Frank M. Anderson
Agricultural Economist & Team Leader
ILCA Highlands Programme, ILCA, Addis Ababa, Ethiopia
Abstract
Introduction
An overview of smallholder feeding systems in Ethiopia
Examples of feeding systems
Implications for research
Acknowledgements
References
The productivity of both crop and animal agriculture in Ethiopia is stagnant. Major increases in both crop and animal production are a prerequisite to the nation achieving food self-sufficiency. Livestock play a vital and catalytic role on Ethiopia's smallholder farms. Oxen, for example, are the main source of agricultural power.
Different farming systems have evolved in Ethiopia over many generations, but many of them will be unable to sustain adequately the increased human and livestock populations certain to exist in the coming years. The middle altitudes of the country are the areas most likely to be required to absorb the bulk of the increased population.
In the future, livestock will depend increasingly upon crop byproducts as their feed source. Pasture areas are decreasing as cropped areas expand. Subsistence-oriented smallholders have limited means with which to boost production and will be unlikely to adopt new feeding systems requiring substantial cash outlays. Furthermore, they will not invest their resources in increasing fodder production if it is at the cost of reducing the output of subsistence food crops.
Researchers must seek innovative ways of boosting livestock production, and recognize that the ultimate beneficiary of the research is the farmer through the links between livestock and crop production. Benefits occurring to the livestock enterprise will carry forward to crop production.
Different technical means exist to boost fodder production on farms and to raise the effective feeding value of crop by-products. However, not all these technologies are well suited to the resource endowments and objectives of smallholders. If gains in fodder production require substantial investments, or the returns on investment are only via a multi-stage production process, they will not be likely to be adopted by farmers. Some strategies to increase fodder production were examined and found wanting in this regard.
The most promising research thrust which addresses the problems of both crop and livestock production concerns crop-legume interactions. This study area deserves priority allocation of research resources. Successful research on this complex of issues, i.e. research producing adoptable results, will require use of the farming-systems approach to research and a commitment to multidisciplinary during its execution.
Agriculture in Ethiopia is dominated by smallholder farms where over 90% of the country's grain is produced. Smallholder farmers cultivate an average of less than 1 ha and the 7 million farm families in Ethiopia cultivate perhaps 6 million ha each year. Although agriculture is organized in a socialist framework, collective farming accounts for less than 5% of the total area cultivated. The bulk of agricultural output is produced by individual farmers who have "farming rights" over the land they till (Gryseels et al 1984). With few exceptions, crop irrigation is not practiced, and under these dryland conditions farmers produce a surplus above subsistence needs equivalent to approximately 10% of total production on each farm.
In adverse years, grain production shortfalls are ordinarily made up by the sale of livestock and surplus grain that farmers may have been able to accumulate from previous years. However, as the Ethiopian tragedy of 1984 and 1985 revealed, even such actions can be pitifully inadequate. Hunger, destitution and death can easily occur in the subsistence farming areas.
Integrated crop and livestock production is an ancient tradition in Ethiopia. However, changes are taking place in the production systems in many parts of the country which are challenging the ability of subsistence farmers to sustain the integration which was easily achieved by past generations when there was much less pressure on land resources. The more important changes occurring include progressively declining average farm sizes in response to rising rural populations, encroachment of cropping land onto erstwhile grazing areas and onto less fertile and more easily erodable lands, and expansion of degraded lands which can no longer support either annual crops or pastures.
These factors are combining to increase the relative dependence of farmers' livestock upon cereal-crop residues and stubble. Unless supplemented, these products are seldom adequate for livestock maintenance and even less so for milk production or as finishing rations. Also, using current technologies, farmers in the coming years will have smaller quantities of crop residues available per farm and be less able to augment these cereal products by pasture grazing or hay. The areas under pasture are being reduced and average stocking rates are increasing.
Existing and future agricultural production systems in Ethiopia will require draught power for cultivation. Livestock will also continue to be farmers' principal buffer in times of crop failure, as well as a vital source of cash for the purchase of farm inputs and consumption goods.
