Economic implications of forage innovations on smallholder farms in western Kenya

F B Nyaribo ¹, J F M Onim ¹ and D L Young ²

¹ Winrock International Institute for Agricultural Development
Small Ruminant Collaborative Research Support Program (SR-CRSP)
PO Box 252, Maseno, Kenya

² Department of Agricultural Economics
Washington State University
Pullman, WA 99164, USA

ABSTRACT

In response to a declining forage base, the Small Ruminant Collaborative Research Support Program (SR-CRSP) has developed a number of low cost forages and forage preservation techniques for the dual-purpose goat technology currently on trial on smallholder farms in western Kenya. Representative whole farm linear programming models incorporating a number of the new forages were constructed for three farm sizes in Hamisi Division. Food and cash crops, existing forages and zebu cattle were also included in the models. Results showed a high degree of complementarily between crop and liverstock production. Livestock stocking rates increased considerably with the smallest farm realising the largest net income gains from the new technologies.

RESUME

Incidences économiques des innovations en matière de technologies fourragères introduites dans les petites exploitations de l'ouest du Kenya

Face à la dégradation de la base de ressources fourragères, le Programme d'appui à la recherche concertée sur les petits ruminants (SR-CRSP) a mis au point un certain nombre d'espèces fourragères et de techniques de conservation des fourrages bon marché destinées à l'élevage des caprins à double fin. Celles-ci sont actuellement évaluées dans de petites exploitations de la région d'Hamisi dans l'ouest du Kenya A cet effet, des modèles de programmation linéaire tenant compte de toutes les activités agricoles, des caractéristiques de certaines nouvelles variétés fourragères et de la taille des exploitations ont été élaborés. Les autres paramètres en compte dans ces modèles sont les caractéristiques des cultures vivrières et des cultures marchandes ainsi que celles des espèces fourragères traditionnelles et de l'élevage bovin. Les analyses effectuées font ressortir une forte complémentarité entre la production végétale et l'élevage. Ces innovations technologiques permettaient d'augmenter considérablement les taux de charge des petites exploitations et partant, de maximiser les bénéfices nets des petits exploitants.

INTRODUCTION

Much of African development research has focused on crop farming to the neglect of the livestock sector (Senga, 1976). However, small ruminants (sheep and goats) play a significant support role in small farming systems of the world (Winrock International, 1983). In Kenya, for example, about 43% (6.5 million) of the country's small ruminants are reared on small farms. Between 1969 and 1978, the goat population increased by 4.8% annually, and the offtake rate rose from 17 to 31% (Schulter, 1984). With increasing prices of other meat sources, such as beef, sheep and poultry, demand for goats will probably continue to rise.

The agricultural sector is expected to continue to play a key role in the Kenyan economy, especially in employment creation. According to World Bank (1989) statistics the growth rate of the agricultural sector has steadily declined from 4.9% in 1965-80 to 2.8% in 1980-86. Several solutions are being explored to revitalise Kenya's agricultural development and growth. Currently research resources are being channelled to the lower strata of livestock farmers who in many cases still employ traditional production practices.

For the past 10 years the Kenya Small Ruminant Collaborative Research Support Program (SR-CRSP) has been working on development of the dual-purpose (milk and meat) goat and supporting technologies in smallholder farms of western Kenya, using a fanning systems research approach With increasing populations and the resultant pressure on land, the dual-purpose goat is an appropriate livestock technology alternative to the zebu cattle which are usually reared on these farms. The dual-purpose goat is low risk and requires fewer capital inputs and smaller land area per unit, and yet under good management conditions produces higher milk yields per unit of required inputs. The overall objective of this study-is to assess the economic impact of the dual-purpose goat and new forage technologies as a means of improving smallholder family incomes, and human nutrition, in Hamisi Division of western Kenya. The specific objectives are to:

· determine the magnitude of economic returns from the dual-purpose goat and forage technologies developed by the SR-CRSP in western Kenya

· determine the level of integration of the new forage technologies in these smallholdings

· determine economic and nutrition constraints to adoption of the dual-purpose goat.

MATERIALS AND METHODS

Study area

The study is being carried out in Hamisi Division, in the western highlands of Kenya in Kakamega District. The area is classified as a high potential agricultural zone. Annual rainfall is 1200-2100 mm, with a bimodal distribution - the long rains season runs from March to June, and the short rains fall between September and November. The soils are humic nitrosols with moderate to high fertility (Hart et al, 1984). Hamisi Division is one of the most densely populated regions of Kenya with about 615 persons/km².

