Table of Contents Next Page


In general, livestock in tropical regions are less productive than those in temperate parts of the globe. They are smaller and have lower yields of milk, wool or meat. The question is: Is this difference inevitable because of the climatic differences, or can it be overcome by changing the animals or by changing the way they are cared for? No simple answer to this question is possible. Introduction of temperate genotypes into the tropics has not usually been successful - the animals either do not survive, or do not breed, or their performance is much poorer than in their country of origin. Tropical animals increase in size and performance when well fed and protected from disease but their output does not reach that of the temperate genotypes. It is clear that a more elaborate approach is needed - a change in both genotype and environment simultaneously or a selection within the local breed under the local environmental conditions. This manual aims to examine some of the breeding programmes which can be applied to achieve the necessary genetic changes. It will not consider the parallel changes in feeding, management and disease control which are a necessary concomitant of some of the programmes discussed.

This introductory chapter will examine how the high performance of temperate breeds, or at least some temperate breeds, has been achieved, in order to consider what lessons can be learned which may be applicable in the tropics.


The selection of breeds for production must have begun very soon after the first domestication of their wild ancestors in the Middle East some 8 000–10 000 years ago. The high yield of the dairy cow, the fleece of the wooled sheep and the Angora goat, the high growth rate of the beef cattle breeds are all the result of selection. Cattle were used for draught long before they were milked. This means that milk was originally only a by-product (the major dairy animals were sheep and goats) and any selection for milk yield could only be subsidiary to their major role as working animals. If Europe had been compared with the tropics in the Middle Ages the differences in the function and productivity of their livestock would not have been nearly so striking as today.

Indeed it is only during the last 200 years starting in western Europe that specialized and high producing breeds of dairy and beef cattle have been formed. Before that time the typical cattle were triple-purpose - draught, milk and meat - and each locality had its own variety. Likewise sheep had to produce wool and meat and often milk as well. Each region and even each mountain valley had its own local breed.

At the end of the 18th century there was a concatenation of circumstances which stimulated the development of livestock breeding and the production of specialized dairy breeds in particular. The ox was giving way to the horse as a draught animal leaving the cow free to become a specialized milk producer. The agricultural revolution was making available improved cattle feed supplies and particularly improved forage for the winter - turnips and hay, for instance, instead of straw. The increasing city markets gave an incentive for higher milk production and more meat production. Thus serious selection for milk yield (and also for meat and wool) began and was successful in several different areas of western Europe where the improvers were fortunate in having available a lucky combination of genes so that they were able to make considerable progress.

Similar developments took place in sheep. The Merino was early selected in Spain from a foundation of finewool sheep which had their origin much further east. In Britain two mutton-wool breeds were produced by selecting local stock - the short-wooled Southdown and the longwooled Leicester.

Once these improved breeds had been formed they naturally attracted attention. The Merino was exported all over Europe and has left a permanent mark on the sheep of eastern and southern Europe. Many of the local breeds have now been graded to Merinos or new breeds have been formed on a Merino × local foundation. The Merino was also exported to South Africa, South America and Australia. It had its greatest success in the last country and by the third quarter of the 19th century it had already developed into a number of strains.

The Southdown and Leicester were used as foundation stock to improve (or to produce) all the shortwool and longwool breeds. These in turn were used in many countries of Europe and elsewhere to form new breeds by crossbreeding.

This is now history. It is only since performance recording began in earnest that the actual increases in productivity can be estimated. Figure 1. 1 shows the increase in milk yield per cow in Sweden since the start of milk recording at the beginning of this century and Figure 1.2 shows the increases in the four Scandinavian countries since the application of AI and progeny testing. The higher rates of increase in Finland and Norway may be due to greater breed replacement.

These changes can best be interpreted as due to the simultaneous improvement in the genetic potential and in the realization of this potential by adequate feeding, management and disease control. We will consider here only the genetic changes.

Figure 1.1 Average milk production of recorded cows in Sweden ( 1900–1980 ) (Drawn from figures in SHS, Eskilstuna. 1981. pp. 23–25)

Figure 1.1

Figure 1.2 Average milk production of recorded cows in Denmark, Finland, Sweden and Norway (1950–1972) (Courtesy Dr. J. Rendel)

Figure 1.2

1.1.1 Improvement by breed replacement (grading up)

This is well exemplified in the history of cattle breeds. The first improved breed, the Shorthorn, spread rapidly in Britain and in 1939 it accounted for about 65 percent of all bulls. The breed was also exported to Europe and was influential especially in forming the dairy breeds of the Scandinavian countries. But its early success did not last. The Friesian, developed in the Netherlands, has a larger size and a greater milk yield than the Shorthorn and has largely displaced it in Britain where over 60 percent of inseminations now come from that breed compared with less than 1 percent for Shorthorns. Average milk production per cow has risen accordingly. The Friesian also spread all over western Europe and is rapidly displacing the local triple-purpose breeds.

