PROCEEDINGS OF THE INTERNATIONAL SYMPOSIUM ON LIVESTOCK PRODUCTION THROUGH ANIMAL BREEDING AND GENETICS
University of Zimbabwe Department of Animal Science
Edited by K. Dzama, F.N. Ngwerume and E. Bhebhe
UNIVERSITY OF ZIMBABWE DEPARTMENT OF ANIMAL SCIENCE Proceedings of the International Symposium on Livestock Production through Animal Breeding and Genetics May 10–11, 1995 Sheraton Hotel, Harare, Zimbabwe International Symposium on Livestock Production through Animal Breeding and Genetics Organised by The University of Zimbabwe, Department of Animal Science in conjunction with The Zimbabwe Society for Animal Production
Coordinator – Dr K Dzama Organising Committee – Editors Dr K Dzama> – Dr K Dzama Dr F Ngwerume> – Dr F Ngwerume Dr E Bhebhe> – Dr E Bhebhe Ms S Moyo> – Dr N Mpofu To obtain the proceedings call University of Zimbabwe Telephone 303211 Ext 1409
Copyright 1995: Organising Committee, International Symposium on Livestock Production through Animal Breeding and Genetics, Harare, Zimbabwe
I.S.B.N. 0–86924–112–5
SUPPORT FOR THE SYMPOSIUM WAS PROVIDED BY
GTZ CATTLE PRODUCERS ASSOCIATION DICK ELLIOT MEMORIAL TRUST NESTLE ZIMBABWE SOCIETY FOR ANIMAL PRODUCTION RAINBOW TOURS IRVINE'S DAY OLD CHICKS COOPERS COLCOM NATIONAL ASSOCIATION OF DAIRY FARMERS PIG INDUSTRY BOARD AIR ZIMBABWE DAIRIBOARD FINHOLD THE OSTRICH PRODUCERS ASSOCIATION OF ZIMBABWE
Table of Contents
Acknowledgement
Opening Address
Dairy herd improvement programmes in Zimbabwe
Future prospects in pig improvement in Zimbabwe
Lactation effects in Zimbabwean Holstein Herds
Environmental factors affecting milk production of Holstein-Friesian cows in Southern Malawi
Utilising test-day information to monitor individual cow performance
Application of biotechnology in genetic improvement, characterisation and conservation of livestock
Conservation strategies for endangered livestock species
The conservation and selection of indigenous beef cattle in Zimbabwe
Poultry Improvement programmes in Zimbabwe
Ostrich production in Zimbabwe
Appendix 1
Some aspects on current and future livestock breeding strategies in livestock in Southern Africa
Pig Improvement Programmes in Zimbabwe
Future prospects in pig improvement in Zimbabwe
A study of the productivity of pigs at the pig industry board farm at ARCTURUS
Crossbreeding for weaner production in pigs
Evaluation of indigenous pigs in Zimbabwe
Goat breeding research and development activities Zimbabwe
Use of molecular genetic techniques in livestock: A case example in goats
Phenotypic differences in digestive parameters between indigenous Malawian and Dorper males
The role of indigenous cattle breeds: Adaptive and production traits.
LINE X environment interaction for fertility in Afrikaner cattle
Selection for improved reproductive efficiency in beef cattle under tropical and subtropical environments
Evaluation of beef breeds for production in Zimbabwe Livestock improvement in South Africa: Performance driven
Beef performance recording in Zimbabwe: the way ahead
Recent developments in livestock breeding in South Africa
Closing Remarks ACKNOWLEDGEMENTS
The organising committee would like to thank everybody who contributed in making this symposium a success. We are especially grateful to the sponsors without whose kind support this gathering would not have been possible. The Chairman of the Department of Animal Science, Dr Simbarashe Sibanda and members of his department provided invaluable support to the symposium. Special thanks to Dr Stanley Makuza and Mr Irvin Mpofu for reading through the manuscript and making very helpful comments. INTERNATIONAL SYMPOSIUM ON LIVESTOCK PRODUCTION THROUGH ANIMAL
BREEDING AND GENETICS: OPENING ADDRESS
P. Nyathi Department of Research and Specialist Services, Ministry of Agriculture, P O Box 8108, Causeway, Harare, Zimbabwe.
Ladies and gentlemen, I want to thank the organisers of this important event for inviting me to come and open this two-day symposium. To all scientists from Kenya, Zambia, Malawi and the Republic of South Africa and to our Zimbabwean scientists representing various institutions and organisations in the livestock industry, I congratulate you all for having been able to support this symposium with papers to be presented in the two days. I hope that the diversity of the subjects under consideration is sufficient to cause the desired discussion to take place. Let me on behalf of the Department of Research and Speciality Services, the Ministry of Agriculture and the Government of Zimbabwe welcome all present, and in particular extend our gratitude to all our visitors for having taken all the trouble to travel to be with us during this symposium.
This symposium on livestock production through animal breeding and genetics has come at the right time when there is an international call for conservation of natural resources. Livestock are an important component of national economies in those countries in which livestock by-products (meat, milk, hides. mohair, tallow etc) are consumed by local industry and are also exported. These products are however, derived from animals whose genetic constitution has been little studied and their potential to survive in the ever changing ecosystems is not well understood.
In Zimbabwe, animal breeders have concentrated on accounting for non-genetic factors as a prerequisite for the assessment of the genetic potential of individual animals. A literature search indicates that factors such as age at weaning, year of calving, feeding regime, age of darn at calving, lactation status etc have been taken into account in animal breeding experiments in Zimbabwe. In addition, the performance (reproductive, growth and carcass attributes) of a wide range of pure bred and crossbred genotypes from representative Bos taurus (Charolais, Sussex, Simmental, Holstein - Friesian, Hereford) Zebu and Sanga cattle (Mashona, Nkone, Tull) has for several years received attention locally and regionally. The body of research results is impressive, although it can be argued that all effort was directed to the improvement of livestock productivity and production with conservation of genetics being incidental. This trend of experimentation has probably led to a loss of important genes necessary for adaptation to changes in the environment.
Animals and the environment in which they survive are both evolving as human intervention aimed at the supply of human needs is intensified. This raises the question whether we have the institutional capacity to develop genetic engineering techniques (such as restriction fragment length polymorphism analysis and Mitochondrial DNA sequencing for differentiating cattle breeds). This form of analysis provides high resolution information with good efficiency and data readily amenable to phytogenetic analysis. The existence of unplanned crossbreeding and crossbreeding for short-term gains when the economics of producing a particular animal type or genotype is favourable, has meant that our livestock populations are now a mixed bag rather than breeds. In the past few decades, there has been interest in the crossing of cattle breeds (Bos indicus x Bos taunts) to exploit interbreed dominance variation, otherwise known as heterosis. However, the selection of suitable breeds and crossbreeding strategies tends to be based on historical records of the breed and information gleaned from existing crossbreeding programmes. Consequently, an estimate of intrabreed and interbreed variation may help in the formulation of rational breeding policies. To that end, the analysis of mitochondrial DNA and Microsatellite markers can contribute information to this form of study and help in the decision making process.
Pedigree records are kept in Zimbabwe by a few breed societies and the Zimbabwe Herdbook and these need to be consulted to ensure that animals which we may want to use in cross- breeding work are unrelated and representative of the desired breeds. There is need to characterise the gene pool of cattle kept in rural areas where no records are kept. In those areas, there could be important genes, relevant to fitness and therefore warranting a concerted effort to harness these genes in order to meet future requirements for survival, particularly because these cattle, and indeed goats and other livestock have evolved with very limited management. The ever-increasing costs of production in commercial livestock enterprises and the frequent droughts that decimate the grazing resource base both indicate the relevance of the need to develop livestock types that can perpetuate the species under conditions of reduced available grazing and feed resources. The question is whether we should change the environment. This requires that careful thought be given to the kind of manipulation needed to achieve the long-term goal of sustainability in the productivity of individual breeds, that of their crosses with other breeds and that of the environment on which livestock production is based. Animal breeding geared towards the satisfaction of short-term gains may be costly in future as it may unintentionally lead to loss of some important genes and this equates to genetic erosion of the livestock resource base.
Technologies to facilitate animal breeding are in place in the world, ie Artificial insemination (AI) Embryo transfer (ET), gene cloning, DNA marker assisted selection and many other newly developed techniques. This symposium is the correct forum for discussion of the relevance of methods that are adoptable in this region. I therefore leave it to you as professionals to look into (a) training needs for the region associated with livestock genetic engineering; (b) to look into adequacy of institutional capacity to undertake livestock breeding work using the new techniques or a blend of both old and new, (c) to look into the breadth and depth of ways and means by which animal breeding and genetics can improve production on a sustainable basis.
Finally, Mr Chairman, I want us to be reminded that the economic success of a production system must be judged by an analysis of returns and costs at the herd level, not on production figures from individual animals. Smularly, our breeding objectives must be detennined by such considerations. Only then may we free ourselves from attempts to breed larger and larger animals with higher and higher production per beast. In Zimbabwe and perhaps in the region, such a breeding policy could be disastrous. The strategy must be one of optimisation rather than maximisation, one of conservation with utilisation rather than conservation per se.
With these few remarks, allow me Mr Chairman, to humbly declare this symposium open.
Thank you. DAIRY HERD IMPROVEMENT PROGRAMMES IN ZIMBABWE
SUMMARY
INTRODUCTION
HISTORICAL DEVELOPMENT
THE 'STATEMENT' SCHEME
CONCLUSION
REFERENCES
C. Banga Zimbabwe Dairy Herd Improvement Association, P 0 Box CY 2026, Causeway, Harare, Zimbabwe.
SUMMARY
The importance of milk recording is highlighted and its historical development in Zimbabwe outlined. Major advances have been made recently to upgrade the milk recording programme in Zimbabwe from an antiquated and cumbersome method of compiling historical production data to an efficient state of the art system. The dairy herd improvement services available in the country today present farmers with a valuable tool for improving profitability through better management and superior genetics.
INTRODUCTION
Herd improvement services comprise milk recording and allied programmes. Milk recording is the objective and systematic measurement of individual cow performance. Performance records serve as a valuable tool for improving overall herd productivity through better management and genetic improvement. In addition, milk recording scheme data may be used for research work, planning and teaching. Hammond (1994) also envisages that milk recording can be used as a key technology to improve livestock production, productivity and sustainability in developing country herds by using it as a vehicle for farm extension services.
HISTORICAL DEVELOPMENT
Official milk recording in Zimbabwe started in 1932, but previously there was organised recording as early as 1929. The milk recording scheme grew gradually in size and became more refined over the intervening years as developments took place in the dairy industry. In 1992, Zimbabwe was granted full membership of the International Committee on Animal Recording (ICAR) to become the third African country after South Africa and Tunisia belonging to the organisation. The so called register system of milk recording that was in use, however, did not keep abreast with developments in technology. Participation of farmers in the scheme remained consistently low, compared to that in other countries (Table 1), probably due to the limited value of the progratnme. Table 2 shows the trend in participation by herd owners in milk recording during the period 1983 to 1994. Table1. Position of milk recorded in selected countries
Country % of recorded herds Argentina 28. 8 Australia 50.6 Canada 54.7 Denmark 76.0 France 42.4 Israel 52.1 Jersey 83.0 Portugal 4.4 South Africa 32.8 United States 30.2 Zimbabwe 20.4
Source: ICAR (1994)
'Register' system
Basically, the method involves analysing individual cow and herd performance records retrospectively. For more than half a century, the scheme was administered by the Dairy Services Branch of the Department of Research and Specialist Services. Virtually all the costs of delivering the milk recording services were borne by the government and minimal fees were charged to the members.
Under the register system herd owners participating in the scheme were required to keep daily and cumulative lactation milk yield records of individual cows. Milk recorders employed by Dairy Services would visit each participating herd at bi-monthly intervals. During their Men twentyfour hour stay on the farm, the milk recorders would carry out the following recording and testing procedures:
(i) Check farmer's milk yield records and measuring devices.
(ii) Sample each cow's milk during each milking and record test day milk yield.
(iii) Determine butterfat content of each cow's milk sample using the Gerber method.
(iv) Collect completed lactation records, summarise and dispatch them to the Central
Milk Records Office.
(v) Ear-tag calves upon the herd owner's request.
Staff at the Central Milk Records Office would process incoming records as follows:
(i) Verify information recorded by the farmer.
(ii) Calculate weighted average butterfat content and calving intervals.
(iii) Prepare records for electronic processing.
All the records would be entered into a computer and at the end of the year, herd averages would be computed for milk yield, butterfat yield and per cent, calving age, days dry and calving interval.
Besides failing to rise to the expectations of herd owners, the milk recording service deteriorated due to inadequate capital and financial support by government.
Recent advances
The year 1993 saw the implementation of the Canadian International Development Agency (CIDA) funded Zimbabwe Dairy Cattle Improvement Project (ZDCIP), which was a milestone in the history of milk recording in Zimbabwe. A major component of the project was the upgrading of the antiquated 'register' system of milk recording to a state of the art 'statement' method. The Zimbabwe Dairy Herd Improvement Association (ZDHIA) had been set up in May 1993 to deliver the new herd improvement services. This major development saw a sharp increase in the participation of herd owners in milk recording (Table 2). The membership of ZDHIA continues to increase and it is anticipated that by the end of 1995, about 40% of farmers will be participating in milk recording.
Table 2. Trend in participation by farmers in milk recording
Year Total no. of dairy herds No. of milk recorded Participation
herds 1983 516 104 20.2 1984 524 105 20.0 1985 531 95 17.9 1986 552 101 18.3 1987 556 104 8.7 1988 511 103 20.2 1989 521 102 19.6 1990 508 99 9.5 1991 479 92 19.2 1992 442 86 19.5 1993 412 84 20.4 1994 410 98 23.9
THE 'STATEMENT' SCHEME
Operation
Essentially, the 'statement' method offers a quick comprehensive profile of the herd by presenting, at regular frequent intervals, reports containing progressive history of all cows in each herd. Two service options exist for herd owners, namely fully supervised and owner sampler. Fully supervised herds receive ten visits per year by ZDHIA milk recorders who carry out all the on farm testing procedures and dispatch milk samples preserved in bronopol pills and input forms to ZDHIA central offices. Owner sampler herd owners on the other hand do their own testing and transportation of samples. Another form of owner sampling is group recording which is used for smallholders. Owner sampler herds are tested twelve times a year, with milk recorders visiting four times a year to check on recording devices and animal identification.
Input forms The forms that are used to record test day information are shown in appendices I to III. Herd owners record events occurring in their herds between test days such as calvings, deaths and disposals on the in barn herd event record forms. In addition, the farmer is required to keep cow identification and parentage information. The milk recorder will transfer this information onto the supplementary test day sheet and barnsheet. Previous test day information is pre- printed on the supplementary test day information sheet and the barnsheet.
Laboratory testing
At the central laboratory, milk samples are tested for butterfat and protein using a Bentley 2000 infra-red milk analyser. Somatic cell counts are determined by a Counter.
Samples are tested within three days of their arrival at the laboratory.
Data processing
Information on the input forms and the lab results are captured into the computer as soon as they reach the data processing unit. Electronic processing of the data then ensues and the reports generated are sent back to the herd owner within ten days after test date. Farmers thus get timely feedback on their herd and individual cow performance, enabling them to make accurate management decisions. ZDHIA reports present the farmer with a valuable tool for enhancing management and consequently improving efficiency of production. Details of how the different reports may be used as management tools are described by Banga (1994) and Banga (1995a,b).
CONCLUSION
The herd improvement services currently available to dairy farmers in Zimbabwe are among the best in the world. Dairy farmers have the potential to increase their profit margins substantially through use of the wealth of information provided by ZDHIA. The biggest challenge to extension agents is to help farmers to get the best out of these services.
REFERENCES
Banga, C. (1992) Cattle Research Network, International Livestock Centre for Africa, Addis Ababa, Ethiopia. Banga, C. (1994) Zimbabwe Dairy Herd Improvement Association Newsletter 1. Banga, C. (1995a) Zimbabwe Dairy Herd Improvement Association Newsletter 2. Banga, C. (1995b) Zimbabwe Dairy Herd Improvement Association Newsletter 2.
Hammond, K. (1994) Proc. 29th Biennial Session Internl. Conunittee Amin. Rec., Ottawa, Canada. FUTURE PROSPECTS IN PIG IMPROVEMENT IN ZIMBABWE
SUMMARY
INTRODUCTION
TRAITS OF ECONOMIC IMPORTANCE
CURRENT STATUS AND FUTURE PROSPECTS
CROSSBREEDING AND SELECTIVE BREEDING
FEEDING SYSTEM DURING TEST
CONCLUSION
REFERENCES
A. Shoniwa and K. Dzama Pig Industry Board, Box HG297, Arcturus, Zimbabwe and Department of Paraclinical Veterinary Studies, University of Zimbabwe, P. O. Box MP 167, Mt Pleasant, Harare, Zimbabwe.
SUMMARY
The Pig Industry Board is responsible for national evaluation of pigs. Pigs are tested using their own performance for litter size at birth, feed conversion efficiency and backfat thickness. A selection index is used to rank the animals. The accuracy of selection in the present system is low and it can be improved by using progeny testing and statistical techniques such as Best Linear Unbiased Prediction to compute estimated breeding values for the traits in question. Introducing progeny testing may also require more extensive use of Artificial Insemination so that boars can be tested widely. Under the current performance testing scheme, testing of animals should be done under the same environment they will experience on the farm.
INTRODUCTION
In pig production there are many traits of economic importance and improving these traits requires the improvement of the environment together with genetic improvement. The genotype of an animal sets a ceiling to which the animal can be improved through manipulation of the environment. It is, therefore, important to compliment advances made in management practices by improving the pigs genetically.
TRAITS OF ECONOMIC IMPORTANCE
The traits of economic importance include growth, carcass traits, reproductive traits and feed conversion efficiency (FCE).
Growth Producers need animals with high growth rates. Fast growing animals result in reduced maintenance feeding and this has a positive effect on the producer's margin. Fast growth, while being desirable for many other reasons, also results in a lower housing charge per pig. Fast growing animals have, however, been reported to show a high incidence of leg problems (English, et al., 1988). It is, therefore, important when selecting breeding stock to take into account the undesirable correlation between growth rate and leg weakness. Good strong legs are essential for the breeding animals. A highly prolific sow is of no use if it cannot stand for the boar and a good boar is of limited use if it cannot mount a sow on heat.
Carcass Traits
The market demands leaner carcasses so the producer and breeder should strive to produce leaner pigs. Not only are leaner carcasses desirable from the consumer point of view but they are also cheaper to produce. Very lean carcasses should, however, be avoided. Genetically very lean pigs may have problems when they enter the breeding herd because they,will farrow with little fat reserves. Sows farrowing with very little fat reserves are more likely to have problems during and after lactation (Mullan and Williams, 1989). A sow suckling a large litter and with little fat reserves will lose a lot of condition during lactation and is more likely to have rebreeding problems. It has been reported that the success achieved in reducing the amount of fat in pigs has been done at the expense of appetite (English et al., 1988). Part of the mechanism by which breeders have achieved a reduction in carcass fat content has been by subconsciously selecting pigs with lower appetites. English et al. (1988) also reported that some genotypes with high lean and low carcass fat contents are unable to grow lean tissue at a sufficiently fast rate because of limited appetite. Breeding females with limited appetite will have problems meeting the requirements of lactation especially if they are nursing large litters. It is therefore important when selecting breeding stock to ensure that the animals selected have reasonable appetites.
Some genotypes which are extremely lean have been reported to have a higher incidence of porcine stress syndrome which causes pale, soft and exudative meat (English et al., 1988). Apart from having meat quality problems these strains of pigs also have lower feed intake and poorer reproductive performance. Care should be taken when importing semen.
Reproductive traits
Because of the low heritability of reproductive traits eg litter size, little effort has been put to try to improve these traits through selection. Traits of low heritability, have been improved through a combination of crossbreeding and manipulation of the environmental factors. Today, heterosis and management influences have been exploited extensively to improve reproductive performance. It is now wise to try to improve reproductive performance through selective breeding. When selecting highly prolific sows it is important to select animals with desirable teat spacing and numbers. Unless the extra piglet produced by the highly prolific sow is able to get a teat to suckle the benefits of (such highly prolific) sows will be lost. For highly prolific lines it may be wise to select animals with 14 or more functional teats.
Feed conversion efficiency
In pig production feed accounts for about 75% of the total production costs. It is, therefore, important to target FCE for genetic improvement. A more efficient animal is desirable because it has the potential to reduce production costs. In addition more efficient feed converters tend to produce leaner carcasses.
CURRENT STATUS AND FUTURE PROSPECTS Currently the Pig Industry Board (PIB) has the mandate for central testing of pigs in Zimbabwe. Pigs are evaluated based on their own performance. The animals sent for testing are preselected for litter size at birth and at weaning by the breeder. At the end of the on- station test, an index incorporating FCE and backfat thickness is used as a basis for selection. Reproductive traits in general, and litter size in particular have low heritabilities (.10). As such the accuracy of predicting the breeding value of an animal for litter size is fairly low. (ie x.10 = .32) The accuracy of predicting the breeding value of an animal for a trait of heritability .10 and repeatability .25 using the animal's own performance and an infinitely large number of its records is 63% (Van Vleck et al., 1987). The accuracy of selection for FCE (heritability 0.3) is .55 and the accuracy for selecting for backfat thickness (heritability .5) is .71 under the PIB performance testing programme. Clearly there is need to incorporate litter size into the index and to improve the accuracy of predicting the breeding values of the animals for the traits selected for. For the traits emphasised in this programme, progeny testing of boars and sows lends itself as a method of choice to increase accuracy of selection.
Prneny testing
Initially progeny testing was used to evaluate the genetic merit of pigs at PIB. Progeny testing was abandoned because it was considered costly and it also took more time to assess the animals compared to using the animal's own performance. Despite these demerits, progeny testing is the best method given the traits being selected for in the PIB test. It substantially increases the accuracy of selection for traits with low heritabilities like litter size. The accuracy of prediction of additive genetic value from records on 30 progeny would be twice as large (.66) as that of predicting breeding value from the animal's own record (.32). In addition progeny testing is important for evaluating traits which are sex limited eg litter size or traits where destructive sampling is involved eg carcass traits.
If progeny testing is reintroduced at PIB it may become necessary to promote extensive use of Artificial Insemination (AI). This will enable boars to be used faster and more widely thus generating larger databases for more accurate evaluations. Currently the use of AI in the industry is almost nonexistent. The economic benefits in terms of increased FCE, improved litter sizes and leaner carcass derived from accurately selecting pigs of high merit at national level far outweigh the disadvantages of progeny testing.
Data management and analysis
Progeny testing will result in generation of large data sets which will inevitably need to be stored on a powerful computer. Data on animals' progeny and relatives will not only be collected on station but also on farm through the existing PIB multiplication scheme. Even though data collection and record keeping at the P1B is meticulous, there is need for computerisation. The database created will be analysed with powerful statistical tools to generate estimated breeding values (EBV) for the traits of interest. An EBV is an estimation of the genetic value of an animal. It indicates its value as a parent. EBVs can be updated as more information on the animal's progeny and relatives become available. EBVs can be used to construct a selection index incorporating the breeders choice of traits. The current selection index in use at the PIB needs to be revamped along these lines.
One tool which can be used to compute EBVs is called Best Linear Unbiased Prediction (BLUP). These are mixed model equations which take into account the heritability of the trait, the amount of information available for each boar or sow, the genetic level of the herd, genetic trend and non-genetic factors such as management groups. Hoste (1994) reported that when BLUP assisted selection is compared with selection based on an animal's records, for traits of low heritability (.10) such as litter size, the response to selection is increased by up to 30% compared with improvements in selection response of 5 to 10 % for traits of moderate (.40) heritability.
CROSSBREEDING AND SELECTIVE BREEDING
Crossbreeding is widely utilised in the pig industry to utilise hybrid vigour and complementarity. However, unless properly planned schemes are put in place economical heterosis will not be achieved thus crossbreeding will be a financial liability (Dzama, 1994). There is evidence from PIB data that in poorly planned schemes where crossing is done haphazardly, the advantage of using crossbred sows to take advantage of maternal heterosis is very small or non existent (Mungate et al., 1995, unpublished). The PIB should develop and recommend various crossbreeding schemes to producers. EBVs obtained from the proposed progeny testing scheme should be used when selecting animals to cross. One way of doing it may be to develop specialised sire and dam lines. In trying to improve reproductive traits it may be beneficial to concentrate on improving reproductive performance in the dam lines while concentrating on improving growth, FCE and carcass traits in the sire lines.
FEEDING SYSTEM DURING TEST
In Zimbabwe selection of breeding stock on the performance testing programme is based on the restricted feeding system.. Most producers raise their fatteners on an ad libitum feeding system. It is therefore important to test animals under a feeding system which their progeny will experience on the different farms. When one tests animals under a restricted feeding system one will be limiting the genetic potential of some animals. Testing of animals on an ad libitum or "to appetite" basis has its disadvantages. As the pig proceeds to satisfy its appetite on an ad libitum system, its FCE will start to deteriorate and its lean content to decline because of increasing fat deposition (English et al., 1988). Thus selection for carcass lean content and FCE on ad libitum feeding will tend to favour pigs with a lower appetite. As reported earlier pigs with lower appetites have serious problems during and after lactation. An alternative system of testing which helps to prevent the reduction of appetite associated with feeding-to-appetite systems of testing is therefore desirable. English et al. (1988) reported a scheme first put forward by Professor Kielanowski and Dr Kotarbinska of Poland to be the most appropriate system upon which to base selection of breeding stock. This scheme is based on a time scale feeding system using a feed scale close to ad libitum. Under this system all pigs receive the same ration each day and are on test for the same period. The pigs which will have gained most weight by the end of the test will have the best combination of ability to consume their ration, FCE and lean tissue growth rate.
CONCLUSION
There is clearly a need to improve the reproductive performance, growth rates and carcass characteristics of pigs in Zimbabwe. Whether this goal is achieved depends on how accurately we select superior pigs for breeding. Certainly PIB provides a sound base from which to launch the proposed improvement programmes.
REFERENCES
Dzama, K. (1994) J. Zimbabwe Soc. Anim. Prod. (in press).
English, P.R., Fowler, VR., Baxter, S. and Smith, B. (1988) Farming Press. Ipswich, Suffolk. Fowler, VR., Bichard, M. and Pease, A. (1976) Anim. Prod. 23:365-387.
