Arid Soi1 Rcsearch and Rchabititation,...
Arid Soi1 Rcsearch and Rchabititation, 14:317-327,2ooO
Copytighl 0 2ooO Taylor & Francis
089@3069/W 512.00 + .oO
Alley Cropping of Maize and Gliricidia sepiunz in the
Sudanese Sahel Region : Some Technical
Feasibility Aspects
MAMADOU NDIAYE
Institut Sénégalais de Recherches Agricoles/Centre National De Recherche
Agronomiques (ISRA/CNRA)
Bambey Senegal
FRANCIS GANRY
ROBERT OLIVER
Centre International de Recherches Agricoles en Coopération/Département
Amélioration des Méthodes pour L-Innovation Scientifique
(CIRAD/AMIS)
Programme Agronomie,
Montpellier, France
An association in un ulley cropping experirnent oja short-rerm maize uuriety and a
tree legume (Gliricidia sepium) adapted to the Sudanese Sahel region wus studied
from the point of view of the nitrogen balance and plot yields. Isotopic lubelling.
applied in thefield enabled the contribution io maize nitrogen nutrition from d$k-
ent nitrogen sources (ferrilizer, prunings, and soii) to be quantijîed. For equul Cu!ti-
vated areus, ulley cropping provides a maize yield greater thun thut of muize grown
in pure stand without nirrogen fertifizer. However this yield is only 40% of thut
obtained in pure stands with fertilizer nitrogen. Alley cropping gives u very favour-
able Lund Equivalent Ratio (LER) of 0.90 and 2.58 compared with N-fertilized and
non-N-fertilized plots, respecrively. These LEKs demonstrate the increased bio-
logical eficiency of the cultivated soi1 in un ugrojïorestry system. The percentage of
che total nitrogen in rhe muize coming from prunings (Nfdp) varies between 30 and
35% and the true coeficient of nitrogen urilization of the prunings (TCUp) varies
from 15-25%. In the environment of central Spnegal, the percentage of total nitro-
gen of G. sepium cominy jrorn N, jxatiorl is yuifp low (NQjjx = 25X), and conse-
quentiy, in the maize, the nitrogen cominy .from N, fixation (Ndfix) is only 8%. It
is therefore necessary to improve the efficicncy ofnitrogenjïxation of Ci. sepium in
this zone to assure Ihe sustuinubility of the uyroforesrry systetn.
Keywords Agroforestry, “N method, truc coefficient of utilization of N, LER,
nitrogen fixation
In spite of the interest which it arouses in regions of high population pressure
(Adegbehin and Igboanugo 1990), alley cropping is still far from being petfect. In
fact, even from simply the technical angle, only its implementation is still difkult.
When the soi1 fertility is low, applications of organic matter originating from hedges
Received 30 August 1999; accepted 10 February 2000.
Address correspondence to R. Oliver, CIRAD/AMIS. Programme Agronomie. TA 40/01-34398
Montpellier Cedex 5. France. E-mail: robert.oliver@cirad.fr
317
318
M. Ndiaye et al.
TABLE 1 Mean analytical properties for the experimental field
Depth
Component
Unit
O-20 cm
20-40 cm
Particle sizea
clay 1< 2 pm)
%
5.3
6.2
slit (2-20 um)
%
2.5
2.7
total sand (20-50 um)
%
92.2
91.1
Organic matter
Organic matter
%
0.47
0.38
Organic carbonb
%
0.27
0.22
Total nitrogen’
g kg-’
0.34
0.29
C/N ratio
8
8
Available P
Olsen-Dabin method
mg kg-’
21.4
19.7
Adsorbant complexd
cmol, k g - ’
Ca’ +
cmol, kg-’
1.12
0.93
Mg2 +
cmol, kg-’
0.21
0.18
K+
cmol, k g - ’
0.05
0.05
Na+
cmol, kg- l
0.03
0.04
Mn2+
cmol, kg-’
0.02
0.03
S(Ca, Mg, K Na)
cmol, kg- r
1.41
1.21
C E C
cmol, kg- ’
1.53
1.40
S/CE!C’
%
0.88
0.82
bF.5) water pH
6.7
6.6
(1/2.5) KCl molar pH
6.0
5.7
--
* Granulometric analysis by sedimentation, pipette method (AFNOR
1999a).
b Carbon by dry combustion method (AFNOR 1996a), organic
matter = 1.727 x C.
’ Total nitrogen by Kjeldhal method (AFNOR 1996b).
d Cobaltihexamonium chloride method (AFNOR Xx31-130, 1985;
Orsini Rémy, 1976).
’ CEC: eation exchange capacity.
’ S: sum of cations.
I pH (1/2.5 soil/water or RC1 1M) suspension (AFNOR 1999b).
of the usual height does not result in an improvement in fertility, and the system
remains dependent in most cases on applications of extra minera1 fertilizer (Ghuman
and La1 1990). Competition for light, minera1 nutrients and water cari be difficult to
manage because cultivation hampers the growth of the Woody species in the estab-
lishment phase. Later on, this situation is reversed, the growth of the plants near the
hedges being greatly reduced, and it is often necessary to do some trimming during
the course of trop growth (Jama, Gehutan, and Ngugi 1990). The role of a refuge
and source of spread for various pests (insects, rodents, birds, various spores fungi)
or, conversely, of a trap for certain fungal diseases should also not be overlooked
(Schroth 1994).
In Senegal, in certain conditions, the presence of trees in (Acacia albida Del.) or
around cultivated fields is a feature which the peasants try to preservc. However, for
about 20 years this association has evolved, with the developmcnt of plantings of
hedges of trees SUC~ as cashew, mango, eucalyptus, and Prosopis, between which are
grown cercals in rotation with groundnut (PRECOBA 1982). Reccnt experiments on
-,III-IY.
-m--I-
--
II)-I-“------
Maize-G. Sepium Alley Cropping in Senegal
319
8 maize lines (0.60 * 0,30)
. .,,
,, . __ Basis of Aliey cropping system
:,
:
total width: 9, 60 m
i i i i i
I
I
i
’ k’
.
.x...,. ...-...-...2 G. C;cp/um neigborough lines
(distance 1,80 in).
Maize in Alley cropping of G. xpium +
Total surface in cereal : 50% of total surface
Maize lines (0,60 * 0.30)
Pure trop of maize * Total surface in cereal : 100% of total surface
FIGURE 1 Schema of experimental design.
the introduction of leguminous tree species such as Gliricidia sepium (Jacq.) Steud.
and Senna siumea (Lam.) Irwin & Bam. (ISRA 1992) have revealed a considerable
potential for agroforestry. However, before introducing agroforestry systems, partic-
ularly in the semiarid zone, we should first consider the sustainability of such
systems and the details of their implementation in different environments (soil,
climate, population pressure) and with different farming systems (farmers’ objectives
and means, and socio-economic constraints). In this area, a11 sources of nitrogen for
fertilizer use are very precious, and the addition of nitrogen by planting trees is a
possibility in a context where the use of fertilizer, paiticularly nitrogen, is less and
less easy for the peasant, and where the nitrogen deficit of cropping systems is
aggravated by the abandonment of the groundnut-cereal rotation. This study is an
artempt to follow Young’s (1989) hypotheses in order to quantify the contribution of
nitrogen in the agroforestry system from a Woody N, fixing plant and the recycling
of nutrient elements, notably nitrogen. For this purpose we quantify, by the use of
1 “N fertilizer, the various potential sources of N : prunings, fertilizer, and soi1 pool.
