixation d’azote dans les nodules de tige.
Parrallèlement, de nombreuses recherches ont été entreprises sur S. rostrata afii
de mieux comprendre l’initiation des 6vènement.s prkoces de ht nodulation et le déve-
loppement des nodules chez la plante hôte. Une &ude r6cente (de Lajudie et Huguet,
1988) a rt%le une différence entre les protiines végétales induites dans les nodules racinaires
par la souche ORS 571 et celles induites par une souche de racine. Bogusz et al. (1987)
ont par ailleurs montre qu’il n’y avait pas de difft%e-nce signifîcative entre les composants
de la leghemoglobine extraits des nodules racinaires et ceux issus des nodules caulinaims.
Les travaux de Trinh (com. pers.) ont permis de mettre au point un système de nodulation
in vitro pour suivre avec précision les difftkentes Ctapes de l’infection a la morphogenkse
des nodules
En ce qui concerne le mutant de S. rostrata sans site de nodulation caulinaire,
nous espkons en continuant sa caracttisation mol&xlaire, parvenir à blucider les mdcanismes
de l’induction et du développement des nodules de tige et a isoler les gènes correspondants.
Ces mécanismes seront n&essaires pour crkr des nodules akiens sur d’autres espi?ces
et particuli~ement celles du genre Sesbania utilisées en agriculture et en agroforesterie.
L’association S. rostrata-Azorhïzobiw et Rhizobium apparait comme l’une des
interactions symbiotiques pr4sentant un interêt majeur pour les études moléculaires. Le
218
développement des recherches fondamentales initiées dans plusieurs laboratoires sur cette
association aura certainement des retombées sur la connaissance de l’ensemble des inter-
actions symbiotiques. Par ses caractkistiques particulieres, ce système apparait de plus
en plus comme un modele pour la biologie moléculaire de la fixation d’azote.
Réfkrences
BOGUSZ, D., KORTT, A.A. and APPLEBY,C.A. 1987 Sesbania rostruta root and Stern
nodules leghemoglobins : purification andrelationship among the seven majorcomponents.
Arch. Biochem. Biophs. 254, 263-211.
De BRUlJN, FJ., PAWLOWSKI, K., RATET, P., HILGERT, U. and SCHELL, J. 1987
The unusual symbiosis between the nitrogen fixing bacterium ORS 571 and its host
Sesbuniu rostrata : regulation of nitrogen fixation and assimilation genes in the free-
living versus symbiotic state. In : Molecular genetics of plant-microbe interactions
(D.P.Y. Verma and N. Brisson, Eds).
De LAJUDLE, P. and HUGUET, T. 1988 Plant gene expression during effective and inef-
fective nodule development of the tropical stem nodulated legume Sesbunia rostruta.
Plant Mol. Biol. 10, 537-548.
DENEFLE, P.. KUSH, A., NOREL, F., PAQUELIN, A. and ELMERICH, C. 1987 Bioche-
mical and genetic analysis of the nijHDKE region of Rhizobium ORS 571. Mol. Gen.
Genet. 207, 280-287.
DREYFUS, B., ALAZARD, D. and DOMMERGUES, Y.R. 1984 Stern-nodulating rhizobia.
Current perspectives in microbial ecology (MJ. Klug and C.A. Reddy Eds.) p. 161.
DREYFUS, B., GARCIA, J.L. and GILLIS, M. 1988 Characterization of Azorhizobium
cuulinodans gen., sp. nov., a stem nodulating nitrogen fixing bacterium isolated fiom
Sesbania rostrata. Znt. J. Syst. Bacteriol. 38, 89-98.
DUHOUX, E. and DREYFUS, B. 1982 Nature des sites d’infection par le Rhizobium
de la tige de la légumineuse Sesbania rostrutu . CR. Acad. Sci. Paris. 294, 407-
411.
JOHN, M., SCHMIDT, J., WIENEKE, U., KONDOROSI, E., KONDOROSI, A. and
SCHELL, J. 1985 Expression of nodulation gene nodC of Rhizobium meliloti in Esche-
richia coli : role of rhe nodC gene product in nodulation. Embo J. 4, 2425-2430.
KAMINSKI, PA., NOREL, F., DESNOUES, N., KUSH, M., SALZANO, G. and ELME-
RICH, C. 1988 Characterization of the frxABC region of Azorhizobium caulinoduns
ORS 571 and identification of a.new nitrogen fixation gene. Mol. Gen. Genet. 214,
496-502.
KONDOROSI, E., BANFALVI, Z.S. and KONDOROSI, A. 1984 Physical and genetic
analysis of a symbiotic region of Rhizobium meliloti : identification of nodulation
genes. Mol. Gen. Genet. 193, 445-452.
PAWLOWSKI, K., RATET, P., SCHELL, J. and De BRUIJN, F.J. 1987 Cloning and
characterization of ni$A and ntrC genes of the Stern nodulating bacterium ORS 571,
the nitrogen fixing symbiont of Sesbuniu rostrutu : regulation of nitrogen futation
(nit) genes in the free living versus symbiotic state. Mol. Gen. Genet. 206, 207-219.
219
STACHEL, S., AN, G., FLORES, C. and NFSlXR, E.W. 1985 A Tn3ZucZ transposon
for the random generation of B-galactosidase gene fusions : applications to the analysis
of gene expression in Agrobacterium. Embo J. 4, 891-898.
TSIEN, H.C., DREYFUS, B.L. and SCHMIDT, EL. 1983 Initial stages in the morphoge-
nesis of nitrogen fming stem nodules of Sesbania rostrata . J. bacteriol. 156, 888-
897.
VAN DEN EEDE, G., DREYFUS, B., GOETHALS, K., VAN MONTAGU, M. and
HOLSTERS, M. 1987 Identification and cloning of nodulation genes kom the Stern
nodulating bacterimn ORS 571. Mol. Gen. Genet. 206, 291-299.
AMELIOR.ATION
DE LA FIXATION SYMBIOTIQUE
DE L’AZOTE
PAR LES PRATIQUES CULTURALES
~Uaxbnisez la FBA pour la Production Agricole et Forestière en AfXque
Amélioration de la Fixation Biologique
de 1’Azote (N,) par les pratiques culturales
GANRY F.
Ingttkiew aé Recherche à IRAWIRAD
L’augmentation de la production d’une 16gumineuse annuelle, telle que l’arachide, fait
appel a des pratiques culturales appropriees qui doivent aboutir à une augmentation du
rendement 21 l’hectare et à un maintien du niveau de fixation symbiotique de l’azote (NJ
le plus élevé possible. C’est en vue d’améliorer la quantité d’azote frxé par la l&umineuse,
qui se differencie du premier objectif (l’augmentation du rendement) en cc sens qu’il
place la fixation de l’azote en priorite, que nous raiso~exons les diff&wes techniques
CUltieS.
La diminution de la quamit. d’azote fmt5 par une culture de 1Qurnineuse r&wlte de
l’effet de contraintes qui affectent M des paramètres suivants ou les deux à la fois :
(i) la croissance vegetative avec une action indirecte sur la fucation de l’azote et (ii) la
fmtion de l’azote avec ou sans action dépressive sur le rendement. Pour lever ou éviter
ces diff&entes contraintes, les agronomes doivent être en mesure de proposer des techniques
culturales applicables dans l’agriculture du pays, et capables de maintenir ou de redmsser
le niveau de fixation de l’azote dans les conditions d’environnement de la légumineuse
cultiv6e. Ceci pose le probleme de l’applicatioon des m6thodes de quantifîcation de la
fixation de l’azote in situ. Ces contraintes ont t% relativement bien analysées dans les
principales zones écologiques du S6n6gal pour l’arachide et le soja (Dommergues et Ganry,
1986 ; Ganry, 1987).
Les m&hodes de quantitïcation in situ se r6sument a deux types : les méthodes fondees
snr l’analyse chimique de l’azote seulement, et les m&.hodes fondees sur l’analyse chimique
et le marquage isotopique. Les premi&res prennent en compte sur une p6riode donn&
de t ann&s la variation du stock d’azote total du sol (AN=NO- NJ ainsi que les inputs
et outputs connus d’azote dans I’agrosysteme : elles permettent seulement de savoir si
le système est le siege de pertes ou de gains d’azote, les gains d’azote (inputs non connus)
Ctaut dus essentiellement a la fixation de l’azote. Les secondes méthodes permettent la
222
quantification pr6ci.w de la fixation de l’azote dans les parties ah-iennes et partant, une
quantikation de cette fixation dans les racines.
Analyse des contraintes pouvant affecter la quantit6 d’azote fixe
Ces contraintes peuvent affecter directement la fixation de l’azote, ou indirecte-
ment celle-ci par l’intermediaire de la plante hôte, essentiellement par r&duction de l’en-
racinement et de l’activit6 photosynthétique. Des techniques culturales appropriees devront
être appliquees pour lever ces contraintes.
La contrainte climatique
Cette contrainte est essentiellement hydrique. Trois voies d’étude sont poursuivies
pour résoudre le problème du deficit hydrique dont l’importance au Sénegal est crois-
sante du sud ou nord :
(i)
Choix des variétés à cycle végétatif adapté à la longueur utile de la
saison des pluies (Dancette 1979) : Il s’agit là d’une action indirecte
sur la fixation de l’azote.
(ii) Amélioration de la résistance a la &cheresse de la I@umineuse, d’autant
plus importante que la du& du cycle est raccourcie (Bockelee-Morvan
er af., 1974 ; Gautreau, 1977) : Il s’agit la également d’une action
indirecte sur la fixation de l’azote.
(iii) Amélioration des techniques permettant une meilleure alimentation hy-
drique (Kuo and Boersma, 1971 ; Sprent, 1972 ; Pankhurst and Sprent,
1975 ; Pate 1975) : Il s’agit Et d’actions directe et indirecte sur la
fixation de l’azote. A titre d’exemple, nous analyserons l’effet d’un
stress hydrique en cours de cycle sur une arachide. On sait d’une part
qu’un stress hydrique en saison de culture provient davantage de l’épui-
sement du stock d’eau disponible dans le sol que de l’accroissement
de la demande Cvaporative, de sorte que ce stress va affecter d’abord
la zone racinaire et les nodosit&. D’autre part, on sait que lorsque les
conditions environnememales ne permettent plus a la plante de satis-
faire à la demande évaporative, elle peut s’adapter en fermant ses sto-
mates mais ce faisant, elle interrompt sa photosynthèse et prive
particuli&rement de chaines carbonn& les 16gions les plus 6loigneeS
de la source, notamment les nodosités. Ainsi, lorsque des symptômes
de fl&rissement temporaire apparaissent au champ a la p&iode la plus
chaude de la journée, la fixation de l’azote a deja 6t6 fortement affectée
par carence hydrique et par carence énergétique. Les interventions
peuvent poursuivre deux objectifs complementaires : l’accroissement
du stock d’eau dans le sol en supprimant les pertes par ruissellement,
par évaporation et par transpiration des adventices et l’amelioration de
la capacité! de la plante a utiliser cette eau par amélioration du système
racinaire
223
La contrainte fertilit6
En zone soudano-sahélienne, on observe depuis une quinzaine d’ann6es une baisse
inquiktante des rendements de l’arachide, imputable en grande partie a la baisse de la
fertilité des sols. Nous savons que le phénomene de dQ5rissement de l’arachide est cause
par une in&Ganw de la fixation de l’azote liée vraisemblablement à une dégradation
de l’environnement edaphique qui n’est pas sans rapport avec cette baisse de fertilité.
Sans prbendre que toute la baisse de fertilité soit due aux pertes min&ales, on peut les
considerer comme essentielles (Blondel, 1970 ; Pieri. 1976 ; Wey et al., 1977).
La contrainte parasitisme
Elle est la charnière entre la contrainte climatique et la contrainte «techniques cul-
turales~. En effet, la virulence d’une maladie donnt% dépend a la fois du climat et des
techniques culturales adopt&s sur l’exploitation (rotation, association d’esp&es ou de variM.s)
favorisant le cycle des parasites. Par exemple, le climat humide de la Casamance dans
le sud du SCn&al favorise la rosette de l’arachide ; de plus, les semis tkhelo~és qui
y sont possibles, accentuent la maladie.
La contrainte *date de semis.
La n&essit& & semer à la date voulue se heurte souvent à un probleme de calendrier
cultural sur l’exploitation ou le manque de mat&iel et de Mail de traction, conduisant
a reporter une partie des semis au delà de cette date la plus favorable. Il en résulte des
pertes de rendement dues essentiellement à un defaut d’enracinement Les pertes de rende-
ment sur l’arachide ont pu être estim6e.s dans le Centre Sud Sénégal (Sine-Saloum) à
50 kg/hzv’jour de gousses dans la première quinzaine de juillet (Tourte, 1974)
La contrainte microbiologique
Parmi les microorganismes impliqués dans la fixation de l’azote des l&umineuses,
il y a bien évidemment les rhizobiums, partenaires obligatoires dans la symbiose, mais
aussi, nous faisons l’hypoth&se que dans les conditions édaphiques des sols pauvres 6tu-
diés, les champignons endomycorhiziens à vésicules et a arbuscules (h4VA) pourraient
jouer un r6le important (Dommergues, 1977).
L’intervention sur la biomasse rhizobiale du sol peut être r6alis6.e pour trois
raisons :
a) Les sols n’ont pas de rhizobiums : l’inoculation est alors nécessaire dans
ces sols ; c’est le cas du soja au St%gal 6tudi6 par Wey (1983).
b) Les sols sont d&àvorables a la nodulation et a la fixation de l’azote bien
que poss&ant encore des rhizobiums : on doit y remedier afin d’bviter
le d@%issement des cultures de 1Cgumineuses ; c’est le cas du d@&is-
sement de l’arachide au Sénégal sur laquelle l’inoculation a Cté 6tudiée
par Wey (1974) et par Drevon (1981).
c) Les sots sont bien pourvus en rhizobiums : dans ces sols où la fixation
de l’azote est normale, on peut theoriquement améliorer I’efficience du
syst2me fixateur en substituant à ces souches autochtones une autre souche
224
à condition de résoudre le problème de la comp&ition : c’est le cas de
l’arachide au Sénégal &udiee par Ganry et Wey (1975) et par Ganry
et NDiaye (1977).
L’intervention sur fa biomasse mycorhizienne peut être envisag6e dans les sols
où le potentiel d’infection mycorhizienne est réduit (Ganry et ~2. 1982. 1985).
L’Appllcatlon de techniques culturales approprkes permet de rbdulre les
effets néfastes des dlffbrentes contraintes
Ces techniques sont essentiellement la culture de vari&% a cycle v6ggétatif adapte,
le semis a une date optimale, la fertilisation minkale, le travail du sol, l’amendement
organique et cakique, l’inoculation avec des rhizobiums et éventuellement des MYA, la
~tatiOll.
Comment mesurer quantitatlvement l’effet des techniques culturales sur la
fixation de l’azote et plus g6nbalement sur I’économle de l’azote ?
Comme nous l’avons tkqué dans l’introduction, seules les méthodes isotopiques
(traçage isotopique et abondance naturelle) permettent d’estimer avec une relative prkision
in situ, la fixation de l’azote. Pour atteindre l’objectif visé d’économie des engrais azotes,
il faut considérer le système de culture dans lequel participe une légumineuse et mettre
en oeuvre la démarche suivante : utiliser au maximum la legumineuse pour enrichir le
pool d’azote du sol et d’autre part déplacer le flux de consommation de l’azote du sol,
de la @umineuse (en Lui assurant cependant le minimum vital pour sa productiviti) vers
la c&eale qui doit exploiter au mieux ce pool mobilisable d’azote du sol en vue d’écono-
miser les engrais azou%. Par ailleurs, il faut chercher à tiuire le plus possible les pertes
d’azote provenant de l’engrais. Cette stratkgie s’applique au systkme de culture composé
de la c&t!ale et de la 16gummeuse en rotation.
Pour d&ïnii ce systkme de culture kconome en engrais azoté, en l’occurrence la
rotation légumineuse &dale, nous devons alors poursuivre trois objectifs :
a) Pour la IQumineuse, il faut réduire le plus possible la fourniture d’azote
par le sol durant la phase où elle est capable de fixer l’azote (NJ.
b) Pour la c&%le, il faut : (i) accroître la fourniture d’azote par le sol
incluant l’azote fixé (en vue de rkhtire l’apport d’engrais azoté) et (ii)
r6duù-e les pertes d’azote engrais.
Cette démarche mkthodologique requiert la quantification des flux d’azote dans
l’agrosystéme, en particulier les pertes d’azote engrais, la fixation de l’azote et la quantid
d’azote sol prélevé par la plante. Les methodes isotopiques 15N permettent ces trois
quantifications. Au Sénégal, cette démarche a 6te appliquke aux systèmes mil-arachide
et maïs-soja (Ganry, 1987).
Conclusion
En guise de conclusion, nous prendrons l’exemple de la culture arachidikre.
L’analyse des contraintes agronomiques et climatiques pouvant affecter la produc-
tion arachidière met en relief : (i) l’importance de la contrainte hydrique, a caractkre
225
non prévisible. avec comme premii%e conskquence un defrcit prononce de la fixation de
l’azote ; (iii la contrainte fertilité, à carackre pr&isible et graduellement menaçante en
culture traditionnelle : la manifestation la plus importante de cette baisse de fertilite est
la carence en azote, attribuée aux faibles rkserves azotées des sols sableux tropicaux
(faible teneur en matik organique) mais aussi et surtout à l’acidit6 du sol dont l’effet
inhibiteur sur la fixation de !’ azote en de@ de pH = $5 se manifeste aussi par une carence
en azote.
Il apparait donc que dans des conditions de stress hydrique et d’acidité du sol
(pH = 55). la quantiti d’azote fixe par une culture d’arachide est doublement r&h&
en raison de l’action conjugwk de ces facteurs d@ressifs sur la croissance vegetative
et sur la fixation de l’azote.
Les interventions culturales appropriees devront d’abord rechercher des types de
plantes aptes a n%ister ou a Cviter la sécheresse et éventuellement r&ister à la toxicité
ahrminique. Une fois le choix variCtal fait, un certain nombre de techniques devront être
appliquées au sol : travail du sol et amendement et éventuellement inoculation avec des
microorganismes. Il importe de savoir maintenant quel est le degr6 d’application et d’ap-
plicabilid de ces techniques en milieu paysan .
En ce qui concerne le choix des vari&% adaptks et de la date de semis la plus
prkoce, il ne soulkve pas de problemes majeurs car chaque agriculteur mesure bien les
risques qu’il encourt en ne respectant pas ces choix.
En ce qui concerne les techniques de labour, d’amen&ment organique et d’amen-
dement calcique, leur application est capitale puisqu’elle conditionne la productivité du
sol et le maintien du patrimoine foncier. Malheureusement, cette application se heurte
a de nombreuses difficultks. En effet, trks peu de paysans labourent leurs champs en
fm de cycle cultural et y apportent de la matik?re organique par enfouissement des pailles
ou apport de fumier. Ce probleme doit être examine dans le cadre du systkme agraire
ayant pour objectif l’int&ation du troupeau dans l’exploitation.
Quant aux diftïcultks d’amendement calcique des sols, elles sont dues à un pr-
oblème de coût et d’approvisionnement en chaux, qui, au SCnégal, malgrk la pr6sence
d’une usine de chaux demeurent un obstacle a son utilisation par l’agriculteur.
En ce qui concerne l’inoculation par le rhizobium, elle n’a pas dkpass6 le stade
exp&imental pour l’instant au SénCgal. Des recherches sont menées en vue de son ap-
plication par l’agriculteur. Neanmoins, nous consid&ons que l’inoculation ne devrait être
envisagee, dans le cas de l’arachide que lorsque les techniques culturales prkonisks
auront Cti rationnellement appliquées.
Rc!f&ences
BLONDEL, D. 1970 Relation entre le « nanisme jaune » de l’arachide en sol sableux
(Dior) et le pH. DKrnition d’un se& de l’activit6 du Rhizobiwn. Agron. Trop. 25,
589-595.
BOCKELEE MORVAN, A., GAUTREAU, J., MORTREUIL, JC. et ROUSSEL, 0. 1974
R&&ats obtenus avec les variCt& d’arachide rksistantes à la sécheresse eu Afrique
de I’0ue.s~ Oléagineux. 29, 309-314.
DANCE’ITE, C.. 1979 Agroclimatologie appliquée à l’économie de l’eau en zone
Soudano-Sah&ienne. Agron. Trop. 34, 225235.
226
DOMMERGUES, Y. and GANRY, F. 1986 Biological nitrogen fixation and soi1 fertility
maintenance. In : Management of nitrogen and phosphorns fertiliiers in sub-saharan
Afiica. Uzo Morwunye, A. and Paul Vlek L.G. (eds.), Martinus Nijhoff Publishers,
Dordrecht, pp. 95-115.
DREVON, JJ. 1981 A deficiency of the symbiotic nitrogen fixation in a dry tropical
agrosystem. The nitrogen chlorosis of groundnut (Arachis hypogaeu) in Senegal.
In : Nitrogen cycling in West Africal ecosystem. Rosswall, T. (ed.), Proc. of a work-
shop held at BTA Ibadan Nigeria 11-15 Dec., Stockholm, SCOPE-UNEP, pp: 209-
213.
GANRY, F. 1987 Application de la méthode isotopique dans la recherche de systèmes
culturaux c&%le-légumineuse économes en azote In : Evaluation des réserves en a-
zote utilisables par les plantes dans les agrosystèmes tempérés et tropicaux~. Egoume-
nides, C. (ed.), GEMOS (Groupe d’étude de la matiere organique des sols), 8ème
r&mion Montpellier les 6 et 7 mars 1987, IRAT - CIRAD, Montpellier, France.
GANRY, F. et NDIAYE, M. 1977 Action de l’inoculation et action de l’@mdage foliaire
d’une solution nutritive (N-P-K-S) sur la fixation symbiotique et le rendement de
l’arachide. Coordinated research on the use of isotopes in fertihxer efficiency studies
on grain legumes. Research contract No RC/1296SEN of the joint FAOZIAEA Division.
In : Report presented at workshop held in Vienna in December 1977, 16~.
GANRY, F. et WEY. J. 1975 Cordinated research program on the use of isotope in studies
on biological dinitrogen fixation. Research contract No RC/1296 - SEN of the joint
FAOZIAEA Division. In : Report presented at workshop held in Vienna. « Action
de l’inoculation et de la fumure azotée sur le rendement et la fixation symbiotique
de l’arachide ». ISRAZAIEA, 20 p.
GANRY, F., DIEM, H.G. and DOMMERGUES, YR. 1982 Effect of inoculation with
Glomus mosseae on nitrogen fixation by field grown soybean. Plant and Soi1.68,421-
429.
GANRY, F., WEY, J., DIEM, H.G. and DOMMERGUES, YR. Inoculation with Glomus
mosseae improves N, fixation by field-grovn soybeans. Biol. Fert. Soils.1, 15-23.
GAUTREAU, J. 1977 Niveaux de potentiels foliaires intervari&aux et adaptation de l’ara-
chide a la sécheresse au Sedgal. Oleagineux. 32, 324-332.
KUO, T. and BOERSMA, L. 1971 Soil water suction and root temperature effects on
nitrogen fixation in soybeans. Agron. J. 63, 901-904.
PATE, J.S. 1975 Physiological studies on the reaction of nodulated legumes to environ-
mental stress. In : Nitrogen fixation and the biosphere. Stewart, W.D.P. (ed). Cambridge
University press.
PANKHURST, CE. and SPRENT, J.U. 1975 Effects of water on the respiratory and
nitrogen fming activity of soybean mot nodules. .Z. Exp. Bot. 26, 287-304.
PIERI, C. 1976 I) L’acidification d’un sol dior cultivé du Sénegal et ses consequences
agronomiques. II) L’acidification des terres cultivées exondées au S&rnégal. Agron.
Trop. 31, 339-368.
SPRENT, J.U. 1972 Nitrogen fixation by legume subjected to water and high stresses
In : Symbiotic nitrogen fixation in pIants. P.S. Nutman (Ed.) Cambridge University
Press, p 385-405
227
SPRENT, J.U. 1972 The effects of water stress on nitrogen-fixing nodules 4. Effects on
whole plants of Vicia faba and Glycine max. New Physiologist. 71, 603-611.
SIBAND, P. et NICOU, R. 1975 Réflexion d’ensemble sur le problème de l’engrais minéral
dans le bassin arachidier. Rapport ronéo. ISRA-CNRA Bambey, 111 pages.
TOURTE, R. 1974 Les Recherches de I’IRAT au Setrégal. Synth&se 1973. Rapport ronéo
ISRA-CNRA de Bambey, 125 p.
WEY, J. 1974 Inoculation bacterienne des légumineuses au Sénégal. Memoire ENITA
Dijon, France, 47 p.
WEY, J., SIBAND, P., EGOUMENIDES, C. et GANRY, F. 1987 Essai de régénkration
d’un sol de la zone arachidikre du Centre-Nord du Sénégal. Agron. Trop. 42, 2%
268.
WEY, J. 1983 (RI) Inoculation du soja par Bradyrhizobium japonicwn. (II) Production d’ino-
culum : mise au point d’une mtthode simplifitk Agron. Trop. 38, 129-136.
hhhiser la FBA pour la production agricole et forestihe en Afrique
Response of Phaseolus vzdgaris I+ to inoculation
with Rhizobium and fertilization with nitrogen
and phosphorus in Northern Tanzania.
AMIJEE, FP’, EDJE, O.T.@‘,
KOINANGE, E.K.c3’, BITANYI, H.F.f4’,
BRODRICK, S.JP), and GILLER, ICE.“’
(I) : Wye College, Univenify of London, Wye.
word, Kent, TN25 5AH. U.K.
(2) : SADCC-CLAT Regicmal Bean Programme.
P.O. Box 2704. Arushn. Tanzania.
(3) : TAR0 Lyanumgu, P.O. Box 3OM. Uoshi, Tanzania.
(4) : INSRC, P.O. Box 4302, Dar es Salaam. Tanzania
Abstract
In this study we aim to identify soil factors limiting nodulation and bean (Ph-
seoZus vulgaris L) production in Nothern T anmnia Investigations car&d out in
1987 indicated that large numbers (10 cells g-’ soil) of Rhizobium legumimsarm
bv. phaseoli were present in most soils, and rwponse to inoculation alone was poor.
Chemical analyses of soil samples collected from these arens showed that pH, or-
ganic matter content and cation exchange capacity were reasonable, but amount of
soil available P was very small (l-2 mg P kg-‘). In 1988, results based on two on-
farm experiments showed that inoculation with Rhizobiwn (strain CIAT 899) and
application of triple supcaphosphate significantly increased nodulation and yield.
Application on Minjingu rock phosphate also increased nodulation and yield but to
a much lesser extent-
2 2 9
Introduction
Bean (Phaseolus vulgaris L.) is an important grain legume widely grown in the
highland areas (1000 meters above sea level) of East and Central Africa The main bean
producers are small farmers for whom bean consumption contributes up to 50% of dietary
protein (Allen, 1986). The majority of these farmers grey the crop on low fertility soils
with no fertilizer input. Generally under these conditions, effective nodulation with Rhizo-
_ biwn is rarely observed and amount of atmospheric nitrogen fixed is probably small.
Many factors (e.g. genetic variability, eclaphic and agronomic factors) are known to-affeci
nodulation in beans (Graham, 1981). Our objective has been to identify soil factors which
limit nodulation and nitrogen fixation in the field. The investigations reported here were
carried out during two growing seasons (March to August 1987 and 1988) in Northern
Tanzania, one of the principal areas for bean production.
Materials and methods
Experiment I (1987)
On station, experiments on two contrasting soil types (volcanic soil at &lien, and
brown earth at Lambo) were done to test response to nitrogen application and Rhiwbium
inoculation. Two bean varieties of large seeded dwarf (Type 1) were chosen : Canadian
Wonder, a popular variety grown throughout Northern Tanzania, and Lyamungu 85, a
variety recently released by the Tanzania Agricultural Research Organisation (TARO)
Bean Programme. Nitrogen was applied at 0 or 30 kg N ha-’ as (NHJ,SO, at time of
planting, and seeds were uninoculated or inoculated (106 cells per seed) with peat-based
mixed Rhizobium inoculant of strains CIAT 632 and 899. A basal dose of
60 kg P,O, ha-’ as triple superphosphate (TSP) was applied to all treatments. Each treat-
ment had five replicates, and a split plot design was used to reduce the risk of conta-
mination between uninoculated and inoculated treatments. Plot size was 4 m x 6 m con-
taining eight rows 0.5 m apart with one seed sown at 0.1 m interval to give
200,000 plants hx’. Nodulation (nodule number, nodule diameter and score for effec-
tiveness) was recorded up to flowering, and seed yield was measured at harvest. Bean
fields owned by research stations and farmers were visited to assess nodulation and plant
vigour. Soil samples were collected for chemical analyses (MAFF/ADAS, 1986) to deter-
mine variation in soil fertility and to identify factors limiting the establishment of ef-
fective nodulation.
Experiment II (1988)
Two sites @rente, Lushoto and Tropical Pesticide Research Institute (TPRI), Amsha)
which had no record of previous fertilizer application were selected for an experiment
to test response to Rhizobium inoculation, nitrogen application and phosphate fertilization.
Lyamungu 85 variety was either uninoculated or inoculated with peat based inoculant
(CIAT 899) to give 106 cells per seed and N applied at 0 or 30 kg ha-’ as nitrochalk
(21% N). Two sources of P were used ; Minjingu rock phosphate (MRP) containing ap-
proximately 30% PZO, or triple super phosphate (TSP) containing 45% PZO, at equivalent
application rates of 0, 60, 120 or 180 kg PZO, ha-‘. The rock phosphate is a sedimentary
guano deposit located at South West of Arusha. N. Tanzania (Scaife, 1968). and has
230
been recommended as an alternative source of P costing a thiid of the price of TSP.
A split plot with two replicates was used to avoid contamination between uninoculated
and inoculated treatments. Plot size was 4 m x 4 m containing eight rows 0.5 m apart
with two seeds sown at 0.2 m interval to give 200,000 plants ha-‘. Measurements included
nodule number before flowering, vesicular arbuscular mycor&zal infection (Phillips and
Hayman, 1970) at pod development, and final seed yield at harvest. Soil pH and bicar-
bonate soluble P (available P) were also determined at final harvest Rainfall was poor
at TPRI, and moisture stress experienced after flowering reduced final yield and may
offset potential effects of the treatments.
Results and discussion
Experiment I (1987)
At &lien, nodulation was excellent (up to 100 nodules per root system, 2-3 mm
in diameter and effective, i.e. pink in colour) in all treatments including uninoculated
plants which nodulated well with indigenous Rhiwbim Earlier on, nodulation was sup-
pressed in treatment receiving N. However this effect was not found at later harvests.
Mean seed yield was 1350 kg ha-’ with no increase due to N application or inoculation,
suggesting that some other factors were limiting yield. A nutrient disorder resemblig po-
tash deficiency was observed, and work is in progress to identify the cause. At Lambo,
nodulation was poor in all treatment (up to 20 nodules, 1 mm in diameter and white
in colour). Mean seed yield was 840 kg ha-’ and did not differ with treatments. This
site experience prolonged moisture stress which would have offset any potential effect
of N application or inoculation. In 1988, Lamb0 received good rainfall, and nodulation
was exceellent.
Enumeration of Phaseolus rhizobia (Vincent, 1970) showed that up to 104 cells
g1 soil were present at both sites. Singleton and Tavares (1986) have shown that unless
the native population of Rhiwbium is less than 102 cells g’, response to inoculation is
not common. Hence it is not surprising to record no response to inoculation in our studies
at Salien or Lamb0 where indigenous population exceeded this number using the strains
CIAT 632 and 899 which had not been shown to have exceptional competitive ability.
Generally, bean plants had good nodulation and plant vigour when raised on fertile
volcanic soils (e.g. Salien, Table 1). In contrast, nodulation and plant vigour were poor
in the more common reddish brown soils. Chemical analysis of these soils (Table 1)
showed that concentration of available P (bicarbona~soluble) was small (l-2 mg kg
i) and that this might be a limiting factor for plant growth and nodulation. No other
single factor could be identified as the cause for pocx nodulation, Soil acidity has been
shown to depress nod&ion, however our data indicated that most soils from this region
were not very acidic @H 5.5, Table 1). organic matter content was average (3%) and
cation exchange capacity also reasonnable (30 meq/lOO g), and there was no indication
that nodulation was related to either of these (Table 1).
In addition to low concentration of soil P, the data clearly showed that soils from
research stations had unrealistically high levels of fertility and did not resemble those
Ii-am farmers’ fields (Table 1). Hence the results of Experiment I, done on sites belonging
to research stations, are not comparable wth the situation in the farmers’ field. It was
also evident that although large numbers of PhaseoZus rhizobia were often present in
the farmers’ fields, nod&&ion and plant vigour were poor and this might be due to the
small amount of available soil phosphate.
232
Experiment II (1988)
At both Erente and TPRI, inoculation with Rhizobium and P application as TSP
markedly (P < 0.01) increased nodulation (Tables 2 and 3), and yield (Tables 4 and 5).
Inoculation alone also increased nodulation and yield, but it was not statistically signi-
ficant Nitrogen application suppressed nodulation (Tables 2 and 3) and sometimes in-
creased yield (Tables 4 and 5). Phosphorus when applied in the form’ of TSP increased
nodulation and yield (Tables 2,3,4 and 5). Similar findings have been reported in Kenya
(Ssali and Keya, 1983). MRP also increased nodulation and yield but to a lesser extent
(Tables 2,3,4 and 5). The highest level of P applied (180 kg P,O, ha-i) gave the maximum
yield (Tables 4 and 5) thus potentially yield might be increased with further applica-
tion of P, although such amounts would not be economical and rarely used by the farmer.
There was no change in pH, but amount of available P had increased following increa-
sing rates of P application (Table 6), more so at TPRI than at Erente. On average, only
20% of the root system was colonized by vesicular arbuscular mycorrhizas, a fungal root
symbiosis which plays an important role in uptake of P from the soil. There was no
real difference in the amount of colonization between sites or treatments. The small extent
of colonization is surprising. For bean, roots are highly susceptible to mycorrhizal infec-
tion (Halley and Peterson, 1979). When the soil was examined, a few spores of Glomus
monosporum were present. It would be justifiable to test response to mycorrhizal inocu-
lation, especially in association with rock phosphate, if such small spore populations are
found to be common.
The preliminary results presented here are now part of a long-term investigation
to study effects of phosphorus, in particular the residual effect of MFW, an important
soil factor limiting bean production.
Conclusions
Large populations of indigenous Rhizobium phaseoli are present and yield responses
due to Rhizobium inoculation are not common in these soils, even though nodulation
may be increased.
Soil phosphate is a major factor limiting nod&ion and Phuseolus production.
By raising the soil available P, it may be possible to form effective nod&ion
from the existing indigenous population of Rhizobium, which appear to be equally effi-
cient as the inoculation strain used here.
In this fmt year of study, TSP was a more effective source of P fertilizer in compa-
rison to h4RP. Future studies will show whether the residual effect from MRP can give
similar yields obtained with TSP.
Acknowledgement
This collaborative project is funded by the U.K. Overseas Development Adminis-
tration (R4279A).
233
Table 2 :
Nodule number (m-2) at Erente (28 OAP) comparing inoculation
(+ Inoc), application of nitrogen (+N = 30 kg N ha -1) and
Minjingu rock phosphate (MRP) or triple super phosphate (TSP).
Level of P
MRP
TSP
Kg P2O5 ha-l
Inoc
-N
+N
-N
+N
-
-
0
721
189
721
189
60
558
319
1031
412
120
1119
281
1812
357
180
779
476
1544
1235
0
+
759
300
759
300
6 0
+
584
256
1970
906
120
t
830
111
1248
596
180
+
952
241
1259
819
Standard Error
82.8
234
Table 3 : Nodule number (m-21 at TPRI (37 DAP) comparing inoculation
(+ Inoc), application of nitrogen (+N = 30 kg N ha -11 and
Minjingu rock phosphate (MRP) or triple super phosphate (TSP).