The farming system in Hamisi is a sedentary mixed crop-livestock system, with two cropping seasons. The primary food crops are maize intercropped with beans. Secondary food crops include sweet potato, cassava, banana and a variety of green vegetables. Tea and coffee are the major cash crops. All household keep some livestock. SR-CRSP (1986-87) surveys indicate mean livestock numbers per farm of 2.10, 0.20 and 0.20 Tropical Livestock Units (TLU: 1 TLU = 250 kg liveweight) for zebu cattle, sheep and indigenous goats. But because more than 40% of participant farmers own 1 ha of land or less, these stocking rates are considered high (Semenye, 1990; Jaetzold and Schmidt, 1982); given existing natural unimproved grazing pastures, Jaetzold and Schmidt (1982) recommend a stocking rate of 0.6 ha/TLU. It is clear that available land is being overgrazed.

Sources of livestock feed

Livestock feeds are derived from off-farm grazing and crop residues such as maize stover, maize thinnings, banana peelings and sweet potato vines. The importance and availability of any specific feed varies by season and off-farm grazing is perhaps the most important source, followed by crop residues. Heavy reliance on off-farm grazing is supported by the long grazing hours reported by Conelly et al (1987).

Table 1 shows the mean cultivable and grazing land, labour and capital endowments for small, medium and large farms in the study area. Project data indicate no significant difference in labour availability per household so the same figure was used for all farms. There is a negligible variability in mean own capital per household. The small area of on-farm fallow land available for grazing on the small and medium farms is the land around the homestead which is not in direct competition with food crop production. On large farms a greater area can be left fallow for grazing. Livestock nutrition budgets constructed by Hart et al (1984) estimate an average of 0.40 ha of off-farm grazing per household for the sample of farmers in Hamisi. Detailed descriptions of the farming system and resource endowments for the study area have been reported elsewhere (Sands, 1983; Nyaribo, 1989; Nyaribo and Young, 1991).

Table 1. Resource constraints by farm size, Hamisi Division, western Kenya

Farm size	Land (ha)		Labour adult equivalent (days)	Own capital (KSh)
	Cultivable	Grazing a
Small	0.59	0.50	1461	3051
Medium	1.19	0.60	1461	3151
Large	2.01	1.39	1461	3737

^a Refers to the sum of on- and off-farm grazing land available. On-farm fallow grazing land is 0.10 ha for the small and medium farms and 0.89 ha for the large farm

KSh 23 = $US 1.00 in 1990

Baseline surveys in the early phase of the SR-CRSP project identified lack of adequate quality feeds as one of the constraints to dual-purpose goat production (Sands, 1983). Seasonal fluctuations in the forage supply are such that in certain months there is excess feed which is wasted if it is not stored. In response to the shortage of livestock feed, SR-CRSP has introduced several forage interventions and forage storage technologies (Hart et al, 1984; Onim et al, 1985).

Study methodology

Linear programming was used to test the economic feasibility of dual-purpose goat and forage production and storage technologies. Representative farm linear programming models were constructed for three farm sizes for the Hamisi study site. The linear programming model maximises gross margin, net of human consumption, from the farm's production activities subject to fixed resource constraints. Gross margin is defined as total revenue minus total variable COSTS. The model solves for the following decision variables:

1. How many hectares to plant to food, cash and forage crops
2. The desired inventory of cattle and dual-purpose goats
3. How many animals to sell during the year
4. How much food to produce, how much to sell and how much to keep for home consumption
5. How much cash to borrow to augment cash balances.

The main constraints are:

1. Maximum cultivable and grazing land available
2. Maximum family labour available
3. Maximum hired labour available
4. Minimum family subsistence requirements
5. Minimum livestock nutrient requirements
6. Maximum available working capital for input and food purchases.

The time unit is one calendar year. The year is divided into four time periods to allow for the transfer of products and resources. The seasonal disaggregation of the livestock component is crucial because of the seasonal changes in the quality and quantity of forages. Also, most farming activities are performed at specific times during the year which leads to distinct seasonal patterns in resource use and supply (Hazel and Norton, 1986).