Hoste, S. (1994) International Pig Breeding Journal 3: 2-4.
Mullan, B.P. and Williams, 1. H. (1989) Anim. Prod. 48:449-457. Van Vleck, L.D., Pollak, E.J. and Oltenacu, E. A.B. (1987) W.H. Freeman and Company N.Y. LACTATION EFFECTS IN ZIMBABWEAN HOLSTEIN HERDS
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSION
ACKNOWLEDGEMENTS
REFERENCES
S. M. Makuza and B. T. McDaniel Department of Animal Science, University of Zimbabwe, Harare, Zimbabwe and Department of Animal Science, North Carolina State University, Raleigh 27695, USA.
SUMMARY
Heritabilities, repeatabilities, and genetic and phenotypic variances for milk yield 3.5% FCM, fat yield and fat percentage in Zimbabwean Holsteins were estimated by lactation using a derivative free; iterative algorithm of REML (MTDFREML). The animal model contained herd- year, calving month, age at calving, permanent environment and animal effects. In all cases the inverse of the additive genetic relationship matrix was included. Separate estimates were obtained from records projected to 305-d basis and unadjusted data for lactations one through four. Heritabilities were higher for the average of first and second lactations than separate first and second lactations. Heritabilities, repeatabilities and variances for milk yield, fat corrected milk yield (FCM) and fat yield were within the range of values from temperate areas, but those for fat percentage were lower. Age at calving follow patterns similar to those of more temperate areas. There is adequate genetic variance in Zimbabwean dairy herds to make large genetic gains in milk production through conventional selection methods.
INTRODUCTION
Makuza and McDaniel (1994) found heritabilities for milk yield from daughter-dam regressions in Zimbabwean Holsteins of .24 for all lactations, .33 for first, and .28 for second lactations. Standard errors of these heritabilities averaged only .03. A problem with previous studies of Zimbabwe data was removal of short records which probably gave biased results for genetic and environmental variances.
Repeatability in the genetic context measures the proportion of the phenotypic variance of a trait due to permanent effects, both genetic (additive and non-additive) and environmental. Reported repeatabilities for null< yield range from .35 to .60 but .45 to .75 for fat percentage Mao (1984). The repeatability value for fat yield quoted by McDowell (1972) is .49.
The objectives of our study were: (a) to determine in Holsteins whether estimates are similar by lactation number; and (b) whether parameters for the four production traits are the same for the first two lactations separately and combined.
MATERIALS AND METHODS
A total of 41,483 Holstein production records for a 10-year calving period (from 1979-1987 and 1989) from the Zimbabwe Milk Recording Scheme (ZMRS), recently renamed the Zimbabwe Dairy Herd Improvement Association (ZDHIA) were the base data. Restrictions required only days in milk (DIM) between 15 and 305 d. All records with missing sire and dam identification were deleted.
First lactation was defined as the first available record without a previous dry period and initiated between 18 and 35 mo of age. Second lactation was defined as any record coded as second and made by cows between 36 and 51 mo of age. Similarly, third lactation was restricted to records coded as third and made by cows between 40 and 85 mo of age. The maximum age at calving for cows in fourth and later lactations was 169 mo with a minimum of 54. Combined first and second lactations were also evaluated. In this set only cows with both first and second records were included, thereby eliminating low producers in first lactation that were culled. The combined first and second lactation set included only cows with both records and a calving interval of greater than 300 d. The traits studied using MTDFREML Fortran programmes by Boldman et al. (1993) were milk yield, 3.5% fat corrected milk yield (FCM), fat yield and fat percentage.
RESULTS AND DISCUSSION
Yields increased up to the third lactation, then decreased (Table 1). Fat percentage decreased from first to fourth lactation in both unadjusted and adjusted records. Dairy cows are most productive during the third lactation in Zimbabwe.
Table 1. Means and standard deviations for yield traits in Zimbabwean Holstein cows by lactation.
Lactation Unadjusted1 Adjusted 2 Trait Mean SD Mean SD 1 Milk yield, kg 5132 1492 5151 1532 3.5_FCM, kg 5195 1455 5215 1508 Fat yield, kg 183 52 184 54 Fat, _ 3.61 .39 3.61 .39 2 Milk yield, kg 5723 1700 5741 1696 3.5% FCM, kg 5767 1645 5785 1639 Fat yield, kg 203 58 204 58 Fat, _ 3.58 .42 3.58 42 3 Milk yield, kg 5900 1777 5921 1768 3.5% FCM. kg 5922 1737 5943 1728 Fat yield, kg 208 62 209 62 Fat, _ 3.55 .41 3.55 .41 >4 Milk yield, kg 5752 1696 5775 1689 3.5% FCM, kg 5751 1683 5774 1675 Fat yield, kg 201 61 202 60 Fat, % 3.51 .40 3.52 .40 1 & 2 Milk yield, kg 5318 1570 5331 1568 3.5_FCM, kg 5386 1537 5399 1535 Fat yield, kg 190 55 191 55 Fat, % 3.61 .41 3.61 .41
1Not projected to 305-4 basis.
2Projected to 305-d basis.
Table 2 shows results using the first two lactations separately and combined. Heritabilitics for all four traits were similar in the first and second lactations. Additive genetic variances for milk and fat percentage decreased from first to second lactation but environmental variances increased, therefore phenotypic variances remained almost constant. Heritabilities for combined first and second lactations were higher than those for separate lactations. Combining first and second records increased all genetic parameter estimates compared to separate lactations. These findings generally agree with those of Makuza (1988), but are slightly lower than those of (Makuza and McDaniel 1994; Makera I 995). Maternal effects and epistatic genetic effects, especially additive by additive gene effects, could explain part of these differences. Hentabilities for milk yield in the first lactation were higher than those from the two previous studies (Makuza and McDaniel 1994: Makuza 1995), but lower for second lactation. Differences may have been due to the algorithm used and the inclusion of all animal additive genetic relationships for estimates in Table 2.
Table 2. Heritabilities. repeatabilities and standard deviations for production traits of Zimbabwean Holstein cows in the first two lactations.
Trait Lactation 1 Lactation 2 Lactation 1 & 2
h2 h2
Unadjusted1 Adjusted2 Unadjusted Adjusted Unadjusted Adjusted
h2 rp h2 rP Milk .25 .22 .25 .27 .35 .59 .35 .60 vield Fat .20 .17 .18 .17 .22 .44 .20 .44 vield 3.5 % .24 .22 .26 .28 .22 .48 .25 .46 FCbl Fat, - .27 .22 .20 .16 .20 50 .20 .50 % Standard Deviations Milk Additive 479 474 568 596 663 663 Yield Environmental 819 887 994 970 712 712 Phenotypic 949 1005 1145 1139 1115 1121 Fat Additive 15 15 17 16 17 17 Yield Environmental 30 33 36 36 27 28 Phenotypic 33 36 40 40 36 37 3.5% Additive 454 467 571 586 467 499 FCM Environmental 803 887 965 934 711 734 Phenotypic 922 1002 1121 1103 987 998 Fat % Additive .18 .16 .16 .15 .16 .16 Environmental .29 .30 .32 .33 .25 .25 Phenotypic .34 .34 .36 .36 .36 .36 n 10 903 7914 5217
1Not projected to 305-4 basis.
2Projected to 305-d basis.
Repeatabilities and heritabilities for fat percentage were not as high as previously reported from temperate areas (Mao, 1984). This possibly is because temperature stress in Zimbabwe tend to dcprcss the intake of forages which can cause considerable fluctuations in fat content (N.R.C, 1`181). Hentabilities and repeatabilities for milk yield, 3.5% FCM and fat yield were similar to those of' temperate zone climates, however (Table 2).
Effects of month at calving of Holsteins within parity are in Table 3. First lactation appears more affected by the environmental effect of month of calving, with second lactation least affcctul. For first and second lactation groupings, cows calving April through September were most productive. In lactations 3 and 4, the period October through December showed negative responses. Dairy fiinncrs should avoid this calving season for older cows if maximum yield is their goal. The period April through September should be the recommended calving season to obtain maximum yields in Zimbabwe.
Table 3. The effect of month of calving on unadjusted' milk yield (kg) in Zimbabwean Holstein cods by lactation.
Location 1 2 1 & 2 3 >_4 Month of Parameter Noof Parameter No.of Parameter No.of Parameter No.of Parameter No.of calving ohs obs ohs estimate estimate estimate obs estimate ohs estimate January 0 621 0 457 0 680 0 386 0 676 February -83 704 118 476 7 784 68 415 23 730 March 53 1049 213 746 102 1037 97 713 204 1117 April 155 970 264 770 201 993 122 718 184 1242 May 170 997 348 745 258 960 197 632 355 1236 June 172 1051 399 744 326 996 226 657 398 1168 July 173 1123 435 811 292 1070 258 692 292 1200 August 58 939 366 794 211 932 236 659 237 1215 September 86 845 333 700 155 872 106 568 171 978 October -25 936 247 592 100 740 -1 532 14 910 November 11 872 201 506 80 667 -102 444 -53 839 December 64 792 274 573 169 702 -122 456 -71 725
1Not projected to 305-d basis.
Table 4 shows the effects of age at first calving on milk yield in Holsteins. Using 24 mo as the base, there is increasing first lactation yield up to 35 mo. This pattern is similar to findings in temperate areas. These results, however suggest that the growth and development rate of heifers in Zimbabwe should be improved. The continual rise shows that those calving at 24 mo or less are required to use a high proportion of their nutrients during lactation to continue growth. It is desirable to have higher performance at younger ages because late calving tends to shorten productive herdlife (Tabbaa, 1993). Tabbaa (1993), working with DHIA records from Southeastern U.S. found that cows calving first at 23 mo of age had the highest lifetime milk yield.
Table 4. Effect of age at first calving on unadjusted' milk yield (kg) of first lactations in Zimbabwean Holstein cows.
Age at first parameter estimate Number of observations
calving(mo) 18 -555 24 19 -607 26 20 -486 50 21 -275 97 22 50 219 23 -68 475 24 0 862 25 186 1213 26 236 1321 27 254 1184 28 404 1056 29 438 905 30 463 815 31 431 716 32 511 574 33 556 550 34 550 439 35 572 377
1 Not projected to 305-d basis.
CONCLUSION
There is adequate genetic variance in Zimbabwean Holstein herds to make large genetic gains in milk through conventional selection methods. Environmental effects such as age at calving follow patterns similar to those of more temperate areas.
ACKNOWLEDGEMENTS
First, the authors thank the Dairy Services Institute (ZDHIA) of Zimbabwe for providing records for this study. Secondly, this work was done while the senior author had an African American Institute (AAI)'s AFGRADIII scholarship. Finally, the first author would like to thank the University of Zimbabwe for granting him the study leave to do Ph.D. studies.
REFERENCES
Boldman, K.G., Kriese, L. A., Van Vleck, L. D. and Kachman, S.D. (1993) User Notes, Version 3.0, University of Nebraska, Lincoln, Nebraska, USA.
Makuza, S.M. (1988) M.S. Thesis. Michigan State University, East Lansing, Michigan, USA.
Makuza, S.M., and McDaniel, B. T. (1994) Proc. 5~' World Congr. Genet. Appl. Livest. Prod.Guelph, Ontario, Canada.
Makuza, S.M. (1995) PhD Dissertation. North Carolina State University, Raleigh. NC, USA.
Mao, I.L. (1984) Proc. of the Natl. Invitational Workshop on Genetic Improv. of dairy cattle. Milwaukee, Wisconsin. pp 25.
McDowell, R. E. (1972) W. H. Freeman and Co. San Francisco, CA, USA.
N.R.C. (1981) National Research Council, National Academy Press, Washington, D.C.
Tabbaa, M. J. (1993) PhD Dissertation. North Carolina State University, Raleigh., U.S.A. ENVIRONMENTAL FACTORS AFFECTING MILK PRODUCTION OF A HOLSTEIN FRIESIAN HERD IN SOUTHERN MALAWI
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSION
ACKNOWLEDGEMENTS
REFERENCES
M. G. Chagunda1, C. Wollny1,2, F. Ngwerume3, L. A. Kamwanja1and T. P. E. Makhambera1 1Department of Animal Science, Bunda College of Agriculture, P 0 Box 219, Lilongwe, Malawi. 2Deutsche Gesellschaft fuer Technische Zusammernarbeit, D-65726 Eschborn, Germany. 3Department of Animal Science, University of Zimbabwe, P 0 Box MP 167, Harare, Zimbabwe.
SUMMARY
The effects of environmental factors on daily milk yield were investigated using weekly milk records of 239 primiparous Holstein Friesian cows of Canadian origin kept in Southern Malawi between 1985 and 1994. A linear model considering fixed effects of season of calving, stage of lactation, year of production, season of test and age at calving as a covariate and random effects of cow was used. Daily milk yield showed a significant variation with season of calving (p < .001). Cows calving in the hot-wet season had a higher averaged daily milk yield of 16.3 kg/d compared to those that calved either in the cold-dry season (14.1 kg/d) or hot-dry season (15.2 kg/d). Differences due to season of test were significant (p < .001). Daily milk yield was higher for cows tested between May and August. Peak milk yield of 18.8 kg/d was attained at 65 days in milk. Daily milk yield also varied with year of production (p < .001). Average daily milk yield decreased in the herd from 22.4 _ 0.6 kg in 1985 to 17.0 t 0.4 kg in 1987 and levelling off in the subsequent years (ranging from 11.6 _0.4 to 15.8 _ 0.4 kg). This study shows the importance of adjusting for environmental effects when computing genetic values for the Malawian dairy cattle.
INTRODUCTION
Dairy farming in Malawi is practised on large scale and smallholder dairy farms. Large scale parastatal farms were established in the early 1980's under the Malawi-Canada Dairy Development Project. Holstein Friesian heifers were imported from Canada to the parastatal farms (then project farms). Breeding in these herds is done using Holstein Friesian semen also from Canada (Malawi Government 1979). Ndata Farm is situated about 35 km to the south east of Blantyre in the Southern region of Malawi. Mean annual temperature in this area is 22°C. The warmest month is October, with an average maximum temperature of 28°C and the coolest month is July with the average maximum temperature of 17°C. Rainfall averages from 1000 mm to 1125 mm per year and is confined to the period from early November to April (Malawi Government, 1979). As one of the three dairy farms established under the Malawi Canada Dairy Development Project, Ndata farm was established to provide Malawi with a foundation herd of Holstein Friesians.
The major sources of variation in milk production are genotype, environment and the interaction between the two. The influence of environmental factors on dairy production has been well documented. Thorpe et al. (1993) showed the effects of season of calving on production performance of dairy cattle in Kenya. Cows born or calving during the season preceding the marked dry season (January to March) had significantly delayed age at first calving (AFC) and longer calving interval (CI). Lactation commencing January to March or during the preceding season were shorter and had significantly lower average daily milk yield (DMY) and lactation milk yield (MY) as compared to lactations beginning in the other seasons. Studies to identify the nongenetic factors that influence milk yield of the dairy cattle population in Malawi have not yet been conducted.
The objective of this study was to investigate the effects of the environmental factors on milk production of Holstein-Friesians in Malawi.
MATERIALS AND METHODS
Data source and Environment Data consisting of 11345 weekly milk records were supplied by Ndata Dairy Farm which is one of the three farms belonging to a parastatal organisation, Malawi Dairy Industries. The data were daily, morning (AMMY) and afternoon (PMMY) milk records of 239 primiparous Holstein Friesian cows calving between 1984 and 1994. Data were edited using the following criteria: i. Age at first calving was restricted to 18-45 months. ii. Cows without milk production records were dropped. iii. Cows with AM or PM milk yield greater than 25kg were dropped.
Statistical analysis
The following mixed model was used for the univariate analysis of DMY:
2 yijklmn = µ+ Ci + SCj + STGk+ YPl + TMm, + bl(AFC) + b2(AFC ) + ijklmn where: yijklmn, = Testday milk yield;
µ = overall mean;
2 Ci = random cow effect with C distributed as N(0,c );
SCj = fixed effect of season of calving with j = 1,2,3;
STGk = fixed effect of stage of lactation on test day with
K = 1,2,3,...,48; Yp1 = fixed effect of year of production with 1= 1,2,3,10;
TMm = fixed effect of season -of test with m = 1,2,3,6;
(AFC)ijklmn =Age at first calving as a covariate with b1, and b2 as the linear and
quadratic regression coefficients, respectively;
2 ijklmn = random residual with s distributed as N(0,e ).
Stage of lactation was formed by 5-day intervals up to 150 days in milk and then 10 day intervals up to 305 days in milk. Three seasons of calving were defined as hot-wet (November to April), cold-dry (May to July) and hot-dry season (August to October). Season of test was defined by 2 months intervals starting with the January-February season. The Statistical Analysis System (SAS) (1988) was used for analysis. Random cow effects were absorbed.
RESULTS AND DISCUSSION
The average age at first calving for this herd was 28.0 months. Milk yield averaged 15.0 kg/day (Table 1). Afternoon milk yield (MYPM) tended to have a higher CV (31.5%) compared to both morning milk yield (MYAM) and DMY (28.2% and 28.5% respectively).
All the non-genetic factors included in the model had a significant influence on DMY yield The model explained 63% of the total variation in DMY. Both linear and quadratic effects of age at calving were significant (p < .05). Effects of age at calving on milk yield have been well documented. Milk yield increases with age at a decreasing rate up to maturity. Yield then decreases as cows become older.
Mean, SD, and CVfor daily yield and age at first calving
Trait Mean SD CV% MYAM (kg) 8.0 2.3 28.2 MYPM (kg) 7.0 2.2 31.5 DMY (kg) 15.0 4.3 28.5 AFC (mo) 28.0 3.2 11.4
Season Effect
Both season of calving and season of test significantly affected DMY (p < .001). Daily milk yield for cows that calved in the hot-wet season was higher than that of cows that calved during the cold-dry and hot-dry seasons (p < .01) as shown in Table 2. Table 3 presents the effects of season of test on DMY. Cows tested from May through August showed significantly higher milk yields as compared to cows that were tested outside this period (p < .001).
Table 2. Effect of season of calving on average daily milk yield at Ndata Farm
Season n DMY SE Hot-wet 7118 2036 16.3a 0.4 Cold-dry 2191 14.16 0.8 Hot-dry 15.26 0.1
Different superscripts denote significant differences (p < .001) Table 3. Effect of season of test on daily milk yield at Ndata Farm
Season n DMY SE
Jan - Feb 1742 14.7 a 0.2
Mar - Apr 1836 14.6 a 0.2 May - Jun 1954 15.2b 0.1 Jul - Aug 1986 15.2b 0.1 Sep - Oct 1977 14.3a 0.1 Nov - Dec 1850 14.2a 0.1
Different superscripts denote significant differences (p < .O1)
The effects of season detected in this study concur with earlier reports by different workers. (Ohashi et al., 1987; Murdia and Tripathi, 1991; Stanton et al., 1992; Thorpe et al., 1994). Ohashi et al. (1987) showed the variation of lactation curves with season of calving. The effects of season of test on lactation curves was also reported by Stanton et al. (1992). Thorpe et al. (1994) reported that seasonal effects significantly influenced milk yield of crossbred Ayrshire and Sahiwal cattle and preweaning growth rate of progeny in the semiarid tropics of Kenya. The relationship of milk yield with season of calving or test is in part caused by the seasonal variations in feeding and care. The quality and quantity of feed or pasture seems to be of particular importance.
State of lactation
Daily milk yield showed a significant variation with stage of lactation (p < .001). Milk yield gradually increased from about 13kg/d during early lactation to a peak of 18.8 kg/day. Peak milk production was attained at 65 days in milk. After peak, milk yield gradually decreased to 12.5 kg/day at the end of lactation.
Year of Production effect
Table 4 shows the effect of year of production on DMY. Daily milk declined from 22.4 ± 0.6 kg in 1985 to 11.6 kg in 1991. The decline in DMY could be due to changes in general management, feeding regime and other environmental factors.
Table 4. Effect of year of production on average daily milk yield (DMY) at Ndata Farm
Year N DMY SE 1985 163 22.4 0.6 1986 997 18.1 0.5 1987 1236 17.0 0.4 1988 1288 15.8 0.4 1989 1142 13.8 0.4 1990 1855 13.5 0.4 1991 1162 11.6 0.4 1992 1287 14.2 0.4 1993 1677 13.5 0.5 1994 538 14.9 0.6 CONCLUSION
The results indicate that, first parity milk yield of a Holstein Friesian herd performing in a semi- arid climate is significantly influenced by environmental factors. In further analysis effect of environment factors on milk yield of multiparous cows will be investigated. 1n ancillary studies, therefore, environmental factors such as feeding and health management should be investigated in more detail and quantified. Furthermore, there is need for comparative studies on the Holstein Friesians and their crosses with indigenous cattle of Malawi to determine their biological and economic efficiency. Further evaluation of the herd should focus on possible genotype x environmental effects.
ACKNOWLEDGEMENTS
The authors would like to thank Malawi Dairy Industries Corporation (MDI) and staff','( Ndata Farm for their fruitful cooperation. Their gratitude also goes to the Department of Animal Science, University of Zimbabwe for assisting in part of the analysis. The financial contribution by SACCAR/GTZ to this work is highly appreciated.
REFERENCES
Malawi Government Report. (1979) Ministry of Agriculture and Livestock Devclopmcnt, Lilongwe, Malawi.
Murdia, C.K. and Tripathi, V. N. (1991) Indian Veterinary Journal 68: 1139-1142.
Nzima, J. (1985) MPhil Thesis, University of Reading, United Kingdom.
Ohashi, T., Katayama, H., Yamanchi, K., Haga, S. and Nakamura, N. (1987) Japanese Journal of Dairy and Food Science 36: 191-195.
Ptak, E., Horst, H. S. and Schaeffer, L. R. (1993) J. Dairy Sci 76: 3792-3798.
SAS Institute Inc. (1988) 6.03 Edition. SAS Institute Inc., Cary, NC, USA.
Sobczynska, M. and Dymnicki, E. (1992) Polish Academy of Sciences, Institute of Genetics and Animal Breeding, Jastrzebiec No.8 39-46.
Stanton, T.L., Jones, L. R., Everett, R. W. and Katchan, S. D. (1992) J. Dairy Sci. 75: 169 1- 1700
Thorpe, W., Kang'ethe, P., Rege, J. E. 0., Mosi, R. 0., Mwandoto, B. A. J. and Njuguna, f. (1993) J. Dairy Sci. 76: 2001-2012. UTILIZING TEST-DAY INFORMATION TO MONITOR INDIVIDUAL COW PERFORMANCE
SUMMARY
INTRODUCTION
UTILIZING TEST-DAY RECORDS
CONCLUSIONS
REFERENCES
F. N. Ngwerume University of Zimbabwe, Department of Animal Science, P 0 Box MP 167, Mt Pleasant, Harare, Zimbabwe.
SUMMARY
Test-day models describe testday milk production. The major advantage of test-day models is that they account for all systematic environmental factors that affect test day milk production thereby minimizing residual error variance. Such factors could be specific to a particular test day. Test-day models have a great potential in accurate genetic evaluations of dairy sires and dams and for monitoring individual cow performance. This paper gives an overview of test-day models.
INTRODUCTION
Dairy producers have a difficult time assessing production change over time. This, in part, is because the group of cows contributing to a day's production vary as cows freshen or dry off between periods being assessed and because physiological factors for other cows change between periods.
A cow's test-day yield is influenced by systematic environmental effects such as age and season of calving, stage of lactation, herd and number of days open. A standardization of test- day yields for all these effects except herd allows a comparison of cows within herd which is a useful tool for management and selection. In the Netherlands, for example, test-day yields of cows recorded prior to 250 days in milk are standardized for age and season of calving and stage of lactation. The standardized tests are averaged to give a herd index which is used as a management guide by the farmer (Wilmink, 1987).
Several studies have shown that the correlation between the different parts of the lactation curve is not one. This is means that the shape of the lactation curve is not the same for all individuals and suggests the effect of non-genetic factors on the lactation curve. Wood (1967) concluded that the lactation curve parameters were affected by season of calving, season of production, parity and the interaction between these factors. It is therefore important to account for these non-genetic effects in models for genetic evaluations or when extending partial records to 305-day production yield by either including such factors in the models simultaneously or by adjusting the records for such factors using their respective correction factors.
Non-genetic factors can be partitioned into two categories, namely, temporary environmental effects and permanent environmental effects. A temporary environmental effect is one which influences only a single observation on the individual. Whether or not an individual receives a particularly favorable or unfavorable influence is simply by chance for each observation. Permanent environmental effects are those which influence all observations made on an individual. For example, the feeding regime used to raise dairy heifers, if extreme (poor feeding or excessive energy) can influence mammary development, hence becoming a permanent environmental effect influencing all lactations.
The general model for the phenotypic expression of a quantitative trait is
Pij=,+Gi+PEi+TEij where
Pij is the jth record of the ith animal is a constant level of performance for all animals which can be thought of as the average value that all animals in the population have in common plus the average level of management,
Gi is the sum of the genetic values,
PEi is the sum of effects of environmental factors which permanently influence the performance of animal i,
TEij is the sum of random environmental effects which affect only the jth record of animal i and are thus temporary.
The term , can be thought of as representing major identifiable non-random environmental effects such as herd, management level, age of the animal at calving, the year and the season when the record was made. In that sense might be different for each animal and each record of each animal. If can be easily accounted for then it is PEi and TEij that would mask the evaluation of Gi.
UTILIZING TEST-DAY RECORDS
In the genetic evaluation of dairy sires and dams, 305-d production is used. Methods have been developed to predict 305-d yield from records in progress. Prediction of 305-d production is not without error. There is still a quest for improved methods for extending part lactations to 305d production. Trus and Buttazzom (1990) proposed a method that describes the lactation curve as a series of correlated traits. This model predicts the residuals for each trait that can be added to the expected values to estimate a missing test-day record which can then be summed with other testday yields to give lactation yield. This method is currently being used in Italy.
Major emphasis is placed on standardized lactation production when selecting dairy cattle. Summarizing test-day records into a single measure is a common practice. However, adjusting this cumulative record for environmental effects such as herd, season and age of calving eliminates the possibility of adjusting for those systematic effects peculiar to individual test-day records. With 305-d yields such effects which are test-day specific are assumed to be random and to average out over the lactation. These effects may be quite different from the average effects for the lactation, hence they may not quite average out (Meyer, 1989; Stanton et al., 1991). Meyer (1985) reported low heritabilities for milk (.17), fat (.15), and protein (.13) yields. These low values were partly attributed to short-term environmental variation affecting daily performance which could not be accurately accounted for by modelling lactation totals.