MRP
TSP
Level of P-
Inoc
Kg P205 ha 1
- N
+ N
- N
+ N
-
-
0
689
300
689
300
60
1090
408
1631
633
120
1022
437
1858
812
180
1159
593
1873
740
0
+
604
226
604
226
60
+
828
259
1626
588
120
+
1013
367
1847
536
180
+
1234
533
2627
951
Standard Error
52.8
235
Table 4 : Seed yield (kg ha-11 at Erente (109 DAP) comparing inoculation
(2 Inoc), application of nitrogen (+N = 30 kg N ha-11 and
Minjingu rock phosphate (MRP) or triple super phosphate (TSP).
Level of P-
MRP
TSP
Kg P205 ha 1
Inoc
- N
+N
- N
+ N
0
421
528
421
528
60
552
596
530
552
120
546
500
668
698
180
398
540
797
866
0
t
367
563
367
563
6 0
t
500
476
622
546
120
+
599
494
586
632
180
t
449
608
790
962
Standard Error
24.6
-’
:‘.
:
‘.
236
Table 5 : Seed yield (kg ha-11 at TPRI (93 DAP) comparing inoculation
(2 Inoc), application of nitrogen (+N = 30 kg,N ha-l) and
Minjingu' rock phosphate (MRP) or triple super phosphate (TSP).
Level of P
MRP
TSP
kg P2O5 ha-l
Inoc
-N
tN
- N
tN
0
158
274
158
274
60
249
221
314
294
120
221
279
323
389
180
276
269
391
438
0
+
181
246
-181
246
60
+
237
264
286
342
120
+
215
279
319
332
180
t
262
408
338
388
Standard Error
10.1
237
Table 6 : Change with application (kg P205 ha-l) of Minjingu rock phosphate
(MRP) or triple super phosphate (TSP) in soil pH,and bicarbonate
soluble P (mg P kg-l) measured at Erente (109 DAP) and TPRI
(93 DAP).
Source and level of applied P
MRP
TSP
0 60 120 180
0 60
120
180
-
-
-
-
-
-
-
-
Erente
pH
5.3 5.3 5.2 5.4
5.3 5.3
5.3 5.3
P
2.2 2.8 4.7 6.2
2.2 4.2
4.7 6.5
TPRI
pH
6.6
6.8
6.6
6.7
6.6
6.6
6.7 6.7
P
1.4
3.9
10.2
11.8
1.4
5.8
14.3
19.0
238
References
ALLBN, DJ. 1986 Bean production systems in Africa CIAT, Cali, Colombia.
GRAHAM, P.H 1981 Some problems of nod&&ion and symbiotic nitrogen fixation in
Phaseolus vulgaris L. A review. Field Crops Res. 4, 93-112.
HOLLBY, J.D. and PETERSON, R.L. 1979 Development of a vesicular arbuscular mycor-
rhiza in bean roots. Can. J. Bot. 57, 1960-1978.
MAFF/ADAS 1986 The analysis of Agricultural materials. RB427. HMSO.
PHILLIPS, JM and HAYMAN, D.S. 1970 Improved procedures for clearing roots and
staining parasitic and vesicular-arbuscular mycorhizal fungi for rapid assessment of
infection. Trans. British Mycol. Sot. 55, 158-161.
SWIPE, A 1968 The effect of a cassava <<fallow* and various mammal treatments on
cotton at Ukiriguru, Tanzania. East A$ican Agric. and Forest. J. 33, 231-235.
SINGLETON, P.W. and TAVARES, J.W. 1986 Inoculation response of legumes in rela-
tion to the number and effectiveness of indigenous Rhizobium populations. Appl. En-
viron. Microbial. 51, 1013-1018.
SSALI, H and KEYA, S.O. 1983 The effects of phosphorus on nodulation, growth and
dinitrogen fixation of beans. Biol. Agric. Hort. 1, 135-144.
VINCENT, J.M. 1970 A manual for the practical study of root-nodule bacteria. IBP Hand-
book W 15, Blackwell Scientific Publications, London.
h4aximiser la FBA pour la production agricole et forest&e en Afiique
Variation in acid-Al tolerance -
of Bradyrhizobium japonicum
ASANUMA, S.(l) and A-ABA kc2)
(1) : Hokkdo National Agricultural Eqwiment Station,
1, Hitsujigaoka, Toyohira-ku, Sapporo. o@$ Japan.
(2) : De1 Monte Corporation, Walnut Creek Research Center.
P.O. Box 9004. Wabu~ Creek, California 94598. USA.
Abstract
Seventy-six stmins of Braa’yrhizobium japonicwn (54 African and 22 exotic)
were examined for tolerance of acidity @H 4.5). low P (5 @VI) and high Al (50
and 100 @f) by an agar plate method. Fortyfour strains were tolerant to acidity
regardless of P levels (1.000 or 5 @vi) of the medium and 22 were sensitive. The
remaining 10 strains differed in tolerance of acidity depending on level P of the
medium;nineweretolelantoflowP(5@vf)andoneofhighP(1.OOOcLM).All
of 21 strains which grew on the most severe stress used consisting of low P and
high Al (100 P.M) at pH 4.5 showed tolerance of acidity at both levels of P. Isolates
from a highly acidic soil @H 4.6, HaO) from Gnne, southeastern Nigeria, showed
different levels of acid-Al tolerance but tended to be more tolerant than those from
a slightly acidic soil (pH 6.5, HaO) from lbadan, southwestern Nigeria However.
20 % of the Omre isolates and 30 % of the Ibadan isolates were sensitive to acidity.
Thus, tolerance of stmins was not necessarily determined by the acidity of the soils
in which they reside. Gum production of certain strains occured at low P. but this
had no relation to acid-Al tolerance.
The great variation in acid-Al tolerance observed among strains of B. japonicwn
supports the strategy to select strains for use as inoculants in acid soils.
240
Introduction
Performance of inoculated strains of Bradyrhizobium japonicum in a highly acidic
soil (Ultisol, pH 4.6, %O) from Gnne, southeastern Nigeria, was good in terms of nodu-
lation of soybeans (Glycine mar (L.) Merr.) in the first year but was poor in the second
promfield and Ayanaba, 1980 ; IITA, 1981). It was suggested that some acid soil factors
are detrimental to the ecological behavior of introduced rhizobial strains and that use
of acid tolerant strains could lead to better utilization of the benefits of biological nitro-
gen fixation to grow soybeans under such acid conditions.
A great variation in tolerance of several acid soil factors has been shown among
strains of Rhizobium leguminosarum biovar tnifolii (Thornton and Davey, 1983 ; Thurman
et al., 1985). R. leguminosurum biovar phaseoli (Munns et al., 1979 ; Graham et al.,
1982 ; Lowendorf and Alexander, 1983), R. meliloti (Howieson et al., 1988) ; R. loti
and Bradyrhizobium sp. (lotus) (Cooper, 1982), Bradyrhizobium sp. (cowpea group) (Date
and Halliday, 1979 ; Keyser and Munns, 1979a and b ; Keyser et al., 1979 ; De Carvalho
v., 1981 ; Munns and Keyser, 1981 ; Ayanaba et al., 1983) and B. japonicum (Keyser
and Munns, 1979a and b ; Ayanaba et al., 1983). Little information, however, is available
on native African strains of B. juponiczun. Ayanaba ef al., (1983) reported that some
African rhizobia isolates nodulating both cow-pea (Vigna unguiculata (L.) Walp.) and soy-
bean cultivar TGm 344 showed wide variation in tolerance.
Tolerance of strains of rhizobii to acid soil factors is generally tested by growth
in defined liquid media (Date and Hall&y, 1979 ; Keyser and Munns, 1979a and b ;
Howieson, 1985). Keyser and MUMS (1979a) reported that in the test of tolerance of
acid soil factors with Bradyrhizobium sp. (cowpea group) and 8. japonicum, Al
(25-50 p.M) was probably a more severe stress than low Ca (50 @VI) and/or high Mn
(200 @VI). Thornton and Davey (1983) showed that in liquid culture media, growth of
R. trjfoliti was more limited by high Al (15 to 40 p.M) than by low P (1 to 6 pM).
Ayanaba et al., (1983) proved that screening of Rhizobium for tolerance to acid-Al stress
could be achieved by an agar plate method and that this method had an advantage in
handling large numbers of rhizobia.
In the present paper we report variation in tolerance of acid-Al stress among strains
of B. jczponicum isolated mostly from African soils. Acidity, low P and high Al were
factors examined by the agar plate method.
Materials and methods
Rhkobial strains
Seventy-six strains of B. juponicum from the culture collection of the Intematio-
nal Institute of Tropical Agriculture @TA), Ibadan Nigeria were used. Fifty-four of them
were isolated from African soils, 20 of which were from either soybean promiscuous
(Malayan, Orba or 4H/149/1) or non-promiscuous cultivars (TGm 294 or Bossier) grown
in a highly acidic soil of Gnne. Twenty were from the same cultivars grown in a slightly
acidic soil @H 6.5, %O) of Ibadan, southwestern Nigeria. Some of the African isolates
nodulate cowpeas (IITA, 1979). Three wild-type strains and their antibiotic resistant mu-
tants were included. One of them is resistant to spectinomycin and the other two to strepto-
mycin. All strains were cultured on slopes of yeast extract-mannitol agar medium (YJZM
agar) (Vincent 1970).
241
Test agar media
Six media prepared according to Keyser and Munns (1976b) were used : full -
defined ; acid ; low P ; low P, acid and two levels of high Al (50 and 100 pM). Each
medium was uspplemented with 15 g/l ionagar (Oxoid Co. LtrL, England). Each medium
contained the basal solution of Keyser and Munns (1979b). Five hundred @vI each of
KH$O, and K.$PO, and 1.1 g/l Na-glutamate were added to the two full-defined media,
whereas 5 pM KH$O,, 1.5 mM KC1 and 1.1 g/l Na-glutamate were added to the other
four media. Fulldefmed and low P media were adjusted to pH 6.8 with lN-NaOH and
the others were adjusted to pH 4.5 with IN-HCl before addition of AlK(SO,), and growth
factors (Geyser and Murms, 1979b). Furthermore, methyl red in ethanol was added as
a pH indicator to a final concentration of 0.002%. The media were autoclaved (121”C,
15 mm) before dispensing into plates (30 ml/plate). Autoclaving and the addition of these
substances slightly changed the pH values of the media but it did not affect the testing.
Testing of acid-Al tolerance
Loopfub of rhizobia grown on YEM agar slopes were suspended in sterilized
0.85% saline to give similar turbidity among strains. The six different agar media, pre-
dried for 3 h at 45°C were inoculated by transferring the bacterial suspensions with steri-
lized cotton swabs. Twenty strains were tested in one plate and the experiment was du-
plicated. Obervations were made daily during a lo-day-incubation at 28°C. Tolerance of
rhizobia to acidity or acid-Al stress was evaluated by comparing their growth on the
stress media with that on the full-defined medium. The tolerance of Onne and lbadan
isolates was as designated in Table 1, bases on the levels of severity of the stress.
Table 1
: Designation (I to VII) of the degree of response to acid-Al
stress by Bradyrhizobium isolates from Onne and Ibadan soils ;
media were at pH 4.5.a
Designation
P(uM)
Al
(uM) I II
III IV v VI
VII
-
-
-
-
-
-
-
1,000
0
S
'S
S
T
T T
T
5
0
S T
T
S
T T
T
5
50
s s
T
S
S T
T
5
100
s s
S
. s
s s
T
a:Growth on each medium was compared with that on control medium
(pH 6.8, 1,000 uM, P and OuM Al) and designated as T (tolerant) or S
(sensitive) accordingly.
242
Results and discussion
Most rhizobia grown on laboratory media are known to produce either acidic or
alkaline substances (Norris, 1965 ; Jones and Burrows, 1%9 ; Cooper, 1982, Hernandez
and Focht, 1984), depending on the preferential use of sugars or organic nitrogen com-
pounds, respectively, as the source of energy (Parker et al., 1977). In evaluating the growth
of Rhizobium on such media, therefore, changes in pH have to be monitored Methyl
red was employed in the present experiment for that purpose. About half of the strains
tested produced alkaEne substances when grown on the agar media (Table 2), but. it was
Table 2 : Bradyrhizobium japonicum strains tolerant to acid-Al (lOO@J, pH 4.5)
stress
IITA
Alkali
Colony type
accession
Origin or source
productiona
no.
High P
Low P
(1,OOO)lM)
(5)lM)
' IRj 2001
cv. Malayan. Onne
"db
nd
2002
cv. Halayan, Onne
+
white
watery
2004
cv. Halayan, Onne
+
white
.watery
2007
cv. Orba, Onne
+
white
watery
2013
cv. Halayan, Ibadan
+
white
watery
2019
cv. Orba, Ibadan
+
white
watery
2020
cv. Orba, Ibadan
--
nd
nd
2025
cv. TGmZ94, Ibadan
+
white
watery
2033
cv. Orba, Yandev
white
watery
2035
cv. Orba, Yandev
nd
nd
2038
cv. Malayan. Samaru
nd
nd
2039
cv. Halayan, Samaru
+
nd
nd
2048
cv. TGm294, Samaru
white
watery
2049
cv. TGm294, Samaru
nd
nd
2052
cv. Bossier, Samaru
nd
nd
nd
2109
Univ. of Wisconsin,
nd
nd
USA (SW 61A76)
2119
ENSA. Ivory Coast (63)
+
nd
nd
2120
ENSA, Ivory Coast (GB)
nd
nd
nd
2123
USDA (3llb138)
+
nd
nd
2128
Thailand (TWA21
+
white
not watery
2129
Thailand (THA5)
+
white
not watery
a
Alkali production : +, producing ; -, not producing ; nd, not determined.
b
nd, not determined.
243
observed that their growth preceded the shift in pH of the media. Ayanaba et al. (1983)
observed that strains of rhizobia tolerant of acid-Al stress on agar medium grew signifi-
cantly in liquid stress medium before a change in pH Thus, it seemed that tolerance
of rhizobia to acid-Al stress might not be directly related to their ability to produce alkali.
Rhizobial strains showed great variation in tolerance of acid-Al stress (Fig.1). Most
strains tolerated acidity stress (pH 4.5) similarly regardless of P levels (1.000 or 5 pM)
pH POJM) Al(pMI
4.5 1,000
Figure I : Effects of acidity and various levels of P and Al on growth of B. japonicum inoculated on
ag~mediaGrowthone~hmediumw~examinedby~mparisonwiththaton~ntrolmedium(pH
6.8.
P 1,000 @VI and Al 0 p.M) and designated T (tolerant) or S (sentitive) accordingly. Results are shown
with num& of strains and their proportions in percent in parentheses.
of the medium ; 44 strains were tolerant and 22 were sensitive. The remaining 10 strains,
however, differed in tolerance of acidity depending on the level of P. Nine required
low P 5 @vi) to be torerant and one required high P (1,000 pM) to be tolerant. Thus,
tolerance of rhizobia to acidity seemed to differ depending on the level of phosphorus
content of the medium. Table 3 shows the effect of P levels on the growth of rhizobia
under two pH conditions (6.8 and 4.5). Over 90 % of the strains grew similarly at pH
6.8 regardless of P levels. By contrast at pH 4.5, 25 % or 19 strains grew less at the
high P than at the low P. Nine strains which required low P to grow at pH 4.5 (Fig.1)
were also in this group. Uptake and utilization efficiencies of phosphorus were reported
to be greatly different among strains of rhizobia, which caused differences in low P tole-
rance of rhizobia (Cassman et al., 1981b ; Beck and Munns, 1984). The result shown
in Table 3 suggests that the efficiencies in phosphorus uptake and utilization might
change depending on the pH surrounding strains of rbizobia.
Growth of rhizobia at pH 6.8 and 4.5 was also compared at different P condi-
tions (1.000 and 5 pM) (Table 4). Most strains responded similarly to pH regardless
of P levels, suggesting that phosphorus is less influential on the growth of rbizobia than
pH. This supports observations with R. kifolii by Thornton and Davey (1983).
As shown in Fig. 1, tolerance of acidity did not always mean tolerance of acid-
Al, indicating that tolerance of acidity stress among rhizobia might be enhanced by the
presence of ahuninum. Thirty three of the 53 strains that tolerated acidity at the low
P level could grow under acid-Al (50 pM, pH 4.5) stress ; 21 of these or 27.6 % were
further tolerant of the most severe Al stress tested (100 J.LM). Similar results were obtained
,-. : :,i . .
..,..
244
Table 3 : Effet of P levels on growth of B. japonicum at pH
-
6.8 and 4.5a
PH
Growth in 1,000 )IMP compared to that in 5 PMP
Larger
Similar
Smaller
6.8
2 (2.6)
70 (92.1)
4 ( 5.3)
4.5
2 (2.6)
.
55 (72.4)
1 9 (25.0)
a Growth of strains was compared after lo-day-incubation. Visual
readings on a scale'& &5 were made of$colonies and the differences
of the minimum 2-point-readings were considered significant.
Results are shown with numbers of strains and their proportions
in percent in parentheses.
by Keyser and Munns (1979b) Who found that three of the eight B. jqwnicum strains
that were tolerant of pH 4.5 also tolerated acid-Al (50 p.M) stress. The most tolerant
strains (Table 2) were not restricted to one geographic region of the world
Does acidity of bulk soil determine the acid-Al tolerance of rhizobia isolated l3om
the soil ? To investigate it, acid-Al toll of isolates from a highly acidic Onne or
Table 4 : Effect of pH on growth of B. japonicum at two different
levels of pa
P (uM1
Growth at pH 6.8 compared to that at pH 4.5
Larger
Similar
Smaller
1,000
42 (55.3)
31 (40.8)
3 (3.9)
5
3 2 (42.1)
4 4 (57.9)
0 (0.0)
a See the footnote in Table 3.
245
slightly acidic Ibadan soil (pH 4.6 and 6.5 respectively) were compared (Fig. 2). The
Cnne isolates showed less varied levels of tolerance than those from Ibadan soil and
tended to be more tolerant of acid-AI than the Ibadan isolates, indicating that the acidity
of a bulk soil could affect the acid-Al tolerance of rhizobia residing in that soil. However,
20 % of the Onne isolates were sensitive to acidity (pH 4.5), and so were 30 % of the
Ihhn isolates. Thus, the tolerance of strains seemed not to be necessarily &tern&d
by the acidity of bulk soil from which they were isolated. Jones and Burrows (1969)
also reported no correlation between the final pH of the culture media in which isolates
of R. lrifolii grew and the pH of the soils from which they originated In soil, acidity
or acid-Al conditions of a microenvironment could affect the acid-Al tolerance of the
inhabitants, therefore, further research needs to be carried ..out in this connection.
Rhizobia nodulating non-promiscuous soybean cultivars are known to nodulate pro-
miscuous ones, too, but the reverse is not always true (IITA, 1979 ; Bromfield and Roughley,
1980). It means there is a difference among strains of rhizobia in terms of compatibility
with host cultivars. Promiscuity of soybean cultivars (Nangju, 1980 ; Pulver et al., 1985)
from which the Onne and Ibadan isolates were obtained and the acid-Al tolerance of
these isolates are shown in Fig. 3. Acid-Al tolerance of the isolates was not dependent
on the promiscuity of host plants from which they were isolated.
Certain strains formed watery and translucent colonies; regardless of P levels (1,000
or 5 pM) in the full-defined medium. However, some strains formed white colonies on
the same medium when P was high, and watery and translucent colonies when P was
low (Table 2). Watery colonies were formed presumably because production of copious
amounts of gummy substances occurred when the P level of the medium was low. Cassman
et al., (1981a) also observed abundant production of external gum by strains of B. juponicum
in defined liquid medium with low and moderate P, but not with the high P concentrations
routinely supplied in laboratory media. These results suggest that gum production of certain
rhizobia may be enhanced when the P level of the medium becomes low. In the study
of cowpea rhizobia isolated from three locations in West Africa, two types of colonies,
dry and pinpoint or large and gummy, were frequentIy observed when grown on YEM
agar (IITA, 1981 ; Ahmad et al., 1981) and Ayanaba et al., (1983) found that rhizobia
strains forming the former type of colonies were more sensitive to acid-Al stress than
those forming the latter ones. Cunningham and Munns (1984) showed that high production
of extracellular polysaccharide is positively related to the acid tolerance of rhizobia. In
this study, the relation between gum production and acid-Al tolerance of rhizobia was
not clear. In this connection, How&on et al., (1988) also reported that with R. trifolii,
there was no distinct~ link between polysaccaride production and acid tolerance.
Three wild type strains and their antibiotic r&st.ant mutants were included in this
study. Mutation induced by two antibiotics seemed not to affect the acid-Al tolerance
of rhizobia : all strains were as sensitive or tolerant as the wild types (data not shown).
This result supports the observation by Ayanaba and Wong (1982) that antibiotic resistant
mutants tolerated acidity to the same extent as their parents.
A consistent and stable property of tolerance of rhizobia to acid and aluminum
stresses (Munns and Keyser, 1981; Munns, 1986) and a great variation in acid-Al tolerance
among strains of B. juponicum observed in this study prove that strains can be selected
for possible use as soybean inoculants in acid soils. It is obvious, however, that there
are some other adverse, abiotic stresses like high temperature and/or desication (Osa-
Afiana and Alexander, 1982) in acid soils to which rhizobial inocula are probably expo-
246
0 Onne
Ibadan
I
I#
.I11 IV
V
V I
V I I
T o l e r a n c e o f a c i d - A l s t r e s s
Figure 2 : Variation in the tolerance of acid-Al stress shown by Bradyrhizobium isolates from
Onne and Ibadan soils. Twenty isolates each were tested. I - VII : See Table 1.
P r o p o r t i o n i n p e r c e n t
A,
0
5 0
1 0 0
I
I
1
I
I
I
I
I
I
I
Onne : Promiscuous
I
V
V I I
c u l t i v a r s
N o n - romiscuous
I v
V I
c u l t r v a r
Ibadan:
p r o m i s c u o u s
c u l t i v a r s
I
I
I’ I
I
I
I
I
I
I
I
Figure 3 : Comparison of acid-Al
stress tolerance between Bradyrhizobium isolates from promis-
cuous and non-promiscuous cultivars of soybean grown in Onne and lbadan soils. Eleven and 9
isolates were tested for the promiscuous and non-promiscuous cultivars, respectively. in
Onne
soil and each 12 and 8 isolates, respectively. in
Ihadan soil.
241
sed, particularly during fallow periods under tropical African conditions. Therefore, tests
are required on the effects of those stresses on the saprophytic competence of rhizobia
(Thornton and Davey, 1984) and their symbiotic effectiveness under field conditions. At
the same time, screening or breeding of soybean cuhivars adapted to acid soil conditions
should not be ignored (Mmms et al., 1981).
Acknowledgments
This study was supported by the United Nations Environment Programme.
References
AHMAD, M.N., EAGLESHAM, AR.S. and HASSOUNA, S. 1981 Examining serolo-
gical diversity of cccowpea>, rhizobia by the ELISA technique. Arch. Microbial. 130,
281-287.
AYANABA, A., ASANUMA, S. and MUNNS, D.N. 1983 An agar plate method for
rapide screening of Rhizobium for tolerance to acid-aluminum stress. Soil Sci. Sot.
Am. J. 47, 256-258.
AYANABA, A. and WONG, W.L. 1982 Antibiotic -. resistauce mutants identified from
nodules of uninoculated soybeans grown in a strongly acidic soil. Soil Biol. Biochem.
14, 139-143.
BECK, D.P. and MUNNS, D.N. 1984 Phosphate nutrition of Rhizobium spp. Appl. Environ.
Microbial. 47, 218 282.
BROMFIELD, E.S.P. and AYANABA, A. 1980 The efficacy of soybean inoculation on
acid soil in tropical Afkica Plant and soil. 54, 95-106.
BROMFIELD, E.S.P. and ROUGHLEY, RJ. 1980 Characterization of rhizobia isolated
from nodules on locally-adaptet Glycine max grown in Nigeria Ann. AppI. Biol. 95,
185190.
CASSMAN, KG., MUNNS, D.N. and BECK, D.P. 198laPhosphorus nutrition of Rhizobium
japonicum : strains differences in phosphate storage and utilization. Soil Sci. Sot.
Am. J. 45, 517-520.
CASSMAN, KG., MUNNS, DN. and BECK, DP. 1981b Grown of Rhiwbium s@ains
at low concentrations of phosphate. Soil Sci. Sot. AmJ 45, 520-223.
COOPER, JE. 1982 Acid production, acid tolerance and growth rate of Lotus rhizobia
in laboratory media Soil Biof. Biochem. 14, 127-131.
CUNNINGHAM, SD. and MUNNS, D.N. 1984 The correlation between extraceIIular
polysaccharide production and acid tolerance in Rhizobium: Soil Sci. Sot. Am. J. 48.
1273-1276.
DATE, R.A. and HALLIDAY, J. 1979 Selecting Rhizobium for acid, infertile soils of
the tropics. Nature. 277, 62-64.
DE CARVALHO, M-M., BUSHBY, H.V.A. and EDWARDS, D.G. 1981 Survival of Rhi-
zobium in nutrient solutions containing aluminum. Soil Biol. Biochem. 13, 541-542.
GRAHAM, P-H., VITERI, SE., MACKIE, F., VARGAS, A.T. and PALACIOS, A. 1982
Variation in acid soil tolerance among strains of Rhizobium phaseoli. Field Crops
Rex 5, 121-128.
248
HERNANDEZ, B.S. and FOCHT, D.D. 1984 Invalidity of the concept of slow growth
and alkali production in cowpea rhizobia. Appl. Environ. Microbial. 48, 206-210.
HOWIESON, J.G. 1985 Use of an organic buffer for the selection of acid to1erantRhiz~bk.m
melibti strains. Plant and Soil. 88, 367-376.
HOWIESON, J.G.,EWING,M.A. andD’ANTUON0,M.F. 1988 Selection for acid tolerance
in Rhizobium meliloti. Plant and Soil. 105. 179-188.
IINTERNATIONAL IINSTITUTE of TROPICAL AGRICULTURE. 1979 The BTA Col-
lection of Cuhures of Rhiwbium. Ibadan, Nigeria.
IINTERNATIONAL INSTITUTE of TROPICAL AGRICULTURE. 1981 Annual Report
for 1980. Ibadan, Nigeria pp. 155160.
JONES, D.G. and BURROWS, A.C. 1969 Acid production and symbiotic effectiveness
in Rhizobium trifolii. Soil Biol. Biochem. 1, 5761.
KEYSER, H.H. and MUNNS, D.N. 1979a. Effects of calcium, manganese, and aluminum
on growth of rhizobia in acid media. Soil Sci. Sot. Am. J. 43, 500-503.
KEYSER, H.H. and MUNNS, D.N. 1979b. Tolerance of rhizobia to acidity, aluminum,
and phosphate. Soil Sci. Sot. Am. J. 43, 519-523.
KETSER, H-H., MUNNS, D.N. and HOHENBERG, J.S. 1979 Acid tolerance of rhizobia
in culture and in symbiosis with cowpea. Soil Sci. Sot. AM. J. 43, 719-722.
LOWENDORF, H.S. and ALEXANDER, M. 1983 Identification of Rhizobium phuseoli
strains that are tolerant or sensitive to soil acidity. Appl. Environ. Microbial. 45, 737-
742.
MUNNS, D.N. 1986 Acid soil tolerance in legumes and rhizobia Adv. Plant Nun. 2,
63-91.
MUNNS, DN., HOHENBERG, J.S.,RIGHETTI, TL. and LAUTER, DJ. 1981 Soil acidity
tolerance of symbiotic and nitrogen-fertiked soybean. Agr. J. 73, 407410.
MUNNS, D.N. and KEYSER, H.H., 1981 Response of Rhizubium strains to acid and
aluminum stress. Soil Biol. Biochem. 13, 115-118.
MUNNS, DN., KEYSER, H.H., FOGLE, V-W., HOHENBERG, J.S., RIGHETII, T.L.,
LAUTER, D.L.. ZAROUG, M.G., CLARKIN, K.L. and WHITACRE, K.W. 1979
Tolerance of soil acidity in symbioses of mung bean with rhizobia Agr. J. 71, 256-
260.
NANGJU, D. 1980 Soybean response to indigenous rhizobia as influenced by cultivar
origin. Agr. J. 72, 403-406.
NORRIS, D.O., 1965 Acid production by Rhizobium a unifying concept. Plant and Soil.
22, 143-166.
OSA-AFIANA, L.O. and ALEXANDER, M. 1982 Differences among cowpea rhizobia
in tolerance to high temperature and desiccation in soil. Appl. Environ. Microbial.
43, 435-439.
PARKER, C.A., TRINICK, M.J. and CHATEL, D.L. 1977 Rhizobia as soil and rhizosphere
inhabitants. In : A Treatise on Dinitrogen Fixation Section IV. Agronomy and Ecology.
Eds. Hardy, R.W.F. and GIBSON, A.H. : A willey-Interscience Publication, USA,
PP 311-352.
249
PULVER, EL., KUENEMAN, E.A. and RANGA RAO, V. 1985 Identification of pro-
miscuous nodulating efficient in N, fixation. Crop Sci. 25, 660-663.
THORNTON, F.C. and DAVEY, C.B. 1983 Acid tolerance of Rhizobium trzfolii in culture
media. Soil Sci. Sot. Am. J. 47, 496-501.
THORNTON, F.C. and DAVEY, C.B. 1984 Saprophytic competence of acid tolerant strains
of Rhizobium tnyolii iu acid soil. Plant and Soil. 80, 337-344
THURMAN, NP., LEWIS, D.M. and JONES, D.G. 1985. The relationship of plasmid
number to growth, acid tolerance and symbiotic efficiency in isolates of Rhkobiurh
trifolii. J. Appl. Bacterial. 58, l-6.
VINCENT, J-M. 1970 A Manual for the Practical Study of the Root nodule Bacteria
IBP Handbook Np 15. Blackwell scientific Publications. oxford and Edingburgh. p.3.
Maximiser la FBA pour la production agricole et fond&e en Afiique
The influence of some parameters
of soil fertility on early growth
of Leucaena ZeucocephaZa and Cassia siama
LIYA, S.M$ MULONGOY, K.c2),
Odu, C.T.P, and AGBOOLA, AA.(l).
(I) : Department qf Agronomy, University of Ibadan, Ibaah, Nigeria.
(2) : International Institute of Tropical Agriculture. PMB 5320. Ibaahn, Nigeria.
Abstract
L~ume~ leucocephala and Cassia siamea were grown in concrete cylinders
containing 0.35 t of a sandy, high base status Entisol. The nutrient content of the
soils varied from one cylinder to another. Fertilizer was added to the rate
11 kg N, 88 kg P and 5 kg Zn per hectare. At 8 months after planting, Lxucm
was well no-&late-d and produced on average seven times as much biomass as the
na-nodulated Casriu. There were significant (P = 0.05) correlations with topsoil
N. With Ca&a, positive correlations were observed for plant height (r = 0.73) and
total stem length (r = 0.75). With Lfxc- there were significant negative correla-
tions for plant height (r = - 0.85) and estimated total biomass (r = - 0.90). L..euasta
plant height was also positively correlated with topsoil K (r = 0.81). Ninety
eight per cent of the variation in height could be accounted for by the equation
Y = 4.88 + 117.37X - 14.94X2 where Y = plant height and X = K/N ratio.
251
Introduction
Trials of alley cropping in farmers’ fields have run into problems on account of
the very uneven establishment of the trees due to poorly understood factors. The work
reported here provides indications of how certain parameters of soil fertility may influ-
ence early growth of Cassiu siamea and Leucaena leucocephala.
Materials and methods
Two month-old plants of LRucae~ leucocephala (K8) and Cassia siamea (bcal
variety) were transplanted into concrete cylinders (60 cm i.d. and 90 cm deep) sunk in
the ground and containing a sandy, high base status Entisol which had been previously
used for various experiments for over a decade at Intemational Institute of Tropical Agri-
culture (BTA), Ibadan. Before planting, composite topsoil (15 cm) samples were taken
by bulking five cores per cylinder, and their nutrient content determined at the university
of Ibadan. At planting time, the cylinders were fertilized with 88 kg P ha-’ as single
superphosphate and 11 kg N ha-’ as (NH&SO,. Two months later 5 kg ha-r Zn was
applied as ZnSO,.Plant height, total stem lenght (plant height plus the sum of branch
lenghts) and leaf number were recorded 6 months after transplanting and a rough estimate
of total aboveground biomass was made, using the formula of Anderson and Ingram (1987)
to estimate stem and branch biomass. Leaf biomass was estimated by multiplying leaf
number by the average weight per leaf of a sample of leaves.
Results
Soil analysis (Table 1) indicated marked differences in nutrient content between
the cylinders with coefficients of variation ranging fi-om 11 to 64%. Soil physical pro-
perties (not reported here) varied little. The soils could be described as having medium
to high organic matter, low available P. and medium to high exchangeable K. The plant
growth data show that Lmcaena produced on average about seven times more biomass
than Cassia. There was considerable variation from one plant to another. For Leucaena,
coefficients of variation for leaf number, total stem length, plant height and estimated
total biomass were 51,58, 17 and 48% respectively. For Cursiu the variation was lower :
17, 40, 18 and 36% respectively. The form of different trees of the same species also
varied appreciably. Leucaena had 3 to 24 branches, Cassia 1 to 7 branches. Leucaena
was well nodulated in all the cylinders, but Cu.&u did not nodulate.
Regression analyses indicated that Cassia and Leucaena were differently affected
by topsoil properties.
Cassia stem and leaf production were positively correlated with soil organic C,
total N, available P, exchangeable Ca and Mg but negatively correlated with exchan-
geable K. Except for total N, these correlations were not significant. Total N was sign&
tautly correlated at the 5% level with total stem length (r = 0.74) and there was a high
though not significant positive correlation with leaf number (r = 0.66). The correlation
coefficient for leaf number and available P was not significant (r = 0.41). Stem and leaf
production were positively correlated with exchangeable Ca (r = 0.58 and 0.19 respectively),
and stem production was positively correlated with exchageable Mg (r = 0.34). On the
other hand, plant height was negatively correlated with all these soil properties except
Table 1 : The range and variation of soil properties in cylinders planted with Leucaena leucocephala and Cassia
siamea and the range and variation of plant growth parameters of the two species.
Characteristic
Leucaena leucocephala
Cassia siamea
Mean
Range
cv (%I
Mean
Range
cv (%I
Soil properties
Organic C (4;
1.8
1.8- 210
1.5-2.1
11
Total N 1%)
0.13
0.07-0.24
2
O.lO-0.14.
Extractable P (ug g-1)
1.8-8.4
ki
K (meq/lOO g)
4.5
1.3- 8.4
ii
0.2-0.4
55
Ca(meq/lOO g)
F:i
0.2-
5.5-13.0 4.6
35
4.3-12.0
44
Mg(meq/lOO g)
K/N(meq/lOO g)
;:i
2.6-
1.5- 4.9 5.8
El
2.4-5.3
1.3-3.5
ii
Plant growth parameters
Leaf No. plant-'
113
23-190
51
39-62
17
Height (cm)
174
123-201
17
43-69
r18
Total stem length (cm)
517
123-935
58
137
69-251
40
Estimated total biomass (g)
737
79-1056
48
96
74-157
36
Table 2 : Correlation coefficients between soil properties and plant growth parameters
Tree
Plant
Soil property
parameters
C
N
P
K
Ca
MCI
K/N
~
-
-
-
-
-
Leucaena
Leaf No.
