“.--A”&‘m, *-.----. Journal of ...
“.--A”&‘m,
*-.----.
Journal of NativèanaAgnCtiltural Envirhments
A Cooperating Journal of-tbl&snationai
Union of Soi1 Science
IUSS - UISS - IBU
L
Decomposition of a Native Shrub, Pîliostigma
reticulatrcm, Litter in Soils of Semiarid Senegal
M. DIACK
M. SENE
A. N. Bz4DIANE
M. DIA’l-TA
Agricultural Kesearch Institute of Senegal (ISRA)
Senegal
R . F. DI[CK
Department of Crop and Soi1 Science
Oregon State University
Corvallis, Oregon, USA
I)etvIopinq effective management strategies that restorc dc<grad& SOI~S requires gn
evaluution of the quality of the litter residues. This study relates the chemical com-
position of the biomass components to the decomposition rates Jor Piliostigma reti-
culatum (DC.) Hochst., a native shrub, under Jield and laboratory conditions. The
ratt’b were determined by mass 10s~. The chamqcs in the specific surface ureu of the
resufrrr ii1 relation to mass loss ranged from 13 x 1O- 5 to #5 x 10 ’ which was
srmrlar tcj trop residues in other studies. At Jic4d conditions, P. reticulatum mass
ic~.s~ WI.\\ highrr (80% of the initial mass lest ovrr eight months) rhan thar under
( on::.::!M c,onditinns (50%). Such fast decornpovr;, St! of rG4rrr.; +r\\ !hr p4writrl
j<;-Jilrmrrs to stop burmng these residues be( ause hiyh amounts C$ residues Will net
likely acrxmulate and cause interferencc
with tillqtj and planting operations.
t.‘urthrr .;tudies are needed on the role oj‘ soi/ fauna un dvcomposition, mineral-
ization cv’nutrients from these residues, and thr potentialfor incorporating residues
into rhe systtm without burning.
Keywords
:soil quality, soil organic matter, litter residues, mass loss, manage-
ment strategjes
Maintaining trop resid*%n the SO~I surface is an effective meth;d of controlling
~--
soi1 erosion and degra.dation
provided that there is enough plant residue. However,
in many areas of the world, insuffkient amounts of residue are produced to provide
adequate soi1 caver and protection. In Senegal, there is little information available
on how plant resiclues, left on the soil surface, affect soi1 quality (Chopart et al.
1979). Several studies have focused on the incorporation of green manure in the soi1
Received 20 May 1999; accepted 20 Sept. 1999.
The authors wouid 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 resea& liûisûri bctneenfSRA-ancl
Oregon State University.
Address corresponience to Dr. Richard P. Dick, Department of Soil Science, 3017 Agricultural
Life Science Building, Oregon State University, Corvallis. Oregon 97331-7306, USA. E-mail:
Richard.Dick@orst.edu
205
ommergues 1956), trop residues (Chopart, Nicou, and Vacéaü&?97$%iwcnhardt-
Reversat 1987; Feller and Ganry -1982) and composting (FeUe:r and Ganry 1982;
--~
Ganry and Sarr 1985; Feller, Chopart, and Daucette 1987; Seck 1987). Most of this
work aimed to increase trop yield in thc short term, where the major contribution
cornes from input of nutrients. However, another important nonnutrient contribu-
tion of organic amendments is its role in building soil structure, organic matter and
maintaining biological activity. This is important in tmproving and sustaining long-
ter-m trop yields.
Although trop residues cari have a beneficial eifect on soils, very little of the
trop residue in Senegal and most of Sub-Sahara is returned to the soi1 because it is
too valuable as a food source for livestock (persona1 communication, A. Badiane,
ISRA, Senegal). Consequently, other organic matter sources are needed in Senegal.
Even though manure and composting are available, there are insuflïcient amounts
that cari be readily collected to apply in ail helds of a typical Senegalese farm
(Ganry and Guiraud 1979, Pichet, Truong, and Bcunard 1979).
Organic matter management is of critical importance in Senegal because of the
low fertility status of soils and the declining vields that may be due to degradation
of soils. This is likely occurring because of mcrcascs in the rural population and
associated intensification of agriculture, e.g., marked decreases in land fallowing for
regeneration (Diack et al. 1998). In the absence of appropriate residue management
technologies, the parkland system with optimal tree or shrub species, or both, may
be an alternative approach for restoring degraded soils. Currently, there is Iittle
information available on how traditional shrubs of the parkland systems improve
degraded soils.
Piliosrigma reticulcttum (DC.) Hochst (Breman and Kessler 1995), a shrub
species, is commonly found in the major peanut and millet growing regjons in
Senegal. P. reticulatum is a nonnodulating legume and could be considered as a
“ratoon caver trop” after the cropping season. In farmers’ fields, the aboveground
biomass of P. reticulatum is tut at the soi1 surface and burned just prior to the rainy
season. As it regrows during the rainy season, it is tut again at every weedmg. After
the growing season, it regrows to hcights of 0.70 to 1 .OO m and canopy diameters of
0.75 to 1.75 m and has rooting depth of 2.5 !o ! tn with as many as 700 shru.bs ha- ’
(Diack et ai. 1998). P. reticulatrrn~ biomass;
Lauld bu: dli Important organic matter
source for soils in Senegal with preliminary estimates of >0,5 Mg ha-’ above-
ground dry biomass (Diack et al. 1998). However, to use these materials rather than
burning them, it would be important to dctermine the litter quality and rate of
decomposition to effectively manage it from an agroromic perspective. Since P. rcri-
culutum is a Woody species, it may decompose too slowly and interfere with trop
cultural activities. Therefore, understanding how rapidly P. retic&zturn
residues
decompose and are lost from a field site, is a prerequisite to the design of effective
management strategies for soi1 improvement.
The objectives of this study were: (1) to determine decomposition ratés for I>.
reticulatum aboveground residues and roots by mass loss under field and laboratory
conditions; (2) to determine the relationships between chemical and physical charac-
teristics of the residues and their rates of decomposition; (3) to determine whether
the previous trop affects the residue quality as measured by chemical analysis and
rates of decomposition; and (4) to provide guidance for future studies-on manage-
ment strategies of P. reticulatum in farmers’ fields to improve the quality of the soils.
Materiais and Methods
S i t e
The experimental site was located in Paoskoto, Kaolack, a semiarid agroecological
zone in the peanut basin, between 13” 35’ and 14” 30’ of northern latitude and 14”
-...i ‘. 1m
a
.y ‘c-J::,$-.
_-.
. +-~~,~~wyw=
__
- == .@p&:~~:<---
ï ,..-_ _ ..L :
-: - Shrub ~t$er, ‘Decotiposition
207
--.-.-.
-- ,-&& -,-.... r
35’ and 16:’ 45’ of western long$~&egion is characterized by a tropical Suda-
nian climate with an annual rainfall of700 mm. Temperatures vary between 16°C in
I>ecember -January and 39°C inÂpÏÏE:-
A Deck Dior loamy-sand (fine-sandy, mixed Haplic Ferric Lixisol), leached ferru-
gcneous tropical $,oil (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
undcr millet (Pcrwisrrum glaucum L). The soil was air-dried, crushed to pass 2 mm
mesh screen. then stored until use. This field and soi1 was used for subsequent field
and lab incubation studies.
Plant MateriaIs
The 1’. rctirulatur~~ materials were randomly collected from each of two farmcr’s
fields which are under a 2-year rotatîon-of peanut (Aruchis hJlpogrtea, L.) and millet.
One field was under peanut and the other under millet the previous year before
collecting thc residues. P. reticuLztum residues were separated into different residue
types: leaves, stems, and roots.
Chcmical Analysis ami Mass Ratio of Plant Residue
The relative mass ‘ylias 17%, 26%, and 56% for leaves, stems, and roots, respectively.
The relative mass of aboveground biomass for leaves and stems was 40% and 60%,
respectively. Each plant residue component was analyzed for total C and total N
contents by dry combustion (Mode1 CHN-600; Leco Corp.). Fructose was deter-
mined spectrophoto-colorimetrically (Davis and Gander 1967). Cellulose, hemi-
cellulose and lignin contents were measured by sequential fiber analysis (Goering
and van Soest 1970).
Mass LOS.~ Experinzent
under Laboratory Condition.~
The experiment consi.sted of a randomized complete design with two trop rotations,
and three residue treatments with three replications. The three treatments wcre
leavcs, stems, and roots. Stems were tut into pieces of 4 to 5 cm long and 5 g of
residue were spread out on 100 g air-dry soi1 in a 15 x 9 cm polystyrene dish and
then covered with another 100 g of air-dried soi]. The soi1 was brought to a mois-
turc content of -- 33.3 kPa water potential which equalled 60% water holding
capacity, plus 300% of the residue mass (Myrold et al. 1981). After addition of the
appropriate amount of water, the inctibEEn’clish was covered with a perforated cap
to aliow aeration. l”he. samples were incubated at 30°C + 1°C.
