Previous Page Table of Contents Next Page

Crossbreeding for milk production in Sri Lanka

by V. Buvanendran and P. Mahadevan

The indigenous cattle of Sri Lanka, the Sinhala, in common with most other cattle of the tropics, are poor dairy animals. They are small, with an average adult weight of 160 kg. Their mean 305-day lactation milk yields are around 450 kg. This level of production, while it may satisfy the needs of subsistence farming, is too low for commercial dairying. It is therefore necessary to improve the production potential of the cattle in order to make them profitable.

The climatic conditions under which cattle are raised in Sri Lanka vary widely. The agroclimatic zones may be broadly classified into: dry zone, coconut triangle, midcountry and hill country. The ecology of these areas is shown in Table 1. About two thirds of the cattle are in the dry zone, 25 percent in the coconut triangle, and the remaining 8 to 10 percent in the midcountry and hill country areas where dairy managerial skills are well developed and high levels of concentrate feeding are adopted under systems of intensive husbandry. The coconut triangle, although hot and humid, is also regarded as an area of high dairy potential, partly because dairy farming has been a traditional enterprise of the farmers in this zone and partly because of the success in cultivating pasture grasses such as Brachiaria miliformis and Brachiaria brizantha as inter-crops. The coconut palms also provide a certain amount of shade and offer some measure of protection from solar radiation. In the dry zone the cattle forage on poor natural pastures with virtually no inputs of concentrate feed; managerial skills are limited, and milk is regarded as a by-product of animals that are raised mainly for beef.

The Government has large cattle farms of European breeds—primarily the Ayrshire, Jersey and Friesian — in the hill country areas that have a high dairy potential. The success that has been achieved during the last two to three decades with purebred herds of these breeds and with the upgrading of local cattle shows that high levels of European blood are not inimical to high milk production in this zone. The wellknown “Hatton” or “Cape” cattle found in the hill country are believed to be descendants of European cattle brought to Sri Lanka by the early Dutch settlers some 300 years ago. They acquired a certain amount of indigenous blood by crossing with the local cattle and were developed without any proper breeding policy or plan. The dairy merit of these animals lends support to the present policy of upgrading all the indigenous cattle in the hill country to European breeds.

V. Buvanendran is Animal Geneticist at the Veterinary Research Institute, Peradeniya, Sri Lanka; P. MAHADEVAN is Animal Production Officer (Research and Education), Animal Production and Health Division, FAO, Rome, Italy.

Sinhala heifer at Karagoda-uyangoda

In the coconut triangle and the dry zone, climatic conditions are unfavourable for the adoption of a straightforward upgrading programme. The dry zone has the further disadvantage of poor nutrition, because forage in terms of both quantity and quality is limited mainly to the rainy season. Crossbreeding research has therefore been directed at evolving breeding policies that would be suitable for these two zones. The present article discusses the results of research carried out over the last 20 years and their possible application to national programmes. The experiments were carried out at the government livestock stations at Karagoda-Uyangoda, Undugoda and Wirawila (see Figure 1).

Karagoda-Uyangoda experiment

Karagoda-Uyangoda is located in southern Sri Lanka, at an elevation of 30 metres above sea level with a mean annual temperature of 26.7°C and a relative humidity of 78 percent. Annual rainfall averages 2 500 mm. Crossbreeding work at the government livestock station was initiated by Mahadevan (1953). The experimental design is shown diagrammatically in Figure 2. Sinhala cattle were crossed with either Jersey or Friesian sires. The F1 generation was bred inter se to produce F2 generations of both breed crosses. In the Jersey X Sinhala crossbreeding programme, part of the F1 population was backcrossed to the Jersey to produce a B1 generation with 75 percent Jersey blood. In the Friesian programme, the F2 animals were carried through to the third filial generation by inter-se mating among the F2. The results obtained during the periods 1956-66 (Wijeratne, 1970) and 1968-73 (Buvanendran, 1975, unpublished) are summarized separately in Table 2.

It will be observed that F1 animals from both the Jersey x Sinhala and the Friesian X Sinhala matings showed a remarkable increase in milk yield over the indigenous Sinhala cattle, but the F2 animals exhibited a marked decline in yield from F1 levels. Further interbreeding to the F3 stage in the Friesian programme did not produce any significant change in performance from the F2. Comparison of F1 performance in the Jersey programme with contemporary F2 and B1 production levels respectively suggests that although backcrossing to the European parental breed gave relatively better results than the inter-se mating of F1 animals, the levels of production attained in the B1 were nevertheless considerably lower than in the F1.