0.05
-0.76*
-0.22
0.05 -0.04
-0.53 0.30
leucocephala
Total stem
length
0.41
-0.69
-0.02
0.19
0.22
-0.41
0.60
Height
-0.58
-0.85"
-0.31
0.81*
0.01
-0.55
0.70
Estimated biomass
-0.13
-0.90*
-0.34
0.36
0.12
-0.49
0.53
Cassia siamea
Leaf No
0.02
0.66
0.41
-0.41
0.19
0.001
-0.57
Total stem
length
0.30
0.73*
0.32
-0.2
0.58
0.34
-0.18
Height
-0.45
0.75*
-0.56
-0.01
0.25
-0.43
-0.03
Cstimated biomass
0.09
0.47
0.03
-0.40
0.20
-0.09 -0.25
* Significant at P = 0.05
254
total N. There was no correlation between plant height and total biomass production
( r = 0.03).
With Lf~~aena, the results were in almost total contrast. There were negative cor-
relations between plant height, total stem length or leaf number and total N with the
following coeficients of correlation : -0.85 (significant at P = O-05), -0.69 and -0.76 respec-
tively. Also, correlations between exchangeable Mg and plant height, total stem length
or leaf number (r = -0.55, -0.41 and -0.53 respectively) were negative but not significantly.
There was however a significant correlation between plant height and exchangeable K
(r = 0.81). Plant height was also significantly correlated with total biomass production
(r = 0.87) and leaf number (r = 0.63). Plant height seemed to be governed by the soil
K/N ratio. The relationship was best described by a quadratic equation, and 98.9% of
the variation in Leucaena plant height could be dkcribed by the equation
Y = 4.88 + 117.37X - 14.94p, where Y = plant height in cm and X = K/N ratio in
the soil (Fig.l), with K expressed in meq/lOO g soil and N in percent.
Figure 1: Relationship betweenkucasla height and K/N ratio of the soil
Discussion
In spite of the deep rooting habit of Cassiu and Leucaena, it was interesting to
note some significant correlations with topsoil properties. The results highlight the very
different response to N of the two species. Growth of the non-nodulated Cassia was signifi-
cantly correlated with topsoil N and clearly limited by this element, so that plant biomass
production was very low compared with that of Leucaena. The results further suggest
that Cussia has a high requirement for P and exchangeable bases. This would explain
the negative correlation with K because of the antagonism between K and Mg. Our results
contour with observations by foresters that Cassiu siumea grows well only in fertile soils,
a fact which limits its use in re-aforestation programs (Anonymous, 1981).
254
total N. There was no correlation between plant height and total biomass production
( r = 0.03).
With L..eucuenu, the results were in almost total contrast. There were negative cor-
relations between plant height, total stem length or leaf number and total N with the
following coeficients of correlation : -0.85 (significant at P = O-OS>, -0.69 and -0.76 respec-
tively. Also, correlations between exchangeable Mg and plant height, total stem length
or leaf number (r = -0.55, -0.41 and -0.53 respectively) were negative but not signifkantly.
There was however a sign&ant correlation between plant height and exchangeable K
(r = 0.81). Plant height was also signikantly correlated with total biomass production
(r = 0.87) and leaf number (r = 0.63). Plant height seemed to be governed by the soil
K/N ratio. The relationship was best described by a quadratic equation, and 98.9% of
the variation in Leltcaena plant height could be decribed by the equation
Y = 4.88 + 117.37X - 14.94X, where Y = plant height in cm and X = K/N ratio in
the soil (Fig.l), with K expressed in meq/lOO g soil and N in percent.
Leucaenahdght (cm)
2 0 0
1 5 0
100
Y - 4.80+117.37X-l 4.942
0
1.0
2 . 0
3 . 0
4 . 0
5 . 0
6 . 0
W N
Discussion
In spite of the deep rooting habit of Cussiu and Leucaena, it was interesting to
note some signifkant correlations with topsoil properties. The results highlight the very
different response to N of the two species. Growth of the non-nodulated Cussiu was signifi-
cantly correlated with topsoil N and clearly limited by this elemenf so that plant biomass
production was very low compared with that of Leucue~~~. The results further suggest
that Cussia has a high requirement for P and exchangeable bases. This would explain
the negative correlation with K because of the antagonism between K and Mg. Our results
contour with observations by foresters that Cussia siamea grows well only in fertile soils,
a fact which limits its use in re-aforestation programs (Anonymous, 1981).
2 5 5
In contrast, all parameters ‘of Leucuenu growth showed negative correlations with
soil N, an indication that this woody species relied for its N supply on biological N,
fixation, a process inhibited by increasing levels of soil N. Apparently the N supplied
in the experimental soils enabled L.etia+nu the plant to make rapid and prolific growth
compared with cussia. ::.3:
L.eucaenu growth J&S also* shown to be influenced by available K in the soil
(r = 0.81 and 0.53 for plant height and total biomass respectively). These relatively high
correlations were likely due to the rather high soil Mg levels. There was a negative corm-.
lation though non-significant between plant height and exchangeable Mg (r = -0.55). As
K controls meristem growth, fast growing trees may be expected to have high K requi-
rcments. This has been demontrated for Leucuena which can respond to up to 800 ppm
K and also appears to be rather inefficient at extracting the element from the soil (Schmidt
and Saefuddin 1987). The K requirement of a plant is increased by an abundant N supply
(Mengel and Kirby 1987). Thus the K requirements of N,-fixing trees are likely more
pronounced than for non-N,-fixing trees. Indeed numerous studies have demonstrated that
K is required in greater quantities than any other soil mineral for total N accumulation
in legumes. It has also been establish that K-stimulated increases in legume dry matter
production are accompanied by increases in ,N, fiition when no fertilizer N is used.
This is believed to be caused by K-induced increases in photosynthesis, the transport
of photosynthates to the root nodules and increased root growth (Duke and Collins, 1985).
Fifty six percent of the variation in height of a related N,-fixing woody species AZbizia
fdcaturia over 28 different sites has been reported to be determined by soil K levels
(Delmacio 1987). Soil K levels of unfertilized soils were positively correlated with nodule
productivity of the N,-fixing Ingu jinicuil in Mexico (Van Kessel and Roskoski, 1981).
Relatively low levels of K fertilization stimulated N, fixation by this tree (Van Kessel
and Roskoski, 1983). The critical K level for maximum growth and N2 fixation of legu-
minous trees appears to be much higher than that required by cereals and needs to be
further investigated.
In this experiment, low soil N levels promote N, fixation and thus plant growth.
However, this condition increases plant K demand, and plant growth in turn becomes
limited by the soil supplied K. There appears to be an optimum K/N ratio (Fig-l).
It can be concluded that while Cussia is dependent on general soil fertility, particu-
larly N availability for adequate growth, nodulated Leucuena can grow well on poor soils,
if supplied with enough K and possibly P and Zn, which were adequately supplied in
this experiment. The relative growth of the two species at any one place will depend
on site factors. Cussie is likely to compete with interplanted food crops for N, but Leucuena
may not compete for K with cereals because legumes are usually less efficient than cereals
at extracting K from soils.
Aknowledgements
The authors thank the Federal Ministry of Education of Nigeria for the award of
a Commonwealth Scholoarship to S. LIYA.
256
References
ANDERSON, J.M. and INGRAM, J.S.I. 1987 Tropical soil biology and fertility methods
handbook. IUBS/UNESCO-MAB, Paris.
ANONYMOUS, 1981 Memento du Forestier. 2nd edition. Paris. Republique Fran&se,
Ministere de la Cooperation et du Developpement.
DELMACIO, M.I. 1987 Relationship between some site factors and growth of Albizia
falcataria. NFT Res. Rep. 5, 26-28.
DUKE, S.H. and COLLINS, M. 1985 Role of potassium in legume nitrogen fixation.
In : Munson R.D. (ed.) Potassium in Agriculture. pp 443446. American Society of
Agronomy, Madison, USA.
MENGEL, K. and KIRBY, E.A. 1987 Principles of plant nutrition. 4th edition. International
Potash Institue, Bern, Switzerland.
SCHMIDT, DR. and SAEFUDDIN, A. 1987 Response of Gliricidia and Leucaena to
potassium and lime. NFT Res. Rep. 5, 46-48.
VAN KESSEL, C. and ROSKOSKI, JP. 198 1 Nodulation and N,-fixation by Inga junicuil,
a woody legume in coffee plantation II. Effect of soil nutrients on nodulation and
N, fixation. Plant and soi1.59, 207-215.
VAN KESSEL, C. and ROSKOSKI, J.P. 1983 Nodulation and N, fixation by Inga junicuil,
a woody legume in coffee plantations III. Effect of fertilizers and soil shading on
nodulation and nitrogen fixation (acetylene reduction) of I. junicuil seedlings. Plant
and soi1.72, 95105.
MESURE DE LA FIXATION D’AZOTE
Maximisez la FBA pour la Production Agricole et Forest&e en Afrique
Evaluation of biological
nitrogen fixation in plants
DANSO, S&A.
’
Joint FAOJIAEA Division, P.O. Box 100, A-1400 Vienna, Austria
Abstract
Measuring the contriiution of atmospheric nitmgen (N2) to the growth of plants
is an kportmt step in understanding factors that influence the process, and for in-
creasing N2 fixation iu various systems. Although the uN techniques generally appear
to be superior to the other methods of measuring N, fixed, the facilities are not
available in many laboratories, and like all the others, it has its weaknesses as well
; undez some situations, or depending on the level of accuracy required. some of
the non-isotopic methods, many of which are less expensive, may be equally or
more suitable. The situations under which methods often regarded as inferior give
near accurate results have seldom been adequately highlighted. In this paper
thexfa the most fkequently used methods, total nitrogen difference ml9 acety-
lene reduction, xylem exudate analysis and the *W techniques have been examined
withthisinmind.Itisconcludedthatassimpleandinexpensiveasitis,theTND
approachcanundexmany cinxmsm give satisfactory estimates of N, fixed. The
acetylene reduction assay, on the other hand, has several drawbacks and cannot be
used to obtain quantitative measures of N2 fured. However. it should at least be
very useful in the initial scmening for wide diffexexes in N2 fuation among geno-
types and for distinguishing between absolute non-furers and potential N, furers.
259
Introduction
Increases in the contribution of nitrogen (N2) fixation to crop and tree growth,
and to high productivity of many agricultmaJ soils are attainable. It is however neces-
sary to lmow the current levels of N, fixed, so that any improvements and the conditions
necessary to achieve them can be ascertained. Several methods already exist, and have
been used to measure N, fixed in many crops and ecosystems. It is certain from all the
reviews on these methods that each has some disadvantages (Knowles, 1981 ; Chalk,
1985 ; Danso, 1985 ; Vase and Victoria, 1986). Judging from the diversity of N; fixing
crops and cropping systems, the prospects of an ideal method for ail situations are slim.
We are therefore compelled to make the best use of existing methodologies, insofar as
we are aware of their limitations, and to use the most appropriate for each particular
situation. This requires a thorough understanding of the strengths and weaknesses of each
methodology. Methods have unfortunately been on many occasions adjudged as the most
or least appropriate, without due regard for the situation. In this paper therefore, the
greatest strengths and weaknesses of the most commonly used methods for measuring
N, fixation will be examined, and as much as possible, the situation(s) under which each
is best suited will be assessed.
Total Nitrogen Difference (TND) Method
This is one of the oldest and simplest methods, and has provided us with many
estimates of N, fixed, upon which several useful management practices have been
devised. In its simplest form, the TND method is based on the difference between the
total N contents of plants or ecosystem in which N, is fmed and those that do not derive
N from N, fixation. The method in essence is based on an assumption that both the
N2 fixing and non-fming control plants absorb equal amounts of soil N for growth (Ren-
nie and Rennie, 1983). This assumption may not however hold under most situations,
as this would require that the different plants, in addition to being similar in root morpho-
logy and in several physiological attributes, should absorb their N from similar depths
and horizon. This assumption is therefore often the greatest limitation of the TND method
(Danso, 1985). In addition, where N, fixed in the whole ecosystem rather than only in
plants is being measured, the inability to recover completely all the different forms of
N using the usual Kjeldahl digestion procedures may result in erroneous estimates. For
measuring N, fnation in an ecosystem using the TND approach, it is also difficult to
quantify accu- rately all the many inputs and outputs of N into the ecosystem ; the errors
in these TND estimates therefore tend to be high, and can be very serious when N, fixa-
tion is low (because the analytical errors in the determination of total soil N alone may
sometimes be greater or close to the measured inputs from N, fixation). To overcome
this, especially in N-rich soils, N2 fixed is usually allowed to accumulate for a few to
several years before measurement. In this way, the TND method tends not to be only
tedious, but also requires long durations.
Despite the stated limitations, the TND method has often provided estimates that
are not significantly different from those obtained using more sophisticated and expen-
sive techniques (e.g., Ham, 1978 ; Phillips et al., 1983). These similarities therefore jus-
tify a close examination of the strengths of the TND method, and when it is most suitable.
As suggested by Danso (1988), and shown in the study of Paterson and LaRue (1983)
260
and Retmie (1984). the TND method reliably estimates of N, fixed in plants grown in
soils or systems in which the initial N content is low. Under such circumstances, almost
all the N would naturally be expected to be derived from N, fixation, with soil N contri-
buting only negligible amounts. Several such situations exist, e.g., in areas consisting
of sand dunes, systems undergoing primary colonization, and in many semi-arid areas.
It is therefore not surprising that the values of N, fixed in Casumina stands on the sand
dunes in Cape Verde (Dommergues, 1963) are still quoted universally, and Gauthier et
al. (1985) in their N, fixation studies on a sandy soil in Senegal frequently found no
sign&ant differences between TND estimates of N, fixed and those obtained using the
more expensive and more sophisticated (both mathematically and analytically) 15N me-
thodologies. The situation could, however, be different even where the soil N level is
low, depending on how much N, is fmed ; when this is low, the estimates made could
like when soil N is high, be highly erroneous. In this czule, the otherwise small errors
in the estimates of soil N uptake could be high, relative to N, fixed, and this would
account for many of the negative estimates of N, fixed that have been reported using
TND method (e.g. Patterson and LaRue, 1983 ; Rennie, 1984). All the same, it may
not be worth bothering much about the magnitude of such inaccurate estimates since in
my opinion, the real fact is, that the negative data only indicate that even if any N,
was fixed, it was so low that it could be regarded as negligible and not worth taking
note of.
The chances of very serious errors in TND estimates of N, fmed are certainly
high when inorganic N is high in the soil, unless the control non-N, fixing crop and
the N,-ftig crop absorb closely similar amounts of soil N. As reported by Rennie (MM),
this ideal situation is attained when the fertilizer utilization efficiencies (FUR) of the fixing
and control plants are equal, and this would have to be ascertained using ‘5N fertihzers.
However when the FUEs are unequal, fairly accurate estimates of N, fixed may
sometimes be obtained. As shown by Hatdamon et al. (1988), errors in N, fixation es-
timates become smaller and less important at higher N2 fmation rates, and although this
principle was demonstrated using only the 15N isotope dilution technique, the rationale
should hold for all the other methods. Thus, for some forage legumes, many of which
have been reported to derive around 80 to 90% of their N from Nz fixation (e.g., Herd]
et al.. 1968 ; Hardarson et al, 1988), it is likely that any erroneous estimate of N, fmed
asaresultoftheinaccurate
assessment of soil N uptake could be minimal, in contrast
to several grain legumes, such as many commonbean (Phusseotus vzdguris) varieties that
are known to be poor N, fixers.
Acetylene Reduction Technique Assay (ARA)
This methodology, used originally for N, fixation studies in pure microbial cul-
tures (Dilworth, 1966), and which was later developed for measuring N2 fixation in green-
house and field-grown crops (Hardy et al., 1968), is a paradoxical case of a method
which has, at one stage or another, been used in almost every laboratory for measuring
N, fixation, and which a little over a decade ago, stimulated an unprecedented interest
in N2 fixation studies, but which today, as shown in many reviews, is almost at the point
of becoming “the” unacceptable or obsolete way of measuring N, fixation (e.g. Witty
and h4inchin, 1988).
261
The early, fast and wide acceptance of the ARA technique for assessing N, fixa-
tion was due to its being a simple, rapid and highly sensitive method (Hardy et al., 1968),
and a reasonably lower instrumentational cost. Also, since ethylene is stable, the final
product could thus be stored (allowing large samples to be collected at a time) and analysed
when most convenient.
Although the early and popular acceptance of the ARA technique, for measuring
N, fixation was almost universal, many of its drawbacks soon became apparent, and various
attempts were made to overcome or minimize their effects. For example, several reports
called attention to the instantaneous nature’of ARA assays, and the doubtful validity in
extrapolating these short-term assays to cover N, fmed over whole growing seasons (in-
volving large diurnal, daily and sometimes seasonal variations in rates of N2 fixation)
(Rennie et al., 1978). Attempts to reduce this problem have- included making several se-
quential assays rather than a single or few measurements (e.g. Hardy et ol., 1968). Because
nitrogenase activity declines significantly in detached nodules, roots with attached nodules
more frequently replaced the use of detached nodules. Also, studies showed that the digging
of roots (especially in field-grown plants) resulted in unavoidable nodule losses, and the
subsequent preparation of root samples for ARA assays, such as washing of roots (e.g.
Carta and Sheaffer, 1983) or the drying of nodules in the field or after sampliig, adver-
sely affected the diffusion of gases, such as oxygen and acetylene into the nodules (Minchin
et al., 1983 ; Weisz et al., 1985 ; Durand et al., 1987). Also, it was noticed that the
levels of acetylene used sometimes inhibited the process being measured (Witty and
Minchin, 1988), while nodulated root segments exhibited a lower ARA than intact plants
(Wych and Rains, 1978). Efforts made to solve or minimize these problems included
conducting nondestructive in situ assays on intact plants in the field or greenhouse rather
than on decapitated roots (Wych and Rains, 1978 ; Pfeifer et al., 1983), and the use
of a flow-through gaz system (Minchin et al., 1983 ; Fellows et al., 1987) which allowed
a low and non-toxic concentration of acetylene to be continuously passed over the N,-
fixing system. None of these modifications was still able to overcome the need for artifi-
cial enclosures for the plants, and it was also often necessary to correct for the endo-
genous production of ethylene from soils.
Furthermore, it became known early that the validity of the factor of 3 originally
proposed for converting C&Id produced into values of N, fared (Hardy et al., 1968) was
not valid for all crops and systems (Rennie et al., 1978 ; Hudd et al., 1980). This discre-
pancy was largely attributed to all the electrons becoming available to nitrogenase in
the ARA assay, whereas Hudd et al. (1980), for example, reported that only about half
the electron supply was used to fix N, in field bean (Viciu f&u). resulting in a true
conversion ratio of 5.75 rather than 3. This value was also not universally acceptable,
since experimentally determined ratios have been reported to differ widely among diffe-
rent N, fixing plants and systems, with some of the reported ratios ranging from 1.5
to 8.4 (Bergersen, 1970 ; Hardy et al., 1973 ; Rermie et al., 1978), and perhaps greater.
Reunie et al. (1978) therefore recommended the experimental determination of the ratios
to use under each situation using lsN assays for calibration of the ARA technique.
However, lsN, gas is expensive, and th$ approach would require determining the correct
ratio to use for each study, as for the same crop, the electron flux (which the ARA tech-
nique essentially measures, rather than direct N, fmation) may differ under different experi-
mental conditions or assay periods (Witty and Minchin, 1988). It was therefore not
surprising that conversion of ARA assays into kg N/ha of N, fixed became less popular,
262
and was more often expressed as ethylene produced (either in nM or pM) per plant or
nodule, and per unit time. The shift allowed seemingly valid relative comparisons of treat-
ment effects to be made, by avoiding the need to use controversial conversion ratios.
This approach held much promise for comparing N, fixation abilities (where the actual
amounts of N, fmed were not essential or critical) until later evidence suggested that
insofar as electron fly, could be influenced differently by different treatments, such rela-
tive comparisons were of questionable validity (Witty and Minchin, 1988).
For those still dedicated to using the ARA technique, it seems that it may soon
become difficult for their results to be accepted, unless the method can somehow or the
other be imporved beyond its present capabilities. These adherents may however argue
that, despite all these stated problems, numerous AFCA estimates of N, fixation, statis-
tically similar or highly correlated to values obtained by other “acceptable” methods
abound in the literature (e-g. Hardy, et al., 1968 ; Paterson and LaRue, 1983). These
results lead us to question whether all the methods used in these cases gave wrong results
or that the limitation of ARA have in reality been exaggerated. These are valid ques-
tions for all scientists interested in N, fEation studies. It would be necessary to closely
examine how often ARA estimates have deviated significantly from those obtained by
other methods ; this would be particularly importaut in studies to rank N, fixation capa-
bilks. for which the ARA technique at one stage held much promise. If these deviations
are found to be only occasional, a case may have been established for Examining when
the ARA technique could be used to obtain valid results, and what precautions are neces-
sary. The method at least holds much promise for identifying potential N, fixing crops
in novel communities, and for grouping plants into broad categories in terms of N2 fixation
ability. It is also currently the only method that has consistently been used to establish
that a plant intended to be used as a reference crop (m the 15N methodology, and inci-
dentally often the method of choice now) is truly a non-N, fixer (e.g. Fried er al., 1983).
Relative Urelde Production
The xylem sap of N, fixing plants contains N compounds originating not only
from soil N uptake by roots, but in addition, contains N of nodule origin, through the
process of atmospheric N, fixation. In some N,-fming tropical legumes, differences in
the N compounds carried by the xylem sap have been found. This is influenced largely
by whether the N is of soil or nodule-origin (Pate et al., 1980 ; Herridge. 1982s) ; the
soil derived N is exported largely as free NO,- and amide amino acids (e.g. glutamine
and aspam&@, while N, fixation results largely in the formation and export of ureides
(e.g. allantoin and allautoic acid). The relative composition of ureides versus NO,- + amides
in these cases therefore serves as a useful index of the sources of the plant’s N. ‘Ibe
composition of the xyIem sap changes progressively from one dominated by ureides to
one composed largely of NO,- and amino compounds as N, fixation declines (Herridge,
1982a).The assay of ureides, contrary to what many expect, is not that sophisticated, and
involves only a simple calorimetric assay. However, like most calorimetric assays, it needs
to be calibrated, and both the ARA and ‘%I techniques have been used for this purpose
(Pate et al., 1980 ; Heniedge, 1982a).
The initial handicaps of the methodology included (i) obtaining a representative
exudate, since the ureide content of the plant was not. uniform in all organs (Herridge,
Maximiser la FBA pour la Production Agricole et Forestière en Afrique
Influence du déficit hydrique
sur la fixation symbiotique
de l’azote atmosphérique chez le soja
SALL, K., DREXON, J.J. et OBATON M.
Laboratoire de Recherches sur les Symbiotes des Racines,
lNRA. 1, place viala, 34060 Montpellier cedex 1, France
RQsumé
L’effet du déficit hydrique sur l’activité nitrogt?nase est étudié chez le soja par
application de mannitol. Ce travail a Bté Alis& en serre en culture hydroponique.
sur deux variktés de soja pnkentan~ dans les essais agronomiques. des sfmsibilit.&
diffkrentes à la sécheresse : Kingsoy n5sistante et Hodgson sensible. Sur ces plantes
de soja en symbiose avec une souche de Bradyrhizobium jqponicum, sont mesures
I’activitr? fïxahice d’azote et I’effet de pressions partielles d’oxygène croissantes,
sous diffhntes intensitks de contrainte osmotique : - 0.2, -0.4. -0.6 et -0.8 MPa
De fortes pressions partielles d’oxygene permettent de lever sen.sibkxnent,
mais pas
totalement, l’inhibition de la fixation due B la contrainte hydrique. L’infbrence de
la contrainte osmotique sur la nitrog&urse varie suivant la teneur en nitrate du milieu
de culture. Ces nkultats suggèrent que le déficit hydrique limite l’alimentation des
bactkroïdes en oxygi?ne mais egalement affecte la quantid de pouvoir r&hrcteur et
la quant% d’enzyme active @sente dans les nodositks pour la fixation de l’azote.
La fixation de l’azote est affect& diff&ennnent chez les deux cultîvars : elle est
moins dkpritnee chez Hodgson que chez Kinsoy. apparenunent parce que Hodgson
posskle une réserve énergétique plus importante dans ses nodosit&.
273
Introduction
Dans les conditions de la culture du soja en plein champ chez les plantes inocult5es
avec Brudyrhizobium juponicum, la fixation symbiotique de l’azote atmosphkrique fléchit
rapidement, pouvant même s’annuler lorsque le déficit hydrique est s&&e. (Obaton et
al., 1982 ; Bouniols et al., 1982 ; Wt5ry et al., 1986).
La baisse de l’activité nitrogénase par le déficit hydrique est observée par la méthode
de l’activité réductrice de l’ac&yEne (ARA) (Pankhurst et Sprent, 1975 ; Sheehy et ul.,
1983 ; Benett et Albrecht, 1984) et par la mCthode 15N (Sinclair et Goudriaan, 1981).
Alors que certains auteurs considknt que la diminution de l’activité nitrog&we est due
A l’inhibition de la photosynthbe (Huang er al., 1973 ; Sprent, 1976), Finn et Brun (1982)
concluent que la nitrogenase est plus sensible au &fkit hychique qu’à la photosynth&se.
Cette interprétation est soutenue par Bennet et Albrecht (1984) qui ont montré que la
fucation de l’azote atmosphkrique est étroitement corrt5Ee au-potentiel hydrique des nodules,
ce dernier &ant plus sensible au déficit hydrique que le potentiel hydrique foliaire et
la résistance stomatique.
Selon Weisz er al. (1983, l’inhibition de l’activid nitrogknase par le déficit hydrique
est détectable au champ trois jours seulement aprRs l’installation de la contrainte hydrique
et avant qu’il y ait une différence pour la surface foliaire, Ie poids des plantes ou des
nodules et le nombre de nodosités entre plantes stressks et plantes témoins.
Pankhurst et Sprent (1975) notent qu’un accroissement de la pression partielle d’oxy-
gène (pOJ lors d’un déficit hydrique, augmente la fixation de l’azote mesur& par l’activité
r&luctrice d’acétylène (ARA) et la respiration des nodules. Ce r&ultat obtenu sur nodules
excisks nous a conduit à réaliser des exp&iences similaires sur plantes entières pour essayer
de mieux comprendre ce phénomène, tout en perturbant au minimum la physiologie des
plantes.
Matbriels et méthodes
Matbrie biologique et culture des plantes
Deux vari&% de soja (GZycti max L. Merrill) sont cultiv&s : Hodgson du groupe
1 demi-prkoce, sensible à la skheresse, et Kingsoy du groupe II demi-tardif, retenue
pour sa bonne tol&ance à la skheresse (Bouniols, con-un. pers.).
L-es graines de soja, après dtifection A lhypochlorite de calcium (30 g 1-l pendant
30 mn), sont mises à germer a I’obscuriti durant 3 jours A 28°C sur vermiculite sttkile.
Les plantules de soja sont placées sur des plaques de polystyr?ne prkilablement pwc&s
pour laisser croître la partie racinaire. L’ensemble est ensuite placé sur des bacs de 40
litres (60 cm x 40 cm) contenant la solution nutritive a&& dont le pH est maintenu
à6J.Lesplantulesdesojasontinoc~~d~lamiseenplaceaveclasouchedeBradyrhizobium
japonicum G 49 (SB 16, IARI, Inde). Les cultures sont conduites en serre &Clair& pendant
16 h (klairement additionnel pendant l’hiver, 100 W mw2 par des lampes Philipps), à
une tempkrature variant entre 15 et 35°C. Pendant la phase obscure de 8 heures. la température
varie entre 15 et 20°C. L’humiditk relative, augmentée grâce A l’action d’un brumisateur,
est supérieure 3 70% et varie suivant les saisons.
274
La solution nutritive contient : qP0, : 025 p.M ; ,CaClr : 33 p.M ; K$O, :
125 plvl ; MgSO,. 7qO : 2.05 @I ; %BO, : 4pM ; MnSO, : 6,6 pM ; ZnSO,,7%0:
1,55 p.M ; CuSG4,5$0 : 1,56 p.M ; Na$foG,, 2qO : 0,12 p.M et du fer sous forme
d’EDTA (0,015 g 1-i de produit). Pendant les 2 premikes semaines, l’azote de complement
est apporte sous forme duree (1 mM). Lorsque la nodnlation est install& les plantes
ne reçoivent aucune autre source d’azote que la fixation sauf dans les exp&iences avec
nitrate où l’apport d’ut% est remplace par des doses variables de NO,-. Le milieu de culture
est retrouvele toutes les semaines, et r@likrement rkajusté avec de l’eau distillee.
Intensité! et mode d’application du dbficit hydrique
La contrainte hydrique est obtenue par abaissement du potentiel osmotique de la
solution par apport de mannitol pendant 24 h. Le mannitol est peu ou pas absorbe par
les cellules (Heath, 1977) et ne prksente pas de phytotoxicité. Pour eviter la multiplication
des microorganismes dans la solution nutritive, un antibiotique, la t&racycline, est apporte
(10 mg 1-l) avec le mannitol. Quatre traitements sont rt!.ali&-s : -0.2 MPa (0,074 M) :
deficit faible ; -0,4 MPa (0,153 M) : déficit moyen ; -0,6 MPa (0,233 M) : deficit fort
mais non destructif ; -08 MPa (0,312 4 M) : déficit tres fort et plus ou moins destructif
suivant les conditions climatiques.
Mesures des échanges gazeux
Les diffkentes mesures effectukes dans ce travail, sont faites à partir du qnarantieme
jour de culture ; les plantes sont alors en pleine floraison, stade le plus sensible a la
sécheresse (Bouniols ef ul. 1982). Les mesures de l’activité nitrogenase sont rt%is&s par
la technique de rkluction de l’ac&ylkne en dthylene (Hardy et cd., 1968) appliquée a des
racines nodulees en place SUT la plante. L’incubation est faite en flacon s&um de 500
ou 1000 ml, ce qui permet de r&liser des mesures non destructives de l’activité nitrog&ase.
LWanchei~ au niveau de la tige est assur& par un joint de coton imbibe de Rhodorsil
RTV 1502 (Rhône-Poulenc). Les incubations en presence de C,H, sont réalisees soit en
systkne ferme, soit en continu avec balayage du milieu racinaire par un mélange gazeux
contenant 10% CA.
Mesures en systbme fermé - Le protocole appliqué est celui dkrit par
Kalia et Drevon (1985). La partie racinak de la plante est herm&iquement enfermé
dans le flacon s&um, avec 10% (v/v) de solution nutrive pour maintenir la turgescence
et l’activid des organes. La nitrogtkw est saturee par une concentration en ac&yl$ne
de 10%. La quantité d’&hyEne accumulée dans ce compartiment mcinaire est &ahK?e
par des pr&vements d’&antillons de l’atrnosphere interne qui sont ensuite analy& B
l’aide d’un chromatographe Girdel30. muni d’un détecteur à ionisation de flamme (colonne
de Porapak T: 1,5 m de long et 3.2 mm de diam&re, tempkature du four : 80°C. gaz
vecteur azote à 30 ml mn-‘). Les mesures d’activité nitrogenase (ARA) obtenues dans
ces conditions sont plus élev&s que celles obtenues sur les racines dparees car l’excision
entraîne une baisse de 50 3 80% de l’activitt enzymatique @ilston et Imsande, 1982).
Toutefois, en systkme fermé, la rkluction d’ackylène ne reste constante que pendant environ
60 mn (Kalia et Drevon, 1985). Les mesures doivent donc être effectuées rapidement
275
Mesures en systtkne continu - Le principe de mesure de l’activité de
la nitrogénase est le même qu’en systeme ferme, mais cette technique a l’avantage de
permettre le contrôle du mélange gazeux qui balaie en permanence l’atmosphère du
système racinaire. Ainsi, grâce à des débitmètres Tylan, le flacon est balayé par un me-
lange gazeux a debit constant et contenant de l’adtylène (NS), de l’oxygène et de l’azote
avec la possibilité de moduler les concentrations respectives de ces deux gaz. L.e systkme
comprend le flacon s&um qui porte la plante, les débitmètres connectés aux différentes
bouteilles de gaz et un tuyau de sortie du milieu d’incubation qui est connecté au chroma-
tographe et au catharomètte à conductivité thermique. Le catharomètre permet de doser
les quantites d’oxygene et d’azote contenues dans les 6chantillons. Les deux appareils sont
munis chacun d’une vanne d’injection automatique qui fonctionne alternativement à la
fréquence programmk. Un autre débitmètre à bulles de savon, relié à la sortie du catha-
romètre, permet de vérifier le débit du mélange gazeux. Il est ainsi possible & suivre
l’evolution de l’activité nitrogénase en fonction de la pression d’oxygène (Drevon et al.,
1988). Le dispositif permet de réaliser quatre incubations simultanément Des prélève-
ments à la seringue d’échantiIlons de l’atmosphère racinaire permettent de mesurer le déga-
gement de CO, par des dosages avec un IRGA ADC. 225 MK3.
R&ultats
Influence de la contrainte osmotique sur I’ARA à 21% 0,
Les donnees présentées à la Fig.lA montrent que chez Kingsoy, l’ARA diminue
des les premieres heures qui suivent l’application du manuitol dans le compartiment ra-
cinaim. Cette inhibition de l’activite nitrogénase est d’autant plus forte que l’intensiti du
déficit osmotique est e1ewS.e. Apres 24 h en prksence de mannitol, l’ARA exprim& en
pourcent de celle du tkmoin n’ayant pas rqu de mannitol, n’est plus que de 72, 63, 52
et 29% respectivement pour une pression osmotique de -0,2, -0,4, -0,6 et -0.8 MPa. Après
avoir 1evC la contrainte osmotique, I’ARA continue a dkcliner pendant au moins 4 jours
chez les plantes ayant Cté soumises a la contrainte osmotique de -0,8 h4Pa. Bar contre,
chez les plantes soumises à -0,4 et -0,6 MPa, ce déclin d’activité cesse apparemment
après 2 jours, et chez celles soumises à une contrainte de -0,2 MPa la stabilisation de
SARA s’effectue aprks environ 24 h. Chez ces demieres, l’ARA augmente même signi-
ficativement durant le troisieme jour.
Chez la variété Hodgson (Fig.lB), l’ARA est apparemment moins affect& que chez
la varieté Kingsoy. En effet, apr& 24 h de contrainte osmotique, I’ARA exprimk en
pourcent du témoin n’est que de 82 a 62% pour les trois traitements. Il n’y aurait pas
de différence significative de l’inhibition de l’activité nitrogénase chez cette vari&& pour
des concentrations de mannitol correspondant a -0,4 et -0,6 M?a. Apres retrait du man-
nitol, la baisse de la fixation se prolonge encore pendant environ 3 jours pour les plantes
ayant subi les traitements Ct -0.4 et -0,6 MF’a alors que la stabilisation de I’ARA est plus
rapide a -0,2 MPa comme pour la variété Kingsoy. Ainsi, 4 jours après la fii de la con-
trainte osmotique, les plantes précédemment soumises a -0,2 MPa retrouvent 84 % de
leur activité fixatrice d’azote, contre 52 et 40 pour -0,4 et -0,6 MPa respectivement.
276
80
60
4 0
6 0
80
100
120
140
- 2 0
p
20)
4 0
6 0
8 0
1 0 0
1 2 0
140
mn!ralnte hydripue
TEMPS en HEURES
Fiiure 1 : Jnfluence du déficit hydrique sur la fixation symbiotique de l’azote mesur& par la méthode
ARA, chez les vari&tés de Soja KINGSOY (A) et HODGSON (B). Les résultats sont exprim& en
pourcentage du témoin. Chaque point est la moyenne de 5 rép&itions. Le traitement au mannitol dure
24 heures. Les mesures sont faites 1 heure avant le traitement, 5 h et 24 h après. Les mesures sui-
vantes sont effectuées au retour à un régime hydrique non limitant. 2 bars A ,4 bars A ,6 bars l
,
Sbarso
277
Influence du d6ficit hydrique sur la repense de I’ARA à la p0,
Les donnkes de la Fig.2 montrent qu’après environ 800 mn (environ 13 heures)
d’application d’une contrainte osmotique a l’aide de mannitol, 1’ARA devient à peu pr&s
stable. Cette stabilité relative permet donc d’étudier l’influence de p0, croissantes appliqukes
progressivement pendant plusieurs heures.
