Agricultural Systems 79 (2004) 83–107
www.elsevier.com/locate/agsy
Spatial carbon, nitrogen and phosphorus budget
in a village of the West African savanna—II.
Element flows and functioning of
a mixed-farming system
Raphaël J. Manlaya,b,*, Alexandre Ickowiczc,
Dominique Massea,1, Christian Fellerd,2, Didier Richarde
a
Institute for Research and Development (IRD, ex-ORSTOM), BP1386, Dakar, Senegal
Institute of Forestry, Agricultural and Environmental Engineering (ENGREF), BP 44494,
34033 Montpellier Cedex 5, France
c
Ecosystèmes Naturels et Pastoraux (ECONAP), CIRAD-EMVT, ISRA-LNERV, BP 2057, Dakar, Senegal
d
Centro da Energia Nuclear na Agricultura (CENA-USP), Institute for Research and Development
(IRD, ex-ORSTOM), Caixa Postal 96 13400-970 Piracicaba, SP, Brazil
e
Projet Concerté de Recherche et Développement pour l’Elevage en Afrique de l’Ouest (CIRDES),
01 BP 454, Bobo Dioulasso, Burkina Faso
b
Received 4 April 2001; received in revised form 12 February 2003; accepted 5 March 2003
Abstract
Management of organic resources plays a decisive role in the viability of mixed-farming
systems in West African savannas. For this reason in this study carbon (C), nitrogen (N) and
phosphorus (P) flows initiated by the different uses of organic resources were quantified in the
different land-use systems of a village in southern Senegal. Livestock, crop harvest, and wood
and straw collecting were responsible for, respectively, 59, 27 and 14% of the C outflows from
the area exploited by the village. Livestock accounted for nearly 80% of C, N and P returns to
the soil. As a result of these transfers and of on-site recycling of herbaceous biomass, high C
inputs were brought to staple crops in the compound ring. Nitrogen and P depletion of the
system amounted to 4 kg N and 1 kg P ha1 year1 only when other abiotic flows were
included. This study also suggests that population growth is likely to rapidly alter the sus-
* Corresponding author. On leave from ENGREF, France. Present address: IRD, MOST, BP 64501,
34394 Montpellier Cedex 5, France. Tel.: +33-467-416-267 ; fax: +33-467-416-294.
E-mail address: [email protected] (R.J. Manlay).
1
Present address: IRD, 01 BP182, Ouagadougou, Berkina-Faso.
2
Present address: IRD, MOST, BP 64501, 34394 Montpellier Cedex 5, France.
0308-521X/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0308-521X(03)00054-4
84
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
tainability of this mixed-farming system at least due to a modification in the balance between
the supply of and demand for organic resources.
# 2003 Elsevier Ltd. All rights reserved.
Keywords: Plant biomass; Carbon; Fertility transfer; Flux; Manure; Mixed-farming system; Nitrogen;
Phosphorus; Savanna; Senegal
1. Introduction
One of the main features shared by traditional, self-sufficient tropical farming
systems especially in West African savannas (WAS)—is the use of organic matter
(OM) as a multi-purpose tool, as it plays not only structural and energetic roles in
the agro-ecosystem, but also conveys nutrients (Ruthenberg, 1971; Kowal and
Kassam, 1978). Organic matter is a valuable output (economic goods) of the system
and represents one promising way—among others—to increase productivity and
ensure viability through biological maintenance.
Assessing the dynamics of carbon (C) and organically mediated nitrogen (N) and
phosphorus (P) resources thus enables the sustainability of local agro-ecosystems to
be evaluated (Woomer et al., 1998; Dugué, 2000). To be valid, this assessment must
be carried out at the village scale, since common land tenure and social organisation
are the rule in West Africa (Landais and Lhoste, 1993; Izac and Swift, 1994; Defoer
et al., 1998). Studies should include an appraisal of the supply of organic resources
through analysis of their spatial distribution. This has been accomplished for a
mixed-farming system in Southern Senegal (Manlay et al., 2003) and is complemented here by a study of the uses made of these resources and their impact on
the nutrient balance of the system.
Few studies have included all the different uses of organic matter in quantifying
organic matter fluxes at the level of the farming system (Woomer et al., 1998;
Dugué, 2000; Ngamine and Altolna, 2000). However, the potential of crop–livestock
integration for fertility transfers at the village scale has now been widely demonstrated (Landais and Guérin, 1992; Landais and Lhoste, 1993; Fernandez-Rivera et
al., 1994; Buerkert and Hiernaux, 1998), though accurate spatialisation and quantification remain rare in sub-Saharan Africa (Murwira et al., 1994; Hiernaux et al.,
1997; Achard et al., 2000).
From a methodological point of view, a fundamental feature of carbon dynamics
in tropical farming systems is its highly seasonal pattern. Production of plant biomass occurs mainly during the wet season, even in perennial vegetation, leading to
peak storage of organic matter at the end of the rains (Kowal and Kassam, 1978).
Plant productivity remains very low in subsequent months, while the continuous
activity of human beings and animals results in the progressive exploitation of the
newly created resource. As a consequence, substantial vertical flows (exchanges
between plant, soil and the atmosphere) and horizontal transfers of carbon and
nutrients occur until the return of the rains. Farmers manage some of these flows
directly; the main flows controlled being related to food harvest (grain and haulm),
livestock-mediated organic fluxes, and collection of wood in fallows and savanna.
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
85
This paper has three objectives: first to quantify and compare flows of C, N and P
in plant biomass caused by human activity within and between the different land-use
systems (LUS) of the village, second to assess their impact on the nutrient balance
of each LUS, and third to estimate future trends in carbon flows for the village
system.
2. Methods
2.1. Site characteristics
The study was carried out in the village territory of Sare Yorobana (12 490 N,
14 530 W), southern Senegal, located in the Region of High Casamance, Department
of Kolda. Although high spatial diversity is found in soil conditions and land use,
the village is fairly representative of the mixed-farming systems in this region.
A detailed description of the climate, soil and vegetation of the study site appears
in Manlay et al. (2002a–c). The main natural features are the following: (1) tropical
subhumid climate with 960 mm of mean annual rainfall during the 1978–1997 period, which lasts from May to October, (2) flat landscape comprising ferric and haplic
Lixisols on the plateau and glacis and clayey Gleysols in seasonally flooded lowlands
(FAO, 1998).
The farming system exhibits ring-like organisation scheme that is typical of West
African human settlements and is described in Manlay et al. (2003). It includes
(from upslope to downslope) a bush ring under semi-permanent cropping (groundnut Arachis hypogea L. and cotton Gossypium hirsutum L.), a compound ring under
continuous cereal cropping (pearl millet Pennisetum glaucum L., maize Zea mays L.
and sorghum Sorghum bicolor L. Moench), and a palm grove (Elaeis guineensis
Jacq.) and paddy fields (Oryza sativa L.).
In 1997, the village managed 410 tropical livestock units (TLU; one TLU being
equal to 250 kg of live weight or LW). Taurine cattle consist of the Bos taurus species. The local Ndama breed is trypano-tolerant but has rather poor zootechnical
performance (Coulomb, 1976). Cattle are usually run by herdsmen in the savanna/
forest ring during the cropping period, and are left to stray during the dry season
(Richard et al., 1991). Common grazing is the usual rule; however, at night manuring animals are corralled on the plots of their owner only, and no manuring agreement exists between households.
