HAL Id: hal-00886375
https://hal.archives-ouvertes.fr/hal-00886375
Submitted on 1 Jan 2007
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
sub-Saharan Africa
D.D. Onduru, C.C. Du Preez
To cite this version:
D.D. Onduru, C.C. Du Preez. Ecological and agro-economic study of small farms in sub-Saharan
Africa. Agronomy for Sustainable Development, Springer Verlag/EDP Sciences/INRA, 2007, 27 (3),
pp.197-208. �hal-00886375�
Agron. Sustain. Dev. 27 (2007) 197–208 Available online at:
INRA, EDP Sciences, 2007c www.agronomy-journal.org
DOI: 10.1051/agro:2007003
Original article
Ecological and agro-economic study of small farms in sub-Saharan Africa
D.D. O
a*, C.C. D
P
baETC-East Africa, PO Box 76378-00508 Yaya, Nairobi, Kenya
bDepartment of Soil, Crop and Climate Sciences, University of the Free State, PO Box 339, Bloemfontein 9300, South Africa
(Accepted 12 December 2006)
Abstract – Land degradation, rising population and poverty in sub-Saharan Africa threatens the agricultural sustainability and productivity, quality of the environment and socio-economic wellbeing of rural populations. We studied farm ecological, economic and social sustainability, productivity and production risks in the Mbeere District of Eastern Kenya. We used a soil nutrient monitoring methodology to collect data from 30 households. Ecological sustainability was threatened by soil nutrient decline at rates of 1.7 kg P and 5.4 kg K ha−1half year−1while N was nearly balanced in soils. Soil phosphorus and potassium stocks, in the cultivated soils, declined at rates of 0.3% and 0.1% half year−1, respectively. Farm economic returns were positive, albeit low, and could not sustain the livelihoods of the households. All the 30 households were living below the poverty line of 1 US dollar a day. Farm productivity was low, with livestock and yields of major staple food crops below on-farm target yields. To spread out the risks of production, farming households were cultivating an average of 4.7 crop fields, keeping more than two types of livestock and practising intercropping systems. Intercropping maize-beans reduced nutrient decline and raised household incomes compared with monocropping of either of the two crops. Despite the low rates of nutrient decline, high risks of production and the low crop yields, the livestock productivity and farm economic performance put the sustainability of these farming systems into question. The low levels of nutrient decline in small farms averaging at 1.7 kg P and 5.4 kg K ha−1 half year−1 contrasts with the high nutrient depletion rates on macro-scale levels, e.g. 20–40 N, 3.5–6.6 kg P and 20–40 kg K ha−1year−1for Eastern African countries and 22 kg N, 2.5 kg P and 15 kg K ha−1year−1 for sub-Saharan Africa. These findings indicate that the extent of nutrient decline and conservation differs across sub- Saharan Africa. The positive contribution of intercropping to nutrient balances suggests the need to encourage farmers to adopt such systems rather than monocropping.
agro-ecological sustainability/ drylands / production risks / socio-economic sustainability / sustainable agriculture
1. INTRODUCTION
Human activity in the drylands of sub-Saharan Africa is characterised by poverty and malnutrition; food insecurity;
high population growth and environmental degradation; va- garies of climate; poor infrastructure; neglect in national re- search and development priorities; high rates of unemploy- ment; changing food habits and dynamic changes in both food demand and production patterns (Ryan and Spencer, 2001). In these areas, the balance between natural resources and agri- cultural production is precariously threatened and there is a growing concern about long-term sustainability of agriculture and food production. This is partly because of the low and de- clining agricultural production, which is intrinsically linked to land degradation and poverty (Cleaver and Schreiber, 1994).
Land degradation, in the form of nutrient decline, soil erosion, soil compaction, waterlogging and surface crusting, among others, is partly responsible for declining food pro- duction in sub-Saharan Africa. Seminal studies covering 37 sub-Saharan Africa countries (excluding South Africa) indi- cate that an average of 660 kg N ha−1, 75 kg P ha−1, and
* Corresponding author: d.onduru@etc-eastafrica.org;
ddonduru@yahoo.com
450 kg K ha−1has been depleted during the last 30 years from about 200 million hectares of cultivated land. The same study reported nutrient decline of 42 kg N ha−1, 3 kg P ha−1 and 29 kg K ha−1 per year for Kenya (Stoorvogel and Smaling, 1990). Because agriculture in the region is a soil-based indus- try that extracts nutrients from the soil, effective and efficient approaches to slowing nutrient decline and returning nutrients to the soil are required to maintain and increase crop produc- tivity and sustain agriculture in the long term (Gruhn et al., 2000).
Reversing nutrient decline and agricultural stagnation in sub-Saharan Africa remains a challenge because not only eco- logical but also social, economic and policy issues are in- volved. The recognition of these issues and their integration into a holistic agricultural development paradigm is the ba- sis for striving for agricultural sustainability. The five pil- lars of sustainable agriculture and land use are inter alia:
(i) maintenance and enhancement of production and services (productivity); (ii) reduction of production risks (security);
(iii) protection of the production potential and capacity of nat- ural resources and preventing the degradation of soil and wa- ter quality and biological diversity (protection); (iv) economic viability (viability); and (v) social acceptance (acceptability) (Smyth and Dumanski, 1995).
Article published by EDP Sciences and available at http://www.agronomy-journal.org or http://dx.doi.org/10.1051/agro:2007003
Development and assessing sustainability of agricultural practices require the use of measurable indicators that diag- nose agro-ecosystem functions over a given time frame on a defined scale of study (Syers and Rimmer, 1995). Agri- cultural sustainability indicators are variables for measuring change and for diagnosing the underlying status and condi- tion of the agricultural system. In this study, agro-ecological and socio-economic dimensions of the dryland farming sys- tem in Mbeere District, Eastern Kenya were assessed using farm-level indicators to gauge the direction of sustainability.
The objectives of the study were to investigate farm ecological sustainability; farm economic and social sustainability; and to determine farm productivity and risks of production.