Chez Kingsoy, sur les plantes témoins n’ayant pas subi de contrainte osmotique,
l’ARA augmente jusqu’à 30% d’oxygkne puis baisse pour des p0, supkeures (Fig.3A),
En présence d’une contrainte osmotique de -0,2 et -0.4 MPa, l’AlU maximale (ARA
4
8BARS
6 0 0
1000
1 2 0 0
1400
1 6 0 0
1 8 0 0
-TEMPS EN MINUTES
Figure 2 : CinCtique de I’ARA chez le soja variéti HOLXXON. en fonction de l’intensité de la
contrainte osmotique (appliquée à l’aide de mannitol) : 0,4 , et 8 bars. Les mesures sont faites eu
continu à 20% d’oxygène. Chaque point est la moyenue de 5 à 8 n?p&tions.
ARA X du TE à 20% d’02
ARA X du TE à 20% d’02
w
0
0
B
:
0
0
0
279
correspondant a la p0, optimale) est atteinte pour une p0, plus élevée, voisine de 40%
et sup&ieure a 60% pour -0,6 IvlPa. La p0, optimale est donc damant plus élevée que
la contrainte hydrique est plus s&ère. Toutefois l’augmentation de la p0, ne permet pas
d’obtenir une ARA maximale comparable à celle du témoin.
Chez la variete Hodgson, en l’absence de contrainte osmotique, la p0, optimale observ&
dans fa Fig.3B est plus élevée que chez Kingsoy. En effet TARA est multipliée par 2
lorsque la p0, augmente de 20 à 40% d’oxygene ; elle baisse au delà de SO% d’oxygene.
De même que chez Kingsoy, l’ARA maximale en presence d’une contrainte osmotique
est observée pour des p0, d’autant plus élev&s que la concentration en mannitol est plus
forte. L’ARA a 60% 0, est alors d’environ 80 à 90% de celle du temoin n’ayant pas
subi une contrainte osmotique pour les traitements -0,2 et -0,4 MPa ; mais pour un deficit
hydrique sévère (-0,6 MPa) l’effet de la p0, est beaucoup plus faible (environ 25% de
1’ABA du témoin pour une p0, de 60% 03.
Influence du deficit hydrique sur la respiration nodulaire
En syst$me fermé chez Kingsoy, la respiration du système racinaim nodule est
influencée par l’intensif de la contrainte hydrique (Fig. 4A). Pendant le déficit hydrique,
la respiration diminue d’autant plus que la contrainte osmotique est plus intense. Au retour
a un régime hydrique non limitant, la respiration reprend legèrement, surtout pour la faible
intensité, mais ne retrouve pas le niveau du tkmoin. Hodgson se comporte de la même
façon que Kingsoy, mais est legèrement plus affecte que celui-ci, surtout pour le trai-
tement -0,6 h4Pa (FigAB). Après 24 h. de contrainte osmotique, la respiration chez
Kingsoy est de 95, 70 et 67% du témoin contre 88, 62 et 55% chez Hodgson respec-
tivement pour -0,2, -0,4 et -0,6 MPa.
Dans les expkïences de r@onse B l’oxygkne en continu, aussi bien chez Kingsoy
que chez Hodgson, le dégagement de CO, augmente avec la p0, au moins jusqu’à 60%
Or Contrairement à I’ARA, il n’y a pas Ctablissement d’un plateau et la respiration semble
être limitée par la p0, quelle que soit la variété ou l’intensité du déficit hydrique (Fig.5).
Ces mésures de l’émission de CO, réaMes simultanément a celles d’ARA permettent
d’etablir une relation linéaire entre ces deux activités (Witty er aI., 1983). La pente de
la droite de régression ainsi obtenue correspond à la quantité de CO, émise par molécule
de C,H, réduite. L’ordonnée a l’origine représente la respiration des tissus racinaire et
nodulaire qui n’est pas associée a la fixation. Les donnees présentées dans le
Tableau 1 ne rév&lent aucune difference significative entre les traitements pour ces deux
paran&tres, excepté pour la contrainte de -0,2 MPa. Les r&ultats obtenus chez Hogdson
(donndes non prkentkes) sont similaires. Ainsi la contrainte osmotique de -0,2 MPa aug-
menterait le coût Cnergetique de la fixation puisque le rapport vitesse d’krnission de
COJARA est voisin de ,lO p.mol CO,.p.mol-r C,H, pour -0,2 Mpa contre environ 5 chez
le temoin. Cette observation mériterait d’être vCrif& dans des travaux ultkieurs.
Influence de l’azote minkal sur I’lnhlbition par le deficit hydrique
L’action du nitrate est dépressive sur la nitrogenase : en effet la fixation est faible
chez les plantes avec nitrate et sans déficit hydrique comparees a celles soumises au même
traitement maïs sans nitrate (Tableau 2). Apres 24 h en présence de mannitol, les plantes
sans nitrate maintiennent une activité de plus 60% du témoin contre 12% chez les plantes
i., -i
~j.‘,.. ---
‘.
20
40
60
80
100
120
140
-20
f
27
40
60
80
T;;ps,,;~~,,;o
contrainte hydriaue
0 bar4 2 barsa 4 b a r s II
6 bars ü
8 bars x
Figure 4 : Influence de la contrainte osmotique sur la respiration des racines
nodulkes. Les mesures
sont effectuées simultakment
aux mesures d’ARA en système fermk, chez les vtidtis de soja
IUNGSOY (A) et HODGSON (T3). Le traitement au mannitol dure 24 heures. Les mesures sont
faites 1 heure avant le traitement, 5h et 24h après. Les mesures suivantes sont effectuées après retour
à un régime hydrique non limitant.
281
200
g 150
-ü-oc3d 100
E
3
pc
50
z
u
0
10
20
30
40
PO* m 02)
Obar A 2barsA 4 b a r s n 6barst-J
Figure 5 : Influence de la contrainte osmotique sur la respiration des racines noduM.es. Les
mesures sont effectuées simultan6ment aux mesures d’ARA en systhme famt, chez les vari6tis de
soja
KINGSOY (A) et HODGSON (T3). Les dsultats sont exprimks en pourcentage du tbrnoïn à
20% d’oxygène. Chaque point est la moyenne de 8 B 16 mesures sur une plante. L’expb-ience est
rép&?e deux fois.
282
Tableau 1 : Influence du déficit hydrique sur la respiration du système
racinaire nodulé et sur le coût énergétique de la fixation
de l'azote (ARAI chez le soja, variété Kingsoy. Les mesures
d'ARA sont faites en système continu. La respiration basale
correspond à l'ordonnée à l'origine de la droite de régres-
sion simple entre 1'ARA et la respiration, et le ooüt éner-
gétique est la pente de cette droite.
Intensité de la contrainte
Respiration basale
Coût énergétique de
osmotique en MPa
nitrogénase
'
pmol CO2 h-' pl-'
pmol CO2 umol -' C2H4
0
255 + 68
5,5 + 1,o
-
-
-0,Z
146 + 5 8
9,4 + 2,1
-
-
-0,4
224 + 89
4,3 + 2,6
-
-
-0,6
234 t 44
6,l + 2,l
-
:,
-.
.
.<
.
.
.
i
-1.
.
.
283
Tableau 2 : Effets cumulés du nitrate et du déficit hydrique sur la fixation
symbiotique de l'azote (ARAI du soja, variété Hodgson ; intensité
de la contrainte osmotique -0,4 MPa, pendant 24 h. Le point -1
correspond aux mesures faites une heure avant application du
mannitol, et le point 120h à celles réalisées 96h soit 4 jours
après suppression de la contrainte osmotique.
Heures
-1
ARA : umoles C2H4h -' Plante
après stress
Sans déficit hydrique
Avec déficit hydrique
-NO-3
+NO-3
-NO-3
+NO- 3
-1
52,0
11,o
52,0
11,o
+5
62,5
13,2
42,4
172
+24
59,7
25,0
37,1
330
+120
86,4
10,8
45,4
4,O
284
cultivkes avec 3 mM de nitrate. L’inhibition de la fixation par le nitrate et le déficit hy-
drique sont donc approximativement cumulatives.
A la suppression de la contrainte hydrique, les valeurs de l’ARA sont très faibles
chez les plantes avec nitrate, mais augmenteut assez rapidement. Chez les plantes sans
nitrate, l’inhibition de fa nitrog&rase n’est que partiellement levke.
Discussion et conclusion
D’après les travaux initiaux de Sprent (1976), la contrainte osmotique inhibe la
fition syrnbiotique de l’azote en limitant la quantiti d’oxygkne disponible dans les no-
dosit6s pour la synthkse de I’ATP qui est indispensable au fonctionnement de la nitro-
ghase. Les rthltats obtenus dans ce travail sur nodosites intactes en place snr la plante
sont en accord avec cette hypothèse. En effet l’inhibition de l’ARA par la comrainte os-
motique, due à l’apport du mannitol dans le milieu racinaire à 20% d’oxigeue, est lev6e
partiellement par l’augmentation de la p0, externe. Le d&placement de la p0, optimale
vers des valeurs d’autant plus Clevtks que la concentration de mannitol est plus forte,
ainsi que les observations sur la respiration nodulaire, renforcent cette interpretation et
suggerent que la diffusion de l’oxygene est perturbee par la contrainte osmotique.
Cependant, la manipulation de la p0, externe ne permet pas de lever totalement lkhi-
bition. En effet les valeurs dARA maximale en prksence de la contrainte osmotique sont
infkieures a celles observees snr les plantes tkrnoin ne recevant pas de mannitol. La
limitation de I’ARA par l’oxygène ne serait donc pas le seul mécanisme implique dans
l’inhibition de la fixation par une comrainte osmotique.
L’ARA maximale est vraisemblablement déterminée par le flux de pouvoir
réducteur aux bactkoïdes (Heckmann et Drevon, 1988) et par la quantité d’enzyme
active. L’irr&ersibilid de l’inhibition pour des pressions osmotiques inferieures à
-0.2 MPa suggère qu’tme fraction des protéines îmiispensables a fa fixation a &t5 dkruite
durant les 24 h d’incubation en presence du mannitol. Ceci est apparemment confkme
par l’observation de la Fig.2 : l’inhibition de l’AFUI durant les premières heures d’incu-
bation en prtkence de mannitol serait due à un ralentissement de la diffusion de 0, tandis
que la baisse plus lente d’activite nitrgenaso observée entre 14 et 20 h s’expliquerait par
une d&krioration du tissu nodulaire, due elle-même à un manque prolonge d’oxygene ou
encore a un manque d’énergie. Il est toutefois peu probable que œ manque d’énergie
au niveau nodulake provienne d’une diminution de la photosynth&se sous l’effet de la
contrairite osmotique car une diminution dela réduction foliaire de CO, par limitation
du flux de photons (Schweitzer et Harper, 1980) ou une stimulation par la pC0, ak-ienne
(I%n et Brun, 1982) n’affectent pas I’ARA avant au moins 24 h. Il est par contre pos-
sible que la contrantte osmotique ait ralenti le transfert des photosynthats aux nodositks,
œ qui d’après Sloger er al., (1975), diminue I’ARA à court terme, vraisemblablement
en limitant la quantid de pouvoir rkhtcteur disponible pour le fonctionnement des bac-
t&oides. Afin de vkritïer ces hypothkses, ii conviendrait de tester l’effet a plus court terme
du mannitol sur l’inhibition de la fixation et la reversibilid de cette inhibition, et de mesurer
l’activit6 ex-planta de bactkïdes prklev6.s à des moments successifs de la cidtique d’inhi-
bition. La présence de nitrate dans la solution nutritive pourrait, par son accumulation
vacuolaire, diminuer l’inhibition en permettant à la plante et aux nodositks de maintenir
leur turgescence. Les résultats pn%entés dans le Tableau 1 ne soutiennent pas cette hy-
285
pothbse puisque le pourcentage d’inhibition de I’ARA par une contrainte osmotique de
5,4 MPa est suphieure en présence (88%) qu’en absence (38%) de NO,-. L’inhibition
plus forte en prksence de cet ion serait due à son effet inhibiteur propre qui augmenterait
la rt!.sistance corticale a la diffusion de 0, dans les nodosités (Minchin et ai., 1986) et
diminuerait la quantité de pouvoir rkiucteur disponible pour les bactéroïdes (Heckmann
et Drevon, 1987). Ceci pourrait également être dû Zr une accumulation dans la nodosité
de nitrite issu de la réduction du nitrate (Rigaud, 1976 ; Streeter, 1982) d’autant plus
que cette accumulation de nitrite est plus forte en condition d’aliientation lim$ante de
0, (Hedmann et Drevon, 1987).
La comparaison des denx cultivars en l’absence de pertwbation osmotique, suggère
que Hogdson aurait d’avantage de rkserves glucidiques facilement mobilisables que
Kingsoy puisque sa p0, optimale pour l’ARA est plus élevée. Ceci pourrait expliquer
la meilleure r&wsibilite & l’inhibition chez Hogdson et soutiendrait lbypothkse d’un rôle
des réserves glucidiques dans le maintien de l’int&rité des nodositks en prksence d’une
contrainte osmotique de moins de 24 h.
Du point de vue agronomique, cette étude entraîne deux remarques :
(i) Il est reconnu que l’apport de nitrate à une culture permet B celle-ci
de mieux rkster à la sécheresse, car cet anion joue le rôle d’osmoticnm
dans la vacuole et permet à la plante de rester turgescente. Cependant,
cet anion diminue egalement la fixation et son rôle dépressif s’ajoute
ii cehri de la contrainte hydrique. Dans ce cas, le soja perd son intkêt
comme plante f-ce. Il serait utile de faire la même étude sur les
1Qumineuses tres résistantes à la skcheresse, telles celles des zones
dksertiques.
(ii) Le cuhivar de soja Kingsoy, dont le rendement est moins diminue que
celui d’Hodgson par une irrigation insuffisante (Paul, com. pers.), pn%kente
cependant une fixation d’azote très sensible au déficit hydrique. Cette
apparente contradiction est sans doute due au fait qu’un defîcit hydrique
de plusieurs jours ne provoque pas de perturbations physiologiques pro-
fondes chez Kinsoy du fait que la &ulahon stomatique est efficace
chez cette vari&.& (Salt, 1984). Par contre, œ déficit hydrique dkprime
plus vite et plus fortement son activite nitrog&tse, cette varieti pos-
skdant des nodositks ayant une plus faible rtkerve tkrgétique que celles
d’Hodgson.
Il y aurait donc sans doute intkrêt it rechercher des vari&& peu permrbees par
le delïcit hydrique, a la fois pour le rendement et la fixation de l’azote, ou additionner
œs deux crit&es par croisement.
Ce travail pr&minaire qui demande confiiation, notamment en poursuivant la
même recherche sur d’autres cultivars de soja ou même sur d’autres espèces de légumi-
neuses, montre l’int&t d’une étude physiologique sur les différentes étapes de Klabo-
ration du rendement pour optimiser la production d’une culture.
286
Remerciements
Les auteurs remercient Nathalie OLLAT pour son aide lors des premières exp&
riences sur la @onse de I’ARA a l’oxygkne. Ce travail a été partiellement fmcé par
la CEE-DG Xll (Contrat STD 1, No TS-A-180 F) et le Minist&re de llndustrie et de
la Recherche (Contrat 84 X 0881).
RQfbrences
BENETT, J.M. et ALBRECHT, S.L. 1984 Drought and flooding effects on N, fixation,
water relation, and diffusive resistance of soybean. Agron. J. 76, 735-740.
BOUNJOLS, A., PUECH, J. et MONDJES, M. 1982 Muence d’un déficit hydrique ap
plique durant la fructification sur la production du soja (Glyck mar L. Merrill).
Bull. Groupe français HumiaTimktre Neutronique. 11, 39-55.
DREVON, J J., KALJA, V.C., HECKMANN, M.O. et PEDELAHORE, P. 1988 In situ
open-flow assay of acetylene reduction activity by soybean nodules : influence of
acetylene and oxygen. Plant Physiol. Bbchem. 26, 73-78.
FJNN, A.G. and BRUN, A.W. 1980 Water stress effects on CO, assimilation, photo-
synthate partitioning, stomatal resistance and nodule activity in soybean. Crop Sci.
20, 431-434.
FJNN, A.G. and BRUN, A.W. 1982 Effect of atmospheric CO, enrichment on growth,
nonstructural carbohydrate content and root nodule activity in soybean. Plant Physiol.
69, 327-331.
HARDY, R.W.F., HOLSTElN, R.D., JACKSON, E.K. and BURNS, R.C. 1968 The ace-
tylene-ethylen assay for N2 fixation : laboratory and field evaluation. Plant P.hysioZ.
43, 911-914.
HEATH, R.L. 1977 Penetration of mannitol into the intracellular space of chlrorella soro-
kiniana. Plant Physiol. 59. 911-914.
HECKMAN, M.O. and DREVON, JJ. 1987 Nitrate metabolism in soybean root nodules.
Physiol. Plant. 69, 721-725.
HECKMAN, M.O. et DREVON, JJ. 1988 Effet du malate sur la @onse de l’activité
nitrogtkase (reduction de J’acétylene) des nodositks de soja à l’oxygene. CR. Acad.
Sci. Paris. 306, 559-564.
HUANG, J.C., CHEN, C.H. and BURRJS, R.H. 1973 Inhibition of nitrogenase-catalysed
reductions. Biophys. Biochim. Acta. 292, 256-270.
KALJA, V.C. et DREVON, JJ. 1985 Variation de l’activité nitrogenase (reduction de
l’acetylène) pendant l’incubation in situ des nodosités de Soja (Glycine max (L) Merr.).
CR. Ad. Sci. Paris 301, 591-596.
MJNCm, RF., INES MINGUEZ, M., SHEEHY, J.E., WJl-TY, J.F. and SKOT, L. 1986
Relationships between nitrate and oxygen in symbiotic nitrogen fixation by white clover.
J. Exp. Bot. 37, 1103-1113.
287
OBATON, M., MIQUEL,M.,ROBIN,P., CONEJERO, G., DOMENACH, A.M. et BARDIN,
R. 1982 Influence du déficit hydrique sur l’activité nitrate r&iuctase et nitrogénase
chez le soja (Glycine mi (L). Menill ; cv. Hodgson). CR. Acad. Sci. Paris. 294,
1007-1012.
PANKHURST, CE. and SPRENT, J-1. 1975 Effets of water stress on the respiratory
and nitrogen-ftig activity of soybean root nodules. J. Exp. Bot. 26, 287-304.
RALSTON, E.J. and IMSANDE, J. 1982 Entry of oxygen nitrogen into intact soybean
nodules. J. Exp. Bot. 33, 208-214.
RIGAUD, J. 1976 Effet du nitrate sur la fixation d’azote par les nodules de haricot (Phuseolus
vulgank L.). Physial. Vég. 14, 297-308.
SALL, K. 1984 Influence d’un déficit hydrique sur les développement végétatif et fruc-
tif& et sur l’alimentation azotee chez trois vari&& de Soja : Kinsoy, Jiangedou et
Quingmoushidou. Diplôme d’études approfondies, INP-Toulouse ; 94 pp.
SCHWEITZER, L.E. and HARPER, J.E. 1980 Effect of light, dark and temperature on
root nodule activity (acetylene reduction) of soybeans (Glycine max L. Mer-r.) Plant
Physiol. 65, 51-66.
SHEEHY. JE., MINCHIN, F.R. and WTITY, JF. 1983 Biological control of the resistance
to oxygen flux in nodules. Ann. Bot. 52, 565-571.
SINCLAIR, TR. and GOUDRIAAN, J. 1981 Physical and morphological constraints on
transport in nodules. Plant Physiol. 67, 143-145.
SLOGER, C., BEZDICED, D., MILBERG, R. and BOQNKERD, N. 1975 Seasonal and
diumal variation in N, (C~H.&fixing activity in field soybeans. Inter. Biol. Prog. 6,
271-284.
SPRENT, J.I. 1976 Nitrogen fixation by legumes subjected to water and light stress.
In : Nutman, P.S. (e&) Symbiotic nitrogen fixation in plants. pp. 405420 Cambridge
Univ. Pre.ss.
STREETER, J.G. 1982 Synthesis and accumulation of nitrite in soybean nodules sup-
plied with nitrate. Plant Physiol. 69, 1429-1434.
WEISZ, R.W., DENISON, R.F. and SINCLAIR, T.R. 1985 Response to drought stress
of nitrogen futation (acetylene reduction rates) by field-grown soybeans. Plant Physiol.
78, 525-530.
WERY, J., TURC, 0. and SALSAC, L. 1986 Relationship between growth, nitrogen frxa-
tion and assimilation in a legume (Medicugo sutivu (L.). Plant and Soil. 96, 17-29.
WITI’Y, JF., MINCHIN, FR. and SHEEHY, JE. 1983 Carbon costs of nitrogenase ac-
tivity in legume mot nodules determined by using acetylene and oxygen. J. Exp. Bot.
34, 951-%3.
Maximiser la FBA pour la Production Agricoie et Forest&e en Afiique
Effect of 16N-labelled mineral-N
and strains of Bradyrhizobium on biological
nitrogen fixation and N-partitioning
in cowpea (Vigna unguiculata) (L.) Walp.
LUYINDULA, N.(l) and WEAVER, R.W.t2)-
(1) Centre Rkgional &Etudes Nucldaires de Kimhasa.
BP. 868. Kinshasa Xl zaiie.
(2) Tear Agricdtmd Experiment Station ~FUI Texav A & M University,
College Station, Texas 77843, USA.
Abstract
‘SN-labelling technique was used to study the effects of different strains of
Bradyrhizobium and two doses of mineral-N on the fate of biologically fixed and
fertiker derived N on the partitioning of N in cowpea. Cowpea plants treated with
40 mg N plant-’ at planting had greater N-content and fixed more N2 than plants
receiving 240 mg N plant? in split applications. Plants treated with the lower dose
of mineral-N contained more N in all parts except the pods and roots as compared
to the. higher mineral N treatment. The high dose of mineral-N inhibited BNF of
cowpea, but N accumulation in the pods was not reduced For this treatment 52-
58% total N fixed and 32-58% of fetilizer-N absorbed were accumulated in the
pods. The nitrogen content of nodules was influenced by the strain of Bradyrhizobirun.
The strain of Bradyrhizobium also influenced the partitioning of N derived Erom
mineral fertilizer among plant parts.
289
Introduction
Cowpea is an important crop for tropical developing countries due to its high pro-
tein content. Scope for improving cowpea yield by applications of nitrogen fertilizer is
limited due to high costs of fertilizer nitrogen, especially in developing countries. Great
attention must, therefore, be paid to inoculation of cowpea with efficient strains of Bru-
dyrhizobium.
Many studies have been done concerning the effects of combined N on the phy-
siology of the Rhizobiumllegume symbiosis and the partitioning of f&r-N in .plants
during reproductive development (Summerfield et al., 1977 ; Eaglesham et al., 1983 ;
Douglas and Weaver, 1986 ; Weaver et al., 1987)
Many authors recommended the use of low doses of N to stimulate nodule growth
and N, fixation (Pate and Dart, 1961 ; Dart and Willon, 1970 ; Subba Rao et al.. 1974 ;
Eaglesham er al., 1983, Weavez et al., 1987). However, the determination of starter-N
level is complex and may depend on many factors, such as the form of the N-compound,
strain of Rhizobim, season, temperature, time of application, plant cultivar, etc...
Some experimental data indicated that applied N at a level lower than 360 mg
N/plant may have synergistic effects on N2 fixation by vigorously growing cowpea and
soybean (Broeshart, 1974, Eaglesham er al., 1983). Morever, Weaver et al. (1987), demons-
trated with glasshouse experiments that the addition of 60 mg plant-’ mineral-N (com-
pared with no addition) to cowpea changed the ranking of strains for effectiveness, in-
fluenced the total N content on shoots and whole plant and N concentration in leaves
and pods.
Partitioning of N during reproductive development of grain legume is largely a
genetic trait of the plant (Minchin et al., 1978) ; some authors found a Rhiwbium strain
effect on N partitioning in reproductive soybean plant parts (Morris and Weaver, 1983 ;
News et al., 1985).
In the previous research, rhizobial strains influenced N partitioning in plants, only
when they differed in effectiveness. The present work utilized four rhizobial strains of
equal effectiveness to determine their influence on N partitioning in cowpea in a long-
term greenhouse experiment under two doses of mineral-N.
Materials and Methods
Plant and bacteria
Cowpea (Vigna Imguiculuta) (L.) Walp. “Bush Purple Hull”) seeds were preger-
minated and planted in plastic pots of expanded vermiculite (Weaver, 1975). Four preger-
minated seeds were planted in each pot and were inoculated at’planting with 1 x lo7
cell I-’ per seed of Bradyrhizobium (strains BDC 22,6A, 24B from Texas A 8r M University
culture collection isolated from Barbados soils and strain 32lXl from Nitragin Co., Inc.).
One week after planting, the seedlings were thinned to one per pot. Plants were
watered with a N-free nutrient solution @vans et al., 1972). every 1 to 2 days and excess
nutrient solution drained from holes at the bottom into containers placed under the pots
(Eaglesham et al., 1983).
Mineral-N application
There were 10 treatments, five (four inoculated and one uninoculated) supplied
with 40 mg N/pot of 15N-labelled KNO, (1.2786 atom % “N excess) and five (four inocu-
lated and one uninoculated) supplied with 240 mg N/pot of lsN-labelled KNO, (0.4783
atom % 15N excess). The low dose was applied as a single application 1 week after planting,
and the higher dose was applied in split application 1.4, 5, 6 and 7 weeks after planting
at rate of 40, 40, 80, 40 and 40 mg N/pot respectively at each time. Also an unino-
culated treatment without N was included as a control to determine the quantity of N
coming from the plant support medium. These uninoculated plants without miner&N dis-
played early N-deficiency symptoms and yielded a maximum of 6.5 mg N/plant. Each
treatment was replicated four times and treatments were arranged in a completely rando-
mized design.
Harvesting and Nitrogen analysis
Plants were harvested at maturity (when 3-4 pods on each plant were completely
purple). Plants were separated into leaves, stems, peduncles, pods, roots and nodules and
oven dried (65-7O’C) for at least 72 h before dry weights were determined. Samples
of leaves, stems, peduncles, pods and roots were ground and N analysed by kjeldhal
method as described by Nelson and Sommers (1973) but was modified to include NO,
nitrogen.
l*N Determination
r5N content was determined by the method described by Hauck (1982) using a
Micromass 602-D mass spectrometer. From the atom % rsN excess in sample and atom
96 15N excess of mineral-N applied, the fraction of N in the plant sample derived from
fertilizer was determined by the formula :
Atom %15N excess in plant part
% N derived from fertilizer
=
X
100
Atom % r5N excess in fertilizer
Results and Discussion
Performance of rhizobial strains was not significantly different within each N
treatment with respect to total N and dry matter contained in various plant parts except
for nodules. The main effects are presented in Table 1. Starter N (40 mg plant -‘) resulted
in more dry matter being partitioned into nodules and less into roots as compared to
the split application of 240 mg N plant -I (Table 1). However, application of only starter
N resulted in more total N being contained in leaves, stems, peduncles and nodules
(Table 1). Whole plant dry matter was not significantly different (p = 0.05) for the two
N treatments (28.5 vs 29.8 g plant-‘) but total N content was higher for the starter N
treatment (624 vs. 596 mg plant “) as indicated by the F test for significance.
291
292
The dry weight of nodules was reduced by the continuous N treatment but within
N treatments rhizobial strains performed similarly (Table 2). Rhizobial strain bad a signi-
ficant influence on the N content of the nodules (Table 2). Nodules induced by strains
22 and 32Hl contained substantially more N than nodules induced by the other two strains.
The effect of strain was less pronounced when more mineral N was provided. The N
content of the nodules was a significant pcxtion of the whole plant N and for the starter
N treatment, the nodules contained more N than the roots (Table 1). The level of N
provided in the contirmous N treatment not only reduced nodulation (Tables 1 and 2)
but dinitrogen fixation (Table 3) even though the percentage of total N in the plant, derived
from the fertilizer was only 15%. Eaglesham et a2. (1983) reported a sigmficant decrease
in N, fxation when plant uptake of NO, provided more than 20% of the total N in the
plant The quantity of biologically fmed N contained in plant parts other than roots was
reduced by the continuous N treatment (Table 3) but the quantity of fertilizer N in all
plants was greatly increased.
Table 2 : Total dry weight and nitrogen content of nodules on cowpea as
influenced by strain of rhizobia and mineral nitrogen.
Dry weight
Total N
Strain
----___---__-_--_--________
------------_-------___________
Starter
Continuous
Starter
Continuous
----------g plant -1 ------__-
- -----__--_ mg plant -1 _-___-__-_
22
0.490
0.347
35.3
25.9
248
0.484
0.325
24.7
16.8
32Hl
0.535
0.340
40.2
24.7
6A
0.445
3
0.320
20.4
15.9
293
Table 3 : The influence of starter N and continuous N on partitioning
of biologically fixed N (BNF) and fertilizer N into various
plant parts.
BNF
Fertilizer N
Plant
part
Starter
Continuous
Starter
Continuous
- - -------______ -__-__-__ mg N plant-1 -____________--_-__-______
Leaf
154.0 (12.8)+
113.0
2.63
17.5
Stem
44.2 (5.27)
31.0
0.62
5.31
'Peduncle
39.2 (4.75)
23.7
0.32
4.55
Pod
326.0 (22.9)
291 .o
2.45
50.1
Root
23.5 (3.77)
23.0
1.15
6.32
Nodule
30.2 (5.83)
20.8
0.36
0.63
Total
617.1 (27.7)
502.5
7.53
84.41
+ The number in parenthesis is the LSD 0.05 for making comparisons on
the effect of mineral N treatments on the quantity of biologically
fixed N in the particular plant part. A test of significance was not
made for fertilizer N comparisons.
294
Fertilizer N treatments influenced the relative partitioning of fertilizer N into various
plant parts (Table 4). Distribution of starter N was largely into vegetative organs probably
because it was depleted from the plant growth medium relatively soon whereas for the
other N treatment the split applications made the N available throughout the growth period
and half of the featihzer N was contained in the pods (Table 4). The largest percentage
of the biologically fmed N was contained in the pods regardless of the fertilizer N treatment
In the case of the split application N treatment, the same proportion of the fertihzer N
and biologically fixed N in tbe whole plant partitioned into the pods. Eaglesham et al.
(1977) reported similar results for cowpea continuously provided with NO,. It appears
that the plant did not have a preference for the source of N for pods when both sources
of N were available throughout plant development. However, the data in Table 4 are
averages across four strains of rhizobia. The nodules only received 3% of their N from
mineral N in contrast to the roots which received 24% (calculated from Table 3). Eaglesham
et al. (197’7) also reported that nodules on plants fertilized with NO,- received 96% of
tbeii N from dinitrogen but roots received only 86% of their N from dinitrogen and concluded
that nodules were largely self-sufficient. Nineteen days after plants were pulsed with r5N
labelled NO,- the roots contained 19% of the NO, that was taken up and the nodules
only contained 2% (Douglas and Weaver, 1986).
Table 4 : Distribution of biologically fixed N (BNF) and fertilizer N
(FN) in plant parts of cowpea as a percentage of the total
plant N derived from each source.
Plant part
Leaves
Stems
Peduncles
Pods
Roots
Nodules
Nitrogen.
-
-
treatment
BNF
fN
BNF FN
BNF FN
BNF FN
BNF
FN BNF FN
-
-
-
-
-
-
-
-
m
m
-
-
-
Starter
25.1
35.2
7.3
8.3
6.3
4.3
5 3
3 2
3.8
1 5
4.8
4.9
Continuous
22.7
21.2
6.2
6.3
4.7
5.5
5 8
5 8
4.6
7.6
4.0
0.4
LSD.O5
2.3
'3.5
0.85
1.6
0.76
0.87
3.0
5.0
0.66 1.7 0.60 0.71
295
The strain of rhizobia did not influence uptake or distribution of starter N in the
plant (data not shown) but rhizobial strain influenced distribution of N from the continuous
N treatment (Table 5). The actual quantity of fertilizer N contained in various plant parts
was only signifkantly different for leaves, stems, and nodules. The coefficient of variation
was quite large for the reproductive plant parts and although large differences occurred
for pods they were not statist&ally significant. However, the data provide quite strong
evidence that the strain of rhizobia influenced the relative distribution of fertilizer N into
the pods (Table 5). For two of the strains approximately 52% of the fertilizer N was
Table 5 : The influence of rhizobial strain on the quantity and percentage
distribution of continuous N into plant parts of cowpea
Plant part
Strain
Leaves
Stems
Peduncles
Pods
Roots
Nodules
Total
-
-
-
-
-
-
---------------------------mg N plant-1 _____________--__--_------
6A
15.7
4.98
4.19
51.4 4.76 0.41 83.4
2 2
18.5
3.83
3.79
64.6 6.20 0.47 97.4
246
13.8
5.00
4.51
36.0 6.87 0.61 66.8
32Hl
22.1
7.42
5.72
48.3 7.46 1.04 92.0
LSD 0.05
4.4*
T.86*
2.21
23.9 2.96
0.24* 30.8
--_-___-____--__-__L--- ----i distribution------------------------
6A
19.7
6.32
5.15
'
62.5.. 5.85 0.53
2 2
20.1
3.90
3.95
65.2 6.45 0.48
-24B
20.8
7.50
6.62
53.7 9.88 0.93
32Hl
24.4
8.10
6.13
52.1 8.10 1.13
LSD 0.05
5.60
1.94*
1.76*
9.5*
2.18*
0.28*
* indicates significant difference by F-test (P = 0.05) between strains.
296
partitioned into the pods but for the other two strains approximately 63% of fertilizer
N was partitioned into the pods. These results indicate that the strain of rhizobia influenced
partitioning of fertihzr N in cowpea as has been reported for pigeon pea (Hernandez
et al., 1987). However, the data do not confii that cowpea strains influence uptake
of mineral N as report& by El Hassan and Focht (1986) but different strains of rhizobia
were utilized in our experiments and El Hassan and Focht (1986) grew their plants in
soil rather than a defined medium.
Evidence is accumulating that the strain of rhizobia may have an effect on nitrogen
partitioning in grain legumes. Our results indicated that four strains of rhizobia of equal
effectiveness based on total dinitrogen accumulated produced similar pod yields but the
relative distribution of fertibxer N partitioned into pods was different (fable 5). The strains
also influence the quantity of N contained in nodules frable 2). Depending on the mineral
N treatment the quantity of N in nodules was more or less than that contained in roots
(Table 1) which accounted for approximately 5% of the total plant N. Averaged across
strains of rhizobii the plant partitioned N from dinitrogen and continuously supplied mineral
N into pods in the same relative proportions (Table 4). This agrees with the data of
Eaglesham er ul (1977) in which 66% of the N from each N source was partitioned
into cowpea fruit. Thus, there was no evidence that the plant preferentially milked biologi-
cally fixed N over mineral N for pod development or that the rhizobial strain influen-
ced partitioning of biologically fixed N into pods.
Acknowledgements
This work was supported by funds from Regional Project S-170, the International
Atomic Energy Agency and US AID grant 84-CRSR-2-2521. The technical assistance of
Mrs. Heidi J. Mjelde is gratefully acknowledged.
References
BROESHART, H. 1974 Quantitative measurement of fertilizer uptake by crops. Netherland
J. Agric. Science. 22, 245.
DART, PJ. and WILDGW, D.C. 1970 Nodulation and nitrogen fixation by Vignu sineti
and Viciu atropurpuneu. The inlluence of concentration, form and site of application
of combined nitrogen- Australian J. Agric. Research. 21. 45%.
