.- .-- Journal of Native and...
.- .--
Journal of Native and Agricuitut’ai:-Enuironmen,ts
A Cooperating Journal of the Interrk~~?&Ziim.of
Soü
EDITOR-IN-CHIEF
Prof. J. Skuj$ Dept. of Biolo~, Utah Srnte Universit): Logan, UT 84322-530 j, USA
Fax: (435) 797-1575; e-mail: skujirls~~cc.usLI.eciu
EDITORIAL BOARD MEMBEKS
----I
A.S. Abdel-Ghaffar
Univcrsiry
of A l e x a n d r i a .
Alexandria, &ypr
R. Aguilu
Lndia NaGmal
Laboraroncr.
Albuqucrquc,
New Mrxico. UiA
J. Amnson
CEFUCNRS.
M~~p~llicr Gdcx. Frnncc
A.G. Babxv
!>cr:rc I..stin.lc,
4shgahar.~~ur~~r”i,~a~~
.-.
I.M. Barca
Fxa1ion t:xparimr”r~l rlçl L.,,,:,”
:ki”Ada. Spxin
..&Y.R. Lhnmergues
Uirç, France
$J$: Guy
CIRAI)-CA, Mo”<pellicr.
Franr<
Carbon, Nitrogen and Pho,$Phtirtis Mineralization
Potential of Native Agroforestry Plant Residues in
Sails of Senegal
F . I Y A M U R E M Y E
V. ‘GEWIN
R. P. DICK
Department of Crop and Soil Science
Oregon St ate University
Corvallis, OR, USA
M. DIACK
M. SEXE
A. BADIANE
M. DIATTA
Institut Sénégalais de Recherches Agricoles
Dakar, Senegal
The objecriues of this study were to inuestigate Piliostigma reticulatum (shrub) ond
Cordyla pinnata (tree) residues for chemical çomposition, and C, N, and P minerul-
izcltion in Senegalese soils. Soit samples (Sols ferrugineux tropicaux) were collected
j-om a rlative agroforestry system beneath and outsidr u C. pinnata c~~vzop~. Syils
::‘L’Ï~* incub~:rcd [Pr 11 weeks with P. reticulatum (!LUC~~), C. pinnata (Stern, LVFL..
or stem t leaues), peanut (Arachis hypogea), or peur1 rmllet (Pennisetum glaucum)
residues. Ni’trogen and P mineralization for thc soil-plant mixtures was determiaed
by periodic leaching with a 0.01 M CaCl, solution. An additional separate incu-
bation was conducted to investigate C mineralization. The results showed that only
peanut residues caused net N mineralization, while N immobiiizution occurrer! in
the remoining treatments in both the soils derived from beneath or outside the
canopy. This indicates that, ut least in the short-term, these agroforestry residues
would n,ot likety be-u source of N for crops. Net P mineralization varied among
plant re.sidues and soils sites but P. reticulatum amended soits increused soluble
PO, in both soils,_Xhïs suggested that it coutd bc useful for improving P rrvail-
ability. Peanut residues had the highest CO, evolution in both soils suggestinq a
probable relationship between C and N mineratization.
Keywortls
Piliostigma reticulatum, Cordyla pinntttu, peanut, millet, parkland
Depletion of soi1 quality and dësertification of sub-Saharan Africa is increasing and
‘hz -2 Dewnber 1999; accepted 28 Febmary 2000.
The authors would like to thank the ISRA Natural Resource Based Agricultural Research
(NRBAR) project which was funded by USAID/CID Senegal project #685-0285-C-00-2329-00 to
support this~rcsearchlbison
hetween TSEA and Oregon State University. Oregon Agricultural Experi-
ment Station. No. 116808.
Address correspoadence to Dr. R. P. Dick, Dept. of Crop and Soi1 Science, 3017 Ag’ and Life
S :ience Bldg’, Oregon State University, Corvallis, OR 97331-7306. Email: Richard.Dick@orst.edu
359
. .
.
. . ,~~.i -3 i
,,
_
zi.x
;
-.’
Y>.
_ _. +sp=;w
.&z;
.I~~z&c-z
~ -:
-.--
._ > ,mp
‘.’
-T-T .’
j(&; ..:- -$5$Te, ;-1_ .: .,-
,.,,
Ai.iil~$*-
-.$ -__- ‘:.+&;“=;$-, : -_- .- _--
F. lynmuremye ci al.
., ,::- CT-g:-&: f-
._
:-
.!, .:, -__ -.
( :s.
.is.-~el~r-iasing rural populations ,snd agricultÜr~~~~t~m~ and
-__..__
reduC‘ii&&f woociyspecies on the landscape. The residues of native-$7 specres
co%ld%?iip?%&
as soi1 nutrient sources and for improving soil._~wiy, .Early
studiesiwgd Faidherbia (Acacia) albida (Del.) A. Chev. improved physicochemical
properties beneath the tree canopies (Charreau and Vidal 1965; Dancette and
Poulain 1968).
This mode1 of the site enrichment by the tree was the basis for the development
of alley farming. In this system continuous addition of prunings from hedgerow
maintains nutrient status (Kang, Wilson, and Lawson 1984), especialiy nitrogen (N}
when the tree or shrub is a leguminous species. Nitrogen release and mineralization
in alley cropping systems, following addition of prunxngs to the soil, has been exten-
sively studied (Constantinides and Fownes, 1993; Constantinides and Fownes 1994;
Hundoyanto, Godisch, and Giller, 1994; Fox, Meyers, and Vallis 1990; Palm and
Sanchez 1991). However, in many countries the most common agroforestry system
is the parkland system. In this system trees are randomly allowed to grow in trop
fields.
Parkland systems of F. albida and other species have been reported in West
Africa (Samba 1997). In Burkina Faso, Depommier, Jadonet, and Olivier (1992)
confirmed the previous results obtained in Senegal by Charreau and VidaU (1965)
and Dancette and Poulain (1968) where millet or sorghum yield increased beneüth
F. albida (relative to yield outside of the F. &idu canopy). Besides showing the
improvement in physicochemical properties beneath F. ohida, Dancette and
Poulain (1968) and Depommier et al. (1992) sugge.sted that the canopy lcrowns
created microenvironments with higher relative humidity, a lower evapotranspira-
tion potential, reduced maximum temperature, increased soi1 humidity, and greater
rain interception.
Unlike other agroforestry systems such as alley cropping, little attention has
been paid to parkland systems. In West Africa and especially in Senegal the tree
species Cordyfa. pinnata [(A. Rich.) Milne-Redh] is an important species in peanut
and millet fields (Samba 1997). Besides trees, other Woody shrub species cari be
found in these regions as well. A widely distributed and dominant shrub in central
and southern Senegal that is commonly found with ic’. pinnatn És Piliostigma r~li-
cufutum
[(DC.) Hochst.] (M. Diatta, ISRA, Dakar, Sc1ie~~11, persona1
communication). The traditionai management consists of cutting the shrubs and
burning the aboveground residue before soi1 cultivation. Although burning con-
serves cations, significant amounts of N and sulfur are volatihzed and it does not
contribute to the improvement of soi1 quality. No information is available on min-.
eralization of the residues from P. reticulatum and C, @mata or whet.her soi1 frorn
beneath the canopies of C. pinnata cari affect mineralization rates. TO determine
whether these residues could be managed to provide nutrients, chemical character-
ization and-nutrient rnineralization studies are needed. We included in the study,
two of the most common trop residues in Senegal [Pearl millet, Pennisstum glaucurn
(L). R. Br,and.peanut, hachis hypogea L.]. This would provide a relative point of
comparison for the potential agronomie effectiveness of the agroforestry residues to
provide nutrients.
The objectives of this study were to investigate two agroforestry (1’. rrtlculatunl,
C. pinnatu) and two trop residues (Pearl millet and peanut) for: (1) chemical com-
position; and (2) carbon, N, and phosphorus (P) mineralization in Senegalese soil.
--.--_
Materials and Methods
Soils ad Plant Materials
--~~-The cxperimental site was located in Paoskoto, Kaolack, a semiarid agroecoiogical
zone in the peanut basin, between 13” 35’ and 14” 3O’N and 14” 35’ and 16” 4,5’W.
