Agriculture, Ecosystems and Environment 187 (2014) 171–182
Contents lists available at ScienceDirect
Agriculture, Ecosystems and Environment
journal homepage: www.elsevier.com/locate/agee
Conservation agriculture in African mixed crop-livestock systems:
Expanding the niche
Frédéric Baudron a,∗ , Moti Jaleta a , Oriama Okitoi b , Asheber Tegegn c
a
b
c
CIMMYT-Ethiopia, Shola Campus, ILRI, P.O. Box 5689, Addis Ababa, Ethiopia
Kenya Agricultural Research Institute, P.O. Box 169, Kakamega, Kenya
Ethiopian Institute for Agricultural Research, Ethiopia
a r t i c l e
i n f o
Article history:
Received 19 February 2013
Received in revised form 1 July 2013
Accepted 20 August 2013
Available online 17 September 2013
Keywords:
Trade-offs
Maize stover
Yield gap
Intensification
Dairy
Traction
a b s t r a c t
Competition for crop residues between livestock feeding and soil mulching is a major cause of the low and
slow adoption of conservation agriculture (CA) in sub-Saharan Africa. Retaining crop residues in the field
is not only a prerequisite for CA but may also be the most viable option for African farmers to retain their
fields in a productive state. In this paper, (1) we explore the possibility of increasing the quantity of crop
residue available by closing the maize yield, (2) we propose interventions that can reduce crop residue
demand for livestock feed, and (3) we quantify the optimum amount of crop residues required as mulch,
using empirical, secondary and modeling data from Western Kenya and the Ethiopian Rift Valley. Residue
retention can also be increased by reducing livestock demand. Closing the maize yield gap–i.e. achieve
90% of the water-limited yield potential–and intensifying dairy production–which would promote the use
of rations that are more energy-dense than cereal residue-based rations–would increase the estimated
proportion of farmers retaining at least 1 t ha−1 of crop residues from the current 36% to 97% in Western
Kenya. In the Ethiopian Rift Valley, closing the maize yield gap and substituting mechanization to animal
draught power would increase the estimated proportion of farmers retaining at least 1 t ha−1 of crop
residues from the current 3% to 83%. We conclude that the question is not ‘if’, but ‘how’ cereal residues
can fulfill the demand of both the soil and the livestock.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
Retaining crop residues as surface mulch, together with minimum soil disturbance and crop rotation and association, forms the
basis of conservation agriculture (CA, Kassam et al., 2009). The environmental benefits of CA in terms of soil and water conservation
are well documented (e.g., Lal, 1998; Scopel et al., 2004). CA may
also generate a number of short-term benefits for farmers (Knowler
and Bradshaw, 2007; Rusinamhodzi et al., 2011). Reduction in
machinery and fuel costs has been one of the major incentives for
the large-scale adoption of CA in North America, South America
and Australia (Kassam et al., 2009). In the less mechanized systems in developing countries, CA may enable early planting, as the
number of operations required to prepare the land are reduced
(Haggblade and Tembo, 2003). Moreover, reduced soil water loss
from runoff and evaporation (Rockström et al., 2009; Thierfelder
and Wall, 2009) may result in more efficient use of rainfall by the
crop and yield stabilization, particularly in dry areas (Friedrich,
2008; Erenstein, 2002, 2003).
∗ Corresponding author. Tel.: +251 116 462324; fax: +251 116 464645.
E-mail address: [email protected] (F. Baudron).
0167-8809/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.agee.2013.08.020
Feeding crop residues to livestock is an alternative use particularly common in the developing world, where 75% of the milk
and 60% of the meat are produced in mixed crop-livestock systems (Herrero et al., 2010; Valbuena et al., 2012). The crop residue
demand by the livestock sector is unlikely to decrease in this part
of the world, as meat and milk consumptions are projected to more
than double by 2050 from their values in 2010 (Thornton, 2010).
A trade-off arises when a particular stakeholder faces more than
one objective towards a resource that cannot simultaneously be
achieved (Grimble and Wellard, 1997). Due to the multiple benefits livestock generates (Schiere et al., 2002; Powell et al., 2004;
Rufino et al., 2006; Herrero et al., 2010; McDermott et al., 2010),
mixed crop-livestock African farmers generally feed the bulk of
their crop residues to livestock and sacrifice on soil mulching i.e.
they trade soil mulching for livestock feeding (Valbuena et al.,
2012). When insufficient quantities of crop residues are retained
as surface mulch, minimum tillage alone may lead to lower yields
compared with the current farm practices, particularly on soils that
are prone to crusting and compaction (Baudron et al., 2012). Based
on those observations, some authors have concluded that CA would
only fit in a limited set of socio-ecological niches in Africa, which
is dominated by mixed crop-livestock systems (Giller et al., 2009,
2011; Andersson and Giller, 2012).
172
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
Table 1
Sub-objectives and methods used in the study.
Sub-objectives
Methods
Calculations from farm survey data
Quantifying current crop
residue use
Exploring alternatives to
Yield gap analysis from farm survey
increase cereal grain and
data
residue yield
Exploring alternatives to reduce livestock demand for crop residues
By stimulating livestock
Feeding trial to establish the
productivity
relationship between cereal residue
feeding and livestock productivity
By providing substitute to
Calculations from farm survey data to
the main functions livestock
quantify livestock functions
plays
On-station trial to establish the
Quantifying the short-term
benefits of soil mulching on
relationship between mulching rate
and grain yield
crop productivity
Several studies have quantified and explained crop residue
trade-offs in mixed crop-livestock systems (e.g. Erenstein, 2002,
2003; Valbuena et al., 2012) but few have explored alternatives to
feed both the livestock and the soil, and thus expand the niche in
which CA would fit. Retaining part of the crop residues in African
fields is not only a prerequisite to CA adoption: it may also be
the most viable option to maintain them in a productive state,
whether they are ploughed in or retained as surface mulch. It is
widely acknowledged that increasing the use of mineral fertilizer, currently estimated at an average 8 kg ha−1 (Groot, 2009),
is necessary to improve crop productivity in Africa (e.g. Vitousek
et al., 2009). Soils with a low soil organic carbon content, however,
generally respond poorly or not at all to mineral fertilizers (Lal,
2010; Vanlauwe et al., 2010). An adequate annual input of organic
material is crucial to maintain soil organic carbon, especially in
coarse-textured soils (Chivenge et al., 2006; Stewart et al., 2007;
Luo et al., 2010). The quantity of manure and plant biomass (other
than crop residues) available to African farmers as organic input
is generally low (Vanlauwe and Giller, 2006; Tittonell and Giller,
2012). Manure and plant biomass, when available, are also bulky
materials: their collection, transport and application are labourintensive, and they are only applied to a minority of fields (Tittonell
et al., 2005). On the other hand, crop residues represent the most
abundant organic material available to farmers (Lal, 2005) and their
retention in situ does not require any labour. Thus, retaining part
of the crop residues produced in African fields is a priority to sustain agricultural intensification in the continent. In circumstances
where the retention of these residues as surface mulch leads to
short-term benefits for farmers, it may also expand the adoption of
CA.
The objective of this study was to quantify crop residue
trade-offs and explore alternatives to overcome them in two
study sites known for their high livestock densities: Western Kenya and the Ethiopian Rift Valley. Both sites are
part of the maize mixed farming system as defined by FAO
(http://www.fao.org/DOCREP/003/Y1860E). Western Kenya corresponds to the wetter part of this domain, and the Ethiopian Rift
Valley to the drier part.
2. Materials and methods
livestock demand. These explorations were made using farm survey
data and feed trial data. Third, the impact of crop residue mulching
on crop productivity was established using on-station trial data.
2.2. Study areas
2.2.1. Western Kenya
In Western Kenya, the study was conducted in the District of
Bungoma, located between 34◦ 22 and 34◦ 03 East and 0◦ 25 and
1◦ 08 North, and the District of Siaya, located between 33◦ 57 and
34◦ 33 East and 0◦ 18 and 0◦ 26 North. Bungoma lies between 1,300
and 3,100 m above sea level, whilst Siaya lies between 1,100 and
1,500 m above sea level. Bungoma receives bimodal rainfall ranging
from 1,200 mm to 1,800 mm per year, with the long rains falling
between March and July and the short rains between August and
November. Higher regions of Siaya receive 1,800 mm to 2,000 mm
of rainfall per annum, while the lower regions receive 800 mm to
1,600 mm per annum. Average temperatures in Bungoma range
from 16 ◦ C to 30 ◦ C with an average of 23 ◦ C. In the two districts,
maize (Zea mays L.), common bean (Phaseolus vulgaris L.), cassava
(Manihot esculenta Crantz), sorghum (Sorghum bicolor (L.) Moench),
pearl millet (Pennisetum glaucum (L.) R. Br.), sweet potatoes (Ipomoea batatas (L.) Lam.), and bananas (Musa sp. L.) are the main
food crops grown, whilst sugarcane (Saccharum sp. L.), tea (Camellia sinensis (L.) Kuntze), cotton (Gossypium hirsutum L.) and tobacco
(Nicotina sp. L.) are the main cash crops. The major livestock enterprises are dairy cattle and poultry. This study area is referred to as
‘Western Kenya’ in the rest of the paper.