DOUGLAS, LA. and WEAVER, R.W. 1986Partitioning of nitrogen-15 labelled biologically
fvred and nitrogen-15 labelled nitrate in cowpea during pod development. Agronmy
J. 78, 499-502.
EAGLESHAM, ARJ., MINCHIN, F.R., SUMMERFIELD, RJ., DART, PJ., HULEY,
P.A. and DAY, JM. 1977 Nitrogen nutrition of cowpea Vigna unguicuha III. Distri-
bution of nitrogen within effectively nodulated plant Exp. Agric. 13, 369-380.
EAGLESHAM, ARJ., HASSOUNA, S. and SEEGERS, R. 1983 Fertilizer-N effect on
N, fixation by cowpea and soybean. Agronomy J. 75, 61-66.
EL HASSAN, G.A. and FGCHT, D.D. 1980 Comparison of inoculant and indigenous
rhizobial dinitrogen fmtion in Cowpeas by direct nitrogen-15 analysis. Soif Sci. Sm.
Am. J. 50, 923-927.
297
EVANS, HJ., KOCH, B., and KLUCAS, R. 1972 Preparation of nitrogenase from nodules
and separation into components. Methock Enzymology. 24, 470-476.
HAUCK, R.D. 1982 Nitrogen-isotope-radio-analysis. In : Page, A.L., Miller, R.H. and
Kenny, D.R. eds. Methods of Soil Analysis. Part 2. Ed ZAmericau Society of Agro-
nomy Soil Science Society of America Journal Madison. pp. 735-779.
HERNANDEZ, B.S., POTH. M. and FOCHT, D.D. 1987 Increased effectiveness of com-
petitive Rhizobium strains upon inoculation of Cajanus cajan. Appl. Environ. Microbial.
53, 2066-2068.
MINCHIN, FR., SUMMERFIE LD, R.J. and EAGLESHAM, ARJ. 1978 Plant genotype
Rhizobium strains interaction in cowpea Vigna unguiculatu (L.) Walp. Trop. Agric.
55, 107-11s.
MORRIS, DR and WEAVER, R.W. 1983 Mobilisation of 15N from Soybean leaves as
influenced by rhizobial strains. Crop Science. 23, 1111-l 114.
NELSON, D.W. and SOMMERS, LE. 1980 Total nitrogen analysis for soil and plant
tissues. Association Analytical Chem. 63, 770-778.
NEVES, M.C.P., DlDONET, A.D., DUQUE, F.F. and DOBEREINER, J. 1985 Rhizo-
bium strain effects on nitrogen transport and distribution on soybeans. Exp. Bot. 36,
1179-l 192.
PATE, J.S. and DART, PJ. 1981 Nodulation studies in legumes. IV. The influence of
inoculum strain and time of application of ammonioum nitrate on symbiotic response.
Plant and Soil. 15, 329-346.
SUBBA RAO, PAHWA, MR. and LAKESHMI KUMARI, M. 1974 Effects of combined
nitrogen on legume root nodulation. Acta Botunica Zndica. 2, 54.
SUMMERFIELD, R J., DART, PJ., HUXLEY, PA., EAGLESHAM, ARJ., MICHIN,
F.R. and DAY, J-M. 1977 Nitrogen nutrition of cowpea Vigna unguiculata L. effects
of applied nitrogen and symbiotic nitrogen fixation on growth and seed yield Exp.
Agric. 13, 129-142.
WEAVER, R.W. 1975 Growing plants for Rhizobium effectivenes tests. Soil Bid. Biochem.
7, 77-78.
WEAVER, R.W., DIPANKAR SEN, KYAN MOE and GRAHAM, RA. 1987 Effective-
ness of Bdyrhizobium sp. on cowpea provided with starter nitrogen. XI North Arne-
rican Rhiwbium Conference, August 9-15, Quebec, Canada.
Maximiser la FBA pour la Producrion Agricole et Forestibe en Afiique
Quantification of the contributkn of BNF
to field grown plants : the use of the 16N isotope
dilution technique - problems and some solutions
BODDEY, R M. and URQUIAGA, S.
EMBRAPA-UAPNPBS. Km 47. Seropkdica, 23851, Rio & Janeiro, Brazil
Summary
When the ‘?N isotope dilution technique is used to estimate the contribution of
biological N2 fixation (BNF) to plants, the assumption is made that both the “N2-
fixing” (test) and control crops remove N from the soil with the same “N enrichment.
The most practical method of labelling the soil with “N is to apply soluble labelled
featilkr but this @ces in the soil mineral N a rapid fall in lSN enrichment with
time. Hence plants with different uptake patterns can acquire soil-N with different
‘-9-l enrichments. While it is xzcornmended that control crops with the same uptake
patteznsasthetestcropsareused,thisrequirementisexcebdingly~~tttot
for, and in each study uptake patterns will vary with different edaphic and environ-
mental cm&ions. In these circumstances the use of several control crops in conjunc-
tion with slow release forms of lSN can be a successful strategy for the reliable quantifi-
cation of BNF. In a soil vjith a stable ‘?I label any control crop may be used or
even dispen& with altogether. These problems and possible solutions are discu-
ssed with examples of the quantification of BNF in both grain legumes and
grasses.
299
Introduction
In recent years the 15N isotope dilution technique has been widely used for the
quantification of the contribution plant-associated biological nitrogen fixation (BNF) to
both nodulated legumes and other crops. The acetylene reduction technique (Hardy et
al., 1968) has made a gmat contribution to knowledge in BNF research, but its use is
severely limited for the definitive quantification of contributions of N, fixation to higher
plants. The limitations of the application of this technique to the quantification of BNF
associated with nodulated legumes have been thoroughIy discussed in a very recent paper
by Witty and Minchin (1988), and the problems in its use to quantify possible N, fixation
associated with grasses and cereals have also been reviewed recently (Boddey, 1987).
While techniques using total N measurements can be very ‘valuable and should not be
discounted (Weber, 1966 ; Bell and Nutman, 1971 ; Nelson and Weaver, 1980 ; App
et al.. 1980, 1986 ; Lima et al., 1987), they are often of limited vahre in field expe-
riments unless very long term (Jenkinson, 1977, 1982).
It may be thought that the cost of using the stable isotope ‘5N is very high, but
a few hundred dollars (US) can buy sufficient of the Iabelled salts to perform one or
two field experiments. While the instruments to analyse ‘5N enrichment are very expensive
(a mass spectrometer is likely to cost upwards from US$ IOO,OOO, the less sensitive em-
mission spectrometer approximately 30 to 50% of this), commercial b&oratories in the
US and elsewhere will perform analyses for $10 to $20 per sample or even less if many
hundreds of analyses are required. However, if it is essential for the aims of the project
in question that the contribution of BNF to a crop be quantified, then the isotope dilution
technique can provide the answer if used with great care.
The principle of the 15N isotope dilution technique
The 15N technique requires that “N, fixing” crop under study (the test crop) is grown
in a soil enriched with ‘SN along with a control (“non-N,-fixing”) crop. The ‘5N enrichment
of the plant material of the control crop is regarded as representing the enrichment of
the plant-available (mineral) nitrogen in the soil. Hence if the test crop has a lower enrichment
than this it is concluded that unlabelled nitrogen from the air was incorporated into the
tissue of the test crop via N, fixation. It is apparent that the underlying assumption in
the use of this technique is that both the control and test crops remove nitrogen from
the soil with the same 15N enrichment.
Assuming this to be the case the calculation of the proportion of nitrogen in the
test crop derived from BNF (%NF) can be calculated from the equation :
Atom % 15N excess in test crop
. ..Eqn 1.
.’ Atom % *5N excess in control crop
>
It may be pointed out that the technique quantifies the contribution of N, Iixa-
tion specilically associated with the test crop, and that any input of unlabelled N derived
from rainfall, pollution or from BNF associated with cyanobacteria or other sources, if
equally available to both test and control crops, will not be detected by this technique.
For the acetylene reduction or total N balance techniques this may not be true.
300
Enrichment of the soil with 15N
It is inevitable that the test and control crops will remove N from the soil with
& me 15N enrichment if the soil is labelled uniformly in area and depth, and is stable
with t&e. III this case any crop which certainly has no associated BNF will we as
a valid control for any test crop. This approach has been used by several authors to
quantify BNF associated with both legumes and grasses (Kohl and Shearer, 1981; Broadbent
er al., 1982 ; Rennie et al., 1982 ; Butler and Ladd, 1985 ; Miranda and Boddey, 1987).
h fact the control crop in this case may be dispensed with altogether and the 15N enri-
chment of the N derived from the soil determined directly from measurements on the
soil mineral N, (Chalk ef al., 1983).
The most practical and usual method for enrichment -of a soil with 15N is by the
application of a solution of labelled nitrogen fertil&r (e.g. urea, ammonium sulphate
or potassium nitrate) to the surface of the soil. If the fertiKzer is uniformly distributed
over the area of soil to be planted then lateral variations in 15N enrichment should not
occur. Injection of labelled fert&er at points within the planted area such as that per-
formed by Rennie and Thomas (1987) are not to be recommanded as differences in lateral
root distribution between test and control crops may cause differences in the 15N en-
richment of the N derived from the soil.
The application of soluble fertilizer to the surf& of the soil does result in consi-
derable gradients with depth @add et al., 1981 ; Ledgard et al., 198.5). Boddey and Vic-
toria (1986) found that soil amended with repeated (small) applications of labelled am-
monium sulphate showed mean enrichments of mineralizable nitrogen of 0.140. 0.069,
0,048 and 0.030 atom % 15N excess in four successive 11 cm layers. They also found
evidence to suggest that this caused plants with different vertical root distributions to
obtain different “N enrichments in the soil-derived N.
However, the most important problem associated with the use of soluble labelled
fertiGzers is that immediately after application there is a sharp decline in lsN enrichment
in the mineral N in the soil, due to continuing mineralisation of unlabelled soil organic
N into a mineral pool continuously depleted by plant uptake, leaching and other losses
(Fried et al., 1983 ; Witty, 1983). This is illustrated by recent data from a study by
our group performed near Brasialia (Boddey, R.M., Urquiaga, S., Suhet, A., Peres, J.R.
and Neves M.C.P. in preparation). In one treatment of this experiment, inoculated
soybean was grown with four different control crops (a non-nodulating genotype of soy-
bean, sorghum, sunflower and rice) in soil amended with 10 kg N ha-l of lsN-labelled
potassium nitrate (10.0 atom 96 15N excess) at planting. The data show a continuous
decrease in “N enrichment in the plant aerial tissue due to the fall in the enrichment
of soil mineral N (Fig. I).
The reason for the initial incry in lsN enrichment of the soybean treatments
is that at the early harvests the unlabelled nitrogen derived fi-om the large N-rich soybean
seeds constitutes a large proportion of the N in the young plants. This contribution of
unlabelled N decreases as it becomes diluted with labelled N from the soil and the 15N
enrichment of the plants then increases to reflect that accumulated from the soil. This
effect is not significant for the sorghum and sunflower as the seeds are smaller and of
lower N content. It should be noted that if a small seeded cereal plant (e.g. wheat, barley,
sorghum etc) is used as a control for soybean or other large-seeded legumes, then at
early harvests the contribution of unlabelled seed N may be confused with that derived
from BNF.
I-
I-
I-
i--
0
ID
80
m
DAySAFIERGERhB(ATIoN
Figure 1 : lSN enrichment of nodulated soybean (
a ), non-nodulating soybean ( .+ ),
sorghum ( Q ) and sunflower ( d ) grown in a soil amended at planting with soluble lsN Iabelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % “N).
302
As the patterns of soil nitrogen uptake in this study were different (Fig. 2) this
resulted in the different control plants accumulating nitrogen from the soil with different
15N enrichments. This problem has been discussed by many authors and is particularly
well illustmted by the data reported by Witty (1983).
Most isotope dilution studies have been performed using single additions of soluble
labelled fettilizer in conjunction with the use of a single control crop (see reviews of
Chalk, 1985 and Danso, 1988). As the data displayed above, and that of Witty, illustrate,
in such studies it is dangerous to assume that different crops will accumulate N from
the soil with the same “N enrichment, and in many of these studies the different “N
enrichment observed in the nod&ted legume may have been partly due to differences
in N uptake patterns introducing an error of unknown magnitude to the resulting BNF
estimates.
It has been suggested that certain control crops are appropriate for specific legume
crops for use with this technique to quantify BNF (Rermie 1986). There is little evidence
from the sources cited by Rennie that these control crops were necessarily “appropriate”
(i.e. accumulated N with the same 15N enrichment as that derived from the soil by the
legume). Furthermore, as edaphic and environmental conditions will cause differential
changes in soil N uptake patterns, even if a control proved to be “appropriate” in one
experiment there is no guarantee that this would be true for other experiments on the
same legume at different sites, seasons or years.
Slow release forms of labelled nitrogen.
As Witty (1983) showed, if the decrease in r5N enrichment of soil mineral N is
slowed down by applying a slow release form labelled N fertilizer then the differences
in enrichment between different controls are attenuated. In our recent study near Bra-
silia a further treatment was included in the experiment where the same quantity of ‘5N
labelled potassium nitrate (10 kg N ha-r) was mixed with sugar to give a CzN ratio of
1O:l and added 7 days before planting. The patterns of uptake of soil N were not altered
significantly by the addition of sugar to immobilize the mineral nitrogen (Fig. 3) but
the rate of temporal decline in 15N enrichment was attenuated (Fig. 4) resulting in a narrower
range of enrichments of the four different control crops (Table 1).
Many different techniques have been utilized to slow down the decline in the 15N
enrichment of soil mineral N with time. Some workers have mixed various carbon substrates
(sucrose, glucose, cellulose, straw etc.) with labelled N fertilizer to immobilize the mineral
N (Legg and Sloger, 1975 ; Broadbent et al.,1982 ; Wagner and Zapata, 1982 ; Talbott
et al.,1982 ; Giller and Witty, 1987). Another similar technique adopted by some has
been to incorporate *sN-labelled plant material (especially prepared or resulting from other
studies) into the soil (Henzell et al.,1968 ; Boddey et al.,1984).
In a recent study on 10 cultivars of sugar cane, lsN-labelled ammonium sulphate
and potassium nitrate were composted with bagasse and filter cake from a sugar cane
mill to form a labelled organic matter which was mixed with soil to fill a concrete tank
(20 x 6 m) to a depth of 60 cm (Urquiaga et al.,1989). There was still a dramatic fall
in r5N enrichment of the soil mineral N with time (Fig. 5) although it should be emphasized
that the measurements were made over a period of almost 2 years. Data for the r5N enrichment
of the sugar cane cultivars and the control crop (Bruchiatiu rdcanr) at the first complete
harvest (250 days after emergence) suggested that there were considerable BNF contri-
303
0
1D
4l
60
80
m
DAYSAFllZCZ!RMNAllON
Figure 2 : Total N accumulation of nodulated soybean ( n ). non-nodulating soybean ( f ), sor-
ghum (Q ) and sunflower ( h ) grown in a soil amended at planting with soluble lSN labelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % 15N).
304
Figure 3 : Total N accumulation of nodulated soybean ( D ), non-nodulating soybean ( j- ), sor-
ghum (Q ) and sunflower (A ) grown in a soil amended 7 days before planting with 15N labelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % “N) mixed with sugar to give a C:N ratio of 10: 1.
3 0 5
22 T-
2
la
16
I4
12
1
a8
llfi
a4
02
0
Figure 4 : “N enrichment of nodulated soybean ( m ). non-nodulating soybean ( + ). sor-
ghum ( 0) and sunflower (h ) grown in a soil amended 7 days before planting with ‘W labelled am-
monium sulphate (10 kg N ha-‘, 10.0 atom % ‘?J) mixed with sugar to give a C:N ratio of 10:l.
306
Table 1. 15N enrichment of nodulated soybean and 4 control crops grown
in soil amended with 15N-labelled fertilizer, with and without
addition of sucrose. Means of 4 replicates.
ATOM % 15N EXCESS
CROPS
SOLUBLE FERTILIZER
FERTILIZER + SUGAR
Non-Nod Soybean
0.4039
0.3538
Sorghum
0.4367
0.3679
Sunflower
0.6209
0.4522
Rice
0.5044
0.4355
Nodulated soybean
0.3131
0.3208
Honest significant
Difference (P=O.OSl*
0.2083
0.1925
* Tukey test, Coefficient of variation 20.5%
L3
Figure S : ‘SN enrichment of soil mineral nitrogen extracted with 2M KC1 from Itaguai series soil
L2-
mixed with “N labelled compost. Data from Urguiaga et al (1988) and Urguiaga S., Cruz K.H.S.
and Boddey R.M. (in preparationj.
11 -
I-
0.9-
0.8 -
ti
CA7 -
?.i82
0.6-
8
E*
05 -
<
0.4 -
* 1.
03 -
0.2 -
0.1 -
u
El
0 (
I
I
I
I
I
1
I
1
I
0
loo
300
400
!3nJ
DAYS AFTER APPLICATlON
308
butions to several of the sugar cane cuhivars, notably Krakatau and SP 70-1143, but
because of the large fall in the 15N enrichment in the soil mineral N during this period
the estimates could not be regarded as accurate (Table 2) However, as subsequent to
this harvest there was very litde further drop in the enrichment of the soil mineral N,
the estimates of the BNF contribution to the ratoon crop can be regarded as being much
more reliable.
From the above disr;ussion it is apparent that in order to accurately quantify plant
associated BNF using the “N isotope dilution technique it is desirable that slow release
forms of labelled nitrogen are used. Nearly all of the slow release forms of N when
incorporated into the soil immobilize soil mineral N and hence change soil N availability
to the plants, and/or cause other changes in soil fertility. It is important therefore to examine
the objectives of the study in question to decide whether this will negate them. In our
work performed on grasses and sugar cane the interest has been in determining under
optimum conditions whether these plants are able to obtain significant contributions from
plant associated BNF and in the comparisons of cult&us with regard to this trait. In
this case the immobilization of nitrogen by the addition of slow release forms of labelled
N does not adversely affect the objectives of the experiment This may be true for many
studies with legumes, but there will be circumstances where it is necezzuy that the soil
N availability is not significantly affected by application of the isotope dilution technique.
One form of slow release labelled N which would probably only marginally affect
soil N availability is that based on the gypsum pellets described by Witty (1983). However,
more recent data suggests that release of N from such pellets is not much slower than
soluble N (Giller and Witty, 1987 ; Boddey et al.,1989). One other solution that may
be possible is the addition of small quantities of soluble labelled N at regular (e.g. weekly)
intervals. This thechnique has been adopted by many authors (Vallis et aI.,1%7 ; Goh
et al, 1978 ; Boddey et al., 1983, 1984 ; Ledgard et al., 1985) but as Rennie (1985)
has pointed out this technique has its dangers if sequential harvest of plants are to be
made and soil N availability is high. If in this case the test and control crops remove
different quantities of N from the soil then the residual mineral N in the soil at the sub-
sequent harvest will be different. Further addition of labelled N to the soil will then re-
sult in a different initial enrichment of mineral nitrogen for the two crops at the start
of the next growth period. Experience to date suggests that the use of small frequent
additions of labelled fertilizer may be the best technique of labelling the soil if soil N
availability is not to be radically altered (Boddey et af.,1983. 1984 ; Ledgard er al., 1985b)
but further, more detailed, studies on this subject are required.
A strategy for the field application of the isotope dilution technique
Unless the 15N enrichment of the soil mineral N is known to be reasonably stable
with time it is advisable that more than one control crop should be used within the same
experiment. This strategy was recently recommended following experiments performed
by our group at two different sites in Brazil (Boddey et u1.,1989). In both experiments
the objective was to quantify BNF contributions to inoculated and non-inoculated soybean,
cowpea and groundnut using three different control crops (a non-nodulating genotype of
soybean, sorghum and rice).
At the fiit site (Rio de Janeiro) the soil was enriched with 15N by addition of
labelled gypsum pellets (Witty, 1983). The data showed that there were considerable and
Table 2 : Estimates of the contribution of biological nitrogen fixation (BNF) to 10 sugar cane varieties
1st HARVEST 250 DAET
2nd HARVEST 380 DAE1
Varietv/
species
Total N
15N
Estlmate of the
Total N
T5N
Estimate of the
accumulation
enrichment
contribution. of BNF2
Accumulation
Enrichment
contribution of
(kg.ha-1)
(ATM % 15N ext.)
%
(kg.ha-1)
(kg.ha-1)
(ATM % 15N ext.) BNF2
%
(kg N.ha-1)
CB 47-89
265 ab
0.316 bc
41.9
111
209 bc
0.108 bc
19
41
CB 45-3
257 ab
0.305 bed
44.1
110
258 b
0.120 ab
10
27
NA 56-79
246 ab
0.313 bc
42.9
106
188 bed
0.126 a b
6
11
IAC 52-150
271 ab
0.300 bc
46.2
125
173 bed
0.115 abc
14
24
SP 70-1143
245 ab
0.261 bc
52.3
128
216 b
0.109 bc
19
40
SP 71-799
244 ab
0.298 bc
46.1
112
200 bc
0.109 bc
19
37
SP 79-2312
202 abc
0.344 bc
41.7
84
277 b
0.127 ab
5
14
Chunnee
152 c
0.349 b
37.2
56
119 bed
0,124 a b
7
9
Caiana
49 d
0.309 bc
42.2
21
26 d
0.123 ab
8
2
Krakatau
294 a
0.255 c
55.6
163
452 a
0.097 c
28
125
B. radicans
189 bc
0.546 a
43 cd
0.134 a
Coefficient
of variation (%I
27.0
11.4
35.6
7.6
'Days after emergence
2Values calculated using the isotope dilution technique using B. radicans as control.
310
significant (P < 0.05) differences between the 15N enrichments of the three different control
crops but all were far higher in i5N enrichment than any of the nodulated legumes which
indicated that the legumes were all obtaining considerable contributions from BNF
(Table 3). For each nodulated legume, three estimates of the BNF contribution could
be made both by using the total N difference method (total N in nodulated legume -
total N in control crop) and the isotope dilution technique and these estimates are dis-
played in Fig 6. It is apparent that the three different estimates of BNF for each nodulated
legume calculated from the total N difference are more diverse than the three estimates
derived from the isotope dilution technique. Furthermore, the estimates for cowpea are
negative and hence obviously erroneous. This was due to the fact that the cowpea grew
very poorly due to fungal disease and accumulated very little N from the soil. These
data show that, at least in this case, the isotope dilution technique yielded far better results
than the total difference N technique. However, the date show that despite the considerable
differences in 15N enrichment between the different control crops the estimates of BNF
contributions fall within a fairly narrow range. The probability that the true estimate of
the BNF contribution lies within this range increases with increasing numbers and diversity
of control crops.
At the second site (Brasilia) 15N-labelled compost (Urquiaga et al.,1989) was used
to label the soil. At this site there were smaller differences between the enrichments of
the different control crops suggesting that this form of slow release N was superior to
the gypsum pellets. This resulted in relatively small differences between the isotope dilution
estimates of BNF contributions (Fig. 7) and leads us to recommend this technique of
combining slow release fertilizer (or perhaps repeated small additions of soluble fertilizer)
with the use of several (3 or more) control crops. The advantage of using several control
crops of a diverse nature is that it gives the experimenter an idea of the range of possible
15N enrichments that crops of different growth habits are able to obtain from a soil with
.
a non-uniform 15N enrichment.
Aknowledgements
Financial support for many of our recent studies on the quantification of biological nitrogen
fixation has been provided by the Financiadora de Rstudos e Rrojetos (FINEP) of Brazil,
and by the U.S. National Academy of Sciences, National Research Council through a
grant from the U.S. Agency for International Development.
Table 3 : final grain yield, and dry matter nitrogen and 15N9)recovery at mid-pod fill stage of 3 cqntrol crops and
inoculated and non-inoculated soybeans, cowpea and groundnuts. Experiment 1, Riz de Janeiro. Means of
4 reolicates.
Dry matter
15N recovery
Grain(3)
Treatment
accumulation(2)
Total N(2)
% 15N1
g.m-2
mg 15N excess
yield
g.m-2
excess
m-2
kg ha-l
Soybean inoc'd(4)
29 w
621
20.6a.(5)
2.56bY5)
Soybean inoc'd(4)
638
20.3a
0.0104e
2.15bc
Soybean non ir#oc'd(4)
589
18.8ab
0.0125e
2.32b
Cowpea inoc'd
202
3.6de
0.0289cd
1.02cd
304
wz
Cowpea non-inoc'd
144
2.5e
0.0315cd
0.78d
334
Groundnut inoc'd
669
16.3ab
0.0143e
2.14bc
1710
Groundnut non-inoc'd
593
14.7b
0.0219de
3.29b
2322
Sorghum
1142
9.7c
0.0574a
5.08a
Soybean, non-nodu-
lating
408
7.7cd
0.0398bc
3.04b
Rice
443
5.2de
0.0484ab
2.67b
Coefficient of
Variation %
18.3
21.4
35.5
32.2
1. 15N added as slow release gypsum pellets (Witty, 1983) at 2.15 kg labelled N/ha
2. Mid-pod fill stage, 110 days after planting, Harvested area = 1 m2/plot.
3. Nature grain stage, 117 and 141 days after planting for cowpea and groundnut, respectively. Harvested area 2 mZ/plot.
4. Grain yield lost due to weevil attack.
5. Means followed by the same letter are not significantly different at P = 0.05 (Duncan).
Figure 6 : Total N difference ( El ) and isotope dilution ( n ) estimates of the contribution of N,
fixation to inoculated and non-inoculated soybean, cowpea and groundnut using 3 different control
crops ; sorghum (S), non-nodulating soybean (Y) and rice (R). Experiment 1, Rio de Janeiro. Means of
4 replicates. Data from Boddey et al (1989).
S O Y B E A N
C O W P E A
G R O U N D N U T
Inoculated
Inoculated
Non
Inoculated
Non
Inoculated
Non
2 9 w
C.8 1 8 0 9
Inoculated
BR 53
inoculated
I I A
inoculated ’
i
I I
I
I
I
I
I
I
11:
S Y R
SYR S Y R
S Y R S Y R O’?
In
313
z-uI
Contrlbutlon of
’
N6
0
UOllOX()
N,
flxatlon
‘N
40
Lo
UOl+l’lq(l(UO3
aN
.
rn-*
314
References
APP, A.A., WATANABE, I., ALEXANDER, M., VENTURA, W., DAES, C.,
SANTIAGO, T. and D DATTA, SK. 1980 Nonsymbiotic nitrogen fixation associa-
ted with the rice plant in flooded soils. Soil Science. 130, 283-289.
APP, A.A., WATANABE, l., VENTURA, T.S., BRAVO, M. and JUREY, C.D. 1986
The effect of cultivated and wild rice varieties on the nitrogen balance of flooded
soil. Soil Science. 141, 448-452.
BELL,F., and NUTMAN, P.S. 1971 Experiments on nitrogen fixation by nod&ted luceme.
Plant and Soil. Special Volume, 231-264.
BODDEY, RM., CHALK, PM., VICTORIA, R-L., MATSUI, E. and DOBEREINER.
J. 1983 The use of the 15N isotope dilution technique to estimate the conhibution
of associated biological nitrogen fixation to the nitrogen nutrition of Paspahm po-
tatum cv. batatais. Can. J. Microbial. 29, 1036-1045.
BODDEY, R.M., CHALK, P.M., VICTORIA, RL. and MATSUI, E. 1984 Nitrogen fma-
tion by nodulated soybean under tropical field conditions estimated by the ‘W iso-
tope dilution technique. Soil Biol. Biochem. 16, 583-588.
BODDEY, R.M and VICTORIA, R.L. 1986 Estimation of biological nitrogen fixation
associated with Bra&aria and Paspalum grasses using ‘SN labelled organic matter
and fertiIizer. Plant and Soil. 90, 265-292.
BODDEY, R.M. 1987 Methods for quantification of nitrogen fixation associated with
gramineae. CRC Critical Reviews in Plant Sciences. 6. 209-266.
BODDEY, R.M., URQUIAGA, S., SUHET, AR., PERES, JR. and NEVES, M.C.P. 1989
Quantification of the contribution of N2 fixation to field-grown grain legumes : A
strategy for the practical application of the ‘% isotope dilution technique. Soil Bid.
Biochem. (Submitted)
BROADBENT, F.E., NAKASHIMA, T. and CHANG, G.Y. 1982 Estimation of nitrogen
fixation by isotope dilution in field and greenhouse experiments. Agronomy J. 74,
625-628.
BUT&R, J.H.A. and LADD, JN. 1985 Symbiotically-fixed and soil-derived nitrogen
in legumes grown in pots in soils with different amounts of available nitrate. Soil
Biol. Biochem. 17, 47-55.
CHALK, P.M. DOUGLAS, L.A. and BUCHANAN, S.A. 1983 Use of 15N enrichment
of soil minezaIizable N as a reference for &ope dilution measurements of biolo-
gically fixed nitrogen. Can. J. Microbial. 29. 1046-1052.
CHALK, P.M. 1985 Estimation of N2 fixation by isotope dilution : an appraisal of tech-
niques involving lsN enrichment and their application. Soil Biol. Biochem. 17, 389-
410.
DANSO, S.K.A. 1988 The use of lsN enriched fertilizers for estimating nitrogen fixation
in grain and pasture legumes. In : Nitrogen fixation by legumes in Mediterranean
agriculture. Beck,D.P. andMateron,L.A.Eds. pp, 345-358. MartinusNijhoff, Dordrecht,
Netherlands.
315
FRIED, M.. DANSO, S.K.A. and ZAPATA, F. 1983 The methodology of measurement
of N, fixation by non-legumes as inferred from field experiments with legumes. Can.
J. Microbial. 29, 1053-1062.
GILLER, K.E. and WITTY, J.F. 1987 Immobilized 15N-fertilizer sources improve the
accuracy of field estimates of N,-fixation by isotope dilution. Soil Biol. Biochem.
19, 459463.
GOH, K.M., EDMEADES, D.C. and ROBINSON, B.W. 1978 Field measurements of
symbiotic notrogen fixation in an established pasture using acetylene reduction and
a 15N method. Soil Biol. Biuchem. 10, 13-20.
HARDY,R.WF.,HOLSTEN,R.D.,JACKSON,E.K.andBURNS,R.C.l968Theacetylene-
ethylene assay for N,-fixation : Laboratory and field evaluation. Plant Physiol. 43,
11851207.
HENZELL, E.F., MARTlN, A.E., ROSS, PJ. and HAYDOCK, K.P. 1968 Isotopic studies
on the uptake of nitrogen by pasture plants. IV. Uptake of nitrogen from labelled
plant material by rhodes grass and siratro. Ausr. J. Agric. Res. 19, 65-77.
JENKINSON, D.S. 1977 The nitrogen economy of the Broadbalk experiments. I. Nitrogen
balance in the experiments. In : Rothamsted Experimental Station Report for 1976,
part 2, Harpenden, Heats, U.K., pp; 103-109.
JENKINSON, D.S. 1982 The nitrogen cycle in long-term field experiments. PhiZsophicul
Transactions of the Royal Society (London) B. 296, 563-571.
KOHL, D.H. and SHEARER, G. 1981 The use of soils highly enriched in 15N to screen
for N,-fixing activity. Plant and Soil. 60, 487-489.
LADD, J.N., OADES, J.M and AMATO, M. 1981 Distribution and recovery of nitrogen
from legume residues decomposing in soils to wheat in the field. Soil Biol. Biochem.
13, 251-256.
LEDGARD, SF., SIMPSON, JR., FRENEY, JR. and BERGERSEN, F-J. 1985 Field
evaluation of lsN techniques for estimating nitrogen fixation in legume-grass associations.
Aust. J. Agric. Res. 36, 247-258.
LEGG, J.O. and SLOGER, C. 1975 A tracer method for determining symbiotic nitrogen
furation in field studies. In : Proceedings of the 2nd-International Conference on Stable
Isotopes. Klein, E.R. and Klein, P.D. Eds. pp. 661666. Oalbrook, Illinois, USA.
LIMA, E., BODDEY, R.M. and DOBEREINER, J. 1987 Quantification of biological ni-
trogen fmtion associated with sugar cane using a 15N aided nitrogen balance. Soil
Biol. Biochem. 19, 165-170.
MIRANDA, C.H.B. and BODDEY, R.M. 1987 Estimation of biological nitrogen fixation
associated with 11 ecotypes of Punicum maximum grown in isN labelled soil. Agronomy
J. 79, 558-563.
NELSON, AN. and WEAVER, R.W. 1980 Seasonal nitrogen accumulation and fmtion
by soybean grown at different densities. Agronomy J. 72, 613-616.
RENNlE, R.J. 1985 Theoretical errors in sequential Nrs labelling of soil to estimate N,
fmation by isotope dilution. Crop Science. 25, 891.
316
RENNTE, RJ. 1986 Advantages and disadvantages of nitrogen-15 isotope dilution to quau-
tity dinitrogen furation in field-grown legumes - A critique. In : Field Measurement
of Dinitrogen Fixation and Denitrification. Soil Science Society America Special Pu-
blication 18. pp. 43-58. American Society of Agronomy, Madison, Wisconsin, USA.
RENNIE, RJ. and THOMAS, J.B. 1987 15N-determined effect of inoculation with Nz fixing
bacteria on nitrogen assimilation in western Canadian wheats. Plant and Soil. 100.
213-223.
RENNIE, RJ., DUBETZ, S.. BOLE, J.B. and MUENDEL, HI-L 1982 Dinitrogen furation
measured by 15N isotope dilution in two Canadian soybean cultivars. Agronomy J.
74, 725.
TALBOTT, H., KENWORTHY, WJ. and LEGG, J-0. 1982 Field comparison of the ni-
trogen-15 and difference methods of measuring nitrogen fixation. Agronomy J. 74,
799-804.
URQUIAGA, S., BOTI’EON, P.B.L. and BODDEY, R.M. 1989 Selection of sugar cane
cultivars for associated biological nitrogen fixation using ‘sN-labelled soil. In : Nitro-
gen fixation with non-legumes. Skinner, F., Boddey, R.M and Fend& I. Eds. pp
311-319. KIuwex Academic Publishers, Dordrecht, Netherlands.
VALLIS, I., HAYDOCK, KS., ROSS, PJ. and HENZELL, E.F. 1967 Isotopic studies
on the uptake of nitrogen by pasture plants. III. The uptake of small additions of
15N labelled fertilizer by rhodes grass and Townsville luceme. Aust. J. Agric. Res.
18, 865-877.
WAGNER, G.H. and ZAPATA, F. 1982 Field evaluation of reference crops in the study
of nitrogen fixation by legumes using isotope techniques. Agronomy J. 74, 607612.
WEBER, CR. 1966 Nodulating and non-nodulatmg soybean isolines. II. Response to ap-
plied nitrogen and modified soil conditions. Agronomy J. 38, 46-49.
WITTY, JF. 1983 Estimating N2-fixation in the field using ‘5N-labekl fertilizer : some
problems and solutions. Soils Biol. Biochem. 15, 631-639.
WITTY, J.F. and MlNCHIN, F.R. 1988 Measurement of nitrogen fixation by the ace-
tylene reduction assay ; myths and mysteries. In : Nitrogen furation by legumes in
Mediterranean Agriculture. Beck, D-P. and Materon, L.A. Eds. pp. 331-343. ICARDA,
Aleppa Syria
Maximiser la FBA pour la Production Agricole et Forestibe en Afrique
Utilisation de la m&hode 16N pour estimer
la fixation symbiotique de l’azote
chez les plantes herbakes et ligneuses.
DOIMENACH, A.M., KURDALI, F. et BARDIN, R.
Universitk Lyon 1. laboratoire de Biologie des So&
VRA CNRS 71. bat. 741, 43 bid ak 11 novembre 1918,
69622 VILLElJRBAhWE Ceder, France.
La m&ode de mesure de la fixation symbiotique de l’aznte bask sur les varia-
tions isotopiques naturelles de l’azote est un outil privilkgiB pour des 6tudes sur le
terrain et en conditions naturelles. En effet elle ne n&e&e aucune intervention
de la part de rexpkrimcntateur
et n’entraine pas de perturbation du systkme ttudib.