Modelling individual test-day records for both genetic evaluations and management purposes might eliminate some of the problems of extending records to 305-day yield, as well as the problems associated with accurately modelling 305-day yields. When modelling individual testday records, a linear model that is assumed to explain test-day records is important. This model shall be referred to as a "Test-day Model". By definition, a test-day (TD) model is a method of evaluating daily production of milk, fat, protein and somatic cell count considering effects for each test-day in place of one set of fixed effects over the 305-day lactation. A TD model would need to incorporate the general shape of the lactation curve (Wood, 1967; Trus and Buttazzoni, 1990; Stanton et al. 1991) and accurately account for the test-day environmental effects affecting all cows on the same test-day and effects specific to a particular cow on a given test-day such as days carried calf, days open and disease. Analyzing test-day records may provide a valuable tool for herd management as well as genetic evaluations. In terms of management, dairy producers are interested in accurate evaluation of their feeding and management practices so their best programs can be repeated On the other hand, geneticists desire accurate estimates of these environmental effects and management programs so that they can be eliminated or properly adjusted for when evaluating animals for breeding purposes.
Meyer (1989) used a TD model to compute genetic parameters using test-day records of first lactation cows. In this study, test-day records were split into 30-day intervals and yield in each interval was analyzed separately by either using a model with herd-year-season (HYS) subclasses or a model with herd-test-day effects (HTD). Fitting HTD effects which accounted for the environmental effects specific to the day of test reduced residual variances as compared to fitting HYS. The proportion of total sums of squares for milk yield explained by HTD ranged from 36% to 86% for the three regions. Ptak and Schaeffer (1992) used a test- day model for genetic evaluation of 576 sires. The breeding values for the same sires were also estimated using 305-day lactation yield. The rank correlations between the two methods was ranged from .889 to .96. Although these results do not suggest which method is better, at least the results show that using a test-day model ranks the animals differently. However, Ptak and Schaeffer showed that when using HTD effects in the model, residual variances were greatly reduced as compared to adjusting for HYS effects only. Further research using simulated records is needed.
Everett and Schmitz (1994) developed a within herd test-day model which corrects for age on test-day, days carried calf, days in milk, month of calving and herd test-day milk on an individual herd basis. In this model an auto-correlation structure was assumed for the residual (co) variance matrix structure. The advantages of the test-model developed by Everett and Schmitz (1994) over the conventional 305-d production models is that it permits age and DIM effects to vary by herd and includes a herd-testday effect that adjusts for differing effects of sampling dates. Since this is a fixed effect model, the residuals are summed up and used for genetic evaluation.
Ngwerume (1994) developed a multitrait testday (MTTD) animal model that incorporated temperature hunudity index (THI) and season of test as factors affecting testday milk yield. Including THI reduced residual error variances. In the same study, a method of predicting a cows next test-day production was developed. The method compares test day production of different cows under the same test day conditions after adjusting for age at calving and days in milk. The MTTD animal model method is more accurate than utilizing extended 305-d yield to compare cows within a herd. CONCLUSIONS
Advantages of modelling actual test records using test-day models can be summarized as follows: i. Methods to project records to 305-day yield will not be necessary. ii. If cows are grouped according to production on a test-day, such grouping, if known, can be included in the models describing test-day records. iii. For genetic purposes, cows can be evaluated as long as they have at least one-test-day measurement. iv. The use of BST, if recorded, can be accounted for as an effect on a specific test-day. v. Comparison of performance of cows within herd based on test-day will be more accurate as animals will be compared on the same test-day and as such would have experienced the same environment. vi. Using a TD model would account for variable amounts of information from different lactations. vii. TD models permit estimates of fixed effects to vary across herds, stages of lactation. viii. Differing effects of sampling date can be accounted for. viii. A test-day model has the potential to reduce residual variances which may lead to better genetic estimates.
The disadvantages of using test-day yields would be: the need to adjust for days in milk; the need to store all of the individual test-day yields on a cow; the computation of genetic evaluations may take more time due to the increased number of observations (test-day yields) and more complex statistical models that might be used for test-day yields.
REFERENCES
Meyer, K., Graser, H. U. and Hammond, K. (1989) Livestock Prod. Sci. 21:177-199. Meyer, K. (1985) Biometrics. 41: 153-158.
Ngwerume, F. N. (1994) PhD Dissertation, Michigan State University, East Lansing, Michigan, USA.
Ptak, E. and Schaeffer, L. R. (1992) Unplished.
Stanton, T.L Jones, L.R., Everret, R.W. and Kachman, R.D. (1992) J. Dairy Scin.75: 161. 1700
Trus,D.and L. Buttazzoni, (1990) Proc. 4th World.congr.Genet.Appl. Livest.Prod. Edinburgh, Scotland.
Wilmink,J.B.M. (1987) Livestock Prod. Scin. 16:343-348.
Wilmink,J.B.M. (1987) Livestock Prod. Scin. 17: 1-17.
Wood, P.D.P. (1967) Dnim. Prod. 11:316-319.
APPLICATION OF BIOTECHNOLOGY IN GENETIC IMPROVEMENT,CHARACTERIZATIONAND CONSERVATION OF LIVESTOCK
SUMMARY
INTRODUCTION
REPRODUCTIVE PHYSIOLOGY
SYSTEMS OF SELECTION AND BREEDING
MOLECULAR CHARACTERIZATION OF ANIMAL GENETIC RESOURCES
CONSERVATION OF ANIMAL GENETIC RESOURCES
REFERENCES
J. E. 0. Rege International Livestock Research Institute P 0 Box 5689, Addis Ababa, Ethiopia.
SUMMARY
Among agricultural and allied fields, animal production and health have probably benefited the most from biotechnology. Successful application of biotechnology has generally been limited to developed countries. Specifically, there are hardly any success stories of the application of biotechnology in the improvement of livestock production in Africa. This paper reviews available biotechnologies with current and/or potential application in genetic improvement, characterization and/or conservation of domestic animal genetic resources and attempts to identify those technologies which have been, or may be, applied in developing countries.
INTRODUCTION
Developing countries are faced with the challenge to rapidly increase agricultural productivity to help feed growing populations while also sustaining the natural resource base. Biotechnology is regarded as a means to meet both objectives. It is generally recognized that biotechnology has a role to play in addressing the production constraints of small scale or resource-poor farmers who contribute more than 70 percent of the food production in developing countries.
Biotechnology can be defined as any technique that uses living organisms or substances from such organisms to make or modify a product, to improve plants or animals or to develop microorganisms for specific purposes. Biotechnology is not new. Man has used these techniques for thousands of years to manufacture products such as beer, wine and bread. Conventional plant and animal breeding which involve selection and mating of phenotypically preferred individuals is a good example of age-old application of biotechnology. What is new about biotechnology has come about as a result of more recent breakthroughs such as recombinant DNA technology and associated techniques, monoclonal antibody techniques, embryo manipulation technology, etc. These have enhanced possibilities for manipulating biological systems for the benefit of mankind.
REPRODUCTIVE PHYSIOLOGY
One of the challenges for genetic improvement is to increase reproduction rates. Several reproduction techniques are available. The commonest of these are artificial insemination (AI), embryo transfer and associated technologies. Measurement of progesterone in milk or blood which is a widely used technique for monitoring ovarian function and for pregnancy tests is also an important technology for managing the reproductive function of the animal.
Artificial insemination
No other technology in agriculture except hybrid seed and fertilizer use has been so widely adopted at a global scale as AI. Progress in semen collection, dilution and cryopreservation now enables a single bull to be used simultaneously in several countries for up to100,000 inseminations a year (Gibson and Smith, 1989). This implies that a very small number of top bulls can be used to serve a large cattle population. Additionally, each bull is able to produce a large number of daughters thus facilitating accurate progeny testing of bulls. The high intensity and accuracy of selection arising from AI can lead to a four-fold increase in the rate of genetic improvement in dairy cattle relative to that from natural mating (Van Vleck, 1981). This aspect is addressed fur ther in the next section.
In addition to its effect on intensity and accuracy of selection, Al facilitates a wider and rapid use of selected males thereby accelerating the rate of improvement. Additionally, use of AI can reduce transmission of venereal diseases in a population, reduces the need for farmers to maintain their own breeding males, facilitates more accurate recording of pedigree and is a cheaper means of introducing improved stock. However, success of AI technology depends on accurate heat detection and timely insemination. The former requires a certain level of awareness among farmers while the latter is dependent on a good infrastructure including transport network and availability of reliable means of transport.
AI is one of the most widely available biotechnologies in developing countries. Howcver, its use in these regions, particularly in Africa is far much less than in the developed countries. Although the technology is available in most African countries, it has remained generally unexploited, being only used for "exploratory" purposes mainly by research institutions. A few countries have taken the technology to the field, mostly for programmes of "upgrading" indigenous stock and as a service to a limited number of commercial farmers keeping exotic dairy cattle breeds. Countries such as Nigeria, Ethiopia, Uganda, Ghana, Botswana, Malawi, Senegal, Mali and Sudan, among others, are in this category. However, there are also countries which have used the technology more widely. Kenya and Zimbabwe, for example, have more elaborate AI systems which include national insemination services incorporating progeny testing schemes. However, even these have gone through periods of collapse or serious degeneration and have had to go through "rehabilitation" phases. The Republic of South Africa is probably the biggest user of AI technology in terms of number of inseminations. Additionally, the country has what is perhaps the best organised progeny testing scheme in the continent.
Although AI technology is available for other domestic livestock species, its use is still more generally associated with dairy cattle. The limitation of AI use in beef cattle has mainly been due to the diffculty in detecting heat in large beef herds kept on ranches and where individual cows are handled only occasionally. 1n sheep and goats there is scope for improvement of the tcchnology. The failure to develop a simple, non-surgical insemination procedure has prevented extensive exploitation of the technology in sheep (Robinson and McEvoy, 1993). However, the technical success of laparoscopic intrauterine insemination has prompted research into less invasive transcervical procedures (Halbert et al., 1990; Buckrell et al., 1992). In Africa, research to improve the freezing-and-thawing properties of sheep semen is underway in the Republic of South Africa. AI in pigs is hampered by the inability to successfully cryopreserve boar semen.
AI is credited for providing the impetus for many other developments which have had a profound impact on reproductive biotechnology. Foote (1982) noted that studies of ocstrus detection and ovulation control which evolved out of a need to correctly time inseminations, led to the development of embryo transfer technology.
Embryo transfer
Although presently not economically feasible for commercial use on snrrll farms, embryo technology can greatly contribute to research and genetic improvement of local breeds. Advances in this area are mainly applicable in cattle. There are two procedures presently available for production of embryos from donor females. One consists of superovulation, followed by A I and then flushing of the uterus to gather the embryos. The other, called in vitro fertilirrtion (IVF) consists of recovery of eggs from the ovaries of the female then maturing and fertilizing them outside the body until they are ready for implantation into foster fcmales. I VF facilitate recovery of a large number of embryos from a single female animal. This reduces the cost of embryos, thus,making ET techniques economically feasible on a larger scale. Additionally, 1VF makes available embryos at the early stage suitable for such manipulations as cloning.
The principal benefit of embryo transfer is the possibility to produce several progeny from the female, just as AI produces many offspring from one male animal. For example the average lifetime production of a cow can be increased from 4 to 25 calves. Increasing the reproductive rate of selected females has the following benefits: Genetically outstanding animals can contribute more to the breeding programme, particularly if their sons are being selected for use in AI; the rate of genetic change can be enhanced if specially designed breeding schemes arc set up, which take advantage of increased intensity of female selection combined with increased generation turnover; transport of embryos is much cheaper than live animals; risk of importing diseases is avoided; facilitates rapid expansion of rare economically important genetic stocks; the stress to exotic genotypes can be avoided by having them born to dams of local breeds rather than importing them as live animals.
Embryo transfer is still not widely used despite its potential benefits. In developing countries this is mainly due to absence of the necessary facilities and infrastructure. However, even in developed countries commercial embryo transfer is still used only in specialized niches or for a small proportion of best cows in the best herds. This is mainly a cost consideration. Thus, in North America and Europe, only about one out of 500 calves born in the last decade was from ET (Seidel and Seidel, 1992). The great majority of commercial embryo transfer is with cattle. This is mainly because ET is relatively easier in cattle than the other species and also because it is more economical in cattle lie cattle are worth more). Additionally, the low reproductive rate and the long generation interval of cattle make ET much more advantageous in the species.
Embryo transfer is of particular importance in research: Production of several closely related and hence genetically similar individuals can make critical contribution to research. For example, a project at the International Livestock Research Institute (ILRI) to locate the genes responsible for tolerance of some cattle populations to trypanosomiasis required large numbers of closely related crosses of trypanotolerant and trypanosusceptible cattle. Use of ET has made it possible to generate such families thereby facilitating the search for genetic markers of trypanotolerance. Additionally, ET could be useful in studying the extent to which a trait is influenced by the embryo (direct component) or the reproductive tract (maternal component).
Embryo sexine and cloning
Although embryo sexing may not have dramatic effects on rates of genetic gain (eg Colleau, 1991; Kinghorn et al., 1991) it can have considerable increases in efficiency. Taylor et al (1985) concluded from a study that an all-female heifer system using ET was 50% more efficient than the highest achievable in a traditional system. It has been suggested that, if multiple sexed-embryo transfer became a routine operation such as AI, beef operations based on this system could become competitive with pig and poultry production in terms of efficiency of food utilization:
Clones may be produced by embryo splitting and nuclear transfer (Macmillan and Tervit, 1990). These offer possibility for creating large clone families (Woolliams and Wilmut, 1989) from selected superior genotypes which, in turn, can be used to produce commercial clone lines (Smith, 1989). However, some studies have concluded that cloning of embryos will not increase rates of genetic progress in the nucleus, but that it offers considerable advantages in increasing rate of dissemination of superior testing genotypes in commercial populations leg Woolliams, 1989). Other potential applications of cloning include efficient evaluation of genotype x environment interactions and testing and/or dissemination of transgenics. From a research standpoint, production of identical siblings should, by eliminating variability among animals, greatly reduce the size and hence the cost of experiments.
Hormonal intervention
As has been pointed out, use of hormonal assays to monitor reproductive function can be a rewarding practice for both research purposes and commercial livestock operations. However, reproduction can also be manipulated using hormonal treatments. Although hormonal treatments have produced desirable results in some studies in Africa leg Aboul-Naga et al., 1992), lack of awareness about their use and the fact that they are not economically viable under most prevailing production circumstances limit their use. Progesterone and PMSG treatment and immunization against androstenedione increased ovulation rate in Ossimi sheep and exogenous melatonin treatment of barren Rahmam ewes resulted in increased proportion of ovulating ewes and a higher ovulation rate (Aboul-Naga et al., 1992). However, these responses did not result in increased litter size because of increased ova wastage. Thus, in addition to the impracticability arising from prohibitive prices of hormonal preparations and problems with hormonal administration at farm level, there are other technical problems with these technologies. Indeed, technologies aimed at increasing litter size in traditional small ruminant production systems should not be applied unless management, including nutrition, can be improved in concert to ensure the survival of the additional progeny.
Reproduction can also be manipulated without application of exogenous hormones. Hassan et al. (1988) have reported that exposing ewes to rams one week prior to mating ("the ram effect") increased the percentage of ewes in oestrus (and hence the percent mated) by 27%. Such management approaches offer practical options for increasing annual lamb (or kid) production in situations where other technologies are either not available or not appropriate. "Accelerated lambing" - increasing the number of lambings per year can also be used to increase annual productivity. However, in tropical and subtropical situations where animals depend on seasonally available natural pastures, this practice may not be feasible. Under such circumstances the reproductive cycle tend to be dictated by availability of feeds.
SYSTEMS OF SELECTION AND BREEDING Genetic improvement of livestock depends on access to genetic variation and effective methods for exploiting this variation. In developed countries, breeding programmes are based upon performance recording. These programmes have led to substantial improvements in animal production. Developing countries have distinct disadvantages for setting up successful breeding programmes: Infrastructure needed for performance testing is normally lacking because herd sizes are normally small and variability between farms, farming systems and seasons are large; reproductive efficiency is low, due mainly to poor nutrition, especially in cattle; and communal grazing preclude implementation of systematic breeding and animal health programmes.
Animal evaluation technology
One of the areas in which artificial insemination has made a major contribution is in sire evaluation: Progeny testing programmes in conjunction with AI have provided exceptional opportunities to improve the accuracy of evaluation of dairy sires, especially in developed countries. AI facilitates simultaneous use of a sire in many herds (and on many dams) thereby providing reliable prediction of the genetic potential of the sire as reflected in performance of daughters.
From a genetic improvement standpoint AI has made impact both in terns of improved accuracy of predicting the genetic worth of sires and in facilitating rapid propagation of superior sires. A critical step in progeny testing is the prediction of an animal's genetic worth or breeding value from the progeny information collected in different environments. Breeding value of an animal is defined as the value of the animal as judged by the mean value of its progeny (eg Falconer, 1981). Application of this concept in selection of animals predates its formal definition and theoretical articulation of its estimation:
"The bull gives no milk, of course, yet will not a bull descended from several generations of high producing dams produce, wizen mated with a highly productive cow, calves which possess this characteristic to a still higher degree?" - Bergen, 1780.
Whether selection is based on individual's own performance or on the performance of relatives or combinations of these, the underlying principle is the same: only the best animals animals with higher breeding values should be selected. Throughout the history of animal breeding, especially diary cattle breeding, there has been a continuous evolution in the complexity of methods for animal evaluation ie breeding value estimation. The amount and type of information used in evaluation has increased from the use of physical appearance of the animal, through recording of animal performance and inclusion of records on selected traits, to inclusion of records of relatives. The substantial, but variable effects of environmental factors on tile underlying genotype has necessitated development and continual refinements of quantitative methods for accounting for (and removing) the environmental noise when estimating breeding values.
This century has witnessed significant developments and refinements in techniques for breeding value estimation. Major contributions include those by Lush (1931a, b; 1933; 1935; 1944 and 1949), Henderson (1953; 1966; 1973; 1975a, b; 1976); Henderson et al (1959); Hcndcrson and Quaas (1976), among others. Henderson's (1973) formal derivation of the mixed model equations and elucidation of the desirable properties of best linear unbiased prediction (BLUP) procedure was particularly instrumental in the development of this technology. The pace in developments in breeding value estimation procedures was accelerated in the 1980s and carly 1990s by rapid developments in computer technology. More recent developments in computer programming strategies and capacity of computers have made it possible to use better, but more computationally demanding, procedures. For example, the individual animal model is currently used as the method of choice in animal evaluations in over 14 countries compared to only three in 1988 (INTERBULL, 1992).
The impact of selection on different livestock species has been quite variable (Table 1).The most systematic genetic improvement arising from selective breeding of animals identified through formal genetic evaluation has been witnessed in dairy cattle where annual genetic gains of1 % or more are achieved.
Table 1. Examples of annual rates of genetic change achieved in selection experiments (E) and in practice (P)
Species Trait Annual Where Number of Reference response achieved years (%) Poultry 5-9 week weight 4.1 E 10 Pym (1982) (Broilers) gain 6.5 P 20 Chambers et al 7-8 week weight (1981) gain Poultry (Layers) Number of eggs 1.7 P 10 Flock (1979) Pigs Low fat depth 2.1 E 12 Hetzer and Miller Index of 6 traits 1.8 P 7 (1972) Litter size 1.5 P 4 Mitchell et al. (1982) Bichard and Seidel (1982) Sheep Weaning weight 1.5 E 10 Pattie (1965) Index of 7 traits 1.2 P 13 Eijke (1975) Litter size 2.9 P 8 Hight et al (1975) Beef cattle Weaning weight 0.7 E 7-26a Barlow (1978) Yearling weight 0.6 E 9 Koch et al (1974) Yearling weight 0.3 P 13 Willham (1982) Dairy cattle Milk yield 2.2 E 13 Young and Miller Milk yield 2.0 E 14 (1982) Milk yield 1.0 P 12 Legates and Myers Milk yield 0.4 P ? (1972) Milk yield 0.07 P 19 Van Vleck (1977) Szkotnicki et al. (1978) Rege(1991 aFrom 7 experiments of varying durations
Source: Adapted (with additions) from Smith (1984)
Table 2 summarizes estimates of genetic gains achieved in practice in selected countries. Although these gains are small, they are continuous and cumulative. Thus, changes of 1 to 2% could amount to 10 to 20% in ten years. Moreover, the methods are reliable and hence associated with minimal risk. Additionally, when returns are considered at a national level, the cost of genetic improvement is small (Smith, 1978). Open nucleus breeding system and multiple ovulation embryo transfer
Multiple ovulation embryo transfer (MOET) is a composite technology which includes superovulation, fertilization, embryo recovery, short-term in vitro culture of embryos, embryo freezing and embryo transfer. Benefits from MOET include increasing the number of offspring produced by valuable females, increasing the population base of rare or endangered breeds or species, ex situ preservation of endangered populations, progeny testing of females and increasing rates of genetic improvement in breeding programmes. Genetic improvement of ruminants in developed countries has made much progress in the last 35 or so years through the use of large scale progeny testing of males. As has been pointed out, the general failure of extensive use of AI in developing countries has implied that progeny testing schemes cannot be operated with much success. In any case the gcncrally small herds/flocks and uncontrolled breeding in communal grazing situations preclude implementation of progeny testing. Smith (1988a) has suggested that the Open Nucleus Breeding System (ONBS) may be especially valuable for developing countries where the use of AI has been a failure due to the reasons given above.
Table 2. Examples of genetic gains (expressed as % of 1980 performance levels) obtained in practice in dairy cattle in selected countries
Country Breed Period covereda Average Trait Annual Phenotypic 1980° no. bulls tested gain (%)b per year Phenotypi Genetic average(kg) c Phenotypi Genetic c Germany German Holstein 1980-86 1975-86 1627 Milk 1.47 1.11 5143 Friesian yield Fat 2.31 1.46 203 yield Protein 1.13 1.01 174 yield Ireland Holstein/Friesian 1980-86 1975-86 37 Milk 5.50 1.02 3210 yield Fat 5.31 1.47 117 yield Protein 5.34 0.97 105 yield Israel Israeli Holstein 1980-87 1975-87 41 Milk 1.19 1.56 8178 yield Fat 0.74 0.77 262 yield Italy Black and White 1980-86 1975-86 67 Milk 2.24 2.13 5901 yield Fat 1.41 1.98 216 yield Protein 2.25 2.25 183 yield The Black and White 1978-88 1978-88 331 Milk 3.01 2.65 4778 Netherlands yield Fat 4.03 2.67 201 yield Protein 3.45 2.56 159 yield USA Holstein - 1980-86 1977-86 907 Milk 1.98 0.99 7871 yield Fat 1.89 0.87 283 yield Protein 1.40 0.78 250 yield Ayrshire 1980-85 1977-85 14 Milk 1.77 0.70 5808 Yield Fat 1.58 0.60 226 yield Protein 1.32 0.63 195 yield Jersey 1980-85 1977-85 47 Milk 2.12 0.90 5267 yield Fat 1.78 0.85 252 yield Protein 1.62 0.71 198 yield aPeriod when data used in estimating phenotypic and genetic gains were collected. Year for genetic data refer to bulls' year of birth while that for phenotypic data refer to year of calving of bulls' daughters. bAs % of 1980 adjusted production records cNumber of lactations included in the evaluations and hence the averages vary: Germany - 3; Ireland - 1; Israel - 6; Italy - all; The Netherlands - 3; USA - 5. USA evaluations also require first lactation data inorder for any lactation to contribute to evaluations.
Source: Calculated by author from sire evaluation information of INTERBULL (1992)
The ONBS concept is based on a scheme with a nucleus herd or flock established under controlled conditions to facilitate selection to be carried out. The nucleus is established from the "best" animals obtained by screening the base (farmers') population for outstanding fema1es. These are then recorded individually and best individuals chosen to form the elite herd or flock of the nucleus. If ET is possible, the elite female herd is used through MOET with superior sires to produce embryos which are carried by recipient cows from the base population. The resulting offspring are reared and recorded and the males among them evaluated using, as appropriate, performance of their sibs and paternal half sibs and their own performance. From these, an elite group of males with high breeding values for the specific trait is selected and used in the base population for genetic improvement through natural service or AI. It should be noted that, while MOET improves the rate of progress substantially, it is possible to operate an ONBS without ET technology, especially in species, such as small ruminants, with high reproductive rates. Such schemes are being tried for sheep in West Asia by FAO (Jasiorowski, 1990) and in Africa (cg Yapi et al., 1994). However, availability of Al and ET, in addition to increasing rates of genetic gain, enhance the flexibility of the system. For example, germplasm from other populations can be introduced easily through semen and/or embryos. One of the advantages of a nucleus herd is that it provides opportunity to record information on more traits than is possible in a decentralized progeny testing scheme. The ONBS can be used for the improvement of an indigenous or exotic breed. It can also be used to improve a stabilized crossbred population. The level of the genetic response depends on the size of the scheme (that is, number of participating herds or flocks and total number of animals) and the selection intensity. Additionally availability and effective use of Al will determine the impact of such a scheme, especially for dairy cattle. An ONBS can initially be developed to form a focus for national sire breeding activities. In time, and with experience, the capacity can be expanded and ET introduced to increase the rate of genetic progress.
At one time it was suggested that application of MOET in nucleus breeding schemes could increase annual genetic gains by 30-80% (eg Nicholas and Smith, 1983). More recently it has been concluded that the earlier figures were over-predictions (eg Keller et al, 1990). The over- prediction arose partly because the assumed average number of progeny (eight) per donor female was unrealistically high and partly because of wrong assumptions made about genetic parameters (Keller et al., 1990). The realistic average number of live progeny per donor flushed is in the range of 2-3 in sheep and cattle and 6-8 in goats (Macmillan and Tervit, 1990). Consideration of these figures suggest that MOET could increase annual genetic gains by 10-20% in large nucleus breeding schemes. However, costs of operating such schemes in developing countries need to be evaluated before they can be recommended.