Sedentary Fulani (Peulh) people have been devoting themselves to cropping for
more than a century, but they are basically herdsmen. Ownership of animals varies
considerably between holdings (Manlay et al., 2003).
2.2. Quantification and spatialisation of carbon, nitrogen and phosphorus fluxes
2.2.1. Flows at harvest
Harvested crop biomass was measured on 91% of the cropped area of the village,
the remaining percentage was estimated from enquiries among farmers (Manlay et
86
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
al., 2003). Conversion to C, N and P amounts was undertaken using element content
data presented in Manlay et al. (2002b). Organic matter flows from the fields to the
farmyard included the following components: panicles of millet, sorghum and rice;
grains and cobs of maize; groundnut haulms. Exports from the village territory
consisted of groundnut pods and cotton bolls (home consumption not included).
Returns from the farmyard to the compound ring were: non-edible components of
millet, sorghum and rice panicles, maize cobs, and haulm recycled as faeces of small
ruminants, calves and oxen after consumption as feed supplement. Biomass partitioning was carried out at the laboratory using traditional techniques, i.e. manual
crushing with a pestle and mortar.
2.2.2. Livestock-mediated transfers
2.2.2.1. Space mapping. Two kinds of mapping were performed. Firstly, the vegetation of the subregional district of Dioulacolon, to which Sare Yorobana belongs,
was mapped (as a photo-interpretation of the physiognomic types of vegetation) and
scaled to 1:12 000. For this study, six vegetation types were distinguished: woody
savanna (upper woody strata or UWS over 7 m high); bush savanna (2 m < UWS < 7
m); grass savanna (UWS < 2 m); rice fields; rainfed fields and ponds (hereafter
pooled under ‘‘Other’’). This typology was only used to characterise the flows
occurring in the peripheral area not appropriated by the village (except for the
palm grove, which was pooled with bush savanna). Secondly, plots owned by the
village were mapped using field measurement methods as described in Manlay et
al. (2003).
2.2.2.2. Livestock location. The day-straying movements of three out of the 10 herds
managed by the village were monitored for 1 day out of every 15 days throughout
the 1995–1996 dry season. These three herds represented roughly half the cattle
population of the village. Their trajectory was recorded using a global positioning
system or a magnet (direction recording) coupled with a topofil survey device (distance recording; one recording every 5 min).
Land-use maps and herd position data were managed using the Atlas Geographic
Information System software (SMI, 1993). The time spent by animals on each LUS
(vegetation type outside the village, basic plot inside the village) was computed by
superimposing herd trajectories on land-use maps. Browsing activity was quite
steady during straying irrespective of land use (Ickowicz et al., 1999); spatial distribution of plant biomass intake was thus inferred linearly from the length of time
spent. Because little interannual change in land use and climate occurred during the
study period, figures obtained for the 1995–1996 dry season were applied to the
1996–1997 dry season.
2.2.2.3. Estimate of plant biomass uptake and dung deposition by livestock.
Methods used to assess the quantity and quality of night and day faecal production
were detailed in Ickowicz et al. (1998, 1999). Faecal indexes were used to estimate
OM intake (OMI) and nitrogen intake from faecal organic matter excretion
(FOME) values (Guérin et al., 1989). Dung C content was determined using
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
87
chromatography after burning at 850 C (Thermoquest NC soil 2000); P content
was measured after acid attack (HCl) on ashes followed by ICP spectrophotometry.
Carbon intake was estimated assuming that only herbaceous biomass was ingested
by cattle, since tree fodder represented only 7–14% of the forage consumed by cattle
on the study site (Delacharlerie, 1994). Carbon intake was computed from mean
carbon contents reported for herbaceous biomass of cropped and fallow fields in
Manlay et al. (2002a,c) and from dry matter intake (DMI) values (DMI was derived
from OMI assuming ash content to be 10%). This approximation could not be
used to estimate phosphorus intake. Cattle do indeed select nutrient-rich plant
components, and the P content of herbaceous biomass must consequently be lower
than that of intake by cattle. Total P intake during the dry season was thus estimated as being equal to P faecal excretion, corrected for the variation in P stored in
the biomass of animals between the beginning and the end of the dry season
(assuming the P content of animals to be 11 g kg LW1 according to Winter, 1993–
2000). This was then dispatched among LUS, proportionally to the time spent on
each unit.
2.2.2.4. Night corralling practices. Kraaling was studied throughout the 1996–1997
dry season. The plot location, date and number of animals in each corral were
recorded. Dung deposition was computed (in t DM ha1 and t OM ha1) and used
for the establishment of a map of dung deposition density, and as an explanatory
variable to predict millet yield measured in 25 plots (14 in the compound ring; 11 in
the bush ring).
2.2.3. Energy and construction needs
Fuel wood needs were estimated using population census (Manlay et al., 2003)
and the work of Bazile (1998) carried out in a village in southern Mali, which has
climatic conditions and living habits comparable with those in Sare Yorobana. In
Mali, the author related wood consumption C (in kg DM hab1 year1) to the size S
of the population of the holding in the following manner: C=5.68S0.73. In the
present study, firewood was assumed to be harvested from the closest wood resources, i.e. bush ring fallows. Full combustion at the farmyard was hypothesised, leading to full loss of C and N, while P was returned to the compound fields as ashes.
Needs for construction wood were almost impossible to assess, because of the
diversity of uses.
Fuel wood consumption was then compared to net annual production of live and
dead wood in fallows over two years old. Live wood production was estimated as
the increase in trunk biomass (TB) for each fallow stand in Sare Yorobana, estimated by the following logistic-like model M1 relating TB (in t ha1) accumulation
to length of fallow t (in year) (Manlay et al., 2002a):
TBðtÞ ¼
581
557
1 þ 0:0431 e0:211t
ð1Þ
88
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Dead wood production was estimated at 0.7 t ha1 year1 in fallow plots aged 10
years or more, and at 0.4 in younger ones (model M2) (Shackleton, 1998; Manlay et
al., 2002a). Models M1 and M2 were applied to the plot database presented in
Manlay et al. (2003) to obtain wood production estimates for the whole village
territory.
Stalks needed for roof construction were quantified. The mean weight of local
bundles was determined from 20 replicates. A regression relationship was determined between the number of bundles (NB) needed for roofing (obtained through
enquiries) and hut diameter (10 replicates), using NLIN procedure from SAS Software (Hatcher and Stepanski, 1994). The number of bundles needed to roof a hut
with a circumference C (in m) was estimated by the following relationship:
NB=0.289C^1.664 (R2=0.96; p{Fobs > Fth}< 0.001), with bundle weight averaging
8.96 kg DM. Straw biomass stored on the roofs of all the dwellings was then computed, and a turnover rate—estimated from investigations among farmers—applied
for computation of the annual flux needed to renew this spoiled biomass, which is
subsequently returned to the compound fields as household waste.
2.3. Outlook on carbon flows
The ratio of the uptake of the C resource to the amount of the C resource in the
standing biomass is one indicator of the viability of the farming system. To predict
how this uptake may change over the next 15 years, a simplified representation of its
dynamics was made using a spreadsheet linking C amounts to land use, which in
turn was linked to manure availability and human needs (for a detailed description
of the model see Manlay et al., 2003). Output variables monitored were C flows (C
intake by livestock, crop, wood and straw harvest) and the ratio of C flow to C
amount.