Samples were withdrawn on day 3, 7, 14, 28, 56, and 84 of the incubation for
mass loss measure.nent. At each destructive sampling, the incubation mixture was
oven-dried at 5o”C, for 48 h. When dry, the residues were carefully separated from
the soi], gently washed to remove the soil particles, and put back into the oven at
TABLE 1 Soi1 physical and chemical characteristics
-~
Clay
Silt
Sand
Total organic C
Total N
Snil --
--.
-
--__
---
sample
P H
g.kg-'
0 -10 cm
10.1
5.2
84.7
6.7
4.68
0.45
--.- -.--..
__- -_. . __ - _,-.. ^
-..
. .x:7.;: ,... L -.
50°C for 48 h. T&%%&&$e&veighed,
then piaced in crucibles for asfiing~ “‘.
500°C for 3 h, and reweighed after ashing. The residue remaining after ashing-wv
assumed to be of ao~-rïïmëraraloti$n, and this was subtracted from rhe weight of the
pre-ashed plant residue.
Mass L.oss Experiment
under Field Conditions
Leaves, stems, and roots constituted the three residuc types for this experiment. One
hundred g of air-dried residues were placed into a 25 x 25 cm bags (2 mm mesh)
and sealed as litter bags. The samples were incorporated (3 July 1997, beginning of
rainy season) into the top 10 cm of a field in a 3 x 2 factorial experiment with three
1’. rcticuhum residue types (stems, leaves, and roots) and two P. rrriculatum residue
sources (from peanut or millet field) with three replications. This field was previously
under millet (source of P. reticulatum) but was under peanut when buried bags were
in the field.
Samples were withdrawn monthly and eight months afler placement of residue
bags into the soil. At each destructive sampling, the residues within the litter bags
were oven-dried at 50°C for 48 h. When dry, the rcsidues were taken off from the
litter bags and processed in the same way as the iaboratory experiment to determine
the mass 10s~.
Soi1 water potential during the season was -- 1000 MPa in .July, 1997, then
increasing to its highest level in September at ---OC, MPa, and then decreasing to
-50 MPa in October, followed by a steady decline to
120 MPa b y F’ebruary,
1998.
, . .
5pcc$k Sarface Area-to-Mass Ratio
Spccihc surface area-to-mass ratios were represcutc~l
hy a k value uith a dimension
of ha kg ’ of residue. Specific surface amas for I~~;L~L; antl >.t+,,; we:r measzuec!
using a digital-planimeter (Mode1 Ottplan 7(Q’71ii, A. UT?‘ GMBH, Kempten,
Germany). As decomposition proceeds, the ratio between the specific surface area
ami the mass remaining was calculated over time The equation used to çonvert
residue mass to surface caver (Gregory et al. 1985) is expressed ;as:
where C is thc fraction of the surface caver remalning; m is the mass (kg ha- ‘) of
residue present on the surface; and k a constant. The constant k was derived from
the following equation :
k = -Iog(l -- C)/m.
Statistical Analysis
Statistical analysis of the data wasdone to determine differences
amongtreatmcnts,
..-..
using MSTAT-C, Version 1 (MSTAT-C, 1991, MSU, Ml). Comparisons between
treatment means were computed at the P = 0.05 and I’ = 0.01 levels, using the
Duncan’s multiple range test procedure.
..-_”
_.? .---
TABLE 2 Initial chemical composition of Piliostigma reticulutunn residues collected from a peanut or
millet field
Total C
Total N
Fructose
Cellulose
Hemicellulose
Lignin
Previous
R e s i d u e - -
field trop
type
gkg- *
_-
Peanut
Leaves
483b*
13.8a
18.7c
241~
22c
242b
Stems
536a
7.5b
42.3a
448a
62b
22oc
Roots
529a
6.3b
30.4b
382b
96a
264a
Millet
Leaves
474c*
12.3a
2O.lc
202c
5oc
247b
Stems
500b
7.7b
40.8a
379a
90b
228c
Roots
545a
7.0b
37.7b
420b
123a
2 9 4 a
.--
I
* In each column. values followed by the same letter, are not significantly different by the Duncan’s multiple range /
test at P = 0.05.
l
Results and Discussion
initial Chenrical Composition for the P. reticulatum Residues
The mean concentrations of total C and N, fructose, hemicellulose, and lignin (Table
2) were signifkantly different between P. reticulatum
residue types as expectcd.
However, the prcvious trop generally had a significant (I’ < 0.05) elfect <on the
chemical properties for most residue types. The analysis of the structural componers.
cellulose, hemicellulose, and lignin, although showing some differences
in means
between the field source of the shrub residue these were not seatistically different
when comparing each individual residue (Table 2).
Changes in Mass LOS~ under Fieid Conditions
Under field conditions, 82% of the roots (Fig. la) was lest during an &month
period whereas 78% of stem residues was lost and only 40% cf the leaves disap-
peared for thc residue that originated from a tîeld where peanut was pervious trop.
During the first month of decomposition, the stem mass loss was greater (- 42%)
9 0
Leaves
8 0
Stems
Roots
7 0
60
Y-,
F
50
b-
--
??
?
Leaves
??Stems
0 Roots
Ju97 A u S e OC No Dë Ja98 Fe Ma
Period of
__-.- _--- incubation.(nlonths)
-__ --
--.
FIGURE 1 Decomposition of P. reticulutum residues that originated from a peanut
field (a) or millet field (b). under field conditions. Bars renresent standard deviations.
\\
,,
r---
!
- .--.-----T----Shrub Litter L?ccompo.yitiort
A..
7”
8 baves from peanut field
o baves from millet field
80 C AT
70
60
50
40
30
m Stems from peanut lïelc
0 Stems from millet tïeld
90
Roots from peanut tïeld
80
Roots from millet h-Id
70
60
50
40
30
20
10
01
Ju97 Au Se Oc No De Ja98 Fe Ma
Period of incubation (months)
FIGURE 2 Field decomposition of P. reticulatum (a) leaves, (b) stems, and (c) roots
originating from either peanut or millet fields. Bars represent standard deviations.
than that for the roots (-- 28%) and the leaves (- 25%). P. reticulntum residue stems
and roots originating from the millet field had equal cumulative ~s.lasses of 82th -
(Fig. lbj cumpared to~leaves whï~h~had~?~% losses. Mer one month, stems and
roots lest 42% of their mass, while the leaves lost only 32%.
The previous trop affect (Fig. 2a) was a significant (P < 0.05) of the origin of leaf
residue with a 70% loss of leaves from the millet field compared to 60% loss of
100
90
80
70
60
50
40
30
?? baves
20 (4
?? Stern8
10
o Rools
90
80
70
60
50
40
?? Leaves
30
’ S t e m s
20
. w
o Roots
10
0
I
I
I
I
I
I
I
I
0
10 20 30 40 50
60 70 SO
90
hcubation tinte (daysj
FIGURE 3 Laboratory decomposition of 1’. rt~ti’c~htwn residues, originating L
from peanut field or (b) millet field. Bars represent S;tandard deviations.
leaves from the peanut field. No significant difference was noted (Fig. 2b) betwe::
loss of stem mass originating from the millet field (82%) or the pcanut tîeld (79”:
Root mass loss also (Fig. 2c) was not significantly different from millet fi.eld (83’ i
and peanut field (80%).
Under field conditions, the decomposition rates for a11 the P. reticufatm reo.-
dues (Fig. la, lb, 2a, 2b, and 2c) followed an exponential 3-phase pattern. The rapt
mass loss during the first month was probably due to a high total N content, a hi5
level of readily available C in the fonn of extractable sugars or, a combination .:;
both. Kinetically, the mass loss from the residues exhibited a linear dependence c:
the chemical composition of the residue during this period. The rapid disalppearancr
of these soluble compounds was probably due to a quick build up of the microblé
activity, which would incresse tb? rnacq 1~s. 4lso, the readily available C and !i
components in the trop residues might provide the initial energy and nutrienu
necessary to activate the microorganisms that are responsible for the degradation 6
.theless readily available components of the residue.
The leveling off phase for mass loss occurred from two to four months whic:
is likely when mainly recalcitrant materials remains. During this period, hemicelluloi-.
is likely the main fraction available to the microorganisms (Collins et al. 1,990; Star.
-.-
;
-‘y.--
m
::,<re.
3:;
-
“7-~~-~.!-.~,..“~y--‘.
-
...--*,.~“~~~;r~~~~~~~
e%g ,1 <-.
~ ,_- A~.-&. -2..
.-
---Shr&
r,r”
-orI
213
-
-,f q..; .I? ;.:.. :,.; i
_, &~~&.C&. i __
1993). As the decomposition protiX:-Erate of mass loss slows down, fohowing
an exponential curve., probably_.~..~~~alcitrance
of the remaining residue.
After four months of decomposition, the remaining residues entered a third phase
during which the slowly available residue components dominated in the substrate.