Indicator traits
Indicator traits are characteristics which are genetically correlated to traits of economic importance and are easier to measure than the latter. Such traits are usually not the target of genetic improvements but provide an indirect means of improving a targeted trait. Blair et al (1990) have reviewed some physiological and/or metabolic characteristics which might be considered as potential indicator traits. Traits such as testicular size in rams or bulls or FSH in ewe lambs (Bodin et al., 1986) have potential as indirect predictors of fertility. Indicator traits can improve genetic response by increasing accuracy of selection and reducing generation interval. The value of an indicator trait will depend largely on magnitude of co-heritability (square-root of product of heritability of the indicator and of the target trait) and the genetic correlation between the two traits (Woolliams and Smith, 1988). Woolliams and Smith (1988) have concluded that, with high co-heritability, selection for the indicator trait alone can result in greater rates of response than is possible with progeny testing, especially when breeding values are not accurately measured by progeny testing.
Packed cell volume (PCV), an indication of the extent of anaemia, is widely used as an indicator trait for pathological conditions associated with anaemia. For example, PCV is currently used at International Livestock Research Institute (ILRI) as an indicator of the effect of trypanosomiasis and hence of trypanotolerance and as an indicator of effect of the endoparasite Haemonchus contortus and hence as an indirect measure of parasite resistance.
Genetic markers and marker assisted selection
A genetic marker for a trait is a DNA segment which is associated with, and hence segregates in a predictable pattern as, the trait. Genetic markers facilitate the "tagging" of individual genes or small chromosome segments containing genes which influence the trait of interest. A genetic marker need not have an effect on performance. Its value is simply that it 'marks' chromosome segments containing genes affecting performance. Markers have already been identified for some traits controlled by single genes. These include double muscling gene in Belgian Blue cattle, polled gene and Weaver syndrome in Brown Swiss cattle (see Archibald et al., 1994). Availability of large numbers of such markers has enhanced the likelihood of detection of major genes influencing quantitative traits. The method involves screening the genome for genes with large effect on traits of economic importance through a procedure known as linkage analysis (eg Paterson et al., 1988). The chances of major genes existing for most traits of interest and of finding them is considered to be high (Mackinnon, 1992). Thus, Archibald et al (1994) have considered that, for mapping of QTL (where the traits are controlled by a number of loci whose individual effects may not be greater, than environmentally induced variation) the availability of suitable pedigrees; rather than inadequacy of markers, is the limitation. The process of selection for a particular trait using genetic markers is called marker assisted selection (MAS). MAS can accelerate the rate of genetic progress by increasing accuracy of selection and by reducing the generation interval (Smith and Simpson, 1986). Meuwissen and Van Arendonk (1992) have reported a selection design (granddaughter design) for dairy traits in which MAS may yield 10-20% extra genetic progress. Zhang and Smith (1992; 1993) have shown that MAS, used in combination with phenotypic information, can increase the efficiency of selection by 10-20% over and above the efficiency of selection based only on best linear unbiased prediction (BLUP) procedures. However, the benefit of MAS is greatest for traits with low heritability and when the marker explains a larger proportion of the genetic variance than does the economic trait. Lande and Thompson (1990) suggest that about 50% additional genetic gain can be obtained if the marker explains 20% of the additive genetic variance and the economic trait has a heritability of 0.2. MAS also facilitates increased rate of genetic gain by allowing measurement in young stock thereby reducing generation interval.
Genome maps of the domestic species are already beginning to become available (see Archibald et al., 1994). This should accelerate identification of markers and their use in genetic improvement. In mice, more than 20 genes conferring resistance to infectious diseases have been mapped (see Visscher and Haley, 1995). Marker identification and use in domestic livestock should enhance future prospects for breeding for such traits as tolerance or resistance to environmental stresses, including diseases. Already, identification of carriers of genes for resistance and introduction of such genes into a population seems feasible for resistance against Trichostrongyhus colubriformis and Haemonchus contortus (Gogolin- Ewens et al., 1990). It should also be possible to eliminate factors predisposing sheep to Listeriosis or Salmonellosis (Blancou, 1990). As has been alluded to, research is currently underway at ILRI (International Livestock Research Institute) to identify genetic markers for tolerance to African trypanosomiasis in N'Dama cattle and for resistance to endoparasites in Red Maasai sheep. Marker technology should provide opportunity for selecting for resistance or tolerance to these important parasites and diseases.
Transgenic animals
A transgenic animal is an animal whose hereditary DNA has been augmented by addition of DNA, through recombinant DNA techniques, from a source other than parental germplasm. Transfer of genes or gene constructs allows the manipulation of individual genes rather than entire genomes. There has been dramatic advances in gene transfer technology in the last two decades since the first successful transfer was carried out in mice in 1980 (Palmiter et al., 1982; Jaenisch, 1988). The technique has now become routine in the mouse and resulting transgenic mice are able to transmit their transgenes to the offspring thereby allowing a large number of transgenic animals to be produced. Successful production of transgenic animals has so far been reported in pigs, sheep, rabbits and cattle. The majority of gene transfer studies in livestock have been carried out in the pig. Although transgenic cattle and sheep have been successfully produced, the procedure is still inefficient in these species (Niemann et al, 1994). Niemann and Reichelt (1993) have reviewed the problems associated with transgenic technology in cattle.
Transgenesis offers considerable opportunity for advances in medicine and agriculture. In livestock, the ability to insert new genes for such economically important characteristics as fecundity, resistance or tolerance to other environmental stresses would represent a major advance in the breeding of commercially superior stock. Indeed, disease resistance alleles are the classical example of candidates for transgenic technology (Smith et al., 1987). Another opportunity that transgenic technology could provide is in the production of medically important proteins such as insulin and clotting factors in the milk of domestic livestock. The genes coding for these proteins have been identified and the human factor IX construct has been successfully introduced into sheep and expression achieved in sheep milk (Clark et al., 1990). Moreover, the founder animal has been shown to be able to transmit the trait to its offspring (see Niemann et al., 1994). To date, the majority of genes transferred into sheep have been growth hormone encoding gene constructs. Unfortunately, in most cases the elevated growth hormone levels have resulted into a clinical diabetes situation leading to an early death of the transgenic sheep (Rexroad et al., 1990). Transgenic sheep have recently been generated which express the visna virus envelope gene (Clements et al.; 1994).
The first reports of the production of transgenic animals created a lot of excitement among biological scientists. In the field of animal breeding, there were diverse opinions on how the technology might affect livestock genetic improvement programmes. Some (eg Ward et al., 1982) believed that it would result in total reorganization of conventional animal breeding theory while others (eg Schuman and Shoffner, 1982) considered the technology as an extension of current animal breeding procedures which, by broadening the gene pool, would make new and novel genotypes available for selection. Application of the technology in animal improvement is still far from being achieved. However, consideration need to be given to its potential role in this field. Smith et al. (1987) have presented a comprehensive evaluation of strategies for developing, testing, breeding and disseminating transgenic livestock in the context of quantitative improvement of economic traits.
An important contribution of transgenic technology is in the area of basic research to study the role of genes in the control of physiological processes. The understanding of the molecular control of life process has important implications for both medicine and agriculture. For example, the generation (through mutation of an endogenous gene) of an organism which lacks a specific gene is a powerful tool to investigate the function of the gene product. This type of genetic analysis has been facilitated by the availability of in vitro cultures of embryonic stem cells from mice (eg Bradley, 1994).
Recent advances in in vitro technology (in vitro fertilization and maturation) will increase the number of zygotes available for gene transfer purposes. This, plus utilization of embryonic stem cell (Stice et al., 1994) and primordial germ cell (Stokes et al., 1994) technologies should enhance the efficiency of gene transfer in cattle and sheep considerably.
Maior genes in livestock
For the application of marker technology, and indeed, for transgenesis, it is necessary to be able to find genes or chromosome segments which explain a significant proportion of the variation in the trait of interest. Virtually all major genes currently known in livestock were first detected by "eyeballing", ie examination of recorded data. Methods for searching for major genes are increasingly getting more sophisticated. For example, statistical methods now exist for screening populations for major genes without the use of genetic markers (Hoeschele, 1988; LeRoy et al., 1989; LeRoy and Elsen, 1992; Knott et al., 1992). However, use of marker technology (see earlier) is perhaps the most rigorous method for searching for major genes. This involves evaluating large numbers of animals for the (quantitative) trait of interest and genotyping them for a number of markers (eg Soller et al., 1979). Table 3 presents some examples of major genes in livestock. Identification of genes with large effects (using modern gene mapping methods) will allow us to unravel the molecular basis of production differences in livestock. This should have important implications for genetic improvement. It is considered that the more recent technologies such as MAS will not replace existing animal breeding techniques; rather, they will complement them: That is, genetic evaluation methods which combine both known and unknown quantitative trait loci are likely to be the most effective.
Table 3. Examples of major genes in livestock
Species Major gene Main advantage Reference Pigs Halothane gene increased lean Fujii et al. ( 199 1 )
meat yield Moishan gene Increased litter Rothschild et al.
size (1994) Sheep Booroola, higher ovulation Montgomery et al.
hwerdale and rate (19)3); Piper w id
Javanese genes Bindon (1982);
Bradford (1980,);
Davis et al. (1991) Callipyge Higher lean meat Cockett et al. (1994)
muscling gelle yield Cattle Double muscling Higher lean meat Hanset and
gene yield Mlcllallx (1985a,b)
MOLECULAR CHARACTERIZATION OF ANIMAL GENETIC RESOURCES
Developing countries are endowed with the majority of the global domestic animal diversity in the form of landraces, strains or breeds. Some livestock breeds in these countries are in immediate danger of loss through indiscriminate crossbreeding with exotic breeds. The importance of indigenous livestock breeds lies in their adaptation to local biotic and abiotic stresses and to traditional husbandry systems. The indigenous genetic diversity constitutes a buffer against changes in the environment and is a key in selection and breeding for adaptability and production on a range of environments. However, most of these animal genetic resources are still not characterized and boundaries between distinct populations are unclear. In such cases breeds are defined on the basis of subjective data and information obtained from local communities. Reliance on these criteria as the basis for classification or development of improvement and/or conservation strategies may be misleading. Additionally, historical evidence is not always accurate, relying as it often does on subjective judgements. Archival research can reveal much about the original type of a breed or strain but it is molecular genetic evidence which is factual and precise. It is in this sphere that biotechnology has an important role.
Genetic uniqueness of populations is measured by the relative genetic distances of such populations from each other. Polymorphism in gene products such as enzymes, blood group systems and leukocyte antigens which have traditionally been used for measuring genetic distances are rapidly being replaced by polymorphism at the level of DNA, both nuclear (eg Jeffreys and Morton, 1987) and mitochondrial (eg Loftus et al., 1994) as a source of information for the estimation of genetic distances. The first DNA polymorphism to be used widely for genome characterization and analysis were the restriction fragment length polymorphisms (RFLP) (Southern, 1975) which detect variations ranging from gross rearrangements to single base changes. Minisatellites sequences of 60 or so bases repeated many hundreds or thousands of times at one unique locus within the genome have been used to generate DNA fingerprints typical of individuals within species (Jeffreys and Morton, 1987). Microsatellites (Weber and May, 1989) repeats of simple sequences, the commonest being dinucleotide repeats - are abundant in genomes of all higher organisms, including livestock. Polymorphism of microsatellites takes the form of variation in the number of repeats at any given locus and is generally revealed as fragment length variation in the products of PCR amplification of genomic DNA using primers flanking the chosen repeat sequence and specific for a given locus (Kemp and Teale, 1991). Ease of identification and of sequence determination (Moore et al., 1992) and need for only small amounts of DNA, are some of the advantages of microsatellites. Additionally, because microsatellite polymorphism can be described numerically, they lend themselves to computerized data handling and analyses (Teale etal., 1994). Microsatellites can be used in non-PCR systems in a similar way to numsatellite probes (Haberfeld et al., 1991).
Randomly amplificd polymorphic DNA (RAPD) (Williams et al., 1990) has been extensively used for genetic characterization of a wide range of organisms. The technique uses short (up to 10 base) primers to amplify nuclear DNA in the PCR. The procedure does not require knowledge of the sequence of DNA under study: primers are designed randomly. The basis of the polymorphism detected by this method is that products are either generated in PCR or not. Complete sequencing of the genome is the ultimate form of genetic characterization. Sequencing has traditionally been expensive and laborious, but with the advent of automated sequencing this is changing rapidly. However, sequencing is unlikely to be used as a technique of choice for genetic characterization.
Regardless of which method is used, the ultimate goal in genetic characterization for conservation is to obtain a measure of available diversity. Nei and Takezaki (1994) have reviewed statistical methods for estimating genetic distances and for constructing phylogenetic trees from DNA sequence data and have concluded that different analytical methods may produce different results. Tcale et al. (1994) have commented on what considerations should be made in using DNA polymorphism data for genetic distance estimation and have cautioned that great care his to be taken in selecting characterization methods and in interpreting the resulting data. While recognizing the importance of the uniparental mode of inheritance of mitochondrial DNA in detecting underlying population structure not discernible from analyses of nuclear DNA, Loftus et al. (1994) have concluded that mitochondrial DNA analysis may not be sufficient to resolve breed differences within Africa. MacHugh et al. (1994) have suggested that microsatcllite polymorphism may be more suitable when trying to discriminate between closely related populations.
CONSERVATION OF ANIMAL GENETIC RESOURCES
The terms conservation, preservation, ex-situ and in-.situ are used here according to the definition given by FAO (1992). There arc several ways, differing in efficiency, technical feasibility and costs, to conserve animal genetic resources. Developing and utilizing a genetic resource is considered the most rational conservation strategy. However, there are cases where ex situ approaches are the only alternatives. Ex-situ approaches include: Maintenance of small populations in domestic animal zoos; cryoprescrvation of semen (and ova); cryopreservation of embryos; and some combinations of these. Brem et al. (1989) have reviewed biotechnologies for ex situ conservation.
Cryopreservation of gametes, embryos or DNA segments can be quite effective and safe approach for breeds or strains whose populations are too small to be conserved by any other means. The safety of these methods has been demonstrated by background irradiation studies. For example, studies based on irradiation of mouse embryos exposed to the equivalent of hundreds of years of background mutation showed no detectable damage (Whittington et al., 1977).
Regeneration of offspring following transfer of frozen-thawed cmbryos has been successful for all ma ior domestic species, except buffalo (Tcale ct al., 1994). In cattle, the transfer of frozen- thawcd embryos is now a commercial practice and embryo survival rate after thawing can be as high as 80% with a pregnancy rate of about 50%. Cryopreservation of oocyfies followed by successful fertilization and live births have been achieved in the mouse. Cryopreserved bovine oocytes have been successfully matured and fertilized in vitro and zygotes developed to blastocyst stage (Lira et al.. 1991). These trends strongly suggest that long-term cryoprcservation of mammalian oocytes will be possible (Tcale et al.. 1994). Respective pregnancy rates of 58 and 50% for fresh and frozen-thawed in vitro produced embryos have been reported (Lu et al.\\, 1990) and calves have been produced from transfer of both split and frozen-thawed in vitro produced cmbryos.
Economic aspects of genetic conservation in farm animals has been assessed by Bren et al (19 84). The study concluded that costs of ex,situ live animal conservation wits moderate to high while costs of long term cryopreservation of gametes were low. Developments in genetic engineering, cryobiology, cell biology and embryology will provide techniques that may enhance our ability to preserve germplasm in vitro. Techniques such as transfer of DNA within and between species and the production of viable transgenic animals are far from practical application. However, biotechnology will certainly contribute newer and cheaper methods for preservation such as storage of catalogued DNA. At present, other than live animal and embryo preservation, the other techniques do not allow preservation of genomes in a form which can be reactivated in tolo at a later stage. but they permit the preservation of individual genes or gene combinations for possible future use.
Conservation of indigenous animal genetic resources should be one of the priority livestock development activities for developing countries. The critical importance of these resources to their owners in developing countries need not be emphasized. Their importance to developed countries is also becoming evident: There is an increasing importation of tropical germplasm by these countries. It is highly likely that these resources will become of increasing importance to the industrialized countries either as sources of unique genes or when environmental concerns necessitate change in production systems. Developed countries should, thus, assist in the conservation and development of these resources. Technology for cryopreservation of semen and embryo is sufficiently developed to be applied in developing countries. What is missing is financial support to implement conservation programmes. Such support has been provided for world-wide conservation activities for plant germplasm. There is also a strong case for support of animal genetic resources conservation.
REFERENCES
Aboul-Naga, A.M., Aboul-Ela, M.B. and Hassan, F. (1992) Small Ruminant Research 7: 151- 160.
Archibald, A. L., Burt, D. W. and Williams, J. L. (1994) In: Proc. 5th World Congress Genetics Appl. Livest. Prod., Guelph, Canada.
Barlow, R. (1978). Animal Breed. Abstracts 46: 469-486.
Bichard, M. and Seidel, C.M. (1982) In: 2nd World Congr. Genetics Applied Livest. Prod., Madrid, Spain.
Blair, H. T., McCutcheon, S.N. and Mackenzie, D.D.S. (1990) In: Proc. 4th World Congr. Gen. Appl. Livest. Prod., Edinburgh, Scotland. Blancou, J. (1990) Rev. Sci. Tech. Off. int. Epiz. 9: 641-659.
Bodin, L., Bibe, B., Blance, M.R. and Ricordean, G. (1986) Genetics, Selection, Evolution 18:55-6.
Bradford, G.E., Quirke, J.F., Sitorus, P., Inounu, I., Tiesnamurti, B., Bell, F.L.. Fletcher, I.C., and Torell, D.T. (1986) J. Anim. Sci. 63: 418-431.
Bradley, A. (1994) In: Proc. 5th World Congr. Gen. Appl. Livest. Prod., Guelph, Canada.
Brem, G., Graf, F. and Krausslich, M. (1984) Livest. Prod. Sci. 1: 65-68.
Brem, G. Brenig, B., Miiller, M. and Springmann, K. (1989) FAO Animal Prod. Health Paper No. 76. FAO, Rome, Italy.
Buckrell, B.C., Buschbeck, C., Gartley, C.J., Kroetsch, T., McCutcheon, W., Martin, J., Penner, W.K. and Walton. J.S. (1992) In: Proc. 12th International Congr. Anim. Reprod., The Hague, Holland.
Chambers, J.R., Gavora, J.S. and Fortin, A. (1981) Canadian J. Animal Sei. 61: 555-563.
Clark, A.J., Archibald, A.L., McClenghan, M., Simons, J.P., Whitelow, C.B.A. and Wilmut, 1. (1990) In: Proc. New Zealand Soc. Animal Production 50: 167-180.
Clements, J.E., Wall, R.J., Narayan, 0., Haver, D. Sheffer, D., Powell, A.M., Zink-, M.C. and Rexroad, C.E. (1994) Theriogenology 41: 180 (abstr).
Cockett, N.E., Jackson, S.P., Shay, T.L., Neilsen, D., Moore, S.S., Steele M.R., Barendse, W., Green, R. D., and Georges, M. (1994) In: Proc. National Academy Sciences USA 91(8):3019- 3023.
Colleau, J.J. (1991) J. Dairy Science 74: 3973-3984.
Davis, G.H., McEwan, J.C., Fennessy, P.F., Dodds, K.G., and Farquhar, P.A. (1991) Biology of Reproduction 44: 620-624.
Eijke, E.D., (1975) Acta Agric. Scandin. 25: 253-260.
Falconer, D. S. (1981) 2nd Ed., Longman, London and New York.
FAO. (1992) The management of global animal genetic resources. Proc. FAO Expert Consultation. FAO Animal Prod. Health Paper No. 104, FAO, Rome, Italy.
Flock, D. (1979) Genetic improvement of egg production in laying type chickens. In: Selection Experiments in Laboratory and Domestic Animals. Roberton, A. (ed), Commonwealth Agricultural Bureaux, Slough, UK.
Foote, R.H. (1982) Journal Andrology 3: 85-100.
Fujii, J., Otsu, K., Zorzato, F., De Leon, S., Khanna, VX., Weller, J.E., O'Brien, P.J. and Maclennan, D.H. (1991) Science 253: 448-451.
Gibson, J.P. and Smith, C. (1989) In: L.A. Babiuk, J.P. Phillips and M. Moo-Young, (eds), p203231. Pergamon Press, Oxford.
Gogolin-Ewens, K.J., Meeusen, E.N.T., Scott, P.C.. Adams, T.E., and Brandon, M.R. (1990) Rev. Sci. tech. Off. int. Epiz. 9: 865-896. Guillaume, J. (1976) World's Poultry Science Journal 32: 285-304.
Haberfeld, A.,Cahaner, A., Yoffe, 0. Plotsky, Y. and Hillel, J. (1991) . Animal Genetics 22: 299305.
Halbert, G.W., Dobson, H., Walton, J.S., Sharpe, P. and Buckrell, B.C. (1990) Theriogenology 33: 1231-1243.
Hanset, R. and Michaux, C. (1985a) Genetics selection Evaluation 17(3): 359-368.
Hanset, R. and Michaux, C. (1985b) Genetics Selection Evolution 17(3): 369-385.
Hassan, F.A. El-Nakhla, S.M., Aboul-Ela, M.B. and Aboul-Naga, A.M. (1988) In: Proc. I lth flit. Congr. Anim. Reprod. and AI, Dublin, June 24-26, 1988, Vol 4, p487.
Henderson, C.R. (1953) Biometrics 9:226.
Hemderson, C.R. (1966) Proc. Symp. on Estimating Breeding Values of Dairy Sires and Cows, Washington, D.C., USA.
Henderson, C.R. (1973) Proc. Animal Breed. Gen. Symposium in Honor of J.L. Lush. ASAS and ADSA, Champaign, Ill., USA.
Henderson, C.R. (1975a) Biometrics 31: 423.
Henderson, C.R. (1975b) J. Dairy Sci 58: 1910.
Henderson, C.R. (1976) Biometrics 32:69.
Henderson, C.R., Kempthorne, 0., Searle, S.R. and von Krosigk, C.M. (1959) Biometrics 15:192.
Henderson, C.R. and Quaas, R.L. (1976) J. Anim. Sci. 43: 1188.
Hetzer, H.O. and Miller, R.H. (1972) J. Animal Sci. 35: 730-751.
Hight, G.K., Gibson, A.E., Wilson, D.A. and Guy, P.L. (1975). Proc. 2nd Conf. New Zealand Federation Livcst. Breed. Groups. Massey University, Palmerston North, NZ.
Hoeschcle, I. (1988) Theor. Appl. Genet. 76: 31 1
INTERBULL. (1992) Bulletin No. 5 of International Bull Evaluation Service (A joint venture of IDF, EAAP and ICRPMA) Dept. Anim. Breed. Genet., Uppsala, Sweden.
Jasiorowski, H.A. (1990) In: Proc. FAO Conf. Open Nucleus Breeding Systems, Bialobrzegi, Poland, June 1989, p7-12.
Jacnisch, R. (1988) Science 240: 1468-1473.
Jeffreys, A.J. and Morton, D.B. (1987) Animal Genetics 18: 1-15.
Keller, D.S., Gearheart, W.W. and Smith, C. (1990) J.Anim. Science 68:1553-1561.
Kemp, S.J. and Teale, A.J. (I 991) Animal Genetics 22: 435.
Kingliorn, B.P., Smith, C. and Dekkcrs, J.C.M. (1991) J. Dairy Science 74: 611-622.
Knott, S.A., Haley, C.S. and Thompson, R. (1992) Heredity 68: 299. Koch, R.M., Gregory, K.E. and Cundiff, L.V. (1974) J. Amin Sci. 39: 459-470.
Lande, R. and Thompson, R. (1990) Genetics 124: 743-756.
Legates, J.E. and Myers, R.M. (1972) J. Dairy Sci. 55: 681-682.
LeRoy, P. and Eisen, J.M. (1992) Theor. Appl. Genet. 83: 635.
LeRoy, P.. Eisen. J.M. and Knott, S.A. (1989) Genetics Selection Evolution. 21:341.
Lim, J.M., Fukui, Y. and Ono, H. (1991) Theriogenology 35: 1225-1235.
Loftus, R.T., MacHugh, D.E., Ngere, L.O., Balain, D.S., Badi, A.M., Bradley, D.G. and Cunningham, E.P. (1994) Animal Genetics 25: 265-271.
Lu, K.H., Jiang, H.S., Wang, W.L. and Gordon, I. (1990) Theriogenology 33: 278(abstr.).
Lush, J. L. (1931 a) J. Dairy Sci. XIV: 209.
Lush, J.L. (1931b) Proc. 24th Annual Meeting Amer. Soc. Amm. Prod. November 1931. p51.
Lush, J.L. (1933) J. Dairy Sci. XVI: 501.
Lush, J.L. (1935) J. Dairy Sci. XVIII: 1.
Lush, J.L. (1944) J. Dairy Sci. XXV: 937
Lush, J.L. (1949) Proc. Eighth International Genetics Congress.
MacHugh, D.E., Loftus, R.T., Bradley, D.G., Sharp, P.M. and Cunningham, E.P. (1994) In: Proc. Royal Society of London, Series B 256: 25-31.
Mackinnon, M.J. (1992) In: Proc. Australian Assoc. Anim. Breed. Genet. 10: 245-251.
Macmillan, K.L. and Tervit, H.R. (1990) In: Proc. Australian Assoc. Anim. Breed. Genet. 8:9- 18.
Mcuwissen, T.H.E., and Van Arendonk, J.A.M. (1992) J. Dairy Sei. 75: 1651-1659.
Mitchell, C., Smith, C., Makower, M. and Bird, P.J.W.M. (1982) Animal Production 35: 215- 224.
Montgomery, G.W., Sise, J.A., Penty, J.M., Tou, H.M., and Hill, D.F. (1993) Animal Genetics 23:411-416.
Moore, S.S., Barendse, W. Berger, K.T., Armitage, S.M. and Hetzel, D.J.S. (1992) Animal Genetics 23: 463-467.
Nei, M. and Takezaki, N. (1994) Proc. 5th World Congr. Gen. Appl. Livest. Prod., Guelph, Canada.
Nicholas, F.W. and Smith, C. (1983) Animal Production 36: 341-353.
Niemann, H., Halter, R. and Paul, D. (1994) In: Proc. 5th World Congr. Gen Appl. Livest. Prod., Guelph, Canada. Vol 21, p339-346.
Niemann, H. and Reichelt, B. (1993) Reprod. Fert, suppl. 48:75-94
Palmiter, R.D., Brinster, R.L., Hammer, R.E., Trumbauer, M.E., Rosenfeld, M.G., Birnberg, N.C. and Evans, R.M. (1982) Nature 300: 611-615.