3. Results
3.1. Carbon, nitrogen and phosphorus flows
3.1.1. Crop harvest
The outflow from plant biomass harvest amounted to 69.4 t C (Fig. 1a); 75%
originated from the bush ring, and 36% was exported as groundnut pods and cotton
bolls for sale, the rest being transferred to the farmyard in the form of panicles and
ears of cereals, and groundnut haulms. An optimistic estimate of recycling of crop
residues and weeds gave 45 t C (excluding loss to fire). Eighteen tons of carbon from
non-edible plant biomass and recycled haulm were transferred from the farmyard to
plots adjoining dwellings. In the food-crop fields of the compound ring N and P
amounts contained in these returns (63 kg N ha1 and 5.2 kg P ha1) counterbalanced N and P exports due to harvest (Table 1). As a whole, 46% of N and 35%
of P harvested were exported (Figs. 2a and 3a), and 511 kg N and 99 kg P lost in
septic tanks (Table 1).
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
89
Fig. 1. Anthropogenic flows of carbon established from November 1996 to November 1997 in Sare Yorobana.
Livestock flows include the dry season only. Arrow width is proportional to flow value. All values in tons.
90
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Table 1
Dry matter, carbon, nitrogen and phosphorus budgets of the land-use systems exploited by smallholders
of Sare Yorobana in and around the village territory, related to crop harvest, livestock-mediated transfers,
wood and straw harvest, and residue recycling
Vector
Savanna
ring
Bush ring
In Out
Fallow
Cash
In
In
Out
Out
Compound
ring
Farmyard
Rice
field
Food
Food
In
In
In
In
Out
Out
Out
Out
1
Dry matter (t ha )
Harvest
Cattle
0.1
0.3 0.1
Wood
Straw
0.0
Residue
nd
Total
0.1
0.3 0.1
0.3
0.7
1.0
0.4
2.6
1.0
0.0
2.0
1.0
2.2
2.9
4.6
0.0
0.4
3.5
2.5
4.5
3.3
0.4
2.2
10.2
0.2
1.0
0.3
0.0
0.8
0.4
0.8
1.0
1.9
1.2
0.1
0.8
3.8
1.2
2.7
128.9
80.7
86.0
8.1
86.0
8.1
4.0
223.0
174.8
0.4
1.0
43.9
27.3
0.3
1.3
1.3
1.2
1.6
2.6
0.1
0.4
0.4
0.4
0.5
0.8
1
Carbon (t ha )
Harvest
Cattle
0.0
Wood
Straw
Residue
Total
0.0
0.1
0.1
0.3
0.0
0.1
Nitrogen (kg ha1)
Harvest
Cattle
2
6
Wood
Straw
0
Total
2
6
Balance
4
Phosphorus (kg ha1)
Harvest
Cattle
0.2
0.4
Wood
Straw
0.0
Total
0.2
0.4
Balance
0.2
Surface (ha)
0.0
445
nd
0.0
3
3
0.2
0.2
0.4
7
2
8
5
0.1
0.2
12
12
0.4
0.5
0.9
0.9
0.7
0.9
117
1.3
56
21
77
65
3.6
1.3
4.8
3.9
42
0.9
1.8
93
93
5.2
5.2
13
48
61
+33
2.5
2.9
5.4
0.1
28
63
146
3
212
5.2
11.8
3.1
0.3
20.4
1.4
15
59
74
+138
3.0
3.5
6.5
+13.8
32.2
2.8
32.2
2.8
79.0
62.4
1704
1193
11
200
57
1961
200
57
1450
+511
198
99
11
0.9
59
6
262
19
59
6
163
+99
0.9
8
28
36
25
1.8
1.7
3.5
2.6
16
All data in amount per hectare, except for farmyard (computed in absolute amounts).
3.1.2. Livestock-mediated flows
3.1.2.1. Uptake. Estimated consumption ranged from 46 to 103 g OM kg LW1 day1
(data not shown). The total area explored by animals was 812 ha, including 256 ha
owned by the village, corresponding to a mean stocking rate of 51 TLU km2.
The influence of land tenure on the behaviour of cattle herds as illustrated by
grazing habits varied depending on herd size and the area owned by the holding
(Table 2). The smallest herd spent five times as long on its owner’s fields than on all
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
91
Fig. 2. Anthropogenic flows of nitrogen established from November 1996 to November 1997 in Sare
Yorobana. Livestock flows include the dry season only. Arrow width is proportional to flow value. All
values in kilograms.
92
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Fig. 3. Anthropogenic flows of phosphorus established from November 1996 to November 1997 in Sare
Yorobana. Livestock flows include the dry season only. Arrow width is proportional to flow value. All
values in kilograms.
93
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Table 2
Grazing habits of the cattle of three holdings during day straying influenced by land tenure, herd size and
surface area of land owned by the holding (dry season 1995–1996)
Behaviour features of the herd during straying
Owned surface (ha)
Herd size (TLU)
Available surface per animal (ha TLU1)
Time (day ha1) spent on the plots of
the cattle owner
the village:
all plots
plots clustered per compound (n=17)
Rank
Cattle owner
(1)
(2)
(1)/(2)
Mean ( SE)
Diao
Mama
Mamo
6.6
21.1
0.31
19
36.2
0.27a
14.4
99.6
0.14
2.58
1.28
0.76
0.52
0.98
2/17
0.4
0.57
2/17
0.45
0.72
8/17
TLU: tropical livestock unit (1 TLU=250 kg of live weight).
a
Includes the size of another herd managed by the holding (34.5 TLU).
village plots as a whole. The ratio dropped to less than two for the largest herd. It
was also observed that the bigger the available area owned by the holding per TLU,
the higher the preference of animals for the owner’s fields.
Previous cropping and land use considerably influenced frequentation of LUS by
animals during the day, as illustrated by the contrasted organic matter flow densities
recorded among land uses (crop type, Table 1) and LUS (ring management, physiognomy of the vegetation of uncropped areas, Fig. 4). The most intensively grazed
crops and LUS were respectively, millet or maize, and the compound ring; fallow
stands and savanna were the least frequented. In absolute values, main forage
sources were the bush ring and the savanna ring (63%) (Fig. 1b); the compound ring
ranked behind the rice fields. Nitrogen and phosphorus flows showed spatial patterns similar to C, and amounted to 9.4 t N and 580 kg P (Figs. 2b and 3b).
The herbaceous DM ‘‘annual uptake:annual production’’ ratio would slightly
exceed 50% in the cropped fields in the compound and bush rings, but would be less
than 10% in fallow stands (Table 3).
3.1.2.2. Faecal excretion. Night corralling was mainly practised close to the dwellings (Fig. 5). Marked contrasts in the intensity of manuring (ranging from 0 to 13.4 t
DM ha1) were recorded between plots. Night corralling took place in plots planned
for cereal cropping (Table 4). The most manured plots were mixed stands of millet
and maize, while groundnut received the smallest quantities. With regard to manuring rates, other cereals were ranked as follows: maize> millet > sorghum.
In Sare Yorobana, FOME ranged from 19 to 48 g OM per kilo of metabolic
weight or MW throughout the year, with a peak recorded at the beginning of the dry
season (Table 5). Dung deposition was almost equally distributed between night
(53%) and day (47%). Total faecal excretion amounted to 239 t DM, half of which
was dropped during night tethering. This was equivalent to 98 t C, 7.5 t N and 610
94
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Fig. 4. Organic matter inflows and outflows initiated by intake and faecal excretion of three herds during
the 1995–1996 dry season.
kg P (Figs. 1b, 2b and 3b). Overall dung deposition mainly occurred in the bush and
compound rings (85%); in these rings carbon inputs from manure offset C output
(input:output ratio=1.4); this was not the case for the savanna ring and rice fields,
in which only a third of C loss was recovered (Fig. 1b, Table 1). The same trends
were recorded for N and P, with P inflows compensating for half the losses (Figs. 2b
and 3b).