The region is characterized by a.tw$~ni&&imate with an annual rainfall of
700 mm and evapotranspiration of~l$OO mm yr-‘. Temperatures range from aver-
ages of 16°C in December-Januaiy fajverages of 39°C m April-June.
A Deck Dior loamy-sand [fine-sandy, mixed Haplic Ferric LixîsOl (FAO 19!?1)],
leached ferrugeneous tropical soi1 (probably an Ultisol), was used in this study
(Table 1). It was collected from the Ap horizon (O-10 cm depth) at a farmer’s .field
that had been under millet (Pennisefum glaucum). Soîl samples were taken randcmly
with a shovel either beneath the canopy-(3 m radius from trunk; approximately
one-half canopy dia.) or outside the canopy influence (30 m radius from trunk) of a
C. pinnata tree in the millet field. The soil was composited, homogenized and then
crushed to pass Z-mm mesh screen and maintained at field moisture and stored at
22°C. Soils and plant residues of P. reticulatum and C. pinnata were collected in
March 1998. Al1 materials came from farmers’ fields near Kaolack, Senegal during
the dry season. The peanut (Arachis hypogea) and pearl millet (Pennisetum glaucum)
residues were collecre’d at trop maturity in September 1997. The C. pinnata litter
was senescent residue that had recently fallen from the trees. The P. reticulntum
residue (approximately 1 m in height at hary& time) was tut green at ground level.
Leaves and stems of P. reticulatum woody &ecies were separated and then a11 plant
rcsidues were dried at 35°C for five days and individually chopped to pass a l-,cm
sieve and kept in sealed plastic bags.
Mineralization
Incubation Experiments
Mineralization of N and P was conducted following the methods of Stanford and
Smith (1972) with slight modification of the nutrient solution added to the soi1
samples incubated to study P mineralization. Seventy-five g of soi1 samples were
thoroughly mixed with 1.5 g of plant material and transferred into the leaching
tubes with a bottom packed with glass wool to retain the soil. A thin layer of glass
wool was placed on top of the soi1 to minimize dispersion during leaching. At time
zero 100 mL 0.01 A4 CaCl, was added in four increments under a suction of 600
MPa to remove re,adily soluble N (NH: and NO;) and phosphate. Then 30 mL of
the nutrient solution as Lsqcribed by Stanford and Smith (1972) was added and ylut
~~nclc”r %O MPa tensiuu to removc excess solution. In the case of P mineralization,
phosphate was omitted from the nutrient solution and replaced with 0.022 iz1
NH,NO, . This was repeated at each sampling interval.
Each tube was capped with cellophane, with a small hole was made in it to
allow exchange of gas, and put into an incubator at 35°C for 91 days (7, 14, 21, 28,
35.42,49, 63,77,9J). Duplicate samples were used for each treatment.
The C mineralization study was done sntbe same soils and plant material as
those used in the N and P mineralization described above by the static procedure of
Zibilske (1994). Fifty g of soi1 were thoroughly mixed with 1 g of plant material artd
this mixture was transferred into a glass iul%&èaléd with septa at each end. Soi1
moisture was adjusted to 8% (wt/wt) and then the tubes were incubated at 25°C.
TABLK 1 Characteristics of soils
~--
Total
Soi1 location
~-
P H
N
C
P
-~--. g kg-*- - -’
Soi1 under canopy
6.07
0.449
4.01
0.090
Soi1 outside canopy
6.05
0.233
2.96
0.054
_,-. ~.-“.,~,-..----.-. .-
i__.
.L
v--I-n.- .___ ___-_
-E-~--~~
-
--ee--G.z
-
:
‘-..“.
..:=-
.-
._
r;.
--r:
,
_
--.
,,iX1.7
-;
--
_..I__.-I..I
.< .-~ ..-.
eu - w.yeiL;*.
362
-.. -
CO, samples were c
1, 28, 35, 42, 56, 70, 84 days and stored in
~.-.r’.Z ~-.-
vacutainer tubes. Samples ~$5~ PL .were anaiyzed on a gas chromatograph. After
each sampling the tu-d--and allowed to equilibrate with ambient: CO,
levels. Ail incubations were carried out in duplicate.
Both experiments employed completely randomized designs where the treat-
ments were: (1) P. reticulatum Iéaves; (2) P. reticulatum stems; (3) P. rt.vi.cultîtum stem
plus leaves (in same proportion as found under field conditions); (4) (J. pinnata
leaves; (5) aboveground mi&& residue; (6) aboveground peanut residue; and (7)
control (soi1 only). Because soîl sampling came from oniy one tree, statistical
analysis was done separately for each soil.
Laboratory Analysis
pH was determined with a glass electrode (soil: water ratio 2: 1). Total soi1 C was
determined by combustion on a C analyzer (Dorhman, Santa Clara, California) and
total C in residue was measured by LECO WR-12 C autoanalyzer (LECO Cor~.. St.
Joseph, Missouri). Total N. &soils and organic residues was done by Kjelda.hl
digestion followed by steam distillation according to Bremner and Mulvaney (1982).
The leachate was analyzed for NO;-N and NH:-N by steam distillation procedure
of Bremner and Keeney (1965). Phosphorus in the leachate of mineralization was
measured colorimetrically (Murphy and Riley 1962). Total P in plant residue and
soils was determined by the methods of Cresser and Parsons (1979) and Dick and
Tabatabai (1977), respectively. Lignin, cellulose, and hemicellulose were determined
by the method of Goering and Vati Soest (1970).
Calculation of Kinetic Constants
The mineralization potential and the rate constants were estimated using bath non-
linear and linear regression. The exponential equation used to calculate potentially
mineralizable N (No) and first rate order constant k, is as follows:
N, = No[1 - exp(- k, 01,
(1)
where N, = cumulative amount of N removed by leaching at a specltic urne (t). f’he
statistical package SAS was used to fit the data to the equation using nonlinear
regression procedure. The linear regression or zero order mode1 was of the [OMI
N,i, = BO + k,t,
(2)
where B, = the intercept and k, is the slope.
Results and Discussion
.---
Chemical Composition of Soils and Plant Material
Soils characteristics are shown in Table 1. Nitrogen, C, artd P contents tend to be
higher in soils beneath tree çanopies than in soils outside tree canopy. The pH
values are the same in both soils. Table 2 shows the chemical composition of the
plant material used in this expetient. The total C content ranged from 34% I:O
39%. The N content was the lowest in the millet (0.46%) and the highest in the
peauut residue (1.~4%). The- P content varies in the following order: C. pinna,rc~
leaves > Arachis hypogea > Pennisetum glaucum L > P. reticulatum leaves :> P. reti-
culutum Stern.
Tht? lî@ïîÏ~~ontënts of P. retlculatum stems and C. pinnata leaves were the
highest at > 28% lignin compared to peanut residues and P. reticulatum leaves,
which had lowest lignin values.
i
-.z --!
_TA&3 Z..Characteristi& of residues
.-.-~..i_ - - - -
-
Total N
C
P
Lignti----Gktt&Se
-HemiceHlulose
-.-
Species
g kg-’
C/N
g:kg-’
-
-
----.
Amchis hypoyeo (peanut)
15.4 374 1.01
24
20
313:
7 1
Pennketum glaucum (millel)
4.6 388 0.85
8 3
171
431
249
P. reliculatum leaves
12.4 345 0.74
2 7
6 2
392.
3 3 3
P. recicdatum stems
06.2 343 0.35
5 5
280
459
6 5
C. pinnata leaves
11.4 390 1.33
3 4
295
178
2 2
--.--~--.
-
Before the incubation, all samples amended with the different plant residues
were leached with 0.01 M CaCI, solution to extract soluble inorganic N. The results
are presented in Figure 1. Interestingly, the soluble inorganic N content in the first
extraction was highest for the millet-amended soils, whereas the leaf material of
both the tree and shrub had the lowest levels. Compared to N, considerably less P
was extracted fr-om these residues (Figure 2). Nonetheless, these values were high,
suggesting there are significant levels of soluble N and P in the materials, which
should be readily available for plant uptake.