Paterson, A.H., Lander, E.S., Hewitt, J.D., Peterson, S., Lincoln, S.E., and Tanksley, S.D. (1988) Nature 335: 721-726.
Pattie, W.A. (1965) J. Exp. Agric Amm. Husb. 5: 353.
Piper, L.R. and Bindon, B.M. (1982) In: Proc. 1st World Congr. Sheep and Beef cattle Breeding, Vol 1 p395-400.
Pym, R.A.E. (1982) Selection for food efficiency in broiler chickens: recent findings. 24th British poultry Breeders Roundtable, Edinburgh. Lehmann Tierzucht, GMBH, Cuxshaven, Germany.
Rege, J.E.0. (1991) J. Animal Breed. Genetics 108: 424-433.
Rexroad, Jr, C.E., Hammer, R.E., Behringer, R.R., Palmiter, R.D. and Brinster, R.L. (1990) J. Reprod. Fertil., Suppl. 41: 119-124.
Robinson, J.J. and McEvoy, T.G. (1993) Animal Production 57: 335-352.
Rothschild, M.F., Jacobson, C., Vaske, D.A., Tuggle, C.K., Short, T.H., Sasaki, S., Eckardt, G.R. and McLaren, D.G. (1994) Proc. 5th World Congr. Gen. Appl. Livest. Prod., Guelph, Canada.
Schuman. R. and Shoffner, R.N. (1982) Proc. 2nd World Congr. Gen. Appl. Livest. Prod. Madrid, Spain.
Seidel, G.E. and Seidel, S.M. (1992). In: H. Niemann, H. Franzen and D. Smidt (eds), Mariensce, Germany, May 1992. Part 1, p68-80.
Smith, C. (1978) Animal Production 26: 101-110.
Smith, C. (1984) Research Development in Agriculture 1(2): 79-85
Smith, C. (1988a) World Animal Review 6: 2-10.
Smith, C. (1989) Animal Production 49: 49-62.
Smith, C. and Simpson, S.P. (1986) J. Anim. Breed. Genet. 103: 205-217.
Smith, C., Meuwissen, T.H.E. and Gibson, J.P. (1987) Animal Breeding Abstracts 55: 1-10.
Soller, M., Brody, T. and Genizi, A. (1979) Heredity 43: 179.
Southern, E.M. (1975) J. Molecular Biology 98: 503-517.
Stice, S., Strelchenko, N., Betthauser, J., Scott; B., Jurgella, G., Jackson, J.. David, V., Keefer, and Mathews, L. (1994) Theriogenology 41: 301 (abstr.)
Stokes, T.M., Cherny, R.A. and Brandon, M.R. (1994) Theriogenology 41: 303 (abstr.)
Szkotnicki, W.J., Tong, A.K.W., Sharaby, M.A., Krotch, K.M., Johnson, L.P., and Schaeff L.R. (1978) J. Dairy Sci. 61:497.
Taylor, St. C.S., Moore, A.J., Thiessen, R.B. and Bailey, C.M. (1985) Animal Production 4 401-440.
Teale, A.J., Tan, S.G. and Tan, J-h. (1994) In: Proc. 5th World. Congr. Gen. Appl. Livest. Pro Guelph, Canada.
Van Vleck, L.D. (1977) In: Proc. International Conf. Quant. Genetics. Pollak, E, Kempthorne, and Bailey, T.B. eds). Iowa State Univ. Press, Ames, Iowa, USA.
Van Vleck, L.D. (1981) In: B.G. Brackett, G.E. Seidel Jr and S.M. Seidel eds, New technologies animal breeding, pp 221-242. Academic Press, New York.
Visscher, P.M. and Haley, C.S. (1995) Amm. Breed. Abstracts 63:1-8.
Ward, K., Sleigh, M.J., Powell, B.C. and Rogers, G.E. (1982). In: Proc. 2nd World Congr. Ge Appl. Livest. Prod. Vol 6, p146-156.
Weber, J.L. and May, P.E. (1989) American J. Human Genetics 44: 388-396.
Whittington, D.G., Lyon, M. and Glenister, P.H. (1977) Genet. Research 30: 287-299.
Willham, R.L. (1982) J. Animal Sci. 54: 659-666.
Williams, J.G., Kubelik, A.R., Livak, K.J.C., Rafalski, J.A. and Tingey, M.S.V. (1990) Nucle Acids Research 18(22): 6531-6535.
Woolliams, JA. (1989) Animal Production 48: 31-35.
Woolliams, J.A. and Smith, C. (1988) Animal Production 46: 333-345.
Woolliams, J.A. and Wilmut,1. (1989) Animal Production 48: 3-30.
Yapi, C.V., Oya, A., Rege, J.E.O. (1994) In: Proc. 5th World Congr. Gen. Appl. Livest. Pro( Guelph, Canada. Vol 20, p421-424.
Young, C.W. and Miller, P. (1982) The advanced Animal Breeder 30(4): 6-8.
Zhang, W. and Smith, C. (1992) Theoret. Appl. Genet. 83: 813-820.
Zhang, W. and Smith, C. (1993) Theoret. Appl. Genet. 86:492-496. CONSERVATION OF DOMESTIC ANIMAL DIVERSITY
SUMMARY
INTRODUCTION
CONCLUSION
REFERENCES
S. Moyo Matopos Research Station, P Bag K 5137, Bulawayo, Zimbabwe.
SUMMARY
The world is facing a challenge to understand the causes of the decline in genetic diversity and to initiate programmes to address this issue. Conservation of domestic animal diversity (CDAD) encompasses characterisation, identification, monitoring and utilisation to ensure management for best short term use and longer term ready availability. Methods of conservation include ex situ (cyropreservation of embryos or semen or DNA segments) and in situ (live animal conservation). This paper also presents a summary of CDAD activities in Zimbabwe. A case for the establishment of a national network in order to achieve concerted efforts is presented. National governments are encouraged to take a leading role in conserving "adaptive" traits of the indigenous livestock which are not necessarily the same as "commercial" traits. This will ensure long term ready availability of this genetic material.
INTRODUCTION
There is an increased demand for animal protein in the world due to the expanding human population. This puts pressure on producers to improve efficiency of production from individual animals. The different livestock breeds and populations which have evolved and adapted to various environmental conditions represent an important genetic resource for improving efficiency of production now and in the future. Recent reports show that many animal genetic resources are endangered and, unless urgent concerted efforts are taken to conserve them, may be lost even before they are described and documented. Diversity refers to genetic variation within and between breeds. In this paper conservation of domestic animal diversity (CDAD) includes identification, monitoring, characterisation and utilisation for best short term use and to ensure management for longer term ready availability.
Background
In Southern Africa as in other countries of the world there is a change in the genetic variation of animal resources due to selection criteria to meet changing production and consumer demands. Environmental and climatic changes are other factors contributing to the change. Furthermore, developments in communication and advances in biotechnology (Artificial Insemination, AI) have accelerated the international movement of germplasm thereby shifting local attention to the more specialised exotic breeds. Consequently, indiscriminate crossbreeding with, and or replacement by exotic germplasm represent serious threats to indigenous populations. In most cases breeds that are not very profitable under current production and market conditions are left out and run the risk of extinction and once lost, genetic material is irreplaceable. Most recent results of the Food and Agriculture Organisation (FAO) global survey show that about 30 percent of the 4000 livestock breeds or so which are identified in the world are endangered (Hammond, 1994a). Thus, the challenge to try and understand the causes of the decline in genetic diversity and to initiate programmes to address this issue.
Several organisations (Food and Agriculture Organisation (FAO); United Nations Environment Programme (UNEP); Organisation of African Unity (OAU) and International Livestock Centre for Africa (ILCA) have addressed the issue of endangered species both at regional and global level. More recently (1992), in Rio De Janeno the Convention on Biological Diversity became international law with ratification from more than 70 countries. Some of the principle requirements of the Convention relating to animal production were summarised by (Hammond, 1994b). Briefly, under the Convention each country is obliged to develop and implement strategies which conserve and sustainably use the diversity both between and within its current species. This treaty also requires design of an international financial mechanism to achieve equal sharing of the benefits gained utilising biological diversity from national resources. Countries will also be required to regularly report on their national plans and progress with their implementation. In 1993 FAO initiated a Special Action Programme on Global Animal Genetic Resources which seeks to design and implement a management operation as well as activities to reduce the decline and to better utilize the available diversity.
Obiectives
The primary objective of this paper is to highlight issues involved in Conservation of Domestic Animal Diversity (CDAD). This paper draws heavily from lecture notes for an FAO/UNEP/ILCA training course on "Management of animal genetic resources, data banks and gene banks in Africa" held in Addis Ababa, Ethiopia (May 16 - 25, 1995). This presentation is timely in that papers presented in this symposium touch on CDAD issues. In addition a case for the establishment of a national network of scientists involved in animal genetic resources is presented.
Breeds and criteria for the endangerment of a breed
Breed is the term most commonly used to describe livestock populations or varieties and has been defined in various ways. For example the World Watch List for domestic animal diversity (WWL for DAD) prepared by the Food and Agriculture Organisation (FAO), 1993 defines a breed as:
"either a homogenous, subspecific group of domestic livestock with definable and identifiable external characteristics that enable it to be separated by visual appraisal from other similarly defined groups within the same species, or it is a homogenous group for which geographical separation from phenotypically sinular groups has led to general acceptance of its separate identity"
Conservation need is greater when size of population is getting smaller, especially when the rate of decrease is rapid. The first issue of WWL-DAD (FAO, 1993) lists the following as criteria for determining breeds at risk:
A breed is considered critical if the total number of breeding females is less than 100 or total number of breeding males is not more than five, or the overall population size is slightly above 100 but decreasing and percentage of females being bred pure is below 80 percent.
A breed is considered endangered if the total number of breeding females is between 100 and 1000 or the total number of breeding males is between six and 20 or the overall population size is slightly above 1000 and decreasing and the percentage of females being bred pure is below 80 percent.
A breed is considered extinct if it is no longer possible simply to recreate its original genetic base and most of its genetic variation has been lost. This is highly likely when there are either no breeding males (semen) and breeding females (oocytes) or no embryos remaining.
In many instances conservation simply means storing semen and embryos. The WWL-DAD (FAO, 1993) notes that preservation alone will not provide effective programmes for making the best use of animal genetic diversity and presents general action of conservation as follows:
1. Identification and listing of all breeds and their description and characterization, pointing out their differential qualities and possible contribution to future biodiversity.
2. Monitoring successive breed census and regularly reporting a list of those breeds currently endangered.
3. Facilitating the short-term use of as many breeds as possible. Well managed utilization of a breed is likely to be the most cost-effective way of maintaining its gene-pool for future use.
4. Storing as many representative samples of as many breeds as possible, usually as semen doses and embryos kept in liquid nitrogen to make revival of an extinct breed possible.
5. Implementing education and training programmes in the field of conservation genetics and effcient on the farm conservation techniques. Contributing to the development of effcient international and national policies of legal instruments both equally necessary for a sound domestic animals diversity conservation.
6. Establishing a coordinated global system to manage this set of operations and get maximum involvement of each of the many intervening agencies which are necessary to the success of this programme.
Methods of conservation
Methods for the conservation of animal genetic resources include; la.
Conservation of animal genetic material in the form of living ova, embryos or semen stored cryogenically in liquid nitrogen (ex situ).
1 b. Preservation of genetic infonnation as DNA, stored in frozen samples of blood or other animal tissues or as DNA segments (ex .situ)
2. Conservation of live populations in their adaptive environment or as close to it is practically possible. (In sins).
The major advantage of cryopreservation is that the relative cost of collecting, freezing and storing frozen material, as compared to maintaining large scale live populations, has been estimated to be very low and once the material has been collected the cost of maintaining a cyrogenic store is minimal (FAO, 1989). The disadvantages are related to the availability of the necessary technology and access to the frozen populations.
The principal advantage of live animal conservation is that it allows for breeds to be characterised and evaluated in their own related localities and at the same time allows for comparative trials, research, selection and improvement and adaptation to the changing environmental conditions. While the disadvantages of live animal conservation arc brought about by lack of complete control over many factors that contribute to the survival of the animal and therefore the genetic makeup of the population (disease threat, unpredictable financial and political change, selection pressures imposed by man and the environment). Furthermore, international movement of animal genetic resources is more complicated compared to movement of frozen material (FAO, 1992).
CDAD activities in Zimbabwe
Indigenous livestock breeds are well adapted to the harsh climatic conditions which prevail in almost two thirds of the country's land area (Vincent and Thomas, 1960) thus, represent an important genetic resource for improving production efficiency in these areas. This section briefly highlights CDAD activities in Zimbabwe with special reference to indigenous livestock, details of which were presented by Moyo (1994). Furthermore, some of the papers presented in this symposium cover topics on CDAD activities for different species.
Zimbabwe's three indigenous cattle breeds (Mashona, Nkone and Tull) have been described in several reports (Oliver. 1966; Goodwin, 1976: Brownlee, 1977). In the 1940s, the then government established herds for the three breeds with the objective of characterising, preserving and improving (through selection) these breeds. The recommended strategy for conservation is that of proper utilization of indigenous species within the different production systems in which they prevail. Programmes for rapid multiplication and genetic improvement within the indigenous breeds should be promoted and supported. Furthermore. it is necessary to determine whether these indigenous breeds and the other southern African Sanga cattle belong to one or different genetic groups.
The only known purebred indigenous "Sabi" sheep population is at Matopos Research Station (approximately 500 breeding ewes). There are small pockets of the Sabi type found in some areas of the country. This breed needs further detailed characterization and multiplication and evaluation of its performance under farm conditions. It is with this aim that a conservation programme has been initiated at Matopos Research Station in collaboration with the International Livestock Research Institute (ILRI).
Two indigenous types of goats are found in Zimbabwe. There is the type which resembles the Small East African goat and is predominately found in the central and eastern areas of the country, and the larger Matebele goat, which is found in the southern parts of the country. There is a need to adequately characterize indigenous goats in order to obtain information on the number of distinct breeds and strains to facilitate identification of animal genetic resources threatened by extinction and for which urgent conservation measures should be taken. It is necessary to understand the existing diversity to facilitate development of rational utilization and conservation strategies. There is also a need to evaluate indigenous goats under semi- extensive systems of production.
The indigenous pigs and poultry as with goats, need to be adequately characterized to understand the existing diversity to facilitate development of rational utilization and conservation strategies.
Organisations involved in CDAD
CDAD programmes can be administered by national government agencies, especially in establishing programmes to carry out further research into the indigenous breeds and their improvement and promotion. This is essential, especially, if they have an advantage over exotics (eg adaptability) which might not be shoring immediate economic benefits under current production and market conditions. Other organisations to administer CDAD programmes can include non government organisations, private organisations, cooperative groups of farmers and private individuals.
Networking
Success in CDAD can be achieved through coordinated efforts by all involved in animal genetic resources activities. This could be achieved through the establishment of a network, the objectives of which are to:
1. Create public awareness on the importance of CDAD through information exchange and organisation of workshops.
2. Identify available resources and promote activities in support of CDAD
3. Facilitate collaboration among scientists thereby breaking the isolation in which most of them operate.
4. Facilitate training in the relevant aspects of CDAD
5. Link with other regional CDAD networks, ILRI animal genetic resources programme and FAO global programme.
Effective networking will need funding to facilitate communication to meet the objectives listed above.
CONCLUSION
The need to conserve genetic variation in domestic animals has been recognised for many years. But, the rapid decline of this diversity under changing environmental and economic climate make conservation needs more urgent than ever before. All people working with animal genetic resources need to be mobilized to promote activities in support of CDAD. The strength of the national groups in collaboration with groups in other regions and other organisations (eg FAO) will lead to the success of the concerted efforts in CDAD. For Zimbabwe there is greater need to promote conservation with utilization and promote sustainable use of the diverse animal genetic resources under the differcnt production systems and consumer demands. National governments should take the leading role in conserving "adaptive traits" of the indigenous breeds which are not necessarily the "commercial traits" to ensure long tern ready availability.
REFERENCES
Brownlee, JW1 (1977). Rhod. Agnc. J. 74: I-9. Convention of biological diversity. (1992). Rio De Janeiro, Brazil. FAO, 1989. FAO Animal Production and health paper No. 76. Rome, Italy. FAO, 1992. FAO Animal Production and health paper No. 99. Rome, Italy. Goodwin, C.P.D. (1966). Modern Farming 4: 9-13. Hammond, K. (1994) World Animal Review 78 (1) Comment. Hammond, K. (1994) FAO, Rome, Italy. Moyo, S. (1994). FAO/UNEP/ILCA training course on "Management of animal genetic resources, data banks and gene banks in Africa" Addis Ababa, Ethiopia. 16-25 May, 1994. Oliver, J. (1966). Exptl. Agric. 2: 119-128. Vincent, V. and Thomas, R.G. 1960. Government Printer, Federation of Rhodesia and Nyasaland. FAO, (1993) World Watch List for Domestic Aiumal Diversity. 1 st Edition. Rome, Italy.
THE CONSERVATION AND SELECTION OF INDIGENOUS BEEF BREEDS IN ZIMBABWE
SUMMARY
INTRODUCTION
CHALLENGES
INDIGENOUS BREED CONSERVATION INITIATIVES
CONCLUSION
REFERENCES
C. T. Khombe Ministry of National Affairs, Employment Creation and Cooperatives, P Bag 7762, Causeway, Zimbabwe.
SUMMARY
The future of beef cattle production in Zimbabwe will rely on a wider use of breeds and genotypes that can survive and produce on free range grazing. A majority of these breeds will comprise of indigenous breeds or their crosses with the larger exotic breeds. It therefore comes as a surprise that, other than the few registered herds in commercial farms and research stations, there is very little information on the structure and performance of the large pool of indigenous cattle in small holder farms.
Challenges encountered in the formulation of conservation and selection strategies of indigenous beef cattle include; the lack of information on population dynamics of the different breeds; inadequate information on their development and characteristics; absence .of a performance recording scheme in the majority of small holder herds; defining the environment and system of management to be used in the conserved herds; and the lack of resources to manage conservation programmes.
Mashona cattle are selected at Nyombi farm by Indibreed (Pvt) Ltd and at Makoholi Experiment Station. The Tuli and Nkone are being bred at Matopos Research Station. It is advocated that the indigenous gene pool in both small holder and commercial herds should be maintained in their respective environments and management systems. The conservation of indigenous genes in crossbred cattle in commercial environments should be encouraged under the strategy of 'conservation with utilisation'.
INTRODUCTION
The frequent dry spells, that are now common in Zimbabwe, have resulted in the deterioration of rangeland resulting in the loss of over 1 million cattle in the small holder farming areas (Madzima, 1993). The prevailing economic environment has made the pen finishing of livestock to be both uneconomic and unethical. In the near future, successful beef production systems will be those that rely on range grazing. Such production systems will promote the use of indigenous breeds, which are renowned for their production on free range grazing due primarily to their superior reproductive performance, relatively low maintenance requirements and high viability of progeny (Tawonezvi, et al., 1988). These characteristics also render the indigenous breeds ideal damlines for use in crossbreeding with larger exotic sire breeds in the commercial sector (Moyo, 1994).
Efforts to utilise these breeds in commercial agriculture have been frustrated by the unavailability of large numbers of selected breeding animals. More than 90 percent of the indigenous cattle population is in the small holder fanning sector. Before the destructive 1992 drought there were an estimated 3.5 million (Holness, 1992) cattle in this farming sector, but the population is likely to have declined to 1.8 million due to the large losses suffered during and after the drought. These losses might increase if the current season (1994/5) does not improve. After each drought, small holder farmers rebuild their herds using exotic type animals acquired from commercial farms, because of the unavailability of adapted indigenous breeds. These restocking periods have contaminated the indigenous gene pool in small holder farms highlighting the need for the establishment of conservation programmes.
Indigenous cattle in commercial farms are threatened by the insatiable quest to improve on their conformation and size by crossbreeding with the larger exotic breeds (Khombe, et al:, 1994). A majority of these improved indigenous animals form the breeding herds of many commercial farms. It is also necessary to initiate within breed selection programmes that will allow the indigenous breeds to compete favourably with other breeds, without the infusion of exotic genes, under improved management production systems.
CHALLENGES
Lack of information on population dynamics of the different breeds
The known populations of indigenous breeds are those in registered commercial herds and Government institutions. There is a lack of population statistics (ie numbers, structures and distribution of genotypes) of the large pool of indigenous cattle in small holder farms. The Mashona breed is the only indigenous breed Nvith known population estimates (Table 1). Figures from the Mashona Cattle Society indicate that there are 22 000 adult animals. (Indibreed, 1994) and Moyo (1994) estimate a total population of half a million Mashona cattle in the small holder herds. Nothing is known about the numbers of Tull and Nkone cattle in this fanning sector.
Table 1. Estimated numbers1 of indigenous beef cattle in the different herds
Breed Herd Stations Commercial Small holder Mashona bulls 50 160 12 000 cows 450 3 200 10 000 Tuli bulls 50 250 unknown cows 220 1 431 unknown Nkone bulls 35 10 unknown cows 140 73 unknown
'Estimates obtained from Indibreed (1994) and Moyo (1994) The absence of this information makes it difficult to plan appropriate conservation strategies and also identify populations that are declining. It comes as no surprise that while the Nkone is in list of breeds considered by FAO as endangered, because of the small numbers of registered animals, their populations in the small holder herds might be greater than that of the Tull (Mario Beffa, personal communication). The degree of crossbreeding that has occurred during the restocking periods is also unknown and so is the related loss of indigenous genes.
Inadequate information on development and characterisation
There is very little information, except on growth and reproduction, on the attributes of indigenous breeds. There is no information on their tick and internal parasite resistance, heat tolerance and adaptation to grazing low quality forages. The loss of adaptive traits in these animals by selection under improved management production systems have not been quantified. So are the effects of the continued selection for high body weights on fitness traits like reproduction, viability and metabolic efficiency.
There is a paucity of information on the genetic differences between indigenous cattle in commercial farms and the large pool in small holder farms. Such information is important for the creation of nucleus herds to improve animal performance in both commercial and small holder environments. Khombe et al. (1994) showed that Mashona cattle in commercial farms had breeding values that were higher by 1.6 and 10 kg for birth and weaning weight, respectively, than bulls from small holder herds. The latter authors speculated that differences were due to the residual heterosis in commercial herds. The relationships between these strains of cattle need to be established. Similarly, the genetic relationships between the three indigenous breeds and other indigenous breeds like the Nguni, Tswana and Barotse need to be established. Oral historical evidence suggests that these breeds have strong ties. In 1896 the Rinderpest and East Coast Fever epizootics wiped out the cattle herd in the whole country and only 50 000 cattle survived. Cattle of Angom type were imported from Zambia and bulls were brought in from South Africa and abroad to build up and 'upgrade' the national herd (Holness, 1992). This indiscriminate cross breeding that occurred is supposed to have created the indigenous breeds that are currently known in Zimbabwe. It is therefore likely that some populations though isolated geographically, are not genetically different. Some of these populations may have some special genes that are similar. Information on the degree of relationships and genetic distances between breeds or genotypes is necessary (Setshwaclo, 1990).
Absence of a performance recording scheme
The lack of a performance recording scheme in the bulk of the indigenous breed population in small holder herds limits the accuracy with which high performing animals can be identified. It is then risky to utilise animals identified in these herds in registered herds since their genotypes and pedigrees are unknown. Commercial herds that have closed themselves from the smallholder gene pool and will face the handicaps associated with selection in small populations and the high inbreeding coefficients. No meaningful conservation of the indigenous breeds can be accomplished without the inclusion of the indigenous gene pool. The challenge facing animal breeders in Zimbabwe is to identify strategies in which animals of high genetic merit could be identified from the small holder gene pool without compromising the genetic gains already attained in the commercial herds (Khombe and Hayes, 1995).
Defining the environment and system of management Indigenous breeds are adapted to produce under conditions of low levels of nutrition, high loads of both internal and external parasites and low water availability. It is not known how this adaptation changes when these animals are reared under high levels of management, le high levels of feeding, regular dipping and dosing and frequent watering. Currently there exists the following schools of thought.
1. That the conservation and selection of indigenous cattle should be carried out in the same environment in which the cattle are supposed to perform. Subscribers to this philosophy suggest that nucleus/conservation herds should be located in smallholder farms. Since logistic limitations restrict the establishment of conservation herds in small holder farms, it has been argued that these conditions can be mimicked in research stations. The herds at Makoholi and Matopos are now run on minimal supplementary feeding, with strategic dosing and dipping. However, these stations are not able to use high, stocking rates of 2 hectares per livestock unit, that have been reported to be the norm in the small holder farming sector.
2. That you cannot set up selection objectives for poor management. It has been argued that conservation strategies should put emphasis on the future utilisation of indigenous cattle in commercial farms, since natural selection is the only active force in small holder farms. Consequently, these cattle should be selected under high levels of management. However, there is a paucity of information on the magnitude of the genotype by environment interaction that is likely to occur when animals selected under high management level are subsequently reared under low management production systems.
Suggestions have been made that within breed, one population should be selected under high management in breeding stations and another selected in smallholder farms.
Lack of resources to manage conservation programs
The policy makers do not see the immediate benefits accrued from the conservation of indigenous breeds. Otherwise they would facilitate conservation efforts by;
1. Funding the creation and operation of conservation/nucleus herds,
2. Enforcing laws that deliberately promote conservation like (i) availing indigenous bulls to small holder farms and only allowing the restocking of small holder herds with indigenous breeds, (ii) giving an indigenous breed subsidy to the Cold Storage Company of Zimbabwe to allow a bonus payment to producers of indigenous breeds.
Technologies such as artificial insemination, embryo transfer and cryopreservation are not widely used in the beef industry, although the former has gained widespread use in the dairy industry. While the benefits accrued from the use of these technologies are well known, the financial resources that are required are not available.
Another option of conservation is to 'attach' indigenous genes to popular breeds like the Brahman through crossbreeding. Such a strategy would ensure that indigenous genes survive from genetic deaths caused by the short term market demands for carcasses from particular genotypes. However, very little research has been carried to establish the minimum proportion of genes that are necessary in a crossbred to allow for the regeneration of the minority genotype. It is still not clear whether conservation should be made on the genes or on the genotypes. In Zimbabwe there is scope to conserve the genotypes in breeding stations and the genes in production herds, both commercial and small holder herds.