Rlaterials and Methods
The study was carried out at the research station of Nioro du Rip (13” 45’ N-15” 47’
1:) in south-central Senegal. In this region, the annual mean temperature is 28”C,
annual rainfall is currently 650 mm with a short rainy season between June and
320
M. Ndiaye et al.
October, the daily evapotranspiration vary between 1 to 3 mm in the raining season
and 4 to 11 mm in the dry season. These conditions are representative of the
southern re:gjon of the Senegal groundnut growing area.
The soils of this area (Table 1) are, according to Bertrand (1971) slightly
leached ferruginous tropical soils (CPCS 1967) or Alfisols (USDA 1992) deveioped
on quaternary sands of fluvial and marine origin. These soils, which are very wide-
spread in the study zone, are deep, but also poor in fine fractions and depleted in
organic matter. Consequently, their fertility is low, and they are prone to acidifica-
tion. The available P (Olsen-Dabin method; Dabin 1967) content is acceptable in
view of the low clay content but there is a serious risk of K deficiency.
In the agroforestry system studied, maize is grown, between G. sepium hedges.
Before setting up the experiment, the uniformity of the experimental plot was
checked by analyzing samples of soi1 (taken from the O-20 and 20-40 cm depths
following a systematic grid pattcrn), SO that the results gave an indication of the
natural variability of the soil. An unfertilized maize “homogeneization trop” was
grown in the year before setting up the experiment, and harvestcd by quadrats
corresponding to the future experimental plots, to measure the natural yield varia-
tion of this txreal on plot.
Experimentd Design
The details of the experimental design are presented in Figure 1. Three parallel
strips corresponding to the three treatments tested were established: Pure Crop of
Maize (PCM), Alley-Cropped Maize (ACM), and Pure Crop of Gliricidia sepium
(PCG). Four independent “measuring plots” 4.50 m long were marked out within
each strip, thus providing four interna1 replications of the treatments. Seedlings of
G. sepium wr:re grown in a neighbouring nursery. In the PCM strip the maize was
.sown at 0.60 m between rows and 0.30 m within the row at 3 seeds per hill, and
thinned to one plant per hiil 15 days after sowing. The PCG strip was made up of
four hedges :spaced 1.80 m apart, with a spacing between trees within the hedge of
0.50 m. This strip, together with the strip set aside for the hedges of the ACM, was
not fertilized. In the ACM strip the paired hedges were spaced 1.80 m apart and
were estabhshed from plants taken from a nursery at the end of the rainy season
preceding the two years of the experiment. The distance between trees within a
hedge was 0.50 m. The hedges were trimmed at the time of sowing the maize to a
height of about 0.50 m (9 months after transplanting the Young trees) and only the
branches less than 10 cm circumference and the leaves were spread unifomlly on the
soil, the rest of the prunings being used for domestic purposes. This application
amounted to 3 Mg ha-’ of prunings with an average nitrogen content of 3.5%.
The areas used for maize’received phosphate fertilizer (13 kg P ha ’ as triple
superphosphate) and potassium (25 kg K ha - * as KCl) at sowing. Nitrogen fertil-
ization was done at the sowing time at a rate of 100 kg ha-’ of N (as ammonium
sulphate) for the PCM strip and at 20 kg ha ’ of N in the same form for the ACM
strip. This application of nitrogen fertilizer on the ACM strip was necessary
for the
use of “N experimentation, but it also acted as a minimum investment of nitrogen
starter recommended by Blonde1 (1971). Within each plot, this nitrogen application
was made over an area of 2.6 m2 in liquid form with a solution of (*‘NHJ2S0, at
1% and 5% ’ 5N atom excess respectively for PCM and ACM strips. This applica-
tion of ’ ’ N allows, thanks to the “A value” rnethod, calculating the contribution of
the different sources of N (soil, prunings, fertilizcr) to the maize nitrogen content.
The maizr was harvested by separating the stems + leaves, the grain, and the
rachis + spathes for the determination of dry matter production and N and “N
content. The results shown here are for two successive crops. After previous verifica-
tion of the field homogeneity (see results section), data are treated by a Xlstat pro-
Maize-G. Sepium AIley Cropping in Senegal
321
cedure (Famhy 1995) for the determination of means and standard deviation of the
replicates in the same strip.
Results and Discussion
Preliminary
Fertility Variability TriaI to Act as a Control for the Experiment
The thorough soi1 sampling carried out over a systematic grid pattern in the alleys
of the experiment enabled the variability in the physico-chemical characteristics of
the soi1 to be assessed at the time of establishment of the agroforestry test (Table 2).
The mean values observed for each of the strips may be regarded as identical if one
takes account of the fairly large standard error (intrastrip variability) observed for
each of the variables. This large intrastrip variability could be due to the Upper
slope position of the field and to the perpendicular direction of each strip to the
general slope. This physico-chemical analysis results are completed by the yields
obtained for homogenization trop (Table 3).
The yields observed on each of the strips are similar to one another; the
maximum errer of the means is 25% for the grain and less than 5% for the straw.
TABLE 2 Mean values of analytical parameters and standard errors of the mean,
( ), for each of the three strips of the experiment
-
-
Exch. K
CEC
Olsen Dabin
Total C
stlip
Cmol, kg- l
Cmol, kg-’
0
‘“(Y’
0
mg P kg-’
Y”
ACMt
0.0~0.02)
1.63 (0.02)
7.8 (0.3)
8.9 (1.9)
0.29 (0.02)
PCMS
0.06 (0.03)
1.59 (0.46)
8.4 (0.5)
11.2 (5.0)
0.26 (0.03)
PCC$
0.05 (0.03)
1.19 (0.24)
7.8 (0.4)
7.8 (2.4)
0.29 (0.02)
_I_-
* (A + L): particles of siz,e ~20 ~III.
t ACM : Alley cropped maize.
$ PCM: Pure trop of maize.
$ PCG: Pure trop of Glyiricidia.
TABLE 3 Spatial variation in maize yield for the homogenization trop. Mean
values and standard errors of means, ( ). The strips are identified by the
abbreviation for the treatment applied to them during the experiment
I_-
-~.--
Total experimental
PCM*
ACMT
PCC$
arean= 12
n=4
n=4
n=4
--~
Components
Yield (kg ha-‘)
-
-
-
-
~--..--..-
_---.-~
Grain
759 (81)
634 (84)
667 (76)
603 (62)
Straw
1400 (77)
1427 (91)
1368 (53)
1.393 (62)
Total
2159 (141)
2223 (171)
2076 (96)
2 146 (89)
-
-
-
-
* ACM : Alley cropped maize.
t PCM : Pure trop of maize.
1 PCC : Pure trop of Glyriridla.
n = number of plots (8.4 m x 4.5 m) harvested for each strip (PCM, ACM and ACG).
-.-
.,-- -_._
_____~ ._~
Maize-G. Sepium Alley Cropping in Senegal
323
Comparîson
qf Effectiveness of Prunings and Nitrogen Fertilizer
The concept of the “A value” (Fried and Dean 1952) enables one to quantify the
respective contribution of fertilizer (Ndff), prunings (Ndfp), and soi1 (Ndfs) in the
total N of the maize:
A kg cquivnlcn! fcrtiliror ha-1
where N, is amount of N fertilizer in kg ha -I, E, is the isotopic excess of used
fertilizer, and Ep is the isotopic excess of cultivated plant. Notice that the “A value,”
expressed in equivalents of ammonium sulphate nitrogen (the fertilizer used in the
study), is an essential link in the calculation chain for Ndff and Ndfp, if it does not
measure the size of the source compartment it provides a good representation of it.