In central and eastern Europe it is the improved dual-purpose breeds of Switzerland which have replaced most of the local breeds. But the Friesian is now coming in because of its higher milk yield. However, the European Friesian is under pressure from the American Holstein-Friesian which, since it was exported to USA, has been subjected to a much more intensive selection for yield.

Corresponding changes of breed have been responsible for increasing milk yield in the New Zealand cattle herd. The early Shorthorns were displaced by Jerseys because of the export butter market so that by 1947 about 85 percent of New Zealand dairy cows belonged to this breed. With the world butter surplus a change was made to the Friesian which now accounts for 56 percent of inseminations. The first to be imported was the British Friesian and now more and more American blood is being incorporated.

However, breed replacement - especially by the Friesian - is now a cause for concern in many countries because of the reduction in genetic diversity which this replacement represents.

1.1.2 Improvement within breeds

Increase in the average yield of a country's herd by breed replacement is only possible if there are suitable breeds to use for the purpose. These arose through selection by the breeders themselves. The process continues and has been refined scientifically since the development of quantitative genetic theory in the 1920s and its application to livestock breeding beginning in the 1930s. Calculations of heritability have ensured that effort is not wasted on selecting traits which are not heritable. The accuracy of selection can be improved by correction for environmental variables and by the use of information from close relatives. The importance of the reproductive rate in increasing the intensity of selection has been recognized. For selection for several characters selection indices can be constructed based on the heritability and genetic variation of the characters, the genetic correlation between them and their relative economic values.

These techniques could first be applied to dairy cattle because the presence of milk recording and artificial insemination made available large numbers of animals to include in the selection scheme. All the countries of western Europe, North America and New Zealand, and several others, now operate sophisticated selection programmes based on the progeny testing of bulls on the milk yield of their daughters. Progeny testing is necessary for milk yield because this characteristic cannot be measured on the bull himself. The use of AI means that many bulls can be tested simultaneously and the tests can be based on a large number of daughters. Thus the needs of accuracy of test and a high selection differential can both be met.

AI is also essential to spread the daughters of each bull over several herds. If each bull has daughters in only one herd then the environmental differences between herds (feeding and management levels) will have more influence on the difference between progeny groups than the genetic differences between the bulls themselves. A nation-wide recording scheme is needed to furnish the milk records on which the progeny testing is based.

Using such methods it is theoretically possible to increase milk yield at the rate of 2 percent per year but this assumes that all the selection pressure is devoted to milk yield and none is wasted on fancy points, e.g. horn shape, colour and markings. However, there are other considerations in a selection programme - milk composition, fertility, calf viability, conformation of udder and legs, milk let-down (“milkability”) and freedom from disease. So the optimum rate of improvement for milk alone can never be reached. Nevertheless there have been steady improvements in the yield of the major breeds in the countries operating these programmes.

For beef cattle selection is easier, since the important characteristics - rate and efficiency of growth - can be measured in both sexes before breeding age. The use of bull performance testing stations where bulls from different herds can be tested for growth rate under standard conditions has made possible rapid increases in the size and growth rate of the major beef breeds. There has been less attention to efficiency.

In sheep the outstanding example of improvement in performance has been the increase in fleece weight of the Australian Merino (see Figure 1.3). Official overall figures of greasy fleece weight per head for sheep plus lambs in New South Wales are available back to 1879–80. These show an increase of 0.030 kg per year up to about 1930–31. Not all of the increase is genetic but it would represent an upper limit to genetic gain. Some of the genetic gain would be due to slight infusion of genes from British breeds, some to selection, the latter being based mainly on visual appraisal. From 1930–31 to 1950–51 there was a marked slackening off with little change in fleece weight. From 1950–51 to 1979–80 fleece weight has again been increasing, at the rate of 0.016 kg per head per year. Some of this increase may be due to the introduction of fleece measurement in recent years, but if measurement were fully utilized the rate of gain could be at least doubled.

Figure 1.3 Average greasy wool weight - sheep + lambs NSW, Australia. Official figures, Australian Bureau of Statistics.
( Courtesy - Dr. Helen Newton Turner )

Figure 1.3


The history of breeding in Europe can be summarized as consisting of periods of pure breeding in populations isolated geographically or by breed society barriers, alternating with periods of crossbreeding - usually with the breeds which have operated the most successful selection programmes. Most of this importation has been between countries of Europe or between these and other temperate countries overseas. So problems of climatic adaptation have not arisen. Also the new breeds have only been used as local feed resources increased so that it was possible to satisfy the exigencies of the higher producing animals.