The livestock component generates demand for feed on a quarterly basis which in turn determines the optimal stocking rate. This permits isolating livestock nutrient needs that may be at risk in certain periods of the year. Dry matter and crude protein supplied by the different sources are balanced with the demands for livestock feeding. The requirements are specified by livestock category and by period.

Livestock production activities

There are two dual-purpose goat breeding activities. The unit of measure is a doe-kid unit. Nutrient requirements for the doe-kid units are the sum of those for the lactating doe, the kid and a yearling replacement doe. The first doe kids in April and the second in November. Weaner kids from the dual-purpose goat production activities are available at five months of age. Nutritional requirements are withdrawn each period the animal remains on the farm. Kid rearing activities produce animals which are transferred to kid sale activities. Weaner kid nutrient requirements are drawn only up to 14 months. To allow for flexibility in sale of kids within the year, sale dates are not predetermined. Accordingly four kid sale activities, one corresponding to each quarter, are specified in the model. The breeding stock replacement activity allows for yearling females to be kept for breeding stock replacement. The labour and nutrient requirements are included in the doe breeding activity.

One cattle production activity is specified. Zebu cattle produce milk for home consumption as well as for sale. They also produce calves and culls for sale. The milk sale activities allow for sales of milk from both dual-purpose goats and cows in all periods.

The model includes three livestock feeding categories; April kidding dual-purpose goat feeding, November kidding dual-purpose goat feeding, and cattle feeding. Livestock derive their feed from eight sources: maize stover, fresh pigeon peas, Sudan grass, off-farm and on-farm fallow grazing, fence row forage, pigeon pea hay, Sudan grass hay and mixed grass hay. In the model this results in 78 seasonal feeding activities. The model determines the optimal mix of feed consumed subject to dry-matter and crude-protein requirements. It also determines quantities of feed to be transferred to feed-deficit months.

Other production activities of the model include intercropped maize and beans, sorghum, vegetables and bananas. Along with maize stover, grass hay and on- and off-farm grazing, the new forages intercropped with food crops are Sudan grass and pigeon peas. Cash crops considered are tea and coffee.

Data sources

The data used in this study came from surveys conducted by SR-CRSP scientists (SR-CRSP, 1986-87, 1987; Hart et al, 1984). SR-CRSP work covers smallholder farmers in two agro-ecological zones and conducts on-farm trials in six villages with a total of 150 participating farmers. In this study modelling results from the Hamisi village with 23 farmers are reported.

MODELLING RESULTS

The different model runs are classified by type of technology and farm size. The technologies were introduced sequentially to the traditional (base) farm. Thus:

· TI denotes traditional technology, referred to as the base model
· TII denotes the base model with the dual-purpose goat and no new forages
· TIII denotes the addition to TII of new forages without forage storage technology
· TIV denotes the full technology package which includes forage storage.

Up to 42% of the SR-CRSP participant farmers own 1 ha of land or less. It was thus important to evaluate the impact of these technologies on various farm sizes. Accordingly the sample was stratified into three farm sizes and the means of the land, labour, and capital endowments for each farm size category (see Table 1) were used in the linear programming models.

Table 2 summarises the impact of the new technologies on stocking rates and on farm income by farm size. As anticipated, the new technologies have had the greatest impact on the smallest farms, where the full technology package (TIV) increased stocking rates by 767% and farm income by 497%, compared with the base model. Results for the medium and large farms showed similar trends, but at much lower levels; for example, farm income increased by only 22% on medium farms, and by a mere 8% on large farms, with the full package.

Table 2. Impact of dual-purpose goats and new forages on stocking rates and farm income by farm size

Technology	Small farm (0.69 ha)		Medium- size farm (1.39 ha)		Large farm (3 ha)
	Stocking rate (TLU) a	Income (KSh)	Stocking rate (TLU) a	Income (KSh)	Stocking rate (TLU) a	Income (KSh)
TI	0.49	1509	0.49	9738	1.22	13792
TII	0.73	1600	0.70	9813	1.22	13792
TIII	3.87	6946	2.73	11554	2.41	14773
TIV	4.25	9003	3.35	11906	2.95	14949

^a TLU = Tropical Livestock Unit (1 TLU = 250 kg)

KSh 23 = $US 1.00 in 1990

Figure 1 shows the shares of sales from livestock products (milk and meat) and crop sales on small and size farms. (Relative magnitudes for the large farm follow the trend of the medium farm. Thus conclusions regarding the medium farm are also relevant to the large farm.) As each element of the dual-purpose goat and forage technology package was added to the base model the share of crop sales declined while that of livestock product sales increased. This reversal of the contribution of livestock and crop sales to total farm sales can be explained by the fact that farm-gate prices for meat and milk are higher than that for maize grain, the major competing commodity. The results indicate that while crop-livestock associations are biologically complementary, economically they can be competitive because of the relative crop/livestock output price ratios.