FIGURE 1. Agroclimatic zones of Sri Lanka


NOTE: The “other areas” marked on the map have a hot, wet climate, and do not fall into the broad classification of agroclimatic zones (see also Table 1).

Table 1. Agroclimatic zones of Sri Lanka

ZoneElevationMean monthly temperaturesRainfallHumidity
Dry zone023.8–32.2890–1 90075–90
Coconut triangle0–45723.8–29.41 900–5 08062–80
Midcountry305–91418.3–23.81 900–5 08065–75
Hill countryAbove 91410.0–23.82 160–3 17558–75

These results suggest that hybrid vigour is responsible for part of the higher yields in the F1 and for the subsequent depression in the F2 and B1. The F2 and B1 animals lose 50 percent of the hybrid vigour present in the F1, but the relative superiority of the B1, animals over the F2 may be attributed to the additional amount of Jersey blood. The data on calving intervals suggest that the additional Jersey inheritance in the B1 did not lower reproductive potential.

Undugoda experiment

This farm is located in the midcountry zone at 180 metres above sea level. The average annual rainfall is 2 500 mm. Mean monthly temperatures vary from 23.0 to 31.1°C and mean relative humidity is 71 percent. The cattle are housed in sheds and zero grazed on Brachiaria brizantha pastures. Red Sindhi cows were mated to Jersey sires to produce F1 animals. F2, B1 and 5/8 Jersey offspring were also produced by mating F1 animals inter se, backcrossing to Jersey or mating F1 to B1 animals. A purebred herd of Ayrshire was maintained contemporaneously and this served to provide a control population during the major part of the experimental period.

Data on milk yield and other characteristics of economic importance are presented in Table 3 (Buvanendran, 1974).

The results are essentially similar to those of the Karagoda-Uyangoda experiment. The yields of the B1 animals were lower than those of the F1 but closer to the latter than to the F2. The difference between the Karagoda-Uyangoda and Undugoda experiments is probably due to the favourable climatic and managerial conditions at Undugoda that permitted a fuller expression of the Jersey genotype in these animals. It may also be noted that the performance of the Jersey F1 in this experiment is similar to that of the purebred Ayrshire. Reproductive performance as measured by the length of the calving interval varied with the different crosses, but there was a general tendency for calving intervals to increase with an increase in Bos taurus blood.

The B1 and F2 animals in this experiment were born and commenced lactation at about the same time, so that the treatments accorded to them were similar. The differences in performance may therefore be regarded as largely genetic.

Wirawila experiment

This government farm is located in the typical dry zone of Sri Lanka. Mean annual temperature is 27.1°C and average precipitation 1 075 mm. Breeding work at this station commenced with Sinhala animals, but these were first mated to Red Sindhi sires to produce females with varying levels of Red Sindhi blood (ranging from 1/2 to 7/8), and these were then crossed to either Shorthorn or Jersey sires.

The production characteristics of Red Sindhi X Sinhala crosses obtained during the period 1961–69, and of Shorthorn and Jersey crosses with the grade Sindhis during 1970–73, are summarized in Table 4 (Buvanendran and Tilakaratne, 1974, unpublished data). The contemporary records of Sinhala cattle are not available, because these animals had been disposed of at the time that the F1 and higher grades had come into milk. Nevertheless, the results are indicative of the increase in yield obtainable by crossing the Sinhala with the Red Sindhi.

Figure 2. Design of experiment at Karagoda-Uyangoda

Figure 2

The yields of the F1 are close to the expected mean of the two parental breeds (if it is assumed that the yield of the Sinhala parents at Wirawila would have been similar to that obtained at Karagoda-Uyangoda during the same period). This suggests a simple additive genetic effect on milk yields when two zebu breeds are crossed. Any heterotic effect is probably of a small order. Further upgrading to the Red Sindhi did not result in any further increase in yield. However, the introduction of European blood resulted in a marked improvement in milk production, the crossbred animals giving nearly 50 percent higher yields than the Red Sindhi. The reproductive performance of the European crosses was also superior to that of the Red Sindhi.

It is relevant to note that the performance of the Jersey crosses at Wirawila is higher than that of the Jersey crosses in the Karagoda-Uyangoda experiment during the same period, even though climatic and managerial conditions were less favourable at Wirawila. This is probably due to the superiority of the foundation zebu stock (Sindhi × Sinhala crosses) used at Wirawila. The Undugoda results also lend support to this interpretation.