The impact of manuring on millet yield was noticeable in plots in both rings; it
was significant for those in the compound ring, or when data from both rings was
pooled (Fig. 6). However, the slope of the regression was higher for the fields in the
bush ring than for those in the compound ring.
95
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Fig. 5. Manuring intensity from night corralling in the village of Sare Yorobana during the 1996–1997
dry season.
Table 3
Annual uptake: annual production ratio (in%) for woody and herbaceous resources for some farming
activities in Sare Yorobana
Plant above-ground component
Herbaceous biomass
Land use system
Bush ring
Compound ring
Rice fields
Whole territory
Wood
Bush ring
Land use
Crop
Fallow
Activities
Livestock
Harvest
61
8
56
50
29
16
0
8
18
14
Wood collect
42
3.1.3. Other anthropogenic flows
Wood consumption per permanent inhabitant was estimated at 280 kg DM per
year (320 kg DM when added to that of the temporary workers needed for cropping). It represented less than half net annual wood production on the village territory (Table 3). Wood harvest generated a considerable flow of carbon between the
bush ring and the farmyard, equivalent to 200 kg N and 58 kg P (Figs. 1c, 2c and 3c).
Total herbaceous biomass stored on roofs of the village was estimated at 40.7 t
DM. Turnover rate was 0.2 year1, equivalent to a yearly input to the compound
ring of 30 kg of straw per capita.
96
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Table 4
Input of dry matter to fields from manuring during night corralling as influenced by the decision concerning the next crop to be grown
Millet(x maize) Maize
Millet
Sorghum
Cotton
Groundnut Fallow All plots
Manuring rate (tDM ha1)
Compound 3.921.28
2.57 1.22 1.580.44 1.100.53 0.24 0.16 0.020.01 0.01
(n)
(12)
(6)
(18)
(14)
(11)
(16)
Village
4.10
4.02
2.28
1.91
0.35
0.03
0.01
0.930.24
(19)
0.59
Amount in absolute values (tDM)
24.6
12.8
(%)
(20.3)
(10.6)
121.2
(100)
69.0
(56.9)
9.0
(7.4)
3.0
(2.5)
1.2
(2.5)
1.5
(1.2)
: standard error. DM: dry matter.
Table 5
Dry matter, carbon, nitrogen and phosphorus intake and excretion by livestock measured during the
1997–1998 dry season
Month
November
December
January
February
March
April
May
June
Organic matter
Nitrogen
Intake
Excretion
Intake
Excretion
74.4
103.2
66.8
61.8
64.5
54.5
51.9
53.4
34.9
47.6
31.8
31.2
32.5
27.9
26.4
25.2
1.92
2.42
1.74
1.32
1.35
1.31
1.19
1.40
1.63
1.98
1.45
1.11
1.19
1.12
1.00
1.20
Carbon excretion
Phosphorus excretion
18.0
24.6
18.0
16.1
16.1
14.7
14.0
14.7
0.111
0.152
0.114
0.093
0.102
0.098
0.081
0.095
All data expressed in g of element per day per kilo of metabolic weight.
Source: ISRA/CIRAD-EMVT, Program ABT (Ickowicz et al., 1998, 1999).
3.1.4. Global carbon and nutrient balance
Livestock activity accounted for more than half of the anthropogenic outflows of
C, N and P, the proportion of crop harvest accounting for only 25–32% of these
flows (Table 6). The proportion of animal activity was even higher when considering
C and nutrient returns to the soil (79–86%). Wood- and straw-mediated transfers
were significant for carbon only.
C, N and P balances resulting from organic input/output due to human activity
were very contrasted between LUS (Table 1)). Highest carbon inputs per hectare
occurred in fields cropped for food production in the compound and bush rings (3.8
and 1.8 t C ha1, respectively). Carbon input in these rings was mainly due to dung
deposition (45 and 55% in the compound and bush rings, respectively) and crop
biomass recycling (mainly residues). In both rings exogenous N and P inputs during
the dry season mainly originated from manuring. The situation in the village rice
97
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Fig. 6. Millet yield related to manuring practices in the compound and bush rings. p{RSpearman=0}:
*< 0.05; **<0.01.
Table 6
Contribution of crop harvest, livestock, and collecting of wood and straw to anthropogenic carbon,
nitrogen and phosphorus transfers due to farming activities
Crop (%)
Livestock (%)
Wood and straw (%)
Total
(%)
(abs)
Removal
C
N
P
27
25
32
59
73
61
14
2
7
100
100
100
255 t
12.8 t
956 kg
Return
C
N
P
15
14
13
83
86
79
2
1
8
100
100
100
119 t
8.7 t
773 kg
fields was slightly different, with 75% of the C inputs originating from residue
recycling.
Carbon uptake in the bush ring was mainly related to harvest and browsing and
was more or less evenly distributed between the two activities. In the compound
ring, browsing accounted for 67% of C withdrawal. In the rice fields, intake was
evenly distributed between animals and humans beings.
Nitrogen and phosphorus balances were strongly positive in the compound fields
cropped with cereals: +117 kg N ha1, and +11.8 kg P ha1, respectively. A net
98
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Fig. 7. Changes in anthropogenic carbon outflows and ratio of C outflow to amount of C stored in plant
above-ground biomass for the territory of the village of Sare Yorobana for the 1997–2012 period. Outflows considered were: harvested crop biomass, livestock uptake during the dry season, and wood and
straw collected. * Ri: initial C outflow: C amount ratio. See description of model in Manlay et al. (2003).
positive nitrogen balance was also recorded in the food-crop fields in the bush ring,
while the P budget was close to equilibrium. All other land-use units had N and P
deficits, with the highest nutrient depletion in cash crops in the bush ring.
The overall uptake from crop harvest and grazing would amount to nearly half
herbaceous resources (Table 3), with major variations (8–77%) between LUS.
3.2. Outlook on future carbon outflows
In the village territory, carbon uptake due to crop and wood harvest and browsing
by livestock is expected to increase by 44% within the next 15 years (Fig. 7). The
ratio of C outflows to amounts of C stored in plant AGB would increase much
faster and would more than double within the next 12 years. Carbon needs are
expected to reach nearly half of available plant C disposal by 2012.
4. Discussion
4.1. Control of stocks over flows: livestock-mediated transfers
4.1.1. Dry matter intake as influenced by organic matter quantity and quality
At the end of the rains, herding of animals becomes less restricted and cattle are
left free to feed wherever they wish at the lowest metabolic cost. Preferential grazing
on crop residues has already been reported for this study site by Richard et al.
(1991) and for North Cameroon by Dugué (1998b). In Sare Yorobana, maize stover
and weeds of cropped fields have a higher N content than the herbaceous layer of
fallows (Manlay et al., 2002a,c) as a result of exogenous nutrient inputs to food crop
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
99
fields (Powell, 1986; Lamers et al., 1996; Buerkert et al., 1997). Biomass availability
also influences grazing trajectories of animals, as testified by the low frequentation
of groundnut fields, in which the removal of haulm leaves only little edible biomass
for livestock (Fig. 4).