Table 5 allows the comparison of “A values” for the soi1 (A,), the soi1 + prunings
(A, + J, and the prunings alone (A,).
The TCU (‘Truc Coefficient of Utilization’) of the fertilizer in pure stands is of
the order of 3O%, and is in agreement with the values normally found (25-45%) for
the maize (Ganry 1990) and with those in the literature for the semiarid zone of
West Africa (Pieri 1989). These values are low because considerable losses (Cissé
1984) of nitrogen in form of nitrate may occur.
The Ndfp is 36% in the first year and 27% in the second. It is close to the
Ndff% for similar amounts of N applied, however, it is less than that of fertilizer
because the yield of the aerial parts is lower. The TCU, of the prunings (TCU,)
(Table 6) are 25% and 16% respectively,
values which are lower than for fertilizer
(30%). This last result is explained by the fact that the prunings contain a non-
mineralizable part estimated at 20% (Ndiaye 1997).
The “A value” provides a quantitative comparison between minera1 fertilizer
and prunings in terms of yield and nitrogen (Zapata .1990). In terms of yield, 1 kg of
ammonium sulphate fertilizer is equivalent to 4.8 kg dry matter (DM) (‘year ‘1) and
8.0 kg DM (year 2) of prunings (i.e., a mean of 6.4 kg DM), and in terms of N, 1 kg
N as ammonium sulphate is equivalent to 0.8 kg N (year 1) and to 1.3 kg N (year 2)
of prunings (i.e., a mean of 1.0 kg N). Thus in terms of nitrogen, in the field, the
effciency of Ghicidia sepium prunings is identical to that of ammonium sulphate.
Contribution O~N, Fixation by G. sepium, via its Prunings, to the Nitrogen Nutrition
of Maize
The percentage: of nitrogen fixed (Ndflïx %) by G. sepjum is about 25% of total N
plant content at Nioro (Ndiaye and Ganry 1997). On the basis of this data,. the
TABLE 6 Ndlf and TCU of fertilizer and prunings
-~-
--~~
--1-
Fertilizer
Pruning
(ammonium sulphate)
(Gliricidia sepium)
Ndff*
NdfpS
TW t
T C U ,
Year
%
kg ha-’
%
%
kn ha-’
%
First year
31
31
31
36
26
25
S e c o n d year
33
28
28
2-l
17
16
~-
-
-
-
* Ndff: nitrogen derived l’rom fertilizer.
7 T C U : T r u e coefficient of utilisation of nitrogen (TCU,) from fertiiizer (TCU,) from
pruning.
$ Ndfp: nitrogen derived from emonds.
>- _l._.______I
- __...__ .-_-
--
----
----“.“.--- m-m-*
324
M. Ndiaye et al.
Ndf fixation Hypothesis: 25% -* maize N derived fiom fixation : 7.7%
Ndfs %
63%
6 %
Ndf fixation Hypothesis: 75% + : maize N derived fiom fixation 23.2%
/- /--
Ndfs %
Nfdp %
63%
31%
-\\
N d f f % - - -
6%
FIGLYRIi: 2 Nitrogen fixation contribution at nitrogen maize supply according to
hypothesis of nitrogen fixation efficiency.
proportion of fixed N, contained in the prunings cari be estimated, and, consequent-
ly, SO cari its contribution to the nitrogen nutrition of the maize (Figure 2). This
contribution is in the order of 7.5%, and is thus quite small in the Nioro soil.
However, we have also shown (Ndiaye 1997) that the fixation of N, cari reach 75%
in the presence of effrcient specific strains of Bradyrhizobium, in a more favourable
environment such as southem Senegal. In these conditions, the contribution of
nitrogen from the prunings of G. sepium coming from N, nitrogen fixation cari reach
20-25X of the total N in the maize, which would be very significant for the intro-
duction of a sustainable agroforestry system.
Estimation of the Land Equivalent Ratio (LER)
By considering the production of the cereal and the production of the tree at the
same time in the maize-tree association, we cari calculate the Land Equivalent Ratio
(LER) which is an indicator of the eflïciency of the intercropping system. The results
(Table 7) are based on total DM yields (without roots). The maize pure trop
without N (KM) is distinguished from that given N (PCM + Nj’, the values for the
latter coming from a plot grown nearby outside the experiment.
Pure Crop of Maize with N
The mean relative DM of maize in ACM system as a PCM with N is 0.44 and that
of Glirkidia sepiurn is 0.46, giving a LER of 0.90. The biological elkiency of land use
----
_.-
-.----
TABLE 7 DM yields of G. sepium and of maize and the LER calculated from the total vield (field exwriment; years 1994 and 1995)
G. sepium
Maize
ISE
PCM*
ACMt
WI
PCM+N
CCM
PCM
Lm, **
Lmnrtt
Year
Total DM$ kg ha-’
Total DM kg ha-’
LERrff
LE%§?3
First year
6198
2729
0.44
5585
2573
-
0.46
-
0.91
-
Second year
6121
2898
0.47
5606
2388
2154
0.43
1.11
0.90
1.58
Mean
6159
2814
0.46
5595
2480
-
0.44
-
0.90
-
* PCM : Pure trop of maize.
t ACM : Alley cropped maize.
$ PCG: Pure trop of gliricidia.
$ Total DM = total dry matter; G. sepium (prunings + branches), maize (grain + straw).
11
Lg = relative yield of G. sepium.
** Lm, = relative yield of maize (in pure stand of maize with fertihzer).
tt J-mnr = relative yield of maize (in pure stand of maize without fertilizer).
$$ LER, = LER calculated using yields of maize grown in pure stand with fertihzer.
@ LER,, = LER calculated using yields of maize grown in pure stand without fertilizer.
326
M. Ndiaye et al.
is almost as good in an agroforestry system without fertilizer as in pure crops with a
large dose of fertilizer (100 kg N ha- ‘).
Pure Trop of Maize without N
In a tria1 carried out on the same site in 1995, the yield of maize without N was 511
and 1643 kg ha-’ for grain and straw respectively. The relative yield is thus 1.11
and that of Giiricidia is, for that year, 0.47, making a LER of 1.58. This result shows
that the biological eflïciency of land use is much greater with an agroforestry system
than with a traditional cropping system: an increase in dry matter production of
58% is found by comparison with pure stands. This advantage is considerable in
region:; of high population density such as the groundnut growing region of Senegal,
where land is a scarce resource.