The challenge to forage researchers will be to reduce this emerging competition between crop and livestock production and enable new, more productive and stable farming systems to develop. Improvements in animal-feed production from declining land areas are urgently and will be a critical means of addressing this problem.
This paper highlights issues concerned with research on the improvement of animal feed resources for the livestock of Ethiopian smallholders. Farmers' circumstances directly influence the acceptance and adoption of new technologies and are stressed here as necessary and desirable influences on forage-related research if the research is to produce relevant results.
The following section of the paper gives an overview of important smallholder feeding systems in Ethiopia with special reference to the farmers' objectives in managing these systems. The problems confronting farmers in achieving these objectives are emphasized. The penultimate section relates research needs to farmers' forage-production problems and opportunities. Finally, this is followed by a summary and conclusions.
The General Production Environment and Livestock Management Systems
Mixed crop-livestock farming in Ethiopia is dominantly on farms located above 1,500 m above sea level and with annual rainfall in excess of 700 mm. Crops are grown at altitudes up to about 3,300 m. Some highland areas have a bimodal rainfall pattern. Most crops are sown in the main rains, which fall between June and October over most of the country. The short rains from February to May are unreliable and only in some areas can farmers crop routinely in that secondly season. Nationally, some 5% of the grain crop is produced in the short rains.
The complex physical environment and ecology of the Ethiopia highlands has resulted in the evolution of distinct agricultural production systems in different parts of the country. Some 85%, or 35 million of the 43 million people in the country, are resident in the highlands. A similar percentage of the country's estimated 30 million cattle, 25 million sheep and 20 million goats are kept by smallholder mixed farmers. Farmers in these systems also practice different animal-feeding and management strategies, reflecting differences in crops grown, in the relative importance and productivity of crops and pastures, in the proximity of the cropping areas to pastoral areas (meaning access to replacement stock and especially draught oxen) and differences in the intensity of-land use by farmers and their livestock.
Domestic livestock in the central highlands are grazed communally on fallow and permanent pasture lands for the duration of the cropping season and on crop lands after harvest. Livestock of indigenous breeds are seldom fed, with this practice being restricted to the few thousands of crossbred dairy cows owned by smallholders. Selected classes of livestock, especially draught oxen and lactating cows, receive supplements to grazing when they are housed at night. Production levels are low. Cows of indigenous breeds, for example, calve about every two years on average and produce some 300 litres of milk per lactation over and above that consumed by the suckling calf.
The Objectives of Livestock Keeping
The primary purpose of cattle keeping in most of the highlands is to produce oxen. The low rates of reproduction and milk production of indigenous cows, both of which are directly influenced by nutrition levels throughout the cows' production cycles, result in low overall levels of conversion efficiency of feedstuffs into draught power. Draught power is the primary product of cattle enterprises. At present farmers keep up to ten cattle to produce and maintain the pair of oxen conventionally used for cultivation.
Oxen are only used for some 60 pair days of work each year. Thus the feed demands of this subsystem (oxen plus cows and followers) in relation to the direct contribution of oxen to production are burdensome and favour changes in management to reduce aggregate feed demands per unit of draught power and the draught-power needs to crop production. The single-ox cultivation system developed by ILCA is one way of reducing the feed demands of the draught animals on farms with up to 1.5 ha under cultivation (Gryseels et al 1984). Using one ox rather than a pair can effectively have the total feed requirements of the cattle kept to provide-draught power.
Cattle manure provides an important fraction of domestic fuel needs in areas with limited fuel-wood supplies. Manure sales make significant contributions to cash incomes. Neat and milk are relatively less important products of cattle enterprises. In some farming systems, such as in western Shoa Province, selected fields are intensively manured by livestock penned overnight in the fields. However, little of the animal manure produced in the country is used intentionally as fertilizer for crop production.
According to location, farmers keep sheep and/or goats as producers of cash and meat. The high rates of flock growth, relative to cattle herds, and the different grazing behaviour of small ruminants make them appropriate complements to cattle through most of the Ethiopian highlands. Small ruminants are the best stock to be sold in times of crop failure They seldom receive special feeds and ordinarily are grazed together with cattle. Special-purpose fattening of smallstock is not practiced by Ethiopian smallholders.
