Last update: March 27, 2012 12:14:32 PM E-mail Print


Why group breeding is struggling in SA

 C Pettit


THE initial high tempo of development of group breeding in South Africa has not been maintained. This can almost certainly be ascribed to the lack of spectacular results and the difficulty of demonstrating any improvement which may have taken place. Also, the unimpressive phenotypic appearance of group-bred rams has led to scepticism and poor acceptance of this system of breeding in the traditional stud sector. A poor knowledge of the basic principles of group breeding, especially at contributor level, also affected the overall efficiency of some group breeding schemes. In this article, originally delivered last year at the second World Congress on Sheep and Beef Cattle Breeding in Pretoria, Mr Cecil Pettit of the Fleece Testing Centre at Grootfontein, Middelburg, discusses some of the more important factors which are likely to influence the future development of group breeding in this country.

SINCE the establishment of the first two sheep group breeding schemes in the RSA in 1971, this co-operative system of breeding continues to attract the attention of sheep as well as other livestock breeders. The circumstances which prevailed and acted as prime incentives for the adoption of this relatively unconventional system of breeding are still largely present. The purpose of this article is to review the current position of group breeding in the RSA since the survey by McMaster (1980) and to discuss some of the most important developments, results and problems pertaining to these aspects.

Since the survey by McMaster fop cit) there has been little change in the statistical aspects of group breeding in the Republic. Of the present 18 sheep Group Breeding Schemes (GBS), 11 are Merino and 2 Dohne Merino groups comprising commercial sheep, 4 comprising registered Dohne Merinos and a newly formed stud Corriedale group.

The total number of breeding ewes in GBS which have been stabilized is approximately 150000 with an average nucleus-to-contributing-flock ratio of 1: 15 which is within the limits suggested by Turner and Jackson (1972) and later by Rae (1974).

With the exception of five groups which are still increasing their nucleus numbers, adequate rams are being bred to supply the replacement requirements of all members. Despite this fact ram distribution figures supplied by nucleus managers from the 1982/83 season indicate that some members are still drawing some of their ram requirements from traditional studs.

Assuming a ram generation interval of 3,5 years in the contributing flocks, (believed to be a conservative estimate), and 3% of rams used at joining, existing GBS are supplying about 70% of group members' annual ram requirements, the 30% shortfall being made up for the greater part from traditional studs.

Assuming an average fertility of 90% (Iambs born/ewes mated) and a 50% culling rate at selection, about 34% of the remaining rams were used in the nucleus and group members' flocks leaving a surplus of approximately 800. Since only 8% of these rams were taken up by the traditionally bred flocks, it is obvious that surplus rams from GBS have as yet had no significant influence on this sector of the industry.



Since this system of breeding is a joint venture involving a number of breeders, it is understandable that problems may occur which will affect the efficiency of the group at different levels of operation. The following are some of the more important of these problems:

Effectiveness of current selection practices - The high selection intensity which can be achieved by screening large numbers of females in order to identify and regroup the most outstanding animals in a nucleus or breeding flock, is the most important advantage of group breeding. The genetic superiority of the nucleus over the foundation population can be quite considerable.

According to Hight and Rae (1970) the nucleus flock could have an average production of 1,1 kg more wool, and Iambs weighing 8,6 kg heavier at weaning, than the average of the foundation population when considering one trait only. Selection advantages of this magnitude were made possible in New Zealand GBS only through the availability of production records supplied by a national flock-recording scheme which had been in existence for a number of years prior to the establishment of the first GBS in 1967.

In South Africa and Australia, where no such flock recording schemes exist, Merino GBS are restricted to selection in maiden ewes, since environmental differences between older ewes made the identification of genetically superior animals impossible without production records.

Although no formal flock-recording scheme is available to Dohne Merino breeders, GBS comprising registered animals were able to select nucleus breeding material on records kept on an individual basis by stud breeders.

Selection restricted to maiden ewes by the absence of records, excludes about 60% of the available foundation population during the initial and subsequent annual selection. The loss to the breeding group of this untapped potential, together with the questionable reliability of selection on two-tooth performance only is a serious handicap to GBS founded and operated on this basis.

A major problem with which the first Merino GBS in South Africa had to contend was the choice of sires of comparable genetic merit to be mated to the newly-formed nucleus of elite ewes.

Since the breeding value of sires in the traditional Merino studs was unknown, rams with the best phenotypic appearance were selected from those currently being used in the contributing flocks.

Although some of the selection advantage of the nucleus was almost certainly sacrificed through the use of unproved rams during the first two matings, the time consuming alternative of identifying suitable sires by means of progeny testing may well have stifled the initiative and enthusiasm of group members in the critical formative stage of the venture.

The achievements of the two pioneer Merino breeding groups were widely exploited by many of the newly established GBS by the leasing of one or more rams from the older groups. Most attempts to evaluate these sires by means of progeny tests were unfortunately invalidated by inadequate progeny numbers or too few sires in the test.