Lignin, which is one of the most resistant plant compounds, was probably the major
remaining componcnt during this phase of dewmposition.
The rate of iignin degradation is less affected by changes in temperature and
nitrogen availabiliiy than other major plant components during decomposition
(Sarkanen and Lucwig 1971). Another factor for slow decomposition is that lignin
decomposers are relatively slowly growing microorganisms (Witkamp and van Der
Drift 1963). As restilt, the rate of mass loss follows a steady-state for the test of the
decomposition proc:ess.
Root and stem residues from the peanut field (Fig. la) have undergone greater
cumulative mass loss than the leaves. This is probably due to higher fructosc
content in the roots and stems than in the leaves of P. reticulatutn.
The physical and
chemical compositi’on
of the residues and, particularly soluble C in the form of
fructose, constitute the most important--factors controlling decomposition rates
(Knapp, Elliott, and Campbell 1983). Residue decomposition is a process in which
the rate transforrna:ion is proportional to the qualitative amount of residue avail-
able to thc rnicroorganisms. Our decomposition patterns (exponential decreases)
were sirnilar to other decomposition studies and Papendick (1978) Knapp et al.
(I 983), Stott et al. (I986), Stroo et al. (1989), Collins et al. (1990) and Jensen (1994).
C ‘hunps in Mass Lo.>s undcr Laboratory Conditions
Thc 1’. tc>ricrrl~r~r>l leaves that originated from the peanut field (Fig. 3a) lest ;i
greater amount of mass (50%) than the roots (35%) and the stems (24%) in eight
months. During the first month, leaf residue mass loss was significantly different
( -28%) than both sterns and roots (-- 13%). Leaf mass loss was 48% whereas for
the roots and the strms originating from the millet field, mass Ioss was only 30%
(l‘if. 3b). l‘he first ml>nth of decomposition showed a greater mass loss for the leavet;
i
?SfLl than cither <item< or roots (- 15%) for residues of the mi!le: field.
CiiijkG the iield study, no significant effect on the residue cumulative mass loss
was notcd as a funct: on of the source of residue. Leaf residue mass loss from peanut.
lield was 5 1% while tha.t from the millet field was 49% (Fig. 4a). Mass loss of stems
from the peanut field WBS 25% which was not significantly different from that of the
millet tîeld (28”/0) (Fig. 4b). Root mass loss from the peanut field was 35% while that
from the millet field was 30% (Fig. 4~).
Under laboratory conditions, the same pattern of rates of decomposition was
obscrved as the held study. However, the decomposition rate of the leaves seemed
greater than the stems or roots, which is the opposite of the field data. This may be
‘----‘Y--
relatcd to several factors such as prehandlmg of the soi1 by air drying and sieving
the soi\\ in the lab strdy. Another factor is that lab conditions excluded large fauna.
Morcover, the soi1 is sandy and has very low total C and N contents (Table 1).
Hecuuse of this 10~1 fertility level, soi1 microorganisms are probably relatively
dormant. The discont:inuous
distribution of organic resources in soils causes soils to
have a large dormant population, which is ri& in species and has an ability to
survive stress (Jenkinson and Ladd 1981). According to Lavelle (1996) this resting
tirne v.4~ !ast rrom months to years.
Therefore, redistribution of microorganisms and organic substrates
by
invertebrates and roots that mix the soi1 and add water and readily available sub-
strates to the soi1 may beimgorta n t in activa&-microbiaf
activity and decomposi-
tion. These invertebrates could be divided into three major functional groups:
micropredators, litter transformers, and ecosystem engineers (Lavelle 1996). We did
net check whether any of these major functional groups were present in the soil.
214
?? Leaf (Peanut as previous trop
60
50
40
30
20 - ?? Stern (Peanut as previous trop
10 _ EI Stern (Millet as previous trop)
50
40
30
20 t + Root (Peanut as previous trop
10
Q Root (M%et as previous trop)
o__
~.
’
’
.‘.
’
’
’
’
’
0
10 20 30 40 50 60 70 80 90
Incubation time (days)
FIGURE 4 Laboratory decomposition of P. reticuhrm (a) leaves, (b) stems, and (c)
roots originating from peanut or miRët fields. Bars represent standard deviations.
iiowever, based on their size, the micropredators are the smahest invertebrates,
protozoa and nematodes of the microfauna (average size ~0.2 mm). At the next
level, litter transformers include the small Oligoc&eta Irnd~ytracidne and arthropods
_~~ -~ ~-
of the~ mesofauna (~2 mm) and macrofauna (>2 mm) that serve as incubators for
microbial activities. The third group is composed of few large invertebrates, mainly
earthworms, (Lumbricus terrestris) and social insects, e.g., ants (Formicw .sp.j and
-.-’
;,.“.Y&,&L
.-..
.-
,. . 7s.
+- L’i
2
’
_,
-
2::’ ::
z-
.,.
_
Shrub Litter Decomposition
-._ 215
l;..i:>-’ :.
- -4;;+7f -
-~._
‘l~h<Termiti.s
terrestris). Qualitative field observations indieateed~e presënte of
termites.
.---.-~
-- ---Y@t?a?ed on the screen mesh size (2 mm) at which the soif bas been sieved, our soif
could certainly contain micropredators and litter transformers as functional groups.
That may explain the loss mass loss observed under laboratory conditions as com-
pared to higher decomposition rates obtained under field conditions. Leaf residues
degraded much faster than stems or roots in the lab study. In the absence of eco-
system engineers in the lab study, high N content of the leaves may have been more
important in the iab study because decomposition was largely done by the micro-
bial community (Fig. 3a and 3b).
Ganges in Specific Surface Area-to-Mass Ratio of the Residues
Changes in the rate constant (k) developed from data shown in Fig. 5a and 5b were
highly significant for the leaves for both peanut and millet fields. For the stems, the
k value was rather low and stayed at a steady-state over time. This is duc to a
greater decomposltion rate of the leaves than the stems under iaboratory conditions.
?
Leaves
* Stems
g
/
2 0
(0” 15
??
kaves
m Stern5
1 0
0 10 20 30 40 50 60 70 80 90
Incu batiqn thm.(cjays)
FIGURE 5 Change in specific surface area-to-mass (k) value for P. reticulutum res,i-
dues originating (a) from peanut or (b) millet fields.
. .-
:
. .
.
. I;
-4:--._ ,..
i_c.__
_‘?
)
.
“--
..-- -L;.
/.,
.---.
-
..,-_
~.~+~-AyY”‘---
:
-LMGcz. i
_ .
M*
j$@p&
-MT.-
-.-.
zs++;
.S>-’
--..
-.
Values of k for P. reticulaturn -residû~~~~d--t~tween
15 x 10-” and 45 x iO--’
and are close to those obtained for most-trop residues. Stott (1994) foun,d a mean k
value of 23 x 10--5 for m&ize wlïiED?El?-(@94j
noted for sorghum, peanut, and
cotton, k values of 19 x 10-‘, 29 x 10ms and 48 x 10mms respectively. The use of a
specific surface area-to-mass ratio is important for developing soi1 management stra-
tegies.
Indeed, an increase in k during decomposition indicates that the rate of organic
matter input into soi1 is changing which should mean an increase in soi1 quali ty and
greater protection against erosion. An increase in k over time means a fast mass loss
of the residue with a relatively slow decrease in specific surface area. When k
changes slowly over time, this means that not only mass loss is nonsigmfiçar~t,
duc
in part to the quality of the residue, but also because the specific surface area
changes little. Since we placed the residues just under the soit surface, previous work
(Stott 1994) would indicate that decomposition rates would be the same il‘ we placed
the residues on surface. Thus when k values are low, the soi1 surface wilE, provide a
good caver and therefore better protection from erosion. The surface area-W-mass
relationship cari be closely related to th.e !.Q& of residues and, considerable decom-
position of mass may occur before a large decrease in area is tneasured (Steiner,
Schomberg, and Morrison 1993).
For residues having a high proportion of 1ea.f material, there may IX tremen-
dous loss in mass with Little loss in caver, because leaf material decomposes rapidly
and is light compared to stem material (Stott 19933. Stems Will lose mass, not surface
area. Soi1 caver cari be achieved by an optimal quantity of biomass required to
caver the soil, a quantity based on the qualitylevel of that biomass because the Ioss
of residue mass and surface caver are simultaneously a function of the quality of
biomass. For the type of degraded soi1 in Senegal, soi1 caver is critical &Cause the
long dry season of eight to nine months. As one assures a soi1 caver with 1’. r’c’ti-
c~&ztum residues, one cari accumulate organic material in the soit and thereby
increases soi1 organic matter over time.