Conclusion
The prunings have a marked effect that cari be compared to that of minera1 nitrogen
fertilizer on the yield and total N content of the maize, which is explained, at least
partly, by the increase in nitrogen absorption by maize from the soil N pool
enriched by the prunings applied. These effects on yield and on plant total N are in
agreement with the nutritional properties of these prunings, already demonstrated
by’our work (Ndiaye 1997). The prunings supply 30-35% of the total N in the
maize, values close to those obtained for fertilizer for a similar amount of applied
nitrogen; the value of the N, fixing tree is to contribute “free” to this percentage,
and to the soi1 N pool. Basing the N, fixation at 25% in G. sepium as obtained in
the environment of central Senegal (Ndiaye and Ganry 1997), the contribution of
this N, fixation to the total N in mahe is 8%. Hence it is not much; we suggest that
the reason lies.in the limited fixing ability of the rhizobial strains. If this constraint
could be removed, the fixed contribution could reach 20-25% in maize if the N,
fixation of G. sepium reached the value of 75% in south Senegal (Ndiaye 1997). This
value of 2O-25% is within the range of values found by numerous authors notably
Kang, Wilson, and Lawson (1984) in a more humid climate. They showed, more-
over, that this benefit increased progressively because of the increase in productivity
and the cumulative effect of the prunings.
Finally, the agroforestry system, with a LRE of 1.58, is a system which has a
biological land use eflïciency greater than that of the sum of the two individual
crops each tested as pure stands. This advantage is vital in areas where the soi1 is a
scarce resource as in the groundnut zone of Senegal. During the rainy season, the
agroforestry system would also offer the possibility of growing cow pea (Yigna
unguiculata
L.) in relay cropping before maize harvest between the G. sepiurn hedges,
which could be an additional source of income for the farmer. This trop, together
with the wood production, would make up for the “10s~” of cultivable land.
References
Adegbehin, J. O., and A. B. 1. Igboanugo. 1990. Agroforestry practiçes in Nigeria. Ayrcl-
joresfry Systems 10: l-22.
AFNOR. 1996a. Dosage du carbone organique et du carbone total après combustion sèche.
(Association Française de Normalisation) NF ISO 106 94, pp. 189-200 in AFNOR,
Qualité des sols. Paris, France.
AFNOR. 1996b. Dosage de l’azote total-Méthode de Kjeldahl modifiée. (Association
Française de Normalisation) NF ISO 1 l-261, pp. 257-264, in AFNOR, Qualité des sols.
Paris, France.
AFNOR. 1999a. Analyse granulométrique par sédimentation. Méthode de la pipette.
(Association Française de Normalisation) NF X 31-107, pp. 57-72 in AFNOR, Qualifi
des sols. Paris, France.
. -
Maize-G. Sepium Alley Croppimg in Senegal
327
AFNOR. 1999b. Détermination du pH. (Association Française de Normalisauon) NF ISC)
103 90, pp- 339-34X in AFNOR, Qualité des sols. Paris, France.
Bertrand, R. 1972. Morphopédologie et orientations culturales des régions soudaniennes du
Siné !%oum (Sénégal). L’Agronomie. Tropicale, 27: 1115-l 190.
Blondel, 1). 1971. Contribution a la connaissance de la dynamique de I’azole minéral ensoi
sableux (Dior) au Sénégal. Agronomie Tropicale 26: 1303-1333.
Cissé, L. 1984. Estimation des pertes moyennes en eau et en éléments fertilisants dans les
systèmes de cultures arachide-mil de la zone soudano-sahélienne du Sénégal. Réunion de
synthise de la Division Conjointe FAO/AIEA sur “l’emploi des radioisotopes et des
rayonnements dans les études pour une utilisation rationnelle de l’eau et des engrais dans
les zones semi-arides”, Vienne 6-11 Février 1984, Document Interne. Institut Sénégalais
de Recherches Agricoles. Dakar, Senegal.
Commission de Pédologie et de Cartographie des Sols (CPCS), 1967. Classification des sols---
Edition 1967. 96 pages.
Dabin, B. 1967. Sur une méthode d’analyse du phosphore dans les sols tropicaux., I., pp.
99-l 15 in Colloque sur iafertilité des sols Tropicaux. Tananarive, Madagascar.
Fahmy, T. 1995. XLSTAT, Software add-in Microsoft Excel. http://www.xlstat.com.
Fried, M., and L. A. Dean. 1952. A concept concerning the measurement of available soi1
nutrients. Soi1 Science 73 : 263-214.
Ganry, F. 1990. Application de la méthode isotopique à l’étude des bilans azotés en zone
tropicale sèche. Thèse de doctorat d’Etat, Université de Nancy 1. Nancy, France.
Ghuman, B. S., and R. Lai. 1990. Nutrient addition into soi1 by leaves of Cassio siamea and
Gliricidia sepium grown on an ultisol in Southern Nigeria. Agroforestry Systems 10: 131~-
133.
ISRA. 1992. Rapport Technique, Direction de Recherches sur les Productions Forestières. Insti-
tut Sénégalais de Recherches Agricoles, Dakar, Senegal.
Jama, B., A. Gehutan., and D. N. Ngugi. 1990. Shading effects of alley cropped Leucaena
leucocephola on weed biomass and maize yield at Mtwapa, Coast Province, Kenya.
Agroforeswy Systems 13: l-l 1.
Kang, B. T., (3. F. Wilson, and T. L. Lawson. 1984. Alley cropping sequentially cropped
maize and cowpea with Leucaena on a sandy soi1 in southern Nigeria. Plant und Soi/
85: 267-277.
Ndiaye, M. 1997. Contribution des légumineuses arbustives à l’alimentation azotée du maïs
(Zea mays L.k-Cas d’un système de culture en allées dans le Centre Sud Sénégal, Thèse
de doctorat de I’INPL, Nancy, France.
Ndiaye, M.. and F. Ganry. 1997. Variation in the biological N, fixation by tree legumes in
three ecological zones from the north to the south of Senegal. Arid Soi1 Rcsearch ~lnd
Rrhabi!itation 11:245- 254.
Orsini, L., and J. C. Rémy. 1976. Utilisation du chlorure de cobaltihexammine pour la déter-
mination simultanée de la capacité d’échange et des bases échangeables des sols. Hullcrin
de I’AFES 1976:269-216.
Pieri, C . 1989. Farti(ité des terres de savanes: bilan de trente uns de rechtzrchc, ct de,
dk~eloppement (IiIricole nu Sud du Sahara. Paris, Ministère de la Coopération et du Dével-
oppement et CIRAD-IRAT, France.
PRECOBA. 1982. Etude sociologique: la motivation et les actions des populations à l’égard
de l’arbre. Rappor: Technique No. 1 Projet de Reboisements C~~rnr?lut~nut~~~rtl.s riuns /C
Bassin ,4rachidier du S>négul. ISRA, Dakar, Senegal.
Schroth, G. 1994. Above and below-ground interactions in alley cropping with G‘ilricidia
sepium as Icompared to conventional and mulched sole cropping on a high base status
soi1 in the African rainforest zone. Ph.D. dissertation, University of Bayreuth, Ciermany.
Soil Survey Staff. 1992. Keys to Soi1 Taxonomy. SMSS Technical Monograph No. 19. Fifth
Edition Pocahontas Press, Inc., Blacksburg, Virginia. 556 pages.
Young, A. 1989. Dix hypothèses pour la recherche sols/agroforesterie I’/l~/ro~~~r~,sf~ri~~
Aujourd’hui 1 : 13-16.
Zapata, F. 1990. Isotope techniques in soi1 fertility and plant nutrition studies In U.SL> o/
Nuclear T,chrtiques i n Studios oJ Soi/-Plunt R~,l<ltiollsllip,s.
7‘rrrininy Cour.~ .sr’r~~.s PRO. 2,
edited by Ci llardarson (pp. 61-127) International Atomic Energy Agency, Vienna,
Austria.