2. MATERIALS AND METHODS
2.1. Study area and selection of farm households The study was conducted in the Mbeere District, Eastern Kenya in the year 2002. Mbeere District (Latitude 0◦ 20’ and 0◦ 50’ South and Longitude 37◦ 16’ and 37◦ 56’ East) cov- ers an area of 209 700 hectares and has a human population of 170 953 persons living in 37 164 households (CBS, 2000).
This results in a population density of 82 persons km−2with an average of 4.6 persons per household. The district has an altitude range of 500–1200 metres above sea level and a mean annual temperature range of 20◦C to 30◦C depending on al- titude. Rainfall in the district is bimodal, unpredictable and unreliable. The annual average rainfall in the study site is 800–
1100 mm. Total rainfall received during the study period was 903 mm. The district has two growing periods with a total length of 90–119 days (Kassam et al., 1991).
Land for faming is held under freehold tenure. Farming takes place mainly under rain-fed conditions. Soils are well drained, shallow to deep, yellowish brown, loamy sand to sandy loam, Luvic Arenosols (Muya, 2003). They are strongly acid to slightly acid and low in organic C, total N and ex- tractable P (Tab. I). Farmers grow maize, beans, cowpeas, sorghum, sweet potatoes and cassava for subsistence and raise indigenous breeds of cattle, goats and poultry. The main prob- lems in the district include declining soil fertility, decreasing arable land per capita, unpredictable and unreliable rainfall, unproductive livestock and limited use of agricultural inputs.
Thirty farm households were selected from a representa- tive catchment in the district following a community meeting.
Farms were selected in the same catchment to minimise vari- ability. Selection criteria comprising biophysical and socio- economic factors typical of smallholder farmers in the district were used to include farms in the study: land size (0.2–5 ha), mechanisation (low), soil conservation status (arable land ter- raced), and market orientation (low), use of external inputs (none to less than 50% of cultivated land under inorganic fer- tilisers), and irrigation (rain-fed farming only). Other criteria used include willingness to participate in the study, willing- ness to share information with others, gender and farming as the main activity. In the community meeting held in the catch- ment, the farmers who met these criteria were requested to
participate in the study on behalf of the community. Data on household socio-economic characteristics were collected us- ing a semi-structured questionnaire, while the values of pro- ductive assets were estimated using local market prices and opportunity costs (Tab. I). The farmers were stratified accord- ing to livestock endowment (tropical livestock units) since livestock is considered the main household asset, important for tillage, manure production and other socio-economic and cultural roles.
2.2. Indicators for sustainability assessment
Selected indicators are presented in Table II. A nutrient balance as an indicator of ecological sustainability is a mea- surement of physical difference (surplus/deficit) between nu- trient inputs into, and outputs from, an agricultural system (van den Bosch et al., 1998). Nutrient balance establishes link- ages between agricultural nutrient use, changes in environ- mental quality and sustainable use of soil resources. A per- sistent deficit (nutrient decline) indicates potential agricultural sustainability problems, while surplus shows potential envi- ronmental pollution depending on local farm conditions, nutri- ent management practices, soil types and agro-ecological con- ditions (OECD, 2001).
Livestock and crop yields (kg ha−1) were selected as pro- ductivity indicators. Yield levels are influenced by farmers’
biophysical and socio-economic conditions as well as man- agement decisions and production practices. The latter may result in sub-optimal use of land resources or even a deterio- rating soil resource base. Thus, crop yield levels are partly a reflection of the quality of the soil resource base.
Eleven economic indicators for tracking farm financial per- formance and profitability were selected. Farmers need to achieve a balance over time between the cost of capital and profits realised from agriculture’s use of natural resources and the environment. This balance provides a link between the environment, economic and social dimensions of sustainable agriculture. The timing, certainty and level of financial re- source flows affect farmers’ ability and actions with respect to the type, level and intensity of input use as well as level of production and acquisition of new technologies.
Labour demand for crop and livestock production, and poverty levels, were used as social indicators of sustainability.
For the latter, the Absolute Poverty Line drawn by the Ministry of Finance and Planning (CBS, 2000) and the poverty line pro- posed by the World Bank (World Bank, 2000, 2002) were used as threshold values in the analysis of poverty levels.
Farming involves risks. The following indicators were se- lected to assess production risks: crop diversity (number of plots sown to different crops or crop combinations, and num- ber of cultivated plots put under mono- and intercropping for major crops) and livestock diversity (number of livestock groups per farm and livestock species).
Ecological and agro-economic study of small farms in sub-Saharan Africa 199
Table I. Characteristics of the 30 farm households studied (standard deviation in parenthesis).