Cet article veut mettre en 6vidence les exigences, d&e&nation de l’identiti isoto-
pique de l’azote fix6 et assinG. et les contraintes de cette mkthode, en particulier
au niveau de l’kchantillonnage,
tant pour les plantes annuelles que pour les arbres.
Les quelques exemples d’application don& illustrent les grandes possibiitis de cette
m&ode qui pennet, en particulier, d’aborder l’ktude des systkmes ligneux.
Introduction
L’estimation de la fixation symbiotique d’azote est we &ape essentielle dans toute
&de des systkmes fixateurs. En effet, aussi bien au niveau globaliste, pour connaitre
les rentrkes d’azote dans un tkzosyst~me et pour les optimiser, qu’au niveau rkduction-
nke pour tester des cultivars ou des souches microbiennes s&xXio~&~ ou transform&s,
318
la quantification s’impose. Aussi est-il nt?cessaire d’avoir des outils de mesures fables
a sa disposition.
Une nouvelle generation de spectrometres de masse a double introduction et double
collection dune tres graude sensibilite, a permis de penser qu’il Ctait possible d’utiliser
les differences isotopiques namrelles de l’azote entre fair et le sol pour estimer, chez
les plantes fixatrices, la part d’azote venant de l’une et l’autre source (Delwiche and Steyn,
1970 ; Rennie et al., 1976).
Mais si les variations isotopiques naturelles de l’azote ont Cte mises en evidence
d?s 1939 par Schoenheimer et Rittenberg, variations qui, d’apr& Hoering (1955), auraient
leurs explications dans les processus biologiques, leur utilisation darts la comprehension
et la quantification des processus biologiques a &5 plus difficilement admise. Edwards
(1975) trouve ces variations trap faibles pour Ctre exploitables. Les r&ultats de Kohl
et al., 1971 sur la quantitr5 de nitrate polluant la nappe phr&ique calculee a partir des
variations isotopiques naturelles ont souleve une vive polemique (Hauck ef al., 1972 darts
“a questionable approach”). Bremner (1977) s’etonne, que les critiques formulees n’aient
pas d&courage les recherches a ce sujet et pour lui, mCme si ces variations existent, elles
ne peuvent dormer que des informations qualitatives.
Pourtant, depuis 1977, la m&hode de mesure de la fixation symbiotique bas& sur
les variations nature&s des abondances isotopiques de l’azote s’est ClabonZe (Amarger
er al., 1977, Bardin et al., 1977, Kohl er al., 1980). Si cette m&rode prrjsente certaines
contraintes et limites, son champ &application reste t&s vaste et inclut m&ne les arlxes
fixateurs d’azote. Sa sp&ificite reside essentiellement darts le fait qu’elle n’implique au-
curie intervention de l’exp&imentateur et peut done Ctre appliqut5e sur les systi?mes na-
turels sans les modifier ni les perburber.
Principe
Rappelons que l’abondance isotopique dun compose azou5 est le pourcentage du
nombre d’atome 15N par rapport au nombre total d’atomes d’azote. L’azote atmospherique
mot-me une composition isotopique t&s stable qui est estimee a 0,3663 f 0,0002% r5N
(Mariotti 1984) et celle des composes azotes varie entre 0,364O et 0.37% 15N.
Ces variations sont l&s a l’existence de fractionnements isotopiques. On peut de-
fmir le fr-actionnement comme &ant la repartition diffr5rent.e des isotopes dun element
contenu darts le produit initial et final au tours dune reaction chimique ou biologique.
Les liaisons etablies par l’isotope leger seraient plus ais&nment bris&s que celles &&lies
par I’isotope lourd et les mol&tles contenant l’isotope leger r&$raient plus vite que celles
contenant l’isotope lourd Chaque processus biologique est une suite de phrsieurs reactions
et la teneur isotopique finale est une rrZ.sultante de l’ensemble des fractionnements.
Le principe de la m&rode de mesure de la fixation symbiotique bast5e sur les
variations isotopiques nature&s de l’azote est simple : une plante fixatrice est aliment&
par deux sources d’azote, l’azote mol&zulaire de l’air et l’azote min&-al du sol. Si ces
deux sources ont une abondance isotopique differente, iI peut ttre possible, comme on
le fait apms un marquage, de connaitre par un simple calcul de dilution isotopique, le
pourcentage d’azote venant de chacune de ces deux sources. Tout consiste done a determiner
la valeur isotopique de l’azote venant du sol et celle venant de lair.
319
Valeur isotopique de I’azote fix4
MCme si on sait que lair a une valeur isotopique constante (Mariotti, 1984). celle-
cipeutCtremodifi~lorsqueI’azotemoI~ulaireestm~bo~paruneplante.Cettemodification
t&s faible, negligeable lorsqu’on utilise le marquage, peut entraker, si on n’en tient pas
compte, une grossikre erreur darts l’estimation du taux de fixation pouvant aller jusqu’a
donner des valeurs aberrantes. Cette m&ode nkessite done la determination de l’abon-
dance isotopique de l’azote provenant de la fmation. Cette determination est faite a partir
dune plante fmatrice poussant sur un milieu depourvu d’azote et qui ne dispose que de
l’air comme source d’azote.
Dans la pratique, la mkthode utilise les valeurs isotopiques des parties akiennes
des planks fixatrices car souvent les racines sont difficilement accessibles. Darts ce cas,
il est n6cessaire de tenir compte egalement du fractionnement isotopique possible de l’azote
se produisant a I’intkrieur mCme de la plante. On a pu voir en effet, que les nodosites
du soja et du robinier s’emichissaient en 15N pendant que la partie aerienne s’appauvrissait
en 15N (Reinero et al., 1983, Domenach, 1985). Ce fractionnement lit a la fixation peut
Ctre different suivant l’espece de la plante h8te (Mariotti et al., 1980, Steele et al., 1983)
ou de la souche microbienne (Yoneyama et al., 1984). Si on a pu montrer que, si darts
le genre Aims, lksphx (A. incana et A. ghtinma) et le type de souches de Frankia
influencent peu la composition isotopique de l’azote fv;C (Dommenach et al., 1988), la
valeur trouv~ (615N = -2) est differente de celle du soja (PN = -3). 11 est done rkessaire
de determiner pour chaque syst&ne fixateur Ctudie, le fractionnement isotopique se produisant
a l’intirieur de la plante avant de pouvoir appliquer cette m&rode. Ce fractionnement
evoke au tours du temps au mains durant les premiers stades de developpement de la
plante (Shearer et al., 1980 ; Fig. 1) puis semble se stabiliser. En effet, les feuilles d’auhres
ages de 5-6 ans transplantf5s sur un milieu sans azote ont une valeur isotopique identique
B celle de jeunes planks de 8 mois (Domenach et al., 1988). Aussi il est nt5cessair-e
de kisser pousser les plantes sur milieu sans azote pendant phtsieurs mois avant de pou-
voir d&z-miner la valeur isotopique de l’azote fmt.
6-
4 ..
Figure 1: Evolution en
faction du tunpa du
3 .
6% des nod&t&, par-
tie3 ntriamcs ct plantes
2 ‘.
entiircs d e Robinin
psf?udoacacia palssant
sur milieu saris azote
(d’aprkr Domcnach.
1985).
320
Les valeurs trouvks de ce fractionnement varient d’un laboratoire a un awe et
semblent en park likes aux procedures de preparation des kchantillons et au type d’appareil
utilisk (Shearer and Kohl, 1986). Cette difficult& vient egalement du fait qu’il n’existe
pas de gaz standard dont on connait la valeur absolue. Ces variations d’un appareil a
un autre ne semblent pas gkuuu.es pour la mesure du taux de fixation si l’ensemble des
mesures se rapportant a une experience sont faites aver la mEme m&ode analytique
et sur le mCme appareil. Ceci impose a chaque laboratoire de d&erminer p&ciskment
la valeur de ses param&res et de ne pas utiliser pour ses propres piantes les valeurs
dorm&s par ia littkrature.
Devoir d&erminer les fractionnements isotopiques ii& a la fmtion de chaque syst&me
symbiotique est tme contrainte props a cette m&rode qui vient de l’exploitation de faibles
variations isotopiques qui sent de l’ordre de lo4 (une unit5 8). En respectant cette contrainte
la mkthode peut Ctre gCn&alisable ?I presque tous les systemes symbiotiques fixateurs.
Valeur isotopique de I’azote assimilb
Mesurer directement la valeur isotopique de l’azote du sol s’avkre difficile dans
le choix de la forme a mesurer (organique ou minkale). On sait en effet que le passage
dune forme ?I une autre provoque un fractionnement isotopique. De plus, la mesure des
diff&-entes formes d’azote du sol a un instant dorm6 ne permet pas d’intigrer la vakur
isotopique de l’azote assimilC par la plante durant tome sa vie. Enfin. un fiactionnement
isotopique pcut parfois se prod&e au moment de l’assimilation et modi!kr l’abondance
isotopique de l’azote provenant du sol (Mkiotti et al., 1980). Aussi comme darts la methode
de marquage du sol de Fried et Middleboe (19771, on apprehende la valeur isotopique
de l’azote assimild par le biais dune plante non furattice poussant dans lcs m&nes conditions
que l’on appelle plante de rkf&ence ou plante t&mom. Lorsque ies &&es sont faites sur
des systkmes annuels, cette plante timoin sera choisie en fonction de l’enracinement et
de la p&iode de croissance comparables aux plantes fixatrices &udi&s. Dans les kcosystkmes
naturels on chosira les planks temoins ayant un developpement equivalent a celui de
la plante fixatrice. Toutefois, le probEme de la plante temoin est moins crucial dans cette
m&ode que dans celle du marquage du sol (Rermie, 1982), car ii n’existe pas d’effets
lies au dkclin prononce de l’emichissement du sol avec le temps, aucun engrais marquC
n’ayant ktc5 rajoute. L’Cvolution de la valeur isotopique naturelle de l’azote dun sol est
lente car elle resulte de l’ensemble des processus lids a la formation de ce sol et la valeur
isotopique mesurke de l’azote assimile est peu differente dune arm& sur l’autre. Dans
les articles (Domenach and Corn-tan, 1984 ; Domenach and Kurdali, sous press& on peut
se rendre compte que les mesures faites sur tme aulnaie montrent la grande stabiliti de
la valeur isotopique de l’azote assimile et que le m5me champ de culture en donne la
mCme valeur sur deux ans d’exp&iences.
Cheng er al. (1964) et Karamanos and Rennie (1980) ont mis en evidence une
variation isotopique de l’azote en fonction de la profondeur du sol, aussi il est important
que la plante non fixatrice ait un systeme racinaire exploitant la m&me zone de sol. Ceci
est surtout vrai pour les kcosystkmes naturels install6 sur des sols t&s evoluCs qui peuvent
presenter ces variations. En systeme agricole, grace a la pratique des labours, la valeur
isotopique de l’azote assimile s’avere beaucoup plus homogene (Selles et al., 1986). Ainsi,
nous avons pu montrer, que si un soja gCnCtiquement non nodulant representait la plante
de reference la plus satisfaisante pour une culture de soja, le sorgho pouvait egalement
321
remplir ce role (Domenach and Corman, 1984), comme le ray-grass peut Etre pour le
m3le (Bergersen and Turner, 1983) ou la luzerne (Ledgard et al., 1985). Les plantes
fixatricesnepos~entpastoutesuneisolignegCn~quementnonnodulante,aussilapossibili~
d’utihser une graminke cornme plante de rkference permet d’appliquer la mt5thode a ces
dims plantcs f~trices et done globalement permet une grande utilisation de la m&ode.
Ceci donne Cgalement la possibiliti de choisir une plante non fixatrice bien adapt&e a
la region. Dans un systkme naturel, les planks de reference sont imposkes par le milieu.
Elles seront done des arbres si on veut estimer la fixation d’arbres symbiotiques. Ces
arbres non fixateurs peuvent avoir des valeurs relativement homogenes (Fig. 2) ou comme
S1TE 1
SITE 2
SITEJ
SITE 4
. . . . . . . . . . .
._._..__.._.... i
. . . . . . *-***
*“~...*............““““““‘i
1
1
1
I
1 I
Figure 2 : 6”N de feuilh d’A1nu.s incaM et de diffbxxes
plantes non fuauiccs rt?coltis dans quatrc stations d’une
sulnaie naturelle. * correspond P la vrdeur mesurke SUT un
aulne poussant sur milieu ssns aznte 250 jours zip&s
t
inoadation d’aprk Domenach et Ku&Ii, 1989).
322
le montrent Shearer et al. (1983), pour le desert de Sonoran, des valeurs assez h&Crogenes
(Fig. 3). 11 est alors n&ssaire de prelever le maximum de plantes non fixatrices afin
de rendre compte de l’ht%rogCneitk isotopique du milieu et de collecter les kchantillons
par pa&. Chaque paire sera form&z dune plante fixatrice et dune non-fixatrice les plus
proches afin d’optimiser la similaritt5 isotopique de l’azote que chacune de ces plantes
assimile (Shearer and Kohl, 1986).
J?ignre 3 : RQmition du 6-N
des plantes non fixatrices du
dCsert d e Sonomn (d’aprks
Shearer et al, 1983).
i23456789
SUNof leaf tissue
La m&ode bask sur les variations isotopiques naturelles de l’azote nkcessite l’exis-
tence dune difference entre la valeur isotopique de la plante ttmoin et celle de l’azote
fix& La precision depend directement de cette difference. Cette exigence marque une
limite dans l’application de la mkthode. La valeur de cette difference est fonction du sol
et kchappe a notre cornrole. Aussi, suivant les conditions du milieu et sans que nous
puissions le savoir a l’avance, le taux de fixation ne pourra pas ttre mesure ou au contraire,
aura une valeur tres precise. La prtkision des mesures peut se situer entre 5 et 25%.
Les sols agricoles foumissent en g&kal une valeur isotopique de l’azote assimile largement
plus haute que celle de l’azote fix& Par con&e, en sol de for-Et tempMe, l’azote assimile
aunevaleurisotopique~uventbasseetm~en~gativevenantdel’importancedel’immobilisation
comme le montre le modele de Shearer et al. (1974). Cette valeur est alors peu differente
de celle de l’azote fixe et entraine une estimation plus grossiere de la fmation et parfois
m&me entraine I’impossibilitC d’utiliser cette m&ode.
La valeur isotopique de la plante li &udier est une combinaison des valeurs isotopiques
des dew sources azotkes prkckdemment d&ermin&zs, l’azote assimilt et I’azote fix&
Le probleme se pose alors de savoir comment realiser l’kchantillonnage de ces
plantes. Les rksultats de Reinero ef al. (1983) et de Yoneyama er al. (1986) ont montre
une grande hCt&og&rtW dans les valeurs isotopiques trouvkes pour les differems organes
dun mCme soja au champ. Si on considere uniquement les feuilles de sojas nodules et
323
de sojas non nodules poussant dans les memes conditions (Fig. 4), on remarque une grande
h&krogtnCite IiQ a la position des feuilles (El correspondant aux feuilles les plus basses,
F15, aux feuilles les plus hautes) et li6es a leur stade phenologique (les valeurs isotopiques
des mCmes feuiks sont diff&entes a la formation des gousses et iI la maturation des
graines). &chant que des sojas poussant sur milieu sans azote ont un WN t&s homo-
gene (Mariotti, 1982), les variations observ&es dans la Fig. 4A sont likes ?r l’assimilation
(fractionnement isotopique lors de l’assimilation ou evolution dans le temps du 615N de
l’amte assimile ?). On peut peruser que la complete pertubation des valeurs mew&s a
maturation des graines (Fig. 4B) proviendrait de la remobilisation de l’azote au profit
des graines. Cette remobilisation affecterait plus spkialement l’azote fix& En effet, si
a15N 2 t
1
-%&-
F 2
F 3 4
FS6
Fll 12
-I
Figure 4 : 6”I-J de feuilles de
nivsau dc6 feuiir
sojn (Hodgson)iuocul~s et non
inoculCs en fonction de leur
position sur la plante (Fl MT-
A la maturation des gahes
respond BUX fcuiks lcs plus
basses), B dew stades phtno-
logiques.
B
a 15N
niveau des feuilles
324
on considkre la valeur isotopique de la partie aerienne en&e, la difference per&e entre
les sojas nodulQ et non-nodules alors que les valeurs isotopiques de leurs feuilles devien-
nent semblables. Cette remobilisation preferentielle de l’azote fme a d’ailleurs Cte misc
en Cvidence grace au r5Nz par Warembourg et al. (1984.)
Ces r&&ats montrent que, s’il existe une variation isotopique de l’azote assimilC,
celle-ci n’est pas g&rante darts la mesure oti elle est prise en compte aussi bien par la
plarite fixatrice que par la phmte non-fixatrice. De plus, ces r&&ats mettent en &kience
l’importance de lkhentillonge dans l’application de la mtthode bask sur les abondances
isotopiques naturelles. En effet, pour unmake le taux de fixation d’une plante fixatrice
herback, il est necessaire de p&lever la partie a&ienne en&z de ces planks : le fait
de mesurer exclusivement les graines entraine une surestimation du taux d’azote prove-
nant de la fixation et la mesure des feuilles pourrait entrainer, suivant la p&Me de p&E-
vement, une sous estimation de ce taux.
Pour estimer le taux de fixation d’un arbre, il parait difficile d’analyser tome la
partie aerienne de cet arbre. Les feuilles, production annuelle de ces arbres apparaissent
comme le mat&iel de choix qui peut Cue utilise pour estimer le taux de fixation dune
p&iode vegetative de cet arbre. Ainsi, si on choisit les feuilles, on se refire ZI la production
vCgCtale de km& alors que si on choisit le tmnc ou les tiges, le ph&tom&ne d’accumu-
lation interfere sur le rt5suhat. La valeur isotopique de l’azote mesurke sur le tronc devrait
reprksenter l’historique de la nutrition azot& de l’arbre alors que la vakur isotopique de
I’azote des feuilles reprksente celle de l’anntk Mais dans ces feuilles, une certaine quantiti
d’azote a Ctk remobilisk a partir des reserves situ&s darts les racines (Domenach et Kurdali,
1989) : pour des auks agb de 5-6 ans les r&serves reprksentent encore 10% de l’azote
des feuilles en fin de p&ode vCg&ative (Fig. 5). Cet azote remobilid constitue avec
l’air et le sol, une troisikme source de nutrition dont il faut tenir compte si on veut estimer
le taux d’azote fixC darts ces feuilles. Ce distingo perd de son importance si lXchantil-
lon est constitue des dernikres feuilles form&es, mais on risque ainsi de n’avoir qu’une
Flgure 5 : Pcurcentagc d’nzote provenant
dcs r&s-
ewes des ncines dans les feuilles d’Ahs gluhosa
e t d e PopullLF alha iges de 5 - 6 111s e t poussant e n
conditions naturelIes dumnt tome Ia pCriodc de
croissance. (d’apr2s Domenach et Kurd&i 1989).
--___.
325
vue partielle de ce qui s’est passk dans la saison. Le mEme probkme se pose dans la
mesure de l’activitt fixatrice des l~gumineuses herbacks pkrennes dont les parties akrien-
nes sont composks de 20% de I’azote venant des rkerves apri?s 40 jours de repousse
(Lescure et Chalamet, 1985).
Sur des aulnes en milieu naturel, les feuilles peuvent p&enter une valeur isoto-
pique relativement homog&ne ou au contraire, hCdrogtne comme le montre la Figd. L.es
variations peuvent etre le signe de l’existence de p&iodes de furation et d’assimilation
plus ou mains intenses. 11 est done important, si on veut avoir une estimation globale
de l’azote fix& de prklever, antant que possible, les fenilles snr l’ensemble de la hauteur.
Les arbres constituent un type de materiel biologique malaise? & &udier, mais quelles que
Gent les difficult& d’application de la m&ode bask sur les variations isotopiques na-
tuelles, il convient de noter que c’est la seule utilisable sur les arbres en conditions na-
turelles aujourd’hui.
M&hode de calcul du pourcentage de fixation (%Ndfa)
La base de la mkthode est simple et se rksume B une application du principe de
la dilution isotopique. Le calcul est le mCme pour les plantes herb&s ou ligneuses.
WN pl. non fixatrice - PN pl. fixatrice
%Ndfa =
x 100
ZPN pl. non fiiatrice - PN pl. fixatrice sur milieu sans N
Cette kquation est de la forme : %Ndfa = (X-Y) / (X-C). L’kcart-type peut Ctre
dond par la formule utiliske par Shearer and Kohl (1986).
o--C)
CAY)
6-w
A % N d f a = -
I I A x + - - -
I I AC
(C-xy
I(C-x>l + (C-X)2
Comme il n’est pas mk-ssaire de rajouter une source d’azote comme on le fait
dans la m&ode de marquage du sol, on n’utilise pas le concept de la valeur “A” ce
qui supprime les problkmes Ii& ii cette valeur. Toutefois lorsque l’on introduit des engrais,
la mkthode isotopique des abondances naturelles reste applicable mais peut poser quel-
ques probkmes. En effet, les engrais, &ant fabriqub a partir de l’azote de l’air, ont souvent
un 6”N t&s bas ce qui a pour cons@uence de diminuer la valenr isotopique de l’azote
assimilC. I1 en r&&e une moins bonne prkision dans la mesure de fixation. Cet incon-
vtkient peut Ctre palliC par le controle du rapport isotopique de l’engrais utili& et si n.5
cessaire, la modification de ce rapport par un infme apport de produit marquk.
Comparaison des m&hodes marquage du sol et abondance isotopique na-
turelle
La possibilitk de quantifier la furation par la seule connaissance des variations iso-
topiques naturelles entre une plante fixatrice et une plante non fixatrice est confirm&
par la comparaison des rksultats obtenus avec czux que donne la m&ode de marquage
du sol sur des plantes similaires.
326
Sit* 3
I IEv)E 6
----a
,
Figure 6 : 6”N de fcuilla d’Afnrrs
incnca (A) et de Acer psewiopkz-
L
fanw (B) r6colties au dkpan Ed 1
c
l’extr=SmirC dcs branches d’un
meme arbre sur les bois des
armies pr&ddentes (___ 1 et
I
327
Ces deux m&hodes sont en effet comparables car tomes deux sont indt$endantes
des quantit&s d’azote assimilkes, utilisent une plante de reference, le concept de la dilution
isotopique, et donnent une valeur globale de la fixation. Seul le principe de base change
: dans un marquage, on admet que les microorganismes et les plantes ne penvent pas
faire la difference entre 15N et 14N lorsqu’ils metaboiisent l’azote alors que c’est sur l’existence
de cette reconnaissance des deux isotopes par les microorganismes qu’est baske la m&hode
WN. LAX deux principes ne sont pas contradictoires et I’un ou l’autre s’applique suivant
l’kchelle de precision darts laquelle on se place, les variations isotopiques naturelles n’&ant
sensibles qu’a 1O4% 15N. Cette comparaison de-s deux m&hodes est essentielle pour &ablir
la validit de la mkhode bask sur les variations isotopiques naturelles meme si peu de
publications (Bergersen and Turner, 1983 ; Domenach et Chalamet, 1979; Ledgard et
al., 1985) en font &at.
Le Tableau 1 montre que les deux methodes donnent des rt5sultat.s 6quivalents aussi
bien sur les plantes her-backs que ligneuses justifmt l’emploi des variations isotopiques
naturelles pour mesurer la futation symbiotique de l’azote.
Si l'utilisation de la methode WN prksente quelques limites et contraintes que nous
avons soulign&s tout au long de cet article, elle offre egalement beaucoup de possibilit6s.
Comme la mCtkxie de marquage du sol, elle fait partie des m&odes globales intQrant
Tableau 1 - Comparaison de deux mGthodes d'estimation
du taux de fixation symbiotique : marquage du sol et
variations isotopiques naturelles de l'azote.
% FIXATION
TRAITEMENTS
plantes
marque
non marqw?
NH4+
Soja
6 0
68
5 6
60
Trefle
42
42
N03-
Soja
7 2
69
Aulne
98
100
N-organique
Trefle
45
4 7
Domenach et Chalamet 1979, Domenach and Kurdali 1989.
328
la fixation de mute la vie de la plante et permet ainsi d’avoir une estimation chiffrk
de la contribution de la fixation dans un systkme donne. Elle est Cgalement indepen-
dame de la production et n’utilise pas l’hypothke d’egalid des quantitks d’azote assimile
entre une plante fixatrice et une plante non fixattice. Mais conuairement a la m&hode
de Fried and Middleboe (1977), cette m&ode n’utilise pas de produits marqds. Cette
diffi&nce irnplique de nombrenx avantages :
(i)
il est possible de realiser des mesures sur l’ensemble du systkme 6tudiC
et non plus de se contenter de microparoelles comme on le fait avec
le marqnage, ce qui donne une meilleure image de la fixation rklle,
dun champ par exemple ;
(ii)
tous les problemes lies It l’hCt&ogedid du marquage due a la difti-
cult6 dune application homogene de l’engrais disparaissent ;
(iii) le choix de la plante temoin est plus large car le rapport isotopique
de l’azote du sol est homogene et ne varie que t&s lentement au tours
du temps, diffkemment du d&Am t&s rapide du %15N dun sol apr?s
un marqnage ;
(iv) enfin et surtout, elle n’implique aucune intervention de l’expkrimen-
tateur et peut done &e appliquke sur les kosystemes naturels saris
les perturber ni les modifier. Cette m&ode apparait ainsi la mieux
adapt& a la mesure d’activite fixatrice des arbres sur qui par le fait
mCme de leur taille, la m&ode a l’adtylene ou les mtthodes de mar-
quage semblent delicates a apphquer.
La grande difficult6 tenant a la methode vient des valeurs isotopiques de l’azote
assimil6 souvent basses en fotit dont la consequence est de n’avoir qu’une estimation
grossikre de la fixation.
Champs d’application
Si on peut disposer dun spectrometre de masse, l’emploi de cette m&rode in situ
est t&s aisk puisque la seule intervention de l’expkimentateur consiste a p&lever des
6chantillons au moment opportun. Son champ d’application est t&s vaste et quelques exemples
peuvent rendre compte des nombreuses possibilit6s qu’elle offre :
0 Cette m&ode pourrait permettre, a la mar&e de Virginia and Delwiche
(1982), de r&lker des criblages de plantes peu connues darts des milieux
vierges. En effet, ces auteurs, en r&disant des mesures isotopiques sys-
t&natiques sur les planks prksumkes non fixatrices, ont pu mettre en
evidence qu’une rosa&, Chamaebatia foliolosa, etait en fait fixatrice,
ce qui, par la suite, a pu Cue confirm&
8 Estimation des entr6e.s d’azote dans un systeme agronomique ou natu-
rel. Des estimations de taux de fixation de plusieurs legumineuses cul-
tivkes dans differentes regions de France (Tableau 2) montrent la plus
ou moins bonne adaptation du pouvoir fixateur de ces plantes au climat,
ce qui ne signifie pas que les dites plantes n’aient pas un bon rendement
m&me si leur taux de fixation est bas.
329
Tableau 2 - Estimation des rentrees d'azote dans differents systPmes
par le biais de la fixation symbiotique.
% d'azote fix&
Stations
Aulne*
Soja
Luzerne
TrPfle
Ornon
9 8
Montpellier
40-75
Dijon
10-20
58
56
* Apport de 1itiPre : 700kg N ha - 1
(d'aprPs Bardin et al 1977, Domenach and Corman.1984, Domenach and
Kurdali 1988).
On pourrait ainsi pr&ser, pour les differentes esp?ces de plantes, les
zones geographiques oti se situe I’optimum de leur capacite fixatrice.
Un exemple de la tres grande adaptation au milieu et au &mat est dorm6
par des aulnes, arbres a actinorhizes dont la repartition des espkes depend
naturellement des conditions de milieux. Par exemple en France, Alnus
incanu est localise en montage et Alms gbtimsu, au bard des rivieres.
Les mesures faites sur ces arbres croissant dans leur milieu naturel, soit
pour A. incanu, poussant en altitude sur un support constitd par des
moraines et pratiquement dQourvu d’azote, soit pour A. ghtinosu poussant
en plaine sur un sol riche en azote, montrent la t&s grande capaciti
de ces arbres a fner l’azote de fair (Tableau 3). On comprend mieux
ainsi le role pionnier que ces arbres fixateurs peuvent avoir dans la consti-
tution dun sol et l’inttk~t de boisements mixtes permettant le develop-
pement d’essences plus nobles.
8 Optimisation des entrees d’azote par la fixation symbiotique d’azote.
0)
par la meilleure connaissance de l’impact des facteurs de l’en-
vironnement sur cette activittk Un exemple est don& par Etude
de l’influence du stress hydrique sur la fixation du soja (Obaton
et al., 1982 ; Domenach and Corman, 1985). Cette estimation
est faite sur des cultures en champ ayant support6 des stress
hydriques durant l’optimum de leur cycle nitrate-r&luctase (stress
AN& arret de l’irrigation de mi-juin a mi-juillet) ou nitrogenase
(stress A.R.A., a&t de l’irrigation au mois d’aout) ou les deux
(stress A.N.R. + A.R.A., arret de l’irrigation de la mi-juin a
.
...*
330
Tableau 3 - Estimation du taux d'azote de fixation de deux espkes
d'aulnes dans leur milieu naturel, Alnus glutinosa (arbre de plaines,
de bard de riviere), Alnus incana (arbre de montagne, poussant sur
des moraines),
Espkes
PH
% C organique % N total % CaC03
% N fix@
A glutinosa
7.1
2.7
0.33
0
95-tl3
A incana
8.3
0.5
0.06
6 4
98+20
la fm aofit). Les mesures obtenues a partir de la m&ode des
abondances isotopiques naturelles montrent la grande sensibilite
de la fixation au manque d’eau alors que l’assimilation est moins
touch&z par ces stress (Pig.7, Tableau 4).
w par la recherche du couple symbiotique le plus efficace et le
mieux adapt6 aux conditions de milieu S&ction de la souche
microbienne : s’il est aid de tester des souches au laboratoire,
il est difficile de prejuger de leur capacitk en conditions naturelles
de culture. Pour connditre les modifications que ces souches
peuvent apporter dans la nutrition azotie de leur plante hbte,
la mesure de fixation s’impose (Tableau 5). La m&rode M’ac&ylene
per-met de classer les souches en fonction de leur activite
nitrog6nasique mais ne peut pas donner une estimation quan-
titative de l’azote fm6 et sumstime le pouvoir fixateur des souches
comme la G49 qui pr&ente une faible capacite hydrogenasique.
(iii) s&ction de la plante h6te : par la connaissance du rendement
et des capacitks futatrices de differems cultivars de sojas (Tableau
6), ou d’aulnes de difft?rentes origines (Fig.8), il est possible
de choisir, pour un sol donne ou une region don&e, la plante
h&e qui offrira le meilleur compromis qui tient compte a la
fois du rendement et des capacitks de fixation de I’azote.
331
Figure 7 : AcfivitCs nitrate rC-
ductase (A.N.R) u rcductrice dc
l’achyline (AKA) chcz le soja
(HOdgson) au cows du cycle
vCgitatif B 1’Cvapotranspiration
maximale @T&I) (d’aprks Obaton
cf AI., 1982).
100 iours spr;sscmis
V
V R
1
16 8 stadts phenol.
Tableau 4 - Influence du stress hydrique provoque lot-s de
l'optimum de l'activite nitrogenasique (stress A.R.A.) et de
l'activite de la nitrate-reductase (stress A.N.R.) sur la
fixation et l'assimilation de l'azote.
Conditions
% N fix@
Quantites assimilees
hydriques
(mg.pl-'1
M.E.T.
41
750
Stress A.N.R.
2 5
540
Stress A.R.A.
1 6
491
Stress A.N.R.+ A.R.A.
0
410
M.E.T. : Maximum d'evapo-transpiration. (Domenach and Corman 1985)
332
Tableau 5 - Soja (hodgson) poussant sur sol et inocules avec
differentes souches de Bradyrhizobium japonicum.
Souches
%Nfixe QNfixe
ARA.
i5N nodosites
mg/plante
nm./pl./h.
Gl
2 5
149
202
2.5
GA
3 6
293
1105
5.2
GMBl
81
1140
1614
8.7
649
6 9
1736
2374
9
G3S
87
1948
1777
9
GPsp
7 7
1990
2317
9.8
Avec la collaboration de J.C. Cleyet-Marrel, INRA, Montpellier,
laboratoire des Symbiotes des Racines.
Tableau 6 - % d'azote fixP par differents cultivars de sojas
inocules avec la m&w souche (63 Sp.)
I‘ 11 1 I he
quaIli 6 ir II i*t
E.T.M.
par plante (mg)
-tardives
106-17
93
1120
Amsoy
7 5
857
Al00
7 0
831
-precoces
Hodgson
5 7
579
Dobrudza
70
258
20-18-82
5 9
230
Avec la collaboration de M. Obaton, INRA, Montpellier,
Laboratoire des symbiotes des racines.
333
80-
60-
Figure 8 : Pounrntagc d’amtc fuC dam
les feuilles d’Alw g1dm.w de di6mtes
2
origines r&colt&es dans une plantation
s
d’arbres igts de 5 am.
* 40 .
20 -
Conclusion
Quoique spectaculaires, ces rbultats sont obtenus pour la plus grande part, B partir
d’essais rkalisCs en pays temphk Pour connaitre toutes les potentialit& de la n&&ode
d’estimation de la fixation symbiotique d’azote bask sur les variations isotopiques na-
tureks, il serait nkcessaire de la tester sous d’autres types de chats (tropical, semi-
dbertique...). Egalement, peu de travaux utilisant cette mkthode ont &15 r&tlis& sur les
tkosyst&mes naturels et il est difficile de se rendre compte des limites que le milieu peut
imposer (en particulier sur la valeur isotopique de l’azote assimilC 3 partir du sol).
Si les travaux qui seront effectuh dans le futur conferment et complhent les larges
possibilitks de cette mkthode, celle-ci pourrait pcrmettre une meilleure connaissance des
kosysthnes naturels, et Ctre particulikrement utiliske en foresterie, domaine dans lequel
les mCthodes classiques ne peuvent ou que difficilement Ctre appliqukes.
Remerciements
L,e CEA de Syrie a octroyk une bourse d’&nde B Monsieur Fawaz Kurd&.
334
Wfbences
AMARGER, N., MARIOTTI, A. et MARIOTIT, F. 1977 Essai d’estimation du taux
d’azote fix& symbiotiquement chez le lupin par le tracage isotopique nature1 (%).
CR. Acad Sci. Paris. 284, 2179-2182.
BARDIN. R. DOMENACH A.M. et CHALAMET A. 1977 Rapports isotopiques naturels
de l’azote. II - Application a la mesure de la fixation symbiotique de l’azote in situ.
Rev. Ecol. Biol. Sol. 14, 395402.
BERGERSEN, FJ. and TURNER, G.L. 1983 An evaluation of lsN methods for estimating
nitrogen fixation in a subterranean clover-perennial ryegrass sward. Aust. J. Agric.
Res. 34, 391-401.
BREMNER, J.M. 1977 Use of nitrogen tracer techniques for research on nitrogen fi-
xation. In : Biological nitrogen fmation in farming systems of tropics. Ayanaba, A.
and Dart, P.J. Eds. Pub., John Wiley and Sons, pp 335-352.
CHENG, H.H. BREMNER, J.M. and EDWARDS, A.P. 1964 Variations of is-N abun-
dance in soils. Science. 146, 1574-1575.
DELWICHE, CC. and STEYN, P.L. 1970 Nitrogen isotope fractionation in soils and
microbial reactions. Environ. Sci. Technol. 4, 929-935.
DOMENACH, A.M. 1985 Comparaison des abondances isotopiques naturelles de l’azote
chez deux legumineuses, l’une ligneuse, Robiniapsetuioacacia, l’autre hexlmct5e, Glycine
max. Acta Oecologica, Oecol. Applic. 6. 23-29.