Livestock are also a major source of cash income. In large parts of the highlands, trade in livestock and the sale of livestock and livestock products provide up to 80% of farm cash income in a normal year (Gryseels and Getachew 1985). In summary, livestock produce a range of intermediate and final products in the traditional farming systems of Ethiopia. Regional differences exist in the relative importance of these products, but in all cases the presence of livestock on smallholders' farms enables them to be more productive and stable over time than would be the case in their absence (Rodriguez and Anderson 1985).
Some Problems of Existing Smallholder Feeding Systems
Seasonality of Fodder Production
A production problem common to most Ethiopian livestock feeding systems is the seasonality in animal feeds supplies. Some parts of the country have dry seasons of up to 200 days each year. Such areas farmers must harvest and store fodder for dry-season use when feed supplies in the fields are extremely limited. Communal grazing of crop and pasture lands after harvest is the norm, so individual farmers cannot calculate with certainty the contribution of such grazing to their needs. Farmers usually stack all cereal straws near the homestead after threshing, selling only that portion needed to provide for basic household needs.
Effective short rains in a particular year can result in pasture growth enabling farmers to sell surplus straw. As noted previously, farmers have competing demands on their labour during the cropping season and "ideal" conservation practices for hay, especially, are not often used. The development and adoption of labour-saving technologies for crop production could release labour for fodder conservation, in turn improving livestock production.
The low frequencies with which cows calve oblige farmers to maximize annual conception rates by not limiting breeding to any time of the year. The resultant spread of calving over the year means that farmers have difficulties in matching feed supplies and feed demands. Farmers maximize their use of communal grazing lands (for which there are no cash payments due) and then utilize stored feeds for their stock on a priority basis. Oxen have first call on conserved feed, followed by lactating cows and the classes of cattle and smallstock if supplies are adequate. Informal surveys by ILCA in the Debre Birhan area of Ethiopia (about 2,800 m above sea level) suggest that farmers trade in smallstock (sheep in that area) both to increase cash incomes and to balance feed demands with available feed supplies.
Fodder Quality
The second major problem in smallholder feeding systems in Ethiopia is that the quality of most harvested and conserved feedstuffs is such that when fed alone it is often unable to provide even for the maintenance needs of livestock. Thus, on an annual basis, livestock must make sufficient liveweight grains during the flush period of the year to be able to offset losses occurring during the following dry season. The regular cycle of liveweight gain and loss each year is the principal cause of the low annual average conversion efficiencies of fodder into meat, milk and draught power.
In commercially-oriented mixed farming systems farmers could if it were profitable, negate this cyclical pattern by feeding of purchased supplements. Such a strategy is rarely possible in Ethiopia at present and is limited to the few farmers with easy access to major urban centres. The majority of dairy and fattening operations are in such situations. (Some 75% of all smallholders are more than one day's return walk from even an all-weather road.) Alternative means of overcoming the annual cycle of gain and loss will have to be developed for these farmers.
Land Tenure
Some aspects of the social context of crop and livestock production in Ethiopia have been noted. All land is the collective property of the Ethiopian people and individuals have user rights only over the land they till each year. While, in principle, an individual farmer can be assigned new plots each year, in practice this does not happen. However, it does foster discernible reluctance in farmers to invest in the maintenance and improvement of any plots of land they use. Also the current system whereby crop land becomes available for communal grazing after harvest encourages farmers to leave crop lands bare of cover after harvest. This predisposes such fields to erosion and has negative long-term impacts on soil fertility and structure.
Also, the individual's fragmented holdings require farmers to haul crop by-products to the homestead for safe keeping. From there they are given to their stock as required. The labour costs of this operation are substantial. Although the traditional method of threshing by trampling using cattle and equines (horses and donkeys) is relatively inefficient from the point of view of grain extraction, it does favour livestock production through the relatively higher grain and husk fraction remaining than would occur if the grain where threshed by mechanical means.