Perspectives
Under Gtld conciitlûz, P. rericulurutn iost u p tel 8b-
“;D GÎ 1l.S rt’siCilL iliLS>. 1 hi>
decomposition rate was higher than that under laboratory conditions. This might be
explained by the role of soi1 fauna, which may have been*involved
under &:Id condi-
tions but were excluded under laboratory conditions. The study suggests an inl,por-
tant role that the large soi1 invertebrates play in decomposition. The resul-ts show&
that a11 components of P. ~tic~rlutum are rapidly decomposed under Licld conditions
and should be a good source of organic matter for restoring degraded soils. It
follows that addition of these residues would stimulate microbiat activity and
improve soi1 structure. By-products of decomposition which favor the stable organic
matter are important for forming aggreggafes;‘thuS
improving S;>i1 quality. This
would aid water infiltration and retention. The rapid decomposition of residues
offers the potential for farmers to stop burning these residues because high lekel of
residues Will not likely accumulate and cause interference with tillage and planting
operations. However, further studies are needed on the role of soi1 fauna in decom-
position, mineralization of nutrients from these residues, and poten.tial to incotpor-
ate residues into the system without burning.
References
~~
Bernard-Reuersat,
F. 1987. Incorpratinn de !a !itière-8 la matie’re organique du sol dans des
peuplements forestiers naturels et plantés du Sénégal, Pedolobioloyk 30(6):401-417.
Breman, H., and J.-J. Kessler. 1995. Woody plants in agro-ecosy.stm~ (I/ ~wr-~~rid rtv~ions.
Berlin: Spring-Verlag.
i
Chopa& 3. &.;~~%&Oll, an4.d. Vachaud. 1979. Le travail dusoil et le mulch pada&ii:
ences comparees SUC I’éau dans Ie systéme arachide mil au Sénégal. In Isotopes &-ru&-
ution in. rdwn soll-plant relationships, eds. International Atomic Energy-Agen+-
(AIEA), 199-221. Vienna, Austria: AIEA-SM.235122.
~O~IS, H. P., L. P. Elliott, R. W. Rickman, D. F. Bezdicek, and R. 1. Papendick. 1990.
Decomposition and interactions among wheat residue components. Soi/ Science Sociery
of America Journal 54:780-785.
Davis, J. S., and J. E. Gander. 1967. A re-evalualion of the Roc procedure for the determi-
nation of fructose. Analytical Biochemistry 19: 72-79.
Diack, M. 1994. Residue decomposition for cotton, p~~nur urlrl sorghum. M. S. thesis. Purdue
IJniversity, West Lafayette, Indiana.
Diack, M., N. A. Badiane, M. Sène, M. Diatta, and R. 1’. Dick. 1998. Cordylu pinnnta en
association avec Piliostigma reficulntum: effet sur la restauration des sols dégradés au
Sénégal. Rapport scien@que jînal Li&on Internutiode (if’ Recherche. Dakar, Senegal:
ISRA.
Dommergues, Y. 19.56. Influence de l’engrais vert sur l’activité biologique du sol. In Comptes
rendus du 6’ congrès Internationul de la Science du Sol, Vol : C, 389-392. Paris.
Felier, C., and F. Ganry. 1982. Décomposition et humification des résidus végétaux dans un
agro-système t.Topical. III-Effet du compostage et de l’enfouissement de divers résidusde
rbcolte sur la répartition de la matière organique dans différents compartiments d’un sol
sableux. Agronomie Tropicale 37(3):262-269.
l,eller, C., J. L. Chopart, and C. Dancette. 1987. Effet dc divers modes de restitution des
pailles de mil sur le niveau et la nature du stock organique dans deux sols sableux
tropicaux (Sénégal). Cuhiers ORSTOM, série P&loloyir 23: 237-252.
Cjanry, F., and G. Guiraud. 1979. Mode d’application du fumier et bilan azoté dans un
système mil-sol sableux du Sénégal: étude au moyen de l’azote. In Isotopes and radiation
111 reseurch on SOI~-plant relationships. procetdInys of on Illrernrctional Symposium on the
Use 01 Isotopes and Radiation in Resrarch 011 Soi/-Plu!~t Kc[lltiorlships, eds. International
Atomic Energy Agency (AIEA), 445-464. Vienna, Austria: AIEA-SM.235/22, AIEA.
(ianry, F., and P. L. Sarr. 1985. Gestion des résidus de récolte et économie de l’azote au
Sénégal, Etudes techniques du CNRA de Bumh~~~~, Dot. No. lOOj83, 20. Bambey, Senegal:
ISRA.
(iocring, H. K., and P. J. van Soest. 1970. Forczye jiht,r rmdvsis. ARS U.S. Department of
Agriculture Handbook No. 379. Agricultural Research Service, USDA. Washington,
D.C. : US. Government Printing Office.
:;reg’.xry, J. M.. T. R. McCurty, F. Ghldc;, 2nd E. E. Alberts. 1%‘. Dcïi~iitiou iitid z4aluat;on
of a residue decay equation. Trcmsucr ion.~ qf A rwri~,~~n Soc%~r~~ oj Ayrr~wltural Engineers
28:98--101.
Jenkinson, D. S., and J. N. Ladd. 1981. Microbial biomass in soil: measurement and turnover.
In Soi1 hiochemistry, vol. 5, eds. J. N. Ladd and E. A. Paul. 41 S-471. New York: Marcel
Dekker.
Jensen, E. S. 1994. Mineralization-immobilization of nitrogen in soi1 amended with low CflU
ratio plant reside:s with different particles sizes. Soi/ Bioloq~~ trrd Biochvnistry 26: 519~-
521.
Knapp, E. B., L. F. Elliott, and G. S. Campbell. 1983. Microbial respiration and growth
during the.deG.omlzosition of wheat straw. Soi/ Bioloy!~ cmd Bioclwrnistr!. 15:3 19-323..
Lav&e, P. 1996. Diversit:y of soi1 fauna and ecosystem funçtion. Hiohq,v Imert~c~tiond 33:3-
16.
Myrold, D. D., L. F. Elliott, R. 1. Papendick, and G. S. Campbell. 1981. Water-potential
water content characteristics of wheat straw. Soi[ Scitw~c Socieiy oj‘ Americu Journal
45:329-333.
Parr, J. F., and R. 1. Papendick. 1978. Factors affecting the decomposition of trop residues by
micro-organisms, In Crop residue manugement systems, ed. W. R. Oswald, 101-129.
Madison, Wisconsin: American Society of Agronomy.
Pichet, J., B. Truong, and P. Beunard. 1979. L’action des phosphates naturels d’Afrique de
l’Ouest sur l’absorption du phosphore par l’agrostis et sur le phosphore isotopiquement
diluable (valeur L*) dans cinq sols tropicaux, In Isotopes and radiution in research on
snifipiurli r~ictiiorish@: Proceedings of‘~n intrrnationat Sy»lposium oil fhr Use of-isotopes
(2nd Radiation in ~!esenrclz on Soi/-Plant Rekztionships, eds. International Atomic Energy
Agency (AIEA), 463-477. AIEA-SM.235122. Vienna, Austria: AIEA.
., -.--.-.
-..-
L.
1.
t.
M. Diack et a1-._ &-Iv-.-
1.1
. ..-.
-_._.
.Y. X-T :.i > ..z
_,y3c-yj
J ..~.
.___ ~___. - -
Sarkanen, K. V., and C. H. Ludwig. 1971. Z,ignins: occurrence, formation, .rtrn~tt~re an<l rea;‘-
rioy. New York: Wiley-Interscience.
Seck, M. A. 1987. Le compostage des coques d’arachide: une technique permettant a”amt;lior~~r
la qua&& des sols sahéliens. Thèse de 3e cycle, Université de Nancy 1, Nancy, France
Steiner, J. L., R. E. Schomberg, and J. E. Morrison. 1993. Kesidue decomposition and redistri-
bution, In Crop residue management technology transfer, Southern Plains, ed. II. A,
Stewart, 65-76. Washington D.C.: USDA-ARS/SCS.
Stott, D. E., L. F. Elliott, R. 1. Papendick, and G. S. Campbell. 1986. Low temperature or Iou
water potential effects on the microbial decomposition of wheat residues. Soi/ Bi~!c)~/\\
and Biochemistry 18 : 577-582.
Stott, D. E. 1993. Mass and C loss from corn and soybean residues as associated with thcu
chemical composition, In American Society of Agronomy (ASA) Meeting, Proceeding~,
Abstract, 267. Madison, Wisconsin: ASA.
Stott, D. E. 1994. Changing relationship between mass and surface area of decompcosing
residues, In American Society of Agronomy (ASA) Meeting, Proccedings, Abstract, 261.
Madison, Wisconsin : ASA.
Stroo, H. F., K. L. Bristow, L. F. Elliott, R. 1. Papendick, and G. S. Campbell. 1989 Predict-
ing rates of wheat residue decomposition. Soil Science Society of Amerisa Journo~ 53:31
99.
Witkamp, M., and J. van Der Drift. 1963. Breakdown of forest litter in relation te envirorr-
mental factors. [%rlt ccnd Soi/ t 5 : 295 .3 1 1.