DOMENACH, A.M. CHALAMET, A. et PACHIAUDI, C. 1979 Estimation de la fmation
d’azote par le soja a l’aide de deux m&odes isotopiques. CR. Acad Sci. Paris 289,
29 l-293.
DOMENACH, A.M. and CORMAN, A. 1984 Dinitrogen fixation by field grown soy-
beans : statistical analysis of variations in 15N and proposed sampling pmedure. Plant
Soil. 78, 301-313.
DOMENACH, A.M. and CORMAN, A. 1985 Use of “N natural abundance method for
LIIC: btuuJ UI by~ti~~uuz mauon 01 rrckr-grown soyneans. Iniluence 01 tixauon on assi-
miiation and effect of water conditions on these two functions. Soil Sci. Plant NW.
31, 311-321.
DOMENACH, A.M. et KURDALI, F. 1989 Influence des reserves azo&s sur la for-
mation des feuilles &Alms glutinosa (L) Gaerm. et ses consequences dans l’estimation
de la fixation d’azote. Can. J. Bot. sous presse.
DOMENACH, A.M. and KURDALI, F. 1989 Estimation of symbiotic dinitrogen fixation
in alder forest by the method based on natural 15N abundance. 7* IntemationaI Meeting
on “Frankia and Actionorhizal plants” Stmrs, Conr~~ticut. Plant Soil, sous presse.
DOMENACH, A.M. KURDALI, F. DANNIERE, C. et BARDIN, R. 1988 Determination
de I’identite isotopique de l’azote fix6 par Frank51 associe au genre Alms. Can. J.
Bot. 66, 1241-1247.
EDWARDS, A.P. 1975 Isotope effects in relation to the interpretation of 15Np4N ratios
in tracer studies. In : Symp. on isotope ratios as pollutant source and behaviour in-
dicators. IAEA SM. - 191/32.
FRIED, M. and MIDDLEBOE, V. 1977 Measurement of amount of nitrogen fixed by
a legume crop. Plant Soil. 47, 713-715.
335
HAUCK, R.D. BARTHOLEMEW, V.W. BREMNER,J.M. BROADBENT, F.E. CHENG,
H.M. EDWARDS, AS. KEENEY, D.R. LEGG, J.O. OLSEN, SX. and PWORTER
L.K. 1972 Use of variations in the natural nitrogen abundance for environmental stu-
dies : a questionable approach. Science. 177, 453456.
HOERING, T. 1975 Variations of i5N abundance in naturally occuring substances. Science.
122, 1233-1234.
KARAMANOS, R.E. and RENNIE, D. A. 1980 Changes in natural i5N abundance asso-
ciated with pedogenic processes in soil. II Changes on different slope positions. Can.
J. Soil Sci. 60, 365-372.
KOHL, D.H.. SHEARER, G.B. and COMMONER, B. 1971 Fertilizer nitrogen conhiiution
to nitrate in surface water in a corn belt watershed. Science. 174, 1331-1334.
KOHL, D.H. SHEARER, G. and HARPER, J.E. 1980 Estimates of N, fixation based
on differences in the natural abundance of 15N in nodulating and normodulating iso-
lines of soybeans. Plant Physiol. 66, 61-65.
LEDGARD, SF. SIMPSON, J.R. FREENEY, J.R. BERGERSEN, F.J. and MORTON,
R. 1985 Assessment of the relative uptake of added and indigenous soil nitrogen
by nodulated legumes and reference plants in the i5N dilution measurement of N,
fixation : glasshouse application of the method. Soil Biol. Biochem. 17, 323-328.
LESCURE, C. et CHALAMET, A. 1985 Etude de l’influence des nkerves azot&e.s dans
l’estimation par 15N de la fEation symbiotique dune plante pkenne, Trrfolium prutense.
Can J. Bot. 63, 991-994.
MARIOTI’I, A. 1982 Apports de la gkochimie isotopique a la connaissance du cycle
de l’azote. These d&at, Universitk Paris VI, 476~.
MARIOTTI, A. 1984 Natural r5N abundance measurements and atmospheric nitrogen stan-
dar calibration. Nature. 311, 251-252.
MARIO-F-I-I, A. MARIOTTI, F. AMARGER, N. PIZELLE, G. NGAMBI, J.M. CHAM-
PIGNY, M.L. and MOYSE, A. 1980 Fractionnements isotopiques de l’azote lors des
processus d’absorption des nitrates et de la fmation de l’azote atmospherique par les
plantes. Physiol. Veg. 18, 163-181.
OBATON, M. MIQUEL, M. ROBIN, P. CONEJERO, G. DOMENACH, AM. et BARDIN,
R. 1982 Influence du deficit hydrique sur l’activite nitrate rkluctase et nitrogthase
chez le soja (Glycine max L. Merr. cv. Hodgson). CJI. Acad. SC. Paris 294, 1007-
1012.
RElNERO, A., SHEARER, G., BRYAN, B.A., SKEETERS, J.L. and KOHL, D.H. 1983
Site of namral 14N enrichment of soybeans nodules. Piant Physiol. 72, 256-258.
RENNIE, RJ. 1982 Quantifying dinitrogen (N,> fixation in soybeans by i5N isotope dilu-
tion : the question of the non fixing control plant. Can. J. Bot. 60, 856861.
RENNIE. D.A., PAUL, E.A. and JONES, LE. 1976 Natural nitrogen-15 abundance in
soil and plant samples. Can. J. Siol Sci. 56, 43-50.
SCHOENHEIMER, R. and RlTlENBERG, D. 1939 Studies in protein metabolism : I-
general considerations in the application of isotopes to the study of protein meta-
bolism. The natural abundance of nitrogen isotopes in aminoacids. J. Biol. Chem.
127, 285-290.
336
SELLES, F., KARAMANOS, R.E. and KACHANOSKI, R.G. 1986 The spatial varia-
bility of nitrogen-15 and its relation to the variability of other soil properties. Soil
Sci. Sot. Am. J. 50, 105110.
SHEARER, G., DUF’FY, J., KOHL, D.H. COMMONER, B. 1974 A steady-state model
of isotopic fractionation accompanying nitrogen transformations in soil. Soil Sci. Sot.
Amer. Proc. 38, 315-321.
SHEARER, G., KOHL, D.H. and HARPER, JE. 1980 Distribution of r5N among plant
parts of nodulating and nonnodulating isolines of soybeans. Plant Physiol. 66. 57-
60.
SHEARER, G. and KOHL, D.H. 1986 N,-fixation in field settings : estimations based
on natural “N abundance. Aust. J. Plant Physiol. 13, 699-756.
SHEARER, G., KOHL, D.H., VIRGINIA, R.A., BRYAN B.A., SKEBTBRS, J.L., NJLSEN,
E.T., SHARIFI, M.R. and RUNDEL, P.W. 1983 Estimates of N, fixation from varia-
tion in the natural abundance of lsN in Sonoran desert ecosystems. Oecologia (Berlin)
56, 365-373.
STEELE, K.W., BONISH, P.M., DANIEL, R.M. and OHARA, G.W. 1983 Effect of
rhizobial and host plant on nitrogen isotopic fractionation in legumes. Plant Physiol.
72, 1001-1004.
VIRGINIA, RA. and DBLWICHE, C.C. 1982 Natural 15N abundance of presumed N2-
fling and non-fixing plants from selected ecosystems. OecoZogiu (Berlin) 54, 317-
325.
WAREMBOURG, FR., HOEGEL, B., FERNANDEZ, M.P.et MONTANGE, D. 1984 Dis-
tribution et utilisation des assimilats carbon& en relation avec la fixation symbio-
tique de I’azotc chez le soja (Glycine max. L. Mcrr.). Plant Soil.082, 163-178.
YONEYAMA, T., YAMADA, N., KOJIMA, H. and YAZAKI, J. 1984 Variation of natural
lsN abundance in leguminous plants and nodule fi-actions. Pkznt Cell. Physiol. 25,
1561-1565.
YONEYAMA, T., FUJLTA, K., YOSHIDA, T., MATSUMOTO, T., KAMBAYASHI, I.
and YAZAKI, J. 1986 Variation in natural abundance of “N among plant parts and
in “Np’N fractioation during N, fmation in the legume-Rhizobia symbiotic system.
Plant Cell. Physiol. 25 no 8 : 1561-1565.
Maximisez la FBA pow la Production Agricole et Forestihe era Afrique
Nitrogen fixation in tropical trees :
Estimations based on 15N techniques
SANGINGA, N.(l), ZAPATA, (l) and DANSO, S.K.A@)
(I) : FAOIIAEA Programme, IAEA Seibersdorf Laboratory, A-2444 Seibersdor$ Autria ;
(2) : Joint FAOIIAEA Division, IAEA Wapmerstrarse 5, A-1400 Vii, Austria.
Abstract
Because of the importance of trees in agroforestry. there is a crucial need to
assess the magnitude of N2 fixed by them, directly in the field The long-term evalu-
ation of the N,-fixing potential and actual amounts of N, fixed in a given tree raise
“rr\\h,,~mr :_ q”v,.-.?:l, n,.,...ti ^r_.. .: .,I. 1, .-. .I..-“.:-_. .r .- 1
I
.ij -...-.. .,.. .,_ D’
I.. .I .-7 ‘.,-P 7. ;,,
. . . . -. ‘,
obtaining reference crops over several seasons. For young or small trees,
‘sN proce-
dures similar to those adopted in gram or pastnre legumes would be expected to
give equally satisfactory results. The greatest problems with
N2 fixation estimations
using “N will occur in mature trees, due to their perennial nature and massive s&s.
leading to logistic and sampling difficulties or differences in tsN/?N ratio of soil
due to diierences in nitrogen turnover processes that occur under the fixing and
reference crops with time. The influence of these effects may differ depending on
how the 15N is applied. An examination of which of the existing ‘sN procedures,
e.g. isotope dilution, A-value or the natural “N abundance and what method of ‘?J
application should be adopted under different situations is therefore urgently needed.
In the paper we have considered the strategies for obtaining representative
samphng
(as against the whole destructive plant sampling) that would closely reflect the overall
r5N enrichment in the whole plant, and the selection of the appropriate reference
tree. The advantage in using the
lSN labelled plant samples for additional agronomic
studies besides N, fixation, e.g. contribution of tree litter to plant N uptake and
soil organic matter are briefly discussed.
338
Introduction
In recent years the potential of leguminous and actinorhizal trees has aroused con-
siderable interest., especially as regards to their socioeconomic importance and increased
use as firewood, paper pulp, forage and timber (NAS, 1979 ; Dommergues, 1982 ; Torrey,
1982). These nitrogen fming trees (NFTs) contribute to restoration and/or maintenance
of soil fertility (Kang, 1985) and are very important in sound tropical soil management
for sustained agricultural productivity, and controlling soil erosion and degradation.
Until recently however, few studies have been conducted on the examination of
root nodules (Allen and Allen, 1981) and the magnitude of atmospheric N, fixed by dif-
ferent Rhizobium or Fmnkia strains, key factors in the successful growth of NFTs in
nitrogen deficient soils. To maximize the contribution of this natural and inexpensive
source of N in agroforestry systems, reliable methods are needed for quantifying biologi-
cally N, fixed in trees.
Various methods, such as the total nitrogen difference (Comet et al., 1985 ; Gauthier
et al., 1985 ; Sanginga et al., 1985,1986 ; Ndoye and Dreyfus, 1988), acetylene reduction
assay (Roskoski, 1981; Hogberg and Kvarnstrom, 1982 ; Lulandala and Hall, 1988) nodule
number and mass (Roskoski, 1981 ; Hogberg and Kvamstrom, 1982) and ureides pro-
duction (Kessel et al., 1987 ; Sanginga et al., unpublished data) have been used to assess
N, fixed by NFTs. Most of these techniques, analyzed in several critical reviews (Fried
et al., 1983 ; Rennie and Rennie, 1983 ; Danso, 1985, 1986 ; Dommergues, 1987),
are based on indirect criteria, and are either qualitative or cannot distinguish between
the various sources of nitrogen in a faring plant.
Nitrogen-15 techniques, that involve the application of “N labelled materials to
soil have been suggested in many studies to be the most reliable method for quantifying
N,-fmed (Legg and Sloger, 1975 ; Rennie, 1982 ; Vose et al., 1982 ; Rennie and Kemp,
1983 ; Hardarson et al., 1984 ; Danso, 1985 ; Danso et al., 1986). These 15N techniques
are particularly useful as unlike many others, they provide an integrated measurement
of the atmospheric N, fixed by field-grown crops during the experimental period and
can discriminate among the various N sources available to the crop.
T-1.- 15%. 1 . .
I . -1.
_
.
.,.. 1 . ...-1 U..i .tiLA...:.; ~~L~~~L~~u~.J +iiL III>, Y3LU L” L>LIIII&LL lh2 ilXc;cl III
pasture and gram legumes. They have however not been used much in tree N, fixation
studies. Experience gamed from grain and pasture legumes would therefore be valuable
for extending this technique to trees.
This paper examines critically the advantages and difficulties reported for the various
techniques involving the application of lW-labelled fertilizers to soil, and the reliability
of the method based on the differences in the natural abundance of lsN between soil and
atmospheric N, to measure N, fixed and how these could be used in devising suitable
procedures for tree N2 fixation assessment Some estimates of N, fixed in various NFTs
under greenhouse and field conditions using these methods will be also reported.
Basis of the lSN methods in NFT studies
There are two main approaches of the r5N techniques for assessing N,-fixation :
the natural abundance of 15N and the W-labelled fertilizers. The “N natural abundance
is based on the small differences in 15N abundance which frequently occur in nature between
atmospheric N, and soil N, in contrast to r5N substratum labelling techniques where the
339
larger differences in 15N between the soil N and atmospheric N, are induced by applying
15N labelled materials.
Both approaches are however in essence, based on the same principle, i.e. the
higher 15Np4N ratio absorbed from soil N by the N,-fixing plant decreases during fma-
tion by incorporation of the lower 15N enrichment of atmospheric N, ; the more N, that
is fixed, the greater this dilution. Since it is not possible to assess the integrated
i5Np4N ratio of the soil N over a growing season by chemical extraction a non-fixing
reference plant is used for this purpose.
Several articles have reviewed the concepts, equations and procedures used in ap-
plying the lsN methods for measuring nitrogen fixation of grain and pasture legumes (Shearer
et al., 1978 ; Knowles, 1981 ; Fried et al., 1983 ; Rennie and Rennie, 1983 ; Danso,
1985, 1986 ; Dommergues, 1987 ; Hardarson et al. 1987). The researcher must be aware
that there are however as yet no generally accepted procedures which can be claimed
to be valid under all conditions. NITS present specific problems which deserve special
attention.
Factors affecting the validity and precision of N, fixation estimation by trees
using 15N-labelled materials
Reference tree
The accurate determination of the actual amounts of N, fixed in the field is crucial
only in some instances, such as in N-balance studies in agro-ecological systems i.e. agrofo-
restry, or in comparing N, fixed in different seasons, years and environments. In these
cases the reference crop constitutes the main potential source of error in the 15N tech-
nique. The criteria used in the selection of suitable reference crops have been discussed
by Fried et aL, (1983) and Danso (1985, 1986). However, according to Danso (1986)
and Danso et al. (1986) the need for the precise quantification of N, fixed may not be
that compelling in studies to simply compare treatment effects or rank plant genotypes
or RhizobiumlFrankia strains for N, fixing abilities.
Uninoculated N,-fixing legume or actinorhizal trees have often been used as ref-
erence crops (Table 1). Since no indigenous rhizobia or Frunkia were present in the experi-
mental soil the uninoculated nitrogen fixing trees were found to be suitable reference
crops (Comet et al., 1985 ; Gauthier et al., 1985
; Sanginga et al., 1986 ; Sougoufara
et al., 1988). However, care must be taken to thoroughly examine roots of such unino-
culated trees to ensure that they are not nodulated or run an acetylene reduction test
to confirm the absence of nitrogenase activity. Cross-contamination of uninoculated control
from inoculated tmatments has heen observed in our pot experiment involving NITS (unpu-
blished data) and has also been reported by others (Sougoufara et al., 1988). Using such
controls, N2 fixation measured by the isotope dilution and the difference methods were
underestimated, (Sanginga et al., unpublished data). It may be difficult to avoid cross-
contamination in the field and even in the greenhouse unless special precautions are taken
(Bromfield and Ayanaba, 1981). If possible, uninoculated control plants should be com-
pared with other potential reference crops such as known non nitrogen-fixing trees. The
non-fixing trees, however, have to fulfill the conditions outlined by Fried et al. (1983)
and Danso (1986) to be an appropriate reference crop. The validity of such selections
can be established by comparing the isotopic composition of nitrogen with that of non-
nodulating isolines (Fried et al., 1983). Since non-nodulating tree genotypes have not
340
yet been identified, further search for non-nodulating fixing trees in natural populations
or their development (e.g., through mutation induction) are most valuable for the rsN metbo-
dology and attempts have commenced. A simple procedure can be planting of several
hundreds of Camarina and Allocasuarina photosynthetic branches in a sandbath irrigated
with heavy suspensions of Frunkiu. After a period of growth, the roots are examined
for the absence of nodules. The identified potential non-nodulating isolines can be vegeta-
tively propagated and again re-inoculated with Frunkiu to ‘ensure that this trait is stable.
Screening fning trees for host and strain specificity could also provide suitable controls
in some situations.
Table 1 : Reference plants for estimating N2 fixation using 15N techniques
Authors
Type of
Fixing
Reference
experiment
plant
plant
GAUTHIER et al.
-
-
Field
Casuarina
Casuarina
(19851
=-F-i?-
v
CORNET et al.
Field
Acacia
Acacia
(19851 - -
holosericea
lX?%%icea
(II
(Ul
SANGINGA (19851
Field
Leucaena
Leucaena
leucocephala
leucocephala
ZAHARAH et al.
-
-
Field
Leucaena
Setaria ancepts
(19861
'FT?-
(U)
NDOYE and
Pot
Sesbania
Sesbania
DREYFUS (19881
lm3
rostrata (11
rostrata (UI
5
.
n
(I)
-
-
S.n (U1
-
-
SOUGOUFARA et al.
-
-
Pot
C. equisetifolia
C. equisetifolia
(19881
(15,50,1000 11
--Tr----
-0
SANGINGA et al.
Pot
C. equisetifolia
Allocasuarina
IunpublisriS;d~
(5 kg)
c. cunmnghamiana
Allocasuarina
SANGINGA et al.
Pot
Leucaena,
(1987. 193X3-
(5 kg)
(;llncldia
Mmea
-
-
Acacia albida
I: Inoculated
U: Uninoculated
341
Genetic variation in 15N fixation
Large plant-to-plant variation in nodulation and N, fixation have been reported
for NFIs (Gauthier el al., 1985 ; Sanginga et al., 1988b; Sougoufara et al., 1987). The
genetic heterogeneity of NFTs has hardly received the attention it deserves. Danso er
al. (in preparation) indicated that this genetic variation may contribute to the poor preci-
sion in the estimates of N, fixation using the 15N methodology. Duhoux and Dommergues
(1985) observed that average percentage N, fixed based on tree sire, varied from 16 (in
the smallest trees) to 70 (in the biggest-sized trees). In such cases, it is essential to select
well-matched reference crops for each plant to lessen errors. Where plants can be grown
by vegetative propagation, this variation could certainly be reduced, and this could be
practical with many NJ!Ts.
Labelling techniques
There are several practical questions on the techniques of applying the labelled
15N materials which need to be addressed before designing “N experiments to measure
N, fixed. These include the N rate of application and lsN enrichment, chemical and phy-
sical form and time of application. There is evidence that labelling techniques can have
a significant effect on estimates of N, fixed. These techniques, described in several re-
views (Chalk, 1985 ; Danso. 1985, 1986), have mainly been used for grain and pastures
legumes. The limited data available for NITS indicate that much more work still needs
to be carried out to develop adequate techniques for measuring N, fixation by NFTs.
(i) Type and rate of “N labelled fertilizer
The chemical form of the labelled fertilizer is important because of its influence
on N, fixation. This is due to the difference in solubility, uptake by the reference and
fming crops, relative movement and biological transformations in the soil. The isN fertilizers
used so far for NFfs have been in the form of NH,, i.e. ammonium sulphate (Comet
et al., 1985 ; Gauthier et al., 1985 ; Zaharah et al. 1986 ; Ndoye and Dreyfus, 1988 ;
Sanginga et al., 1988a) although no specific reasons can yet be adduced for this preference.
Lower cost of the ammonium sulphate, less interference with fmations and lower potential
losses of N may be one of the reasons. Only recently, Sougoufara et al. (unpublished
data) used urea but there is a high risk for significant N losses.
The amount of “N applied for labelling a tree will depend upon the N rate and
the ‘W enrichment of the labelled material utilized. The N rates and i5N enrichment used
for grain and forage legumes can be readily adopted for small tree plants grown in the
greenhouse or in field conditions.
If isotope-aided experiments are performed with large-sized trees, the amount of
N already present in tree may constitute an extra-dilution factor of the 15N applied. Under
such conditions, it is highly advisable to conduct a preliminary experiment with few trees,
to ascermin several questions, e.g. total N and its partitioning among tree organs, labelling
techniques, as well as sampling procedures.
Single application rates of 2g N/m2, or 0.5 g N/tree enriched with 10 atom %
N have been used to NETS (Comet et al., 1985 ; Gauthier et al., 1985 ; Ndoye and
Dreyfus, 1987). Sanginga et al. (1988) in a single application used twice this rate. Appli-
cation rates will however vary for different NITS, type of studies, locations and soil charac-
teristics, and 15N measuring equipment. As a general guide, the 15N rate must be enough
not to significantly interfere with N, fixation in NFTs (Danso, 1986), and also capable
342
of being detected within the sensitivity range of the isN measuring equipment. For an
emission spectrometer assay, a higher 15N level is needed for detection than by a mass
spectrometer.
In most studies, the isotope dilution method, which involves the application of
the some amount of N-15 labelled material to both the reference and fixing NETS has
been adopted (Table 2). Comet et al. (1985) and Gamhier er al. (1985) compared the
“A value” with the isotope dilution method for estimating N, fixed by Acacia holosericeu
and Cusuarinu equisetifoliu grown in a N-deficient sandy soil. To prevent N-deficiency
in the reference crop so as to ensure comparable satisfactory growth as the N,-fixing
plant, a higher rate of N. 100 kg N/ha was added to the plots of the non-fixing reference
crops, while a lower rate, 20 kg N/ha that would not significantly inhibit N, fixation
was applied to the fixing crops. These authors found that the “A value” technique gave
slightly lower estimates of N, fmation than the isotope dilution and the difference method.
Altbough the “A value” expression is not dependent on fertilizer N addition rates and
fertilizer recovery by both fixing and non-fixing plants, errors can be introduced with
factors affecting differentially the applied fertilizer, e.g. losses due to leaching, deninifi-
cation, N turnover, etc.
Table 2 : Labelling techniques for N2 fixation studies using 15R methods
Authors
Type of
Form
N Addition Atom % 15N Comparison
experiment
rate
excess
other methods
GAUTHIER et al
- -*
Field
(NHqk?S04
;;OUby)
10.5(I.U)
Difference N
11985)
1.91 U ) A-Value
CORNET et al.
(1985) - -
Field
lNH412SO4
2O(I,UI
10 1I.U)
Difference
SANGINGA
Field
lNH412SO4
401I.U)
10 lI,U)
Difference
t 1985)
ZAHARAH et al.
-
-
Field
lNH4)ZS04
2011,R)
10
lI,R)
-
119861
NDOYE and
Pot
lNH4)2SO4
2011,U). 1 0 lI,U)
Difference
DREYFUS
11987)
SOUGOUFARA
Pot
Urea
201I.U)
9.19(I.U) Difference N
et al.
115.50.
100 (U)
1.89 1lJ)
A-Value
7l9aE)
1000 1)
SANGINGA et
Pot
KNOX
2OlI,UI
Difference N
al. 1198r
(5 kg)
100(u)
'! IY'
A-Value
T!%8)
SANGINGA et
Pot
KN03
20 (I)
10 1I.U)
Difference N
~18\\1987~
15 kg)
20,100(uI
2 (u)
A-Value
343
(ii) Methods, frequency and time of “N application
Several workers have reported that the decline in the soil 15NF4N ratio with
time influences the accuracy of Njfixation estimates ; the lower the rate of decline, the
less serious the errors associated with any mismatch between the N uptake by reference
and fixing crops (Fried et al., 1983 ; Witty, 1983). This could be very important in field
experiments involving NITIs because of the long period of growth. In this case a single
initial application of labelled fertilizer may prove to be inadequate for measuring N,-
flation over a long period. Multiple additions of small concentrations of highly labelled
fertilizer to the soil, use of slow-release N fertilizers and ‘SN-labelled organic matter offer
great promise in N,-fixah‘on with NFfs, since the i5NP4N ratio would remain relatively
stable due to the rather slow release of a small but fairly constant amount of N with
time. The greatest problem with these approaches however is that when applied N rates
are lower than needed to support a non-fixing crop over several seasons, the reference
crop may grow poorly or even die.
The lsN fertilizer has been applied to NITS at seeding (Comet et al., 1985) or
soon after germination (Sanginga et al., 1988) or at transplantation of the 3.5 month-
old plants (Gauthier et al., 1985). However, it has been proved by several workers that
early growth of many NITS is slow and that nodulation and N,-fmation is delayed some-
times for several months (Reddell et al., 1986 ; Sanginga et al., 1988a). It is therefore
advisable to delay 15N application until both fixing and reference crops are well-matched
in growth and probably in N uptake.
i5N labelled fertilizer has generally been applied in solution, although a few in-
vestigators (Sanginga et al., 1988a) have broadcasted (NHJ,SO, on field plots and then
incorporated it into the soil with water additions. Several other methods of 15N appli-
cation have been tried for grain legumes (Fried et aI., 1983 ; Chalk, 1985). They have
been reported to give similar results and are thus all satisfactory ways of adding i5N to
soil for N,-fixation studies. Methods of application and their influence on N,-fixation esti-
mate by NITS have not been investigated yet. In some cases the soil injection may provide
a satisfactory approach to apply the “N labelled fertilizer to similar root depths of both
fixing and non-f=ing trees. The choice of one over the other for NITS will be dictated
largely by the homogeneity of application, necessity of diluting the labelled materials
and the objectives of the experimental work.
Cropping systems and nitrogen cycling
The integration of trees, especially NFTs, into stable agroforestry systems can con-
tribute significantly towards restoring and maintaining soil fertility, combating erosion
and desertification and in providing fuelwood. The potential of such trees has already
been shown in alley farming systems in the humid tropics (Kang et al., 1981, 1985).
In this system in which rows of NFTs alternate with several rows of food crops,the NFTS
are periodically lopped, the foliage is used as green manure for food crops or as forage
for animals and the stems provide fuel. However much of scientific background required
for the development of agroforestry technology is still in its initial stage and almost
nothing is known about the magnitude of factors that influence N, fixation and N cycling
processes in this system.
344
The cropping systems could also affect N, fixation as shown in grain and pasture
legumes (Danso et al., 1986) by influencing the suitability of the reference crop and therefore
the validity of the estimates made.
Variations in N, feting activity with the age of trees, or interference by different
processes such as litter fall and its decomposition, and the redistribution of nitrogen in
the different compartments of the tree/soil system may present some difficulties for measuring
N,furation using lsN isotopic methods. Significant amounts of N, fixed by NITS are undoubtedly
finally incorporated into the soil, but no attempts have been made so far to quantify
their magnitude and significance on soil N pool. This could be very important in agroforestry
systems such as in alley cropping where NFTs and non-NFTs are grown in alternate
rows. The soil nutrients under these contrasting plants have been reported to mineralize
at different rates due to differences in litter quality (Swift, 1984, 1985). NITS, because
they contain more N than non-NFTs tend to contribute more N to the soil. This may
lead to conditions where the fixing and non-fixing crops are not absorbing soil N from
the same isotopic composition (15NP4N ratio) during the growing period and thus invalidate
one of the most important assumptions of N, fixation estimates (Chalk, 1985). These
factors can significantly affect the application of the r5N methodology. The periodic collection
and an estimation of the N in fallen leaves in a given area could be used to assess the
error of measurement caused by N in falling and decomposing litter.
Transfer of N, fared by NFfs to associated reference non-NFTs could be another
source of error for N,-fixation estimates (Fried et al., 1983 ; Chalk, 1985 ; Danso et
cd., 1986).
Sampling of plant material
Witty (1983) emphasized that N, fixation estimates based on isotope dilution are
not estimates of the amount of N, fixed by a crop, but rather estimates of the amount
of fixed N contained in the harvested portion of the crop. The r5N enrichment in different
plant parts of crops grown on “N enriched soils frequently differ (Fried et al., 1983 ;Rennie
et al., 1978). Measurements of N,-fixation, based on only one plant part are therefore
not adequate because the “N enrichment of one plant part is not representative for the
whole plant. This difference in enrichment among plant parts has been cited as a problem
in the “N-isotope technique for estimating N,-fmed (Heichel et al., 1984 ; Chalk, 1985).
According to Danso et &(1986) and Fried et &(1983), errors due to differences in
r5N enrichment can be minimixed by sampling each of these plant parts with fairly uniform
N-isotopic composition separately for N and 15N determinations and using theweighted
atom % lsN as a basis for the N, fixation estimate for the whole plant (Danso et aI.,
1986). This seems to be a standard procedure adopted in most N2 fixation studies that
use 15N and is feasible for tree seedlings and young NFTs. Problems may occur when
sampling big trees, because of their high above and below-ground biomass. In contrast
to grain legumes, NFIs have significant proportions of their total biomass and N content
below ground (Young 1985). In this case, quantification of N, fixed that disregards roots,
nodules and crowns would result in serious errors and the proportion of N, fixed may
be largely underestimated. In addition, for many perenniel legumes the N stored in these
underground structures and unharvested portions of foliage may, after decomposition exert
a significant effect on the dilution of the soil 15Nr4N ratio. This is a practical problem
345
which may not be easy to resolve, given the difficulty in recovering all roots and litter.
As suggested by Danso (1986), it is advisable to estimate this stored N and make
a correction for its effect on the isotopic composition of N at harvest, An examination
of procedures for obtaining a representative sampling (as against whole plant sampling)
that would closely reflect the overall r5N enrichment of the whole plant is urgently
needed.
In alley cropping and most the agroforestry systems however ‘W assay in leaf sam-
ples appears to be the practical method of sampling. The validity of the technique would
be based on the assumption that any defined type of leaf sample would probably provide
the same answer with regard to N, fixation of the whole plant. Leaves of similar age
and morphological position should be sampled from the tree at various intervals after
application of lsN. The method of sampling however, should be established after a series
of preliminary experiments. Another approach may involve the forester methods of using
formula to estimate the standing biomass of trees. In this case, it should be possible sample
a few trees to estimate small portions of different organs of the tree for lsN enrichment
and using the formula to estimate total N, fixed.
Sampling of NITS constitutes the greatest problem in N,-fixation estimation and
ways to overcome this needs to be seriously addressed. These sampling procedure problems
are further confounded by the practicality to use single trees usually as experimental units
to cut down cost, a practice that introduces high variability in the estimates.
N, fixation estimates by some NFT using 15N isotope techniques
The isotope dilution method has predominantly been used to evaluate N, fixation
by a few trees in the field. Estimates of N,-fixed of some NFI’s by different methods
used are summarized in Table 3.
The natural 15N abundance based on the study of small differences between the
natural abundance of 15N in non-fixing and fixing trees has been used for tree by few
workers (Khol and Shearer, 1981 ; Domenach, 1987). One of the first studies using this
method was carried out on Prosopis in the Sonoran desert. The natural 15N abundance
in the tree was significantly lower than in the soil, indicating that it had fixed about
40 to 60% of its nitrogen though no nodules were found. Prosopis was presumed to
develop nodules on deep roots which are not normally harvested (Virginia et al., 1981).
More recently, Hogberg (Pers. comm.) reported in a pilot study of N,-fixing potential
of legume trees in a mixed tropical woodland in Tanzania that the N:P ratio and the
15N natural abundance appeared to be more reliable indicators on N,-fixation than the
N content alone. However, its application is limited by the availability of a high precision
massspecm>meterandspecialsamplepreparationprocedures(Amargeretal.,
1979; Domenach
et al., 1979). Other problems include the natural variations in atom 96 15N in soil with
depth (Steele et al., 1981) the spatial variability in the field (Shearer et al., 1978), and
the soil enrichment in 15N during litter decomposition and subsequent plant i5N uptake
(Turner et al., 1983 ; Domenach, 1987). This could be important in agroforestry systems
such as alley cropping where NFTs are sequentially cut and used as N source after decom-
position (Kang et al., 1985).
The feasibility of using 15N-depleted material for assessing N, fixation of legu-
minous treeshasonlybeeninvestigatedonce(KesselandNakao, 1986). Significantdifferences
in atom % 15N and percentage N derived from N2 fixation were measured in Albizziu
Table 3 : N2 fixation estimates (Proportion <rtd actual amounts of N2 fixed) by NFT's
N2 fixation estimates
Authors
Fixing plant
% Nitrogen derived from Atmosph.
N2 fixed from Atmosphere (g/pL)
Difference
Isotope
dilution
A-value
Difference
Isotope
dilutton
A-Value
-
~:;UM;ER E 2.
C. equisetifolla
49
55
3
9
3.07
3.27
2.31
CORNET et al. (1985)
-
-
A. holosericea
29
33
32
1.12
1.35
1.26
SANGINGA et al.
C. leucocephala
39
4Q
ND
98*
134*
ND
(1985) - -
ZAHARAH et al.
L. leucocephala
-
ND
78
ND
ND
230*
ND
(1986) - -
,-
NDOYE and DREYFUS
S. rostrata
35
ND
0.59
0.62
ND
(1987)
S. sesban
18
:i
ND
0.12
0.13
ND
-
-
SOUGOUFARA et al.
C. equisetifolia
62
45
46
3.8
2.5
2.8
(1989) - -
SANGINGA et al.
Casuarina,
NO
30-80
30-80
ND
NO
ND
(unpublisliZdT-
Allocasuarina
SANGINGA et al.
;: law;z;ephala
ND
6-52
37-72
4-37
21-47
(unpublisl%dr
ND
14-44
14-37
;;
3-9
3-11
GliYiXKa sepium
ND
19-78
28-79
ND
13-40
28-52
ND * not determined
l
= kg/ha/yr
347
l&beck and leucaena indicating the possibility of 15N depleted ammonium sulphate for
measuring N,-fixation and N-allocation in NITS. However. difficulties such as those des-
cribed for the natural abundance methods may limit the usefulness of this method for
measuring N,-fixed by trees.
Acknowledgements
We thank G. Hardarson for his careful reading of the manuscript and his helpful
criticisms and suggestions. We also acknowledge the assistance of M. Tadjbakhsh for
secretarkd services.
References
ALLEN, O.N. and ALLEN, E.K. 1981 The leguminosae. A source book of characteristics,
uses, and nodulation. Madison, University of Wisconsin Press.
AMARGER, N., MARIO-I-II, A., MARIOTTI, F., DURR, J.C., BOURGUIGNON, V and
LAGACI-IERIE, B. 1979 Estimate of symbiotically fixed nitrogen in field-grown soy-
beans using variations in 15N natural abundance. Plant and Soil, 52, 269-280.
BROMFIELD, E.S.P. and AYANABA, A. 1980 The efficacy of soybean inoculation on
acid soil in Tropical Africa. Plant and Soil, 54, 95-106.
CHALK, P.M. 1985 Review. Estimation of N, fixation by isotope dilution : An appraisal
of techniques involving 15N enrichment and their application. Soil Biol. Biochem.
17, 389410.
CORNET F., OTTO, C.. RINAUDO, G., DIEM, H.G. and DOMMERGUES, Y. 1985
Nitrogen fixation by Acacia holosericea grown in field simulating conditions. Acta
Oecologica, 6 (20), 211-218.