Shoa Province
Farmers in the central Ethiopian highlands in Shoa Province, have a tradition of hay-making from permanent (usually valley-bottom) pastures. The hay is usually cut well after its protein content has peaked, and the protein percentage in harvested hays on a dry-matter basis seldom exceeds 5%. If harvested at the "right" time, the protein content on a dry-matter basis can exceed 11%. However, at the "right" time for hay making, farmers favour using their scarce labour for weeding of cereal and pulse crops. Subsistence food crops have the highest priority for farmers' time. Furthermore, as most of the hay is used as a feed supplement for oxen, farmers report that they maximize dry-matter yields per hectare by harvesting later in the season, although they are aware that the feeding value of hay made at that time is reduced. Hay making is relatively more important in the higher altitude areas (say above 2,500 m) than in lower altitude areas. The latter tend to be more intensively cultivated and have less pasture land available per farmer.
Frost is a major crop hazard toward the end of the main growing season in the higher altitudes and farmers there attempt to avoid frost damage to crops by sowing on land away from the more frost-prone valley bottoms. Even in the long term it is unlikely that main-season crop production will be important in these high-altitude valley bottoms. Routine cropping on them would require development of frost-tolerant, short-duration cereal and pulse varieties. Maturation periods as short as 90 days under relatively low temperature and radiation conditions will be required. Crop-breeding research in Ethiopia will be less problematic and have greater impact on national production if targeted on the needs of the less rigorous agricultural environments in the middle altitudes of the highlands.
Crop production in the higher and cooler mixed-farming systems will continue to be risky and farmers are likely to become increasingly dependent upon livestock production as human populations increase. However, without significant improvement in overall soil fertility (perhaps largely rectifiable by rock phosphate application), the quality of feedstuffs available to stock will not be increased. In turn this places upper bounds on the per-unit productivity of livestock. Gains in productivity per head will then only be possible if strategic supplements are provided and these will not be likely to be produced within these systems. Also, the prospects of identifying a productive legume shrub for the higher altitude areas are remote (S. Jutzi, personal communication).
Under traditional management regimes both crops and pasture are more produced for livestock. Leguminous plants are not cultivated specifically to provide supplements for the cereal-straw-based diets. These straws are now complemented only to a limited extent by pasture grazing and hay. Purchased feed supplements are used routinely by an insignificant percentage of farmers and fallow land is a diminishing proportion of the land used regularly for cropping. In the Debre Zeit area, for example, the proportion of cultivable land under fallow in any year is less than 5% (Gryseels and Anderson 1983).
Hararghe Province
In other parts of Ethiopia, such as the highlands of Hararghe Province in the south-east of the country, limited areas of permanent grazing are available and livestock depend more than in the central highlands upon thinnings from annual crops during the growing season and crop stover and stubble during the dry season. The proportion of arable land left fallow is minimal. Substantial areas of land on more than 50% slopes are cultivated, reflecting the severe shortage of arable land, even though net returns from crops on such land are at best, marginal. Sorghum and maize are the dominant annual crops. Perennial crops, including coffee, provide a substantial proportion of the cash income of farmers in the area.
All cattle and most small ruminants are tethered in the field during the day and hand fed selected crop parts according to the season. As elsewhere throughout Ethiopia, all animals are penned at night, usually at the owner's homestead. The Hararghe highlands are close to extensive rangeland areas and the working oxen in Hararghe Province come mainly from the rangelands. Typically smallholders purchase oxen from the rangelands (through traders), use them as draught animals for some years and then fatten them prior to sale. The oxen are fattened successfully on farm products alone. The regional success of this strategy is reflected in the price premium offered to fatstock from Hararghe Province on the Addis Ababa market, the most important domestic meat market in the country. The relatively close proximity of the Province's smallholders to pastoralists in the rangelands areas enables Hararghe farmers to keep relatively more efficient herds and flocks (in terms of rates of conversion of animal feed into draught power and other livestock products) than is the case in the central highlands. More than half the cattle kept by Hararghe farmers are working oxen. The interregional links between cattle breeders and smallholders who have different production objectives are advantageous to both groups.