N Mineralization
At any given time of incubation, the minera1 N (NO;-N + NH:-N) content was
dominated by nitrate N fraction (data not shown) regardless of the type of treat-
ment. The changes of cumulative NH:-N with time were variable. In the peanut
1000 T------
AMENDMENTS
FIGURE É Amount of N leached at time zero after amending soil with plant resi-
dues (bars with the same letter within a column are not significantly different at
P < 0.05 according to Duncan’s Multiple Range test).
J
0.05
i
AMENDMENTS
FIGURE 2 Amount of P leached at time zero after amending soi1 with plant reqi-
dues (bars with the same letter within a column are noé significantlr different at
P < 0.05 according to Duncan’s Multiple Range test).
treated soils (the only treatment to have net N mineralization), NH:-& Or, :he leach-
ate increased hetween the third and sixth week of incubation and decreajc 1 from the
ninth tieek. ln the control samp!ss NI%:-N in the leachats showect s p.-ik ::f the
second week of incubation and decreased for the remainder of the incubc::ion (data
net shown). Low values of NH: -N in the leachate compared to NO; -2; *Lalues are
in agreement with the findings of Cassman and Muns (1980). Overall ths camulative
NH,-N values were greater in the soils collected beneath the tree canq-a Lhan the
soi1 samples outside the tree canopy.
The relationships between cumulative inorganic N (NO;-N + NI-Ii -‘i I and thc
time of incubation are shown in Figure.3. A.visual check of the curves sb;:wed that
peanut amended soi1 and the control were curvilinear whereas the othsr ‘.r;arments
were linear or near linear.
The results show that net N mineralization occurred only in pear.:’ Imcnded
soils, whereas net immobilization occurred in the other treatments. Cor.Tarison of
the treatments where N immobilization occurred showed that soi1 from btneath the
tree canopy, amended with millet residue or P. reticulatum leaves, itnm~?:iized 11:s~
N than the remaining treatments. Conversely,.jn the soi1 collected outil: the tree
canopy, treatments showed no significant differences in N immobilizing capacity.
This difference in soils was also reflected in the fact that more inorgL5c N was
accumulated in soils beneath the tree canopy than those outside the trti :anopy in
a11 treatments.
These differences in N dynamics between soi1 collected beneath o: ,.::side 1 he
tree canopy may be related to the higher N content of soil-beneath thz ::E canopy.
Furthermore, C content was greater beneath the canopy (0.4 vs. ci.:: : for soi1
outside the tree). Thus the greater amount of N and C as an energy SOUTZ Zay have
--
.Soils outside-of canopy
Peanut
/./a
,.E-’
m.----.
4000
,l.-
3000
Control
L
-a
Tree leaf
1
_-.I
I
Soils under canopy
Contro
3000
/A
2000
1000
0
8
10
0
2
4
6
12
.
WEE c
FIGIJRE 3 Cumulative N leached in soi1 amende‘I: with plant materials in soi1
bene;.th and outside the canopy of C. pinnata tree.
reduced the amount of N immobilization and increi ie:d total N accumulation from
soi1 below the tree canopy. More organic C in the SI 3 likely enablled greater micro-
fi
bial activîty.
Chrr results are very consist with other studie: in that net N mineralization
occurred oolyin soils amended with a herbaceous lej me (neanut) and net immobil-
ization dominated with nonleguminous herbaceot
(millet) ami nonleguminous
trees and shrub material used in agroforestry (C
pinnata and P. reticulaturn).
Kaboneka, Sabble, and Mauromoustakos (1997)
Jorted net mineralization in
..”
.:
.--
l
F. lyamuremye et â
~~~ -.‘k-leguminous
herbamus
dues (maize amended s( ): Also, comparing leguminous +rees used in agroforestry,
Constatinïdes a n d F o w s (1994) concluded thtit frêsïë@Ïï%ëG.n general had net
accumulation whereas tl nonlegume fresh leaves had net depletion.
The plant materiais sed in our study varied in quality N, C, C/N, lignin, and
other properties (Table
. Initial N content, lignin and polyphenol concentrations
have been reported to c .relate with N release from plant material. But the results
obtained by Melillo, i er, and Muratore (1982), Fox -et al. (1990), Palm and
Sanchez (1991) and COI
antinides and Fownes (1994) differed in which parameter
correlated best. Along * th P. rcriculutum leaves, peanut residue material had the
highest initial N conten
and the Iowest C/‘N ratio, lignin/N ratio {Table 2). But f’.
reticulatum leaves with
relatively low C/N ratio of 27 still had net N immobil-
ization. These results 07 -a11 agree with the conclusion reported by Constantinides
and Fownes (1994). 01 :r factors such as polyphenol concentration might hnve
contributed to the low ,elease of N from P. reticulatum leaves. The poor per-
formance of C. pinrîata
:aves residues is similar to Aber and Melillo (1982) who
reported the highest im: lbilization rates occurred in the nonleguminous tree litter
(hardwood leaves) high : both lignin and N.
Another possible e: lanation for the net N immobilization for millet and 1’.
reticulatum was the higl Z/N ratios of these residues, 81 and 53, respectively. This
might have even been i àctor for C. Pinnata litter N mineralization with a C/N
ratio of 21. According tc Stevenson (1985) net N immobilization lasts until the C/N
ratio of the decomposi ; material has been lowered to about 20. Only peanut
residue material had C, ratio close to 20 (23). It may be that the time of incu-
bation was too short to
wer C/N ratios to 20 or that lower residue quality may be
inhibiting N release.
N Mineralization
Model
Data from the peanut ( lended samples and the control fit the first order mode1
regardless of the soi1 si j (data not .shown). Data from soils outside tree canopy
sites treated with P. rcti
latum leaves also fit the first ordel mode1 with the :em? ;II-
ing treatments fitting th’ inear zero order model. Peanut amended sampies had the:
greatest mineralization 1 potential (N,,).
We did not find th mode1 fitting to be very useful for interpreting the data.
which indicates these r Idels are poorly suited for developing N mineralization
parameters for soils am ded with organic residues. In some cases, N, was greatel
than the total cumulatil mineralized N (e.g., peanut amended samples) and others
were smaller than the tc 11 cumulative N. Dou and’colleagues (1996) attributed the
values of N, smaller th L the cumulative mineralizable N to the weakness of lhe
single model, which sho d be greater than the observed cumulative minera1 N. On
the basis of the studies 1 Juma, Paul, and Mary (1984), Cabrera and Kisscl (1988).
and Sierra (1990), it was lggested tha.t because of the dependence of goodness of t?t
and the magnitude of t
parameters on specific models and incubation time, the
parameters derived fron nodels do not predict N mineralization for a given soil; if
such potential exist.s un< r a given set of conditions (Dou et al. 1996). These values
are, however, merely
ithematically defined quantities obtained by nonlinear
I
regression analysis (Fau an and Bonde 1987).
Phosphorus Mitzeralizati
The changes of cumula re inorganic P mineralization wi& &-ne are presented in
Figure 4. Contrary to 1 mineralization, P mineralization investigations have limi-
tations because of the I ceptibility of PO:- (mineralization product) to chemical
_...
i
_._
..~:‘~:~
Z-T;;:
.-.-.-
-.---------
w-m----
èsidues
1.25
Millet
“y:=zrf!j
..ys-=-$~ea”ut
w-- - - ?%uGY,t,&
_ -..$em. t-!qw
P
9
CZontrol
s
EP-;IP-.l;;;es
D,,
‘.
<’
Peanut
0’
/~.-----.-a
:
:
,4----+
* - -
Tree leaves
stems + leaves
0
2
4
s
8
1 0
1 2
WEEI c
FlG1JRE 4 Cumulative P leached in soi1 amended with plant materials in soi1
beneat h and outside the canopy of-e:-pinnata tree.
sorpticn and precipitation reactions (Sharpley and SI lith 1989). Thus the P deter-
mined in the leachate is not quantitative for P miners .ization, but rather is the net
efl’ect of P mineralization and degree of P fixation reactions. Consequently, the
results are most useful from a comparative basis amonl the treatments.
Al:hough the soifs used in this experiment are cha acterized as very sandy soils,
these soils at~. boaieU witi! ke and Al oxides or hydr jxides (persona1 communica-
tion, A. Badiane) that are well known to strongly sorb ; dded P to soif.
The curve of cumulative P in the leachate vs. tim of incubation is curvilinear.