- - I I - - -
-.
k .
Copytighl 0 2ooO Taylor & Francis
089@3069/W 512.00 + .oO
Alley Cropping of Maize and Gliricidia sepiunz in the
Sudanese Sahel Region : Some Technical
Feasibility Aspects
MAMADOU NDIAYE
Institut Sénégalais de Recherches Agricoles/Centre National De Recherche
Agronomiques (ISRA/CNRA)
Bambey Senegal
FRANCIS GANRY
ROBERT OLIVER
Centre International de Recherches Agricoles en Coopération/Département
Amélioration des Méthodes pour L-Innovation Scientifique
(CIRAD/AMIS)
Programme Agronomie,
Montpellier, France
An association in un ulley cropping experirnent oja short-rerm maize uuriety and a
tree legume (Gliricidia sepium) adapted to the Sudanese Sahel region wus studied
from the point of view of the nitrogen balance and plot yields. Isotopic lubelling.
applied in thefield enabled the contribution io maize nitrogen nutrition from d$k-
ent nitrogen sources (ferrilizer, prunings, and soii) to be quantijîed. For equul Cu!ti-
vated areus, ulley cropping provides a maize yield greater thun thut of muize grown
in pure stand without nirrogen fertifizer. However this yield is only 40% of thut
obtained in pure stands with fertilizer nitrogen. Alley cropping gives u very favour-
able Lund Equivalent Ratio (LER) of 0.90 and 2.58 compared with N-fertilized and
non-N-fertilized plots, respecrively. These LEKs demonstrate the increased bio-
logical eficiency of the cultivated soi1 in un ugrojïorestry system. The percentage of
che total nitrogen in rhe muize coming from prunings (Nfdp) varies between 30 and
35% and the true coeficient of nitrogen urilization of the prunings (TCUp) varies
from 15-25%. In the environment of central Spnegal, the percentage of total nitro-
gen of G. sepium cominy jrorn N, jxatiorl is yuifp low (NQjjx = 25X), and conse-
quentiy, in the maize, the nitrogen cominy .from N, fixation (Ndfix) is only 8%. It
is therefore necessary to improve the efficicncy ofnitrogenjïxation of Ci. sepium in
this zone to assure Ihe sustuinubility of the uyroforesrry systetn.
Keywords Agroforestry, “N method, truc coefficient of utilization of N, LER,
nitrogen fixation
In spite of the interest which it arouses in regions of high population pressure
(Adegbehin and Igboanugo 1990), alley cropping is still far from being petfect. In
fact, even from simply the technical angle, only its implementation is still difkult.
When the soi1 fertility is low, applications of organic matter originating from hedges
Received 30 August 1999; accepted 10 February 2000.
Address correspondence to R. Oliver, CIRAD/AMIS. Programme Agronomie. TA 40/01-34398
Montpellier Cedex 5. France. E-mail: robert.oliver@cirad.fr
317
318
M. Ndiaye et al.
TABLE 1 Mean analytical properties for the experimental field
Depth
Component
Unit
O-20 cm
20-40 cm
Particle sizea
clay 1< 2 pm)
%
5.3
6.2
slit (2-20 um)
%
2.5
2.7
total sand (20-50 um)
%
92.2
91.1
Organic matter
Organic matter
%
0.47
0.38
Organic carbonb
%
0.27
0.22
Total nitrogen’
g kg-’
0.34
0.29
C/N ratio
8
8
Available P
Olsen-Dabin method
mg kg-’
21.4
19.7
Adsorbant complexd
cmol, k g - ’
Ca’ +
cmol, kg-’
1.12
0.93
Mg2 +
cmol, kg-’
0.21
0.18
K+
cmol, k g - ’
0.05
0.05
Na+
cmol, kg- l
0.03
0.04
Mn2+
cmol, kg-’
0.02
0.03
S(Ca, Mg, K Na)
cmol, kg- r
1.41
1.21
C E C
cmol, kg- ’
1.53
1.40
S/CE!C’
%
0.88
0.82
bF.5) water pH
6.7
6.6
(1/2.5) KCl molar pH
6.0
5.7
--
* Granulometric analysis by sedimentation, pipette method (AFNOR
1999a).
b Carbon by dry combustion method (AFNOR 1996a), organic
matter = 1.727 x C.
’ Total nitrogen by Kjeldhal method (AFNOR 1996b).
d Cobaltihexamonium chloride method (AFNOR Xx31-130, 1985;
Orsini Rémy, 1976).
’ CEC: eation exchange capacity.
’ S: sum of cations.
I pH (1/2.5 soil/water or RC1 1M) suspension (AFNOR 1999b).
of the usual height does not result in an improvement in fertility, and the system
remains dependent in most cases on applications of extra minera1 fertilizer (Ghuman
and La1 1990). Competition for light, minera1 nutrients and water cari be difficult to
manage because cultivation hampers the growth of the Woody species in the estab-
lishment phase. Later on, this situation is reversed, the growth of the plants near the
hedges being greatly reduced, and it is often necessary to do some trimming during
the course of trop growth (Jama, Gehutan, and Ngugi 1990). The role of a refuge
and source of spread for various pests (insects, rodents, birds, various spores fungi)
or, conversely, of a trap for certain fungal diseases should also not be overlooked
(Schroth 1994).
In Senegal, in certain conditions, the presence of trees in (Acacia albida Del.) or
around cultivated fields is a feature which the peasants try to preservc. However, for
about 20 years this association has evolved, with the developmcnt of plantings of
hedges of trees SUC~ as cashew, mango, eucalyptus, and Prosopis, between which are
grown cercals in rotation with groundnut (PRECOBA 1982). Reccnt experiments on
-,III-IY.
-m--I-
--
II)-I-“------
Maize-G. Sepium Alley Cropping in Senegal
319
8 maize lines (0.60 * 0,30)
. .,,
,, . __ Basis of Aliey cropping system
:,
:
total width: 9, 60 m
i i i i i
I
I
i
’ k’
.
.x...,. ...-...-...2 G. C;cp/um neigborough lines
(distance 1,80 in).
Maize in Alley cropping of G. xpium +
Total surface in cereal : 50% of total surface
Maize lines (0,60 * 0.30)
Pure trop of maize * Total surface in cereal : 100% of total surface
FIGURE 1 Schema of experimental design.
the introduction of leguminous tree species such as Gliricidia sepium (Jacq.) Steud.
and Senna siumea (Lam.) Irwin & Bam. (ISRA 1992) have revealed a considerable
potential for agroforestry. However, before introducing agroforestry systems, partic-
ularly in the semiarid zone, we should first consider the sustainability of such
systems and the details of their implementation in different environments (soil,
climate, population pressure) and with different farming systems (farmers’ objectives
and means, and socio-economic constraints). In this area, a11 sources of nitrogen for
fertilizer use are very precious, and the addition of nitrogen by planting trees is a
possibility in a context where the use of fertilizer, paiticularly nitrogen, is less and
less easy for the peasant, and where the nitrogen deficit of cropping systems is
aggravated by the abandonment of the groundnut-cereal rotation. This study is an
artempt to follow Young’s (1989) hypotheses in order to quantify the contribution of
nitrogen in the agroforestry system from a Woody N, fixing plant and the recycling
of nutrient elements, notably nitrogen. For this purpose we quantify, by the use of
1 “N fertilizer, the various potential sources of N : prunings, fertilizer, and soi1 pool.