*-.----.
Journal of NativèanaAgnCtiltural Envirhments
A Cooperating Journal of-tbl&snationai
Union of Soi1 Science
IUSS - UISS - IBU
L
Decomposition of a Native Shrub, Pîliostigma
reticulatrcm, Litter in Soils of Semiarid Senegal
M. DIACK
M. SENE
A. N. Bz4DIANE
M. DIA’l-TA
Agricultural Kesearch Institute of Senegal (ISRA)
Senegal
R . F. DI[CK
Department of Crop and Soi1 Science
Oregon State University
Corvallis, Oregon, USA
I)etvIopinq effective management strategies that restorc dc<grad& SOI~S requires gn
evaluution of the quality of the litter residues. This study relates the chemical com-
position of the biomass components to the decomposition rates Jor Piliostigma reti-
culatum (DC.) Hochst., a native shrub, under Jield and laboratory conditions. The
ratt’b were determined by mass 10s~. The chamqcs in the specific surface ureu of the
resufrrr ii1 relation to mass loss ranged from 13 x 1O- 5 to #5 x 10 ’ which was
srmrlar tcj trop residues in other studies. At Jic4d conditions, P. reticulatum mass
ic~.s~ WI.\\ highrr (80% of the initial mass lest ovrr eight months) rhan thar under
( on::.::!M c,onditinns (50%). Such fast decornpovr;, St! of rG4rrr.; +r\\ !hr p4writrl
j<;-Jilrmrrs to stop burmng these residues be( ause hiyh amounts C$ residues Will net
likely acrxmulate and cause interferencc
with tillqtj and planting operations.
t.‘urthrr .;tudies are needed on the role oj‘ soi/ fauna un dvcomposition, mineral-
ization cv’nutrients from these residues, and thr potentialfor incorporating residues
into rhe systtm without burning.
Keywords
:soil quality, soil organic matter, litter residues, mass loss, manage-
ment strategjes
Maintaining trop resid*%n the SO~I surface is an effective meth;d of controlling
~--
soi1 erosion and degra.dation
provided that there is enough plant residue. However,
in many areas of the world, insuffkient amounts of residue are produced to provide
adequate soi1 caver and protection. In Senegal, there is little information available
on how plant resiclues, left on the soil surface, affect soi1 quality (Chopart et al.
1979). Several studies have focused on the incorporation of green manure in the soi1
Received 20 May 1999; accepted 20 Sept. 1999.
The authors wouid 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 resea& liûisûri bctneenfSRA-ancl
Oregon State University.
Address corresponience to Dr. Richard P. Dick, Department of Soil Science, 3017 Agricultural
Life Science Building, Oregon State University, Corvallis. Oregon 97331-7306, USA. E-mail:
Richard.Dick@orst.edu
205
ommergues 1956), trop residues (Chopart, Nicou, and Vacéaü&?97$%iwcnhardt-
Reversat 1987; Feller and Ganry -1982) and composting (FeUe:r and Ganry 1982;
--~
Ganry and Sarr 1985; Feller, Chopart, and Daucette 1987; Seck 1987). Most of this
work aimed to increase trop yield in thc short term, where the major contribution
cornes from input of nutrients. However, another important nonnutrient contribu-
tion of organic amendments is its role in building soil structure, organic matter and
maintaining biological activity. This is important in tmproving and sustaining long-
ter-m trop yields.
Although trop residues cari have a beneficial eifect on soils, very little of the
trop residue in Senegal and most of Sub-Sahara is returned to the soi1 because it is
too valuable as a food source for livestock (persona1 communication, A. Badiane,
ISRA, Senegal). Consequently, other organic matter sources are needed in Senegal.
Even though manure and composting are available, there are insuflïcient amounts
that cari be readily collected to apply in ail helds of a typical Senegalese farm
(Ganry and Guiraud 1979, Pichet, Truong, and Bcunard 1979).
Organic matter management is of critical importance in Senegal because of the
low fertility status of soils and the declining vields that may be due to degradation
of soils. This is likely occurring because of mcrcascs in the rural population and
associated intensification of agriculture, e.g., marked decreases in land fallowing for
regeneration (Diack et al. 1998). In the absence of appropriate residue management
technologies, the parkland system with optimal tree or shrub species, or both, may
be an alternative approach for restoring degraded soils. Currently, there is Iittle
information available on how traditional shrubs of the parkland systems improve
degraded soils.
Piliosrigma reticulcttum (DC.) Hochst (Breman and Kessler 1995), a shrub
species, is commonly found in the major peanut and millet growing regjons in
Senegal. P. reticulatum is a nonnodulating legume and could be considered as a
“ratoon caver trop” after the cropping season. In farmers’ fields, the aboveground
biomass of P. reticulatum is tut at the soi1 surface and burned just prior to the rainy
season. As it regrows during the rainy season, it is tut again at every weedmg. After
the growing season, it regrows to hcights of 0.70 to 1 .OO m and canopy diameters of
0.75 to 1.75 m and has rooting depth of 2.5 !o ! tn with as many as 700 shru.bs ha- ’
(Diack et ai. 1998). P. reticulatrrn~ biomass;
Lauld bu: dli Important organic matter
source for soils in Senegal with preliminary estimates of >0,5 Mg ha-’ above-
ground dry biomass (Diack et al. 1998). However, to use these materials rather than
burning them, it would be important to dctermine the litter quality and rate of
decomposition to effectively manage it from an agroromic perspective. Since P. rcri-
culutum is a Woody species, it may decompose too slowly and interfere with trop
cultural activities. Therefore, understanding how rapidly P. retic&zturn
residues
decompose and are lost from a field site, is a prerequisite to the design of effective
management strategies for soi1 improvement.
The objectives of this study were: (1) to determine decomposition ratés for I>.
reticulatum aboveground residues and roots by mass loss under field and laboratory
conditions; (2) to determine the relationships between chemical and physical charac-
teristics of the residues and their rates of decomposition; (3) to determine whether
the previous trop affects the residue quality as measured by chemical analysis and
rates of decomposition; and (4) to provide guidance for future studies-on manage-
ment strategies of P. reticulatum in farmers’ fields to improve the quality of the soils.
Materiais and Methods
S i t e
The experimental site was located in Paoskoto, Kaolack, a semiarid agroecological
zone in the peanut basin, between 13” 35’ and 14” 30’ of northern latitude and 14”
-...i ‘. 1m
a
.y ‘c-J::,$-.
_-.
. +-~~,~~wyw=
__
- == .@p&:~~:<---
ï ,..-_ _ ..L :
-: - Shrub ~t$er, ‘Decotiposition
207
--.-.-.
-- ,-&& -,-.... r
35’ and 16:’ 45’ of western long$~&egion is characterized by a tropical Suda-
nian climate with an annual rainfall of700 mm. Temperatures vary between 16°C in
I>ecember -January and 39°C inÂpÏÏE:-
A Deck Dior loamy-sand (fine-sandy, mixed Haplic Ferric Lixisol), leached ferru-
gcneous tropical $,oil (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
undcr millet (Pcrwisrrum glaucum L). The soil was air-dried, crushed to pass 2 mm
mesh screen. then stored until use. This field and soi1 was used for subsequent field
and lab incubation studies.
Plant MateriaIs
The 1’. rctirulatur~~ materials were randomly collected from each of two farmcr’s
fields which are under a 2-year rotatîon-of peanut (Aruchis hJlpogrtea, L.) and millet.
One field was under peanut and the other under millet the previous year before
collecting thc residues. P. reticuLztum residues were separated into different residue
types: leaves, stems, and roots.
Chcmical Analysis ami Mass Ratio of Plant Residue
The relative mass ‘ylias 17%, 26%, and 56% for leaves, stems, and roots, respectively.
The relative mass of aboveground biomass for leaves and stems was 40% and 60%,
respectively. Each plant residue component was analyzed for total C and total N
contents by dry combustion (Mode1 CHN-600; Leco Corp.). Fructose was deter-
mined spectrophoto-colorimetrically (Davis and Gander 1967). Cellulose, hemi-
cellulose and lignin contents were measured by sequential fiber analysis (Goering
and van Soest 1970).
Mass LOS.~ Experinzent
under Laboratory Condition.~
The experiment consi.sted of a randomized complete design with two trop rotations,
and three residue treatments with three replications. The three treatments wcre
leavcs, stems, and roots. Stems were tut into pieces of 4 to 5 cm long and 5 g of
residue were spread out on 100 g air-dry soi1 in a 15 x 9 cm polystyrene dish and
then covered with another 100 g of air-dried soi]. The soi1 was brought to a mois-
turc content of -- 33.3 kPa water potential which equalled 60% water holding
capacity, plus 300% of the residue mass (Myrold et al. 1981). After addition of the
appropriate amount of water, the inctibEEn’clish was covered with a perforated cap
to aliow aeration. l”he. samples were incubated at 30°C + 1°C.
