IMPROVEMENT OF PIGS IN THE TROPICS:
GENERAL PRINCIPLES

J.W.B. King 1/

1. INTRODUCTION

The principles for the improvement of pigs do not differ around the world but the way in which they are implemented will depend on selection objectives and on the existing population of pigs and the way in which it is to be utilized for production to meet particular markets. Overlying all these considerations are likely to be requirements for adaptation to local climatic conditions and to diseases, and endo- and ecto-parasites. Much judgement will therefore be required in specifying programmes appropriate to particular circumstances and the purpose of the following review is to discuss some of the issues which arise.

2. NATURE OF THE INDIGENOUS PIG POPULATION

Populations of indigenous pigs will undoubtedly vary in their history and breeding structure. At the one extreme there may be large, ill-defined populations made up of individuals without any of the uniformity usually ascribed to breeds, but nevertheless filling a particular niche and performing a particular function. It is unlikely that more than fragmentary knowledge exists about the pedigree of individual animals and yet the population as a whole probably represents a valuable asset to the animal agriculture of the region. At the other extreme there may well be small pedigreed populations that have been carefully husbanded for several generations, recorded in detail and about which a great deal more is likely to be known. When both kinds of population are present in the same region, problems may well arise in defining distinctness and deciding whether performance levels between a small pedigreed nucleus and the population at large are genetic or merely environmental. If trials to establish this are needed, the general experimental principles described by Sellier (1980) are useful for planning appropriate designs.

3. USE OF THE INDIGENOUS POPULATION

The advantages of crossbreeding in the pig appear to be so great that it is most likely that the genetic resource of the indigenous breed would not be utilized in some form of crossbreeding. Although critical evidence is scanty, the general principle is that the amount of heterosis observed is usually greater in unfavourable environments, thus giving a further incentive to some crossbreeding system for use in the tropics. To organize the crossbreeding that is required may present some organizational difficulties and at first crossbreeding may take the very simple form of crossing the females of the indigenous breed with males of an exotic breed as a means of producing pigs for meat production. This first crossing method does not use heterosis in the sow so tht extending the crossing system to use first cross sows for breeding will probably be advantageous providing adaptability to the local environment is not lost. In the interests of simplicity, a backcross to boars of the parental exotic breed is the simplest method of using a crossbred sow and avoiding the necessity for having a third breed.

Because of organizational difficulties, attempts are frequently made to perpetuate the crossbred population from inter se matings, thus producing a synthetic population (or eventually a new breed) which may be multiplied for use in that environment. Although attractive from the administrative viewpoint, the serious loss of heterosis may well mitigate against the system, as well as requiring the use of breeding females not well adapted to the local conditions. Ingenuity in devising ways in which a discontinuous crossing system can be implemented may therefore be rewarding.

In later discussion, it is assumed that the indigenous breed will in fact be used as the mother of slaughter pigs, or as a contributor to a crossbred sow. The improvement of maternal performance is therefore of paramount importance in the indigenous population.

4. SELECTION OBJECTIVES

The potential number of characteristics in which genetic changes might be desired is great and some grouping and simplification may help the task.

4.1 Female Reproduction

-	Piglets weaned per litter (as a convenient integration of numbers born and viability).
-	Weaning weight of the litter (particularly when weaned piglets are sold from one producer to another).
-	Re-breeding interval (leading to a measure of piglets per sow per annum).
-	Sow feed costs.
-	Sow carcass value.

4.2 Male Reproduction

Genetic changes may not be necessary but possibly the ability to produce fertile sperm at high ambient temperatures may be required.

4.3 Slaughter Pig Production

-	Growth rate
-	Food conversion
-	Viability to slaughter
-	Carcass yield
-	Carcass value

All the characteristics listed above may be dependent upon inbuilt genetic resistance to climatic stress, diseases and endo- and ectoparasites. The extent to which these adaptive characteristics show genetic variation will be unknown and the problem of whether or not to attempt selection for them will be considered later.

5. SELECTION METHODS

To achieve the desired objectives, reliance is placed on various measured traits, not necessarily those listed as selection objectives. Nevertheless the more directly it is possible to reflect a selection objective in a measured trait, the greater the response is likely to be. Some indirect measures are valuable since they facilitate the employment of selection methods which would otherwise not be feasible. The prime example of this is the estimation of carcass merit from live animal fat mesurements which makes it possible to carry out large scale performance testing and to dispense with progeny testing.

For the improvement of female reproductive traits which have a low heritability, the use of additional litter records on the same individual and of records from close relatives is valuable. Thus the device of keeping potential breeding stock from a first litter and retaining that sow for a second litter to record litter performance, will almost certainly be a valuable method. Using the records from relatives will require the calculation of selection indices with appropriate weights and, because of the varying numbers of relatives which will be available in any given case, the use of a computer becomes useful.

In measuring the performance of the growing pig, it is appropriate to use weight intervals rather than age intervals for record purposes since this avoids many non-genetic variables. Measurement of fat thickness as a predictor of leanness would also be best done at a fixed weight, but since this may be inconvenient, correction of the measurement to a fixed weight is appropriate. With the higher heritabilities observed for most traits observed on the growing pig, the additional benefits to be gained from use of information on relatives will be small and not a high priority in devising an appropriate selection method.