INDIGENOUS BREED CONSERVATION INITIATIVES
Mashona
1. The Mashona Group Breeding Scheme. The Mashona Group Breeding Scheme is operated by Indibreed (Pvt) Ltd at Nyombi farm. It was established in 1990 as a joint venture between the Mashona Breeders Association and Apex Corporation. The scheme was established to preserve and improve the genetic qualities of the Mashona (Indibreed, 1994). This idea was developed when the breed society recognised the limitations of within herd selection from their small herds and the dangers associated with inbreeding. The breeders combined their genetic resources (160 bulls and 3200 cows) and contributed 300 cows to the scheme. When fully operational the nucleus herd will comprise 400 cows and 25 bulls. The animals in the nucleus herd are selected on weight using a programme that compares each individual animals performance with its group mean to produce a 205-day ratio. The selection index for bulls emphasises on performance test results (feedlot and range) and that selection of heifers includes the dam's productivity (Smith, 1994).
2. The Mashona herd at Makoholi Experiment Station.
Makoholi holds the largest Government herd (300 cows and 20 bulls) of Mashona cattle, that is also registered with the Mashona Cattle Society. The major objective of this herd is to broaden its genetic base through the acquisition and progeny testing bulls from both commercial and small holder farms (Khombe, 1994). The project works on the premise that 35 percent of the bulls in small holder farms have average genetic merit for growth (Khombe et al., 1994) and identifies these bulls through a progeny test run using the Makoholi breeding cow herd. High merit bulls and their offspring are retained in the herd. Brought-in bulls will be given BLUP breeding values based on the pre-weaning growth of their progeny and their own performance. When the screening program is fully operational, high merit animals will be used to form a nucleus herd.
Tuli
1. The Tull herd at Matopos Research Station.
It has been proposed that the closed herd, of 140 cows, at Matopos should be increased and opened through a screening process. The females will be sourced from registered breeders and the small holder herds. Stock bulls will be generated within herd with the proviso that bulls of exceptional merit will be purchased if found necessary. The Tuli Cattle Society will assist during the screening exercise and in ensuring that an acceptable percentage of the best cows available enter the nucleus. This herd will be managed with minimal inputs (prophylactics and supplementary feeding) so that selection of animals will ensure that the hardiness and adaptability of these animals is not lost.
2. The Tuli herd at Grasslands Research Station.
Grasslands has a 200 breeding cow herd that is maintained by purchasing bulls from commercial farms and from Matopos Research Station. This herd is currently kept to service the station's experiments but has a potential of being developed into a conservation herd.
Nkone
Matopos Research Station holds a 140 breeding cow herd that has been closed herd since its inception. A proposal has been made to increase the number of cattle either by expansion of the Matopos herd or by establishing a conservation herd on another site (Beffa and Ncube, 1994). The optimal number of cows in the nucleus herds will be set at 300 and the management and selection will be similar to the Tuli herd.
CONCLUSION Zimbabwe still has an abundant indigenous cattle gene pool and none of the three major breeds can be regarded as being endangered in the true sense of the word. However, the large reserve of indigenous cattle in the small holder farms face the danger of being wiped out by the now frequent droughts or contaminated by the indiscriminate crossbreeding that occurs subsequently during herd building. It is still opportune for Government to fund attempts to conserve these cattle. The capture of as much genetic variability as is possible for conservation and selection will need the application of appropriate sampling techniques, which will depend on the structure of the population (Setshwaelo, 1990). However, the lack of financial resources will limit the degree to which the sampling can be done, thus making it difficult to capture all the genetic variability present. Conservation strategies should also use embryo collections of desirable cow/bull combinations and semen from productive (desirable) bulls. Since conservation is not easily justifiable economically, breeding stations should put emphasis on the strategy of conservation with utilisation. An effort should also be made to establish and characterise other stable genotypes, other than the Tuli, Nkone and Mashona, that have now adapted to stressful environments. It is my belief that some of the crossbred genotypes have adapted to the harsh environments typical of small holder farms and could possess desirable qualities like high milk yields and faster growth rates.
REFERENCES
Beffa, M. L. and Ncube, B. (1994) Matopos Research Station, Zimbabwe.
Holness, D.H. (1992) Mashona Cattle of Zimbabwe. Mashona Cattle Society.
Indibreed (Pvt) Ltd (1994) Description of the Mashona Group Breeding Scheme. Khombe, C.T. (1994) Makoholi Experiment Station, Zimbabwe.
Khombe, C.T. and Hayes, J.F. (1995) Zimbabwe Journal of Agricultural Research (in press)
Khombc, C.T., Hayes, J.F., Tawonezvi, H.P.R. (1994) Journal of Agricultural Science, Cambridge, 122:459-463.
Madzima, W.N. (1993) Department of Veterinary Services, Zimbabwe.
Moyo, S. (1994) FAO/UNEP/ILCA Course on the Management of Databanks and Genebanks in Africa. Ethiopia.
Setshwaelo, L. L. (1990) Proc. 4th World Congr. Genet. Appl. Livestock Prod. 449-454. Smith, R. D. (1994) Mashona Indaba, Mashona Cattle Society.
Tawonezvi. H.P.R., Ward, H.K., Trail, J.C.M. and Light, D. (1988) Anim. Prod. 47:351-359. POULTRY BREEDING IN ZIMBABWE
SUMMARY
INTRODUCTION
BREEDING IN INDIGENOUS FLOCKS
BREEDING IN COMMERCIAL POULTRY BREEDS AND STRAINS
THE FUTURE
REFERENCES
A. T. Faranisi Irvine's Day Old Chicks, P 0 Box 815, Harare, Zlmbabwe.
SUMMARY
The poultry industry in Zimbabwe is based on both indigenous and imported poultry strains. Commercial production is dominated by imported strains and indigenous strains have remained insignificant due to lack of genetic improvement in all commercially important traits. This paper presents a brief account of commercial poultry breeding and traits which are considered important in Zimbabwe.
INTRODUCTION
Poultry breeding in Zimbabwe is based on commercial strains produced by the four main breeders and thousands of indigenous flocks in the communal farming sector and backyard flocks in urban areas. Trends in broiler, egg and day old chick production are given in Table 1.
Table 1. Trends in production of broilers, table eggs and day old chicks 1985 - 1994
Year Sales by volume from commercial producers Table broilers Day old Sexed Eggs
(Millions) Broilers pullets (millions of dozens)
(millions) (millions)
1985 7.2 6.3 0.7 11.7 1986 7.1 8.2 1.6 11.5 1987 8.4 9.5 1.5 12.0 1988 10.5 11.9 1.5 14.1 1989 11.5 15.2 1.9 17.8 1990 13.0 16.1 1.6 18.7 1991 13.7 23.5 2.3 19.4 1992 13.2 15.1 1.3 19.6 1993 11.4 15.6 2.1 17.5
Source: C.S.O. (1994)
Although the production of commercial poultry strains is ever increasing as shown in Table 1, the indigenous flocks still contribute significantly to meat and egg requirements in Zimbabwe, but statistics are not available.
BREEDING IN INDIGENOUS FLOCKS
Breeding and selection in indigenous poultry has been largely left to nature and to date no differentiation into broiler or layer strains has occurred. Consequently, production in both meat and egg production has remained very low when compared with commercial strains. Field observations for egg production, for example, show that in a laying cycle indigenous hens will produce 10 - 50 eggs with an average weight not exceeding 52 grammes. Commercial strains on the other hand can produce up to 300 eggs with an average weight of 63 - 65g. While some of the difference in productivity may be due to non-genetic factors such as feeding regime and management system, most of it is a result of genetic differences.
BREEDING IN COMMERCIAL POULTRY BREEDS AND STRAINS
As mentioned earlier, commercial poultry breeding is conducted by a few companies in Zimbabwe. Each of the companies produces its own broiler or layer stock from either greatgrandparents (GGP) or grandparents (GP) and parents, imported from primary breeders in Europe. Greatgrandparents are essentially purelines. The progeny of the great grandparents that are selected to produce parents are referred to as grandparents and their offspring are called parents. Eggs from parents then hatch to give broiler or layer chicks for commercial production.
The breeding companies generally use three line or four line crosses in their breeding programmes. A three line cross involves crossing two strains with different qualities and then crossing their progeny with the third line. In a four line cross, two of the four lines are crossed and the remaining two are also crossed. The male offspring of the one cross are then mated to the female offspring of the other to produce parent stock. Parent stock is retained by the breeders and only commercial day old chicks are sold to producers, but the breeders remain the largest producers of commercial poultry. The poultry industry can therefore be described as shown in Figure 1.
Figure 1. Structure Of Commercial Poultry Breeding in Zimbabwe
GP' S
GP'S
(From Europe)
Local breeders
Commercial farms
Small scale producers
EGG LAYING STRAINS
Nearly all the genetic improvement in commercial layer chickens is undertaken by primary breeders. Local breeders then choose from the primary breeders genetic stock that will best meet their local market demands. Traits that receive most attention are ease of sexing chicks, feed conversion, egg productions, egg size and quality of eggs.
CHICK SEXING
Layer chick producers depend on colour and feather sexing to separate cocks and pullets at day old. Plumage colour in local brown egg layers is controlled by sex linked genes which produce the gold/brown or silver/white plumage. At the parent level males are brown coloured and the pullets are white coloured. Mating these parents produces pullets with a golden down and males with a silver down. Sexing day old chicks, which all have the same colour, is achieved by use of sex linked fast feathering and slow feathering genes. These genes on some strains, for example, will produce fast feathering pullets in the female line pullets and slow feathering males.
This ability to sex layer pullets at day old significantly reduces the cost of raising layer pullets.
EGG PRODUCTION
Egg production is made up of several associated characteristics such as age at onset of lay, higher peak rate and persistence of lay. Selection or higher egg production is normally directed at these traits (Flock, 1994). Of the above traits, emphasis or selection pressure on age at onset of lay will be reduced as most egg producers feel that genetically early maturing pullets are difficult to bring into lay and manage.
BODYWEIGHT AND FEED CONVERSION
Local breeders continually import strains with lower bodyweights as this reduces pullet maintenance costs. With the ever increasing costs of stockfeeds, strains with the best feed conversion ratios (FCR) are preferred. Over the years, improvements in FCR have occurred as a correlated response to selection for greater egg mass production and lower bodyweight by primary breeders.
EGG QUALITY
On egg quality, breeders are concerned mainly with optimum egg weights for local markets, good shell strength and internal quality factors such as albumen height and absence of inclusion such as blood spots. The Zimbabwean market places the highest demand on eggs that have an average weight of 60g or lower. Breeders therefore prefer strains that produce the largest proportion of eggs in this weight range.
Other important genetic traits considered are chick livability and hatchability which are both critical for reducing chick production costs. Table 3 gives a summary of progress in egg production traits in brown egg strains.
Table 3. Changes in key traits for brown egg strains 1981 to 1993
Strain Years Egg Egg Weight Egg FCR
Ending Number (kg/henhoused) Weight (kg/kg)
(g) Hisex 81-83 277 17.6 63.7 2.56 Brown 91-93 293 18.9 62.7 2.28 Lohmann 81-83 272 17.5 64.5 2.60 Brown 91-93 297 19.8 66.7 2.22
Source: Flock (1994).
BROILER BREEDING
Unlike in layer strains, some selections are done locally to improve productivity in broilers. The current targets for selection in broiler breeding and how they changed are given in Table 4.
Table 4. Targets for selection in broiler breeders
1960 1970 Targets for selection 1990
1980 Weight for age Weight for age Weight for age Weight for age Conformation Conformation Conformation Conformation Reproductive Reproductive Reproductive Reproductive performance performance performance performance Leg defects Leg defects Leg defects Leg defects Yield of parts Yield of parts Yield of parts Livability Livability Livability Feed conversion Feed conversion Feed conversion ratio ratio ratio
Source: Barton (1994)
All the traits listed in Table 4 are determined between 5 and 6 weeks of age with the exception of reproductive performance which is measured over an entire laying cycle. Selection for reproductive traits is mainly applied to female line G.P's. They are selected to produce a fairly large number of eggs that hatch well and genetically they have good growth promoting characteristics. The other traits such as meat yield, FCR and weight for age are selected for in the male line G.P's. The male line G.P's typically have exceptionally heavy fleshing, good breast conformation, rapid growth rate and achieve good FCR. These traits have of course been improved at the expense of good egg production and fertility. The targets for selection in broiler breeders in Zimbabwe are primarily aimed at cost reduction and improving quality of broiler meat.
COST REDUCTION
Genetically improving liveweight gain, FCR, reducing mortality and reproductive performance will lead to a reduction in the cost of broiler meat production. It must also be noted that these improvements have given poultry meat a cost advantage over other meats.
QUALITY TRAITS
Selection here is for improved carcass conformation, absence of blisters, hockburn and lameness. These traits are really linked with growth rate and are considered at 5 - 6 week selections. Breeders also strive to keep their flocks free of vertically transmitted poultry and human pathogens such as mycoplasmas and salmonella.
In some faster growing strains it may be necessary to reduce growth rates genetically in order to reduce physiological disorders such as ascites or structural defects such as splayed legs. Some breeders achieve this by renewing stocks of female G.P's only and retaining the male line G.P's which would have adapted to local conditions.
THE VALUE OF 5 - 6 WEEK SELECTIONS
Selection for broiler traits in Zimbabwe are currently done at 5 - 6 weeks because the weight of G.P's at this age is highly correlated to the weight of broilers at the same age. As most of the selection targets are related to the male line, higher selection pressure is applied to this line. Table 5 gives a summary of genetic progress made in Cobb broilers.
Table 5. Genetic Progress in Broilers 1977-1993
Traits 1977 1985 1993 Liveweight (grammes) 28 days 720 1011 1204 42 days 1360 1793 2092 56 days 2068 2576 2978 FCR 1.6 1.6 1.51
Source: Cobb Breeding Company (unpublished).
THE FUTURE
The future of poultry breeding in Zimbabwe looks very promising, but there are a lot of non- genetic factors that need to be improved before we can fully utilise the genetic resources we have. In breeding operations, for example, introduction of separate sex feeding will enable better control of male body weight which is essential for maintaining good fertility levels.
At all levels of poultry breeding improvements can be made in feeds and feeding. Outbreaks of Newcastle Disease in 1994 decimated a lot of indigenous flocks which represent a vast genetic resource yet simple vaccination controls can prevent these outbreaks. Housing and management standards have also hampered full expression of genetic potential for all important traits. In open sided broiler sheds, for example, daytime temperatures can be as high as 25oc and drop to 5oc at night resulting in high mortality and poor growth rates.
Demands on primary breeders by markets in developed countries may result in some undesirable trait changes for our poultry producers. If the processed egg market, for example, starts demanding thinner shelled eggs, the resulting egg will probably not survive our local transport system. In broilers, developed countries may push for leaner carcasses which may be less acceptable to our local markets. It is hoped that primary breeders will continue to produce products that will fit into various world markets.
Information on genetic make-up and productivity in indigenous breeds is very scarce which makes it difficult to assess their genetic potential for meat and egg production. However, there is little doubt that even the absence of genetic improvement, better feeding and housing will improve productivity in the indigenous flocks.
REFERENCES
Barton, N.F. (1994) Proc. 9th European Poultry Conference, Glasgow, U.K. C. S. 0. (1994) Central Statistics Office, Zimbabwe Government. Flock D. (1994) Proc. 9th European Poultry Conference, Glasgow, U.K.
OSTRICH PRODUCTION IN ZIMBABWE: SUMMARY OF SURVEY RESULTS
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
K. Dzama1, F. Mungate and J. H. Topps 1University of Zimbabwe, Department of Paraclinical Veterinary Studies, Box MP 167, Mt Pleasant, Harare, Zimbabwe. University of Zimbabwe, Department of Animal Science, Box MP 167, Mt Pleasant, Harare, Zimbabwe.
SUMMARY
A survey was carried out to asses the production systems and production levels of ostrich farmers in Zimbabwe. A total of 35 ostrich farmers collectively owning 5411 ostriches participated in this study. A large variation in sizes of enterprises and in productivity were found. The mean flock size per farm was 182.7, ranging from 8 to 1830 ostriches. About 37% of the farmers obtain replacement eggs from other farmers whilst 31% obtain them from within their own stock. There is no correlation between source of eggs and level of inbreeding. There are no standardised mating practices. The proportion of farmers who mate the ostriches in pairs was 43% whilst those who mate in trios and smaller ratios was 57%and the mean breeding season length was 7.06 months, ranging from 4 to 11 months. The average egg fertility was found to be 66%. Even though there was no correlation between egg fertility and inbreeding, eggs from farms practicing inbreeding were less fertile (59%) when compared to those from farms without inbreeding (73.3%). Of the total, 44.4% of the farmers claim their inbreeding level is low, 16.7% say it is high and 36.1% claim it is zero. Farmers reporting high levels of inbreeding had also observed an increase in congenital deformities while those reporting low or no inbreeding reported no increase in congenital deformities. Farmers with complete pedigree records and breeding hen flocks of 15 and above represent 14% of the respondents. These farmers can provide data for meaningful statistical genetic analyses.
INTRODUCTION
The production of ostriches in Zimbabwe on a commercial basis began in the mid 80s and since then it has become a popular alternative or an addition to traditional farming of crops and livestock on several farms. The products of value are the hide, meat and feathers, usually in that order of importance but recently the export of live birds has been the major marketing activity. However, this has had an effect of flooding the European market, resulting in uncompetitive prices being offered (The Herald, 3/11/94). The local tanning industry for the ostrich hides has become popular overseas, with the American ostrich breeders expressing interest in having their hides tanned in Zimbabwe (The Sunday Mail, 26/02/95). Production systems are either intensive or extensive or a combination of both. Returns per unit area are generally high, for example, a 30 hen production unit compares very favourably with many well managed mixed tobacco farms (Sharp, 1992).
The industry is still young, with most farmers having been in production for three years or less. A lot is still unknown about ostriches such that most of the managerial techniques being used are derived mostly from poultry. Specialized breeders who raise high quality chicks to sell to producers, similar to the situation in the cattle industry are not yet available and farmers tend to obtain replacement eggs mostly from other farmers. Usually the genetic and hence production potential of these eggs are unknown and hence productivity is not guaranteed. Problems of congenital deformities, particularly of the legs and neck have been reported. Some ascribe these to management shortfalls during chick rearing, particularly nutrition, whilst others say they could be due to inbreeding. The objectives of this study were to characterize the productivity of the ostrich population in Zimbabwe with special emphasis on the breeding strategies of ostrich farmers and other genetic aspects.
MATERIALS AND METHODS
Questionnaires were sent out to ostrich farmers registered with The Ostrich Producers Association of Zimbabwe (TOPAZ) (see Appendix 1). The returned questionnaires were analyzed using SAS (1987). Statistics for central tendency and dispersion were calculated for the flock size, breeding hen, breeding cock and growing stock populations. The performance parameters assessed were number of eggs laid per bird, number of fertile eggs laid per bird, number of eggs laid by the best and worst birds per season. Simple correlation analyses were carried out to measure the strength of relationships among certain management practices eg the breeding season length and number of eggs produced.
RESULTS AND DISCUSSION
Simple statistics
A total of 174 farmers are registered with TOPAZ. All of them had questionnaires mailed to them. A total of 35 farmers collectively owning 5411 ostriches responded.Figure 1 shows the distribution of the flock sizes among the farms surveyed. The results reveal a fairly uniform distribution of enterprise size on the scale used.
Figure 1. Flock sizes of the different farms surveyed
51-100
>50 Flock size 101-150 151+
While this sample represents a small percentage of the ostrich farming community, it is also fairly random. Table 1 shows the statistics of central tendency and variability for flock sizes and compositions. The farmer with the largest flock size had 1830 birds while the one with the smallest flock had 8 birds with the average being 182.7 ostriches. The median flock size is 115 and the mode is 26 birds. The mean breeding hen population is 25.5 with a median of 14 birds, a standard deviation of 34.3. The breeding hen size ranges from 3 to 189. Among the farms surveyed, breeding cocks averaged 20.2 birds with a median of 10, a standard deviation of 32.27 and a range of 2 to 188 ostriches. These results show that there is a large variation in sizes of enterprises. This is mainly due to the fact that most of these farmers have recently joined this industry and are just starting to build up their flocks. Most of the farmers surveyed have been raising ostriches for three years or less.
Table 1 Statistics of central tendencv and variabilitv of the flock sizes and composition
Item n Mean SD Median Mode Range Flock 35 182.74 323.79 105 26 8-1830 size Breeding 35 25.51 34.32 14 4 3- 189 hens Breeding 35 20.17 32.27 10 6 2- 188 cocks Growing 32 149.31 271.58 78 10 10-1453
stock
Among the farmers responding, 8% mate their birds in groups, 16% in trios (one cock to two hens), 43% mate their birds in pairs and the rest mate them in the ratio 2:3. It is of importance to know which mating ratio is best since it is not known which ratio is the best for the different production systems in Zimbabwe. Thirty seven percent of the farmers said they obtained replacement hens from other farmers within Zimbabwe while 31% said they obtained their replacement hens from within their flocks. There is no correlation between source of eggs and level of inbreeding reported. Only one farmer said he imported eggs in addition to buying from other farmers. However, Byford-Jones (1995) reported that Zimbabwe represents a substantive market for the export of fertile ostriches and their eggs from some South African farmers. Some farmers collect eggs from the wild even though none of the farmers responding mentioned this practice.
There is also a great variation in the length and timing of the breeding season. The minimum is 4 months, the maximum is 11 months and the mean is 7.1 months. There is no correlation between the length of the breeding season and the number of eggs laid. Some farmers mate from as early as May to as late as January whilst others start as late as July and end as early as November. This disparity indicates the divergence of ideas as far as mating season is concerned in the ostrich industry. There is need for research to determine the optimum length of the breeding season. As for the stocking rate, no two farmers had the same stocking rate and the explanation would be that areas vary in resource availability and this influences farmer preference on stocking rate.
Performance indicators
The number of eggs laid per bird per season ranges from 11 to 60 with an average of 29.8 eggs. Of these the mean fertile eggs per bird per season was 19. According to Sharp (1992) the number of eggs per bird averages around 45. In South Africa the average eggs per hen is 80 eggs while in the USA it is 55 eggs (Van Zyl, 1991). Fertility in the wild is reported to be around 70%. The average fertility in this study was found to be 66%. The national average fertility is 63% (Foot, 1994). When only the top birds are considered, the average fertility rises to 69% while the worst performing birds average 47% fertility. Farmers reporting some inbreeding in their operation collectively had an average of 59% fertility while those reporting no inbreeding had 73.3% fertility. However there was no correlation between level of inbreeding and fertility. Fertility is much higher in South Africa and the USA averaging 90% and 85% respectively (Van Zyl, 1991).
Although this study did not include all the farmers in this country, the indication is that there is room for further improvement in reproductive capacity. Productivity in the South African ostrich industry is much higher compared to the local industry. The industry in the USA, even though as young as the Zimbabwean industry, seems to be more productive. It should be noted that most of these ostriches are in their first and second years of production, which, of course, are associated with generally low production. Domesticated ostriches are susceptible to stress and this leads to low reproductive performance especially since most of the management is largely by trial and error.
Genetic issues
The survey revealed that only 58.6% of the farmers keep full pedigree records and 10.3% have some but incomplete records. The rest of the farmers do not keep any pedigree records at all. Most of these farmers do not obtain any performance histories of parent which lay the eggs that they acquire from other farmers as replacements. As mentioned earlier, the majority of farmers breed their birds in trios and smaller ratios and some record laid eggs to each pen and not to each parent. This system makes it difficult to obtain pedigree records. Most farmers reduce the problem by having a stockman watch the birds during the likely laying periods and make the appropriate records. The probability of misrecording the eggs parentage is fairly high especially where the breeding ratio is as small as 1:5.
Of the total, 17.2% of the farmers have observed an increase in congenital deformities in their flock, and virtually all of them (16.7%) thought they had high levels of inbreeding within their flock and 41.4% say their flock inbreeding is low whilst 36.1 % report no inbreeding. Of those farmers reporting low levels of inbreeding, none had observed an increase in congenital deformities. It does not necessarily mean that the farmers do not have congenital deformities in their flocks. As expected farmers reporting zero inbreeding had not observed an increase in congenital deformities. These are mostly farmers with smaller breeding flocks and can manage their matings more carefully. These farmers largely assumed that their base population birds were unrelated. Given the rapid expansion of the industry and movement of birds, this assumption may not hold.
Congenital deformities like twisted necks, deformed legs, twisted beaks and sometimes deformed eggs are regarded as a problem in the ostrich industry. Ideally they must be kept at 3% or less (Foggin, 1992). There seems to be evidence suggesting that inbreeding may be one of the causes. It is equally vital to find out the role of non genetic factors such as management, particularly nutrition during chick rearing. Pedigree records have to be kept by ostrich producers so that exploitation of the genetic potential of these birds can be maximized in the future. In this study farmers with complete pedigree records and a large operation of say 15 breeding hens or more represent 14% of the respondents. The records will make it possible for breeders to determine which cocks to mate to which hens for maximum performance if they are accompanied by production records. In the present study, none of the farmers responding kept any growth records. As gross margins become smaller and sale of live birds becomes less important compared to meat production, farmers will need these records. Records will also provide breeders with a tool to minimize inbreeding and prevent genetic diseases and inbreeding depression. Other issues like temperament in some ostriches need to be assessed for any genetic causes and this is possible only in the presence of pedigree records.
CONCLUSION
This study revealed that the ostrich industry is still very uneven as far as levels of production and enterprise sizes are concerned. In addition productivity is very low compared to the USA and South Africa. There is need to standardize breeding management practices eg length of breeding season and mating ratios. A small number of farmers with pedigree records do exist. However, most of them have been collecting records for less than 3 years. These can form the basis of a statistical genetic analysis and this will help answer questions about current and potential production levels, genetic parameters eg heritabilities, the levels of inbreeding and the associated depression in productivity and the potential for crossbreeding.
REFERENCES
Byford-Jones, C. (1995) The Farmer's Weekly, January, 1995. 8-9.
Foggin, C. M. (1992) Proceedings of the workshop on Ostrich management and production. University of Zimbabwe, Harare.
Foot, C. J. (1994) Proceedings of the Symposium on Livestock Production, Harare, Zimbabwe.
Gwaze, J. (1995) US ostrich breeders impressed by local tanning standards. The Sunday Mail, 5 March, 1995.
SAS. (1988) SAS Inst. Inc. Cary, N.C. USA.
Sharp, G. (1992) Proceedings of a Conference on Alternative Land Use-What Options, Harare, Zimbabwe.