The Approach
Cost-effective agricultural research requires that researchers undertake research only on those problems for which results are expected a priori to have an impact on production. A substantial body of theory and experience under the rubric of the "farming systems approach to research" (FSR) has been developed in the last decade. The philosophy guiding this approach is that adequate knowledge of farmers' production circumstances is a prerequisite to the efficient design and execution of agricultural research. The FSR approach offers relevant guidelines for research on animal feeding systems and forage research.
As argued earlier in this paper, improvements in animal-feed resources in the Ethiopian highlands, both in total and on an individual-animal basis, necessitate substantial changes to traditional practice. Ethiopian agricultural systems are under stress and simple incremental changes to existing systems will be an inadequate means of producing the major increases in production which are necessary.
Subsistence-oriented smallholders, such as those which dominate Ethiopian agriculture, will not adopt technologies which increase the production risks associated with their staple crops. In regard to staple food crops, the annual rates of return on marginal investments to increase crop yields need to be as high as 40% before farmers will adopt them on a large scale. Forage crops as pure stands produce intermediate products and only have their value expressed when they are used by animals. It is arguable, therefore, that the annual rates of return on investments in such fodder crops in subsistence-oriented systems with shortages of arable land must exceed 40% as this will also have to account for the opportunity cost of not producing staple crops which are ordinarily the main users of arable land. Together, the opportunity cost of not producing staples, and the production of intermediate products, only militate against the prospect of widespread use of special-purpose fodder crops by subsistence farmers.
The urgency of solving the feed-supply problem for both people and their animals means that other methods of raising the quality and quantity of feed on offer to livestock must be sought. Several avenues are possible. These are outlined below. The first concerns the strategic use of agro-industrial by-products to raise the effective feeding value of farm produced fodders. The second concerns strategies to increase farm production of animal feeds. The last strategy considered relates to exploitation of the synergistic link between legumes and cereal crops.
Agro-Industrial By-Products
Agro-industrial by-products are in short supply in most African countries. Ethiopia has only limited supplies of various oilseed cakes and molasses which, even when efficiently used, cannot boost production for more than a small proportion of the ruminant livestock in the country.- Research is in progress to identify where these resources would be best used. This will most likely be by enterprises producing meat and milk and will be unlikely to be used by smallholders in support of draught oxen feeding. The general shortages of cash and credit in Ethiopian smallholder systems make it improbable that farmers will purchase such inputs to enhance the working capacity of their draught oxen.
Difficulties of access to markets for farm inputs and products will also limit the use of agro-industrial by-products by farmers in remote situations. Thus, while research on the best utilization of these products is important, at this stage in Ethiopia's agricultural development it has lower priority than the two other avenues of fodder development outlined below.
Intensified Crops and Pasture Production
Crop yields are low in Ethiopia. Nationally, average cereal yields are substantially less than 1 t/ha for most crops. Grain and straw yields of most crops are positively correlated, and increases in grain yields through the use of new crop varieties and fertilizers can be expected to increase the yields of crop by-products. However, significant increases in grain yields at the national level will take many years to achieve. Even at this time the yield gaps between crops on research stations and the same crops on farmers' fields are indicative of the major research effort needed to have farmers adopt even currently available technologies. However, there are potentially conflicting changes to crop production as regards the impact on animal feed supplies and at present it is not possible to predict which changes will dominate. For example, if farmers adopt dwarf cereal varieties in order to increase grain yields they may reduce the quantity of byproducts produced per hectare. Alternatively, if they change from a wheat to a sorghum-based cropping system they would ordinarily increase the byproducts available to their stock.
Concerning pastures, as all permanent pasture areas in Ethiopia have at least some element of communal tenure related to their use, it is improbable that they will be a target for development by individual farmers. If profitable productivity gains are demonstrable, then the proper avenue for investment would be by the Peasants Association which has group control over the use of pasture lands within their boundary. Research on technical and socio-economic issues related to permanent pasture development is not well advanced.
Fallow lands are relatively more important in the higher altitudes of the highlands where the soil-fertility levels are such that fallowing for several years (sometimes up to 15 years) is necessary to support low-input cereal and pulse cropping for three or so years. In these cases farmers will not invest directly in fallow land improvement. Here the principal means of improving pasture yields will be through fertilization of the cereal crops in the cereal-pulse-fallow rotation. In this way farmers will receive the benefits of the purchased inputs directly through the increase in cereal and pulse-grain yield and indirectly through the carry-over of nutrients to the fallow pasture phase. Research on ways and means of improving fallow lands alone without reference to the crops in the rotation has low priority.