Iiowever, these cüfiés--prësent two diisfinctivë Segr lents ~corresponding to two
period!; of incubation (from day 7 to day 21, and from day 21 to day 77). Generally,
a visual check of the curves showed that during the fi ,st three weeks of incubation
8” -- --.
i.
ij,
r,,
i;
-
,,.
ear with time. During the second stage. of incubati&%@l$% to
ship was curvilinear. These results agreed with those of Sha.rp-
v e d an initial phase during residue decotipoSi7iîon
of
14 days with maximum mineraiization occurring at 28
tive P in the leachate followed treatments in the order of:
gea = Pennisetum glaucum > P. reticularum
tum stem = control = C. pinnata leaves for soi1
outside the--tfee
The treatment order for net P release in soi1 beneath the
tree canopy was
culatum leaves > Arachis hypoigea LY mixture P. reticulalm
leaves + stem >
tum glnucum 2 P. reiiculnfum stem z control > C. pinnata
;ed net P immobilization in soi1 beneath tbe tree
innata and P. reticulatum stem + leaf residue caused zeto
e tree canopy. The remaining plant materials caused ‘cumu-
lative P levels that
greater than the control. The highest cumulative P was in
P. reticulntum leaves, or peanut (soi1 outside tree canopy)
lended samples (soi1 beneath tree canopy). 1’. reti-
ntly caused the highest amount of P in the leachates of both
the canopy followed by peanut residues. A careful obser-
ts that soi1 source might have influenced the amount of P
. The consistently higher P content in the leachate of soi1
amended with P .
ticulatum leaves is probably the result of its high P oontent
low C/P ratio. Except millet which had variable trends,
reticulatum leaves that had a relatively low C: P ratilo, had
of P accumulation. Dalal (1977), Fuller, Nelson, and Miller
nd Smith (1989) suggested that for residues with C: P ratio
P mineralization occurs and if the ratio was higher tha.n 300,
. Soi1 characteristics of the two sites could have affected P
. Relatively greater amounts of P accumulation were found
ree canopy, which is consistent with other studies (Kamara
1997; Jung 1970; Dancette and Poulain 1968; and Char-
YRU and Vidal 1
that have report-d higher P content in thp <oil bcneath th:
ted that mineralization of residue P appeared
r function. t o available soi1 P content. The high C: F’ ratio
s the only tree material) may explain why it
P mineralization.
reater levels of soluble P by peanut and P. rctidatum may
e more effective in decreasing P fixation. Organic amend-
to increase 0.01 M CaCl, extractable P and Bray 1 P
Dick 1996; Sharpley and Smith 1989). During decomposi-
ds may be released that reduce potential for P fixation reac-
Dick 1996) and thus less P may have been adsorbed
lized P to remain in solution and be leached.
hat P. reticulatum residue could be useful for increasing P
negalese soil. This could be important because P is a major
limitation for trop
duction in the important agroecological zone of Senegal. Fur-
thermore, farmers
have limited resources
to purchase fertilizers. However, this
needs to be tested
field conditions.
cumulative CO, and time of incubation is shown in
that, at any given time, CO, evolved was greater for
r the control samples.
M ineralization
1000
--
-..--~
Soils outside of canopy
800
600
Soils under canopy
0
---
_.-_-
_.1
I
II-----T-
0
2
4
6
8
1 0
1 2
1 4
EK
HGUKE 5 Cumulative CO, evolved in soil ar :nded with plant materials in soi1
beneath and outside the canopy of C. pinnata trel
Peanut amended samples were associated w h greater CO, evolution. In other
samples, the amount of CO, evolved about the tme, regardlless of soi1 origin. It is
interesting 1.0 notice that peanut residues that Ci rsed net N mineralization had the
roost CO, evolution. This may suggest a relation
hip between C mineralization and
14 mineralization. Overall tix iLdtLti&ip bet.wl
n CO2 evolved and time of incu-
bation is linear.
l‘his study has shown that the millet and Woody
arkland species residues immobil-
i.led N during a standardized incubation up to 7 days. Only peanut residue caused
_.
8.
-_.
-+-s+~rornboth beneath or outside the tree canopy. PiE&~S~-
vqhad ~1a1w li_gnio content and a relatively narrow C: ‘N ratio 2ZS -l
rIet, N immobilization. This may suggest that ocher parameters
~eoneentration played a greater role in deterrnining the net
-
s time of incubation arc curvi-
gments corresponding to two periods of incubation. The
was in the soil amended with millet and P. reticulatrtm leaves
Or wanut residu
soi1 outside the tree canopy) or with P. rrric~ularum leaves c’soi]
The consistently highest P content in the leachate of soiis
3
~lutum leaves is probably the result of its relatively low C/P
content of these residues and their capacity of reducing F’ sorp-
between the cumulative CO2 and incubation time was linear,
reater for a11 residues than the control. Peanut residues
ion had the most CO, evolution suggesting a probable
guerallization and N mineralization. Plant residlres that
ion also caused less CO, evolution or C minelalization.
S. Keeney. 1965. Steam distillation methods for determination of
and itrite. Anaf~*tical Chemical Acta 32:486-495.
Mulvaney. 1982. Total Nitrogen, In M~thods ($soii .~M~~~.~;.s. JMI’; 2.
bi&g;cal properties. edited by A. L. Page, pp. 595 -624, American
al. 1965. Influence de I’.-lcacia alhida Del. sur le $01. cct;i!ion rn;nir::
nliis ï~‘ti::!.<t’! 11ttt aU Sénégal. L Lryl’. *C~liill:’ I.roJJi<al< ?Xl- rji>.
J. H. Fon-nes. 1993. NItrogen minerakzation
patterns of Ieaf tu,ig
1 legnminous trees. ,+lyrqforestry S,~srrrns 24:223-231.
H. Folvnes. 1994. Nitrogen mmeralization from leavcs and litter of
. ICROSAT, Andhra Pradesh, India.
0 World Soc Resources Report 66. FAO, Rome.
Ggkcubation data on nitrogen mineralization
ng soii organic matter dynnmics and soil pro.
Bulletin International Association of E.cdogy, then&&gii-USA.
_ _.--
FOX, R. H., R. J. K. Meyers, and 1. Vallis. 1990. T e nitrogen minerahzation rate of tegume
residues in soil as influenced by their polyphen 1, fignin, and nitroeen contents. Pinnt nnd
Soi1 129:251-259.
a
Prunings of iegume hedgerow tree in
ubation.
Plant and Soi1 160:237-248.
endments and phosphorus sorptiou in
Juma, N. G., E. A. Paul, and B. Mary. 1984. Kine
sis of net nitrogen mineralization in
soil. Soi1 Science Society of America Journal 4
Jung, G. 1970. Variations saisonnieres des
ologiques d’un sol ferru-
gineux rropicai peu lessive (d’or) soum
de I’Acadiu albida (Del).
Oecologia Plantarum 2: 113-136.
Kaboneka, S., W. E. Sabble, and A. Mauro
and its effects on Ethiopian highland
Xang, B. T., G. F . Wilson, and T. L. Lawson. 19
Iley cropping, a stabie alternative to
shifting cultiuation in Nigcriu. Nigeria Interna
a1 Institute of Tropical .Agriculture,
Ibadan, Nigeria.
Melillo, J. M., J. D. Aber, and J. F. Muratore
and lignin control of hardwood
leaf Iitter decomposition dynamics. EcoJo
F4urphy, J., and L. P. Kiley. 1962. A modifïed
ethod for the determination
of
phosphorus in natural waters. halyticnl
Palm, C. A., and P. A. Sanchez. 1991. Nitroge
legumes as affected by their lignin and polyp
Ph.D. thesis, Universite de Laval,
Sharpley, A. N., and S. J. Smith. 1989
leaching of phosphorus from soi1
incubated with surface-applied an
residues. Journal of Enuironmet~tal
Quulity 18: 101-105.
Zibilske, L. M. 1994. Carbon mineralization. In Merho
soi1 analysis. Part 2. Microbioloyi-
cal and biochemical propertics. edited by W. K.
vers. 3. Angle, P. Bottomley, D.
Bezdicck, S. Smith, A. Tabatabai, and A. Wollum,
835-864. Soi1 Science of America,
Madison, Wisconsin, USA.
