Smallholder dairy on-farm research in Burundi

J. B. Berahino

ISABU, Avenue de la Cathédrale BP 795 Bujumbura, Burundi

Background country information
Size of the livestock industry
Importation of cattle for breeding purposes
Objectives of the research programme
References

Summary The research programme evaluates the introduction of the improved Sahiwal-Ankole cattle cross and feed production on-farm as compared to a on-station. To assess the relative success of the project different indices were set which included liveweight of offspring, milk and manure production and production parameters (herd miltiplication, milk production, fecundity and mortality rates) were measured All these variations with the exception of mortality were better on-farm than on-station. Concurrently, several forage species were introduced on-farm to improve animal nutrition and to eventually establish the zero-grazing programme. The adaptability of the introduced forage species and their productivity was studied and finally, the manure collected on-farm was used to fertilise arable crops. A comparative analysis of the impact of manure on crop production was made. In addition a laboratory analysis was made on milk processed by traditional technology. The major benefits realised by the package were increased supplies of milk for home consumption, more income from sales of milk and disposable animals and 9 significant improvement in crop production upon application of farmyard manure.

Background country information

Burundi is a land-locked central African country, located between Rwanda to the north, Zaire in the west and Tanzania to the east. A small part of the south-western part of the country is bordered by Lake Tanganyika. The country covers a land area of 27,834 km² and is divided into three ecozones determined by altitude: low altitude (880 m), middle altitude (900-1500 m), and high altitude (1500-2600 m). With a total of 5.3 million inhabitants, Burundi has the second highest population density (207 people per km²) in Africa.

Burundi has a tropical climate that varies according to altitude. The following are the characteristics of the ecozones:

· Low altitude: Surface area is 1875 km²; temperature ranges from 21 to 23°C; the rainy season is about five months long; and the rainfall average per annum is 900 mm.

· Middle altitude: Has a surface area of 8511 km²; the temperature ranges from 18 to 25°C; the rainy season lasts eight months, totalling 1100 mm per annum.

· High altitude: Covers a land area of 15,511 km²; the temperature is 16 to 20°C; and the rainy season is nine months long with a total precipitation of 1500 mm.

Rainfall is generally not scarce but poor soil fertility due to overexploitation is a major set-back. To overcome this problem, the government initiated a policy to encourage livestock production. The emphasis is meat and milk as primary products and manure as a by-product to improve soil productivity. Currently supplies of both meat and milk are inadequate for the country's needs and the government policy is to increase their production.

Size of the livestock industry

The national livestock herd is estimated at 400,000 cattle, 1,100,000 goats and sheep (Table 1), 90,000 pigs and 700,000 poultry.

Table 1. Cattle, goats and sheep population in Burundi

Province	Cattle	Goats	Sheep
Bubanza	28,767	45,224	15,572
Bujumbura	23,779	35,923	13,729
Bururi	89.984	44,848	33,929
Cankuzo	32,123	36,304	4,979
Cibitoke	20,457	50,789	15,080
Gitega	27,317	106,308	44,990
Karuzi	9,717	41,358	31,768
Kayanza	17,500	59,529	33,716
Kirundo	17,554	71,324	17,461
Makamba	15,032	-	-
Muramvya	72,683	56,135	10,377
Muyinga	11,758	85,586	30,987
Ngozi	15,144	66.079	42,552
Rutana	25,747	46,801	18,639
Ruyigi	15,502	56,285	13,305
	423,064	801,473	327,104

Source Département de la production animale (1969).

The livestock industry contributes to the national and rural household income. Animal by-products such as manure are used to improve the fertility of crop land and milk is usually churned on-farm into butter for sale or for home consumption. Farmers also traditionally sell young and old bulls to generate income.

Importation of cattle for breeding purposes

The indigenous Ankole cattle are known to have poor traits for production especially in milk and therefore importation of cattle breeds has aimed at improving the Ankole type. Sahiwals were imported from Kenya and are the most significant exotic breed in Burundi. They were used to improve the indigenous breed (Ankole) through crossbreeding. Importation and cross-breeding has resulted in a national herd of about 1900 exotics and their crosses (Table 2).

Table 2. Importation of exotic breeds into Burundi.

Breed	Number
Friesian	50
Brune des Alpes	50
Jersey	50
Friesian	250
Friesian	300
Friesian	200
Sahiwal x Ankole	50
Montbeliarde	50
Sahiwal x Ankole	850
Total	1,850

Source: Engels (1955).