Characteristic Household resource characteristics
Total (n= 30) TLU< 0.34
(n= 10)
0.34< TLU < 0.72 (n= 10)
TLU> 0.72 (n= 10) Labour
Consumer units (aeu)a 4.9 (1.5) 4.5 (1.6) 4.8 (1.4) 4.7 (1.5)
Labour units (aeu) 3.3 (1.2) 2.8 (1.4) 3.4 (1.3) 3.2 (1.3)
Household heads (absolute values)
Non-educated 2 0 0 2
Primary/elementary education 7 9 10 26
Post-primary education 0 1 0 1
Secondary education 1 0 0 1
Land characteristics
Total farm area (ha) 1.7 (1.8) 1.3 (0.9) 1.3 (0.6) 1.4 (1.2)
Cultivated area (ha) 1.4 (1.2) 1.1 (0.5) 1.2 (0.6) 1.2 (0.8)
Fallow area (ha) 0.3 (0.7) 0.2 (0.6) 0.1 (0.1) 0.2 (0.5)
Average slope (%) 19.5 18.0 18.0 18.5
Capitalb
TLUc 0.2 (0.1) 0.5 (0.1) 2.4 (2.4) 1.1 (1.7)
TLU person−1 0.04 (0.03) 0.1 (0.04) 0.4 (0.5) 0.2 (0.4)
Ha TLU−1 4.4 (2.2) 2.5 (1.9) 0.9 (0.6) 2.5 (2.1)
Value of livestock (US $) 69.0 (31) 158.0 (39) 391.0 (273) 206.0 (207)
Value of land (US $) 2503.0 (3466) 2184.0 (1486) 2219.0 (1163) 2302.0 (2203)
Value of equipment (US $) 55.0 (67) 35.0 (24) 61.0 (77) 50.0 (59)
Ratios
Land:Labour (ha aeu−1) 0.5 (0.4) 0.5 (0.3) 0.4 (0.2) 0.5 (0.3)
Land:Consumer (ha aeu−1)d 0.4 (0.4) 0.3 (0.2) 0.3 (0.1) 0.3 (0.3)
Consumer:Labour (aeu aeu−1) 1.6 (0.3) 1.7 (0.3) 1.4 (0.3) 1.6 (0.3)
Soilsd
pH-H20 (1:1.2.5 suspension) 5.6 (5.2, 6.0) 5.7 (5.0, 6.3) 5.4 (4.8, 5.8) 5.6 (4.8, 6.3)
Total N (g kg−1) 0.6 (0.3, 0.9) 0.6 (0.4, 0.9) 0.6 (0.4, 0.9) 0.6 (0.3, 0.9)
Organic C (g kg−1) 5.7 (2.9, 9.4) 6.5 (4.0, 11.1) 6.2 (4.2, 8.0) 6.1 (2.9, 11.1)
Extractable P (mg kg−1) 7.3 (5, 10) 9.6 (2, 20) 9.4 (2, 20) 8.8 (2.0, 20)
Exchangeable K (cmol kg−1) 0.5 (0.2, 1.0) 0.4 (0.2, 0.6) 0.4 (0.2, 0.5) 0.4 (0.2, 1.0)
aaeu: Adult equivalent units;b1 US $= 75 Ksh at time of study;cTropical Livestock Units (1 TLU= 250 kg live weight);dminimum and maximum in parenthesis for soil characteristics.
2.3. Quantification of sustainability indicators and NUTMON conceptual model
The nutrient monitoring decision support tool, NUTMON, was used to quantify the selected indicators. NUTMON is an integrated, multi-disciplinary and multi-scale approach used for calculating nutrient (nitrogen, phosphorus and potassium) flows, soil nutrient stocks, nutrient balances and economic per- formance indicators on different scale levels (e.g. plot, farm, district, national, etc.). The NUTMON toolbox consists of a set of questionnaires and computer software for data entry and processing (Vlaming et al., 2001a, b).
NUTMON operates on a simplified framework that concep- tualises a farm. The boundaries of such a farm coincide with
its physical borders (the system is bounded by its physical bor- ders) while livestock grazing outside these borders are still considered part of it. The upper boundary is the atmosphere- soil or atmosphere-plant interface, whereas the lower bound- ary is defined as the depth below which leached nutrients are assumed to be lost from the system. The physical environment beyond the farm borders is considered the “external environ- ment”. The external environment is important for farming ac- tivities (e.g. markets), but not monitored in detail.
The NUTMON model further conceptualises a farm to have four major components: farm section units (static farm units);
nutrient pools (dynamic farm units); entries other than nutri- ent pools, which influence farm management, and nutrient and
Table II. Selected indicators and their relevance to sustainability.
Pillar of sustainability Selected indicator (s)
Protection of production resources (soils; farm-scale study)
(a) Soil nutrient status: total N (g kg−1), extractable P (mg kg−1), total P (kg ha−1), exchangeable K (cmol kg−1) and total K (kg ha−1)
(b) Partial and total soil nutrient balances for N, P and K (kg ha−1) (c) Rate of nutrient stock depletion (%) per unit time
Economic viability and social accep- tance
(farm-scale study)
(a) Economic viability
– Net farm income (US $a; US $ ha−1) – Gross margin PPUsb(US $ ha−1)b – Gross margin SPUsc(US $ TLU−1) – Gross margins RUsd(US $) – Off-farm income (US $; US $ day−1)
– Family earnings (US $; US $ consumer unit−1) – Farm net cash flow (US $)
– Household net cash flow (US $) – Market share (% of produce sold) – Off-farm share in total family earnings (%) (b) Social indicators
Poverty
– Poverty level in relation to family earnings Labour
– General family labour in farm practices (days) – Cropping family labour (days)
– Livestock family labour (days) – Off-farm family labour (days) – Hired labour on PPUs (days) – Hired labour on SPUs (days) – Hired labour on RUs (days)
– Total labour for cropping activities (days) – Total labour for livestock activities (days) – Labour intensity for crops (days ha−1) – Labour intensity livestock (days TLU−1)e – Return to labour (US $ day−1)
Productivity and production risks (agro- economic performance of crops and livestock on activity/plot scale)
Crops
– Number of cultivated plots/fields per farm
– Yields, nutrient balances and economics of intercrops versus monocrops Livestock
– Livestock diversity (Number of SPUs, livestock types) – Livestock productivity
– Economic performance of livestock (gross margins and net cash flow)
a1 US $= 75 Ksh at time of study;bPrimary Production Unit (PPU) – crop activities;cSecondary Production Unit (SPU) – livestock activities;d Redistribution Unit (RU) – nutrient storage activities;eTropical Livestock Units (1 TLU= 250 kg live weight).
economic flows; and nutrient and cash flows. Usually a farm is divided into two or more farm sectional units, with each sec- tion having homogenous soil properties (assumed), slope, and flooding regime and land tenure. Crops growing within a given farm sectional unit acquire the soil and land characteristics of that farm sectional unit.
The six nutrient pools comprise primary production units (PPUs = cropping activities); secondary production units (SPUs= livestock activities – group of animals of the same species within the farm which are managed by the farmer as one unit); redistribution units (RUs= nutrient storage and
redistribution points); stock (STOCK = staple foods, crop residues and chemical fertilisers temporarily stored for later use); household (HH= a group of people who live in the same house or group of houses who share food regularly from the same “cooking pot”); and the outside/external farm environ- ment (EXT= markets and other families and neighbours who are the sources and destination of nutrient and cash flows).