Complementarity of crop-livestock production

The complementarily of crop-livestock production systems on smallholder farms is determined by the degree of dependence of each enterprise on the other. In western Kenya crop residues play an important role in providing forages for livestock on the farms. This is clearly illustrated in Figure 2, which shows the share of each forage type in total feed consumed on the small and medium farms under the full technology package (TIV). On the large farm the new forage and hay storage technologies contributed 37 and 15%, respectively, or 52% of the total feed consumed, while maize stover and grazing each contributed 24% of total. These results demonstrate that maize, which is the most important food crop, is highly complementary with production of dual-purpose goats.

Constraints to livestock production

Four major constraints to livestock production were identified (Table 3). Digestible energy was constraining for all farm sizes under all technology alternatives. This finding is consistent with the study by Blackburn et al (1986) in the same area. Grazing land and cultivable land were also constraining for all farm sizes with the small farm realising the largest shadow price on cultivable land. Capital was constraining for the medium and large farms. This points to the need for credit support for smallholder farms.

CONCLUSIONS

Adoption of the new forage and forage storage technologies significantly increased stocking rates and farm incomes on smallholder farms in western Kenya. While all farm sizes benefited from the new technologies the smallest farm size realised the largest net income benefits from adoption of the new technologies. For all farm sizes maximum benefits were realised by adopting the new forages plus the forage storage technology, that is, the full technology package specified as TIV. Due to limited land endowment for the smallest farm, economic competitiveness between livestock and crop production occurred at lower farm incomes, and lower technology levels (between TII and TIII) than it did on the medium and large farms. Digestible energy, cultivable land and grazing land were identified as constraining resources for all farm sizes. Capital, on the other hand, was constraining for the medium and large farms. For the smallest farms, land constraints imply the need for further research in output-increasing technologies per unit of land. This will require the use of improved management and husbandry practices and the use of high yielding maize varieties as well as high milk and meat yielding dual-purpose-goat genotypes.

Figure 1. Shares of livestock products and crops in farm sales

Figure 2. Share of various forages in total forage consumption

Table 3. Economic results of selected variables by technology combination and farm size, Hamisi Division, western Kenya

Farm size	Total live-stock units (TLU) a	Quantity DM (kg)		Farm income	Dual values
					Nutrient b	Grazing land	Cultivated land	Operating capital
		Fresh	Stored	(KSh)	(KSh/kg)	(KSh/ha)	(KSh/ha)	(KSh/KSh)
I
Small	0.49	1690	na	1 509	0.77	1358	7 475	0.00
Medium	0.49	1705	na	9 738	0.77	1358	7 519	0.00
Large	1.22	4220	na	13 792	0.30	294	2 045	3.90
II
Small	0.73	2 049	na	1 600	0.85	1 509	7 475	0.00
Medium	0.70	1 660	na	9 813	0.65	1 153	6 169	0.96
Large	1.22	4 220	na	13 792	0.30	294	2 045	3.90
III
Small	3.87	9 952	na	6 946	0.85	6349	11 946	0.00
Medium	2.73	8 367	na	11 554	0.16	877	857	4.75
Large	2.41	7 111	na	14 773	0.09	183	117	5.28
IV
Small	4.25	6 161	1 918	9 003	0.32	2346	14 816	0.00
Medium	3.35	9 051	1 164	11 906	0.07/0.67	217	469	5.11
Large	2.95	7 730	1 391	14 949	0.08	161	3	5.36

^a TLU = Tropical Livestock Unit (1 TLU = 250 kg)

^b Digestible energy was the constraining nutrient for all farm sizes under all technology alternatives. Crude protein was also constraining for the medium farm under TIV

KSh 23 = $US 1.00 in 1990

Considering the significant share of maize stover and grazing in total feed consumption, these traditional sources of feed will continue to play a key role in the diets of dual-purpose goats and other livestock. Livestock and crop production were shown to be highly integrated and technically complementary. This no doubt contributes to the sustainability of smallholder farms. However, complementarily of livestock-crop production is tempered by economic competitiveness due to farm-gate livestock/crop price ratios. Given the price responsiveness of farmers the balance of farm activities will continue to shift to livestock production the higher the output price ratios.