Crossing Sahiwal and Ankole and subsequent evaluation of the productivity of F₁ and F₂ was conducted at ISABU (Institut des sciences agronomiques du Burundi) cattle research stations. The study concluded that the Sahiwal and Ankole crosses gave a mean yield of 900 kg of milk in 305 days as compared to the Ankole breed which yielded only 400 kg in 225 days under the same conditions. Weight comparisons showed that the male offspring of Sahiwal x Ankole crosses weighed 86 kg at four months and 187 kg at 12 months, whilst the females weighed 78 kg and 181 kg at the same ages, respectively. The Ankole breed offspring weighed only 156 kg at 12 months.

From 1975 ISABU started a selection programme of the Sahiwal and Ankole crossbreds. At the beginning of the programme, the mean milk yield per lactation was 500 kg in 305 days. Ten years later (1985), the mean milk yield per lactation was 550 kg. In Rwanda, similar studies on the crossing of Sahiwal with the Jersey were made. Table 3 shows milking yields of Jersey and Sahiwal crosses in Rwanda. The Sahiwal and Jersey cross appears to be a good milk producer; however, the herd size of both breeds in Burundi is very small and therefore it would be difficult to multiply them.

Table 3. Milk production of Jersey and Sahiwal cross cows In Rwanda in 1981 (305 days)

Number of lactation	Number of cows	Total milk yield (kg)
1 st	37	2795
2 nd	27	2994
3 rd	20	3103
4 th	14	3305
5 th	10	3448
6 th	4	3412
7 th	2	3388
	114	Mean 3203

Objectives of the research programme

The general objective was to improve animal production systems prevailing in Burundese small-scale farms and to strengthen the research capability of ISABU. The specific objectives were to:

· study the animal production systems practised by the smallholders

· introduce improved livestock technologies and evaluate their acceptance by the smallholders

· study the milk processing techniques used by smallholders

· train scientific staff.

Development of the research programme

As 90% of the Burundi population is based in the rural areas and is dependent on arable and animal farming, the economic future of the country relies on the development of both crops and livestock. In cognizance of this fact the government set up a number of agricultural projects in which livestock are a major component. ISABU, a government institution with a mandate to conduct agricultural research in the country, initiated a project in 1960 of crossing Sahiwal with Ankole. This was followed by a selection programme of the crosses in 1975 at Rukoko and in 1978 at Mahwa.

After many years of on-station research work the scientists felt that it was important to introduce the crosses on-farm. Government statistics indicate that the total national cattle herd has declined in the past 10 years partly due to the shrinkage of grazing areas and the demand for beef In consideration of the positive productive interactions between crops and livestock the government felt that an integrated agricultural approach was the most expedient to ensure continuity in crop and livestock development.

In pursuit of this policy, Kisozi was identified as a suitable site for on-farm research as it was representative of the high-altitude areas; its proximity to the Kisozi Research Station was an added advantage. The target group was the smallholder farmer who grew crops and raised livestock. The ISABU scientists put together a package of technologies that comprised a Sahiwal and Ankole cross, an improved cow shed and a feed production programme. However, before on-farm introduction the scientists made a survey to characterise and identify potential cooperating farmers.

The following questions were put to the farmers:

· Do you have your own land?
· Do you live on your own land?
· Do you derive most of your income from the land?
· Do you have less than three head of livestock?
· Have you established a plot of cultivated pastures?
· Do you agree to participate in the ISABU proposal on livestock improvement and its emphasis on the integration with crops?

Methodology and time sequence

The farmers who were identified were asked to fulfil a set of conditions including construction of a simple cow shed and the establishment of a plot of Tripsacum laxum. In 1986, six farms participated in the programme after adhering to the stipulated conditions. They were given an in-calf Sahiwal and Ankole crossbred heifer on credit at a cost equivalent of US$ 65 which had to be repaid in four instalments. In 1987, fifteen additional participating farmers joined the programme and in 1988 and 1989 the number increased by 12 and by six, respectively.

Burundese farmers have traditionally kept cattle for milk and manure production, therefore their acceptance of the package was not difficult. Government also encourages farmers to keep livestock to improve their standard of living. However, the cost of a pregnant heifer is generally prohibitive for peasants, when this is compared with the per capita income of US$ 250 per year. Furthermore, the scarcity of land does not favour raising large ruminants.

Implementation of the programme

The main objective of the project was to characterise the livestock production system and to identify alternatives that promote livestock and agricultural production.