Samples were withdrawn on day 3, 7, 14, 28, 56, and 84 of the incubation for
mass loss measure.nent. At each destructive sampling, the incubation mixture was
oven-dried at 5o”C, for 48 h. When dry, the residues were carefully separated from
the soi], gently washed to remove the soil particles, and put back into the oven at
TABLE 1 Soi1 physical and chemical characteristics
-~
Clay
Silt
Sand
Total organic C
Total N
Snil --
--.
-
--__
---
sample
P H
g.kg-'
0 -10 cm
10.1
5.2
84.7
6.7
4.68
0.45
--.- -.--..
__- -_. . __ - _,-.. ^
-..
. .x:7.;: ,... L -.
50°C for 48 h. T&%%&&$e&veighed,
then piaced in crucibles for asfiing~ “‘.
500°C for 3 h, and reweighed after ashing. The residue remaining after ashing-wv
assumed to be of ao~-rïïmëraraloti$n, and this was subtracted from rhe weight of the
pre-ashed plant residue.
Mass L.oss Experiment
under Field Conditions
Leaves, stems, and roots constituted the three residuc types for this experiment. One
hundred g of air-dried residues were placed into a 25 x 25 cm bags (2 mm mesh)
and sealed as litter bags. The samples were incorporated (3 July 1997, beginning of
rainy season) into the top 10 cm of a field in a 3 x 2 factorial experiment with three
1’. rcticuhum residue types (stems, leaves, and roots) and two P. rrriculatum residue
sources (from peanut or millet field) with three replications. This field was previously
under millet (source of P. reticulatum) but was under peanut when buried bags were
in the field.
Samples were withdrawn monthly and eight months afler placement of residue
bags into the soil. At each destructive sampling, the residues within the litter bags
were oven-dried at 50°C for 48 h. When dry, the rcsidues were taken off from the
litter bags and processed in the same way as the iaboratory experiment to determine
the mass 10s~.
Soi1 water potential during the season was -- 1000 MPa in .July, 1997, then
increasing to its highest level in September at ---OC, MPa, and then decreasing to
-50 MPa in October, followed by a steady decline to
120 MPa b y F’ebruary,
1998.
, . .
5pcc$k Sarface Area-to-Mass Ratio
Spccihc surface area-to-mass ratios were represcutc~l
hy a k value uith a dimension
of ha kg ’ of residue. Specific surface amas for I~~;L~L; antl >.t+,,; we:r measzuec!
using a digital-planimeter (Mode1 Ottplan 7(Q’71ii, A. UT?‘ GMBH, Kempten,
Germany). As decomposition proceeds, the ratio between the specific surface area
ami the mass remaining was calculated over time The equation used to çonvert
residue mass to surface caver (Gregory et al. 1985) is expressed ;as:
where C is thc fraction of the surface caver remalning; m is the mass (kg ha- ‘) of
residue present on the surface; and k a constant. The constant k was derived from
the following equation :
k = -Iog(l -- C)/m.
Statistical Analysis
Statistical analysis of the data wasdone to determine differences
amongtreatmcnts,
..-..
using MSTAT-C, Version 1 (MSTAT-C, 1991, MSU, Ml). Comparisons between
treatment means were computed at the P = 0.05 and I’ = 0.01 levels, using the
Duncan’s multiple range test procedure.
..-_”
_.? .---
TABLE 2 Initial chemical composition of Piliostigma reticulutunn residues collected from a peanut or
millet field
Total C
Total N
Fructose
Cellulose
Hemicellulose
Lignin
Previous
R e s i d u e - -
field trop
type
gkg- *
_-
Peanut
Leaves
483b*
13.8a
18.7c
241~
22c
242b
Stems
536a
7.5b
42.3a
448a
62b
22oc
Roots
529a
6.3b
30.4b
382b
96a
264a
Millet
Leaves
474c*
12.3a
2O.lc
202c
5oc
247b
Stems
500b
7.7b
40.8a
379a
90b
228c
Roots
545a
7.0b
37.7b
420b
123a
2 9 4 a
.--
I
* In each column. values followed by the same letter, are not significantly different by the Duncan’s multiple range /
test at P = 0.05.
l
Results and Discussion
initial Chenrical Composition for the P. reticulatum Residues
The mean concentrations of total C and N, fructose, hemicellulose, and lignin (Table
2) were signifkantly different between P. reticulatum
residue types as expectcd.
However, the prcvious trop generally had a significant (I’ < 0.05) elfect <on the
chemical properties for most residue types. The analysis of the structural componers.
cellulose, hemicellulose, and lignin, although showing some differences
in means
between the field source of the shrub residue these were not seatistically different
when comparing each individual residue (Table 2).
Changes in Mass LOS~ under Fieid Conditions
Under field conditions, 82% of the roots (Fig. la) was lest during an &month
period whereas 78% of stem residues was lost and only 40% cf the leaves disap-
peared for thc residue that originated from a tîeld where peanut was pervious trop.
During the first month of decomposition, the stem mass loss was greater (- 42%)
9 0
Leaves
8 0
Stems
Roots
7 0
60
Y-,
F
50
b-
--
??
?
Leaves
??Stems
0 Roots
Ju97 A u S e OC No Dë Ja98 Fe Ma
Period of
__-.- _--- incubation.(nlonths)
-__ --
--.
FIGURE 1 Decomposition of P. reticulutum residues that originated from a peanut
field (a) or millet field (b). under field conditions. Bars renresent standard deviations.
\\
,,
r---
!
- .--.-----T----Shrub Litter L?ccompo.yitiort
A..
7”
8 baves from peanut field
o baves from millet field
80 C AT
70
60
50
40
30
m Stems from peanut lïelc
0 Stems from millet tïeld
90
Roots from peanut tïeld
80
Roots from millet h-Id
70
60
50
40
30
20
10
01
Ju97 Au Se Oc No De Ja98 Fe Ma
Period of incubation (months)
FIGURE 2 Field decomposition of P. reticulatum (a) leaves, (b) stems, and (c) roots
originating from either peanut or millet fields. Bars represent standard deviations.
than that for the roots (-- 28%) and the leaves (- 25%). P. reticulntum residue stems
and roots originating from the millet field had equal cumulative ~s.lasses of 82th -
(Fig. lbj cumpared to~leaves whï~h~had~?~% losses. Mer one month, stems and
roots lest 42% of their mass, while the leaves lost only 32%.
The previous trop affect (Fig. 2a) was a significant (P < 0.05) of the origin of leaf
residue with a 70% loss of leaves from the millet field compared to 60% loss of
100
90
80
70
60
50
40
30
?? baves
20 (4
?? Stern8
10
o Rools
90
80
70
60
50
40
?? Leaves
30
’ S t e m s
20
. w
o Roots
10
0
I
I
I
I
I
I
I
I
0
10 20 30 40 50
60 70 SO
90
hcubation tinte (daysj
FIGURE 3 Laboratory decomposition of 1’. rt~ti’c~htwn residues, originating L
from peanut field or (b) millet field. Bars represent S;tandard deviations.
leaves from the peanut field. No significant difference was noted (Fig. 2b) betwe::
loss of stem mass originating from the millet field (82%) or the pcanut tîeld (79”:
Root mass loss also (Fig. 2c) was not significantly different from millet fi.eld (83’ i
and peanut field (80%).
Under field conditions, the decomposition rates for a11 the P. reticufatm reo.-
dues (Fig. la, lb, 2a, 2b, and 2c) followed an exponential 3-phase pattern. The rapt
mass loss during the first month was probably due to a high total N content, a hi5
level of readily available C in the fonn of extractable sugars or, a combination .:;
both. Kinetically, the mass loss from the residues exhibited a linear dependence c:
the chemical composition of the residue during this period. The rapid disalppearancr
of these soluble compounds was probably due to a quick build up of the microblé
activity, which would incresse tb? rnacq 1~s. 4lso, the readily available C and !i
components in the trop residues might provide the initial energy and nutrienu
necessary to activate the microorganisms that are responsible for the degradation 6
.theless readily available components of the residue.
The leveling off phase for mass loss occurred from two to four months whic:
is likely when mainly recalcitrant materials remains. During this period, hemicelluloi-.
is likely the main fraction available to the microorganisms (Collins et al. 1,990; Star.
-.-
;
-‘y.--
m
::,<re.
3:;
-
“7-~~-~.!-.~,..“~y--‘.
-
...--*,.~“~~~;r~~~~~~~
e%g ,1 <-.
~ ,_- A~.-&. -2..
.-
---Shr&
r,r”
-orI
213
-
-,f q..; .I? ;.:.. :,.; i
_, &~~&.C&. i __
1993). As the decomposition protiX:-Erate of mass loss slows down, fohowing
an exponential curve., probably_.~..~~~alcitrance
of the remaining residue.
After four months of decomposition, the remaining residues entered a third phase
during which the slowly available residue components dominated in the substrate.