Non-nutrient pool components are soils, climatic factors and markets. Climatic factors include monthly precipitation (used in a leaching transfer function) and rainfall erosivity (pa- rameter in Universal Soil Loss Equation, USLE). The market
Ecological and agro-economic study of small farms in sub-Saharan Africa 201
is required for determining farm gate prices for calculations of economic indicators. Soil properties are used in pedotransfer functions to calculate leaching, gaseous losses and erosion.
Three categories of flows for calculating nutrient balances are distinguished: inflows (6 flows); internal flows; and out- flows (6 outflows). Flows into the farm (inflows) originate from outside (EXT) the farm and their destination is one of the nutrient pools within the farm (IN 1–6). They are in the form of inorganic fertilisers and feeds (IN 1), imported or- ganic fertilisers/manures (IN 2a), and manure from external grazing (IN 2b), wet and dry deposition from the atmosphere (IN 3), symbiotic (IN 4a) and non-symbiotic biological nitro- gen fixation (IN 4b), irrigation and flooding or sedimentation (IN 5) and sub-soil exploitation (IN 6). Flows out of the farm (outflows) are flows from one of the nutrient pools to a des- tination outside (EXT) the farm (OUT 1–6). They are in the form of harvested products (OUT 1), exported crop residues and manure (OUT 2a) and excretion of manure outside the farm (OUT 2b), leaching from soils (OUT 3a) and redistribu- tion units (OUT 3b), gaseous losses from soil (OUT 4a) and redistribution units (OUT 4b), erosion (OUT 5), and lost hu- man excreta (OUT 6). Internal flows are flows from one nu- trient pool to another (HH, PPU, SPU, RU, STOCK↔ HH, PPU, SPU, RU, STOCK).
In addition to nutrient flows, product flows and economic flows are also considered. Product flows (physical flows of in- puts and outputs, e.g. maize grains) are converted into nutri- ent flows by multiplying their quantities with respective nu- trient contents. They are also converted into economic flows by multiplying their quantities by farm gate prices. At the same time, there are flows which are purely of an economic nature, e.g. off-farm income. Flows used in economic calcu- lations are those that are “visible” to the farmer or “easy-to- quantify flows”: IN 1, IN 2, OUT 1, and OUT 2.
In the NUTMON model, nutrient flows are quantified using four methods: (i) asking the farmer; (ii) using pedo-transfer functions; (iii) using sub-models, e.g. a livestock model; and (iv) assumptions. The calculation rules for nutrient flows and balances and economic performance used in NUTMON have been described by Vlaming et al. (2001a, b) and are available at http://www.nutmon.org. NUTMON calculates the nutrient balance of a unit (Farm, PPU, RU, etc.) by subtracting the sum of all flows out of a unit from the sum of all flows into a unit.
The benefit of this approach is that either a full or partial nu- trient balance can be calculated for any unit:
Full nutrient balance of a unit= Σ (IN 1 + IN 2 + IN 3 + IN 4+ IN 5) – Σ (OUT 1 + OUT 2 + OUT 3 + OUT 4 + OUT 5+ OUT 6).
Partial nutrient balance of a unit = Σ (IN 1 + IN 2) – Σ (OUT 1+ OUT 2 + OUT 6).
2.4. Data collection, processing and analysis
Soil sampling was carried out on each of the 30 selected farms. The soil samples were analysed for particle size dis- tribution (texture), pH, organic C, total N, total P, extractable P and exchangeable K. Analysis of particle size distribution
was done using a hydrometer method (Hinga et al., 1980).
pH was determined with a conventional glass electrode me- ter in a 1: 2.5 soil to water suspension (Hinga et al., 1980).
Organic C was oxidised using concentrated sulphuric acid and potassium dichromate followed by colorimetric determi- nations (Anderson and Ingram, 1993). Total N and P were determined by wet digestion followed by colorimetric mea- suring methods (Novosamsky et al., 1983). Extractable P was determined colorimetrically after extraction with Mehlich I so- lution (Mehlich et al., 1962). Exchangeable K was extracted with ammonium acetate and then determined with flame pho- tometry (Hinga et al., 1980; Okalebo et al., 2002). Only pri- mary data on clay (%), organic C (%), total N (%), total P (%) and exchangeable K (cmol/kg) were used in calculations with NUTMON. Secondary data also gathered for this purpose were rooting depth (m), N mineralisation rate (% per year), bulk density (kg m−3), erodibility (K factor in USLE equation) and nutrient enrichment factor.
Nutrient and economic flow data were collected by admin- istering the NUTMON inventory and monitoring question- naires to heads of households through one-time recall semi- structured interviews (Tab. III). The inventory questionnaire was administered at the beginning of the agricultural season to capture data expected to remain relatively unchanged during the study period. However, the farm monitoring questionnaire was administered at the end of the agricultural season to cap- ture actual farm management practices, including dynamics of nutrient and economic flows.
Farmers’ local units were calibrated into metric units through sampling and weighing. Additionally, local market prices of agricultural and livestock inputs and outputs were collected to authenticate prices of agricultural products and livestock obtained at farm level. Where it was necessary, lim- ited sampling and analysis of farm inputs and products were carried out to establish their nutrient contents to refine the data that come with NUTMON. A literature review was also conducted to collect local data needed for refining “hard-to- quantify” flow calculations (IN 3, IN 4, IN 5, OUT 3, OUT 4, OUT 5, and OUT 6). Climatic data (rainfall) was collected for the period of study from the local weather station in the catch- ment.
The collected data were triangulated, verified, edited, en- tered and processed using NUTMON computer software. The validity and consistency of data were checked through debug- ging options in the software. Processed data were then ex- ported and further analysed using a special program for social scientists, SPSS (SPSS Inc., 2002).
3. RESULTS AND DISCUSSION 3.1. Farm ecological sustainability
The mean values of N, P and K flows and balances for the long rainy season in 2002, according to livestock categories, in the 30 studied farms are presented in Table IV. The results showed that partial N balances were negative. However, the full N balance was slightly positive due to N imports through
Table III. Data collected using NUTMON inventory and monitoring questionnaires.