On-station research trials in Western Kenya were conducted at
Kakamega Agricultural Research Center. The station is located at
34◦ 45 East, and 00◦ 16 North, at an altitude of 1,585 m above sea
level, receives between 1,890 and 1,920 mm of annual rainfall, and
is characterized by a mean minimum temperature of 14 ◦ C, and
a mean maximum temperature of 27 ◦ C. Soils are dominated by
mollic Nitosols.
2.2.2. Ethiopian Rift Valley
In Ethiopia, the study was conducted in four districts of the
Ethiopian Rift Valley: Adama, Adami Tulu, Boset, and Dodota Sire,
located between 38◦ 40 and 39◦ 30 East and 7◦ 50 and 8◦ 40 North.
The area lies between 1,500 and 1,950 m above sea level, and is
characterized by low and erratic rainfalls comprised between 500
and 800 mm and high evapo-transpiration rates. Mean minimum
temperatures range from 7.8 to 14.4 ◦ C and mean maximum temperatures from 27.2 to 28.6 ◦ C. The study area is characterized
by two clearly defined seasons: a main rainy season from June
to October, and a long dry season from November to May. Tef
(Eragrostis tef (Zucc.) Trotter), maize, wheat (Triticum sp. L.) and
common bean are the main crops grown. Cattle, goats, horses and
donkeys are the main livestock types farmers keep. This study area
is referred to as the ‘Ethiopian Rift Valley’ in the rest of the paper.
On-station research trials in the Ethiopian Rift Valley were conducted at Melkassa Agricultural Research Center. The station is
located at 39◦ 12 East, and 8◦ 24 North, at an altitude of 1,550 m
above sea level, receives 763 mm of average rainfall annually, and
is characterized by a mean minimum temperature of 28.6 ◦ C, and a
mean maximum temperature of 13.8 ◦ C. Soil is dominantly loamy
and clay loamy.
2.1. General approach
2.3. Farm survey
A range of methods were used in this study, to fulfill four interrelated sub-objectives (Table 1). First, current crop residue uses
were quantified, using farm survey data. Second, alternatives were
explored to increase the quantity of crop residues retained in the
field by (1) increasing the quantity produced and (2) reducing
A sample of farms was randomly selected in the two study areas:
299 in Western Kenya and 344 farms in the Ethiopian Rift Valley. The head of each farm was interviewed using a standardized
questionnaire addressing size and composition of the household,
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
173
Table 2
Main characteristics of the ingredients used for the feeding trial.
Ingredient
DM (%)
Metabolisable energy
(Mcal kg−1 DM)
Digestibility (%)
Crude protein (g kg−1 DM)
Price (KSH kg−1 )
Maize stover
Residue from shelled maize
Napier
Soya bean meal
Mollasses
97
87
30
88
74
1.6
3.3
2.2
2.2
2.6
43
94
65
89
83
24
90
72
500
0
10
10
10
15
2
DM: dry matter, KSH: Kenyan Shilling.
production capital (e.g. land, equipment), crop and livestock production and management, and income generating activities.
2.4. Effect of soil mulching on maize productivity
On-station trials were conducted in 2011 and 2012 to test the
effect of soil mulching on maize productivity in Kakamega Agricultural Research Center and Melkassa Agricultural Research Center.
In both sites, each plot was 100 m2 , and maize was seeded in rows
spaced 75 cm apart, with a plant-to-plant spacing within a row of
25 cm. Maize grain yield was estimated by harvesting maize ears
at maturity, shelling, and determining grain moisture content by
oven-drying a grain subsample for 48 h at 60–70 ◦ C to determine its
moisture content. Similarly, stover yield was estimated by counting
the number of plants in the harvested area, sampling 15 representative ones, weighing them and determining the moisture content
of a subsample oven-dried for 48 h at 60–70 ◦ C.
In Kakamega Agricultural Research Center, four blocks were
seeded to the maize variety KSTP 94, during the long rains of 2011.
Mineral fertilizer was applied at a rate of 125 kg ha−1 of diammonium phosphate at planting and 125 kg ha−1 of calcium ammonium
nitrate was side-dressed at the six-leaf growth stage. After harvest,
each block was divided into five treatments with 0, 25, 50, 75 and
100% of crop residues retained as surface mulch. During the following seasons–i.e. short rains of 2011 and long rains of 2012–maize
was seeded through the mulch by opening shallow planting stations, without prior land preparation. For each plot, the same level
of residue retention was used during the different seasons.
In the trial conducted at Melkassa Agricultural Research Center,
three blocks were seeded to the maize variety Melkassa 2 during the
2011 season. Mineral fertilizer was applied at a rate of 100 kg ha−1
of diammonium phosphate and 25 kg ha−1 of urea at planting and
25 kg ha−1 of urea was side-dressed at the six-leaf growth stage.
After harvest, each block was divided in four treatments with 0,
33, 67 and 100% of crop residues retained as surface mulch. During
the next season (2012), maize was seeded through the mulch by
opening shallow planting station, without prior land preparation.
2.5. Feeding trials on dairy cows
In 2012, seven lactating Friesian cows were selected in
Kakamega research station, on the basis of similar live weight
(mean of 367 ± 13 kg), lactation phase (mid-lactation) and milk
yield prior to the feeding trial (mean of 9.9 ± 3.1 kg milk day−1 ).
Six experimental diets containing varying levels of maize residues
(maize stover, and residues from shelled maize), napier grass, and
concentrates (soya bean meal, and molasses) were formulated
based on their total neutral detergent fiber (NDF), total metabolisable energy (ME) and total crude protein content (CP). Each diet
aimed at providing enough energy to support a milk yield of
20 L day−1 , for a cow in mid-lactation (i.e. requiring a CP in the
ingested ration between 140 and 160 g kg−1 DM). The nutritive
values of the different ingredients are shown in Table 2, and the
composition of each diet in Table 3.
Table 3 Prior to being mixed in rations, the napier grass (variety
Kakamega 3) and the maize stover were chopped, the residue from
shelled maize was broken using a hammer mill without a sieve, the
soya bean meal was milled to pass through 2 mm sieve, and the
molasses was diluted with water at a 1:2 ratio. Ingredients were
then weighed precisely and mixed thoroughly in a drum mixer.
The feeding trial was of switch-over type, each cow receiving each
ration for a period of 25 days (i.e. 10 days for ration adjustment and
then 15 days for data collection) with a 3 day cross-over period.
The feed was offered twice a day in the morning and afternoon.
The amounts of feed offered and refused were recorded daily. Milk
yield was recorded at each milking.
2.6. Statistical tests, calculations, and models
Normal distribution was tested using Kolmogorov–Smirnov
tests. Means of quantitative data following a normal distribution
were compared using Fisher tests or Student tests (parametric
tests). Medians of quantitative data not following a normal distribution were compared using Kruskal–Wallis tests. These analyses
were carried out with the software Statgraphic (version XV).
2.6.1. Quantification of current crop residue uses
For each farm and each major cereal, the quantity of residues
used for mulching, livestock feed, household fuel and other uses
(e.g. construction, burnt in the field) were calculated by multiplying the fraction for each use - as estimated during the farm survey
- by the estimated quantity of cereal residues. In Western Kenya,
the major cereal was maize. In the Ethiopian Rift Valley, the major
cereals were maize, tef, wheat, barley and sorghum. Other crops
were cultivated on small surface areas relatively to these cereals.
The quantities of cereal residues produced by each farm were calculated from the quantities of grain produced–as reported during
the farm survey–and harvest indexes: we used 0.23 for tef, 0.26 for
wheat, 0.3 for barley, 0.5 for maize, and 0.4 for sorghum (Keftasa,
1987; Nijhof, 1987).
2.6.2. Boundary line models and yield gap analysis
Based on the survey data from the two study areas, maize yield,
and nitrogen (N) and phosphorus (P) fertilization rates of maize
were calculated. A yield gap analysis was performed using boundary line models as described by Fermont et al. (2009), using N
fertilization and P fertilization as independent variables. First, outliers were identified and removed from the analysis (yield data in a
given fertilization class exceeding 1.5 interquartile range above the
Table 3
Composition of the six diet used for the feeding trial.