Animals left to graze freely exploit only a fraction of available plant biomass
(Table 3). Cereal leaves and weeds are eaten first, but much of stalk biomass is left
because of its poor feed value, or due to tainting by urine, and trampling during
browsing. As a result, frequentation of cropped rainfed areas drops rapidly within
the first two dry months; later, animals prefer to explore rice fields (where they have
access to rice straw and limited grass re-growth) and rangelands (fallow, savanna)
(Richard et al., 1991; Ickowicz et al., 1998). The rapidly decreasing feed quality of
the herbaceous layer during the dry season (César, 1992), and removal of large
amounts of plant biomass by uncontrolled fires may well account for the low frequentation of fallow stands.
As a result, only 29% of available herbaceous forage on the village territory (not
including the palm grove) is eaten by animals, which is consistent with findings for
Burkina Faso (Quilfen and Milleville, 1983).
4.1.2. The human factor
Preferential frequentation of animals of plots owned by the holding that herds the
cattle may result from the memorisation of corralling sites by animals. Less pronounced selectivity of herds from holdings with low land availability (compared
with herd size; Table 2) confirms that the larger the size of a holding’s herd, the
higher the taping exerted by this holding on plant biomass and thus on nutrients
located in plots belonging to others (Manlay et al., 2003). Other human factors that
influence cattle behaviour are the need to oversee cattle and the choice of watering
animals at down-slope ponds; both factors decrease the time spent in peripheral
areas.
4.2. Control of flows over stocks: night corralling and cereal yield
The positive response of cereal yield to manuring has been clearly established in
sub-Saharan Africa under controlled conditions (Pieri, 1989; de Ridder and van
Keulen, 1990; Bationo and Mokwunye, 1991) and in farmers’ fields (Powell, 1986;
Derouw, 1998). Trends could only be evidenced for data from the bush ring, which
probably reflects interactions with fallow practices. The variable response of millet
to manuring in the compound ring suggests a significant influence of manure applied
during previous years (de Ridder and van Keulen, 1990; Lupwayi and Haque, 1999),
inputs of fertilisers such as household waste and dung dropped during the day,
agricultural practices and land-use history.
Dung deposition during day straying might alleviate unequal access to manuring
between holdings, but excessive manuring of some plots (Fig. 5, this paper; Manlay
et al., 2003) prevents the village from exploiting the real potential for food production, since cereal yield does not respond linearly to organic manuring at rates
exceeding 5–6 t ha1 (Gueye and Ganry, 1981; Bationo and Mokwunye, 1991),
100
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
while high manuring rates (10 t DM ha1 year1) lead to heavy leaching of C, N and
P (Brouwer and Powell, 1998), and even to a decline in yield under dry conditions
(Williams et al., 1994; Achard et al., 2000). Increased weed encroachment is a further
drawback to manuring (Powell, 1986).
4.3. Global carbon and nutrient balance of the village agro-ecosystem
4.3.1. On-site recycling
From the pastoralist point of view and given the facts mentioned earlier, common
grazing saves labour but not forage resources. From an agricultural viewpoint,
organic matter recycling through animals speeds up biogeochemical cycles (Landais
and Guérin, 1992), but leads to the withdrawal of a third of the carbon from the
system through animal respiration. Plot studies carried out on various agro-ecosystems in Sare Yorobana have underlined the necessity of ensuring steady C flows
through the soil to maintain soil quality (Manlay et al., 2002b,c). Thus proposals
aimed at improving the quality and availability of manure through herd expansion
and stalling of cattle (Bosma et al., 1999) should take into account their cost in
energy (carbon). Organic matter/energy loss is all the more likely to occur, since the
return of manure from stall to field is often limited due to the lack of means of
transport and of labour (Schleich, 1986).
4.3.2. Spatial organic transfers, carbon inputs and nutrient balances
The bush ring represented the main carbon source for the village. Due to the
extension of this ring, C outflows have up to now only represented 8.9% of the
carbon stored in above-ground biomass, not including litter (Manlay et al., 2003).
Inside the bush ring, C uptake was high in cropped fields and low in fallow stands
(75 and 3% of the C-AGB pool, respectively). High values were also recorded for
the food crops in the compound ring (65%) and rice fields (68%). But C redistribution benefits to the compound ring were at the expense of other rings (Table 1).
Carbon inputs in food crops in this ring (3.8 t C ha1 year1) were much higher than
amounts usually recommended to compensate for soil organic carbon mineralisation, which assume a rate of soil organic matter mineralization of 0.06 year1 (de
Ridder and van Keulen, 1990; Berger, 1996). Amounts of carbon measured for
cropped plots in the compound ring and for fallow plots in the bush ring averaged
17.7 and 15.5 t C ha1, respectively (0–20 cm soil layer; Manlay et al., 2002b,c).
Under the reasonable hypothesis that the soil organic carbon stock of cropped
plots would equal that of old fallow stands some 20 years ago (when the village
moved to its current location), the actual soil organic matter decomposition rate
would be 0.21 year1 (neither decay nor exudation of roots taken into account).
This suggests that estimates usually accepted in the literature (0.05; Pieri, 1989) are
much too low for some of the sandy soils of the region.
When considering nutrient dynamics, N and P outflows originated equally from
rainfed cropped fields and other land-use systems. But positive N and P balances
recorded for staple crops would not have been possible without the recycling of
exogenous nutrients, of which nearly all benefited staple crops. Thus, the current
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
101
system acts as an impluvium for carbon and nutrient elements, since it taps organic
resources from peripheral areas to rainfed food crops (Fig. 1). In this way, the ringlike organisation enables sustainable continuous cultivation of cereals at relatively
high yields on 7% of the land owned by the village (Manlay et al., 2003).
In semi-arid Burkina Faso, Krogh (1997) showed how the nutrient balance of
farming systems depends on the spatial scale considered. Most of the N and P balances of staple fields reported in his work were negative. Things were different at the
village level, since outputs due to harvest of staple crops were kept within the village
boundaries. The present study yielded the same kind of results: the nutrient balance
in the bush ring was negative, but that of the village territory as a whole was positive
when N and P stored in septic tanks were integrated. However, our conclusion differs from Krogh’s, since a distinction has to be made between geographic and
functional balances: N and P excreted by human beings is not recycled and is consequently lost for the cropping system.
From a nutrient viewpoint, the system as a whole might be considered close to
sustainability. However, one of the major prerequisites is livestock availability,
which mediates most of C, N and P flows. The potential of higher herd densities to
sustain agricultural systems through manure production has been well demonstrated in West Africa (Schleich, 1986; Williams et al., 1994; Bosma et al., 1999).
But unless there is a switch to more intensified farming patterns such as fertilised
ley and improved fallow (Hoefsloot et al., 1993), forage availability quickly
impedes the maintenance of animal husbandry in crowded areas under higher
population pressure. Another condition ensuring the sustainability of the farming
system in Sare Yorobana is thus satisfactory land availability. The accessibility of
wide peripheral rangelands (1) ensures forage availability during the cropping period, thus avoiding seasonal—and mostly definitive—migration of livestock as
widely experienced in the Groundnut Belt of Central Senegal (Lericollais and
Milleville, 1993), (2) lessens competition between human and animal needs for
plant biomass, since large amounts of fuel wood are stored in the fallow and
savanna ring, and (3) compensates for nutrient losses from the system at a low
mineral depletion rate. Reporting good carbon and chemical status of soils in the
compound ring in villages under various climates in Burkina Faso, Prudencio
(1993) concludes that the move toward more permanent cultivation systems will
mine the fertility of fields in the outer ring, but not that of the chemically well
endowed soils in the compound ring. This conclusion cannot apply to Sare Yorobana, since the soil quality of plots neighbouring the compounds relies on
organic mining flows from the bush and savanna rings. Intensification may thus
reduce the total surface area of nutrient sources and threaten both biological
mechanisms of mineral repletion and crop–livestock integration (Giller et al.,
1997).