Data group collected
Details of data/information collected NUTMON inventory data
General farm data Farm location, respondent and distance of the farm from major market Demographic structure of the
Household
Household members, their sex, age, main occupation, education level and percentage of time they spend in the household
Implements/agriculture-related Machinery
Implements, agricultural constructions on the farm and their value
Farm section units Information on sections of the farm with homogenous characteristics in terms of slope, soil properties, flooding regime and land tenure
Primary production units Various crops and fields within the farm sown to different cropping activities
Crop calendar Presence of a primary production unit every monitoring month
Secondary production units Various animal species/breeds within the farm
Redistribution units Nutrient accumulation and distribution points within the farm Farm sketch/primary production
unit sketch
Sketch map of the farm depicting locations of farm sectional units, cropping activities and farm boundary
Precipitation data Monthly rainfall during monitoring period; and annual average rainfall
Measurements and other data Areas of farm sectional units, primary production units, slope, slope length and other data requirements for Universal Soil Loss Equation
Soil sampling and analysis data Information on soil attributes NUTMON monitoring data
Inputs into primary production units
Sources, quantities and prices of fertilisers, seeds, manure, pesticides, labour, etc. used in crop production
Outputs from primary production units Quantities, destination and prices of harvested products and crop residues
Herd growth Number of animals in the farm, number of animals born, purchased, consumed, died or given as gifts
Inputs in secondary production units
Sources, quantities and prices of fodder, concentrates, labour, etc. used in livestock production
Outputs from secondary production units Sources, quantities, destination and prices of milk, eggs, hides, skins, traction and other livestock products
Livestock confinement Number of days for which a given secondary production unit has been confined to primary production units (fields, pastures, fallows and homestead), redistribution units (kraal) and outside the farm (farm external environment)
Inputs and outputs from redistribution units
Use of external inputs into redistribution units and quantity and destination of re- used manure, compost, garbage and household waste. It also identifies destination of human excreta
Inputs and outputs from stock Sources, quantities and prices of staple foods (grains and legumes) as well as stover that goes into and out of Stock. Home consumption from stock is calculated directly from the NUTMON model
Off-farm income Household members’ engagement in off-farm income and money earned.
Family labour Number of days spent on crops, livestock, general farm activities and off-farm activi- ties for each person in the household
biological fixation. The full as well as the partial P balances were negative, implying that there was P decline. However, the rate of decline of 0.1–2.1 kg P ha half year−1was low to mod- erate according to the classification (<1.7 kg ha−1year−1low;
1.7 to 3.5 kg ha−1 year−1moderate) developed by Stoorvogel and Smaling (1990). Major loss pathways for P were ero- sion and human excreta. In comparison with N, the situation was reversed for K, where imports through grazing partly ac- counted for positive partial balances. However, losses due to leaching, erosion and human excreta made the full K balances
negative. Soil erosion accounted for 26%, 37% and 56% of total N, P and K outflows. The estimated soil loss through ero- sion, using the NUTMON model, was in the range of 1.8–59.4 (mean of 9.7) tonnes ha−1 half year−1 for the farms studied.
Adoption of measures to control soil erosion would thus be important in reducing nutrient decline.
Total emissions, defined as the sum of leaching (OUT 3), gaseous losses (OUT 4), erosion (OUT 5) and human exc- reta (OUT 6), accounted for 66%, 78% and 63% of the N, P and K losses from the studied farms, respectively. Similar
Ecologicalandagro-economicstudyofsmallfarmsinsub-SaharanAfrica203 Table IV. Mean values of nutrient flows and balances for the long rainy season in 2002 (half-year period) according to livestock category in the 30 studied farms (standard deviation in parenthesis).
Flows N (kg ha−1) P (kg ha−1) K (kg ha−1)
TLU< 0.34 0.34< TLU TLU> 0.72 All (n = 30) TLU < 0.34 0.34 < TLU TLU > 0.72 All (n = 30) TLU < 0.34 0.34 < TLU < 0.72 TLU > 0.72 All (n = 30)
(n= 10) < 0.72 (n = 10) (n= 10) n= 10 (n= 10) n=10 n= 10 < 0.72 (n = 10) n=10
Nutrient stock 1818 1928 1744 1830 553 567 555 558 5623 4630 4307 4854
(595) (646) (501) (568) (258) (262) (131) (217) (3044) (20111) (1094) (2197)
IN 1 0.30 3.37 1.40 1.69 0.31 2.04 0.92 1.09 0.01 2.10 0.82 0.97
(0.6) (7.6) (3.2) (4.8) (0.7) (3.4) (1.7) (2.3) (0.0) (6.4) (2.6) (4.0)
IN 2a 0.57 1.28 0.29 0.71 0.08 0.32 0.04 0.15 0.32 0.64 0.14 0.36
(0.9) (2.6) (0.5) (1.6) (0.1) (0.5) (0.1) (0.3) (0.5) (1.1) (0.2) (0.7)
IN 2b 5.01 9.33 23.69 12.68 0.33 0.62 1.58 0.85 6.68 12.45 31.59 16.91
6.3) (9.4) (25.8) (17.7) (0.4) (0.62) (1.71) (1.2) (8.45) (12.6) (34.4) 23.6)
IN 3 6.71 6.71 6.71 6.71 1.10 1.10 1.10 1.10 4.41 4.41 4.41 4.41
(0.0) (0.0) (0.0) (0.0) (0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
IN 4 10.76 13.96 9.27 11.33 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
(6.3) (19.0) (7.2) (12.0) (0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
IN 5 0.00 0.00 0.00 0.00) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
(0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