Diet
Diet1
Diet 2
Diet 3
Diet 5
Diet 4
Diet 6
Maize stover
Residue from
shelled maize
Napier
Soya bean meal
Mollasses
10
64
45
30
65
6
13
44
26
21
42
5
6
14
6
1
23
1
0
29
0
9
16
18
15
17
21
13
22
18
174
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
WESTERN KENYA
ETHIOPIAN RIFT VALLEY
LONG RAINS
MAIZE
SHORT RAINS
TEFF
Soil
Feed
Fuel
2.7. Scenarios
WHEAT
BARLEY
Other
Fig. 1. Mean use of maize stover in Western Kenya during the long rains and the
short rains (estimated from 299 farms), and of maize stover, tef straw, wheat straw
and barley straw in the Ethiopian Rift Valley (estimated from 344 farms).
A number of scenarios were used to predict the consequences of
changes in residue production and/or residue feeding on the quantity of residues retained in the field. For scenarios of changes in
residue production, the quantities of each cereal residue used as
feed, fuel and for other uses were kept to their current value, and
the remainder was assumed to be retained in the field. For scenarios
of changes in residue feeding, the demand for cereal residues per
livestock type was estimated for each farm, by converting livestock
numbers into Tropical Livestock Units (TLU), which is defined as a
hypothetical animal of 250 kg live weight (Le Hourou and Hoste,
1977). Bulls and oxen were assumed to be equivalent to 1.1 TLU,
cows to 0.8 TLU, heifers to 0.5 TLU and calves to 0.2 TLU (Machila
et al., 2003; Wondatir, 2010). Feeding by small ruminants and
equines was ignored, as these are minor livestock types compared
with cattle in the two study sites.
3. Results
3.1. Farm survey data
third quartile). Then, logistic regression models were fitted through
the upper boundary points on a scatter plot, i.e. the maximum
maize yield response for each level of the independent variable.
This analysis was done with Genstat 6.1 (2002). For each maize plot
of each farm surveyed, attainable yield was calculated as the minimum of the N-limited attainable yield (obtained from the logistic
regression model for N) and the P-limited attainable yield (obtained
from the logistic regression model for P). Actual yields were then
expressed as a fraction of the attainable yield.
3.1.1. Analysis of current crop residue use
In Western Kenya, the majority of maize residues produced during the long rains and during the short rains was retained in the
field as soil amendment (mean values of 55 ± 38% and 61 ± 37%,
respectively, Fig. 1). Livestock feeding was the second use during
both seasons (mean values of 28 ± 30% during the long rains and
27 ± 30% during the short rains). A minor fraction of maize residues
was used as fuel, both during the long and short rains (mean values
of 7 ± 16% and 3 ± 11%, respectively). In the Ethiopian Rift Valley,
80%
30%
(a)
20%
36%
20%
0%
0%
80%
(c)
20%
49%
Frequency
60%
25%
20%
0%
0%
80%
(e)
20%
93%
60%
(f)
60%
40%
10%
20%
0%
0%
80%
30%
(d)
40%
10%
30%
3%
40%
10%
30%
(b)
60%
(g)
20%
97%
60%
(h)
83%
40%
10%
20%
0%
0%
0
1 to 501 1001 1501 2001 2501 3001 3501 4001
500 to to to to to to to or
1000 1500 2000 2500 3000 3500 4000 more
0
1 to 501 to 1001 1501 2001 2501 3001 3501
500 1000 to
to
to
to
to
or
1500 2000 2500 3000 3500 more
Quanty of residues retained in the field (kg ha-1)
Fig. 2. Frequency distribution of farms according to their current rate of residue retention in (a) Western Kenya and (b) the Ethiopian Rift Valley; frequency distribution of
farms according to their projected rate of residue retention in Western Kenya assuming a scenario (c) of livestock intensification–i.e. farmers feeding their cows with rations
containing no maize residue, (e) of farmers “closing the maize yield gap” - i.e. achieving 90% of the water-limited yield potential, and (h) of livestock intensification and
closing the yield gap; and frequency distribution of farms according to their projected rate of residue retention in the Ethiopian Rift Valley assuming a scenario of all farmers
(d) substituting mechanization to animal draught power, (f) closing the yield gap, and (h) substituting mechanization to animal draught power and at the same time closing
the yield gap. From (a) to (c), (e), and (g), the proportion of farmers in Western Kenya retaining at least 1 t ha−1 of residues in the field increases from 36% to 49%, 93% and
97%. From (b) to (d), (f), and (h), the proportion of farmers in the Ethiopian Rift Valley retaining at least 1 t ha−1 of residues in the field increases from 3% to 25%, 60% and 83%.
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
8
(a)
7
7
6
6
Yield (kg ha-1)
Yield (kg ha-1)
8
5
4
3
3
2
1
0
0
8
50
100
N (kg ha-1)
150
0
200
9
(c)
8
20
40
P (kg ha-1)
60
80
(d)
7
Yield (kg ha-1)
7
Yield (kg ha-1)
4
1
9
(b)
5
2
0
175
6
5
4
3
6
5
4
3
2
2
1
1
0
0
0
40
80
N (kg ha-1)
120
0
10
20
30
P (kg ha-1)
40
50
Fig. 3. Maize grain yield in Western Kenya as a function of (a) nitrogen fertilization and (b) phosphorus fertilization, and maize grain yield in the Ethiopian Rift Valley as a
function of (c) nitrogen and (d) phosphorus. The black dotted lines represent the boundary lines, and the grey dotted lines the fertilization rate necessary to attain 90% of the
water-limited yield. The boundary line represent the N-limited yield (a and c) or the P- limited yield (b and d). The highest value of the boundary line represents the waterlimited yield potential. The dotted represent regression models for the attainable yield: (a) ATTYLD = 8235 − 3687 × e−0.055×N , R2 = 0.83; (b) ATTYLD = 9000 − 4366 × e−0.047×P ,
R2 = 0.94; (c) ATTYLD = 7438 − 2937 × e−0.010×N , R2 = 0.88; and (d) ATTYLD = 6847 − 2203 × e−0.035×P , R2 = 0.84; with ATTYLD the maize attainable grain yield in kg ha−1 , N the
N application rate in kg ha−1 , and P the P application rate in kg ha−1 .
the majority of maize, tef, wheat, barley, and sorghum residues was
used as livestock feed (mean values of 67 ± 22%, 88 ± 23%, 85 ± 24%,
88 ± 25%, 58 ± 33%, respectively). A small fraction of maize and
sorghum residues was used as fuel (22 ± 17% and 17 ± 19%, respectively), while no tef, wheat, and barley residue was used as fuel. A
minor fraction of maize, tef, wheat, barley, and sorghum residues
was retained in the field as soil amendment (7 ± 14%, 7 ± 20%,
9 ± 16%, 7 ± 16%, 9 ± 19%).
Currently, 19% of the farmers in Western Kenya (Fig. 2a) and 69%
of the farmers in the Ethiopian Rift Valley (Fig. 2b) are not retaining
any crop residue in their fields. Only 36% of the farmers in Western
Kenya and 3% of the farmers in the Ethiopian Rift Valley currently
retain 1 t ha−1 of crop residues or more.
3.1.2. Yield gap analysis for maize
The highest estimated yield attained in Western Kenya was
6750 kg ha−1 during the long rains of 2011 and 2700 kg ha−1 during
the short rains of 2011 (Fig. 3a and b). It was 8120 kg ha−1 in the
Ethiopian Rift Valley in 2011 (Fig. 3c and d). These yields can be
considered the water-limited yield potentials, defined as the maximum yield achievable under rainfed conditions without irrigation
if soil water capture and storage is optimal and nutrient constraints
are released (Tittonell and Giller, 2012). When using N application
rate and P application rate as independent variables, the boundary lines showed an increase in the attainable yield with increasing
application rates of N and P in both sites, illustrating the fact that
maize yield is limited by N and P in both sites. Table 4 gives the
values of N and P application rates required to achieve 90% of the
water-limited yield potentials in 2011 in Western Kenya (during
the short and the long rains) and in the Ethiopian Rift Valley. These
values were calculated using the logistic regression models fitted
through the upper boundary points of maize yield as a function of
N and P fertilization rates.
Table 4 Respectively 22% and 32% of the farmers in Western
Kenya and the Ethiopian Rift Valley achieved 50% or more of the
attainable yield given the N and P application rates used (Fig. 4).
This demonstrates low N and P use efficiency due to limitations in
other production factors and/or high incidence of yield reducing
factors such as weeds, pests and diseases.
3.1.3. Analysis of livestock functions
The density of cattle in the Ethiopian Rift Valley was significantly higher than in Western Kenya (means of 5.69 ± 6.11
Table 4
Nitrogen (N) and phosphorus (P) fertilization rate required to reach 90% of the waterlimited yield potential in Western Kenya during the short rains (October–November)
and the long rains (March–June) and in the Ethiopian Rift Valley.