4.3.3. Other carbon, nitrogen and phosphorus flows in soil
Only organically mediated C, N and P inputs and outputs are reported here.
Actual C inflows to the soil should also take into account root exudation and decay,
and litter production in fallows. Carbon transfers through erosion, runoff and
102
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
leaching should remain limited: assuming such C flows related to the water cycle to
be 20 and 84 kg C ha1 year1 in fallow and cropped fields, respectively (Roose and
Barthes, 2001), C transfers would amount to 10.6 t ha1 over the whole area owned
by the village, which is less than 2% of the flows of fresh plant biomass generated by
harvest and grazing. Factors affecting N and P balances not taken into account here
were: atmospheric deposition, biological fixation of nitrogen, leaching, gaseous losses and erosion. For cropping fields under uncertain rainfall in Senegal, Stoorvogel
and Smaling (1990) estimate the net balance of these factors to be 3.5 kg N and
1.5 kg P ha1 year1, and biological N and P accumulation in fallow to be +2.0
kg N and +0.87 kg P ha1 year1. When adding these values to the flows calculated
for Sare Yorobana in Table 1, the final nutrient balance of the system would be
4 kg N and 1 kg P ha1 year1, which is closer to equilibrium than the findings
of these authors (20 kg N and 3 kg P ha1 year1) for the area with uncertain
rainfall area in Senegal.
4.4. Future trends in the use of carbon resources
The current proportion of carbon amounts withdrawn by farmers (harvest and
grazing) for their needs to the C amounts in AGB of the village territory was low in
1997 (16% according to the model). However, it is likely to increase rapidly during
the coming years as a result of the growing need for cropped land. Among other
consequences, this could easily lead to exclusion of livestock, as experienced by the
Sereer farming system in Central Senegal during the 1965–1985 period, and, more
recently, by farmers throughout the Sine-Saloum region, which is not far from HighCasamance. Coexistence of continuous cultivation and animal husbandry is another
way in which this agricultural system may evolve. The accurate prediction of changes in forage availability as a result of a growing need for land is a difficult task. On
the one hand conversion of savanna to cropland can lead to an increase in plant
productivity (Mortimore et al., 1999) and forage availability (+20% predicted in
the coming 15 years in Sare Yorobana according to Manlay et al., 2003). However,
in the long term, livestock feeding during the cropping season as well as exhaustion
of soil fertility due to increased nutrient uptake by harvest and browsing would
require drastic changes in farming practices, land tenure status, increased chemical
inputs, and thus the existence of financial and technical advisory structures (Dugué,
1998a).
5. Conclusion
Limited nutrient depletion and the low rate of anthropogenic uptake of annual
production of plant above-ground biomass suggest that Sare Yorobana is more
ecologically sustainable than is generally reported for smallholder farming in subSaharan Africa. Levels of availability in livestock and land per capita that are unusual for the sub-region are required (Landais and Lhoste, 1993). From an ecological
viewpoint, and in a context of preeminence of biological maintenance over
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
103
substitutional maintenance (senso Izac and Swift 1994), a parallel can be drawn with
the role of carbon circulation in sustaining the organisation and functions of the
ecosystem at the plot and the village territory scale. Since fluxes of matter and
energy are determined by the conservation of gradients of elements across landscape, the quantitative comparison of organic pools and fluxes (this study and
Manlay et al., 2003) yields more than the assessment of uptake pressure on natural
resources or of the nutrient balance of the farming system.
In all likelihood, rapid changes in the balance between human, livestock and land
factors in Sare Yorobana will alter patterns of C cycling in the future, hence the
need for intensified management of organic resources. Based on this study, realistic
proposals for the region that only requires light investments should include the
recycling—on cropped fields—of (1) manure produced during the dry season, which
represented more than half the total dung production over the year, and (2) human
dejecta that are a major sink of nitrogen and phosphorus. However, the feasibility of
both proposals is likely to be influenced by more than technical considerations since
they would require funding to improve transport (carts), and the abandonment of
certain cultural taboos.
Beyond a certain threshold of population pressure, other options will have to be
considered to ensure crop-livestock integration and thus the sustainability of agricultural systems in the WAS. Some of these options should aim at increasing organic
matter production, including ley-making and live-hedge planting; others should aim
at increasing the efficiency of its use, for instance through hay-making and stalling.
However, it is unlikely that the organic strategies mentioned above will be enough to
maintain the viability of the farming system of Sare Yorobana in the case of high
population pressure. The use of chemical fertiliser combined with organic matter
recycling will be needed, at least to ensure soil nutrient recapitalisation. Adoption of
exogenous inputs is highly reliant on agricultural subsidising policies in sub-Saharan
Africa. But it may benefit from the tradable carbon emission permits market
(Ringius, 2002), depending on the balance between its terrestrial carbon sequestration benefit and the cost of the emission of greenhouse gases (chemical synthesis,
transportation and fate).
Acknowledgements
The authors acknowledge critical and helpful comments by J. Aronson, C. Floret,
A.-M. Izac, C. Millier and J.-C. Remy, as well as D. Friot for providing laboratory
analyses. This work received financial support from the EEC Project ‘‘Reduction
of the Fallow Length, Biodiversity and Sustainable Development in Central and
West Africa’’ (TS3-CT93-0220, DG12 HSMU) (Floret, 1998), and from the following French institutions: Institute for Research and Development (IRD, exORSTOM), International Centre for Agricultural Research in Hot Regions
(CIRAD), and Institute of Forestry, Agricultural and Environmental Engineering
(ENGREF).
104
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
References
Achard, F., Banoin, M., Bartholmey, C., 2000. Gestion de la fumure animale dans un terroir du sud-ouest
du Niger. In: Dugué, P. (Ed.), Fertilité et Relations Agriculture-élevage en Zone de Savane, Montpellier,
France, 5–6/05/1998. Centre International de Recherche pour le Développement, France, pp. 97–106.
Bationo, A., Mokwunye, U., 1991. Role of manures and crop residue in alleviating soil fertility constraints to crop production: with special reference to the Sahelian and Sudanian zones of West Africa.
In: Mokwunye, U. (Ed.), Alleviating Soil Fertility Constraints to Increased Crop Production in West
Africa. Kluwer Academic Publishers, The Netherlands, pp. 217–225.
Bazile, D., 1998. Les Gestions des Espèces Ligneuses dans l’Approvisionnement en Énergie des Populations. Cas de la Zone Soudanienne du Mali. PhD thesis, Université de Toulouse le Mirail, Toulouse.
Berger, M., 1996. Fumure organique : des techniques améliorées pour une agriculture durable. Agriculture et Développement 10, 37–46.
Bosma, R.H., Bos, M., Kante, S., Kebe, D., Quak, W., 1999. The promising impact of ley introduction
and herd expansion on soil organic matter content in southern Mali. Agricultural Systems 62 (1), 1–15.