Rlaterials and Methods
The study was carried out at the research station of Nioro du Rip (13” 45’ N-15” 47’
1:) in south-central Senegal. In this region, the annual mean temperature is 28”C,
annual rainfall is currently 650 mm with a short rainy season between June and
320
M. Ndiaye et al.
October, the daily evapotranspiration vary between 1 to 3 mm in the raining season
and 4 to 11 mm in the dry season. These conditions are representative of the
southern re:gjon of the Senegal groundnut growing area.
The soils of this area (Table 1) are, according to Bertrand (1971) slightly
leached ferruginous tropical soils (CPCS 1967) or Alfisols (USDA 1992) deveioped
on quaternary sands of fluvial and marine origin. These soils, which are very wide-
spread in the study zone, are deep, but also poor in fine fractions and depleted in
organic matter. Consequently, their fertility is low, and they are prone to acidifica-
tion. The available P (Olsen-Dabin method; Dabin 1967) content is acceptable in
view of the low clay content but there is a serious risk of K deficiency.
In the agroforestry system studied, maize is grown, between G. sepium hedges.
Before setting up the experiment, the uniformity of the experimental plot was
checked by analyzing samples of soi1 (taken from the O-20 and 20-40 cm depths
following a systematic grid pattcrn), SO that the results gave an indication of the
natural variability of the soil. An unfertilized maize “homogeneization trop” was
grown in the year before setting up the experiment, and harvestcd by quadrats
corresponding to the future experimental plots, to measure the natural yield varia-
tion of this txreal on plot.
Experimentd Design
The details of the experimental design are presented in Figure 1. Three parallel
strips corresponding to the three treatments tested were established: Pure Crop of
Maize (PCM), Alley-Cropped Maize (ACM), and Pure Crop of Gliricidia sepium
(PCG). Four independent “measuring plots” 4.50 m long were marked out within
each strip, thus providing four interna1 replications of the treatments. Seedlings of
G. sepium wr:re grown in a neighbouring nursery. In the PCM strip the maize was
.sown at 0.60 m between rows and 0.30 m within the row at 3 seeds per hill, and
thinned to one plant per hiil 15 days after sowing. The PCG strip was made up of
four hedges :spaced 1.80 m apart, with a spacing between trees within the hedge of
0.50 m. This strip, together with the strip set aside for the hedges of the ACM, was
not fertilized. In the ACM strip the paired hedges were spaced 1.80 m apart and
were estabhshed from plants taken from a nursery at the end of the rainy season
preceding the two years of the experiment. The distance between trees within a
hedge was 0.50 m. The hedges were trimmed at the time of sowing the maize to a
height of about 0.50 m (9 months after transplanting the Young trees) and only the
branches less than 10 cm circumference and the leaves were spread unifomlly on the
soil, the rest of the prunings being used for domestic purposes. This application
amounted to 3 Mg ha-’ of prunings with an average nitrogen content of 3.5%.
The areas used for maize’received phosphate fertilizer (13 kg P ha ’ as triple
superphosphate) and potassium (25 kg K ha - * as KCl) at sowing. Nitrogen fertil-
ization was done at the sowing time at a rate of 100 kg ha-’ of N (as ammonium
sulphate) for the PCM strip and at 20 kg ha ’ of N in the same form for the ACM
strip. This application of nitrogen fertilizer on the ACM strip was necessary
for the
use of “N experimentation, but it also acted as a minimum investment of nitrogen
starter recommended by Blonde1 (1971). Within each plot, this nitrogen application
was made over an area of 2.6 m2 in liquid form with a solution of (*‘NHJ2S0, at
1% and 5% ’ 5N atom excess respectively for PCM and ACM strips. This applica-
tion of ’ ’ N allows, thanks to the “A value” rnethod, calculating the contribution of
the different sources of N (soil, prunings, fertilizcr) to the maize nitrogen content.
The maizr was harvested by separating the stems + leaves, the grain, and the
rachis + spathes for the determination of dry matter production and N and “N
content. The results shown here are for two successive crops. After previous verifica-
tion of the field homogeneity (see results section), data are treated by a Xlstat pro-
Maize-G. Sepium AIley Cropping in Senegal
321
cedure (Famhy 1995) for the determination of means and standard deviation of the
replicates in the same strip.
Results and Discussion
Preliminary
Fertility Variability TriaI to Act as a Control for the Experiment
The thorough soi1 sampling carried out over a systematic grid pattern in the alleys
of the experiment enabled the variability in the physico-chemical characteristics of
the soi1 to be assessed at the time of establishment of the agroforestry test (Table 2).
The mean values observed for each of the strips may be regarded as identical if one
takes account of the fairly large standard error (intrastrip variability) observed for
each of the variables. This large intrastrip variability could be due to the Upper
slope position of the field and to the perpendicular direction of each strip to the
general slope. This physico-chemical analysis results are completed by the yields
obtained for homogenization trop (Table 3).
The yields observed on each of the strips are similar to one another; the
maximum errer of the means is 25% for the grain and less than 5% for the straw.
TABLE 2 Mean values of analytical parameters and standard errors of the mean,
( ), for each of the three strips of the experiment
-
-
Exch. K
CEC
Olsen Dabin
Total C
stlip
Cmol, kg- l
Cmol, kg-’
0
‘“(Y’
0
mg P kg-’
Y”
ACMt
0.0~0.02)
1.63 (0.02)
7.8 (0.3)
8.9 (1.9)
0.29 (0.02)
PCMS
0.06 (0.03)
1.59 (0.46)
8.4 (0.5)
11.2 (5.0)
0.26 (0.03)
PCC$
0.05 (0.03)
1.19 (0.24)
7.8 (0.4)
7.8 (2.4)
0.29 (0.02)
_I_-
* (A + L): particles of siz,e ~20 ~III.
t ACM : Alley cropped maize.
$ PCM: Pure trop of maize.
$ PCG: Pure trop of Glyiricidia.
TABLE 3 Spatial variation in maize yield for the homogenization trop. Mean
values and standard errors of means, ( ). The strips are identified by the
abbreviation for the treatment applied to them during the experiment
I_-
-~.--
Total experimental
PCM*
ACMT
PCC$
arean= 12
n=4
n=4
n=4
--~
Components
Yield (kg ha-‘)
-
-
-
-
~--..--..-
_---.-~
Grain
759 (81)
634 (84)
667 (76)
603 (62)
Straw
1400 (77)
1427 (91)
1368 (53)
1.393 (62)
Total
2159 (141)
2223 (171)
2076 (96)
2 146 (89)
-
-
-
-
* ACM : Alley cropped maize.
t PCM : Pure trop of maize.
1 PCC : Pure trop of Glyriridla.
n = number of plots (8.4 m x 4.5 m) harvested for each strip (PCM, ACM and ACG).
-.-
.,-- -_._
_____~ ._~
Maize-G. Sepium Alley Cropping in Senegal
323
Comparîson
qf Effectiveness of Prunings and Nitrogen Fertilizer
The concept of the “A value” (Fried and Dean 1952) enables one to quantify the
respective contribution of fertilizer (Ndff), prunings (Ndfp), and soi1 (Ndfs) in the
total N of the maize:
A kg cquivnlcn! fcrtiliror ha-1
where N, is amount of N fertilizer in kg ha -I, E, is the isotopic excess of used
fertilizer, and Ep is the isotopic excess of cultivated plant. Notice that the “A value,”
expressed in equivalents of ammonium sulphate nitrogen (the fertilizer used in the
study), is an essential link in the calculation chain for Ndff and Ndfp, if it does not
measure the size of the source compartment it provides a good representation of it.