6. SELECTION METHODS FOR ADAPTIVE TRAITS

Selection of individual animals for adaptation to climatic conditions will probably not be feasible and is probably not to be regarded as an essential feature of the selection procedure. In cattle, where climatic room exposures and body temperature measurements have been made in the selection of stock adapted to the tropics, the current tendency is for such tests to be dropped and reliance placed on natural exposures.to climatic extremes. Similarly with pigs, Horst (1982) from his review of the literature could find no indicators of heat tolerance.

With resistance to diseases and parasites, although laboratory tests may be feasible, or can possibly be devised for the future, natural exposure will probably be all that can be achieved. For many diseases natural exposure is subject to drawbacks because of the complexities introduced by maternally acquired immunity. For the time being some hard decisions may be necessary in deciding upon those diseases which will be excluded by sanitary measures and those where endemic exposures will be the normal course of events. Realistic assessment of the diseases to which stock may be exposed during the production phase of their use will be needed to avoid undue optimism about veterinary measures which are unlikely to succeed in widespread practice. Selection can then be practised in that environment using what Horst (1982) calls 'productive adaptability' where the performance of the animal is used indirectly as a measure of whether or not it is adapted to the stresses it has encountered.

For the future, great store is set on the possibilities of being able to detect genetic markers which can be used as indicators of resistance actors. Although in other animals a few useful associations have been found, this type of investigation must at this stage be regarded as speculative not a method which can be relied upon for solving present problems.

7. BREEDING PLANS

Breeding plans can be assessed by computation of expected progress using the known dependencies on selection intensity, accuracy of selection, extent of genetic variation and on the inverse relationship to generation interval. The influence of some of these factors can now be noted.

Obtaining reasonable selection intensities can be a problem with, relatively small populations and to maximize the opportunities, it is desirable that all potential breeding animals born should be subject to the selection procedure. While some independent culling levels for particular traits may be ncessary, they should be kept to a minimum and unnecessary selection of uniformity of colour or type avoided. Where there are requirements for special recording, as in the measurement of food conversion, then the allocation of pens should follow the principles laid down by Smith (1969).

The accuracy of selection will depend very much on the traits under consideration. As noted already, there may be scope for using records on relatives or characters with low heritability, but little to be gained where the heritabilities are higher. Weight given to different traits is best determined by the use of a selection index although uncertainties over the parameters to be used in such an index will often make it necessary to use values obtained from the literature. These can be combined with estimates of the variability of the local population to give procedures which will allow a start to be made on selection. For a discussion of these problems, see James (1982).

The extent of genetic variation available in the population will depend on the past history of selection, on past bottlenecks and on effective population sizes. If the indigenous population under improvement is a unique one, then it may be worthwhile devising some form of open breeding plan which allows the immigration of additional breeding animals from outside the nucleus on the basis of superior performance. As has been shown by James (1978), this measure can increase effective population size and reduce rates of inbreeding.

In many practical breeding schemes, the generation interval does not receive as close attention as other selection parameters. One reason for this is that the discard of comparatively young breeding animals may add considerably to the cost of the breeding operation. Arrangements to pass animals from the breeding nucleus to commercial crossing herds may therefore be useful in offsetting this cost. Optimum structures should be computed for alternative circumstances, using the general principles enunciated by Ollivier (1974). It is unlikely that sows should be maintained for more than two litters or boars for more than twenty matings if the generation interval is to be kept at an optimum level.

During the course of selection, unexpected and unwanted correlated responses may need to be monitored. For example, the tropical conditions are such that small body size is an adaptive characteristic, so large increases in adult size may have adverse effects as suggested by Horst (1982). Similarly, the reduction of fatness and increase of lean will not only lead to an animal with reduced energy stores, but may also produce increased heat production with consequent ill effects for climatic adaptability. Such adverse possibilities are not a recommendation for making no changes, but for measuring those that do occur so that some kind of genetic control seems particularly appropriate to monitor the value of selection procedures. The ability to deep freeze sperm makes it possible to provide such controls at reasonable cost and further economies can be made by not taking controls from every generation but by introducing them after a period of selection.

8. RATES OF INBREEDING

The rate at which the improved population becomes inbred will be a major concern in the design of improvement plans. Although some indigenous populations, such as the Meishan from China, appear to have had a long history of inbreeding without serious consequences for fecundity, such isolated experiences are not a general recommendation for neglect of inbreeding. Intense selection will inevitably lead to high rates of inbreeding and some balance will need to be struck. Definitive advice is not possible but attempts to maintain rates of inbreeding at those found in major pig breeds at round 1/2 percent per generation would be a pragmatic choice and one which would certainly be tolerated for a long term programme. For shorter term programmes, high rates of inbreeding could be tolerated.

9. GENERAL REQUIRMENTS

Finally some general observations might be permitted. A major requirement for the success of pig breeding operations is the organizational one both in the conduct of breeding operations in the nucleus herd and in the dissemination of that improvement to the population at large. This requirement should be stressed and placed high on the list of priorities.

Some new technologies may have exciting prospects but probably have low priorities for implementation in many basic schemes of the kind described here. To take a statistical example, the use of BLUP methods will probably add little to a scheme if the breeding population has been well structured for improvement purposes. Similarly, reproductive techniques such as AI and embryo transfer would not appear to be essential ingredients. This is not to deny the value of existing and future research on indigenous populations but to counsel critical appraisal of priorities. Organization and structuring of pig breeding operations should be near the top of that list of priorities.

REFERENCES

^1/ Edinburgh School of Agriculture, West Mains Road, Edinburgh, EH9 3JG, U.K.