Improving milk production in small-scale dairy farms in Botswana: Incorporating legume fodder in the farming systems

W S Boitumelo and W Mahabile

Department of Agricultural Research
Private Bag 0033, Gaborone, Botswana

ABSTRACT

A research project was established to investigate the problems of milk production by small-scale cattle owners in Botswana. The major constraint identified was lack of adequate feed (quantity and quality), particularly during the dry season. The research emphasis was therefore to integrate fodder crops such as lablab (Lablab purpureus) into the farming systems and establish a practical feeding programme based on planted fodders and harvested crop byproducts. Participating farmers planted 1 ha with lablab, with or without single superphosphate fertiliser. On-station trials were established to study the effects of fertiliser application, the optimal stage of harvest and appropriate methods for harvesting, drying and storage.

Feeding packages developed by the project were tested on-farm. A comparative feeding trial was conducted involving Simmental-Tswana crossbreds and Tswana milking cows fed sorghum stover supplemented with different levels of lablab and sorghum bran as protein supplements.

During the four years of the trial (1985-1989), average lablab dry-matter yields from farmers' fields were 1.14, 0.70, 2.82 and 0.87 t/ha. Fertiliser application did not affect (P>0.05) lablab yields on either farmers' or research plots, and there were no differences (P>0.05) in yields between farmers' fields and the research plots. Lablab contained more crude protein than sorghum stover (16.4 vs 6.4%). The lactation yield of the crossbred cows was more than twice that of Tswana cows (3.8 vs 15 kg/day).

Although the major traditional crops in the project area are sorghum and millet intercropped with cowpea and watermelon, the project has succeeded in convincing farmers to allocate one 1 ha of land for lablab production and fodder production has become a component of the cropping system.

RESUME

Amélioration de la production des petites exploitations laitières au Botswana: introduction des légumineuses fourragères dans les systèmes agraires

Un projet de recherche a été lancé pour étudier les problèmes rencontrés dans le domaine de la production laitière par les petits éleveurs de bovins au Botswana. D'après les résultats de ces travaux, les pénuries et la qualité médiocre des aliments, notamment au cours de la saison sèche, constituent certains des obstacles les plus sérieux au développement de la petite exploitation laitière dans le pays. Fort de ces conclusions, il a été décidé de mettre l'accent sur des travaux destinés, non seulement à intégrer des plantes fourragères comme la luzerne (Lablab purpureus) dans les systèmes agraires en place mais également à élaborer un système adapté d'alimentation basé sur la culture de plantes fourragères et les sous-produits agricoles. Les paysans participant à ces travaux ont semé 1 ha de luzerne avec ou sans application de superphosphate simple. Des essais en station ont en outre été conduits en vue d'étudier l'effet de l'engrais et de déterminer le stade optimal de récolte ainsi que les meilleures méthodes de coupe, de séchage et de conservation.

Des paquets alimentaires élaborés dans le cadre de ce projet ont été expérimentés en milieu réel. Un essai d'alimentation comparatif a été réalisé sur des vaches métisses Simmental x Tswana et des Tswana, avec une ration composée de paille de sorgho complémentée avec différents niveaux de luzerne et de son de sorgho comme sources de compléments protéiques.

Pour les quatre ans qu'a duré cette étude (1985-1989), la production moyenne de matière sèche de la luzerne au niveau des paysans s'établissait à 1,14; 0,70; 2,82; et 0,87 t/ha. L'application d'engrais n'avait aucun effet (P>0,05) sur la production de la luzerne, ni au niveau des paysans ou du projet ni entre les parcelles des paysans et celles du projet. La teneur de la luzerne en protéines brutes (16,4%) était supérieure à celle de la paille de sorgho (6,4%). Enfin, les vaches métisses produisaient plus de deux fois plus de lait (3,8 kg/j) que les Tswana (1,5 kg/j).

Bien que les principales cultures traditionnelles de la zone d'étude soient le sorgho et le mil, entre lesquelles étaient généralement intercalés le niébé et la pastèque, le projet a réussi à convaincre les paysans de réserver 1 ha de terre à la culture de luzerne, faisant désormais de la production fourragère une composante à part entière du système agraire.

INTRODUCTION

Increasing milk production by pert-urban small-scale farmers in Botswana has been the major objective of the Ministry of Agriculture during the current six-year National Development Plan (1985-91). The Animal Production Research Unit (APRU), Department of Agricultural Research, established a project to investigate the problems of milk production by small-scale cattle owners. This project is jointly funded by the Government of Botswana and the International Development Research Centre (IDRC), Canada.