DANSO, S.K.A. 1985 Methods of estimating biological nitrogen fmation. In: Biological
Nitrogen Fixation in Africa. PI. Ssali and S.O. Kew. eds.) nn. ??4-?44 Rhivhium
MLRCEN, Nairobi, Kenya.
DANSO, S.K.A. 1986 Estimation of N24ixation by isotope dilution: An appraisal of tech-
niques involving 15N enrichment and their application-Comments. Soil Biol. and Bio-
them. 18, 243-244.
DANSO, S.K.A., HARDARSON. G. and ZAPATA, F 1986 Assessment of dinitrogen
fixation potentials of forage legumes with 15N technique. In : proceedings of a Workshop
on Potentials of Forage Legumes in FarmingSystems of Sub-Saharan Africa, (Haque,
I., Jutzi, S. and Neate, PJ.H. eds.) p. 26-57. ILCA, Addis Ababa.
DOMENACH, A.M. 1987. Estimation de la fixation symbiotique chez des plantes he&a&es
et ligneuses ; utilisation et validite de la m&bode ba& sur la mesure des abondances
isotopiques naturelles de l’azote. These de Doctorat - Lyon I. 189 p.
DOMENACH, A.M., CHALAMET, A. and PACHIANDI, C. 1979 Estimation de la futation
d’azote par le soja a l’aide de deux methcxles d’analyses isotopiques. CR. Acad. Sci.
Ser. D 289, 291-293.
348
DOMMERGUES, Y. 1982 Ensuring effective symbiosis in nitrogen fixing trees. In : Bio-
logical nitrogen fixation technology for tropical agriculture. (Graham, P.H. and Harris,
SC. eds.) p. 395413. CIAT. Cab, Colombia
DOMMERGUES Y. 1987 The role of biological nitrogen fixation in agroforestry. In :
Agroforestry a decade of development. (Steppler, H.A. and Nair, P.K.R. eds.) p. 245-
271 ICRAF - Nairobi, Kenya.
FRIED, M., DANSO, S.K.A. and ZAPATA, F. 1983 The methodology of measurement
of N, fixation by non-legumes as inferred from field experiments with legumes. Can.
J. Microbial. 29, 1053-1062.
GAUTHIER, D.L., DIEM, H.G., DOMMERGUES, Y. and GANRY, F. 1985 Assessment
of N, fixation by Casuurina equisetifolia inoculated with Frankia 0RS021001 using
“N methods. Soil Biol. Biochzem. 17, 375-379.
HARDARSON, G., ZAPATA, F. and DANSO, S.K.A. 1984 Field evaluation of symbiotic
nitrogen fixation by rhizobial strains using 15N methodology. Plant and Soil. 82,369-
375.
HARDARSON, G., DANSO, S.K.A. and ZAPATA 1987. Biological nitrogen fixation
in field crops. In : Handbook of Plant Sience in Agriculture (Christie, B.R. ed.) p.
165-191 CRC Press.
HEICHEL, G.H., HARDARSON, G., BARNES, D.K. and VANCE, C.P. 1984 Dinitrogen
fixation. herbage yield, and rhizobial preference of selected alfalfa clones. Crop Sci.
24, 1093-1097.
H&BERG P. and KVARNSTROM, M. 1982 Nitrogen fixation by the woody legume
Leucuena leucocephalu. Plant Soil 66, 21-28.
KANG, B-T.. WILSON, GF. and SIPKENS, L. 1981 Alley cropping maize (Zea muys
L.) and leucaena (Z.eucuenu leucocephalu (Lam.) de Wit in Southern Nigeria. Plant
and Soil, 63, 165-179.
KANG, B.T., GRIMME, H. and LAWSON, T.T. 1985 Alley cropping sequentially cropped
maize and cowpea with leucaena on sandy soil in Southern Nigeria. Plant and Soil
_,
_
o-)> iv i .-.L / , .
KESSEL, C. VAN and NAKAO, P. 1986 The use of nitrogen-15depleted ammonium
sulfate for estimating nitrogen fixation by leguminous trees. Agronomy J. 78, 549-
51.
KESSEL, C., VAN, NAKAO, P., ROSKOSKI JP. and KEVIN, K. 1987 Ureide production
by N,fixing leguminous trees. Soil Biol. Biochem (in press).
KOHL, D.H. and SHEARER, G.B. 1981 The use of soils lightly enriched 15N to screen
for N2 fixing activity in plants. Plant and Soil 60,, 487-489.
KNOWLES, R. 1981 The measurement of nitrogen fixation. In : Current Perspectives
in Nitrogen Fixation (Gibson, A.H. and Newton, W.E. eds.). pp. 327-333 Australian
Academy of Sciences Canberra, Australia
LEGG, J.O. and SLOGER, C. 1975 A tracer method for determining nitrogen fixation
in field studies. Proc. Int. Conf. Stable Isot. 2nd 1975. pp. 661666.
349
LIJLANDALA, L.L.L. and HALL, J.B., 1988. Leucaena leucocephala’s biological nitro-
gen fixation : a promising substitute for inorganic nitrogen fertilization in agrofo-
restry systems. In : Biotechnology of nitrogen fixation in the tropics (BIGtufT), Pro-
ceedings of UNE!XO Regional Symposium and Workshop, UPM, Malaysia, 25-29
August 1986, p. 259-265.
NAS. 1979. Tropical legumes : Ressorces for the future. National Academy hess Washington
D.C. 331 p.
NDOYE, I. and DREYIFUS, B. 1988 N2 fvtation by Sesbania rosfrutu and Sesbunia sesbun
estimated using 15N and total N difference methods. Soil Biol. und Biochem. 20,209-
213.
RENNIE, R.J. 1982 Quantifying dinitrogen fuation in soybean by 15N isotope dilution:
The question of the non-fixing control plant. Can J. Bot. 60, 856861.
RENNIE, RJ. and RENNIE, A.D. 1983 Techniques for quantifying N2 fixation in asso-
ciation with non-lehumes under field and greenhouse conditions. Can. J. Microbial.
29, 1022-1035.
RENNJE, RJ. and KEMP, G.A. 1983 Dinitrogen fixation in field beans quantified by
15N isotope dilution. I. Effect of strains of Rhizobium phuseoii. Agron. J. 75, 640-
644.
RENNIE, R.J.; RBNNIB, D.A. and FRIED, M. 1978 Concepts of 15N usage in dinitrogen
fixation studies. In : Isotopes in Biological dinitrogen fixation. IAEA, Vienna, Austria,
STI/PUEV478 pp. 107-130.
REDDELL, P.W., BOWEN, G.D. and ROBSON, A.D. 1986 Nodulation of Casuarinaceae
in relation to host species and soil properties. Aust. J. Bot 34, 435-44.
ROSKOSKI JP. 1981 Nodulation and N, fixation by Zngu jinicuil, a woody legume in
coffee plantations. I. Measurements of nodule biomass and field C!& reduction rates.
Plant Soil 59, 201-m.
SANGINGA, N., MULONGOY K. and AYANABA, A. 1985 Effect of inoculation and
mineral nutrients on nodulation and growth of Leucuenu leucocephulu. In : Biological
*,;u”giiL AlLLUII ill AiIILd. \\3kl**, Il. ;IllLI I\\cya,s.u.txs.), p. 4lY-;t30. iwkCM4,P4aIrobI.
SANGINGA, N. 1986 Inoculation of Leucaenu leucocephulu (Lam.) de Wit with Rhizobium
and its nitrogen contribution to a subsequent maize crop. Biol. Agric. Hort. 3, 347-
352.
SANGINGA. N., MULONGOY, K. and AYANABA, A. 1988a Nitrogen of field ino-
culated Leucuena Zeucocephulu (Lam.) de Wit estimated by the nitrogen-15 and the
difference method. Plant and Soil (In Press).
SANGINGA N., DANSO, S.K.A. and BOWEN, G.D. 1988b Nodulation and growth response
of Cusuurinu and Allocusuurinu to P fertilization in an acid soil. Plant und Soil.(Ac-
cepted).
SHEARER, G., KOHL, D.H. and CHIEN, S.H. 1978 The nitrogen-15 abundance in a
wide variety of soils. Soil Sci. Sot. Am. J. 42, 899-902.
SOUGOUFARA B., DUHOUX, E. and DOMMERGUBS, Y.R. 1987 Improvement of
nitrogen fixation by Cusuurinu equisetifoliu through clonal selection. Arid Soil Research
and Rehabilitation 1, 129-132.
350
SOUGOUFARA B., DIEM, H.G. and DOMMERGUES, Y.R. 1988 Response of tield-
grown Cusuarina equisetifoliu to inoculation with Frunkia strain ORS021001 entrap-
ped in alginate beads. Plant and Soil (Accepted).
STEELE, K.W. 1983 Quantitative measurements of nitrogen turnover in pasture systems
with particular reference to the role of lsN. In : Nuclear Techniques in Improving
Pasture Management. IAEA, Vienna, Austria. pp. 17-35.
SWIFT, MJ. 1984 Soil biological processes and tropical soil fertility: a proposal for col-
laborative programme of research. Biol. Internat. Spec. Issue 5, 38 pp.
SWIFT, M.J. 1985 Tropical soil biology and fertility (TSBF). Planning for research. Biol.
Internat. Spec. Issue 9, 24 pp.
TORREY, J.G. 1982 Camurina - Actinorhizal nitrogen-fixing tree of the tropics. In :
Biological Nitrogen Fixation Technology for Tropical-Agriculture (Graham, P.H. and
Harris, S.C. eds.) p. 427-441. CIAT. Cali, Colombia.
TURNER, G.L. and BERGERSEN, F.J. 1983 Natural abundance of N in root nodules
of soybean, Iupin, subterranean clover and luceme. Soil Biol. Biochem. 15, 525-530.
VOSE, P-B., RUSCHEL, A.P. VICTORIA, R.L., SAITO, S.M.T. and MATSUI, E. 1982
15N as a tool in biological nitrogen fEation in soil. In : Biological Nitrogen Fixation
Technology for Tropical Agriculture (Graham, P.H. and HARRIS, S.C. eds.) p. 575
593.
VIRGINIA, R.A., JARRELL, W.M. KOHL, D.H. and SHEARERE, G.B. 1981 Symbiotic
nitrogen fixation in Prosopis (Leguminosae) dominated desert ecosystems. In : Current
perspectives in nitrogen fixation. (Gibson, A.H. and Newton, W.E. eds.) Canberra:
Australian Academy of Science.
WITTY, J.F. 1983 Estimating N-fixation in the field using lSN-labelled fertilizer: some
problems and solutions. Soil Biol. Biochem. 15, 631-639.
YOUNG, A. 1985 The potential of agroforestry as a practical means of sustaining soil
fertility. ICRAF Working Paper 34, 28pp.
ZAHARA, A.R., SHARIFUDDING, H.A.H. and STIBRAMANIAM. R. 19%. Nitrrmn
lmauon by Leucaena leucocephala as measured by N-15 dilution technique. Pertanika
91, 17-22.
CONCLUSIONS ET RECOMMANDATIONS
MaCmiser la FRA pour la production agricole et forestière en Afrique
Les recherches sur la fixation biologique de N,
et leur impact présent et futur sur la production
végétale et forestière en Afkique
DOMMERGUES, Y.,(‘)
BORDELEAU, L.MJ2) et GUEYE, ML3)
(1) : BSSFTKTFT, 4.5 bis Av. de la Belle Gabrielle, Nogent sur Marne, France ;
(2) : Canada Agriculture, Station de Recherche Sainte Foy,
2540 Blvd Hochelaga. GIV 2J3 Canada ;
(3) : MIRCEN-CNRA, BP. 53 Bambey, Sénégal
La troisième Conference de I’AABNF qui s’est tenue a Dakar du 7 au 12 no-
vembre 1988 revêtait, au moins autant que les prt!c&lentes, une importance considérable
pour le développement de l’agriculture africaine. Les experts sont en effet unanimes à
reconnaître que l’Afrique n’a que très faiblement bénéficié de la Révolution Verte en raison
notamment des caractkistiques particulièrement defavorables de beaucoup de ses sols et
de certains de ses climats. Cest seulement en recourant a un matériel veggétal robuste
et frugal, notamment en ce qui concerne l’azote, que l’on pourra rattraper le retard accu-
mulé par rapport a d’autres continents. Il est &ident que l’optimisation de la fixation
de l’azote constitue une voie prometteuse mais encore insuffkmment exploitée pour at-
teindre cet objectif.
Dans cette note finale, on tentera de souligner les domaines dans lesquels les efforts
entrepris devront être poursuivis et intensifiés. Ces domaines concernent plus particulie-
rement la maîtrise de l’inoculation avec les microorganismes fixateurs d’azote, l’kude qua-
litative des effets de l’inoculation et l’&&ration de la fixation de l’azote, les stratégies
d’amelioration des partenaires des symbioses f=atrices d’azote et lIntt5gration des systemes
fixateurs de N, dans les systèmes culturaux.
Outre ces recommandations qui portent sur les programmes de recherche fonda-
mentale ou appliquee, on indiquera les grandes lignes de la politique à suivre pour mener
a bien ces programmes de recherche et assurer le transfert des rbu1tat.s à l’application.
353
Maîtrise de l’inoculation
Identification des caractbristiques des sols
Il est absolument necessaire que, contrairement a ce que l’on observe trop
souvent, l’on fasse pr&%der les essais d’inoculation d’une étude systématique des sols.
Cette étude doit consister essentiellement a déterminer les niveaux de populations de
Rhizobium (ou F~&&I) prksents dans les sols que l’on utilisera. Compte tenu de l’état
actuel de nos connaissances et, sauf cas particulier, si le nombre de Rhizobium com-
pétents dt$asse le seuil de 102 2 IV par g de sol, il ne faut pas s’attendre a une r@onse
significative. & l’inoculation. Cette règle relative aux Rhizobizun ou Frankia s’applique
Cgalement aux champignons mycorhiziens et Cventuellement aux bact&ies fixatrices
d’azote en association. Il importe de mettre au point des méthodes relativement simples
et ne nécessitant aucun appareillage particulier pour determiner approximativement ce
potentiel infectieux des sols.
L’analyse chimique des sols effectuée conjointement devra consister à déter-
miner le pH et si possible en analyser le complexe absorbant, évaluer les teneurs en
C, N, P (assimilable), K et Cventuellement permettre l’identification de carences en
certains f%ments.
S6lectlon des souches
Cette sélection sera effectuée à partir des collections de souches, notamment celles
des h4IRCENs.
Parmi les critères dont il sera nécessaire de tenir compte, le critère de sp6cificité
devra être considéré avec attention. Il est désormais admis par beaucoup d’experts qu’il
vaut mieux developper des souches spkcifiques (c’est-Mire à spectre d’hôte étroit) que
des souches a large spectre.
Conditionnement des inoculums
La tourbe est encoure le support le plus souvent utilise, mais d’autres supports
méritent encore d’être essayés notamment dans tous les pays où il est difficile de se procurer
la tourbe. L’usage des inoculums matriciels - résultant de l’inclusion des microorganismes
dans une matrice de polymère - devrait être développé à l’avenir.
Survie des inoculums
Les inoculants doivent conserver leur pouvoir infectieux le plus longtemps possible
au cours du stockage et même après l’application dans le sol ou sur les graines. Il est
souhaitable qu’une méthode simple puisse permettre, dans un proche avenir, de vkifier
la survie des bactkrks dans l’inocuhun immkdiatement avant l’application de celui-ci au
champ.
Pour répondre à la question de savoir si un inoculum doit être appliqué plusieurs
années de suite, il est nécessaire que l’on mesure, dans différents sols et sous diffkrentes
conditions climatiques, le taux de survie des bactéries apres la ou les premieres appli-
cations d’inoculum.
354
Etude qualitative des effets de l’inoculation et haluation de la fixation
de N,
Effets spkifiques autre que la fixation de N,
Il est certain que l’inoculation d’une plante avec Rhizobium (ou Fmnkiu) peut mo-
difier la distribution de l’azote dans la plante, cet effet étant mesuré par l’indice de récolte ;
il peut y avoir aussi une modification du développement des parties aériennes par rapport
aux racines, cet effet étant mesure par le rapport parties akriennes (shoots) / racines. On
sait aussi que les Rhizobium peuvent, dans certains cas particuliers, réduire l’incidence
de l’attaque des pathoghes.
Il apparaît n6cessaire de developper les recherches de nature phénoménologique
sur ces effets afin de pouvoir en entreprendre ensuite une étude mécanistique, ce qui
implique le developpement de recherches sur la physiologie des symbioses consider6es.
Evaluation de l’azote fixe! par voie biologique dans les hosyst&mes
Il convient de souligner ici qu’avant de choisir la méthode d’estimation à utiliser
(notamment s’il s’agit dune méthode isotopique qui peut être relativement coûteuse) il
est indispensable de définir les objectifs prkis de l’expkimentation. C’est ainsi que la
sélection de clones d’une plante donnée en fonction de son pouvoir fixateur de N, ne
requiert pas obligatoirement le recours a la méthode isotopique.
Il faut être toujours conscient (i) du fait qu’aucune méthode n’est parfaite, (ii) de
la rkcessité de mettre en place une ou plusieurs plantes témoins et (iii) de la nécessité
d’apporter le plus grand soin au marquage du sol (lorsque l’on utilise la methode de di-
lution isotopique).
Les problemes spkifiques aux arbres ont Cd &oqués à plusieurs reprises au cours
de la conference. On ne les &mm&era pas ici, mais il apparaît nkessake d’attirer l’at-
tention sur ces problemes dont certains risquent de passer inaperçus.
Il est bien connu que l’extrapolation au champ d’essais effectues en pot est très
risquée. Toutefois, les n%ultats rapportes suggerent que la culture de plantes dans des
bacs de très grandes dimensions (tanks) pourrait, dans une certaine mesure, se rapprocher
de la culture en plein champ. Il est donc recommandé de faire appel autant que possible
à ces dispositifs expkimentaux.
Stratégies d’am&lioration des partenaires des symbioses fixatrices de N,
Ces stratkgies sont bien COMIKS dans leur principe ; elles portent a la foi sur
la plante hôte et sur le microsymbiote. Jusqu’à pnkent., elles ont Cté fondées essentiel-
lement sur l’emploi de méthodes conventionnelles qui consistent à exploiter la variabilité
naturelle des populations microbiennes ou vegetales. Mais grâce au développement
rkcent de la biologie mol&Aaire, on peut envisager avec optimisme la mise au point
de nouvelles techniques d’amelioration des deux partenaires de la symbiose.
Dans le cas des arbres fixateurs de N,, la stratégie fondke l’expIoitation de la varia-
bilité intraspkcifique des plantes hôtes et la multiplication végétative des individus les
plus performants devrait permettre, dans un tres proche avenir, de mettre à la disposi-
tion des agriculteurs et forestiers un matkriel vegétal à haut rendement et doté d’un
potentiel fixateur de N, considerablement accru.
IrMgration des plantes flxatrices de N, dans les systèmes culturaux
Jusqu’à présent, il est certain que, sauf rares exceptions, peu d’efforts ont éte con-
Sac&s à l’étude de l’int&ation des plantes f=atrices de N, dans les systemes culturaux
qu’il s’agisse de cultures mixtes (multiple cropping), cultures alternées (intercropping),
ou cultures séquentielles (sequential cropping), ou de systèmes agroforestiers. Il apparaît
indispensable : (i) de chercher à identifier et à quantifier les interactions entre les cultures
composant les systèmes et les microorganismes du sol notamment ceux qui intervien-
nent dans le cycle de l’azote ; (ii) d’évaluer avec précision les économies d’azote que
l’intégration des plantes f=atrices de N, permet d’obtenir dans les différents systkmes cul-
turaux ; (iii) de mettre au point les méthodes d’aménagement des differents systèmes de
façon à tirer le profit maximum des apports d’azote provenant de la fixation et de façon
à assurer au mieux la conservation et la rCgén&ation des sols.
Ces différentes investigations se heurteront à des difficultés qu’il faudra tenter de
surmonter, ces difficultks Ctant liées notamment (i) à l’évaluation de la fixation biolo-
gique de N,, (ii) à l’évaluation des transferts d’azote des plantes fixatrices aux plantes
non fixatrices de N,, (iii) aux problemes inhérents aux recherches sur le terrain, re-
cherches qui doivent Ztre conduites dans des conditions simulant aussi parfaitement que
possible celles de la ferme.
Politique de formation, de collaboration et de vulgarisation
A l’unanimité, les s&minaristes sont arrivés a la conclusion que l’effort de formation
entrepris devait être amplifie, étant entendu que cette formation devrait être continue.
Une attention particulière a été accordée à la conception des programmes établis
en collaboration dans le sens Nord-Sud et Sud-Sud, ces programmes devant idéalement
presenter l’un et/ou l’autre des caractires suivants, c’est-à-dire être : (i) focalises sur des
thèmes bien précis ; (ii) liiit6s à des zones géographiques pn5sentant de grandes simi-
laritis sur le plan écologique ; (iii) fondés sur l’exploitation des découvertes récentes
ouvrant la voie a des applications pratiques importantes (ex. découvertes des susbtances
élicites des gènes nod excrétées par la plante hôte) ; (iv) considérés comme prioritaires
dans une région donnée.
La vulgarisation des résultats obtenus au niveau du laboratoire ou de la station
de recherche devrait faire l’objet d’un effort tout particulier. Il s’agit Iii dune activité dif-
férente de celle du chercheur de laboratoire ou de station et à laquelle des chercheurs,
ingenieurs et techniciens sp6cialist% devraient être affectes. Dans le cadre de cet effort
de vulgarisation, il est souhaité que l’on quantifie, partout où ce sera possible, les bénéfices
qui pourront découler, pour le paysan, des nouvelles technologies mises a sa disposition.
Liste des participants
ANGLETERRE
AMIJEE, FEROZ
Wye College London University WYE, KENT
TN25 5AH United Kingdau
ANGOLA
PASSOS TEIXEIRA G. DEIP Biologica Faculdade de Ciêucias Da Uni-
LIDIA
versidade A. Neto Av. 4 de FevereironT3 Luanda
R.P. ANGOLA
C.P. No425 Faculdade de Ciencias Agrarias Uni-
vasidade A. Neto Luanda RP. ANGOLA
MORAS CORDEIRO,
C.P. Np 425 - Faculdade de Ciencias
J. MANUEL
Agrarias Uuiversidade Agostinho Neto
Luanda - RP. ANGOLA
BELGIQUE
DESMADRYL,, DIRK
Laboratoire de Physiologie Végétale Institut
Camoy, Place Croix du Sud, 4, Université
C8thdique&eLwvain,B-l348Lowain-La-Newe,
BELGIQUE
Laboratoire de Physiologie VegCtale Institut
VAN HOVE, CHARJ&S Camoy. Place Croix du Sud, 4, UniversitC
CathdiquedeLouvain,B-1348Lowain-La-Nwve.
BELGIQUE
BRESIL
BODDEY, ROBERT
EMBRAPA - UAPNPBS, Km 47. S&opédica,
MICHAEL
23851 Rio de Janeiro, BRESIL
BURKINA FASO
DIANDA, kfAHAM,%DI lRB~~B.P.7047Guagadougcu, BURKINA FASO
Station de Recherché 2560 Hochelaga, Saiite Foy
CANADA
BORDELEAU, LUCIEN ~utbec GIV 2j3. CANADA
CONGO
PANDZOU, JOSEPH
Centre &Recherches Agronauiques de Loudima,
B.P.28 Lmdima, CONGO
COTE DWOIRE
IUMOU, ANKOMIAN
D6ptemeut d’Agratomie Laboratoire de Micro-
biologie et des Iudustries Agro-alimentaires
ENSAA. B.P.35 ABIDJAN 08
EGYPTE
BADR EL DIN, SAID
NationalResearchCmterDokki.Cairo,EGYPT
KHALAFALLAH, MOHA- National Research CmterDokki, Cairo,EGYPT
MED A.
MOAWAD, HASSAN
National Research Cmter Dokki. Caim. EGYPT
ETATS UNIS
ANGLE, FAY SCOTT
Department of Agronomy University of
D’AMERIQUE
Maryland College Park, Maryland 20742 USA
DODO, HORTENSE
Tuskege University Alabama 626 South PUGH
St # 4 College PA 16801. USA
WEAVER, RICHARD
Dept. Soil and Crop Sciences Texas A CI M
University College Station, Texas USA 77843
FRANCE
BJZJq =FA. mIJA Socitté CALIOPE, 17 Rue SEBASTOPOL
34500 - Béziers - FRANCE
BEUNARD, PIERRE
IRATKIRAD Avenue du Val de Montferrand
B.P.5035 34032 MontpeBier Cedex FRANCE
DOMENACH, ANNE
Université Lyon f, Laboratoire de Biologie des
MARIE
Sols, Bat741, 43 Bld du 11 Novembre 1918,
69622 Vi!.leurbanne Cedex FRANCE
DOMMJZRGUES, YVON
BSSFf/Cl’pT, 45 Bis. Avenue de la Belle
Gabrielle 94736 Nogent/Mame FRANCE
DUHOUX, EMILE
BSSFftCTpT, 45 Bis, Avmue de la Belle
Gabrielle 94736 NogmtlMame FRANCE
FERRET, ROGER
Société CALIOPE, 17 Rue SEBASTOPOL
34500 - BCziers - FRANCE
GALIANA
BSSFT/CFI’. 45 Bis. Avenue de la Belle
Gabrielle 94736 Nogent/Mame FRANCE
MAROC
AURAG, JAMAL
Université Mohamed V. Facultt des Sciences, De-
partement de Biologie, Labo.de Microbiologie -
B.P.1014. Rabat, MAROC
BERRAHO, EL BEICKAY
Université Mohamed V, Faculté des Sciences. Dé-
paxtement de Biologie, Labo.dc Microbiologie -
B.P.1014. Rabat, MAROC
HKLALI, ABDELALI
D+rtemmt des Sci- du Sol, Institut Agm-
nomique et V&kinaire Hassan IIE .P.6202. Rabat,
MAROC
NIGERIA
~ONGOY, WEMANJ
LLT.A., O y o R o a d - P M B 5 3 2 0 Ibadan, NIGERIA
ISICHEI, AUGUSTINE
Depanment of Botany. Obafemi Awolowo
University Ile-Ile, NIGERIA
ODEYEMI, OLO
Department of Microbiology - Obafemi,
Awolowo Univeniry, Ile-Ile. NIGERIA
OUGANDA
San ME (‘MS-
Departmmtof Soil Scienœ.Facultyof Agriculmre
andFomstry.MakerereUniversiry,P.O.BOX 1002,
TINE’
Kampala, OUGANDA
PAYS-BAS
v&q BRU~~, JO~
Uuivaity of Dar Es Splam. Bot=~ Departmfflt.
P.O.BOX 35060, Dar Es Salam. TANZANIA
RWAh’DA
RUGABA, SILAS
Projd Rizicole Butare B.P.76. Butare, RWANDA
HAKIZIMANA, A.
LS.A.R. - RUBONA - B.P. 138 - Butare
RWANDA
SIERRA LEONE
A.MARA, DENIS
soascimœ~mSNyllaUnivnsityCdlege,
PMB. Freetown. SIFRRA LEONE
TAA’ZANIE
MuRuKE, MASOUD
UniversityofDuEaSalam,DepartmmtofBotany,
SALUM
Microbiology Iabarrtory, P.O.BOX 35060. Dar
Es Salam, TANZANIA
FRANCE
OBATON, MICHEL
Laboratoire de Recherches sur les Symbiontes des
Racines INRA, 1 Place Viala 34060 Montpellier Cedex
FRANCE
RINAUDO, GIRARD
Universiti Claude Bernard Lyon 1. Ecologie Micro-
bienne (Bat741) 6%22 Villembme Cedex FRANCE
SALL, KEBA
Laboratoire de Recherches sur les Symbiontes des
Racines INRA, 1 Place Vi& 34060 Montpellier Cedex
FRANCE
GHANA
KALEEM, FEZRHAT
CropsResearchInstituteP.O.BOX55-Educati<mRidge
TamaIe - GHANA
HAITI
FELIX, JEAN FENEL
2290 Rue Ste Foy G, W, A8 Québec, CANADA
JAPON
ASANUMA, SHUICHJ
Hokkaido National Agricultural Experiment Station 1,
Hitsujigaoka, Toyohira-Ko Sappzm, 004 JAPAN
KENYA
Botany Department University of Nairobi, P.O.BOX
mAm’ m*cols
30197 Nairobi . KENYA
ODEE, DAVID W.
Kenya Foresty Research Institue P.O.BOX 20412,
Nairobi. KENYA
ODUOL, PETER AL-
ICRAF. BOX 30677. Nairobi, KENYA
LAN
LJBERIA
KAGABO, WILSON
CentralAgriculmalResearchInstitute.P.O.BOX3929,
EMMANZI
Ma-mvia, LIBE!RIA
MAW
LAHE3IB MESSAOTJD
Ecole Normale Supérieure B.P.241. Bamako, MALI
TRAORE TAHIROU
Bcole Normale Sup&ieure B.P.241, Bamako, MALI
MAROC
ABOUROUH, MOHA-
DivisionRecherchesForestièresB.P.763.Agdel.Rabat,
MED
MAROC
TUNISIE
BAGHDADI, RIVIERE
Ministi?redel’Agticultuulture@irectiatG&tCrale
KHAOULA
de la Production VCgbale). 30, Rue Alain
Savary, Tunis, TUNISIE
ZAIRE
ALrMASI, MWALIBANTU pro ramme National Engrais, B.P.3325.
g
KinshasaKioinbe, ZAIRE
LUYINDULA, NDIKU
Commissariat Gbtral à 1’Energie Atomique,
BP.868. Kinshasa XI, ZAIRE
ZIMBABWE
NYIKA, MICHAEL AN-
Department of Research and Specialist
DREW
Services. Chemistry and Soil Research Ins-
titute,P/BAB37S7,M~~den,Z1MBABWE
AGENCES INTER-
DIARA, HENRI
W.A.R.D.A. Azolla Pmject, ADRAO Project
NATIONALES
Azolla, BP.%, Saint-Lxis. SENEGAL
RIVEROS, FERNAND0
Groupe de Paturage (Grassland Group), FAO,
Via Delle Terme di Caracalla, 00100, Rome,
lTALIA
SCAGLIA, J. ALEXAN-
F.A.O. AGLF. Via Dclle Terme di CaracalIa,
DRE
00100. Rome, lTALIA
WANGARI, ELISABETH
UNESCOMAB, BP.33 11 .Dakar, SENEGAL
BA, AMADOU TIDIANJZ
UNES33 ABN. BP.33 11. Dakar. SENEGAL
DANSO, SETH KOFI
Point FAO/IAEA Division Soil Science
Section, 5, Wagramerstrasse, 100-A 1400.
Vienne, AUSTRIA
SANGINGA, NTERANYA
Soil Science Unit, IAEA Seibersdorf Labo-
ratoty A 2444, Seibersdorf, Vienne, AUS-
TRIA
HAQUE, 1.
ILCA. P.O.5690, Addis Abeba, ETIBOPIA
SENEGAL.
SIAKA SADIO
ISRA/CNRF BP.53 Dakar. SENEGAL
OUSMANE DIAGNE
Mme AMlNATA NIANE
Is-A/BAbIBEY BP.53 Ehbey,
SENEGAL
BABACAR NDAO
”
DEMBA FARBA MBAYE
n
FRANCIS GANRY
”
Mme FATOU GUEYE
h4AMADOU GUEYE
BERNARD DREYFUS
ORSTOM BEL-AIR BP.1386.Dakar.
SENEGAL
BASSJROU SOUGOUFARA
<*
JEAN PAUL COLONNA
MARC DUCOUSSO
MAGGIA LAURENT
PHILsIE’PE DE LAJUDIE
DANIEL THOEN
n
KODJO TOMEKPE
”
PHILIPPE TRAN
I,
Mme MARIE MADELEINE
Université Cheikh Anta DIOP. Dakar,
SPENCER BARRETO
SENEGAL
Mme YAYE KANE GASSA-
MA SEYE
LES PUBLICATIONS DE L’ISRA
Publications non périodiques
“Cahiers d’Information”
Monographies destikes plus particulièrement
aux producteurs agricoles et à leur encadrement
technico-tkonomique immédiat.
“Etudes et Documents”
Monographies éditées a l’intention des cadres du
développement agricole et des chercheurs.
“Actes”
Proposant des comptes rendus des rkunions,
,
colloques et séminaires dans lesquels ont été
impliques des agents de 1’ISRA.
“Réflexions et perspectives”
Monographies sur la politique de la recherche, du
développement, de la liaison entre la recherche et
le développement.. _
Publications périodiques
“Revue Sénégalaise des Recherches Agricoles et Halieutiques” Publication trimestrielle.
“La Lettre d’Information”
Trimestrielle, donnant des renseignements sur la
vie de l’Institution, B l’usage de ses agents et des
utilisateurs situes en amont et en aval de ITSRA.
“Bulletin analytique Documentaire”
Trimestriel, contenant les réferenccs des travaux
publiés par les chercheurs de l’lSRA, qu’il s’agisse
de rapports, de comptes rendus, de mémoires ou
d’articles...
“Bulletin Agroclimatologique”
Paraissant au moment de l’hivernage.
Pour tous renseignements concernant ces diverses publications et en particulier
leur prix de vente, s’adresser a la Direction Gémkale, et à l’attention de 1’ UNIVAL.
S’agissant de la revue, le tableau suivant donne les conditions de cession
4 En FCFA
Abonnement annuel
Numéro séparé
/
Etats d’Afïique
et Madagascar
6000
2 000
Autres uavs
10000
3000
ue n’a que très faiblement bédficie de la
en raison notamment des caractéristiques
défavorables de ses sols et de certains de ses
e de 1’Azot.e (FBA) est une des voies prometteuses
foresterie en Afrique dans le contexte général de l’éco-
n Africaine 5-m la Fixation Biologique de l’Azote
du 7 au l%@ovembre 1988 a particulièrement
dans lesquels les efforts entrepris doivent être poursuivis et
nference comprenait les sessions suivantes :
l
Amélioration de la FBA par la sélection des plantes ;
l
Amélioration de la FBA pour les pratiques culturales ;
*
Inoculation : production d?noculum et essais d’inoculation ;
*
Mesure de la FBA ;
l
FEIA chez les legumineuses à graines ;
l
FBA chez les légumineuses fourragères ;
l
F%A chez les arbres ;
l
Fixation d’Azote associative ;
l
Association AzollaJAnabaena ;
l
Economie de 1’Azote dans les systemes de culture en association et en
l r-._ _. /
l
/ \\ :
i
/
,
I
I
!
--... -,
1
I-L-=,-, -4
Il est reconnu que l’Afrique n’a que très faiblement bénéficié
de la Revolution Verte en raison notamment des caractéris-
tiques particulièrement défavorables de ses sols et de certains
de ses climats.
Maximiser la furation Biologique de l’Azote (FBA) est une des
voies prometteuses pour développer l’agriculture et la foresterie
en Afrique dans le contexte génkal de l’économie de l’azote.
La 3ème conferennce de l’Association Africaine pour la Fixa-
tion Biologique de 1’Azote (AABNF) qui s’est tenue à Dakar
du 7 au 12 novembre 1988 a particulièrement souligné les do-
maines dans lesquels les efforts entrepris doivent être
poursuivis et intensifies. Cette 3ème conférence comprenait les
sessions suivantes :
l Amélioration de la FRA par la sélection des
plantes ;
l Amélioration de la FBA pour les pratiques
culturales ;
l Inoculation : production d’inoculum et essais
d’inoculation ;
l Mesure de la FBA ;
. * u.,
TTn 4 Chi 12s l~gumincuscs à graines ;
l FBA chez les légumineuses fourragères ;
l FBA chez les arbres ;
l Fixation d’Azote associative ;
l Association AzollaJAnabaena ;
l
Economie de 1’Azote dans les systkmes de culture
en association et en agroforesterie.