Journal of Native and Agricuitut’ai:-Enuironmen,ts
A Cooperating Journal of the Interrk~~?&Ziim.of
Soü
EDITOR-IN-CHIEF
Prof. J. Skuj$ Dept. of Biolo~, Utah Srnte Universit): Logan, UT 84322-530 j, USA
Fax: (435) 797-1575; e-mail: skujirls~~cc.usLI.eciu
EDITORIAL BOARD MEMBEKS
----I
A.S. Abdel-Ghaffar
Univcrsiry
of A l e x a n d r i a .
Alexandria, &ypr
R. Aguilu
Lndia NaGmal
Laboraroncr.
Albuqucrquc,
New Mrxico. UiA
J. Amnson
CEFUCNRS.
M~~p~llicr Gdcx. Frnncc
A.G. Babxv
!>cr:rc I..stin.lc,
4shgahar.~~ur~~r”i,~a~~
.-.
I.M. Barca
Fxa1ion t:xparimr”r~l rlçl L.,,,:,”
:ki”Ada. Spxin
..&Y.R. Lhnmergues
Uirç, France
$J$: Guy
CIRAI)-CA, Mo”<pellicr.
Franr<
Carbon, Nitrogen and Pho,$Phtirtis Mineralization
Potential of Native Agroforestry Plant Residues in
Sails of Senegal
F . I Y A M U R E M Y E
V. ‘GEWIN
R. P. DICK
Department of Crop and Soil Science
Oregon St ate University
Corvallis, OR, USA
M. DIACK
M. SEXE
A. BADIANE
M. DIATTA
Institut Sénégalais de Recherches Agricoles
Dakar, Senegal
The objecriues of this study were to inuestigate Piliostigma reticulatum (shrub) ond
Cordyla pinnata (tree) residues for chemical çomposition, and C, N, and P minerul-
izcltion in Senegalese soils. Soit samples (Sols ferrugineux tropicaux) were collected
j-om a rlative agroforestry system beneath and outsidr u C. pinnata c~~vzop~. Syils
::‘L’Ï~* incub~:rcd [Pr 11 weeks with P. reticulatum (!LUC~~), C. pinnata (Stern, LVFL..
or stem t leaues), peanut (Arachis hypogea), or peur1 rmllet (Pennisetum glaucum)
residues. Ni’trogen and P mineralization for thc soil-plant mixtures was determiaed
by periodic leaching with a 0.01 M CaCl, solution. An additional separate incu-
bation was conducted to investigate C mineralization. The results showed that only
peanut residues caused net N mineralization, while N immobiiizution occurrer! in
the remoining treatments in both the soils derived from beneath or outside the
canopy. This indicates that, ut least in the short-term, these agroforestry residues
would n,ot likety be-u source of N for crops. Net P mineralization varied among
plant re.sidues and soils sites but P. reticulatum amended soits increused soluble
PO, in both soils,_Xhïs suggested that it coutd bc useful for improving P rrvail-
ability. Peanut residues had the highest CO, evolution in both soils suggestinq a
probable relationship between C and N mineratization.
Keywortls
Piliostigma reticulatum, Cordyla pinntttu, peanut, millet, parkland
Depletion of soi1 quality and dësertification of sub-Saharan Africa is increasing and
‘hz -2 Dewnber 1999; accepted 28 Febmary 2000.
The authors would like to thank the ISRA Natural Resource Based Agricultural Research
(NRBAR) project which was funded by USAID/CID Senegal project #685-0285-C-00-2329-00 to
support this~rcsearchlbison
hetween TSEA and Oregon State University. Oregon Agricultural Experi-
ment Station. No. 116808.
Address correspoadence to Dr. R. P. Dick, Dept. of Crop and Soi1 Science, 3017 Ag’ and Life
S :ience Bldg’, Oregon State University, Corvallis, OR 97331-7306. Email: Richard.Dick@orst.edu
359
. .
.
. . ,~~.i -3 i
,,
_
zi.x
;
-.’
Y>.
_ _. +sp=;w
.&z;
.I~~z&c-z
~ -:
-.--
._ > ,mp
‘.’
-T-T .’
j(&; ..:- -$5$Te, ;-1_ .: .,-
,.,,
Ai.iil~$*-
-.$ -__- ‘:.+&;“=;$-, : -_- .- _--
F. lynmuremye ci al.
., ,::- CT-g:-&: f-
._
:-
.!, .:, -__ -.
( :s.
.is.-~el~r-iasing rural populations ,snd agricultÜr~~~~t~m~ and
-__..__
reduC‘ii&&f woociyspecies on the landscape. The residues of native-$7 specres
co%ld%?iip?%&
as soi1 nutrient sources and for improving soil._~wiy, .Early
studiesiwgd Faidherbia (Acacia) albida (Del.) A. Chev. improved physicochemical
properties beneath the tree canopies (Charreau and Vidal 1965; Dancette and
Poulain 1968).
This mode1 of the site enrichment by the tree was the basis for the development
of alley farming. In this system continuous addition of prunings from hedgerow
maintains nutrient status (Kang, Wilson, and Lawson 1984), especialiy nitrogen (N}
when the tree or shrub is a leguminous species. Nitrogen release and mineralization
in alley cropping systems, following addition of prunxngs to the soil, has been exten-
sively studied (Constantinides and Fownes, 1993; Constantinides and Fownes 1994;
Hundoyanto, Godisch, and Giller, 1994; Fox, Meyers, and Vallis 1990; Palm and
Sanchez 1991). However, in many countries the most common agroforestry system
is the parkland system. In this system trees are randomly allowed to grow in trop
fields.
Parkland systems of F. albida and other species have been reported in West
Africa (Samba 1997). In Burkina Faso, Depommier, Jadonet, and Olivier (1992)
confirmed the previous results obtained in Senegal by Charreau and VidaU (1965)
and Dancette and Poulain (1968) where millet or sorghum yield increased beneüth
F. albida (relative to yield outside of the F. &idu canopy). Besides showing the
improvement in physicochemical properties beneath F. ohida, Dancette and
Poulain (1968) and Depommier et al. (1992) sugge.sted that the canopy lcrowns
created microenvironments with higher relative humidity, a lower evapotranspira-
tion potential, reduced maximum temperature, increased soi1 humidity, and greater
rain interception.
Unlike other agroforestry systems such as alley cropping, little attention has
been paid to parkland systems. In West Africa and especially in Senegal the tree
species Cordyfa. pinnata [(A. Rich.) Milne-Redh] is an important species in peanut
and millet fields (Samba 1997). Besides trees, other Woody shrub species cari be
found in these regions as well. A widely distributed and dominant shrub in central
and southern Senegal that is commonly found with ic’. pinnatn És Piliostigma r~li-
cufutum
[(DC.) Hochst.] (M. Diatta, ISRA, Dakar, Sc1ie~~11, persona1
communication). The traditionai management consists of cutting the shrubs and
burning the aboveground residue before soi1 cultivation. Although burning con-
serves cations, significant amounts of N and sulfur are volatihzed and it does not
contribute to the improvement of soi1 quality. No information is available on min-.
eralization of the residues from P. reticulatum and C, @mata or whet.her soi1 frorn
beneath the canopies of C. pinnata cari affect mineralization rates. TO determine
whether these residues could be managed to provide nutrients, chemical character-
ization and-nutrient rnineralization studies are needed. We included in the study,
two of the most common trop residues in Senegal [Pearl millet, Pennisstum glaucurn
(L). R. Br,and.peanut, hachis hypogea L.]. This would provide a relative point of
comparison for the potential agronomie effectiveness of the agroforestry residues to
provide nutrients.
The objectives of this study were to investigate two agroforestry (1’. rrtlculatunl,
C. pinnatu) and two trop residues (Pearl millet and peanut) for: (1) chemical com-
position; and (2) carbon, N, and phosphorus (P) mineralization in Senegalese soil.
--.--_
Materials and Methods
Soils ad Plant Materials
--~~-The cxperimental site was located in Paoskoto, Kaolack, a semiarid agroecoiogical
zone in the peanut basin, between 13” 35’ and 14” 3O’N and 14” 35’ and 16” 4,5’W.