The major constraint identified was lack of adequate feed (quantity and quality) to sustain milk production, particularly during the dry season. Large quantities of crop residues are produced and can be harvested for dry-season feeding of lactating and in-calf cows (Mosienyane, 1983; Mosimanyana and Kiflewahid, 1987). However, crop residues are of low quality and cannot satisfy the nutritional requirements of these cows. Therefore, the research emphasis has been to incorporate fodder crops into the fanning system and establish a practical feeding programme based on planted fodder and crop residues.

Agronomic trials to examine various varieties of fodder legumes and pasture grasses were conducted at a number of locations between 1977 and 1980 (APRU, 1980). The results indicated that from among several annual legumes screened for drought resistance, ease of management, grazing potential and dry-matter yield, lablab (Lablab purpureus) was the most productive and the most easily established in the communal areas of Botswana (APRU, 1979). It has soil requirements similar to those of cowpea, and better disease and insect resistance, and can be grazed, ensiled or conserved as hay (Humphreys, 1978). In addition to these qualities and its high nutrient content, it is useful in terms of improving soil fertility in crop rotation in small-scale farm conditions (APRU, 1979).

The specific of objectives of the work described here were to determine:

· productivity and nutritive value of lablab
· optimal methods for conservation
· the potential for including a fodder legume such as lablab into the small-scale mixed farming systems.

MATERIALS AND METHODS

Fodder legume production and conservation

Beginning in 1985, the project introduced lablab to the participating farmers. Each farmer was given 20 kg of lablab seed and 100 kg of single superphosphate fertiliser (10.5% P) to plant 1 ha of land; the fertiliser was applied to half of each hectare. Three 1-ha plots were also planted at Sebele Agricultural Research Station. Superphosphate was applied at the rate of 0, 100 or 250 kg/ha on each plot. An additional 27 ha of land were planted to study the optimal stage of harvest and appropriate methods for harvesting, drying and storage, and for on-station feeding trials

Lablab yield was measured using 2-m radius circular subplots from three random locations in each of the farmers' plots and the research plots. After measuring yield, the plants were harvested, sun-dried and stored for dry-season feeding. After harvest the total quantities of lablab, crop stovers and post-harvest residues stored by each farmer were estimated.

Feeding trial

A dry-season on-farm feeding trial was carried out to determine the voluntary feed intake and performance of Simmental-Tswana crossbred and Tswana lactating cows fed sorghum stover supplemented with various levels of lablab hay and sorghum bran. The objective was to formulate a nutritionally optimal diet for lactating cows based on low quality but widely available sorghum stover and farm-produced lablab hay or sorghum bran. The four treatment diets (see Table 1) were formulated to satisfy the daily nutrient requirements for maintenance and milk production of lactating dairy cows such as those used in this trial (NRC, 1978; Crampton and Harris, 1969).

Table 1. Experimental diets

Treatment	Composition (%, dry-matter basis)			Nutrient content (%)
	Sorghum stover	Lablab bay	Sorghum bran	Crude protein	Total digestible nutrients
Diet 1	60	40	-	10.5	56.3
Diet 2	60	30	10	10.0	58.8
Diet 3	60	20	20	9.5	61.0
Diet 4	60	10	30	9.3	64.1

In diet 1 lablab was mixed with the stover by hand at the time of feeding; in the other three diets, the sorghum bran was fed separately in split oil-drums. The animals were provided with ad libitum bonemeal-salt (1:1 w/w ratio) and vitamins A, D and E. The treatments were randomly allocated to each of 40 farmers. Voluntary feed intake was determined in all treatment groups by measuring the daily refusals over a period of seven days and for a maximum of six lactating cows per farm. Statistical analysis using the t-test (Snedecor and Cochran, 1967) was conducted on the performance data obtained from the lactating cows that completed the trial.

Nutritive value

Three replicate samples of lablab, sorghum stover and sorghum bran were submitted for laboratory chemical analysis.

Samples of fresh whole plants and plant parts were air-dried (60°C) and then ground in a Wiley mill (1-mm screen). The dried and ground samples were stored in air-tight bottles and later analysed for organic matter, crude protein, crude fibre, ash and mineral components according to methods approved by AOAC (1975) and in vitro dry and organic matter digestibility according to Tilley and Terry (1963) procedures.