Western Kenya Short Rains
Western Kenya Long Rains
Ethiopian Rift Valley
N (kg ha−1 )
P (kg ha−1 )
90% of
water-limited
yield
122.8
75.7
25.0
45.5
30.3
20.3
2430
6075
7308
176
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
1
1
(b)
Probability of exceedance
Probability of exceedance
(a)
0.8
0.6
0.4
0.2
0
0.8
0.6
0.4
0.2
0
0%
20%
40%
60%
80%
100%
0%
20%
40%
60%
80%
100%
Yield (fracon of the aainableyield)
Fig. 4. Cumulative probability distribution of farmer’s maize grain yield expressed as the fraction of the attainable yield at a given N and P fertilization rate (a) in Western
Kenya and (b) in the Ethiopian Rift Valley.
TLU ha−1 cropland and 4.04 ± 6.20 TLU ha−1 cropland, respectively,
Table 5). In Western Kenya, the herd was dominated by cows
(mean of 62 ± 32% of the total TLU). In the Ethiopian Rift Valley,
oxen were the predominant cattle type (mean of 42 ± 24% of the
total TLU) followed by cows (mean of 36 ± 21% of the total TLU).
Although there was no difference between the Ethiopian Rift Valley and Western Kenya in the calf/cow ratio (means of 0.60 ± 0.42
and 0.61 ± 0.59, respectively), the heifer/cow ratio was significantly higher in the Ethiopian Rift Valley compared with the ratio
in Western Kenya (means of 0.42 ± 0.54 and 0.17 ± 0.48, respectively). Manure application on cropland was significantly higher in
Western Kenya compared with that in the Ethiopian Rift Valley
(means of 1700 ± 3882 kg ha−1 and 423 ± 1003 kg ha−1 , respectively). The intensity of manure collection from cattle was also
significantly higher in Western Kenya compared with the Ethiopian
Rift Valley (means of 664 ± 1766 kg TLU−1 and 81 ± 172 kg TLU−1 ,
respectively). Milk production was also significantly higher in
Western Kenya compared with the Ethiopian Rift Valley (means
of 421 ± 718 L cow−1 year−1 and 160 ± 171 L cow−1 year−1 respectively).
Table 5 In Western Kenya, the number of dairy cows and the
quantity of manure applied on the farm had a significant influence on the cereal production of the farm (F = 31.78, P < 0.001,
Fig. 5a). In 2011, farmers having no dairy cows and applying less
than one ton of manure in their farms produced an average of
1208 ± 1999 kg cereals, whilst farmers having no dairy cow but
applying at least one ton of manure in their farms produced an
average of 2064 ± 2695 kg cereals, and farmers having at least one
dairy cow produced an average of 4659 ± 7979 kg cereals. In the
Ethiopian Rift Valley, the number of pairs of oxen owned had a significant influence on the cereal production of the farm (F = 51.91,
P < 0.001): in 2011, farmers having no oxen produced an average
of 1934 ± 2227 kg cereals, whilst farmers having one pair of oxen
produced an average of 3989 ± 3108 kg cereals, and farmers having
at least two pairs of oxen produced an average of 6812 ± 6023 kg
cereals (Fig. 5b).
3.2. Trial data
3.2.1. Effect of soil mulching on maize grain yield
During the short rains in Kakamega, the quantity of surface
mulch in the different treatments was between 2.0 and 3.2 t ha−1
(Fig. 6a). However, the quantity of mulch retained had no effect on
maize grain yield. During the long rains in Kakamega, as well as in
Melkassa, maize grain yield was found to increase non-linearly with
the quantity of soil mulch applied (Fig. 6b and c). During the long
rains of Kakamega, a plateau was reached in maize grain yield with
a surface mulch of approximately 1 t ha−1 of maize stover, implying
that quantities of maize stover beyond this threshold did not influence maize grain yield significantly, and could thus be allocated to
other uses–e.g. livestock feeding–with no yield penalty. Similarly
in Melkassa, a plateau was reached with a surface mulch of 3 t ha−1 .
Table 5
Characteristic of the average cattle herd in Western Kenya and in the Ethiopian Rift Valley.
Characteristics
Western Kenya
Ethiopian Rift Valley
P
Cattle density (TLU/ha cultivated)
Herd composition (% TLU)
Oxen
Bulls
Cows
Heifers
Calves
Herd management
Heifer/Cow
Calf/Cow
Manure
Applied (kg/ha)
Collected (kg/TLU)
Milk (L/cow)
4.04 ± 6.20
5.69 ± 6.11
<0.001
±
±
±
±
±
0.24
0.11
0.21
0.15
0.07
<0.001
n.s.
<0.001
<0.001
<0.001
0.17 ± 0.48
0.61 ± 0.59
0.42 ± 0.54
0.60 ± 0.42
<0.001
n.s.
1700 ± 3882
664 ± 1766
421 ± 718
423 ± 1003
81 ± 172
160 ± 171
<0.001
<0.001
<0.001
0.12
0.10
0.62
0.05
0.12
±
±
±
±
±
0.26
0.21
0.32
0.14
0.19
0.42
0.07
0.36
0.10
0.05
Standard errors are given after the signs ‘ ± ’. P values are given for Kruskal–Wallis tests comparing medians between Western Kenya and the Ethiopian Rift Valley.
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
14
(a)
13%
F = 31.78
P < 0.001
12
10
8
6
22%
4
65%
2
0
Cereal producon (t farm-1)
Cereal producon (t farm-1)
14
177
(b)
17%
F = 51.91
P < 0.001
12
10
8
47%
6
36%
4
2
0
0 dairy cow,
< 1 t manure
0 dairy cow,
≥ 1 t manure
≥ 1 dairy cow
0 pair
1 pair
≥ 2 pairs
Fig. 5. Cereal production during the year 2011 (a) in Western Kenya for farmers having no dairy cow and applying less than 1 t of manure in their fields, for farmers having
no dairy cow and applying at least 1 t of manure in their fields, and for farmers having at least one dairy cow; and (b) in the Ethiopian Rift Valley for farmers having no oxen,
for farmers having one pair of oxen, and for farmers having at least two pairs of oxen. Bars represent standard errors. Percentages represent the proportions of the different
farm types. For each graph, a summary of the Fisher test comparing the means is given.
This higher threshold compared to that in Kakamega may be due to
the fact that rainfall in Melkassa is lower and more erratic than in
Kakamega, whilst evapotranspiration is higher. It may also be due
to differences in soil texture and soil organic carbon content.
4
(a)
3.2.2. Effect of maize residue feeding on dairy production
In the feeding trial conducted in Kakamega, the production
cost per litre of milk decreased with decreasing amounts of maize
residues in the ration of dairy cows (Fig. 7a). Moreover, the maximum cost per litre of milk production decreased with increasing
milk yield (Fig. 7b). Therefore, the profitability of milk production
tended to increase with decreasing amounts of maize residues in
the ration and with increasing milk yields.
3
3.3. Scenarios
2
1
0
Maizegrain yield(t ha-1)
0
2
4
6
8
8
(b)
6
4
2
0
0
2
4
6
8
(c)
6
4
2
0
0
1
2
3
4
Surface mulch(t ha
5
6
-1)
Fig. 6. Maize grain yield as a function of the quantity of surface mulch (a) in
Kakamega Agricultural Research Centre during the October 2011–November 2011
(long rains) rainy season, (b) in Kakamega Agricultural Research Centre during
the March 2011–June 2011 (long rains) rainy season, and (c) in Melkassa Agricultural Research Centre during the 2011 rainy season. The dotted lines represent
regression models. The dotted lines in (b) and (c) represent regression models: (b)
YLD = 4673–1938 × 0.997M , R2 = 0.47; (c) YLD = 4328–2564 × 0.499M , R2 = 0.23); with
YLD the maize grain yield in kg ha−1 and M the quantity of maize stover applied as
surface mulch in kg ha−1 .
Assuming a scenario where all farmers in Western Kenya feed
their cows with rations containing no maize residue but with
rations composed of ingredients having lower fiber contents and
higher energy densities, the proportion of farmers not retaining
any residue in their field is predicted to decrease from 19% to 8%
(Fig. 2a and c). Concomitantly, the proportion of farmers retaining at least 1 t ha−1 of crop residue would increase from 36% to
49%. Assuming a scenario where all the farmers in Western Kenya
‘close the yield gap’–i.e. achieving 90% of the water-limited yield
potential–the proportion of farmers not retaining any residue in
their fields would decrease further to 3% and the proportion of
farmers retaining at least 1 t ha−1 of crop residue would increase
further to 93% (Fig. 2e). Finally, assuming a scenario where all farmers in Western Kenya both feed their cows with rations containing
no maize residues and close the yield gap, the proportion of farmers not retaining any residue in their fields would be 1% only, and
the proportion of farmers retaining at least 1 t ha−1 of crop residues
would be as high as 97% (Fig. 2g).