OUT 1 –4.20 –8.66 –3.34 –5.40 –0.52 –0.90 –0.38 –0.60 –2.31 –4.97 –1.86 –3.05
(3.6) (13.5) (5.7) (8.7) (0.6) (1.4) (0.7) (1.0) (1.9) (7.9) (3.1) (5.0)
OUT 2a –0.17 –0.22 –0.10 –0.16 –0.03 –0.08 –0.02 –0.05 –0.13 –0.23 –0.05 –0.14
(4.1) (0.5) (0.2) (0.4) (0.1) (0.2) (0.1) (0.1) (0.3) (0.5) (0.1) (0.3)
OUT 2b –2.36 –3.83 –9.67 –5.28 –0.21 –0.38 –0.72 –0.44 –3.17 –5.14 –12.90 –7.07
(2.5) (2.1) (8.0) (5.8) (0.2) (0.2) (0.6) (0.4) (3.4) (2.6) (10.7) (7.7)
OUT 3 –3.43 –4.87 –4.68 –4.32 0.00 0.00 0.00 0.00 –0.22 –0.47 –1.02) –0.57
(1.0) 3.4) (1.6) (2.2) (0.0) (0.0) (0.0) (0.0) (0.2) (0.5) (0.9) (0.7)
OUT 4 –0.86 –1.30 –1.58 –1.25 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
(0.2) (1.0) (0.8) (0.8) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0) (0.0)
OUT 5 –6.84 –8.77 –9.06 –8.22 –1.46 –1.90 –1.99 –1.78 –14.70 –16.32 –16.49 –15.84
(5.2) 9.3) (9.8) (8.1) (1.0) (2.1) (1.9) (1.7) (11.1) (18.8) (17.3) (15.6)
OUT 6 –6.20 –10.10 –5.75 –7.35 –1.68 –2.75 –1.55 –2.00 –1.22 –1.75 –1.20 –1.39
(3.3) (17.5) (2.4) (10.2) (0.91) (4.8) (0.7) (2.8) (0.6) (3.0) (0.5) (1.8)
Partial balance –7.10 –8.80 6.50 –3.10 –1.70 –1.10 –0.10 –1.00 0.20 3.10 16.50 6.60
(2.9) (15.9) (19.9) (15.9) (1.2) (3.0) (2.6) (2.4) (4.3) (6.4) (24.8) (16.2)
Total balance –0.70 –3.10 7.20 1.10 –2.10 –1.90 –1.00 –1.70 –10.30 –9.30 3.40 –5.40
8.6) (10.0) (15.1) (12.0) (1.4) (2.7) (2.9) (2.4) (13.0) (20.1) (17.4) (17.7)
Depletion of stock (%) –0.04 –0.16 0.41 0.06 –0.38 –0.34 –0.18 –0.30 –0.18 –0.20 0.08 –0.11
IN 1: inorganic fertilisers and feeds; IN 2a: imported organic fertilisers; IN 2b: imported manure from grazing; IN 3: wet and dry atmospheric deposition; IN 4: biological N fixation; IN 5: irrigation and flooding; OUT 1: harvested crop products; OUT 2a: crop residues and manure; OUT 2b: excreted manure outside farm; OUT 3: leaching; OUT 4: gaseous losses; OUT 5: erosion; OUT 6: human excreta;
OUT 6: TLU= tropical livestock unit (250 kg live weight).
studies conducted in semi-arid areas of Eastern Kenya have confirmed that current farming practices result in N, P and K decline (Gachimbi et al., 2002).
Although nutrient decline was observed, the N, P and K stocks in the upper 30 cm soil layer (calculated from the chem- ical analyses of the soil samples) were still large. This illus- trates one of the problems in soil fertility and hence nutrient balance studies: even though the soils are deficient in N and P (Tab. I) and nutrient decline may be taking place (Tab. IV), large quantities may still be present. On average, P and K stocks were declining at rates of 0.3% and 0.1% per half-year period, respectively.
There were low quantities of nutrients extracted in eco- nomic crop products (OUT 1) and crop residues (OUT 2a) due to poor weather and low soil water conditions (affecting nutrient uptake) and poor crop performance: 10.8 kg N ha−1, 1.1 kg P ha−1, and 10.3 kg K ha−1. Similarly, nutrient inputs (IN 1 and IN 2a) into the farming system were low. The rate of application of inorganic fertilisers was low, supplying less than 2 kg of nutrients (N, P and K) per hectare. This was well be- low the recommended basal fertiliser application rates of 30–
50 kg N ha−1 and 30–50 kg P2O5 ha−1 for the staple crop, maize, in the study area (Ouma et al., 2002).
A number of socio-economic factors determined the mag- nitude of nutrient balances in the study area, namely livestock, off-farm income and size of cultivated land. Livestock was the major determinant of nutrient balances in the study area. This is because the free-range livestock in the subsistence-oriented farming system concentrates nutrients from communal pas- tures into areas under crop cultivation. Livestock is kept un- der an open system where livestock are grazed in communal areas or in fallow lands during the day and corralled at night.
Farm N balance (kg ha−1) was correlated with tropical live- stock units (r= 0.65, P < 0.01). Similarly, farm P and K bal- ances, in kg ha−1, were positively correlated with tropical live- stock units (r= 0.29, P < 0.05 for phosphorus balance; r = 0.05, P< 0.05 for K balance).
Farm P balance (kg ha−1) was positively correlated with off- farm income (r = 0.41, P < 0.05). We observed from this study that households with access to off-farm income poten- tially stand a high chance of purchasing phosphorus contain- ing inorganic fertilisers. We also observed from this study that farm K balance (kg ha−1) was negatively correlated with cul- tivated land area in hectares (r= –0.39, P < 0.05), implying that cultivated large farms have comparatively high levels of K depletion. This is probably due to the ability of cultivated large farms to produce marketable crop and residue surplus, which export nutrients off the farm without adequate replenishment.
3.2. Farm economic and social sustainability
The farming system studied was economically viable (pos- itive net farm income), but operating on a subsistence scale with low performance of major economic indicators (Tab. V).
The percentage of farm produce sold was low (22%) and farm- ing was practised mainly as a means of securing household food needs. The share of crops (primary production units) in
net farm income was about 88%, showing its importance in de- termining the economic viability of the farming system stud- ied (while livestock numbers was the key determinant of soil nutrient balances).
The off-farm income share in family earnings was 51%, implying that farming activities were inadequate at meeting household needs. This was further corroborated by the fact that farm net cash flow was negative for 52% of the studied house- holds, prompting them to explore opportunities elsewhere to bridge income and food gaps.