The region is characterized by a.tw$~ni&&imate with an annual rainfall of
700 mm and evapotranspiration of~l$OO mm yr-‘. Temperatures range from aver-
ages of 16°C in December-Januaiy fajverages of 39°C m April-June.
A Deck Dior loamy-sand [fine-sandy, mixed Haplic Ferric LixîsOl (FAO 19!?1)],
leached ferrugeneous tropical soi1 (probably an Ultisol), was used in this study
(Table 1). It was collected from the Ap horizon (O-10 cm depth) at a farmer’s .field
that had been under millet (Pennisefum glaucum). Soîl samples were taken randcmly
with a shovel either beneath the canopy-(3 m radius from trunk; approximately
one-half canopy dia.) or outside the canopy influence (30 m radius from trunk) of a
C. pinnata tree in the millet field. The soil was composited, homogenized and then
crushed to pass Z-mm mesh screen and maintained at field moisture and stored at
22°C. Soils and plant residues of P. reticulatum and C. pinnata were collected in
March 1998. Al1 materials came from farmers’ fields near Kaolack, Senegal during
the dry season. The peanut (Arachis hypogea) and pearl millet (Pennisetum glaucum)
residues were collecre’d at trop maturity in September 1997. The C. pinnata litter
was senescent residue that had recently fallen from the trees. The P. reticulntum
residue (approximately 1 m in height at hary& time) was tut green at ground level.
Leaves and stems of P. reticulatum woody &ecies were separated and then a11 plant
rcsidues were dried at 35°C for five days and individually chopped to pass a l-,cm
sieve and kept in sealed plastic bags.
Mineralization
Incubation Experiments
Mineralization of N and P was conducted following the methods of Stanford and
Smith (1972) with slight modification of the nutrient solution added to the soi1
samples incubated to study P mineralization. Seventy-five g of soi1 samples were
thoroughly mixed with 1.5 g of plant material and transferred into the leaching
tubes with a bottom packed with glass wool to retain the soil. A thin layer of glass
wool was placed on top of the soi1 to minimize dispersion during leaching. At time
zero 100 mL 0.01 A4 CaCl, was added in four increments under a suction of 600
MPa to remove re,adily soluble N (NH: and NO;) and phosphate. Then 30 mL of
the nutrient solution as Lsqcribed by Stanford and Smith (1972) was added and ylut
~~nclc”r %O MPa tensiuu to removc excess solution. In the case of P mineralization,
phosphate was omitted from the nutrient solution and replaced with 0.022 iz1
NH,NO, . This was repeated at each sampling interval.
Each tube was capped with cellophane, with a small hole was made in it to
allow exchange of gas, and put into an incubator at 35°C for 91 days (7, 14, 21, 28,
35.42,49, 63,77,9J). Duplicate samples were used for each treatment.
The C mineralization study was done sntbe same soils and plant material as
those used in the N and P mineralization described above by the static procedure of
Zibilske (1994). Fifty g of soi1 were thoroughly mixed with 1 g of plant material artd
this mixture was transferred into a glass iul%&èaléd with septa at each end. Soi1
moisture was adjusted to 8% (wt/wt) and then the tubes were incubated at 25°C.
TABLK 1 Characteristics of soils
~--
Total
Soi1 location
~-
P H
N
C
P
-~--. g kg-*- - -’
Soi1 under canopy
6.07
0.449
4.01
0.090
Soi1 outside canopy
6.05
0.233
2.96
0.054
_,-. ~.-“.,~,-..----.-. .-
i__.
.L
v--I-n.- .___ ___-_
-E-~--~~
-
--ee--G.z
-
:
‘-..“.
..:=-
.-
._
r;.
--r:
,
_
--.
,,iX1.7
-;
--
_..I__.-I..I
.< .-~ ..-.
eu - w.yeiL;*.
362
-.. -
CO, samples were c
1, 28, 35, 42, 56, 70, 84 days and stored in
~.-.r’.Z ~-.-
vacutainer tubes. Samples ~$5~ PL .were anaiyzed on a gas chromatograph. After
each sampling the tu-d--and allowed to equilibrate with ambient: CO,
levels. Ail incubations were carried out in duplicate.
Both experiments employed completely randomized designs where the treat-
ments were: (1) P. reticulatum Iéaves; (2) P. reticulatum stems; (3) P. rt.vi.cultîtum stem
plus leaves (in same proportion as found under field conditions); (4) (J. pinnata
leaves; (5) aboveground mi&& residue; (6) aboveground peanut residue; and (7)
control (soi1 only). Because soîl sampling came from oniy one tree, statistical
analysis was done separately for each soil.
Laboratory Analysis
pH was determined with a glass electrode (soil: water ratio 2: 1). Total soi1 C was
determined by combustion on a C analyzer (Dorhman, Santa Clara, California) and
total C in residue was measured by LECO WR-12 C autoanalyzer (LECO Cor~.. St.
Joseph, Missouri). Total N. &soils and organic residues was done by Kjelda.hl
digestion followed by steam distillation according to Bremner and Mulvaney (1982).
The leachate was analyzed for NO;-N and NH:-N by steam distillation procedure
of Bremner and Keeney (1965). Phosphorus in the leachate of mineralization was
measured colorimetrically (Murphy and Riley 1962). Total P in plant residue and
soils was determined by the methods of Cresser and Parsons (1979) and Dick and
Tabatabai (1977), respectively. Lignin, cellulose, and hemicellulose were determined
by the method of Goering and Vati Soest (1970).
Calculation of Kinetic Constants
The mineralization potential and the rate constants were estimated using bath non-
linear and linear regression. The exponential equation used to calculate potentially
mineralizable N (No) and first rate order constant k, is as follows:
N, = No[1 - exp(- k, 01,
(1)
where N, = cumulative amount of N removed by leaching at a specltic urne (t). f’he
statistical package SAS was used to fit the data to the equation using nonlinear
regression procedure. The linear regression or zero order mode1 was of the [OMI
N,i, = BO + k,t,
(2)
where B, = the intercept and k, is the slope.
Results and Discussion
.---
Chemical Composition of Soils and Plant Material
Soils characteristics are shown in Table 1. Nitrogen, C, artd P contents tend to be
higher in soils beneath tree çanopies than in soils outside tree canopy. The pH
values are the same in both soils. Table 2 shows the chemical composition of the
plant material used in this expetient. The total C content ranged from 34% I:O
39%. The N content was the lowest in the millet (0.46%) and the highest in the
peauut residue (1.~4%). The- P content varies in the following order: C. pinna,rc~
leaves > Arachis hypogea > Pennisetum glaucum L > P. reticulatum leaves :> P. reti-
culutum Stern.
Tht? lî@ïîÏ~~ontënts of P. retlculatum stems and C. pinnata leaves were the
highest at > 28% lignin compared to peanut residues and P. reticulatum leaves,
which had lowest lignin values.
i
-.z --!
_TA&3 Z..Characteristi& of residues
.-.-~..i_ - - - -
-
Total N
C
P
Lignti----Gktt&Se
-HemiceHlulose
-.-
Species
g kg-’
C/N
g:kg-’
-
-
----.
Amchis hypoyeo (peanut)
15.4 374 1.01
24
20
313:
7 1
Pennketum glaucum (millel)
4.6 388 0.85
8 3
171
431
249
P. reliculatum leaves
12.4 345 0.74
2 7
6 2
392.
3 3 3
P. recicdatum stems
06.2 343 0.35
5 5
280
459
6 5
C. pinnata leaves
11.4 390 1.33
3 4
295
178
2 2
--.--~--.
-
Before the incubation, all samples amended with the different plant residues
were leached with 0.01 M CaCI, solution to extract soluble inorganic N. The results
are presented in Figure 1. Interestingly, the soluble inorganic N content in the first
extraction was highest for the millet-amended soils, whereas the leaf material of
both the tree and shrub had the lowest levels. Compared to N, considerably less P
was extracted fr-om these residues (Figure 2). Nonetheless, these values were high,
suggesting there are significant levels of soluble N and P in the materials, which
should be readily available for plant uptake.