Achievements and failures

The research programme emphasised on-farm introduction of improved cattle and feed production. To assess the programme, different indices on the suitability of the technologies on-farm were set, including liveweight changes of offspring and milk and manure production. Concurrently, several forage species were introduced on-farm to improve animal nutrition and to eventually establish the adoption of a zero-grazing programme. The adaptability of the introduced forage species and their productivity was studied and the manure produced on the farms was used to fertilise arable crops. A comparative analysis of the impact of manure on crop production was made. In addition, traditional milk processing technology was critically analysed during the later phase of the programme. The adaptation of the introduced Sahiwal-Ankole crossbred in the rural area was assessed by measuring production parameters of herd multiplication, milk production, fecundity and mortality rates (Tables 4, 5, 6 and 7) and manure production (Table 8).

Table 4. Herd growth rate and mortality rates on-farm and on-station.

	Station	On-farm
Calving rates (%)	69	77.8
Mortality rates (%)	1.6	2.6
Growth of the herd per annum (%)	29.3	32.1

Table 5. Milk production in the smallholdings of Sahiwal x Ankole crosses (excludes suckled milk by the calves).

	On-farm			On-station
	1987/88	1988/89	1989/90	1990/91	1987/1988
Number of milking cows	20	17	22	33	65
Milk cow (kg)	1,084	1,018	1,055	1,054	693,3
Lactation Period (day)	305	305	305	305	229,4
Kg milk cow/day	3.55	3.33	3.45	3.45	3.00

Table 6. Herd size of introduced types in the smallholdings.

	Total herd	Increase (%)
Number introduced 1987 to 1989	40
Number in April 1991	119	197.5

Table 7. Liveweight of calves and growers at different ages (on-farm and on-station).

	Age
	4 months		12 months		24 months
	Male	Female	Male	Female	Male	Female
On-farm (kg)	80	71	170	142	270	263
On-station (Mahwa)(kg)	88	67	-	-	-	201

Source ISABU (1988).

Table 8. Manure production on-farm with participating and non-participating farmers.

Number observations	Participating farmers	Non-participating farmers
	(t/year/cow)	(t/year/cow)
119	6.7	1.5

All of these indices, with the exception of mortality, showed better performance on-farm than on-station.

The use of animal manure to improve soil fertility has been a long standing practice in Burundi. Participating farmers were encouraged to put bedding (in the form of grass) in the sheds and this was mixed with animal dung to produce farmyard manure. As the manure accumulated it was deposited into a covered pit outside the shed to reduce the level of leaching. The careful handling of manure resulted in high levels of recovery (Table 8).

Tests were conducted evaluate the effect of manure application on the Ndinamaara variety of potato. Manure was applied at 10, 20 and 30 t/ha (Table 9). The subsequent recorded increase in the production of potatoes was 65.5%, 98.8% and 150%, respectively. The percentages were based on the highest achievable yield. It was estimated that five animals could produce adequate manure for one hectare of a potato field.

Table 9. Trial on potato production using farm yard manure.

Manure application (t/ha)	0	10	20	30
Yield of potatoes (t/ha)	4.5	7.45	8.95	11.25
increases in yield(%)	0	65.5	98.8	150.0

This project introduced Tripsacum laxum, Pennisetum purpureum, Setaria sphacelata and Calliandra calothyrsus onto farms. Among these, Tripsacum laxum was the most acceptable as reflected by the total areas planted (Table 10).

Table 10. Area under introduced forages.

Species	Area (ares)	% of total
Tripsacum	1378.96	90.09
Pennisetum	79.94	5.22
Setaria	71.82	4.69
TOTAL	1530 74	100.00

Note: Are is an area of 100 m²

To estimate the productivity of introduced forage species, a survey was made on-farm. The preliminary results on Tripsacum (Table 11) showed that one hectare of this forage species yielded more than 64 tonnes of green matter which was estimated at about 9.6 tonnes DM/ha. Calculations indicate that this quantity would be sufficient to feed, over a period of one year, a herd consisting of one cow weighing 370 kg, one heifer weighing 220 kg and a calf weighing 60 kg. Coincidentally, this herd composition corresponds to that owned by the majority of the participating farmers.

Table 11. Green matter yield of Tripsacum laxum (t/ha).

Number of cuts		Age at forage cut (months)	Mean (t/ha)
10	First	12	32.55
10	Second	18	31.96
Total			64.51

Traditional milk technology

Traditionally, Burundese farmers milk cows in the morning and in the evening. The morning's collection is sold through the markets while the evening milk is bulked in batches for three to four days. The bulked milk sours which makes it ideal for churning. A survey showed that churning is done in 10 to 15 litre containers for 45 minutes, sufficient time for effective separation of skimmed milk from butter. The extraction rate is about 4.9% (Table 12).