Brouwer, J., Powell, J.M., 1998. Increasing nutrient use efficiency in West-African agriculture: the impact
of micro-topography on nutrient leaching from cattle and sheep manure. Agriculture Ecosystems and
Environment 71 (1–3), 229–239.
Buerkert, A., Lamers, J.P.A., Schmelzer, G.H., Becker, K., Marschner, H., 1997. Phosphorus and millet
crop residue application affect the quantity and quality of millet leaves and fodder weeds for ruminants
in agro-pastoral systems of the Sahel. Experimental Agriculture 33 (3), 253–263.
Buerkert, A., Hiernaux, P., 1998. Nutrients in the West African Sudano-Sahelian zone: losses, transfers
and role of external inputs. Zeitschrift Fur Pflanzenernährung und Bodenkunde 161 (4), 365–383.
César, J., 1992. La Production Biologique des Savanes de Côte d’Ivoire et son Utilisation par l’Homme.
PhD thesis, Université de Paris VI, Maison Alfort.
Coulomb, J., 1976. La race N’Dama. Quelques caractéristiques zootechniques. Revue d’Elevage et de
Médecine Vétérinaire des Pays Tropicaux 29 (4), 367–380.
de Ridder, N., van Keulen, H., 1990. Some aspects of the role of organic matter in sustainable intensified
arable farming systems in the West-African semi-arid-tropics (SAT). Fertilizer Research 26, 299–310.
Defoer, T., de Groote, H., Hilhorst, T., Kanté, S., Budelman, A., 1998. Participatory action research and
quantitative analysis for nutrient management in southern Mali. A fruitful marriage? Agriculture Ecosystems and Environment 71 (1–3), 215–228.
Delacharlerie, P.F., 1994. Contribution à l’Étude de l’Alimentation des Bovins sur Parcours Naturel en
Moyenne-Casamance (Sénégal : Composition Botanique des Régimes, Rôle des Fourrages Ligneux.
Première Approche de la Disponibilité Fourragère Ligneuse. Master’s thesis, ENVA, Maison-Alfort.
Derouw, A., 1998. Gestion de la fertilité du sol sur un terroir sahélien. Fumure animale, matière organique et encroûtement superficiel du sol dans les systèmes de culture de mil, étude au Niger. Agriculture et
Développement 18, 63–70.
Dugué, P., 1998a. Gestion de la fertilité et stratégies paysannes. Le cas des zones de savanes d’Afrique de
l’Ouest. Agriculture et Développement 18, 13–20.
Dugué, P., 1998b. Les transferts de fertilité dus à l’élevage en zone de savane. Agriculture et Développement 18, 99–107.
Dugué, P., 2000. Flux de biomasse et gestion de la fertilité à l’échelle des terroirs. Etude de cas au Nord
Cameroun et essai de généralisation aux zones de savane d’Afrique sub-saharienne. In: Dugué, P. (Ed.),
Fertilité et Relations Agriculture-Élevage en Zone de Savane, Montpellier, 5–6 Mai 1998. Centre
International de Recherche pour le Développement, France, pp. 27–59.
FAO, 1998. World Reference Base for Soil Resources. World Soil Resources Reports, Vol. 84. Food and
Agricultural Organisation, Rome.
Fernandez-Rivera, S., Williams, T.O., Hiernaux, P., Powell, J.M., 1994. Faecal excretion by ruminants
and manure availability for crop production in semi-arid West-Africa. In: Powell, J.M., FernandezRivera, S., Williams, T.O., Renard, C. (Eds.), Livestock and Sustainable Nutrient Cycling in Mixed
Farming Systems of Sub-Saharan Africa. Proceedings of an International Conference, Addis Ababa,
Ethiopia, 22–26 November 1993. ILCA (International Livestock Centre for Africa), pp. 149–169.
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
105
Floret, C. (Ed.), 1998. Raccourcissement du temps de jachère, biodiversité et développement durable en
Afrique Centrale (Cameroun) et en Afrique de l’Ouest (Mali, Sénégal) Final Report. Comm. des
Communautés Européennes. Contrat TS3-CT93-0220 (DG 12 HSMU). IRD, Paris.
Giller, K.E., Beare, M.H., Lavelle, P., Izac, A.-M.N., Swift, M.J., 1997. Agricultural intensification, soil
biodiversity and agroecosystem function. Applied Soil Ecology 6 (1), 3–16.
Guérin, H., Richard, D., Lefevre, P., Friot, D., Mbaye, N., 1989. Prévision de la valeur nutritive des
fourrages ingérés sur parcours naturels par les ruminants domestiques sahéliens et soudaniens. In:
XVIth International Grassland Congress, Nice, France, 4–11 Octobre. INRA, pp. 879–880.
Gueye, F., Ganry, F., 1981. Recyclage des résidus de récolte par la voie fermentative. Exemple de deux
pailles: mil et maı̈s. International Foundation for Science—Institut Sénégalais de Recherches Agricoles/
CNRA, Stockholm—Bambey.
Hatcher, L., Stepanski, E.J., 1994. A Step-by-step Approach to Using the SAS System for Univariate and
Multivariate Statistics. SAS Institute Inc, Cary, NC.
Hiernaux, P., Fernandez-Rivera, S., Schlecht, E., Turner, M.D., Williams, T.O., 1997. Livestock-mediated nutrient transfers in Sahelian agro-ecosystems. In: Renard, G., Neef, A., Becker, K., von Oppen,
M. (Eds.), Soil Fertility Management in West African Land Use Systems. Proceedings of the Regional
Workshop, Niamey, Niger, 4–8 March. Margraf Verlag, Weikersheim, pp. 339–347.
Hoefsloot, H., Van der Pol, F., Roeleveld, L., 1993. Jachères Améliorées. Options pour le Développement
des Systèmes de Production en Afrique de l’Ouest. Bulletins de l’Institut Royal des Tropiques, Vol. 333.
Royal Tropical Institute, Amsterdam.
Ickowicz, A., Usengumuremyi, J., Badiane, A., Richard, D., Colleye, F., Dupressoir, D., 1998. Interactions entre jachère et systèmes d’alimentation des bovins en zone soudanienne du Sénégal: choix techniques et dynamique de développement (zone soudanienne, Sénégal). In: Floret, C., Pontanier, R.
(Eds.), Jachère et Systèmes Agraires, Niamey, Niger, 30 Septembre–2 Octobre. ORSTOM, pp. 123–138.
Ickowicz, A., Richard, D., Usengumuremyi, J., 1999. Estimation of organic matter transfers by cattle in a
Senegalese village. In: Proceedings of the VIth International Rangeland Congress, Townsville, Australia, 19–23 July, pp. 500–502.
Izac, A.-M.N., Swift, M.J., 1994. On agricultural sustainability and its measurement in small-scale farming in sub-Saharan Africa. Ecological Economics 11, 105–125.
Kowal, J.M., Kassam, A.H., 1978. Agricultural Ecology of Savanna: a Study of West Africa. Clarendon
Press, Oxford, UK.
Krogh, L., 1997. Field and village nutrient balances in millet cultivation in northern Burkina Faso: a
village case study. Journal of Arid Environments 35 (1), 147–159.
Lamers, J., Buerkert, A., Makkar, H.P.S., von Oppen, M., Becker, K., 1996. Biomass production, and
feed and economic value of fodder weeds as by-products of millet cropping in a sahelian farming
system. Experimental Agriculture 32 (3), 317–326.