Van Zyl, J. (1991) Farmer's Weekly, September, 1991. 18-19.
Ziana. (1995) Ostrich market flooded. The Herald, 3 November, 1994. APPENDIX 1
Questionnaire
Background
Two University of Zimbabwe Department of Animal Science scientists, Dr Kennedy Dzama and Professor John Topps are setting up a research project on aspects of ostrich production in Zimbabwe. As a start they would like to gather as much information as they can on ostrich production in Zimbabwe. They have designed a questionnaire and would be most grateful if they can get as many responses as they can get from farmers. Farmers need not identify themselves and they can make any additional comments when they respond.
Questions
Flock size
No. of breeding hens
No. of breeding cocks
No. of growing stock (non breeding ostriches)
Do you have individual or group mating? Individual/Group
Is your recording system such that every egg's cock and hen are recorded? Yes/No
If the answer to the preceding question was yes, for how long have you had this system?
Do you keep hen fertility records? Yes/No
How long is your breeding season and when is it?
Average no. of eggs per bird per season?
Average no. of fertile eggs per bird in one season?
Average no. of eggs per season by top bird?
Average no. of fertile eggs per season by top bird?
Average no. of eggs per season by worst bird?
Average no. of fertile eggs per season by worst bird?
Do you weigh your birds periodically? Yes/No
How often?
Do you keep any other growth records? Yes/No
Which ones? What is the source of your replacement birds? eg other farmers, from within your stock or imported?
Have you had an increase in congenital deformities in your flock? Yes/No
Would you classify the level of inbreeding in your flock as high, low or zero?
What is your ostrich paddock size?
How many ostriches do you keep per paddock? SOME ASPECTS ON CURRENT AND FUTURE LIVESTOCK BREEDING STRATEGIES IN LIVESTOCK IN SOUTHERN AFRICA
SUMMARY
INTRODUCTION
PRINCIPLES OF DESIGNING A STRATEGIC BREEDING PROGRAM
MAJOR CONSTRAINTS
SELECTION AND MIGRATION
ESTIMATION OF BREEDING VALUES
ACTION PLAN FOR THE ESTABLISHMENT OF EFFECTIVE AND EFFICIENT BREEDING PROGRAMMES
CONCLUSION
REFERENCES
C. Wollny University of Malawi, Bunda College of Agriculture, P 0 Box 219, Lilongwe, Malawi.
SUMMARY
The objective of this paper is to discuss and evaluate principles of breed improvement strategies and programmes to be applied in the Southern African region. Major constraints such as lack of adequate breeding policies, infrastructure, and know-how were identified. Limited information on breed characterization is available, which takes productive and adaptive performance into account. Few long-term comparative studies on indigenous, crossbred and exotic stock are available. There seems to be a considerable potential in using indigenous or local populations. Research findings indicate that indigenous animals are able to produce efficiently in harsh environments of the Southern African region. A need for characterization of available populations is emphasized. Chances and risks of nucleus-MOET schemes are discussed. Simulation studies demonstrated the need to evaluate the genetic and the economic potential before capital is invested into implementation of breeding strategies. In situations of limited foreign currency supply and high cost of capital, the risk of an economic failure could be significant. Breeding strategies based on exotic livestock imports or sophisticated reproductive technologies should be thoroughly assessed before adopted. Open nucleus schemes integrating livestock producers might be an appropriate strategy for several species and countries. A participatory approach is recommended. Continuous testing and implementation of a recording scheme is mandatory, if improvement in terms of biological and economic efficiency is to be realized. Antagonistic genetic relationships between traits of productive and adaptive performance require further analysis. Data processing for estimation of breeding values and to monitor genetic progress could be centralized in one or two countries of the region. Best linear unbiased prediction (BLUP) should become state of the art. Selection indices need to be developed taking into account biological and economic importance of traits under consideration. A strong need for networking among animal breeders is emphasized. An action plan targeting 10 strategically important areas is proposed.
INTRODUCTION
The main objective of improving livestock production through animal breeding strategies is to increase efficiency. Efficiency in its broad sense is defined as the product output per unit of input involving a complex relationship among factors such as feed input, maintenance feed requirements, level of reproductive and productive performance, infrastructure and breeding costs and income per unit of sold product. Breeding strategies to improve efficiency of livestock production comprise of selection schemes amongst populations, crossbreeding and biotechnological techniques to enhance the productive and reproductive performance or combination of such strategies.
In Africa, the production index for livestock products per capita dropped from 100 (1979/81) to 90.3 in 1993. Within the Southern African developing countries (SADC) Malawi showed the lowest value with an index of 68.9. In contrast, Asia improved to 161.7 and SouthAmerica realized a moderate increase to 104.4 in the same period (FAO 1994). The need to improve the efficiency of livestock production is obvious for the region.
Several theoretical and applied breeding programmes are still focusing solely on output such as milk yield ignoring the antagonistic relationships of yield with other important traits such as reproductive performance or health. In Southern Africa the availability of foreign currency is a determining factor in any breeding strategy involving exotic germplasm. In a harsh environment the realized response in improvement schemes may be well below expectations if important traits are not included in the aggregate genotype.
At the level of subsistence farming which is characterized by a generally low input - output system, the sustainability of animal breeding efforts to improve animal productivity becomes a dominant factor. The common multipurpose use of livestock by communal farmers requires that correlation among different traits have to be considered before conventional breeding schemes are adopted nationwide. In several countries of Southern Africa the lack or insufficient development of breeding policies and insufficient characterization of indigenous populations are preventing any, progress on the genetic improvement. More or less effective improvement schemes with exotic breeds of various origin result often in uncontrollable developments. Consequently, endangerment of local populations and unrecoverable loss of genetic material can be observed.
The objectives of this paper are i) to evaluate general principles of breed improvement strategies, ii) to identify major constraints on breeding programmes, iii) to review breeding programmes related to commercial and subsistence production systems and iv) to propose an action plan required to implement an effective and sustainable breeding strategy in the region.
PRINCIPLES OF DESIGNING A STRATEGIC BREEDING PROGRAM
Table 1 shows 7 steps, which have to be logically followed in designing a strategic breeding program. The general strategy in animal breeding should focus on optimization of the genetic potential according to production factors, the needs of the market, the ecological environment and future development. Countries in Southern Africa should develop and identify their own breeding objectives, testing schemes and breeding stock based on their own commercial conditions rather than taking blueprints from elsewhere (Smith, 1988). Characterization of indigenous populations and comparative performance trials require sufficient and accurate data sources since the choice of the foundation stock for any breeding program is of paramount importance.
Table 1. Design of breeding programmes
Principal thrust:
"Improve overall biological and economic efficiency of livestock production through provision of an optimized genetic potential fulfilling the needs of the market or the subsistence farming system" Step 1: Identify production system(s) and economic merit Step step 2: Define breeding objective based on Step 3: Evaluate available populations for breeding purposes and select the best stock Step 4: Develop breeding systems involving selection or migration or both Step 5: Develop improvement schemes based on testing and selection Step 6: Disseminate improved animals to the livestock industry Step 7: Monitor steps 1 to 6 and adjust for future trends
MAJOR CONSTRAINTS
Several countries of Southern Africa face a lack of infrastructure for breeding purposes, relatively small defined populations, little exchange of breeding stock, lack of suitable subpopulations or lines for crossbreeding and high costs for the import of genetic material from exotic breeds. A general feature is the small number of, if any, stud breeders in several countries of the region, which makes the implementation of an effective national breeding policy virtually impossible. After dissemination of crossbred stock to farmers subsequent coatings occur often uncontrolled in various directions and performance data are not available for the estimation of breeding values. It could be stated that a characterization and weighed performance evaluation of available indigenous populations is still missing or incomplete for all important livestock species including local strains of poultry. Major constraints for the improvement of livestock production in Southern Africa are listed below:
Sector policy and breeding programmes
1. Lack of national breeding policies
2. Insufficient conservation strategies for genetic resources
3. Lack of well defined breeding strategies
4. Non-sustainable or lack of continuity of breeding programmes
5. Little support of relevant research activities
6. Insufficient number of qualified animal breeders available
7. Dependency on foreign currency supply situation
Infrastructure l . Lack of performance recording schemes in several countries especially in the smallholder sector and for indigenous breed
2. Breeders organization in many countries insufficient or non-existent
3. Communication, transport and computation facilities insufficient or not available Breeding Programmes
1. Breeding objectives often non-existent or vague
2. Ineffective sire exchange and A.I. programmes
3. Small population sizes, small herd sizes and unreliable animal identification
4. Insufficient characterized indigenous populations
5. Genotype x environment interactions often neglected
Selection and genetic gain
1. Long generation intervals through extended and late maturing animals
2. Low selection intensity through high mortality rates and limited capacity for performance testing
3. Low accuracy of estimated breeding values due to small active population
4. Inbreeding effects
5. Assumed antagonistic relationship between genetic merit for production and adaptation
SELECTION AND MIGRATION
Focus on indigenous populations
Selection within population or breeding for traits of medium to high heritability such as daily gain or lean meat percentage could be a promising and sustainable strategy in developing countries. Breeding costs could be kept low by not importing exotic livestock. The improved product is likely to be acceptable. The strongest argument against this strategy, however, focuses on the low progress over a given time period due to low level of performance of some of the indigenous livestock. Mass selection programmes may also be very difficult to organize due to the existing constraints. Insufficient characterization of local populations prevented in the past to set-up a viable long-term selection program. A strategy based on selection within indigenous populations for economically relevant traits assumes sufficient genetic variation among local tropical breeds, which is the case (Hetzel, 1988; Bondoc et al., 1989; Tawonezi, 1989a; Vacarro et al., 1994). Selection within local populations may facilitate further improvement of crossbreds and viable breeding programmes for local breeds may help to conserve genetic resources.
Several studies have shown, that African large ruminants can be highly productive largely due to their reproductive rates (Hetzel, 1988). The author concludes that the index of biological productivity should refer to the production systems that accounts for age of progeny at sale or utilization. By this, the most relevant trait such as postweaning growth for beef cattle could be considered in the breeding plan. The economic value depends, however, on price differences according to grading if such a system is in place. Similar principles apply to small ruminants, pigs and poultry. Bondoc et al. (1989) reviewed examples of indigenous dairy cattle in Kenya and India, which have been successfully selected after minimum requirements such as well- defined objectives, recording scheme and AI had been established. Further examples of selection within indigenous cattle for beef or dual-purpose production are given for Mashona cattle in Zimbabwe (Tawonezvi, 1989b) and for Sahiwal in Kenya (Mwandotto, 1994). Mwenya (1990) pointed out that considerable overlap in cattle breeding activities exists in Southern Africa. However, research results on indigenous cattle breeds and comparative trials of crossbreeding programmes are often of very limited value due to small numbers, environmental conditions, design, and analysis of data and not a single research study included adaptive traits such as tick resistance or heat tolerance. The adaptation of the breed to the production environment as a determinant of production efficiency can be described by a well defined analysis of environmental influences on economically relevant traits. Studies such as conducted by Tawonezvi (1989b) and Moyo et al. (1994) are important in assessing the potential of local breeds. In the latter work, the comparative maternal productivity of indigenous and exotic cattle were demonstrated and the local breed Mashona ranked first in terms of annual weaner production per 100 kg cow metabolic liveweight. Comparative evaluation of indigenous breeds to Southern Africa under different environments are essential to develop adapted and highly productive populations. African breeds of large and small ruminants represent a unique source of genetic resources as most of them have evolved in tropical environments. Improved disease control, AI and ET technology may enable faster dissemination of suitable breeds over wider areas.
The need for characterization of indigenous populations is of high importance and the comparative evaluation of environmental effects on expression of traits is required to understand the interactions between genotype and environment. In small populations and under the constraint of an insufficient infrastructure, however, it is difficult to establish a breeding program, which can survive. National livestock coordinators should consider to pool their breeding strategies and programmes.
Evaluation of breeding strategies
The predicted genetic response of breeding strategies utilizing regionally or worldwide defined populations depends largely on the differences in genetic merit between subpopulations. The intensity of selection , based on highly accurate estimated breeding values and the generation interval are determining factors for genetic gain. Biological and economic efficiency must be assessed before a specific scheme is finally recommended. The biological efficiency of breeding strategies may increase from a simple mass selection program based on individual culling to various AI programmes including progeny testing and multiple ovulation and embryo transfer (MOET) schemes and may be further enhanced through application of gene technology such as marker assisted selection (MAS) or even gene transfer. At present the genetic progress is completely based on application of quantitative genetic methodology. Relative efficiency of MAS is probably highest in nucleus breeding systems and for traits of low heritability. MAS and similar techniques require a considerable capital investment and know-how input (Brascamp et al., 1993) and present expectations about MAS may be too optimistic since they are based on the simplistic concept of additive gene effects (Simianer, 1995).
The theoretical concepts of closed or open nucleus breeding schemes applying Al and MOET are well developed (Nicholas and Smith 1983). Nucleus breeding schemes have been recommended for developing countries (Smith 1988). Some of the major advantages and disadvantages of a nucleus-MOET program are presented below:
Advantages of Nucleus-MOETprogrammes
improved accuracy of testing and recording due to a better controlled environment recording of additional and secondary traits is easier various reproductive techniques could be applied superior in small populations (< 50 000 cows) versus AT progeny testing scheme reduction of generation interval higher female reproduction rate genetic lift in establishing the unit utilization as conservation technique for genetic resources training centre and base for research
Disadvantages and risks of Nucleus-MOET programmes
risk from concentrating animals and resources in one centre risks due to diseases and epidemics genotype x environment interactions likely inbreeding rate could become a problem in closed schemes high initial investments required recurrent cost recovery difficult ('cost of capital') high dependency on efficiency of bio-techniques (MOET) variability of progeny per donor (MOET) farmer and industry commitment and integration required logistical problems to disseminate stock to production herds
Mpofu et al. (1993a,b) evaluated breeding strategies for commercial dairy cattle in Zimbabwe from a genetic and an economic point of view. Based on the assumptions of their simulation studies the biological and the economic evaluation did not yield the same results. The authors compared local programmes based on progeny testing in a closed population, progeny testing combined with semen importation to sire 30% of the cows, progeny testing with foreign sirs as sires of bulls, and a closed nucleus selection scheme using embryo transfer started from elite imported stock. Further strategies evaluated were based on continued importation of semen for 30, 50 or 100°/0 of the cows and semen from elite foreign bulls used on local elite cows and, finally, bulls from elite imported embryos. Similar selection objectives and negligible genotype x environment interactions are assumed. Imports of exotic genetic material are justified, if genetic merit of the imports is considerably higher than genetic merit of the local stock. Based on a population of 100 000 cows with 40% milk recorded and 80% AI the nucleus MOET scheme ranked first among the local strategies. Based on an initial difference of the two populations of 1.25 standard deviations [phenotypic (ISD = 540 kg)], which was assumed for all strategies, the continuous importations of semen for 100% of the cows ranked highest among the import strategies. The economic evaluation, however, showed different results due to the country specific input factors. The authors conclude, that local progeny testing is an economically viable alternative and continuous importation of above average genetic merit semen is suitable only at bargain prices. On the other hand, imported semen from elite bulls and to be used to sire sons might be an alternative, if reproductive techniques (MOET) are highly efficient in practice. Table 3 presents the net present value for the nine simulated strategies based on achievable genetic response and assuming a genetic correlation between exporting and importing country of r = 0.7, an initial genetic difference of 1.25 SD, a discount rate of 10% and an exchange rate of USD to
$Zof 1:3. A decreasing genetic difference and a lower genetic correlation would benefit the closed progeny testing scheme. In developed countries and in a highly competitive environment, open nucleus-MOET schemes become more and more attractive and operate successfully. Instead of setting-up costly stations for keeping a nucleus herd breeding associations may choose to contract the best herds to create an elite subpopulation. It should be emphasized, that the study by Mpofu et al. (1993a,b) can not be related to evaluate smallholder dairy breeding programmes on communal land but demonstrates the necessity to conduct detailed analysis of the predicted genetic and economic response, before investments into a costly strategy are made. Table 3. Net present value at yr 25 for nine breeding strategies in dairy cattle in Zimbabwe (Mpofu et al., 1993b)
Strategy Net present value (Mill. $Z) Rank Closed progeny testing scheme (PT 1) 4.57 4 PT 1 but 30 % of cows sired by foreign bulls (PT 2) -5.82 5 PT 1 but with imports of for sire of sons (PT 3) 7.01 3 Closed nucleus multiple-ovulation embryo-transfer herd -30.66 8 (NMOET)
Continual semen import for 100 of the population (CSI -36.01 9 100) CSI 50 (50 % import) -15.56 7 CSI 30 (30 % import) -9.69 6 Untested young bull team from imported embryos 24.78 1 Untested young bull team foreign elite bulls x local elite 15. 96 2 cows(ELITE)
Breeding strategies and the environment
It is well established, that selection for improvement in quantitative traits such as milk yield may result in a decline of fitness (Dunklee et al., 1994; Falconer, 1989). Bradford et al. (1994) concluded from a selection experiment conducted over 30 years in Targhee sheep, that selection in the environment of use is preferable to importation of livestock from an environment where heritability is higher. In the semi-arid and arid climatic zones of Southern Africa the selection within the less favorable environment, even taking into account reduced rates of response to selection especially for traits of lower heritability might be more appropriate than selection conducted in a non-representative environment.
In well structured breeding programmes the establishment of crossbreds, composites or new breeds could be viable alternatives to selection within a given population. Cunnigham and Syrstad (1987) concluded from a review of crossbreeding dairy cattle in the tropics based on immigration of various proportions of Bos taurus genes into Bos indicus, that linear improvement in productive traits examined up to 50% Bos taurus but no clear trend at higher proportions was observed. Heterosis was estimated at 28% for milk yield. F2 performance was below expectations possibly due to epistatic effects. Further studies, confirming such findings, were reviewed by Davis and Arthur (1994). Studies should also investigate to what extent specialization of high grade crosses are antagonistic to low-input farming systems in the communal areas. In Venezuela high grade Friesian x Zebu crosses were inferior on lower performing farms compared to medium grade crosses and Zebu-type cows (Vacarro et al., 1994) indicating important genotype x herd interactions. However, if an AI program targeted on smallholder farmers comes to a halt, ambitious crossbreeding programmes collapse. In Malawi, the dairy cow population consists mainly of various grades of Malawi Zebu x Friesian crosses in the smallholder sector (n = 4173 cows, Ministry of Agriculture, 1993) and a couple of thousand purebred Friesians or Holstein Friesians. On private and parastatal estates herds are maintained on continuous importation of semen. An assessment of the efficiency based on evaluation of input-output variables of the various crosses is not available and selection based on performance in the semiarid climate of Malawi does not occur.
In general, any migration strategy based on exotic breeds implies a risk component which increases with the genetic differences between the local and the exotic population. In Malawi, even well adapted improved breeds such as Dorper sheep or Boer goats crossbred Nvith the local small ruminant population in a harsh environment could probably not improve overall efficiency due to high mortality rates (MGLDP, 1992). In this analysed nucleus herd the additive and heterotic effects could4 not account for the huge effects exerted by environmental, factors such as management. Selection within the local populations would have been more appropriate taking into account the scarcity of foreign currency to import exotic breeds. The present situation is that purebred or high grade exotic breeds are not available in sufficient numbers to ensure continuity of the station.
Group breeding schemes based on a participatory approach for the smallholder and communal sector may offer an efficient system to improve livestock in several countries of SADC. In the smallholder sector and for goat and sheep breeding, where no progeny testing and AI scheme exists, breeders may adopt cooperative breeding schemes. A number of interested farmers record their flock, select the best females and send them to one unit forming a nucleus. The nucleus is kept open for highly productive females and selected males are used as replacement sires in the cooperating farms. The maximum rate of gain is achieved, when 5 to 10% of the total number of animals are kept in the nucleus. Ricordeau (1991) describes an example of such a system for sheep in New-Zealand. Group breeding system reduces inbreeding due to regular and two-way exchange of stock. In addition, it provides a "focal point among participating breeders" and the evaluation of performance occurs in the production environment (Niekerk and Schoeman, 1993). Vaccaro et al (1994) described a functioning cooperative breeding system for dual purpose cattle involving 6 arms to produce sires, which are offered tthe 18 members of the project in Venezuela. The monitoring and supervision is provided by the local university.
Simple nucleus schemes could well function, if the participation and integration of farmers or villages is ensured. The dissemination of improved sires may occur 'via AI in a better developed environment or via natural service in poorer conditions. Within herds, however, the simplest form of selection, the culling of the least productive animals could realize an important potential (Vacarro et al., 1994). This is one of the strongest arguments in favor of performance recording under difficult conditions in developing countries. The introduction of regional performance testing schemes relying on owners of the animals is one of the actual challenges to provide sufficient data for animal breeding programmes. Examples on such systems can be found for dairy cattle in South Africa (Loubser, 1994) and remote controlling systems might be available in future (Davis and Arthur, 1994). Independently of the breeding scheme applied, continuous evaluation of productive and reproductive performance in inevitable.
ESTIMATION OF BREEDING VALUES
Any genetic progress is based on selection and the required information for genetic merit is based on estimations. Quantitative methods to estimate the breeding value could utilize data supplied from all levels of livestock production including smallholder and subsistence farmers. The basic requirements are identification of animals in their environment, standardization of meaningful data evaluation schemes and a sufficient number of records to ensure statistical validity. The strategic goal is to improve the rate of gene flow through an organized breeding sector and to disseminate genetic progress to the production segment.
The BLUP is the presently best available prediction of the true breeding value taking into account non-genetic factors such as herd, season, year but also sex, age of dam etc., which have an impact on the expression of the investigated traits, the genetic relationships among relatives and their performance records and the genetic ability. The estimation of the breeding value (EBV) takes into account the genetic values of the mating partners and environmental effects on contemporary groups including their genetic values. Genetic improvement is cumulative in contrast to any other strategy applied. More accurate estimation of breeding values have resulted in realized significant annual increases of annual genetic gains in already high performing herds elsewhere. It could be expected that this effect may be even greater in little or unimproved populations of Southern Africa with probably high genetic variation.
ACTION PLAN FOR THE ESTABLISHMENT OF EFFECTIVE AND EFFICIENT BREEDING PROGRAMMES
Any viable and sustainable breeding program is based on the economic and biological efficiency of its products. Immediate action is required as livestock production can presently not cope with the increasing demand of the human population. In the least developed SADC countries the situation of the livestock sector seems to be extremely poor and may require significant assistance from neighboring states. Based on the reviewed information the following actions are recommended for the SADC region to achieve measurable impact:
1. Analyses of production systems for economic merit in existing and future national, regional and overseas markets for livestock products. Assess needs of the subsistence sector to improve income generation in rural areas. 2. Develop as a matter of urgency breeding policies and implement measures to avoid further uncontrolled developments resulting in extinction of indigenous populations and inefficient crossbreeding programmes. 3. Analyses of existing recorded performance records of indigenous, exotic and crossbred populations and continue to conduct comparative studies under standardized typical environmental conditions involving sufficient numbers of animals. Analyses assuming antagonistic relationships between productive and adaptive traits. Develop cost efficient and effective field performance testing schemes. 4. Define and record secondary traits of importance for multipurpose livestock, such as disease resistance or utilizing locally available feed resources. 5. Conserve valuable genetic resources. Establishment of a regional network on conservation issues might be of assistance. 6. Evaluate genetic and economic merit of planned breeding strategies before exotic stock or prestigious technology is imported. 7. Disseminate improved livestock to producers by applying a participatory approach, ie farmers must be integrated to achieve ownership of the program. Develop, inter alia, group breeding and open, decentralized nucleus schemes. 8. Review the impact of state owned and managed nucleus herds, breeding station and extension organizations. Market development and orientation as well as cost recovery should become a high priority. 9. Pool national efforts and utilize existing facilities, know-how and technologies region- wide, such as estimation of breeding values or testing of innovative technologies. 10. In SADC, professional animal breeders should start a concerted action to implement those actions and define appropriate breeding strategies and programmes to serve the livestock industry. Industry must provide data to enable assessment of breeding programmes.
CONCLUSION
In contrast to any other technology, genetic improvement results per se in sustainable, cumulative and multiplicative effects. In SADC countries breeding policies and programmes should be developed or fine-tuned focusing on the biological and economic efficiency of livestock in the various farming systems. A strong need has been identified to characterize local populations in terms of productive and adaptive performance. National efforts alone may not be successfttl due to the very limited resources of capital and know-how in most of the SADC. On the other hand, regional centers such as nucleus-MOET herds may not be sufficiently efficient in the dissemination of improved genetic material due to the major logistical problems in the region. It is strongly recommended to network and to utilize facilities for testing purposes and data analyses in the region. Large and small scale livestock farmers should cooperate in developing decentralized breeding schemes and standardized recording systems. Centralized data processing to estimate breeding values (BLUP) should be made available by the Universities and natiopal agricultural research stations by providing management information on selection procedures of animals. Importation of exotic breeds or capital intensive technology requires a preceding detailed evaluation of the potential effects. Agro-ecological related decentralized open nucleus schemes applying standardized data recording schemes could be the best compromise in the less developed countries to improve livestock and to conserve genetic resources. The indigenous genetic resources of livestock may have a huge potential, which is not yet exploited. The conservation of such valuable germplasm should be regarded as mandatory for future generations.
REFERENCES
Animal Improvement Institute. (1995) Irene, Pretoria, South Africa.
Bondoc, 0. L., Smith, C., and Gibson J. P. (1989) Anim. Breed. Abstr. 57: 819-829.
Bosman, D.J. (Ed.). (1994) Livestock Improvement Schemes, Irene, Pretoria, South Africa.
Bradford, G., Sakul, H., Neira, R., Famula, T., Dally, M. and Finley, C. (1994) Proc. 5th World Congr. Genet. Appl. Livestock Prod. 18: 95-98.
Brascamp, E. W., van Arendonk, J. A. M. Groen. A. F. (1993) J. Dairy Sci. 76: 1204-1213.
Cunningham, E. P. and Syrstad, 0. (1987) FAO Animal Production and Health Paper No. 68, Rome, Italy.
Davis, G. P. and Arthur, P. F. (1994) Proc. 5th World Congr. Genet. Appl. Livestock Prod. 20: 332-339.
Dunklee, J. S., Freeman, A. E. and Kelley, D. H. (1994) J. Dairy Sci. 77: 3683-3690. Falconer, D. S. ( 1989) 3rd ed. Longmans, Essex, England.