General conclusions

On the basis of the results of the three experiments outlined above, the following general conclusions may be drawn:

  1. Milk production from F1 European x zebu crosses is markedly superior to that obtained from the base zebu population. Progeny from Friesian matings to the zebu produce more milk than those from corresponding Jersey matings.

  2. Interbreeding of the F1 progeny results in a marked decline in milk yield, the reduction in yield at the F2 level averaging 30-40 percent. Further interbreeding to the F3 generation does not lead to any further significant decline. These results point to an appreciable degree of hybrid vigour in the F1 and its loss in the F2 and F3 generations.

  3. Backcrossing the F1 to the European breed results in performance levels that vary considerably with climatic and managerial conditions. Under relatively mild climatic conditions and with minimal exposure to heat stress (as at Undugoda under zero grazing), B1 animals can produce at a level fairly close to the F1, whereas under harsher climatic conditions (as at Karagoda-Uyangoda), B1 production would be lower than F1, although relatively higher than F2.

  4. Inasmuch as the genetic merit of the European parent has an influence on the performance of the F1, the genetic merit of the zebu is also important, as evidenced by the results from Undugoda and Wirawila, where purebred Sindhis or Sindhi × Sinhala crosses were used as the foundation zebu stock.

Table 2. Production characteristics of different breeds at Karagoda-Uyangoda

Milk yieldLactation lengthCalving intervalMilk yieldLactation lengthCalving interval
Friesian × Sinhala:      
1 5733273931 482387442
Jersey ×Sinhala:      
1 2153133701 076377453

Sources: 1956–66 data, Wijeratne (1970); 1968–73 data, Buvanendran (1975, unpublished).
Note: Figures in brackets give number in mean for less than 50 observations.

Table 3. Production characteristics of different breeds at Undugoda, 1957–71

BreedMilk yieldLactationl engthCalvingi ntervalAge at first calving
Jersey × Sindhi:    
F11 92929536833.7
F21 115(49)265(49)430(49)33.0(49)
B11 70031743439.6

Source: Buvanendran (1974).
Note: Figures in brackets give number in mean for less than 50 observations.

Table 4. Production characteristics of different breeds at Wirawila

Milk yieldLactation lengthCalving intervalMilk yieldLactation lengthCalving interval
Sindhi x Sinhala:      
Jersey x grade Sindhi, F1---1 209(37)303(37)389(37)
Shorthorn x grade Sindhi. F1---1 320287397

Source: Data collected by Buvanendran and Tilakaratne (1974, unpublished).
Note: Figures in brackets give number in mean for less than 50 observations.

Jersey × Sinhala crossbred cow at Karagoda-Uyangoda

National breeding programmes

In deciding on a national breeding programme to improve the indigenous zebu stock by the introduction of European blood, two options are possible — either new breed formation or the adoption of a crisscrossing or rotational crossbreeding system. Where new breed formation is desired, it is necessary to decide early on in the programme the European breed that is to be used for crossing with the indigenous zebus, and the level of European blood that should desirably be incorporated into the new breed. Where the evidence suggests that a system of rotational crossbreeding using two or more breeds is likely to be more fruitful than new breed formation, the decision concerns the number and choice of breeds to be used in the rotation.

For new breed formation, the choice of European breed should be made with the aim of maximizing yield in the F1, so that despite the loss of heterosis in the F2 and B1 generations high yields would still be possible when the level of European blood that should desirably be incorporated into the new breed has been reached and inter-se mating takes place. From the experimental results obtained in Sri Lanka, it would appear that the Friesian would be more suitable than the Jersey as the European parent for areas, such as the coconut triangle, which have a high dairy potential notwithstanding the tropical climate. Prevailing meat prices could also influence the choice of breed.

Cows on pasture at Karagoda-Uyangoda

For the dry zone, however, the Jersey has distinct advantages because of environmental and managerial considerations. It would also seem that using an improved zebu such as a grade Sahiwal or Sindhi as the foundation stock would pay dividends in the long run by ensuring higher performance in the F1 and subsequent generations. This may be regarded as impractical on the grounds that it would delay the introduction of European blood into the programme by one generation. Nevertheless there are areas, such as the dry zone of Sri Lanka, where an immediate introduction of European blood is not feasible due to the prevailing poor standards of nutrition and management. In such areas the slower, two-stage process of improvement would be more suitable.