Dairy cattle performance

Each farmer was provided with two calibrated 10-litre plastic milk buckets for recording milk yield. Milk records were collected and summarised by project staff at the end of each month.

RESULTS AND DISCUSSION

Fodder legume production and conservation

Average dry-matter yields (t/ha) of lablab by farmers and on the research plots in 1985/86 and 1986/87 are shown on Table 2. Fertiliser application had no effect (P>0.05) on average dry-matter yields on farmers' plots or on research plots, and there were no significant differences (P>0.05) in dry-matter yield between farmers' fields and the research plots. The results from the two seasons were not greatly different. Low dry-matter yields and lack of fertiliser response could be due to lower than average rainfall in both seasons.

Table 3 shows the total quantities of lablab dry matter harvested and conserved by the project farmers during the four years of the project. The higher lablab dry-matter yields in 1987/88 than in the other three years were attributed to higher rainfall in the project area (651 mm) in that year. The decrease in the total quantity of lablab hay harvested by farmers in 1988/89 season, even though the rainfall was higher than normal, was due to the improved grazing conditions caused by reasonable rainfall.

Table 2. Average dry-matter yield of lablab hay from project farms and research plots as influenced by fertiliser application rate

Year	Dry-matter yield (t/ha) at fertiliser application rate of
	Rainfall (mm)	0 kg/ha	100 kg/ha	250 kg/ha
Farmers' plots
1985/86	328	1.23	1.44	-
1986/87	380	1.37	1.49	-
Research plots
1985/86		1.41	1.56	1.70
1986/87		1.47	1.54	1.79

Table 3. Lablab dry-matter yields from farmers' fields

Year	Mean rainfall (mm)	No of farmers	Average dry-matter yield (t)	Total dry-matter yield (t)
1985/86	328	12	1.14	13.1
1986/87	380	30	0.70	21.0
1987/88	651	40	2.82	112.8
1988/89	565	40	0.87	34.8

The major constraint observed during harvesting and conservation was the time required to dry the stems. Lablab leaves dry and shatter within three days but it takes up to six weeks for the stems to dry completely. Various methods of drying and storage using tripods, stocking on the ground and combing three rows into one row were tried at Sebele Research Station. Method of drying had no effect on drying time but the crop was baled more efficiently from stooks and tripods with minimum dry-matter loss due to leaf shattering, mould or termite damage. In 1987/88 lablab was dried in rows and baled six weeks after harvest.

Feeding trial (dry season)

There were no differences (P>0.05) in the average daily dry-matter, crude-protein and energy (TDN) intakes by lactating cows fed the four diets (Table 4).

Table 4. Average voluntary intake of nutrients by lactating cows

Diet	Voluntary intake (kg/day)
	DM	DM/100 kg liveweight	CP	CP/100 kg liveweight	TDN	TDN/100 kg liveweight
Diet 1	9.43	2.46	0.99	0.26	5.33	1.39
Diet 2	8.77	2.43	0.88	0.24	5.16	1.42
Diet 3	8.93	2.50	0.85	0.24	5.45	1.52
Diet 4	8.22	2.24	0.76	0.21	5.27	1.48

Diet 1: 60% sorghum stover and 40% lablab hay
Diet 2: 60% sorghum stover, 30% lablab hay and 10% sorghum bran
Diet 3: 60% sorghum stover, 20% lablab hay and 20% sorghum bran
Diet 4: 60% sorghum stover, 10% lablab hay and 30% sorghum bran

Nutritive value

The nutrient composition and in vitro digestibility valves for lablab, sorghum stover and sorghum bran are shown in Table 5. Lablab contained more crude protein and had higher dry-matter and organic-matter digestibilities and calcium to phosphorus ratios than sorghum stover. The higher digestibility coefficients are attributed to the higher crude-protein and lower crude-fibre content in lablab than in sorghum stover. The calcium to phosphorus ratios were within acceptable limits reported for cattle (NRC 1978). Due to high crude-protein and mineral contents observed in lablab plants, dry-season diets for lactating and in-calf cows were based on sorghum stover supplemented with lablab hay.