Assuming a scenario where all farmers in the Ethiopian Rift Valley substitute tractors to their oxen, the proportion of farmers not
retaining any residue in their field is predicted to decrease from
69% to 18% (Fig. 2b and d). Concomitantly, the proportion of farmers
retaining at least 1 t ha−1 of crop residues would increase from 3%
to 25%. Assuming a scenario where all the farmers in the Ethiopian
Rift Valley would close the yield gap, the proportion of farmers
not retaining any residue in their fields would further decrease to
13% and the proportion of farmers retaining at least 1 t ha−1 of crop
residues would increase further to 60% (Fig. 2f). Finally, assuming a
scenario where all farmers in the Ethiopian Rift Valley both substitute tractors to their oxen and close the yield gap, the proportion of
farmers not retaining any residue in their fields would be 5% only,
and the proportion of farmers retaining at least 1 t ha−1 of crop
residues would be as high as 83% (Fig. 2h).
178
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
18
(a)
14
12
10
8
6
4
2
(b)
16
-1)
16
Production cost(KSH L
Production cost(KSH L
-1)
18
14
12
10
8
6
4
2
0
0
0
10
12
2
4
6
8
Intakeof maizeresidue(kg day -1)
8
10
12
14
16
Milk yield (L day -1)
18
20
Fig. 7. Production cost of milk as a function of (a) the daily intake of maize residues, and (b) the milk yield, for Friesian dairy cows in Kakamega Agricultural Research Centre.
4. Discussion
4.2. Closing the yield gap to increase maize stover yields
4.1. Current cereal residue use
Maize yield in both sites is limited by N and P, as the attainable
yield increase with increasing application rates in these nutrients
(Fig. 3). However, the efficiency with which N and P are used is low,
as the large majority of farmers (78% in Western Kenya and 68% in
the Ethiopian Rift Valley) do not produce half of the yield they are
expected to achieve given the quantity of N and P applied (Fig. 4).
This may be the results of a high incidence of yield-reducing factors
such as weeds, pests and diseases, and/or of poor response to N and
P.
Simple management options such as increasing crop density
(Olsen et al., 2005) or combining low doses of pre-emergence
herbicide with manual weeding (Versteeg and Maldonado, 1978)
may control weeds effectively. CA may also lead to a reduction
in weed pressure over time, as reduced soil disturbance generally
lead to a reduction in seed banks, rhizomes and tubers in the soil
(Swanton and Weise, 1991). Cover crops, often used in CA systems,
may also control weeds through shading and/or allelopathic effects
(Teasdale, 1996). Pests can be repelled from the crop by exploiting
semiochemicals produced by particular plants (e.g. ‘push-pull’ system controlling stem borer in Kenya, Hassanali et al., 2008). Finally,
the incidence of diseases can be reduced by the use of geneticallyresistant varieties (e.g. maize lines resistant to Streak virus, Rodier
et al., 1995).
Crop response to a given nutrient is in part regulated by the
availability of other limiting resources (Janssen et al., 1990) and
the fulfillment of other plant requirements in terms of e.g. aeration
and physical support (Vanlauwe et al., 2010). Water is probably
a limiting factor in both sites as demonstrated by the positive
response of maize yield to increasing rates of surface mulch (Fig. 6).
In most cases, surface mulching increases water storage in the soil
thanks to reduced water runoff and soil evaporation (Findeling
et al., 2003; Scopel et al., 2004). The use of cover crops or trees
may lead to similar benefits (Burgess et al., 1998; Ong et al.,
2000).
If all farmers in both sites were to close the maize yield gap–by
increasing the quantity of nutrients applied, using site-specific
technologies to control yield reducing factors, and ensuring maximum nutrient use efficiency, the majority would be able to retain
1 t ha−1 of crop residues or more (Fig. 2e and f). From the large basket of technologies available to achieve this goal, adoption will be
the result of complex interactions between the technology itself
and the farming context including e.g. labour demand vs. labour
availability, cost vs. financial capital, information about the technology, knowledge and skills required, input and output market,
In Western Kenya, the majority of farmers retained over 80%
of crop residues in their field (Fig. 2a), as there appear to be little competition with other uses such as feed or fuel (Fig. 1). This
is consistent with previous studies conducted in the study area
(e.g. Jaleta et al., in press). However, only about a third of the
farmers retained at least 1 t ha−1 of crop residue. Other farmers retained quantities of crop residues that may be too low to
have a significant impact on soil organic carbon and other soil
parameters.
In the Ethiopian Rift Valley, the bulk of the cereal residues produced is fed to livestock: over 80% of all the tef, wheat and barley
straw, and about two thirds of the maize and sorghum stover
(Fig. 1). About 20% of maize and sorghum stover was also used
as fuel. As a result, the majority of farmers in the Ethiopian Rift
Valley did not retain any crop residue in their fields (69%, Fig. 2b).
Only 3% of the farmers in this site retained at least 1 t ha−1 of crop
residue. Previous studies have also found that the bulk of the crop
residues produced in the Ethiopian Rift Valley is fed to livestock
(e.g. Wondatir, 2010).
The lower use of crop residue for feed in Western Kenya
compared with the Ethiopian Rift Valley may be the result of
a lower cattle density (Table 5). It can also be attributed to a
more intensive livestock production system–as illustrated by the
higher proportion of cows in the herd, the higher milk production, and the lower heifer/cow ratio. Indeed, reliance on crop
residue tends to diminish when livestock production intensifies,
as more energy-dense rations become necessary (see Section 4.3.1
below).
The adoption of CA sensu stricto in the Ethiopian Rift Valley
(i.e. including crop residue mulching) is thus not possible without increasing the quantity of crop residues retained in the field.
Although the bulk of crop residues is retained in the field in Western Kenya, for the majority of farmers, this translate to quantities
per surface area that may be too small to yield any benefit, and
may even be detrimental if retained as surface mulch (Baudron
et al., 2012). In both sites, expanding the niche where CA fits–and
more generally feeding both the livestock and the soil–requires
to increase the quantity of crop residues retained as mulch. This
can be achieved (1) by improving the quantity and quality of the
available crop residues, and (2) by reducing crop residue demand
for livestock feeding. These alternatives are explored in the next
sections.
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
and policies. Meanwhile, if these technologies can increase the
quantity of crop residues produced, their adoption will only lead to
increased quantities of residues retained in farmers’ fields where
they are accompanied with interventions aiming at reducing crop
residue demand for livestock feeding. This is explored in the following section.
4.3. Reducing crop residue demand for livestock feeding
4.3.1. Providing incentives to increase livestock productivity
The profitability of milk production tended to increase with
decreasing amounts of maize residues in the ration (Fig. 7a). This
is probably due to the high price of maize residue relative to
its low nutrient density and its high fiber content. Indeed, for a
given milk price, the profitability of dairy production is a function of the feeding costs, the genetic potential of cows (in terms
of milk yield), and the extent to which cows fulfill their genetic
potential with the available feed. The latter depends on the daily
energy intake, which is determined by feed intake, feed energy
concentration and digestibility. As the digestion and transit of
fiber is slow compared to e.g. starch, feed intake decreases due
to ruminal fill with increasing total fiber content (measured by
the NDF). Feed intake of cows appears to be maximal with NDF
between 25 and 30% (VandeHaar and St-Pierre, 2006). In comparison, the mean NDF value of maize stover in Ethiopia is 77.9 ± 3.1%
(http://192.156.137.110/feeddb/FeedDB.html). This analysis suggests that dairy producers would respond to increased market
demand (stimulating the goal of profit maximization through
dairy) by shifting from feed rations composed mainly of cereal
residues to feed rations with lower fiber content and higher
energy density. This agrees with the model proposed by Romney
et al. (2004), who state that the quantity of crop residues fed
to livestock as a function of the degree of intensification can be
described by an inverted U-shaped curve. As livestock production
systems intensify, crop residues are first substituted to grazing,
and planted fodders and supplements are later substituted to crop
residues.
Milk production cost per litre also tended to decrease with
increasing milk yield (Fig. 7b). This is due to the fact that fixed costs
(i.e. energy required to maintain the cow’s normal metabolism - e.g.
breathing, maintaining body temperature) decrease relative to total
costs (i.e. energy for maintenance and production) when individual
productivity increases (Vandehaar, 1998). A typical cow producing 15 kg of milk per day exports half of the energy it ingests (the
other half being used for maintenance), whereas a cow producing 45 kg of milk per day exports three quarters of the energy it
ingests (Vandehaar and St-Pierre, 2006). As a result, dairy farmers generally intensify by selecting for animals with increasingly
high genetic milk yield potential. For example, when shifting from
milk production for self-consumption to milk production for the
market, dairy producers of the Kenyan highlands tend to shift from
East African Zebu breeds to high-yielding breeds such as Friesian
and Ayrshire (Bebe et al., 2003). As genetic selection increases
energy requirement but does not increase the ruminal capacity
of improved animals (Coppock, 1985), they require rations that
are increasingly energy-dense i.e. containing no or little cereal
residue.