Lignin, which is one of the most resistant plant compounds, was probably the major
remaining componcnt during this phase of dewmposition.
The rate of iignin degradation is less affected by changes in temperature and
nitrogen availabiliiy than other major plant components during decomposition
(Sarkanen and Lucwig 1971). Another factor for slow decomposition is that lignin
decomposers are relatively slowly growing microorganisms (Witkamp and van Der
Drift 1963). As restilt, the rate of mass loss follows a steady-state for the test of the
decomposition proc:ess.
Root and stem residues from the peanut field (Fig. la) have undergone greater
cumulative mass loss than the leaves. This is probably due to higher fructosc
content in the roots and stems than in the leaves of P. reticulatutn.
The physical and
chemical compositi’on
of the residues and, particularly soluble C in the form of
fructose, constitute the most important--factors controlling decomposition rates
(Knapp, Elliott, and Campbell 1983). Residue decomposition is a process in which
the rate transforrna:ion is proportional to the qualitative amount of residue avail-
able to thc rnicroorganisms. Our decomposition patterns (exponential decreases)
were sirnilar to other decomposition studies and Papendick (1978) Knapp et al.
(I 983), Stott et al. (I986), Stroo et al. (1989), Collins et al. (1990) and Jensen (1994).
C ‘hunps in Mass Lo.>s undcr Laboratory Conditions
Thc 1’. tc>ricrrl~r~r>l leaves that originated from the peanut field (Fig. 3a) lest ;i
greater amount of mass (50%) than the roots (35%) and the stems (24%) in eight
months. During the first month, leaf residue mass loss was significantly different
( -28%) than both sterns and roots (-- 13%). Leaf mass loss was 48% whereas for
the roots and the strms originating from the millet field, mass Ioss was only 30%
(l‘if. 3b). l‘he first ml>nth of decomposition showed a greater mass loss for the leavet;
i
?SfLl than cither <item< or roots (- 15%) for residues of the mi!le: field.
CiiijkG the iield study, no significant effect on the residue cumulative mass loss
was notcd as a funct: on of the source of residue. Leaf residue mass loss from peanut.
lield was 5 1% while tha.t from the millet field was 49% (Fig. 4a). Mass loss of stems
from the peanut field WBS 25% which was not significantly different from that of the
millet tîeld (28”/0) (Fig. 4b). Root mass loss from the peanut field was 35% while that
from the millet field was 30% (Fig. 4~).
Under laboratory conditions, the same pattern of rates of decomposition was
obscrved as the held study. However, the decomposition rate of the leaves seemed
greater than the stems or roots, which is the opposite of the field data. This may be
‘----‘Y--
relatcd to several factors such as prehandlmg of the soi1 by air drying and sieving
the soi\\ in the lab strdy. Another factor is that lab conditions excluded large fauna.
Morcover, the soi1 is sandy and has very low total C and N contents (Table 1).
Hecuuse of this 10~1 fertility level, soi1 microorganisms are probably relatively
dormant. The discont:inuous
distribution of organic resources in soils causes soils to
have a large dormant population, which is ri& in species and has an ability to
survive stress (Jenkinson and Ladd 1981). According to Lavelle (1996) this resting
tirne v.4~ !ast rrom months to years.
Therefore, redistribution of microorganisms and organic substrates
by
invertebrates and roots that mix the soi1 and add water and readily available sub-
strates to the soi1 may beimgorta n t in activa&-microbiaf
activity and decomposi-
tion. These invertebrates could be divided into three major functional groups:
micropredators, litter transformers, and ecosystem engineers (Lavelle 1996). We did
net check whether any of these major functional groups were present in the soil.
214
?? Leaf (Peanut as previous trop
60
50
40
30
20 - ?? Stern (Peanut as previous trop
10 _ EI Stern (Millet as previous trop)
50
40
30
20 t + Root (Peanut as previous trop
10
Q Root (M%et as previous trop)
o__
~.
’
’
.‘.
’
’
’
’
’
0
10 20 30 40 50 60 70 80 90
Incubation time (days)
FIGURE 4 Laboratory decomposition of P. reticuhrm (a) leaves, (b) stems, and (c)
roots originating from peanut or miRët fields. Bars represent standard deviations.
iiowever, based on their size, the micropredators are the smahest invertebrates,
protozoa and nematodes of the microfauna (average size ~0.2 mm). At the next
level, litter transformers include the small Oligoc&eta Irnd~ytracidne and arthropods
_~~ -~ ~-
of the~ mesofauna (~2 mm) and macrofauna (>2 mm) that serve as incubators for
microbial activities. The third group is composed of few large invertebrates, mainly
earthworms, (Lumbricus terrestris) and social insects, e.g., ants (Formicw .sp.j and
-.-’
;,.“.Y&,&L
.-..
.-
,. . 7s.
+- L’i
2
’
_,
-
2::’ ::
z-
.,.
_
Shrub Litter Decomposition
-._ 215
l;..i:>-’ :.
- -4;;+7f -
-~._
‘l~h<Termiti.s
terrestris). Qualitative field observations indieateed~e presënte of
termites.
.---.-~
-- ---Y@t?a?ed on the screen mesh size (2 mm) at which the soif bas been sieved, our soif
could certainly contain micropredators and litter transformers as functional groups.
That may explain the loss mass loss observed under laboratory conditions as com-
pared to higher decomposition rates obtained under field conditions. Leaf residues
degraded much faster than stems or roots in the lab study. In the absence of eco-
system engineers in the lab study, high N content of the leaves may have been more
important in the iab study because decomposition was largely done by the micro-
bial community (Fig. 3a and 3b).
Ganges in Specific Surface Area-to-Mass Ratio of the Residues
Changes in the rate constant (k) developed from data shown in Fig. 5a and 5b were
highly significant for the leaves for both peanut and millet fields. For the stems, the
k value was rather low and stayed at a steady-state over time. This is duc to a
greater decomposltion rate of the leaves than the stems under iaboratory conditions.
?
Leaves
* Stems
g
/
2 0
(0” 15
??
kaves
m Stern5
1 0
0 10 20 30 40 50 60 70 80 90
Incu batiqn thm.(cjays)
FIGURE 5 Change in specific surface area-to-mass (k) value for P. reticulutum res,i-
dues originating (a) from peanut or (b) millet fields.
. .-
:
. .
.
. I;
-4:--._ ,..
i_c.__
_‘?
)
.
“--
..-- -L;.
/.,
.---.
-
..,-_
~.~+~-AyY”‘---
:
-LMGcz. i
_ .
M*
j$@p&
-MT.-
-.-.
zs++;
.S>-’
--..
-.
Values of k for P. reticulaturn -residû~~~~d--t~tween
15 x 10-” and 45 x iO--’
and are close to those obtained for most-trop residues. Stott (1994) foun,d a mean k
value of 23 x 10--5 for m&ize wlïiED?El?-(@94j
noted for sorghum, peanut, and
cotton, k values of 19 x 10-‘, 29 x 10ms and 48 x 10mms respectively. The use of a
specific surface area-to-mass ratio is important for developing soi1 management stra-
tegies.
Indeed, an increase in k during decomposition indicates that the rate of organic
matter input into soi1 is changing which should mean an increase in soi1 quali ty and
greater protection against erosion. An increase in k over time means a fast mass loss
of the residue with a relatively slow decrease in specific surface area. When k
changes slowly over time, this means that not only mass loss is nonsigmfiçar~t,
duc
in part to the quality of the residue, but also because the specific surface area
changes little. Since we placed the residues just under the soit surface, previous work
(Stott 1994) would indicate that decomposition rates would be the same il‘ we placed
the residues on surface. Thus when k values are low, the soi1 surface wilE, provide a
good caver and therefore better protection from erosion. The surface area-W-mass
relationship cari be closely related to th.e !.Q& of residues and, considerable decom-
position of mass may occur before a large decrease in area is tneasured (Steiner,
Schomberg, and Morrison 1993).
For residues having a high proportion of 1ea.f material, there may IX tremen-
dous loss in mass with Little loss in caver, because leaf material decomposes rapidly
and is light compared to stem material (Stott 19933. Stems Will lose mass, not surface
area. Soi1 caver cari be achieved by an optimal quantity of biomass required to
caver the soil, a quantity based on the qualitylevel of that biomass because the Ioss
of residue mass and surface caver are simultaneously a function of the quality of
biomass. For the type of degraded soi1 in Senegal, soi1 caver is critical &Cause the
long dry season of eight to nine months. As one assures a soi1 caver with 1’. r’c’ti-
c~&ztum residues, one cari accumulate organic material in the soit and thereby
increases soi1 organic matter over time.