A similar conclusion was reached by Mahadevan et al. (1962) in their genetic study of the Sahiwal grading-up scheme in Kenya: they stated that the Sahiwal breed in Kenya provided a satisfactory first stage for grading up the East African zebu to a more productive level in areas of medium agricultural potential: the second stage would then consist of one or perhaps two introductions of European blood into the Sahiwal grade stock, depending on the level of management attained by the stock owners at the end of the first stage and on the general climatic conditions of the area. The decision on the level of European blood that should desirably be incorporated in any national crossbreeding programme in Sri Lanka would be governed by similar considerations.

The lessons learned from new breed development in Jamaica (Wellington and Mahadevan, in this issue), concerning the need to ensure the participation of a large number of milkrecorded herds and their full involvement in sire testing, are particularly relevant to Sri Lanka where both milk recording and sire evaluation are largely confined to governmentowned herds. Jamaican experience has demonstrated that a relatively small nucleus of animals of a new breed, developed within one or perhaps a few government herds, is unlikely to make any significant impacton milk production in a time span of 20 to 25 years without the cooperative efforts of a number of participating farmers.

Rotational crossbreeding

The second option for a national breeding programme for the coconut triangle and the dry zone of Sri Lanka — crisscrossing or rotational crossbreeding, using zebu and European breeds — appears to be better according to the experimental evidence (although animals with two thirds zebu blood are likely to be unacceptable to many milk producers — see below). This conclusion is based on two principal considerations. The first is the large decline in performance from the F1 to the F2 generation in both the, KaragodaUyangoda and Undugoda experiments, which indicates the presence of a high degree of hybrid vigour. The second is the relative superiority of the B1 animals over the F2 although the magnitude of this superiority varies with climatic and managerial conditions.

Two practical questions would need to be answered in designing and implementing a rotational crossbreeding programme, namely, the number of breeds that should be used in the rotation and the choice of breeds. A two-breed rotation involving a zebu and a European breed has the advantage that it is simple to operate, and when equilibrium is reached it should be possible to distinguish the cows with two thirds European and one third zebu genes from those with one third European and two thirds zebu genes; this would facilitate the choice of sires that should be used on each type of cow. Furthermore, there is no experimental evidence to suggest that three-breed combinations will offer greater advantages than twobreed systems. But there is one major disadvantage in a two-breed rotation using widely different parental strains. This arises from the fact that there would be two kinds of animals in the population at all times, and since their production levels are likely to be substantially different it may be psychologically and commercially difficult for producers to continue with a two-breed rotational crossbreeding system. It was partly in this context that a three-breed system involving one zebu and two European breeds was recommended by Mahadevan (1970). For a two-breed system, the breeds most suitable for the coconut triangle of Sri Lanka would be the Friesian and the Sinhala, with the proviso that if the Sinhala could be upgraded to an improved zebu before being incorporated in the rotational crossbreeding programme substantially higher yields are likely in the long run. Under dry zone conditions, the choice of the Jersey as the European breed component of the rotation has much to commend it because of the harsher environmental conditions. For the three-breed system, an improved zebu such as the Sahiwal × Sinhala or Sindhi × Sinhala cross may be used, together with the Friesian and the Jersey.

Top picture: Jersey × Sindhi crossbred cow (right) and purebred Sindhi cow at Wirawila. Lower picture: Sinhala bull at Karagoda-Uyangoda

Whatever rotational crossbreeding programme is finally adopted, it would have the advantage that the whole of the national herd in a given zone could become involved in the programme even in the absence of a milk recording system. The role of the government livestock stations would then be to raise nucleus herds providing superior sires of the chosen breeds for crossing, rather than serve as centres for the multiplication of breeding stock for distribution to farmers. The importation of improved sires or their semen from other countries would serve the purpose of maintaining the quality of the nucleus herds and increasing their contribution to the national breeding programme.


Buvanendran, V. 1974. Crossbreeding experiments of Jersey with Indian zebu breeds. Brief Communications. Vol. JE. XIX International Dairy Congress, India: p. 55.

Mahadevan, P. 1953. Crossbreeding experiments with Sinhala cattle at Karagoda-Uyangoda. Trop. Agriculturist (Ceylon), 109:123-127.

Mahadevan, P. 1970. A note on the maintenance of heterosis in crossbred cattle in the coconut triangle of Ceylon. Ceylon Cocon. Quart., 21:92–93.

Mahadevan, P., Galukande, E.B. & Black, J.C. 1962. A genetic study of the Sahiwal grading-up scheme in Kenya. Anim. Prod. 4:337–342.

Wijeratne, W.V.S. 1970. Crossbreeding Sinhala cattle with Jersey and Friesian in Ceylon. Anim. Prod., 12:473–483.

Previous Page Top of Page Next Page