Table 5 allows the comparison of “A values” for the soi1 (A,), the soi1 + prunings
(A, + J, and the prunings alone (A,).
The TCU (‘Truc Coefficient of Utilization’) of the fertilizer in pure stands is of
the order of 3O%, and is in agreement with the values normally found (25-45%) for
the maize (Ganry 1990) and with those in the literature for the semiarid zone of
West Africa (Pieri 1989). These values are low because considerable losses (Cissé
1984) of nitrogen in form of nitrate may occur.
The Ndfp is 36% in the first year and 27% in the second. It is close to the
Ndff% for similar amounts of N applied, however, it is less than that of fertilizer
because the yield of the aerial parts is lower. The TCU, of the prunings (TCU,)
(Table 6) are 25% and 16% respectively,
values which are lower than for fertilizer
(30%). This last result is explained by the fact that the prunings contain a non-
mineralizable part estimated at 20% (Ndiaye 1997).
The “A value” provides a quantitative comparison between minera1 fertilizer
and prunings in terms of yield and nitrogen (Zapata .1990). In terms of yield, 1 kg of
ammonium sulphate fertilizer is equivalent to 4.8 kg dry matter (DM) (‘year ‘1) and
8.0 kg DM (year 2) of prunings (i.e., a mean of 6.4 kg DM), and in terms of N, 1 kg
N as ammonium sulphate is equivalent to 0.8 kg N (year 1) and to 1.3 kg N (year 2)
of prunings (i.e., a mean of 1.0 kg N). Thus in terms of nitrogen, in the field, the
effciency of Ghicidia sepium prunings is identical to that of ammonium sulphate.
Contribution O~N, Fixation by G. sepium, via its Prunings, to the Nitrogen Nutrition
of Maize
The percentage: of nitrogen fixed (Ndflïx %) by G. sepjum is about 25% of total N
plant content at Nioro (Ndiaye and Ganry 1997). On the basis of this data,. the
TABLE 6 Ndlf and TCU of fertilizer and prunings
-~-
--~~
--1-
Fertilizer
Pruning
(ammonium sulphate)
(Gliricidia sepium)
Ndff*
NdfpS
TW t
T C U ,
Year
%
kg ha-’
%
%
kn ha-’
%
First year
31
31
31
36
26
25
S e c o n d year
33
28
28
2-l
17
16
~-
-
-
-
* Ndff: nitrogen derived l’rom fertilizer.
7 T C U : T r u e coefficient of utilisation of nitrogen (TCU,) from fertiiizer (TCU,) from
pruning.
$ Ndfp: nitrogen derived from emonds.
>- _l._.______I
- __...__ .-_-
--
----
----“.“.--- m-m-*
324
M. Ndiaye et al.
Ndf fixation Hypothesis: 25% -* maize N derived fiom fixation : 7.7%
Ndfs %
63%
6 %
Ndf fixation Hypothesis: 75% + : maize N derived fiom fixation 23.2%
/- /--
Ndfs %
Nfdp %
63%
31%
-\\
N d f f % - - -
6%
FIGLYRIi: 2 Nitrogen fixation contribution at nitrogen maize supply according to
hypothesis of nitrogen fixation efficiency.
proportion of fixed N, contained in the prunings cari be estimated, and, consequent-
ly, SO cari its contribution to the nitrogen nutrition of the maize (Figure 2). This
contribution is in the order of 7.5%, and is thus quite small in the Nioro soil.
However, we have also shown (Ndiaye 1997) that the fixation of N, cari reach 75%
in the presence of effrcient specific strains of Bradyrhizobium, in a more favourable
environment such as southem Senegal. In these conditions, the contribution of
nitrogen from the prunings of G. sepium coming from N, nitrogen fixation cari reach
20-25X of the total N in the maize, which would be very significant for the intro-
duction of a sustainable agroforestry system.
Estimation of the Land Equivalent Ratio (LER)
By considering the production of the cereal and the production of the tree at the
same time in the maize-tree association, we cari calculate the Land Equivalent Ratio
(LER) which is an indicator of the eflïciency of the intercropping system. The results
(Table 7) are based on total DM yields (without roots). The maize pure trop
without N (KM) is distinguished from that given N (PCM + Nj’, the values for the
latter coming from a plot grown nearby outside the experiment.
Pure Crop of Maize with N
The mean relative DM of maize in ACM system as a PCM with N is 0.44 and that
of Glirkidia sepiurn is 0.46, giving a LER of 0.90. The biological elkiency of land use
----
_.-
-.----
TABLE 7 DM yields of G. sepium and of maize and the LER calculated from the total vield (field exwriment; years 1994 and 1995)
G. sepium
Maize
ISE
PCM*
ACMt
WI
PCM+N
CCM
PCM
Lm, **
Lmnrtt
Year
Total DM$ kg ha-’
Total DM kg ha-’
LERrff
LE%§?3
First year
6198
2729
0.44
5585
2573
-
0.46
-
0.91
-
Second year
6121
2898
0.47
5606
2388
2154
0.43
1.11
0.90
1.58
Mean
6159
2814
0.46
5595
2480
-
0.44
-
0.90
-
* PCM : Pure trop of maize.
t ACM : Alley cropped maize.
$ PCG: Pure trop of gliricidia.
$ Total DM = total dry matter; G. sepium (prunings + branches), maize (grain + straw).
11
Lg = relative yield of G. sepium.
** Lm, = relative yield of maize (in pure stand of maize with fertihzer).
tt J-mnr = relative yield of maize (in pure stand of maize without fertilizer).
$$ LER, = LER calculated using yields of maize grown in pure stand with fertihzer.
@ LER,, = LER calculated using yields of maize grown in pure stand without fertilizer.
326
M. Ndiaye et al.
is almost as good in an agroforestry system without fertilizer as in pure crops with a
large dose of fertilizer (100 kg N ha- ‘).
Pure Trop of Maize without N
In a tria1 carried out on the same site in 1995, the yield of maize without N was 511
and 1643 kg ha-’ for grain and straw respectively. The relative yield is thus 1.11
and that of Giiricidia is, for that year, 0.47, making a LER of 1.58. This result shows
that the biological eflïciency of land use is much greater with an agroforestry system
than with a traditional cropping system: an increase in dry matter production of
58% is found by comparison with pure stands. This advantage is considerable in
region:; of high population density such as the groundnut growing region of Senegal,
where land is a scarce resource.