Table 12. Characteristics of traditional milk chuming practices.

Number of Observations	Mean milk quantity (l)	Churning time (minutes)	Butter yield (g)
13	8.8	46	331

During the course of the project, samples of cow milk from peasant farms were taken and analysed. The results show that the fresh milk was contaminated with microbes. The presence of E. coli, salmonella and putrid microbes was detected. The next phase of the project will look into the improvement of hygienic production and handling of the milk.

Lessons learned

Given that the Sahiwal cow is originally a subtropical breed (Pakistan), under normal circumstances it can adapt to the Burundese environment without significant difficulties. The research confirmed that it is possible to keep improved Sahiwal and Ankole crosses on-farm and that the farmers accept the technologies required to create a favourable environment for the improved animal.

The package of improved technologies for the Sahiwal cattle (feed production, simple shed and harnessing of manure) appeared to be appropriate for the smallholders as they accepted and adopted the technologies. Probably the main incentives were the probability of increased availability of milk for sale and home consumption and improved crop production enhanced by farm-yard manure. A comparison of production parameters such as milk yield, liveweight and forage production showed that the farmers were more productive than the government farms. This performance was a pointer to the keenness of farmers. A tentative economic analysis was made; the gross income from a one cow unit was estimated at FBU 182,900 (Table 13) which is equivalent to US$ 750. This is above the national per capita income of US$ 250. Significant income was generated through increased potato production after the application of manure.

Table 13. Estimated income of livestock products in the peasant farm.

	Yearly yield	Value in FBU
Potatoes (response to manure)(kg)	2990	104,650
Milk(kg)	1055	52,750
Steer (kg)	170	25,500
Total gross income		182,900

Although farmers produced forage feeds on-farm, the herbage did not provide sufficient quantities of metabolisable energy and protein and consequently some supplementary oil cakes had to be given.

To assess the profitability of the improved technologies package an economic analysis is expected to be conducted in 1994-95, when sufficient data will be available. Given the income-generating potential of the scheme peasants should be in a position to ask for bank credit to improve housing, watering and other farm facilities. Fortunately, in Burundi there is a bank which provides this type of credit.

Besides the improved income, participating farmers also had access to protein rich milk. However, a specific study on the benefits to human nutrition of the increased availability of dairy by-products has not been made. What cannot be disputed, and is evident, are the social benefits and subsequent improvement of household status. The demand for milk will continue to be greater than supply in the foreseeable future (Table 14). Paradoxically, even while the farmers see the opportunity and are keen to participate in the project, there are insufficient numbers of Ankole and Sahiwal crosses to be distributed.

Table 14. National needs of milk and meat (tonnes).

	Needs	Available	Deficit
Milk	275,000	18,500	258,500
Meat	55,000	15,150	39,850

East Coast Fever is the most important disease in Burundi; fortunately there are sufficient supplies of acaricides and other drugs to control it.

Impact of research programme

Lessons have been learned from the project and in response the government has developed an integrated livestock-crop agriculture pilot project in Ngozi and Gitega provinces, both of which are heavily populated.

Future research programmes

Future research programmes will continue to monitor on-farm trials in the two ecological zones of the highlands and lowlands and a detailed economic analysis will be made at the end of the project.

References

Département de la production animale 1989 Rapport Annuel 1988-1989 Direction Générale de L'élevage, Burundi.

Département de la santé animale. 1989. Rapport Annuel 1988-1989. Direction génerale de l'élevage, Burundi.

Engels Dirk E. 1982. Plan de dévelopment d'élevage au Burundi pp 35.

ISABU (Institut des sciences agronomiques du Burundi).1987. Programme Système de Production du Bétail. Rapport Annuel. Bujumbura, Burundi.

ISABU (Institut des sciences agronomiques du Burundi). 1988. Programme Système de Production du Bétail. Rapport Annuel. Bujumbura, Burundi.

ISABU (Institut des sciences agronomiques du Burundi). 1989. Programme Système de Production du Bétail. Rapport Annuel. Bujumbura, Burundi.

ISABU (Institut des sciences agronomiques du Burundi). 1990. Programme Système de Production du Bétail. Rapport Annuel. Bujumbura, Burundi.

ISABU (Institut des sciences agronomiques du Burundi). 1991. Programme Système de Production du Bétail. Rapport Annuel. Bujumbura, Burundi.

ISABU (Institut des sciences agronomiques du Burundi). 1981. Rapport Annuel. Butare.