Therefore, livestock producers of the developing world are likely
to respond to increasing market demand for livestock products by
using rations that are poor in cereal residues but rich in energydense ingredients, and by selecting for cows with high individual
productivity and requiring energy-dense rations to fulfill their
genetic potential. If all farmers in Western Kenya were to intensify
dairy production to a level where cows would be fed on rations
containing no maize residues, the proportion of farmers unable
to retain any crop residue in their fields would halve (Fig. 2c) At
179
high intensification level, farmers may feed their dairy cows with
high-value forage produced on-farm or purchased, agro-industrial
by-products such as cakes from oil crops, and from grain in situation where this is profitable (i.e. value-addition) and does not
threatens food security (Lenné and Thomas, 2005). However, livestock intensification may also result in a reduction of the resilience
of smallholder systems, due to increased reliance on external
inputs and a possible shift from many animals with low productivity to fewer animals with higher productivity (Moll et al., 2007).
Moreover, the reliable supply of improved animal genetics and of
veterinary products and services, however, is a major barrier to
livestock intensification in most of sub-Saharan Africa (McDermott
et al., 2010).
4.3.2. Providing substitutes to the current functions of livestock
In Western Kenya, milk production is an important function
played by livestock, as demonstrated by the facts that the cattle
herd was dominated by cows, and that milk productivity per cow
was almost three times higher than in the Ethiopian Rift Valley
(Table 5). Manure production is another important function played
by livestock in this site, as demonstrated by an average collection
intensity of manure eight times higher than in the Ethiopian Rift
Valley, and by an average application rate of manure four times
higher. The importance of these two functions is further illustrated
by the fact that the number of dairy cows and the quantity of
manure used in the farm have a significant influence on crop production in Western Kenya (Fig. 5a).
Intensification of dairy production in Western Kenya - to a
level where cows would be fed on rations containing no maize
residue–would lead to half of the farmers to be able to retain at least
1 t ha−1 crop residues in their field (Fig. 2c). Such intensification
may lead to a reduction in livestock number, which may ultimately
threatens manure production, the second important function of
livestock in this site. However, manure use efficiency could be
improved by simple technics such as flooring and roofing of the
stall (Rufino et al., 2006).
In the Ethiopian Rift Valley, animal traction is an important function played by livestock, as demonstrated by the fact that the cattle
herd was dominated by oxen (Table 5). Moreover, the number of
pairs of oxen owned by a farm had a significant influence on the crop
production of the farm (Fig. 5b). Therefore, we hypothesized that
the adoption of cheap and low powered tractors–e.g. two-wheel
tractors–to replace oxen would significantly increase the quantity
of crop residues retained in the field. From a sustainability point of
view, mechanization is dependent on fossil reserves deposits whilst
draught animals are sustained by short-term nutrient deposits–i.e.
biomass (Schiere et al., 2002). However, the opportunity cost of
labour, land and capital for keeping draught animals the whole year
while they may only be put to productive work few days per year
should also be considered (Ehui and Polson, 1993). More than half
of the farmers would retain 1 t ha−1 crop residues or more in their
fields if tractors were substituted to oxen in the Ethiopian Rift Valley
(Fig. 2d).
4.4. How much surface mulch is required?
Through a combination of interventions, almost all farmers
would be able to retain at least 1 t ha−1 cereal residues in their
fields (Fig. 2g and h). The quantity of biomass available for mulching
being acknowledged as a major limitation to CA in Africa, this would
expand the niche where CA fit. However, farmers are only likely to
retain this biomass as surface mulch and not plough it in if the
practice leads to a short-term increase in maize grain yield. In the
following section, we question what quantity of mulch is optimal
to increase maize grain yield in the two study sites.
180
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
In Western Kenya, surface mulching had no effect on maize
grain yield during the short rains (Fig. 6a), while quantities of maize
stover applied as surface mulch in excess of 1 t ha−1 did not improve
maize grain yield during the long rains (Fig. 6b). Similarly in Melkassa, quantities of maize stover applied as surface mulch in excess of
3 t ha−1 did not improve maize grain yield (Fig. 6c). This suggests
that (1) soil mulching does not always improve maize grain yield,
and (2) when soil mulching is beneficial, maize grain yield does
not increase linearly with the quantity of surface mulch applied.
These results agree with the ones of Larbis et al. (2002) who found
that increased residue retention resulted in increased crop yield
up to a retention rate of 50%, but that further mulch retention
did not improve crop yield significantly. Therefore, the target of
30% soil cover often used in CA (Erenstein, 2003) may not lead to
the most profitable crop residue allocation. This target originated
from the US Corn Belt Region, a region that cannot be compared
with sub-Saharan Africa (Blanco-Canqui and Lal, 2009). Therefore, there is a need to understand the site-specific crop response
to mulching, from which appropriate recommendations can be
formulated.
Under certain circumstances, mulching may lead to yield
penalty. Cereal residues have a wide C:N ratio and their decomposition may lead to temporary N immobilization (Palm et al., 2001;
Zibilske et al., 2002). In high rainfall areas, mulching may exacerbate water-logging (Rusinamhodzi et al., 2011). Conversely, in
areas receiving frequent and small amounts of rainfall, mulching
may reduce water infiltration as rainwater is intercepted by the
mulch, temporarily stored, and subsequently lost to evaporation
(Cook et al., 2006; Kozak et al., 2007). In areas characterized by
periods of prolonged drying (e.g. where a distinct rainy season is
followed by a distinct dry season) a surface mulch may facilitate
water flow from the soil to the atmosphere through capillarity,
by maintaining the topsoil wetter for a longer period of time, and
increasing the evaporation rate compared with a bare soil (Unger
and Vigil, 1998).
In areas where mulching leads to yield penalty and/or where
the quantities of crop residues available for mulching are too small
(due e.g. to low crop productivity), minimum-tillage with no mulch
may still improve water balance and short-term crop productivity
if surface rugosity is purposefully increased. In the semi-arid area
of Ethiopia, strip tillage was found to result in significantly higher
maize yield and lower surface runoff compared with conventional
tillage systems (Temesgen et al., 2012). For hand-hoe based systems, planting basins have been shown to increase soil moisture
content compared to conventional ploughing in Zimbabwe and
Zambia (Mupangwa et al., 2008). In addition to these tillage practices, a variety of structures – such as contour bunds, grass strips,
pits, furrows, dikes, and terraces – may contribute to soil and water
conservation at plot- and landscape-level in the absence of surface
mulch (Vohland and Barry, 2009).
Thus, the availability of crop residues for soil mulching may
not be as limiting for the wide adoption of CA in Africa as suggested in previous analyses (e.g. Giller et al., 2009, 2011; Andersson
and Giller, 2012). Indeed, crop yield does not appear to increase
linearly with increasing quantities of surface mulch in many situations, implying that crop residues could be shared between soil
mulching and livestock feeding without negative consequence for
crop productivity. Under other circumstances, mulching may not be
desirable and thus competition for residue between CA and livestock may not exist. Last but not least, ‘CA without mulch’ may
yield a number of benefits for smallholders, including cost- and or
labour-savings during land preparation. If implemented in combination with other soil and water conservation measures, it can
gradually raise biomass production up to a level where ‘full CA’ (i.e.
including all three principles) can be implemented (Tittonell et al.,
2012).
5. Conclusions
Through this study, we suggest that a large basket of technologies is available to allow Eastern African farmers to retain
substantial quantities of cereal residues in their fields without
threatening their livestock production. Residue retention is not
only a prerequisite to CA but it is also the most viable option
African smallholders have to maintain their fields in a productive state. As the need for agricultural intensification in Africa
becomes more pressing, fueled by increased demand for agricultural products, ‘feeding’ both the soil and the livestock becomes
a necessity. We argue that the focus on trade-offs and competition for scarce crop residues diverts research efforts away from
proposing alternatives: the question should not be ‘if’, but ‘how’
crop residues can fulfill the need of both the soil and the livestock. In each site, the adoption of the technologies that best fit
local circumstances may be stimulated by putting in place the right
incentives.
Although this paper argues that the retention of crop residues
is possible in most mixed crop-livestock systems of Eastern Africa,
this would not be achieved by any single technology. Modifying
the supply-side of crop residues without altering the demand-side
is unlikely to increase the quantity of crop residues retained in the
fields, particularly in areas where farmers tend to maximize the
number of livestock they keep. The promotion of combinations of
technologies, rather than single component technologies, would be
needed, though this requires a major shift in the way research and
development is being conducted.