Conclusion
The prunings have a marked effect that cari be compared to that of minera1 nitrogen
fertilizer on the yield and total N content of the maize, which is explained, at least
partly, by the increase in nitrogen absorption by maize from the soil N pool
enriched by the prunings applied. These effects on yield and on plant total N are in
agreement with the nutritional properties of these prunings, already demonstrated
by’our work (Ndiaye 1997). The prunings supply 30-35% of the total N in the
maize, values close to those obtained for fertilizer for a similar amount of applied
nitrogen; the value of the N, fixing tree is to contribute “free” to this percentage,
and to the soi1 N pool. Basing the N, fixation at 25% in G. sepium as obtained in
the environment of central Senegal (Ndiaye and Ganry 1997), the contribution of
this N, fixation to the total N in mahe is 8%. Hence it is not much; we suggest that
the reason lies.in the limited fixing ability of the rhizobial strains. If this constraint
could be removed, the fixed contribution could reach 20-25% in maize if the N,
fixation of G. sepium reached the value of 75% in south Senegal (Ndiaye 1997). This
value of 2O-25% is within the range of values found by numerous authors notably
Kang, Wilson, and Lawson (1984) in a more humid climate. They showed, more-
over, that this benefit increased progressively because of the increase in productivity
and the cumulative effect of the prunings.
Finally, the agroforestry system, with a LRE of 1.58, is a system which has a
biological land use eflïciency greater than that of the sum of the two individual
crops each tested as pure stands. This advantage is vital in areas where the soi1 is a
scarce resource as in the groundnut zone of Senegal. During the rainy season, the
agroforestry system would also offer the possibility of growing cow pea (Yigna
unguiculata
L.) in relay cropping before maize harvest between the G. sepiurn hedges,
which could be an additional source of income for the farmer. This trop, together
with the wood production, would make up for the “10s~” of cultivable land.
References
Adegbehin, J. O., and A. B. 1. Igboanugo. 1990. Agroforestry practiçes in Nigeria. Ayrcl-
joresfry Systems 10: l-22.
AFNOR. 1996a. Dosage du carbone organique et du carbone total après combustion sèche.
(Association Française de Normalisation) NF ISO 106 94, pp. 189-200 in AFNOR,
Qualité des sols. Paris, France.
AFNOR. 1996b. Dosage de l’azote total-Méthode de Kjeldahl modifiée. (Association
Française de Normalisation) NF ISO 1 l-261, pp. 257-264, in AFNOR, Qualité des sols.
Paris, France.
AFNOR. 1999a. Analyse granulométrique par sédimentation. Méthode de la pipette.
(Association Française de Normalisation) NF X 31-107, pp. 57-72 in AFNOR, Qualifi
des sols. Paris, France.
. -
Maize-G. Sepium Alley Croppimg in Senegal
327
AFNOR. 1999b. Détermination du pH. (Association Française de Normalisauon) NF ISC)
103 90, pp- 339-34X in AFNOR, Qualité des sols. Paris, France.
Bertrand, R. 1972. Morphopédologie et orientations culturales des régions soudaniennes du
Siné !%oum (Sénégal). L’Agronomie. Tropicale, 27: 1115-l 190.
Blondel, 1). 1971. Contribution a la connaissance de la dynamique de I’azole minéral ensoi
sableux (Dior) au Sénégal. Agronomie Tropicale 26: 1303-1333.
Cissé, L. 1984. Estimation des pertes moyennes en eau et en éléments fertilisants dans les
systèmes de cultures arachide-mil de la zone soudano-sahélienne du Sénégal. Réunion de
synthise de la Division Conjointe FAO/AIEA sur “l’emploi des radioisotopes et des
rayonnements dans les études pour une utilisation rationnelle de l’eau et des engrais dans
les zones semi-arides”, Vienne 6-11 Février 1984, Document Interne. Institut Sénégalais
de Recherches Agricoles. Dakar, Senegal.
Commission de Pédologie et de Cartographie des Sols (CPCS), 1967. Classification des sols---
Edition 1967. 96 pages.
Dabin, B. 1967. Sur une méthode d’analyse du phosphore dans les sols tropicaux., I., pp.
99-l 15 in Colloque sur iafertilité des sols Tropicaux. Tananarive, Madagascar.
Fahmy, T. 1995. XLSTAT, Software add-in Microsoft Excel. http://www.xlstat.com.
Fried, M., and L. A. Dean. 1952. A concept concerning the measurement of available soi1
nutrients. Soi1 Science 73 : 263-214.
Ganry, F. 1990. Application de la méthode isotopique à l’étude des bilans azotés en zone
tropicale sèche. Thèse de doctorat d’Etat, Université de Nancy 1. Nancy, France.
Ghuman, B. S., and R. Lai. 1990. Nutrient addition into soi1 by leaves of Cassio siamea and
Gliricidia sepium grown on an ultisol in Southern Nigeria. Agroforestry Systems 10: 131~-
133.
ISRA. 1992. Rapport Technique, Direction de Recherches sur les Productions Forestières. Insti-
tut Sénégalais de Recherches Agricoles, Dakar, Senegal.
Jama, B., A. Gehutan., and D. N. Ngugi. 1990. Shading effects of alley cropped Leucaena
leucocephola on weed biomass and maize yield at Mtwapa, Coast Province, Kenya.
Agroforeswy Systems 13: l-l 1.
Kang, B. T., (3. F. Wilson, and T. L. Lawson. 1984. Alley cropping sequentially cropped
maize and cowpea with Leucaena on a sandy soi1 in southern Nigeria. Plant und Soi/
85: 267-277.
Ndiaye, M. 1997. Contribution des légumineuses arbustives à l’alimentation azotée du maïs
(Zea mays L.k-Cas d’un système de culture en allées dans le Centre Sud Sénégal, Thèse
de doctorat de I’INPL, Nancy, France.
Ndiaye, M.. and F. Ganry. 1997. Variation in the biological N, fixation by tree legumes in
three ecological zones from the north to the south of Senegal. Arid Soi1 Rcsearch ~lnd
Rrhabi!itation 11:245- 254.
Orsini, L., and J. C. Rémy. 1976. Utilisation du chlorure de cobaltihexammine pour la déter-
mination simultanée de la capacité d’échange et des bases échangeables des sols. Hullcrin
de I’AFES 1976:269-216.
Pieri, C . 1989. Farti(ité des terres de savanes: bilan de trente uns de rechtzrchc, ct de,
dk~eloppement (IiIricole nu Sud du Sahara. Paris, Ministère de la Coopération et du Dével-
oppement et CIRAD-IRAT, France.
PRECOBA. 1982. Etude sociologique: la motivation et les actions des populations à l’égard
de l’arbre. Rappor: Technique No. 1 Projet de Reboisements C~~rnr?lut~nut~~~rtl.s riuns /C
Bassin ,4rachidier du S>négul. ISRA, Dakar, Senegal.
Schroth, G. 1994. Above and below-ground interactions in alley cropping with G‘ilricidia
sepium as Icompared to conventional and mulched sole cropping on a high base status
soi1 in the African rainforest zone. Ph.D. dissertation, University of Bayreuth, Ciermany.
Soil Survey Staff. 1992. Keys to Soi1 Taxonomy. SMSS Technical Monograph No. 19. Fifth
Edition Pocahontas Press, Inc., Blacksburg, Virginia. 556 pages.
Young, A. 1989. Dix hypothèses pour la recherche sols/agroforesterie I’/l~/ro~~~r~,sf~ri~~
Aujourd’hui 1 : 13-16.
Zapata, F. 1990. Isotope techniques in soi1 fertility and plant nutrition studies In U.SL> o/
Nuclear T,chrtiques i n Studios oJ Soi/-Plunt R~,l<ltiollsllip,s.
7‘rrrininy Cour.~ .sr’r~~.s PRO. 2,
edited by Ci llardarson (pp. 61-127) International Atomic Energy Agency, Vienna,
Austria.
- - I I - - -
-.
k .