N Mineralization
At any given time of incubation, the minera1 N (NO;-N + NH:-N) content was
dominated by nitrate N fraction (data not shown) regardless of the type of treat-
ment. The changes of cumulative NH:-N with time were variable. In the peanut
1000 T------
AMENDMENTS
FIGURE É Amount of N leached at time zero after amending soil with plant resi-
dues (bars with the same letter within a column are not significantly different at
P < 0.05 according to Duncan’s Multiple Range test).
J
0.05
i
AMENDMENTS
FIGURE 2 Amount of P leached at time zero after amending soi1 with plant reqi-
dues (bars with the same letter within a column are noé significantlr different at
P < 0.05 according to Duncan’s Multiple Range test).
treated soils (the only treatment to have net N mineralization), NH:-& Or, :he leach-
ate increased hetween the third and sixth week of incubation and decreajc 1 from the
ninth tieek. ln the control samp!ss NI%:-N in the leachats showect s p.-ik ::f the
second week of incubation and decreased for the remainder of the incubc::ion (data
net shown). Low values of NH: -N in the leachate compared to NO; -2; *Lalues are
in agreement with the findings of Cassman and Muns (1980). Overall ths camulative
NH,-N values were greater in the soils collected beneath the tree canq-a Lhan the
soi1 samples outside the tree canopy.
The relationships between cumulative inorganic N (NO;-N + NI-Ii -‘i I and thc
time of incubation are shown in Figure.3. A.visual check of the curves sb;:wed that
peanut amended soi1 and the control were curvilinear whereas the othsr ‘.r;arments
were linear or near linear.
The results show that net N mineralization occurred only in pear.:’ Imcnded
soils, whereas net immobilization occurred in the other treatments. Cor.Tarison of
the treatments where N immobilization occurred showed that soi1 from btneath the
tree canopy, amended with millet residue or P. reticulatum leaves, itnm~?:iized 11:s~
N than the remaining treatments. Conversely,.jn the soi1 collected outil: the tree
canopy, treatments showed no significant differences in N immobilizing capacity.
This difference in soils was also reflected in the fact that more inorgL5c N was
accumulated in soils beneath the tree canopy than those outside the trti :anopy in
a11 treatments.
These differences in N dynamics between soi1 collected beneath o: ,.::side 1 he
tree canopy may be related to the higher N content of soil-beneath thz ::E canopy.
Furthermore, C content was greater beneath the canopy (0.4 vs. ci.:: : for soi1
outside the tree). Thus the greater amount of N and C as an energy SOUTZ Zay have
--
.Soils outside-of canopy
Peanut
/./a
,.E-’
m.----.
4000
,l.-
3000
Control
L
-a
Tree leaf
1
_-.I
I
Soils under canopy
Contro
3000
/A
2000
1000
0
8
10
0
2
4
6
12
.
WEE c
FIGIJRE 3 Cumulative N leached in soi1 amende‘I: with plant materials in soi1
bene;.th and outside the canopy of C. pinnata tree.
reduced the amount of N immobilization and increi ie:d total N accumulation from
soi1 below the tree canopy. More organic C in the SI 3 likely enablled greater micro-
fi
bial activîty.
Chrr results are very consist with other studie: in that net N mineralization
occurred oolyin soils amended with a herbaceous lej me (neanut) and net immobil-
ization dominated with nonleguminous herbaceot
(millet) ami nonleguminous
trees and shrub material used in agroforestry (C
pinnata and P. reticulaturn).
Kaboneka, Sabble, and Mauromoustakos (1997)
Jorted net mineralization in
..”
.:
.--
l
F. lyamuremye et â
~~~ -.‘k-leguminous
herbamus
dues (maize amended s( ): Also, comparing leguminous +rees used in agroforestry,
Constatinïdes a n d F o w s (1994) concluded thtit frêsïë@Ïï%ëG.n general had net
accumulation whereas tl nonlegume fresh leaves had net depletion.
The plant materiais sed in our study varied in quality N, C, C/N, lignin, and
other properties (Table
. Initial N content, lignin and polyphenol concentrations
have been reported to c .relate with N release from plant material. But the results
obtained by Melillo, i er, and Muratore (1982), Fox -et al. (1990), Palm and
Sanchez (1991) and COI
antinides and Fownes (1994) differed in which parameter
correlated best. Along * th P. rcriculutum leaves, peanut residue material had the
highest initial N conten
and the Iowest C/‘N ratio, lignin/N ratio {Table 2). But f’.
reticulatum leaves with
relatively low C/N ratio of 27 still had net N immobil-
ization. These results 07 -a11 agree with the conclusion reported by Constantinides
and Fownes (1994). 01 :r factors such as polyphenol concentration might hnve
contributed to the low ,elease of N from P. reticulatum leaves. The poor per-
formance of C. pinrîata
:aves residues is similar to Aber and Melillo (1982) who
reported the highest im: lbilization rates occurred in the nonleguminous tree litter
(hardwood leaves) high : both lignin and N.
Another possible e: lanation for the net N immobilization for millet and 1’.
reticulatum was the higl Z/N ratios of these residues, 81 and 53, respectively. This
might have even been i àctor for C. Pinnata litter N mineralization with a C/N
ratio of 21. According tc Stevenson (1985) net N immobilization lasts until the C/N
ratio of the decomposi ; material has been lowered to about 20. Only peanut
residue material had C, ratio close to 20 (23). It may be that the time of incu-
bation was too short to
wer C/N ratios to 20 or that lower residue quality may be
inhibiting N release.
N Mineralization
Model
Data from the peanut ( lended samples and the control fit the first order mode1
regardless of the soi1 si j (data not .shown). Data from soils outside tree canopy
sites treated with P. rcti
latum leaves also fit the first ordel mode1 with the :em? ;II-
ing treatments fitting th’ inear zero order model. Peanut amended sampies had the:
greatest mineralization 1 potential (N,,).
We did not find th mode1 fitting to be very useful for interpreting the data.
which indicates these r Idels are poorly suited for developing N mineralization
parameters for soils am ded with organic residues. In some cases, N, was greatel
than the total cumulatil mineralized N (e.g., peanut amended samples) and others
were smaller than the tc 11 cumulative N. Dou and’colleagues (1996) attributed the
values of N, smaller th L the cumulative mineralizable N to the weakness of lhe
single model, which sho d be greater than the observed cumulative minera1 N. On
the basis of the studies 1 Juma, Paul, and Mary (1984), Cabrera and Kisscl (1988).
and Sierra (1990), it was lggested tha.t because of the dependence of goodness of t?t
and the magnitude of t
parameters on specific models and incubation time, the
parameters derived fron nodels do not predict N mineralization for a given soil; if
such potential exist.s un< r a given set of conditions (Dou et al. 1996). These values
are, however, merely
ithematically defined quantities obtained by nonlinear
I
regression analysis (Fau an and Bonde 1987).
Phosphorus Mitzeralizati
The changes of cumula re inorganic P mineralization wi& &-ne are presented in
Figure 4. Contrary to 1 mineralization, P mineralization investigations have limi-
tations because of the I ceptibility of PO:- (mineralization product) to chemical
_...
i
_._
..~:‘~:~
Z-T;;:
.-.-.-
-.---------
w-m----
èsidues
1.25
Millet
“y:=zrf!j
..ys-=-$~ea”ut
w-- - - ?%uGY,t,&
_ -..$em. t-!qw
P
9
CZontrol
s
EP-;IP-.l;;;es
D,,
‘.
<’
Peanut
0’
/~.-----.-a
:
:
,4----+
* - -
Tree leaves
stems + leaves
0
2
4
s
8
1 0
1 2
WEEI c
FlG1JRE 4 Cumulative P leached in soi1 amended with plant materials in soi1
beneat h and outside the canopy of-e:-pinnata tree.
sorpticn and precipitation reactions (Sharpley and SI lith 1989). Thus the P deter-
mined in the leachate is not quantitative for P miners .ization, but rather is the net
efl’ect of P mineralization and degree of P fixation reactions. Consequently, the
results are most useful from a comparative basis amonl the treatments.
Al:hough the soifs used in this experiment are cha acterized as very sandy soils,
these soils at~. boaieU witi! ke and Al oxides or hydr jxides (persona1 communica-
tion, A. Badiane) that are well known to strongly sorb ; dded P to soif.
The curve of cumulative P in the leachate vs. tim of incubation is curvilinear.