Furthermore, considering family earnings, all households were living below the World Bank-defined poverty line of 1 US $ a day (World Bank, 2000, 2001) and the Government of Kenya-defined poverty line of Ksh. 1239 (US $ 16.5) per adult equivalent per month for rural areas (CBS, 2000). Poor people depend heavily on a natural resource base for their basic needs, such as food, energy, water and housing, and in their desperation to survive may use and overuse natural resources, resulting in a vicious cycle of land degradation–
declining productivity–poverty-low income and further land degradation (GTZ, 1995).
The farming system studied realised low returns for labour compared with average wages of unskilled agricultural labour of US $ 1.1 to US $ 1.5 per day. About 97% of labour for farm- ing activities was family manual labour. Cropping activities accounted for 56% of labour allocation to farming activities.
The level of mechanisation was low, with about 7% of house- holds using animal traction for land preparation. The manual operations negate the prospects of intensification and increas- ing productivity on a large scale. Labour supply for farming operations was further undermined by a lack of interest in farming among the youth, who would rather seek alternative employment in urban areas.
3.3. Production risks and agro-economic performance of crops
Farmers’ approaches to managing production risks in- cluded diversification of farm enterprises (crops and live- stock), early planting, intercropping and judicious use of farm inputs (inorganic fertilisers, manure, etc.). The studied farm- ing households were cultivating 4.7 fields as a strategy for spreading out risks associated with climatic variables. How- ever, the mean number of fields cultivated did not differ sig- nificantly with resource endowments: 4.5 for households with TLU< 0.34; 4.7 for households with 0.34 < TLU < 0.72; and 4.9 for households with TLU> 0.72.
Crops were dominantly grown under an intercropping sys- tem. For the sample studied, 69 plots were sown to inter- cropped maize while only five plots were sown to monocrop maize. These figures were seven and two for beans, respec- tively. Maize and beans are the main staple food crops in the study area, as shown in Table VI. Yields of maize and bush beans were less than the target potentials of 5 tonnes ha−1 and 2 tonnes ha−1, respectively (KARI, 1994; Okoko and Makini, 1999). Factors contributing to low yields include use
Ecological and agro-economic study of small farms in sub-Saharan Africa 205
Table V. Farm economic and social performance indicators for the studied farming system over a half-year period (mean values with standard deviation in parenthesis)a.
Economic indicators Mean values of economic indicators per half-year period TLU< 0.34
(n= 10)
0.34< TLU < 0.72 (n= 10)
TLU> 0.72 (n= 10)
All (n= 30)
Net farm income (US $) 93.3 (78.3) 81.8 (57.4) 104.4 (95.2) 93.2 (76.3)
Net farm income (US $ ha−1) 64.0 (51.2) 114.0 (179.0) 94.8 (80.0) 91.0 (114.8)
Gross margin PPUb(US $ ha−1) 77.4 (43.0) 100.8 (142.9) 61.0 (60.6) 79.7 (91.3)
Gross margin SPUc(US $ TLU−1) 142.9 (222.0) 71.8 (25.7) 53.3(29.3) 89.3 (131.6)
Gross margin RUd(US $) –11.1 (6.1) –22.9 (9.2) –45.4 (21.2) –26.5 (19.7)
Off-farm income (US $) 64.5 (63.1) 147.0 (230.0) 77.4 (93.0) 96.3 (147.3)
Family earnings (US $) 157.8 (85.0) 228.8 (218.4) 181.8 (122.1) 189.5 (150.3)
Family earnings (US $ consumer unit−1)e 32.0 (16.0) 55.1 (49.8) 39.6 (26.9) 42.2 (34.2)
Family earnings (US $ person−1) 24.0 (11.8) 33.6 (24.0) 28.8 (21.7) 28.8 (19.6)
Farm net cash flow (US $) 20.5 (43.9) –12.7 (53.8) –20.6 (62.5) –4.3 (55.1)
Household net cash flow (US $) 85.0 (77.1) 134.3 (227.3) 56.8 (124.2) 92.0 (154.0)
Off-farm income (% of family earnings)
41 64 43 51
Market share (% of produce sold) 36 20 11 22
Households below poverty line (%) 100 100 100 100
Labour
General family labour (days) 9.4 (9.6) 2.8 (4.2) 4.0 (5.5) 5.4 (7.2)
PPU family labour (days)2 249.7 (130) 141.5 (66.4) 159.8 (69.1) 183.7 (102.1)
SPU family labour (days) 29.9 (31.0) 53.8 (43.6) 77.1 (72.3) 53.6 (53.8)
Off-farm family labour (days) 64.1 (59.3) 96.4 (157.7) 67.6 (75.1) 76.0 (103.8)
Hired labour on PPUs (days) 8.1 (10.4) 14.5 (24.8) 5.3 (8.9) 9.3 (16.3)
Hired labour on SPUs (days) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0)
Hired labour on RUs (days) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0)
Labour intensity for crops (days ha−1) 263.4 (210.3) 220.7 (253.1) 147.5 (51.5) 210.5 (191.8) Return to family labour
(US $ day−1)
0.4 (0.4) 0.5 (0.4) 0.5 (0.4) 0.4 (0.3)
a1 US $= 75 Ksh;bPrimary Production Unit – crop activities;cSecondary Production Unit – livestock activities;dRedistribution Unit – nutrient storage activities;eConsumer units in adult equivalents; TLU= Tropical livestock unit (250 kg live weight of an animal); Poverty line = 1 US $ per person per day; Household net cash flow= farm net cash flow + off-farm income; Family earnings = net farm income + off-farm income.
of germplasm with low genetic potential, low soil fertility, poor rainfall distribution, diseases and pests.
Although farmers prefer intercropping of maize and beans to monocropping, the yields of the two crops were low in the intercropping system. Legume dry matter production and ni- trogen accumulation are reduced in intercropping systems be- cause of competition from the companion crop (Nambiar et al., 1983). The low maize grain yields of intercropped maize have been corroborated with other studies in Uganda (Kasenge, 2000). This implies that farmers intercrop for other reasons besides grain yields.