The Cereal-Legume Link
Legume crops are an important element in most Ethiopian smallholder farming systems. Only in those systems which are semi-commercial, as in Hararghe Province, do some farmers eliminate legumes from their cropping patterns. Nationally legume production is almost exclusively of grain legumes. Pulses are an integral part of the diet of most Ethiopians. In addition to contributing to soil nitrogen and producing a break in cereal-dominated crop rotations, legume crap by-products contribute modestly to livestock production. Vicia faba, for example, is not fed to livestock except in combination with cereal straws. Ordinarily farmers favour cereal production over legume-grain production because of the higher prices received for the former.
Forage legumes are grown as single-stand crops by an insignificant number of smallholders. The conditions of subsistence production do not favour using crop land for fodder crops instead of grain crops. Fodder crops produce only an intermediate product which must then be transformed into a saleable commodity such as meat or milk for the farmer to receive the benefits of his cash outlay and investment of labour and land. The risks perceived by farmers to be attendant upon this multi-stage process are sufficient to limit its adoption by farmers even though it can be a profitable activity (Gryseels and Anderson 1983). Regular and successful participation by Ethiopian farmers in the market economy is a precondition for the widespread adoption of special-purpose fodder crop production.
Given these considerations, can the well-established benefits which legumes confer upon cereal crops be tapped to advantage by subsistence farmers and incorporated into their farming system?
While a definitive answer to this question must await the results of research now in progress by several agencies in Ethiopia, the evidence to date is that innovative approaches to using legumes will enable smallholders to increase food-crop production (and the quantity of crop by-products) and simultaneously produce significant quantities of high-protein leguminous materials. Intercropping, serial and relay cropping of cereals with legumes are only a few of several ways of including legumes in systems without diverting land from subsistence food crop production. Research at ILCA initiated in 1985 has shown how legume shrubs, in rows at 5 m spacings between rows, can produce up to 2 t/DM/ha/year of 20% crude protein content (at 1,850 m above sea level) without significant reduction in yields of the cereals grown between the rows. This quantity of leguminous material is adequate on a per-hectare basis as a strategic feed supplement to convert a below-maintenance straw diet into one allowing a crossbred cow to produce 58 litres of milk per day.
Such results are indicative of the substantial potential of legumes to boost and stabilize food-grain and fodder production in Ethiopia. Research on ways and means of profitably integrating fodder legumes into Ethiopian farming systems is a high priority. They are especially relevant as they require modest cash inputs in addition to those used to boost crop production and are adoptable equally by farmers in areas of easy and difficult market access. Also, given the degradation of agricultural land which-is now occurring in Ethiopia, such legume technologies will play a central role in ensuring the long-term ability of Ethiopia to be self-sufficient in food needs. Effective research on the crop-legume complex will require multidisciplinary research with an FSR perspective.
The helpful comments of Guido Gryseels, Samuel Jutzi, Barbara Grandin and Jenny Davis on earlier drafts of the paper are recorded with appreciation.
Gryseels, G., Abiye, Astatke, Anderson, F.M. and Getachew, Asamenew. (1984). The use of single oxen for crop cultivation in Ethiopia. ILCA Bulletin No. 18, Addis Ababa, Ethiopia.
Gryseels, G. and Anderson, F.M. (1983). Research on farm and livestock productivity in the central Ethiopian highlands: Initial results 1977-1980. ILCA Research Report No. 4, Addis Ababa, Ethiopia.
Gryseels, G. and Getachew, Asamenew. (1985). Links between livestock and crop production in the Ethiopian highlands. Research Brief in ILCA Newsletter, Vol. 4(2).
Rodriguez, G. and Anderson, F.M. (1985). Farm risks: a case study of a mixed farming system in highland Ethiopia. IFPRI/DSE Workshop on sources of Increased Variability in Cereal Yields, Feldafing, Germany, November 26-29, 1985.