Iiowever, these cüfiés--prësent two diisfinctivë Segr lents ~corresponding to two
period!; of incubation (from day 7 to day 21, and from day 21 to day 77). Generally,
a visual check of the curves showed that during the fi ,st three weeks of incubation
8” -- --.
i.
ij,
r,,
i;
-
,,.
ear with time. During the second stage. of incubati&%@l$% to
ship was curvilinear. These results agreed with those of Sha.rp-
v e d an initial phase during residue decotipoSi7iîon
of
14 days with maximum mineraiization occurring at 28
tive P in the leachate followed treatments in the order of:
gea = Pennisetum glaucum > P. reticularum
tum stem = control = C. pinnata leaves for soi1
outside the--tfee
The treatment order for net P release in soi1 beneath the
tree canopy was
culatum leaves > Arachis hypoigea LY mixture P. reticulalm
leaves + stem >
tum glnucum 2 P. reiiculnfum stem z control > C. pinnata
;ed net P immobilization in soi1 beneath tbe tree
innata and P. reticulatum stem + leaf residue caused zeto
e tree canopy. The remaining plant materials caused ‘cumu-
lative P levels that
greater than the control. The highest cumulative P was in
P. reticulntum leaves, or peanut (soi1 outside tree canopy)
lended samples (soi1 beneath tree canopy). 1’. reti-
ntly caused the highest amount of P in the leachates of both
the canopy followed by peanut residues. A careful obser-
ts that soi1 source might have influenced the amount of P
. The consistently higher P content in the leachate of soi1
amended with P .
ticulatum leaves is probably the result of its high P oontent
low C/P ratio. Except millet which had variable trends,
reticulatum leaves that had a relatively low C: P ratilo, had
of P accumulation. Dalal (1977), Fuller, Nelson, and Miller
nd Smith (1989) suggested that for residues with C: P ratio
P mineralization occurs and if the ratio was higher tha.n 300,
. Soi1 characteristics of the two sites could have affected P
. Relatively greater amounts of P accumulation were found
ree canopy, which is consistent with other studies (Kamara
1997; Jung 1970; Dancette and Poulain 1968; and Char-
YRU and Vidal 1
that have report-d higher P content in thp <oil bcneath th:
ted that mineralization of residue P appeared
r function. t o available soi1 P content. The high C: F’ ratio
s the only tree material) may explain why it
P mineralization.
reater levels of soluble P by peanut and P. rctidatum may
e more effective in decreasing P fixation. Organic amend-
to increase 0.01 M CaCl, extractable P and Bray 1 P
Dick 1996; Sharpley and Smith 1989). During decomposi-
ds may be released that reduce potential for P fixation reac-
Dick 1996) and thus less P may have been adsorbed
lized P to remain in solution and be leached.
hat P. reticulatum residue could be useful for increasing P
negalese soil. This could be important because P is a major
limitation for trop
duction in the important agroecological zone of Senegal. Fur-
thermore, farmers
have limited resources
to purchase fertilizers. However, this
needs to be tested
field conditions.
cumulative CO, and time of incubation is shown in
that, at any given time, CO, evolved was greater for
r the control samples.
M ineralization
1000
--
-..--~
Soils outside of canopy
800
600
Soils under canopy
0
---
_.-_-
_.1
I
II-----T-
0
2
4
6
8
1 0
1 2
1 4
EK
HGUKE 5 Cumulative CO, evolved in soil ar :nded with plant materials in soi1
beneath and outside the canopy of C. pinnata trel
Peanut amended samples were associated w h greater CO, evolution. In other
samples, the amount of CO, evolved about the tme, regardlless of soi1 origin. It is
interesting 1.0 notice that peanut residues that Ci rsed net N mineralization had the
roost CO, evolution. This may suggest a relation
hip between C mineralization and
14 mineralization. Overall tix iLdtLti&ip bet.wl
n CO2 evolved and time of incu-
bation is linear.
l‘his study has shown that the millet and Woody
arkland species residues immobil-
i.led N during a standardized incubation up to 7 days. Only peanut residue caused
_.
8.
-_.
-+-s+~rornboth beneath or outside the tree canopy. PiE&~S~-
vqhad ~1a1w li_gnio content and a relatively narrow C: ‘N ratio 2ZS -l
rIet, N immobilization. This may suggest that ocher parameters
~eoneentration played a greater role in deterrnining the net
-
s time of incubation arc curvi-
gments corresponding to two periods of incubation. The
was in the soil amended with millet and P. reticulatrtm leaves
Or wanut residu
soi1 outside the tree canopy) or with P. rrric~ularum leaves c’soi]
The consistently highest P content in the leachate of soiis
3
~lutum leaves is probably the result of its relatively low C/P
content of these residues and their capacity of reducing F’ sorp-
between the cumulative CO2 and incubation time was linear,
reater for a11 residues than the control. Peanut residues
ion had the most CO, evolution suggesting a probable
guerallization and N mineralization. Plant residlres that
ion also caused less CO, evolution or C minelalization.
S. Keeney. 1965. Steam distillation methods for determination of
and itrite. Anaf~*tical Chemical Acta 32:486-495.
Mulvaney. 1982. Total Nitrogen, In M~thods ($soii .~M~~~.~;.s. JMI’; 2.
bi&g;cal properties. edited by A. L. Page, pp. 595 -624, American
al. 1965. Influence de I’.-lcacia alhida Del. sur le $01. cct;i!ion rn;nir::
nliis ï~‘ti::!.<t’! 11ttt aU Sénégal. L Lryl’. *C~liill:’ I.roJJi<al< ?Xl- rji>.
J. H. Fon-nes. 1993. NItrogen minerakzation
patterns of Ieaf tu,ig
1 legnminous trees. ,+lyrqforestry S,~srrrns 24:223-231.
H. Folvnes. 1994. Nitrogen mmeralization from leavcs and litter of
. ICROSAT, Andhra Pradesh, India.
0 World Soc Resources Report 66. FAO, Rome.
Ggkcubation data on nitrogen mineralization
ng soii organic matter dynnmics and soil pro.
Bulletin International Association of E.cdogy, then&&gii-USA.
_ _.--
FOX, R. H., R. J. K. Meyers, and 1. Vallis. 1990. T e nitrogen minerahzation rate of tegume
residues in soil as influenced by their polyphen 1, fignin, and nitroeen contents. Pinnt nnd
Soi1 129:251-259.
a
Prunings of iegume hedgerow tree in
ubation.
Plant and Soi1 160:237-248.
endments and phosphorus sorptiou in
Juma, N. G., E. A. Paul, and B. Mary. 1984. Kine
sis of net nitrogen mineralization in
soil. Soi1 Science Society of America Journal 4
Jung, G. 1970. Variations saisonnieres des
ologiques d’un sol ferru-
gineux rropicai peu lessive (d’or) soum
de I’Acadiu albida (Del).
Oecologia Plantarum 2: 113-136.
Kaboneka, S., W. E. Sabble, and A. Mauro
and its effects on Ethiopian highland
Xang, B. T., G. F . Wilson, and T. L. Lawson. 19
Iley cropping, a stabie alternative to
shifting cultiuation in Nigcriu. Nigeria Interna
a1 Institute of Tropical .Agriculture,
Ibadan, Nigeria.
Melillo, J. M., J. D. Aber, and J. F. Muratore
and lignin control of hardwood
leaf Iitter decomposition dynamics. EcoJo
F4urphy, J., and L. P. Kiley. 1962. A modifïed
ethod for the determination
of
phosphorus in natural waters. halyticnl
Palm, C. A., and P. A. Sanchez. 1991. Nitroge
legumes as affected by their lignin and polyp
Ph.D. thesis, Universite de Laval,
Sharpley, A. N., and S. J. Smith. 1989
leaching of phosphorus from soi1
incubated with surface-applied an
residues. Journal of Enuironmet~tal
Quulity 18: 101-105.
Zibilske, L. M. 1994. Carbon mineralization. In Merho
soi1 analysis. Part 2. Microbioloyi-
cal and biochemical propertics. edited by W. K.
vers. 3. Angle, P. Bottomley, D.
Bezdicck, S. Smith, A. Tabatabai, and A. Wollum,
835-864. Soi1 Science of America,
Madison, Wisconsin, USA.