In this study, intercropping of maize with beans was more beneficial in terms of reduced nutrient decline and higher eco- nomic gains than monocropping of either crop (Tab. VI). The latter has been corroborated by the work of Francis (1978) and Nadar (1984), who both reported that intercropping of maize
with beans results in higher economic gains than monocrop- ping of either crop, when maize-bean price relations are taken into account. Other studies have also reported that farmers continue to intercrop maize with beans because of the possi- bility of harvesting two crops from a field in one season; min- imising crop failure risks; utilising available land optimally;
and reduced weeds, pests, soil erosion and peak labour de- mands (Jodha, 1981; Okigbo, 1981).
3.4. Production risks and agro-economic performance of livestock
The studied farm households were keeping an average of 2.3 groups of animals per household as a risk management strategy. Goats, chicken, zebu cattle and rabbits were kept by 93%, 100%, 20% and 20% of the households, respectively.
Table VI. Impacts of intercropping on yields and nutrient balances of staple food crops in the Mbeere District of Eastern Kenya1.
Description Crops
Maize monocrop (n= 5)
Maize intercrop2 (n= 69)
Bean monocrop (n= 2)
Bean intercrop3 (n= 7)
Maize-bean intercrop (n= 53)
Yield (kg ha−1) 166.5 (176.2) 159.5 (166.6) 345.1 (148.2) 180.1 (152.9) Maize: 160.7 (169.0) Beans: 137.6 (111.1) Gross value (US $ ha−1) 302.9 (555.2) 198.6 (326.1) 244.6 (284.6) 117.8 (115.6) 163.3 (203.5) Gross margin (US $ ha−1) 265.1 (510.7) 159.4 (314.0) 229.4 (265.5) 91.3 (98.4) 129.8 (176.2) Net Cash Flow (US $ ha−1) 23.0 (38.7) 52.1 (291.0) 10.4 (14.7) 5.2 (34.6) 23.7 (71.6) Variable costs
◦ Seeds (US $/ha) 2.1 (0.9) 6.2 (13.1) 1.0 (0.4) 4.4 (4.5) 7.5 (14.7)
◦ Mineral fertilisers (US $ ha−1) 0.0 (0) 5.2 (12.7) 0.0 (0) 1.1 (2.9) 4.6 (12.5)
◦ Organic fertilisers (US $ ha−1) 35.7 (50.5) 11.5 (33.8) 13.4 (19.4) 10.1 (13.9) 4.8 (12.5)
◦ Hired labour (US $ ha−1) 0.0 (0) 12.4 (26.2) 0.5 (0.7) 10.7 (14.2) 11.9 (25.7)
◦ Traction (US $ ha−1) 0.0 (0) 3.4 (19.7) 0.0 (0) 0.0 (0) 4.4 (22.4)
◦ Others (US $ ha−1) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0)
Total variable costs (US $ ha-1) 37.8 (50.1) 39.2 (58.5) 15.2 (19.1) 26.5 (25.0) 33.5 (52.4) N partial balance (kg ha−1half year−1) –44.1 (90.0) –19.3 (41.9) –57.2 (68.3) –22.3 (19.7) –21.4 (46.0) P partial balance (kg ha−1half year−1) –8.3 (17.6) –0.7 (7.5) –4.9 (5.9) –2.3 (2.3) –1.2 (7.33) K partial balance (kg ha−1half year−1) –24.0 (66.4) –3.8 (40.8) –40.1 (48.8) –12.9 (9.5) –5.6 (42.29) N full balance (kg ha−1half year−1) –48.0 (92.1) –10.1 (30.5) –23.6 (11.6) –16.6 (15.1) –9.3 (29.52) P full balance (kg ha−1half year−1) –8.4 (18.0) –1.4 (7.8) –6.0 (4.4) –3.2 (2.6) –1.7 (7.36) K full balance (kg ha−1half year−1) –30.1 (69.5) –16.8 (46.7) –60.7 (32.1) –25.5 (13.1) –16.4 (43.42)
11 US $= 75.0 Ksh at time of study;2Maize intercropped with other crops than beans;3Beans intercropped with other crops than maize.
Broad-based livestock diversity contributes to sustainable food supply, resilience of agricultural and natural ecosystems, and provides farmers with the opportunity to reap other benefits associated with livestock and farm productivity (Mohamed Saleem, 1998).
The productivity of the animals in the study area was low.
Milk production was in the range of 1–2 litres per day per cow with a lactation period of 150 days. The indigenous goats were kept mainly for meat and other socio-cultural functions.
Indigenous hens lay a batch of 10–12 eggs and then become broody, repeating this 3–4 times a year. Although the gross margins of livestock were positive (due to valuation of manure and other livestock products), the net cash flows were negative and livestock were kept for other reasons (socio-cultural) than economic efficiency (Tab. VII).
3.5. Limitations of the study
The study attempted to assess farm nutrient flows and bal- ances, current farm productivity and socio-economic sustain- ability of the dryland farming systems using only 30 farm households. The sample of 30 households is small, although it generated quite interesting results. Future studies are needed that involve a larger sample of farmers over time. This will allow better estimates of nutrient balances, farm productivity and socio-economic sustainability.
The study duration of a complete cropping cycle was short.
Ideally, studying sustainability of production systems should be temporal (spanning many cropping cycles and calendar years), because some of the nutrient flows may take place over more than one year. The study was further limited in capturing long-term climatic data (e.g. rainfall), which was unavailable in the study site. The impact of inter-seasonal variability on nutrient balances and farm economic performance could not be studied in depth either. However, total rainfall for the long rains was 903 mm (first half-year period), which was within the annual average rainfall of 800–1100 mm. It is thus envis- aged that the results of the rest of the year could probably be comparable, ceteris paribus.
The NUTMON model and indicators used in this study proved useful in diagnosing the ecological and agro-economic sustainability of smallholder farms, but the model was lim- ited in providing an understanding of the dynamics of nutri- ent pools, gender and kinship factors in relation to on-farm and off-farm income, prices and labour and in making sim- ulations for future scenarios of nutrient and farm economic performance. There is a need to link NUTMON to a dynamic model so as to simulate the effects of farm management prac- tices in time as well as to determine the effects of feedback mechanisms. There is also a need to link nutrient budgets with various nutrient pools (mineral pool, solution pool and organic matter pool) to improve the interpretation of nutrient balances.