However, the results of one progeny test processed by the South African Fleece Testing Centre - and which was considered to be meaningful – have shown that the best group-bred ram had a breeding value 10% higher in respect of body mass and 12% higher in respect of clean fleece mass than the best traditionally bred ram.

Furthermore two of the three group bred rams used had above-average breeding values for both body and clean fleece mass, whereas none of the five traditionally bred sires had above-average breeding values for more than one of the traits in question.

Age structure - Most groups, with the exception of those which keep comprehensive ewe records or have low lambing percentages, conform roughly to an age structure of two age groups of rams and three age groups of ewes in the nucleus flocks as suggested by Rae (1974) in the model used for determining optimal structure for GBS.

The use of records for more effective ewe selection in some nucleus flocks has had limited success owing to the lack of data-processing facilities, and any genetic gain which may have accrued from the slight increase in efficiency has in all probability been offset by the increase in generation interval resulting from the retention of older ewes in the nucleus.

In contrast to the nucleus flocks, the age structure in the contributing flocks cannot be regarded as optimal, having remained basically unchanged, with ewes and rams being drafted from the flock only after signs of ageing. This will, however, only have caused a slight increase in the number of age groups of ewes recommended by Rae (1974) and is therefore unlikely to have had any appreciable effect on genetic gain.

The slow turnover of rams, on the other hand, is bound to have had an adverse effect on the overall efficiency of the groups in view of the fact that half the nucleus replacements are made from the contributing flocks.

Low-key participation - lack of complete conformity to the principles of group breeding by group members can usually be attributed to a poor knowledge of basic breeding principles. III-informed members who are easily influenced by outside opinion opposed to group breeding continue to draw on traditional studs for some of their sire replacements. It is noticeable that where this negative approach is prevalent, GBS have little or no success in disposing of surplus rams to commercial breeders.

Low-key participation at group member level is a serious problem which not only adversely affects the GBS concerned, but can also tarnish the image of group breeding in general.

Effect of stress on group-bred sires – A great deal of scepticism levelled at group breeding stems from the uninspiring phenotypic appearance of group bred sires raised under the stress of normal flock farming conditions. Although there has been a tendency by commercially orientated GBS to overemphasise the stress factor by delaying supplementary feeding for too long during critical stages of growth, some degree of stress is nevertheless part of a sound breeding strategy.

Failure of the traditional stud breeders to take cognisance of this important aspect has led to poor adaptability in traditionally bred rams and their offspring, to the harsh environment found in many sheep farming regions of the country. Ironically this has proved to be the most powerful motive for the initiation of GBS in the RSA.

Over-emphasis of subjective selection - Despite the fact that all wool led sheep GBS make use of the performance testing facilities provided by the SA Fleece Testing Centre, there is still a strong tendency, particularly in the Merino groups, to over-emphasise subjective or "hand-and-eye" appraisal during this preliminary form of selection. The hand-and-eye method of selection has for so long been associated with the breeding of an ideal type of sheep that a large proportion of selection pressure is still being sacrificed by the inclusion of traits in the selection programme which are of questionable value to both breed improvement and economic return.



From the time that the Lowestoffe and The Downs Merino GBS were established in 1971, the importance of keeping comprehensive records was recognised both for the purpose of assessing the accuracy and effectiveness of selection techniques as well as for monitoring genetic change.

Selection response in the Lowestoffe Merino GBS -In the absence of a genetically stable control flock, mean values for some of the measured traits in 16 month old ewe progeny of the Lowestoffe nucleus flock were used to estimate the selection response over the eight-year period 1974-81 (Erasmus and Pettit, 1982). When considered in conjunction with the fact that a slight restriction on fibre diameter appears to have kept this trait at a constant level, there is some evidence that there has been some response to direct positive selection for clean fleece mass and body mass. The average phenotypic values of the four traits with their fitted trend lines are shown in Figure 1.



Control flock test - As members of the Performance Testing Section of the Merino Stud Breeders' Society, Merino GBS are encouraged to carry out control tests using ram samples from the state-owned control flock as described by Erasmus (1976).

Effectiveness of two-tooth selection - Doubts concerning the effectiveness of two-tooth selection have resulted in the adoption of a more sophisticated form of selection by two of the Merino GBS. The final selection of nucleus flock replacements is now made from contributors' flocks at approximately 2½ years of age on a combination of the dam's greasy fleece mass and adjusted mass of Iamb weaned. Although selection on total ewe productivity is a new concept to the South African Merino sheep industry, the availability of lambing records in the Dohne Merino GBS comprising registered animals, has made it possible to select both foundation stock as well as nucleus flock ewe replacements on reproductive performance from the outset.