Table 5. Average nutrient composition of diet components

Feed	Composition (% of dry matter)						In vitro digestibilities (%)
	Organic matter	Crude protein	Crude fibre	Ash	Ca	P	DM	OM
Lablab hay	90.8	16.4	27.7	9.2	1.68	0.13	59.9	57.1
Sorghum stover	91.6	6.4	32.5	8.4	0.38	0.10	54.8	49.8
Sorghum bran	97.3	11.5	3.2	2.7	0.38	0.31	54.4	47.2

Dairy cattle performance

Comparative basic data on dairy cattle performance on-farm are given in Table 6. Average lactation lengths for Simmental-Tswana cows were longer (P<0.05) than for Tswana cows in 1985/86, 1986/87 and 1988/89 but similar in 1987/88. Average daily milk yields per cow were similar within breed but different between breeds. On average, milk yield per day per cow for both breeds tended to increase with increasing seasonal rainfall; however, the reason for the low yield in the high rainfall year 1987/88 is not clear.

On average under similar feeding and management systems, Simmental x Tswana cows produced twice as much milk as Tswana cows. These results are comparable with those obtained in on-station trials (APRU 1988).

CONCLUSIONS

The predominant crops in the traditional cropping system of the project area are sorghum, maize and millet intercropped with cowpea and watermelon. The dairy project has succeeded in convincing participating farmers to allocate one hectare of land for lablab fodder production. Fodder production has become a component of the cropping system.

The dry-season feeding strategy using high protein Lablab purpureus legume fodder, planted and harvested by farmers and fed in association with crop byproducts and post-harvest residues, has been a significant intervention in the small dairy farms.

Table 6. Milk production of Simmental x Tswana and Tswana cows

Item			1985/86	1986/87	1987/88	1988/89
No of farmers			12	30	40	40
No of cows per farmer
	Simmental x Tswana		10	15	24	20
	Tswana		24	74	140	111
Lactation length (days)
	Simmental x Tswana		273	285	303	262
	Tswana		223	202	296	181
Milk yield/cow (kg)
	Per lactation
		Simmental x Tswana	760	890	818	988
		Tswana	338	319	474	420
	Per day
		Simmental x Tswana	2.8	3.1	2.7	3.8
		Tswana	1.5	1.6	1.6	2.3
Rainfall (mm)			328	380	651	565

Fertiliser applications on farm plots as well as on station plots did not show any significant effect (P>0.05) on lablab dry-matter yield. Lack of response may be attributed to low seasonal rainfall and distribution.

Chemical analysis showed that lablab contained 2.5 times more crude protein than sorghum stover (16.4 vs 6.4%). The use of lablab as a protein supplement in crop stover based diets is the practical approach for the dry-season feeding of lactating and in-calf cows.

REFERENCES

APRU (Animal Production Research Unit, Department of Agricultural Research). 1979. Livestock and range research in Botswana Annual Report 1979. Government Printer, Gaborone, Botswana. pp 46-49.

APRU (Animal Production Research Unit, Department of Agricultural Research). 1980. Ten years of animal production and range research in Botswana. Government Printer, Gaborone, Botswana. p. 56.

APRU (Animal Production Research Unit, Department of Agricultural Research). 1988. Livestock and range research in Botswana Annual Report 1988. Government Printer, Gaborone, Botswana. pp. 15-45.

AOAC (Association of Official Analytical Chemists). 1975. Official methods of analysis. 12th edition. AOAC, Washington DC, USA.

Crampton D W and Harris L E. 1969. Applied animal nutrition. 2nd edition. W H Freeman, San Francisco, California, USA.

Humphreys L R. 1978. Tropical pastures and fodder crops. Longman, London, UK. 135 pp.

Mosienyane B P. 1983. Crop residues for animal feeding. The Bulletin of Agricultural Research in Botswana. 1:3.

Mosimanyana B and Kiflewahid B. 1987. Feeding of crop residues to milking cows in small -scale farms in Botswana. In: Little D A and Said A N (eds), Utilization of agricultural by-products as livestock feeds in Africa. Proceedings of a workshop held at Ryall's Hotel, Blantyre, Malawi, September 1986. ARNAB (African Research Network for Agricultural By-products). ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia. pp. 127-135.

NRC (National Research Council). 1978. Nutrient requirements of domestic animals, No 5. Nutrient requirements of dairy cattle. 5th edition. NRC, Washington, DC, USA.

Snedecor G W and Cochran W G. 1967. Statistical methods. 6th edition. Iowa State University Press, Ames, Iowa, USA. pp. 54-56.

Tilley S M A and Terry R A. 1963. A two stage technique for the in vitro digestion of forage crops. Journal of the British Grassland Society. 18:104-111.