Landais, E., Guérin, H., 1992. Systèmes d’élevage et transferts de fertilité dans la zone des savanes
africaines. I. La production des matières fertilisantes. Cahiers Agricultures 1, 225–238.
Landais, E., Lhoste, P., 1993. Systèmes d’élevage et transferts de fertilité dans la zone des savanes
africaines. II. Les systèmes de gestion de la fumure animale et leur insertion dans les relations entre
l’élevage et l’agriculture. Cahiers Agricultures 2, 9–25.
Lericollais, A., Milleville, P., 1993. La jachère dans les systèmes agropastoraux Sereer au Sénégal. In:
Floret, C., Serpantié, G. (Eds.), La Jachère en Afrique de l’Ouest. Atelier International, Montpellier,
France, 2–5 Decembre 1991. ORSTOM, pp. 133–145.
Lupwayi, N.Z., Haque, I., 1999. Leucaena hedgerow intercropping and cattle manure application in the
Ethiopian highlands—III. Nutrient balances. Biology and Fertility of Soils 28 (2), 204–211.
Manlay, R.J., Kaı̈ré, M., Masse, D., Chotte, J.-L., Ciornei, G., Floret, C., 2002a. Carbon, nitrogen and
phosphorus allocation in agro-ecosystems of a West African savanna I. The plant component under
semi-permanent cultivation. Agriculture Ecosystems and Environment 88 (3), 215–232.
Manlay, R.J., Masse, D., Chotte, J.-L., Feller, C., Kaı̈ré, M., Fardoux, J., Pontanier, R., 2002b. Carbon,
nitrogen and phosphorus allocation in agro-ecosystems of a West African savanna II. The soil
component under semi-permanent cultivation. Agriculture Ecosystems and Environment 88 (3), 233–
248.
106
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
Manlay, R.J., Chotte, J.-L., Masse, D., Laurent, J.-Y., Feller, C., 2002c. Carbon, nitrogen and phosphorus allocation in agro-ecosystems of a West African savanna III. Plant and soil components under
continuous cultivation. Agriculture Ecosystems and Environment 88 (3), 249–269.
Manlay, R.J., Ickowicz, A., Masse, D., Floret, C., Richard, D., Feller, C., 2003. Spatial carbon, nitrogen
and phosphorus budget of a village of the West African savanna—I. Element pools and structure of a
mixed-farming system. Agricultural Systems. doi:10.1016/S0308-521X(03)00053-2.
Mortimore, M., Harris, F.M.A., Turner, B., 1999. Implications of land use change for the production of
plant biomass in densely populated Sahelo-Sudanian shrub-grasslands in north-east Nigeria. Global
Ecology and Biogeography 8 (3–4), 243–256.
Murwira, K.H., Swift, M.J., Frost, P.G.H., 1994. Manure as a key resource in sustainable agriculture. In:
Powell, J.M., Fernandez-Rivera, S., Williams, T.O., Renard, C. (Eds.), Livestock and Sustainable
Nutrient Cycling in Mixed Farming Systems of Sub-Saharan Africa. Proceedings of an International
Conference, Addis Ababa, Ethiopia, 22–26 November 1993. ILCA (International Livestock Centre for
Africa), Africa, pp. 131–148.
Ngamine, J., Altolna, M., 2000. Flux de biomasse et gestion de la fertilité à l’échelle des terroirs. Etude de
cas dans trois terroirs agro-pastoraux du sud du Tchad. In: Dugué, P. (Ed.), Fertilité et Relations
Agriculture-Élevage en Zone de Savane, Montpellier, France, 5–6 Mai 1998. Centre International de
Recherche pour le Développement, France, pp. 13–25.
Pieri, C., 1989. Fertilité des Terres de Savanes. Ministère de la Coopération—CIRAD, Paris (English
edition: Pieri, C., 1992. Fertility of Soils: A Future for Farming in the West African Savannah. Springer
Series in Physical Environment. Springer-Verlag, Berlin).
Powell, J.M., 1986. Manure for cropping: a case study from central Nigeria. Experimental Agriculture 22,
15–24.
Prudencio, C.Y., 1993. Ring management of soils and crops in the West African semi-arid tropics—the
case of the Mossi farming system in Burkina Faso. Agriculture Ecosystems and Environment 47, 237–
264.
Quilfen, J.P., Milleville, P., 1983. Résidus de culture et fumure animale. Un aspect des relations
agriculture-élevage dans le Nord de la Haute-Volta. L’Agronomie Tropicale 38 (3), 206–212.
Richard, D., Ahopke, B., Blanfort, V., Pouye, B., 1991. Utilisation des zones agricoles et pastorales par
les ruminants en zone soudanienne (Moyenne Casamance, Sénégal). In: Gaston, A., Kernick, M., Le
Houérou, H.N. (Eds.), Actes du Quatrième Congrès International des Terres de Parcours, Montpellier,
France, 22–26 Avril. CIRAD, France, pp. 754–756.
Ringius, L., 2002. Soil carbon sequestration and the CDM: opportunities and challenges for Africa.
Climatic Change 54 (4), 471–495.
Roose, E., Barthes, B., 2001. Organic matter management for soil conservation and productivity
restoration >in Africa: a contribution from Francophone research. Nutrient Cycling in Agroecosystems 61 (1–2), 159–170.
Ruthenberg, H., 1971. Farming Systems in the Tropics. Clarendon Press, Oxford.
Schleich, K., 1986. Le fumier peut-il remplacer la jachère ? Possibilité d’utilisation du fumier: exemple de
la savane d’Afrique Occidentale. Revue d’Elevage et de Médecine Vétérinaire des Pays Tropicaux 39
(1), 97–102.
Shackleton, C.M., 1998. Annual production of harvestable deadwood in semi-arid savannas, South
Africa. Forest Ecology and Management 112 (1–2), 139–144.
SMI, 1993. ATLAS GIS—3.1 Manuel de référence. Le système d’information géographique pour
Windows. Strategic Mapping Inc, Santa Clara.
Stoorvogel, J.J., Smaling, E.M.A., 1990. Assessment of Soil Nutrient Depletion in Sub-Saharan Africa:
1983–2000. Vol. I: Main Report (Rep. No. 28). The Winand Staring Centre, Wageningen.
Williams, T.O., Powell, J.M., Fernandez-Rivera, S., 1994. Manure utilisation, drought cycles and
herd dynamics in the Sahel: implications for cropland productivity. In: Powell, J.M., FernandezRivera, S., Williams, T.O., Renard, C. (Eds.), Livestock and Sustainable Nutrient Cycling in
Mixed Farming Systems of Sub-Saharan Africa. Proceedings of an International Conference, Addis
Ababa, Ethiopia, 22–26 November 1993. ILCA (International Livestock Centre for Africa), Africa,
pp. 393–409.
R.J. Manlay et al. / Agricultural Systems 79 (2004) 83–107
107
Winter, M., 1993–2000. WebElements. The Periodic Table on the WWW. Available from: http://
www.webelements.com/.
Woomer, P.L., Palm, C.A., Qureshi, J.N., Kotto-Same, J., 1998. Carbon sequestration and organic
resource management in African smallholder agriculture. In: Lal, R., Kimble, J.M., Follett, R.F.,
Stewart, B.A. (Eds.), Management of Carbon Sequestration in Soil. CRC Press Inc, Boca Raton, pp.
153–173.