The scope for extensive use of records for more sophisticated selection appears to be limited especially in Merino GBS located in the extensive sheep farming regions of the country. The position is, however, likely to change as results in GBS using updated production records for the selection of ewe replacements become known.

Structure of GBS - Although all sheep GBS in this country operate as two-tier systems, Dohne Merino groups comprising registered animals have integrated their stud and commercial flocks which in effect makes these GBS three-tier systems.

Since the average number of ewes in GBS in this country is only about 9 000, consideration has been given to the establishment of three-tier systems by some groups. However, the advantage of greater numbers associated with this system could be exploited with greater effectiveness and fewer organisational problems by progeny testing sires from groups with similar breeding objectives in a common environment. The breeding values of sires in the test would become known after the comparatively short period of two years, enabling groups to use sires selectively on the basis of their superiority in respect of all measured traits.



The sheep stud breeding industry is characterised by the relatively small stud breeding units. A study of the stud structure of the Merino by Erasmus (1977) has established that the average number of females in the studs is about 270. According to an estimate by J C McMaster (personal communication 1983) the number of registered ewes in the Dohne Merino studs is about 160. Stud sizes of other sheep breeds are not likely to be any larger.

In breed structures typified by small studs, frequent exchange of sires between studs is necessary to avoid inbreeding and to ensure sustained and effective breed improvement. However, for a meaningful exchange of genetic material to be made, the breeding value of sires in individual studs must be known. Although accurate within-flock comparisons are now possible through the performance testing schemes of the Department of Agriculture, the breeding value of sires in different environments remains largely unknown.

Group breeding offers a viable solution to this problem by giving small stud breeders the opportunity to pool their limited resources for the purpose of breeding their own stud sires at reasonable cost with greater predictability of breeding value. The immediate and long-term advantages of the implementation of GBS by this important section of the stud breeders would be effectively transferred to commercial breeders through existing ram distribution channels with considerable benefit to the whole sheep industry.

The course and future development of group breeding in South Africa is likely to be determined by a number of factors. Up to the present time group breeding has been confined almost exclusively to commercial breeders. This tendency has been most prevalent in breeds where the traditional stud breeders have not taken up the challenge of adapting to more scientific breeding methods and complying with current requirements of the commercial sector.

As the pressure and competition from other breed societies using modern breeding methods increases, the complementary role of group breeding may well be recognised by the more innovative breeders. Should this occur the initiative for the establishment of GBS might well shift from the commercial to the traditional stud breeding sector.

Educational programmes designed to improve group members' knowledge of basic breeding principals as well as the principles of the group breeding concept are regarded as essential to the success of GBS. It is abundantly clear that where this knowledge is lacking, low-key participation at contributor level becomes a severe handicap to the operational efficiency of the whole scheme.

Although the South African Federation of Group Breeders (SAFG) is to some extent involved in advising breeders who are interested in establishing new groups, this organisation could play a more active role in the field of extension to existing groups.

The extension service to farmers provided by wool brokers is almost as old as wool broking in the RSA. Although the number of these advisors has been greatly reduced since the consolidation of these brokers into a single wool handling co-operative, they still have a considerable influence on the woolled sheep industry. Members of this service have attended courses on all aspects of modern breeding methods and the practical application of performance testing. This could do much not only to promote the group breeding concept in the future, but also to ensure the overall efficiency of new and existing GBS.



Erasmus, G J. 1976. Sheep breeding programmes in South Africa. In Sheep Breeding, Proc. Int. Congr.

Erasmus, G J. 1977. Breed structure of the South African Merino. M.Sc. Agric. Treatise. Univ. of the OFS (English summary).

Erasmus, G J and Pettit, C. V. 1982. Response to selection in a group breeding scheme for Merino sheep. Karoo Agric. 2 (3) 16.

Hight, G K and Rae, A L. 1970. Large-scale sheep breeding: its development and possibilities. Reprinted from Sheepfarming Annual 1970. 73 - 85.

Jackson, N and Turner, Helen N. 1972. Optimal structure for a co-operative nucleus breeding system. Proc. Aust. Soc. Anim. Prod. 9, 55 - 64 (Quoted by Rae, 1974: paper presented to the first conference of the New Zealand Federation of Livestock Breeding Groups. Executive Officer, PO Box 11137 Wellington, N.Z.).

McMaster, J C. 1980. Co-operative Breeding Schemes in the Republic of South Africa. Proc. Wid. Congr. Sheep and Beef Cattle breeding. Edited by R A Barton and W C Smith. Denmore Press Ltd. Parmeston North. 123 - 132.

Rae. A L. 1974. The development of group breeding schemes: some theoretical aspects. Paper presented at the first conference of the New Zealand Federation of Livestock Breeding Groups. Sheep Farming Annual, Massey University, Parmeston North. 121 - 127.



Golden Fleece Vol 15, No 1, 1985