Finally, competition for cereal residues between livestock feeding and soil mulching may not be as limiting for the adoption of CA
as previously stated. Indeed, the impact of surface mulching on crop
productivity is equivocal, and is not reflected in the blanket recommendation of a 30% soil cover. Thus, there is a need to understand
the site-specific crop response to mulching, from which recommendations can be formulated. There should be situations where
mulching does not improve crop productivity, or even has a detrimental effect on crop yield. Mulching should not be an end by itself,
and the focus should be on processes (e.g. increased water-use
efficiency), not on technologies.
Acknowledgements
This research was funded by EU-IFAD (European Union–
International Fund for Agricultural Development) through the
project ‘Enhancing Total Farm Productivity in Smallholder Conservation Agriculture Based Systems in Eastern Africa’ and by
ACIAR (Australian Centre for International Agricultural Research)
through the project ‘Sustainable Intensification of Maize-Legume
Cropping systems for Foods in Eastern and Southern Africa’ (SIMLESA). We thank Dr. Bruno Gérard and Dr. Fred Kanampiu for
comments on an earlier version. We are also grateful to Jiska van
Ommen for assistance with field work in the Ethiopian Rift Valley.
References
Andersson, J., Giller, K.E., 2012. On heretics and God’s blanket salesmen: contested
claims for Conservation Agriculture and the politics of its promotion in African
smallholder farming. In: Sumberg, James, Thompson, John (Eds.), Contested
Agronomy: Agricultural Research in a Changing World. Routledge, London.
Baudron, F., Tittonell, P., Corbeels, M., Letourmy, P., Giller, K.E., 2012. Comparative performance of conservation agriculture and current smallholder farming
practices in semi-arid Zimbabwe. Field Crops Res. 132, 117–128.
Bebe, B.O., Udo, H.M.J., Rowlands, G.J., Thorpe, W., 2003. Smallholder dairy systems
in the Kenya highlands: breed preferences and breeding practices. Livest. Prod.
Sci. 82, 117–127.
Blanco-Canqui, H., Lal, R., 2009. Crop residue removal impacts on soil productivity
and environmental quality. Crit. Rev. Plant Sci. 28, 139–163.
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
Burgess, S.S.O., Adams, M.A., Turner, N.C., Ong, C.K., 1998. The redistribution of soil
water by tree root systems. Oecologia 115, 306–311.
Chivenge, P.P., Murwira, H.K., Giller, K.E., Mapfumo, P., Six, J., 2006. Long-term
impact of reduced tillage and residue management on soil carbon stabilization:
implications for conservation agriculture on contrasting soils. Soil Till. Res. 94,
328–337.
Cook, H.F., Valdes, G.S., Lee, H.C., 2006. Mulch effects on rainfall interception, soil
physical characteristics and temperature under Zea mays L. Soil Till. Res. 91,
227–235.
Coppock, C.E., 1985. Energy nutrition and metabolism of the lactating dairy cow. J.
Dairy Sci. 68, 3403–3410.
Ehui, S., Polson, R., 1993. A review of the economic and ecological constraints on
animal draft cultivation in Sub-Saharan Africa. Soil Till. Res. 27, 195–210.
Erenstein, O., 2002. Crop residue mulching in tropical and semi-tropical countries:
an evaluation of residue availability and other technological implications. Soil
Till. Res. 67, 115–133.
Erenstein, O., 2003. Smallholder conservation farming in the tropics and sub-tropics:
a guide to the development and dissemination of mulching with crop residues
and cover crops. Agr. Ecosyst. Environ. 100, 17–37.
Fermont, A., van Asten, P.J.A., Tittonell, P., van Wijk, M.T., Giller, K.E., 2009. Closing
the cassava yield gap: an analysis from smallholder farms in East Africa. Field
Crops Res. 112, 24–36.
Findeling, A., Ruy, S., Scopel, E., 2003. Modeling the effects of a partial residue mulch
on runoff using a physically based approach. J. Hydrol. 275, 49–66.
Friedrich, T., 2008. Sustainable production intensification in Africa–a climate change
perspective. Open science conference on Africa and carbon cycle: the CarboAfrica project, Accra, 25–27 November 2008.
Giller, K.E., Witter, E., Corbeels, M., Tittonell, P., 2009. Conservation agriculture and
smallholder farming in Africa: the heretics’ view. Field Crops Res. 114, 23–34.
Giller, K.E., Corbeels, M., Nyamangara, J., Triomphe, B., Affholder, F., Scopel, E., Tittonell, P., 2011. A research agenda to explore the role of conservation agriculture
in African smallholder farming systems. Field Crops Res. 124, 468–472.
Grimble, R., Wellard, K., 1997. Stakeholder methodologies in natural resource management: a review of principles, contexts, experiences and opportunities. Agric.
Syst. 55, 173–193.
Groot, J.J.R., 2009. Update of fertiliser supply and demand - sub Saharan Africa, IFA
Africa Forum, IDFC, Cairo, Egypt.
Haggblade, S., Tembo, G., 2003. Conservation farming in Zambia. International Food
Policy Research Institute, Washington, DC.
Hassanali, A., Herren, H., Khan, Z.R., Pickett, J.A., Woodcock, C.M., 2008. Integrated
pest management: the push-pull approach for controlling insect pests and
weeds of cereals, and its potential for other agricultural systems including animal husbandry. Philos. T. Roy. Soc. B 363, 611–621.
Herrero, M., Thornton, P.K., Notenbaert, A.M., Wood, S., Msangi, S., Freeman, H.A.,
Bossio, D., Dixon, J., Peters, M., van de Steeg, J., Lynam, J., Parthasarathy, R.,
Macmillan, S., Gerard, B., McDermott, J., Seré, C., Rosegrant, M., 2010. Smart
investments in sustainable food production: revisiting mixed crop-livestock
systems. Science 327, 822–825.
Janssen, B.H., Guiking, F.C.T., van der Eijk, D., Smaling, E.M.A., Wolf, J., van Euler,
H., 1990. A system for quantitative evaluation of the fertility of tropical soils
(QUEFTS). Geoderma 46, 299–318.
Kassam, A., Friedrich, T., Shaxson, F., Pretty, J., 2009. The spread of Conservation
Agriculture: justification, sustainability and uptake. Int. J. Agric. Sustain. 7,
292–320.
Keftasa, D., 1987. Role of crop residues as livestock feed in Ethiopian highlands. In
African Forage Genetic Resources, Evaluation of Forage Germplasm and Extensive Livestock Production Systems. In: Proceedings of a Workshop Held in
Arusha, Tanzania, pp. 430–439.
Knowler, D., Bradshaw, B., 2007. Farmers’ adoption of conservation agriculture: a
review and synthesis of recent research. Food Policy 32, 25–48.
Kozak, J.A., Ahuja, L.R., Green, T.R., Ma, L., 2007. Modelling crop canopy and residue
rainfall interception effects on soil hydrological components for semi-arid agriculture. Hydrol. Processes 21, 229–241.
Lal, R., 1998. Soil erosion impact on agronomic productivity and environment quality. Crit. Rev. Plant Sci. 17, 319–464.
Lal, R., 2005. World crop residues production and implications of its use as a biofuel.
Environ. Int. 31, 575–584.
Lal, R., 2010. Enhancing eco-efficiency in agro-ecosystems through soil carbon
sequestration. Crop Sci. 50, S120–S131.
Larbis, A., Smith, J.W., Adekunles, I.O., Agyare, W.A., Gbaraneh, L.D., Tanko, R.J., Akinlades, J., Omokaye, A.T., Karbo, N., Aboh, A., 2002. Crop residues for mulch and
feed in crop-livestock systems: impact on maize grain yield and soil properties in the West African humid forest and savanna zones. Exp. Agr. 38,
253–264.
Le Hourou, H.N., Hoste, C.H., 1977. Rangeland production and annual rainfall relations in the Mediterranean basin and in the African Sahelo-Sudanian zone. J.
Range Manage. 30, 181–189.
Lenné, J.M., Thomas, D., 2005. Integrating crop-livestock research and development in Sub-Saharan Afric. Option, imperative or impossible? Outlook Agr. 35,
167–175.
Luo, Z., Wang, E., Sun, O.J., 2010. Can no-tillage stimulate carbon sequestration in
agricultural soils? A meta-analysis of paired experiments. Agr. Ecosyst. Environ.
139, 224–231.
Machila, N., Wanyangu, S.W., McDermott, J., Welburn, S.C., Maudlin, I., Eisler, M.C.,
2003. Cattle owners’ perceptions of African Bovine trypanosomiasis and its control in Busia and Kwale Districts of Kenya. Acta Trop. 86, 25–34.