Where climatic differences were involved - for instance different altitudes - a stratified crossbreeding system was sometimes developed with the hardy sheep or cattle on the hills being crossed with better meat breeds for producing progeny to fatten in the lowlands. A similar stratified scheme can be based on dairy cows as a foundation; those not being needed for breeding replacements can be crossed with beef bulls.

1.2.1 Constraints on production in the tropics

To what extent can these techniques be used in the tropics?

It is not possible to transfer the improved temperate breeds directly to the tropics because of climatic problems. There may be highland areas where temperate breeds can adapt but there may be other constraints - lack of feed resources, lack of veterinary services to control the endemic diseases. If these difficulties can be overcome then such favoured areas may be ideal for maintaining nucleus herds of improved temperate breeds.

Where the climatic stress is unabated the improved breeds may not be used but the technology of improvement can. The evidence suggests that the genetic parameters are similar in tropical and temperate environments so that the same techniques of selection and crossing should be applicable to both. The chief lack is the infrastructure. Improvement is possible by selecting on eye appraisal within a small herd but it is extremely slow. For effective change accurate recording, large herds and/or cooperation between small herds, and for dairy cattle, an AI system, are needed. The infrastructure should also include a central nucleus herd for breeding superior males, a well trained extension service and some system to monitor progress.

The question of feed resources is crucial. It is true that the improved breeds need more feed but the justification is that feed requirement per unit of product is less than in the low producing breeds. In general breeding programmes should be designed to use the existing resources in the most efficient way. Tropical developing countries often export the valuable ingredients of the concentrate feeds used for milk production in western countries. With more efficient tropical breeds these ingredients might be better used at home.


This manual can discuss general principles in the choice of a breeding policy. It cannot give recipes for all countries or regions. Each country must decide what is the best breeding policy according to ecological conditions (climate, altitude, vegetation), feed resources (pastures and available by-products), management systems and possibilities of disease control.

An example of the sort of plan which is needed is given in Table 1.1 which shows a policy for Kenya worked out by Meyn and Wilkins (1974).

Table 1.1


Ecological zone1ClimateApprox. annual rainfallPercentage of surface areaCattle populationStocking rate2Breeding policy
Subsistence smallholdersCommercial, large-scale
  Millimetres MillionsStock units/ha  
I to IIIAfro-alpine equatorial, humid to semi-arid900213.72Dairying3European breeds(i) Dairying3
(ii) Dual-purpose European breeds4
IVSemi-arid600–90092.53–4Subsistence beef5

European x zebu crosses
(i) Dairying5Ranching6

(ii) Beef
VArid400–600501.94Subsistence: beef-SahiwalBeef-Boran
VIVery aridunder 400201.015Subsistence: beef-indigenous zebu-

1 From Pratt et al., 1966.
2 Stock units, equivalent to 450 kg liveweight.
3 Ayrshire, Friesian, Guernsey, Jersey.
4 Simmental, Brown Swiss, Red Poll.
5 Crisscrossing: Friesian, Simmental, Brown Swiss, Red Poll, with Sahiwal and Boran.
6 Commercial crosses: Charolais, Simmental, South Devon, Sussex, Hereford, Friesian, with Boran and Sahiwal.

Source: Meyn and Wilkins (1974)


In Europe the spread of specialized breeds has led to the loss of local generalized breeds. These (especially beef cattle and sheep) may have important adaptive characters. Their loss may mean that whole areas of marginal land (especially uplands, dry areas and lands of low fertility) may go out of production. Also these lost breeds may have unique genes which cannot easily be replaced if at some future time changed conditions require their return. This applies even more to tropical countries. The local breeds constitute an irreplaceable stock of adapted germ plasm and should be conserved for both present and future use.

Clearly the best way to conserve a breed is to continue to use it commercially. This is certainly the way that most tropical breeds will be conserved because of their adaptation. It is most important that the utility of the local breeds should be demonstrated by comparing them with temperate breeds (in overall productive efficiency, not merely in short-term milk yield or growth rate) so that they are not decimated by premature programmes of crossbreeding and replacement. If breeds are threatened then the government may have to offer financial inducements for their retention. If the worst comes to the worst and breeds are reduced to relict proportions, which has happened to so many in Europe, then it may be necessary to form special herds in order to retain the last representatives on account of their historic, cultural, scientific or possible future economic interest.

Top of Page Next Page