Perspectives
Under Gtld conciitlûz, P. rericulurutn iost u p tel 8b-
“;D GÎ 1l.S rt’siCilL iliLS>. 1 hi>
decomposition rate was higher than that under laboratory conditions. This might be
explained by the role of soi1 fauna, which may have been*involved
under &:Id condi-
tions but were excluded under laboratory conditions. The study suggests an inl,por-
tant role that the large soi1 invertebrates play in decomposition. The resul-ts show&
that a11 components of P. ~tic~rlutum are rapidly decomposed under Licld conditions
and should be a good source of organic matter for restoring degraded soils. It
follows that addition of these residues would stimulate microbiat activity and
improve soi1 structure. By-products of decomposition which favor the stable organic
matter are important for forming aggreggafes;‘thuS
improving S;>i1 quality. This
would aid water infiltration and retention. The rapid decomposition of residues
offers the potential for farmers to stop burning these residues because high lekel of
residues Will not likely accumulate and cause interference with tillage and planting
operations. However, further studies are needed on the role of soi1 fauna in decom-
position, mineralization of nutrients from these residues, and poten.tial to incotpor-
ate residues into the system without burning.
References
~~
Bernard-Reuersat,
F. 1987. Incorpratinn de !a !itière-8 la matie’re organique du sol dans des
peuplements forestiers naturels et plantés du Sénégal, Pedolobioloyk 30(6):401-417.
Breman, H., and J.-J. Kessler. 1995. Woody plants in agro-ecosy.stm~ (I/ ~wr-~~rid rtv~ions.
Berlin: Spring-Verlag.
i
Chopa& 3. &.;~~%&Oll, an4.d. Vachaud. 1979. Le travail dusoil et le mulch pada&ii:
ences comparees SUC I’éau dans Ie systéme arachide mil au Sénégal. In Isotopes &-ru&-
ution in. rdwn soll-plant relationships, eds. International Atomic Energy-Agen+-
(AIEA), 199-221. Vienna, Austria: AIEA-SM.235122.
~O~IS, H. P., L. P. Elliott, R. W. Rickman, D. F. Bezdicek, and R. 1. Papendick. 1990.
Decomposition and interactions among wheat residue components. Soi/ Science Sociery
of America Journal 54:780-785.
Davis, J. S., and J. E. Gander. 1967. A re-evalualion of the Roc procedure for the determi-
nation of fructose. Analytical Biochemistry 19: 72-79.
Diack, M. 1994. Residue decomposition for cotton, p~~nur urlrl sorghum. M. S. thesis. Purdue
IJniversity, West Lafayette, Indiana.
Diack, M., N. A. Badiane, M. Sène, M. Diatta, and R. 1’. Dick. 1998. Cordylu pinnnta en
association avec Piliostigma reficulntum: effet sur la restauration des sols dégradés au
Sénégal. Rapport scien@que jînal Li&on Internutiode (if’ Recherche. Dakar, Senegal:
ISRA.
Dommergues, Y. 19.56. Influence de l’engrais vert sur l’activité biologique du sol. In Comptes
rendus du 6’ congrès Internationul de la Science du Sol, Vol : C, 389-392. Paris.
Felier, C., and F. Ganry. 1982. Décomposition et humification des résidus végétaux dans un
agro-système t.Topical. III-Effet du compostage et de l’enfouissement de divers résidusde
rbcolte sur la répartition de la matière organique dans différents compartiments d’un sol
sableux. Agronomie Tropicale 37(3):262-269.
l,eller, C., J. L. Chopart, and C. Dancette. 1987. Effet dc divers modes de restitution des
pailles de mil sur le niveau et la nature du stock organique dans deux sols sableux
tropicaux (Sénégal). Cuhiers ORSTOM, série P&loloyir 23: 237-252.
Cjanry, F., and G. Guiraud. 1979. Mode d’application du fumier et bilan azoté dans un
système mil-sol sableux du Sénégal: étude au moyen de l’azote. In Isotopes and radiation
111 reseurch on SOI~-plant relationships. procetdInys of on Illrernrctional Symposium on the
Use 01 Isotopes and Radiation in Resrarch 011 Soi/-Plu!~t Kc[lltiorlships, eds. International
Atomic Energy Agency (AIEA), 445-464. Vienna, Austria: AIEA-SM.235/22, AIEA.
(ianry, F., and P. L. Sarr. 1985. Gestion des résidus de récolte et économie de l’azote au
Sénégal, Etudes techniques du CNRA de Bumh~~~~, Dot. No. lOOj83, 20. Bambey, Senegal:
ISRA.
(iocring, H. K., and P. J. van Soest. 1970. Forczye jiht,r rmdvsis. ARS U.S. Department of
Agriculture Handbook No. 379. Agricultural Research Service, USDA. Washington,
D.C. : US. Government Printing Office.
:;reg’.xry, J. M.. T. R. McCurty, F. Ghldc;, 2nd E. E. Alberts. 1%‘. Dcïi~iitiou iitid z4aluat;on
of a residue decay equation. Trcmsucr ion.~ qf A rwri~,~~n Soc%~r~~ oj Ayrr~wltural Engineers
28:98--101.
Jenkinson, D. S., and J. N. Ladd. 1981. Microbial biomass in soil: measurement and turnover.
In Soi1 hiochemistry, vol. 5, eds. J. N. Ladd and E. A. Paul. 41 S-471. New York: Marcel
Dekker.
Jensen, E. S. 1994. Mineralization-immobilization of nitrogen in soi1 amended with low CflU
ratio plant reside:s with different particles sizes. Soi/ Bioloq~~ trrd Biochvnistry 26: 519~-
521.
Knapp, E. B., L. F. Elliott, and G. S. Campbell. 1983. Microbial respiration and growth
during the.deG.omlzosition of wheat straw. Soi/ Bioloy!~ cmd Bioclwrnistr!. 15:3 19-323..
Lav&e, P. 1996. Diversit:y of soi1 fauna and ecosystem funçtion. Hiohq,v Imert~c~tiond 33:3-
16.
Myrold, D. D., L. F. Elliott, R. 1. Papendick, and G. S. Campbell. 1981. Water-potential
water content characteristics of wheat straw. Soi[ Scitw~c Socieiy oj‘ Americu Journal
45:329-333.
Parr, J. F., and R. 1. Papendick. 1978. Factors affecting the decomposition of trop residues by
micro-organisms, In Crop residue manugement systems, ed. W. R. Oswald, 101-129.
Madison, Wisconsin: American Society of Agronomy.
Pichet, J., B. Truong, and P. Beunard. 1979. L’action des phosphates naturels d’Afrique de
l’Ouest sur l’absorption du phosphore par l’agrostis et sur le phosphore isotopiquement
diluable (valeur L*) dans cinq sols tropicaux, In Isotopes and radiution in research on
snifipiurli r~ictiiorish@: Proceedings of‘~n intrrnationat Sy»lposium oil fhr Use of-isotopes
(2nd Radiation in ~!esenrclz on Soi/-Plant Rekztionships, eds. International Atomic Energy
Agency (AIEA), 463-477. AIEA-SM.235122. Vienna, Austria: AIEA.
., -.--.-.
-..-
L.
1.
t.
M. Diack et a1-._ &-Iv-.-
1.1
. ..-.
-_._.
.Y. X-T :.i > ..z
_,y3c-yj
J ..~.
.___ ~___. - -
Sarkanen, K. V., and C. H. Ludwig. 1971. Z,ignins: occurrence, formation, .rtrn~tt~re an<l rea;‘-
rioy. New York: Wiley-Interscience.
Seck, M. A. 1987. Le compostage des coques d’arachide: une technique permettant a”amt;lior~~r
la qua&& des sols sahéliens. Thèse de 3e cycle, Université de Nancy 1, Nancy, France
Steiner, J. L., R. E. Schomberg, and J. E. Morrison. 1993. Kesidue decomposition and redistri-
bution, In Crop residue management technology transfer, Southern Plains, ed. II. A,
Stewart, 65-76. Washington D.C.: USDA-ARS/SCS.
Stott, D. E., L. F. Elliott, R. 1. Papendick, and G. S. Campbell. 1986. Low temperature or Iou
water potential effects on the microbial decomposition of wheat residues. Soi/ Bi~!c)~/\\
and Biochemistry 18 : 577-582.
Stott, D. E. 1993. Mass and C loss from corn and soybean residues as associated with thcu
chemical composition, In American Society of Agronomy (ASA) Meeting, Proceeding~,
Abstract, 267. Madison, Wisconsin: ASA.
Stott, D. E. 1994. Changing relationship between mass and surface area of decompcosing
residues, In American Society of Agronomy (ASA) Meeting, Proccedings, Abstract, 261.
Madison, Wisconsin : ASA.
Stroo, H. F., K. L. Bristow, L. F. Elliott, R. 1. Papendick, and G. S. Campbell. 1989 Predict-
ing rates of wheat residue decomposition. Soil Science Society of Amerisa Journo~ 53:31
99.
Witkamp, M., and J. van Der Drift. 1963. Breakdown of forest litter in relation te envirorr-
mental factors. [%rlt ccnd Soi/ t 5 : 295 .3 1 1.