181
McDermott, J.J., Staal, S.J., Freeman, H.A., Herrero, M., Van de Steeg, J.A., 2010. Livest.
Sci. 10, 95–109.
Moll, H.A.J., Staal, S.J., Ibrahim, M.N.M., 2007. Smallholder dairy production and markets: a comparison of production systems in Zambia, Kenya and Sri Lanka. Agric.
Syst. 94, 593–603.
Mupangwa, W., Twomlow, S., Walker, S., 2008. The influence of conservation tillage
methods on soil water regimes in semi-arid southern Zimbabwe. Phys. Chem.
Earth 33, 762–767.
Nijhof, K. (Ed.), 1987. Staff working paper SOW-87-08. Centre for World Food Studies, Amsterdam-Wageningen, p. 52.
Olsen, J., Kristensen, L., Weiner, J., Griepentrog, H.W., 2005. Increased density and
spatial uniformity increase weed suppression by spring wheat. Weed Res. 45,
316–321.
Ong, C.K., Black, C.R., Wallace, J.S., Khan, A.A.H., Lott, J.E., Jackson, N.A., Howard,
S.B., Smith, D.M., 2000. Productivity, microclimate and water use in Grevillea robusta-based agroforestry systems on hillslopes in semi-arid Kenya. Agr.
Ecosyst. Environ. 80, 121–141.
Palm, C., Gachengo, C.N., Delve, R.J., Cadish, G., Giller, K.E., 2001. Organic inputs for
soil fertility management in tropical agroecosystems: application of an organic
resource database. Agr. Ecosyst. Environ. 83, 27–42.
Powell, J.M., Pearson, R.A., Hiernaux, P.H., 2004. Crop-livestock interactions in the
West African drylands? Agron. J. 96, 469–483.
Rockström, J., Kaumbutho, P., Mwalley, J., Nzabi, A.W., Temesgen, M., Mawenya, L.,
Barron, J., Mutua, J., 2009. Damgaard-Larsen, SConservation farming strategies
in East and Southern Africa: yields and rain water productivity from on-farm
action research. Soil Till. Res. 103, 23–32.
Rodier, A., Assié, J., Marchand, J.L., Hervé, Y., 1995. Breeding maize lines for complete
and partial resistance to maize streak virus (MSV). Euphytica 81, 57–70.
Romney, D., Utiger, C., Kaitho, R., Thorne, P., Wokabi, A., Njoroge, L., Kirui, J., Kamotho,
D., Staal, S., 2004. Effect of intensification on feed management of dairy cows in
the central highlands of Kenya. In: Owen, E., Smith, T., Steele, M., Anderson,
S., Duncan, A., Herrero, M. (Eds.), Responding to the Livestock Revolution: The
Role of Globalization for Poverty Alleviation. Nottingham University Press, pp.
167–177.
Rufino, M.C., Rowe, E.C., Delve, R.J., Giller, K.E., 2006. Nitrogen cycling efficiencies
through resource-poor African crop-livestock systems. Agr. Ecosyst. Environ.
112, 261–282.
Rusinamhodzi, L., Corbeels, M., van Wijk, M.T., Rufino, M.C., Nyamangara, J.,
Giller, K.E., 2011. A meta-analysis of long-term effects of conservation agriculture on maize grain yield under rain-fed conditions. Agron. Sustain. Dev. 31,
657–673.
Schiere, J.B., Ibrahim, M.N.M., van Keulen, H., 2002. The role of livestock for sustainability in mixed farming: criteria and scenario studies under varying resource
allocation. Agr. Ecosyst. Environ. 90, 139–153.
Scopel, E., Macena, F., Corbeels, M., Affholder, F., Maraux, F., 2004. Modelling crop
residue mulching effects on water use and production of maize under semi-arid
and humid conditions. Agronomie 24, 1–13.
Stewart, C.E., Paustian, K., Conant, R.T., Plante, A.F., Six, J., 2007. Soil carbon saturation: concept, evidence and evaluation. Biogeochemistry 86, 19–31.
Swanton, C.J., Weise, S.F., 1991. Integrated weed management: the rationale and
approach. Weed Technol. 5, 657–663.
Teasdale, J.R., 1996. Contribution of cover crops to weed management in sustainable
agricultural systems. J. Prod. Agric. 9, 475–479.
Temesgen, M., Savenije, H.H.G., Rockström, J., Hoogmoed, W.B., 2012. Assessment
of strip tillage systems for maize production in semi-arid Ethiopia: effects on
grain yield, water balance and water productivity. Phys. Chem. Earth, 156–165,
47-48.
Thierfelder, C., Wall, P.C., 2009. Effects of conservation agriculture techniques on
infiltration and soil water content in Zambia and Zimbabwe. Soil Till. Res. 105,
217–227.
Thornton, P.K., 2010. Livestock production: recent trends, future prospects. Philos.
T. Roy. Soc. B 365, 2853–2867.
Tittonell, P., Vanlauwe, B., Leffelaar, P.A., Shepherd, K.D., Giller, K.E., 2005. Exploring
diversity in soil fertility management of smallholder farms in western Kenya:
II Within-farm variability in resource allocation, nutrient flows and soil fertility
status. Agr. Ecosyst. Environ. 110, 166–184.
Tittonell, P., Scopel, E., Andrieu, N., Posthumus, H., Mapfumo, P., Corbeels, M., van
Halsema, G.E., Lahmar, R., Lugandu, S., Rakotoarisoa, J., Mtambanengwe, F.,
Pound, B., Chikowo, R., Naudin, K., Triomphe, B., Mkomwa, S., 2012. Agroecologybased aggradation-conservation agriculture (ABACO): targeting innovations to
combat soil degradation and food insecurity in semi-arid Africa. Field Crops Res.
132, 168–174.
Tittonell, P., Giller, K.E., 2012. When yield gaps are poverty traps: The paradigm
of ecological intensification in African smallholder agriculture. Field Crops Res.
143, 76–90.
Unger, P.W., Vigil, M.F., 1998. Cover crop effects on soil water relationships. J. Soil
Water Conserv. 53, 200–207.
Valbuena, D., Erenstein, O., Homann-Kee Tui, S., Abdoulaye, T., Claessens, L., Duncan, A.J., Gérard, B., Rufino, M.C., Teufel, N., van Rooyen, A., van Wijk, M.T.,
2012. Conservation Agriculture in mixed crop–livestock systems: scoping crop
residue trade-offs in Sub-Saharan Africa and South Asia. Field Crops Res. 132,
175–184.
Vandehaar, M.J., 1998. Efficiency of nutrient use and relationship to profitability on
dairy farms. J. Dairy Sci. 81, 272–282.
Vandehaar, M.J., St-Pierre, N., 2006. Major advances in nutrition: Relevance to the
sustainability of the dairy industry. J. Dairy Sci. 89, 1280–1291.
182
F. Baudron et al. / Agriculture, Ecosystems and Environment 187 (2014) 171–182
Vanlauwe, B., Giller, K.E., 2006. Popular myths around soil fertility management is
sub-Saharan Africa. Agr. Ecosyst. Environ. 116, 34–46.
Vanlauwe, B., Bationo, A., Chianu, J., Giller, K.E., Merckx, R., Mokwunye, U., Ohiokpehai, O., Pypers, P., Tabo, R., Shepherd, K.D., Smaling, E.M.A., Woomer, P.L.,
Sanginga, N., 2010. Integrated soil fertility management. Operational definition and consequences for implementation and dissemination. Outlook Agr. 39,
17–24.
Versteeg, M.N., Maldonado, D., 1978. Increased profitability using low doses of herbicide with supplementary weeding in smallholdings. Int. J. Pest Manage. 24,
327–332.
Vitousek, P.M., Naylor, R., Crews, T., David, M.B., Drinkwater, L.E., Holland, E., Johnes,
P.J., Katzenberger, J., Martinelli, L.A., Matson, P.A., Nziguheba, G., Ojima, D., Palm,
C.A., Robertson, G.P., Sanchez, P.A., Townsend, A.R., Zhang, F.S., 2009. Nutrient
imbalances in agricultural development. Science 324, 1519–1520.
Vohland, K., Barry, B., 2009. A review of in situ rainwater harvesting (RWH) practices
modifying landscape functions in African drylands. Agr. Ecosyst. Environ. 131,
119–127.
Wondatir, Z., 2010. Livestock Production Systems in Relation to Feed Availability
in the Highlands and Central Rift Valley of Ethiopia. MSc Dissertation. Ethiopia,
Haramaya University.
Zibilske, L.M., Bradford, J.M., Smart, J.R., 2002. Conservation tillage induced changes
in organic carbon, total nitrogen and available phosphorus in a semi-arid alkaline
subtropical soil. Soil Till. Res. 66, 153–163.