REFERENCES

Blackburn H D, Cartwright T C, Howard P J and Ruvuna F. 1986. Development of dual purpose goat production for smallholders in western Kenya using computer simulation and systems analysis. Small Ruminant Collaborative Research Support Program Technical Report 89. Texas Agricultural Experimental Station, College Station, Texas, USA. 136 pp.

Conelly W T. Mukhebi A W. Oyugi L and Knipscheer H C. 1987. Household labour allocation in an intensive crop/livestock farming system in western Kenya. Selected Paper, Farming Systems Research Symposium, held at the University of Arkansas, Fayetteville, Arkansas, USA, 18-21 October 1987. 19 pp. [Copies available from SR-CRSP, PO Box 58137, Nairobi, Kenya, or PO Box 252, Maseno, Kenya]

Hart R. Onim J F M, Russo S. Mathuva M and Fitzhugh H. 1984. An analytical framework for feed resources research on mixed farms in western Kenya. Winrock International Institute for Agricultural Development, Morrilton, Arkansas, USA, and Ministry of Agriculture and Livestock Development, Kenya. 43 pp.

Hazel P B R and Norton R D. 1986. Mathematical programming for economic analysis in agriculture. MacMillan, New York, USA.

Jaetzold R and Schmidt H. 1982. Farm management handbook of Kenya, Volume II. Natural conditions and farm management information, Part A, West Kenya, Nyanza and Western Province. Kenya Ministry of Agriculture and the German Agricultural Team (OAT) of the German Agency for Technical Cooperation (GTZ), Nairobi, Kenya. 397 pp.

Nyaribo F B. 1989. Integrating duel purpose goats on small farms in western Kenya: A linear programming analysis. PhD Dissertation. Washington State University, Pullman, WA, USA. 263 pp.

Nyaribo F B and Young D L. 1991. The impacts of capital and land constraints on the economics of new livestock technology in western Kenya. Agricultural Economics (in press)

Onim J F M, Mathuva M, Otieno K and Fitzhugh H. 1985. Simplified hay making for small scale farmers in western Kenya. In: Proceedings of the 5th Small Ruminant-CRSP Kenya Workshop, Kabete, Kenya. SR-CRSP (Small Ruminant Collaborative Research Support Program), Nairobi and Maseno, Kenya. pp. 25-28.

Sands, M W. 1983. Role of livestock on smallholder farms in western Kenya: Prospects for a dual-purpose goat. PhD Thesis. Cornell University, Ithaca, New York, USA. 218 pp.

Schulter M. 1984. Constraints on Kenya's food and beverage exports. IFPRI Research Report 44. IFPRI (International Food Policy Research Institute), Washington, DC, USA. 56 pp.

Semenye P P. 1990. Nutrition and management for dual-purpose goats: Research highlights. Winrock International Institute for Agricultural Development, Morrilton, Arkansas, USA; Kisumu. Lake Printers Ltd. Kisumu, Kenya. 53 pp.

Senga W. M. 1976. Kenya's agricultural sector. In: Heyer J. Maitha J K and Senga W M (eds), Agricultural development in Kenya. Oxford University Press, Nairobi, Kenya.

SR-CRSP (Small Ruminant Collaborative Research Support Program). 1986-87. Rapid rural appraisal survey. Maseno Research Station, Maseno, Kenya. SR-CRSP, Nairobi and Maseno, Kenya.

SR-CRSP (Small Ruminant Collaborative Research Support Program). 1987. Economics project survey. Maseno Research Station, Maseno, Kenya. SR-CRSP, Nairobi and Maseno, Kenya.

Winrock International Institute for Agricultural Development. 1983. Sheep and goats in developing countries: Their present and potential role. World Bank Technical Paper. Washington DC, USA. 116 pp.

World Bank. 1989. Development Report. World Bank, Washington DC, USA. 305 pp.