FAO. (1994) Production Yearbook. Food and Agricultural Organisation, Rome, Italy.
Galal, E. S. E., Ahmed, A. M., Abdel-Aziz, A. 1. and Younis, A. A. (1993) Small Ruminant Research 10: 143-152.
Hetzel, D. J. S. (1988) Anim. Breed. Abstr. 56: 243-255.
Loubser, L. F. B. (1993) In: Annual Report 1993. Department of Agriculture, Livestock improvement schemes, Irene, Transvaal, South-Africa, pp. 20-28.
Malawi German Livestock Development Programme. (1992) Proc. Ministry of Agriculture, Dep. of Anim. Health and Industry, Lilongwe, pp. 5-8.
Ministry of Agriculture. (1995) Statistics for the smallholder dairy farmers 1993. Department of Animal Health and Industry, Lilongwe, Malawi.
Moyo, S., Regc, J. E. 0. and Swanepoel, F. J. C. (1994) Proc. 5th World Congr. Genet. Appl. Livest. Prod. 20: 344-347 .
Mpofu. N., Smith, C. and Burnside, E. B. (1993a) J. Dairy Sci. 76: 1163-1172
Mpofti, N., Smith, C., van Vuuren, W. and Burnside, E. B. (1993b) J. Dairy Sci. 76: 11731181. Mwandotto, B. A. J. (1994) Bull. Anim Prod. Afr. 42: 61- 67.
Mwenya, W. N. M. (1990) International Livestock Center of Africa (ILCA), Addis Ababa, Ethopia.
Nicholas, F. W. and Smith, C. (1983) Anim. Prod. 36: 341-353.
Nickerk, W. A. and Schoeman, S. J. (1993) In: Maree, C., N.H. Casey (Eds.): Livestock production systems. Agri-Development Foundation, Brooklyn, New York, USA: 124-149. Ricordeau, G. (1991) In: Maijala, K. Ed.: Genetic resources of pig, sheep and goat. World Animal Sciences, B8, Elsevier Science Publishers B.V., Amsterdam, pp. 495-515.
Schoeman, S. and van der Merwe, C. (1994) Proc. 5th World Congr. on Genet. Appl. Livest. Prod. 18: 91-94.
Simianer, H. (1995). Perspectives of the use of genome mapping in animal breeding), Mariensee, 1995 (n press)
Smith, C. (1988) World Animal Review 65: 2-10.
Tawonezvi, H. P. R. (1989a) Trop. Anim. Health. Prod. 21: 37-42.
Tawonezvi, H. P. R. (1989b) Trop. Anim. Health. Prod. 21, 170-174.
Vaccaro, L., Vaccaro, R., Verde, O., Mejias, H., Perez, A., Rios, L., and Romero, E. (1994) Proc. 5th World Congr. Genet. Appl. Livest. Prod. 20: 313-318. PIG IMPROVEMENT PROGRAMMES IN ZIMBABWE
SUMMARY
INTRODUCTION
TRAITS OF ECONOMIC IMPORTANCE
CONCLUSION
E. Takaindisa Pig Industry Board. P O Box HG 297 Highlands, Harare, Zimbabwe.
SUMMARY
At the Pig Industry Board (PIB), pig improvement is based on performance testing. Animals are tested at nucleus level at the central testing station. A selection index based on feed conversion efficiency and backfat thickness is used. The quality of legs, carcass and teats is also assessed. Multiplication testing is conducted on farm. Selection criteria is based on the age at 86 kg and backfat thickness. A visual appraisal is also done on the external features. Approved animals are then made available to the other producers.
INTRODUCTION
Genetic improvement within breeds can only be achieved by means of selection. Selection is the act of choosing those animals which will be the parents of the next generation and which have the highest breeding value. Information on the breeding value can be obtained by sib testing, progeny testing or performance testing.
Sib Testing
This is an attempt to assess the value of breeding stock on the basis of the performance of their brothers and/or sisters.
Progeny Testing
Parents are selected on the basis of the performance of their progeny. Progeny testing is the only effective way to assess an animal's ability to transmit high merit in terms of reproductive traits to the progeny. Progeny testing is much more expensive than performance testing and it takes a longer period to assess an animal. Before the animal can complete the test it has to be sexually mature, mated and its offspring go through the test period. In the case of pigs the offspring would be 5 to 6 months of age on completion of the test. Since the gestation period of sows is close to four months and sexual maturity is obtained at about seven months of age, it means the results of a boar undergoing progeny testing become available when the boar is 1 approximately 1 /2 years. During this period the boar has to be housed and fed whilst he is not being used. Progeny testing is however, an accurate way of assessing traits with low heritability. Progeny testing was once used in Zimbabwe but it has since been stopped. Performance Testing
It involves evaluation of an animal by looking at its own performance. Performance testing is more effective for those traits which are of high heritability. This method can also be called individual selection. Performance testing is the selection method which is used at PIB. The testing programme involves the selection of breeding stock based on the animal's own performance. The use of the animal's own performance as a measure of its genetic merit is true for traits of high heritability like backfat thickness and average daily gain. The traits to be selected for are combined into a selection index. Each trait is weighed according to its economic value. The heritabilities of the traits selected for are also considered in the formulation of the index. The advantage of the index method is that exceptional performance in a trait makes up for a weakness in another.
TRAITS OF ECONOMIC IMPORTANCE
The traits of economic importance include feed conversion efficiency, carcass and reproductive traits.
Feed conversion efficiency
Heritability of this trait is around 0.35, thus performance testing results in a reasonable response. Measurement of individual feed conversion requires facilities for feeding the pigs individually hence it is very expensive to assess. Feed conversion is highly correlated to daily gain. As a result average daily gain is used in place of feed conversion efficiency for pigs that are tested on farm where feeding the animals individually is not practicable.
Backfat
The amount of fat in a carcass determines the quality of the carcass. The more the fat the poorer the quality. Backfat thickness is highly heritable and as such can be improved throughperformance testing. Backfat thickness can be measured in the live animal by using an ultrasonic machine. The measurement of the amount of fat in the live animal through an ultrasonic machine enables the inclusion of backfat thickness in the selection index of performance tested breeding stock.
Reproductive Traits
Reproductive performance is one of the most important trait but unfortunately hardly any improvement in this trait can be achieved through performance testing because of its low heritability h2 = 0.10. Improvement of the environment together with cross breeding are the most effective means of obtaining improvement in this trait.
PIG IMPROVEMENT SCHEME
A small number of registered nucleus breeders have their pigs tested at the Pig Industry Board Central testing station. Approved gilts are either retained by the farmer for multiplication or sold to other producers. The progeny of retained boars are then performance tested (multiplication testing) on the farm. Animals which pass the test are made available to commercial producers. This system ensures a constant supply of replacement stock to commercial herds.
Nucleus Testing Animals are centrally tested at the Pig Industry Board station. There are 6 breeders in the country and these bring in their animals for testing. This system necessitates the bringing together of animals from different sources which is contrary to the basic principles of disease control. To reduce the risk of spreading diseases the animals are quarantined for a period of three weeks before being moved into the main station. Whilst the animals are in quarantine a veterinary doctor comes to inspect them on a weekly basis. Any animal which dies in the station is sent to the veterinary laboratories for post mortem.
The breeder pre-selects the animals to be sent for testing. Pigs are only taken in for testing if they are from gilt litters of not less than 9 born and 7 weaned and sow litters of not less than, 10 born and 8 weaned. Control over numbers weaned per litter assist in standardising the pre- test environment.
Performance is measured from 35-86 kg liveweight. During this period the pigs are individually fed and all the feed issued is recorded. The animals are tested on a restricted feeding system. Any treatments done on the pigs are recorded. At the end of the test the feed conversion efficiency is calculated and back fat thickness measured. The amount of fat is measured at the 1 P2 - position which is 7 /2 cm from the midline along the last rib. This is the same position used by Government graders on slaughtered pigs throughout the country. The two measurements, that is feed conversion efficiency and backfat thickness are combined into an index. Pigs which fail to attain an index of 100 are culled.
A visual appraisal on external features is also done. In doing the visual appraisal the following factors are taken into consideration; Pigs with obvious faults such as genetic conditions like hernia and cryptorchidism are rejected and pigs with leg weaknesses are also rejected. The gait of the animal should be free and easy and not stiff. The pasterns should be short and inclined upright. Too open and unbalanced digits are not desirable. The animal to be selected should be wedge shaped. The shoulders should be light in relation to the hams. The hams should be filled down to the hock. The animal should have at least 12 well developed and evenly spaced teats. Animals with blind or inverted teats are culled. Males with poorly developed testicles are culled.
A disadvantage of this system is that its subjective but the best guideline is to keep a picture of an ideal pig in mind. Consideration not to cull animals with an excellent index score are made unless it is unavoidable. The breeder is notified of the results of the animal. Animals which fail the test are sent for slaughter and approved animals are sold back to the breeders. The breeders can either sell the animals or retain them on their farm for further breeding.
Multiplication Testing
The test is conducted on the breeder's farms. The animals tested are the progeny from either nucleus or multiplication tested animals. Multiplication testing is a faster method of performance testing as the number of animals tested is not limited. Approval is based on the animals age at 86 kg which is an indirect measure of average daily gain and also the amount of fat at the P2 position. The system is based on independent culling and the following approval standards are used.
Boars Gilts Age to reach 86 kg < 170 days < 175 days P2 < 19 mm < 20 mm
Any animal above the stated values is culled. Animals which pass the test are then made available for sale to other producers. CONCLUSION
The performance testing scheme which has been discussed need some modifications especially on the feeding system under which the animals are tested on. In future the animals coming for nucleus testing should be assessed on an ad libitum feeding system which is widely used by producers in Zimbabwe. As markets become more and more sophisticated in future there might be need to assess animals on the basis of lean meat content rather than backfat thickness. FUTURE PROSPECTS IN DAIRY CATTLE IMPROVEMENT IN ZIMBABWE
SUMMARY
INTRODUCTION
GENETIC EVALUATION OF DAIRY CATTLE IN ZIMBABWE
NON-GENETIC FACTORS AND PART LACTATIONS
SELECTION METHODS
CONCLUSION
REFERENCES
F. N. Ngwerume1, C. Banga2 and V. Muchenje1 1University of Zimbabwe, Department of Animal Science, P 0 Box MP 167, Harare, Zimbabwe. 2Zimbabwe dairy Herd Improvement Association, P 0 Box CY 2026, Causeway, Harare, Zimbabwe.
SUMMARY
In order to make accurate culling and selection decisions, accurate genetic evaluations are needed. Recent advances in animal breeding theory and increased computer capacity have made animal model evaluation computationally feasible. Animal model evaluations use information from all known relationships among animals to predict each animal's genetic merit. Non-genetic factors affecting production traits must be properly adjusted for when evaluating an animal's genetic merit.
INTRODUCTION
Milk production is influenced by a number of non-genetic factors. When attempts are made to estimate the genetic value of an animal, the effects of some of these factors have to be considered. Adjusting records for known causes of variation is a must in making culling and selection more accurate. The non-genetic factors affecting milk yield are well documented in numerous investigations reported in literature. The sizes of different effects and their mutual interrelations as well as estimating problems have been thoroughly investigated.
Dairy producers need accurate genetic values that estimate an animal's genetic worth for economically significant traits. The ultimate goal for any dairy enterprise is to maximise profits from cows within a herd. It is, therefore, important for producers to have knowledge of genetic evaluations for important traits. Knowledge of which traits are most significant and the overall genetic worth of the animals, all traits being considered is vital.
Major dairy cattle breeding countries have developed systems whereby field data can be analysed inorder that genetic values are available for all animals. Over the past three decades, major changes have occurred in the methodology of analysis and in the reporting procedures. In fact there has been a shift from the actual observations to an estimated breeding value for an animal as it relates to the population that the animal is compared to. This shift has resulted in very significant increases in the accuracy of predicting the animal's genetic worth.
GENETIC EVALUATION OF DAIRY CATTLE IN ZIMBABWE
The greatest problem in obtaining improvement through breeding is to choose animals that have the greatest genetic value. Recent improvements in the processing power of computers and developments in BLUP computing strategies has led to the widespread acceptance of the concept of evaluating individual animals for genetic merit rather than just one portion of the population, traditionally the sires. An accurate method of genetic evaluation of dairy cattle is needed in Zimbabwe.
Sire model
The first genetic evaluation programme for dairy cattle in Zimbabwe was set up in 1986 by the Dairy Services branch of the Department of Research and Specialist Services with the assistance of the Royal Netherlands Cattle Syndicate (NRS). Under the programme, the BLUP sire model procedure is used to evaluate local and imported sires for milk yield, fat yield and fat per cent annually. Estimation of breeding for type traits commenced in 1989. Only first lactation records were used in the estimation of breeding values. The Sire Model used is:
Yijkl = HYSi + Gj + S(i)k + eeijkl where: yijk = first lactation milk yield;
HYSi = fixed effects of the ith herd-year-season;
Gj = fixed effects of the jth genetic group;
S(j)k = random effects of the kth sire nested within the jth genetic group; eijkl = random residual.
The yield traits are pre-adjusted for age at calving and season of calving. Sire evaluations are published as predicted differences (PD). The average of sires with a repeatability of at least 50% is used as the base.
Cow genetic evaluation are done separately using an individual's first three lactation yield records and information from relatives. Records are also adjusted for age and season of calving.
Animal Model
Individual animal model evaluation has become an accepted method of genetic evaluation world wide. The animal model predicts the genetic merit of each animal in a population from the animal's own production records (if available) and the production records of all related animals. In the USA the Animal Model evaluation method replaced the Modified Contemporary Comparison (MCC) in 1989 (Wiggans and VanRaden. 1989). Major benefits of the animal model are use of all relatives rather than just certain classes of relatives and use of exact statistical procedures (Best Linear Unbiased Prediction) rather than approximations. The animal models allows simultaneous genetic evaluation of bulls and cows with all relationships included. In matrix notation, an animal model can be represented as y=Mm+Za+ZAgg+Pp+Cc+e where y = vector of standardised milk, fat or protein; m = vector of effects of management group; a = vector of random additive genetic effects; g = vector of unknown parent genetic groups; p = vector of permanent environmental effects; c = vector of herd-sire interaction;
M, Z, ZAg, P and C are incidence matrices for these effects; e = vector of random residuals.
The matrix Ag. relates animal to unknown ancestor groups and is equivalent to A,10Q as reported by Wiggans et al. (1988). Vectors a, p, c, and e are mutually uncorrelated with 2 2 2 2 variances A aa , 1P , Ic , Re respectively. The matrix A is the additive relationship matrix among all animals in a. In Zimbabwe, the animal model evaluation method has not yet been implemented. There arc plans by the Zimbabwe Dairy Herd Improvement Association to implement the animal model by September 1995.
NON-GENETIC FACTORS AND PART LACTATIONS
Age and Season adiustment factor
Among the numerous identified non-genetic within-herd influences on milk yield and milk components, age at calving of a cow is one of the main factors affecting milk, protein and fat yields in dairy cattle. The effect of cow's age on milk production was recognised long ago and studied to facilitate comparison of cows of differing ages. Yield increases with age at a decreasing rate and reaches a maximum at maturity. Yield then decreases as cows become still older. Auran (197 3) reported that age explains about 20-40 `% of the total variation in milk production. Influence of month or season of calving on production records is also well established. In Canada, an age-season adjusted record is known as a Breed Class Average (BCA) and in the United States the adjusted records are known as Mature Equivalents (ME). Age season adjusted records are used to compare lactation yields of cows that calve at different ages and in different seasons. No age and season adlusttnent factors have been developed for the dairy population in Zimbabwe. Currently the age and season factors developed in Canada are being utilised to adjust records to a mature basis. Using such factors may causes biases in making culling and selection decisions. Genetic parameters estimated using these adjustment factors may not be accurate thereby causing biases in genetic evaluations. Such factors have been developed for breeds and environments that are different from the ones in Zimbabwe. There is therefore a need to estimate age and season factors using field data collected locally inorder to precisely evaluate the local dairy cattle.
Part Lactations Genetic evaluation of dairy sires has been based, for many years, on the analysis of 305 day (305-d) lactation yields. The basis of 305-d yield is a set of test-day yields taken at approximately 30 day intervals. As such, the measurement of lactation milk production in dairy cattle has been accepted as the first 305 days following freshening. This standard length, allows records to be compared without concern for the length of the production period. However, one difficulty is that a cow must have the opportunity to complete 305 days in milk before this measure of her productive ability exists. For cows that are culled, sold or die, this information is never available meaning that the 305-d yield must be estimated. In many cases 305-d lactation yields are estimated from lactations that are in progress. The advantages of extending records to 305-d production are numerous. The prediction of lactation total is important for early breeding values and farm management purposes. For example, a producer is able to identify his low producing cows earlier and make culling decisions sooner. Prediction of 305-d records for lactations in progress and culled cows provides data from more daughters in evaluating dairy sires (Congleton and Everett, 1980; Wilmink, 1987; Danell, 1981). In his review, Danell (1981) pointed out that extending part lactations to 305-d offers the potential of shortening the generation interval. Also, it is possible to reduce breeding program costs by culling progeny tested bulls with low breeding values for milk up to a half year earlier than when using completed 305-d lactations. However, the accuracy of extending records to a 305-day yield will depend on the number of test-days involved and the method used to project these test-day records.
Extension factors have not yet been computed for the Zimbabwe dairy records. Research focusing on development of such extension factors for the Zimbabwe dairy populations is, therefore, essential. Test day models eliminate the need for extension factors (Ngwerume, 1994). Research should be conducted to determine the application of test day animal models in Zimbabwe.
SELECTION METHODS
Selection based on an index incorporating many traits is increasingly becoming more important world-wide. In Canada a Lifetime Profit Index is currently in use (Chenais 1995. personal communication). The LPI includes several traits. The weighting on each trait in the index is based on its economic importance. The Canadian LPI is
LPI = 6(6PY + 5FY) + 4(4MS + 3FC + 2FL + 1C) where:
PY =Protein yield
FY =Fat Yield
FC =Final Score
FL =Feet and Legs
C =Dairy character.
A selection index will give farmers an option of either selecting for more than one trait or single traits at a time depending on the farmer's selection goals. Future research should also focus on the development of a multitrait selection index appropriate for Zimbabwe.
CONCLUSION
The animal model allows simultaneous genetic evaluation of bulls and cows with all relationships included. For data that follow the assumptions of the model, evaluations computed with an animal model offer the best prediction of future performance. Implementation of an animal model in Zimbabwe will improve the accuracy of evaluation of the Zimbabwe dairy cattle.
REFERENCES
Auran, T. (1973) Acta. Agric. Scand. 23: 189-193.
Congleton, W. R. and Everett, R. W. (1980) J. Dairy Sci. 63: 109-119.
Danell, B. (1981) PhD Dissertation, Swedish University of Agricultural Sciences. Uppsala, Sweden.
Ngwerume, F. N. (1994) PhD Dissertation, Michigan State University, East Lansing, Michigan., USA.
Wilmink, J. B. M. (1987) Livest. Prod. Sci. 47: 85-90.
Wiggans, G. R., Misztal, I. and Van Vleck, L. D. (1988) J. Dairy Sci. 71:1319-1322.
Wiggans, G. R. and VanRaden, T. (1989) National Co-operative Dairy Herd Improvement Program Handbook, USA. A STUDY OF THE PRODUCTIVITY OF PIGS AT THE PIG INDUSTRY BOARD FARM AT ARCTURUS
A STUDY OF THE PRODUCTIVITY OF PIGS AT THE PIG INDUSTRY BOARD
FARM AT ARCTURUS
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
F. Mungate1, K. Dzama2, A. Shoniwa3 and K. Mandisodza3 University of Zimbabwe, Department of Animal Science, Box MP 167, Mt Pleasant, Harare, Zimbabwe. 2University of Zimbabwe, Department of Paraclinical Veterinary Studies, Box MP 167, Mt Pleasant, Harare, Zimbabwe. 3Pig Industry Board, P O Box HG297, Arcturus, Zimbabwe.
SUMMARY
Effects of non-genetic factors on growth and reproductive traits of pigs at the Pig Industry Board Farm at Arcturus were studied using data collected at the farm over a period of 13 years from 1980 to 1992. Litter size, birth weight and preweaning growth were higher for earlier to mid-parity (2-8) sows than for first-parity and later parity (9 and beyond) sows (p < .05). Farrowing mortality was higher in piglets born from parity 7 onwards (p < .05). Productivity in terms of litter size and birth weights was higher (p < .05) in summer than in winter contrary to widely held beliefs. The advantage of crossbred sows over purebred sows for litter size was not demonstrated. It is recommended that farmers keep their sows up to parity eight not six as currently practiced.
INTRODUCTION
The pig industry is becoming more competitive in Zimbabwe and the world over (Holness and Chabeuf, 1991). The success of the industry depends very much on the scale and efficiency of production of weaner pigs for fattening to slaughter weights. The preweaning stage of production is vital, and several factors interact to determine the final productivity at weaning. These factors encompass sow productivity, piglet and litter weights at birth and at weaning, daily piglet and litter weight gains and farrowing mortality.
These components of productivity are influenced by non-genetic factors. Some important nongenetic factors include parity of the sow, month of farrowing and weaning, year of farrowing and general management which includes nutrition, health and others (Maburutse, 1992). The breed of the sow and boar, though a genetic component, can be manipulated by management. Elsewhere numerous studies on the effects of non-genetic factors on pig productivity have been carried out (see Dufour and Fahmy, 1975; Campbell and Dunkin, 1982; Omtvedt et al., 1986 and Yen et al., 1987).
In the only study done in Zimbabwe, Maburutse (1992) examined the use of crossbreeding for weaner production using a mixed model and concluded that breed differences exist for most traits contributing to weaner production and can be exploited to enhance productivity. It is clear that there is need for further research on productivity of exotic pigs under local conditions. The objectives of this study were to characterise productivity of the commercial herd of pigs at the Pig Industry Board (PIB) experimental farm and to specifically examine the effects of non-genetic factors on their growth and reproduction.
MATERIALS AND METHODS
The experimental animals were part of the commercial herd at the PIB experimental farm. The Large White, Landrace and Large White-Landrace, Hampshire-Landrace, Large White-Large White-Landrace and Landrace-Landrace-Large White crosses were used as dam lines and Duroc, Hampshire, Large White and Landrace were used as sire lines. Boars used at the farm were locally bred or bought-in from other pig breeders. The management of the pigs was described by Maburutse (1992).
Data
A total of 5167 records which were collected between 1980 and 1992 were used in this study. Litter size records outside the range 5-15 were pooled, so were weaned litter size outside the range 5-12 and parities beyond 10. Birth weights below 0.5 kg and above 2.5 kg and the number of dead piglets at birth of 4 or more were deleted. Sow breeds with less than 100 records and sires with less than 5 litters were also deleted. In total, 8.7% of the data was lost and 4716 litters were finally available for statistical analysis.
Statistical Analysis
Data was analyzed using the General Linear Model (SAS, 1989). Average birth weight and total birth weight were analyzed using the following fixed effects model:
Yijklmn = µ + Li + Pj + Mk + Dl + Xm + Sn + Eijklmn where Yijklmn = the average or total birth weight; µ = the overall mean; Li = the fixed effect of
th th th the i litter size at birth; Pj = the fixed effect of the j parity; Mk = the fixed effect of the k
th th month of birth; Dl = the fixed effect of the 1 breed of sow; Sn = the fixed effect of the n breed
th of sire; Xm = the fixed effect of the m year of birth and Eijklmn = the random error associated with the ijklmnth record.
Piglet and litter preweaning average daily weight gain (ADG) was analysed using the following covariate model: Yijklm = µ + Pi + Rj + Dk + Sl + Wm + b0(BWT) ijklm + Eijklm
th th where Rj = the fixed effect of the j month of weaning; Wm = the fixed effect of the m number of piglets weaned alive and b0(BWT) ijklm is the birth weight as a covariate in which b0 is the regression coefficient of birth weight on daily gain (piglet birth weight in the case of piglet daily gain and litter birth weight in the case of litter daily weight gain). The other terms are as in the first model. Average weaning weight was analysed using the same model as for average daily piglet weight gain. Litter size was analysed using the following model;
Yijklmn = µ + Pi + Mj + Dk + S1 + Xm + b1(BWT)ijklmn + Eijklmn
th where Xm is the fixed effect of the m year of birth; b1 is the regression coefficient of piglet birth weight on litter size and all other terms are as in the preceding models. The model used to analyse mortality at farrowing was as follows:
Yijklmn = µ+ L i+ Pj + Mk + D1 + Xm + Sn + b2(BWT)ijklmn + Eijklmn where b2 is the regression coefficient of birth weight on mortality at farrowing and all other terms are as in the earlier models.
RESULTS AND DISCUSSION
Table 1 shows a summary of the analysis of variance (ANOVA) results for all the traits studied. Preliminary analyses of variance showed that all two factor interactions were non- significant (p < .05) and were removed from the models.
Table 1. Summary of the F statistics for average birth weight, litter birth weight, average piglet daily weight gain, total litter daily weight gain, average weatung weight, litter size and mortality
Item df Average Total Piglet Litter Average Litter Mortality birth birth ADG ADG weaning size weight weight (kg/d) (kg/d) weight (kg) (kg) (kg) Parity 9 27.7** 25.4** 1.87** 2.2* 12.4** 61.9** 27.2** Litter size 10 63.5** 992.2** - - - - 18.3** Sire breed 3 10.1** 11.3** 3.1* 3.7* 2.4 2.2 0.2 Sow breed 5 3.2** 2.2* 0.5 0.6 8.4** 7.4** 3.3** Month 11 5.3** 6.3** - - - 2.2* 1.6* Year 12 43.3** 49.2** - - - 3.6** 5.2** Average birth weight 1 - - 0.3** - 100.6** 205.6** 79.5** Weaning month 1 - - 0.9 0.9 41.7** - - Weaned litter size 7 - - 1.6 17.1** 28.3** - - Total birth weight 1 - - 20.6** - -- 1.5 14.6 0.2 1.8 7.9 10.1 0.06
*p < 0.05
**p < 0.01 Weight traits
Average piglet birth weight and litter birth weight were significantly affected by parity, litter size at birth, month and year of farrowing, sire breed (p < .01) and sow breed (p < .05). As shown in Table 2 sows in parities two to eight had higher piglet and litter birth weights than those in their first and ninth and older parities (p < .05). Very young sows are still physiologically immature and hence have to partition nutrients between their own growth requirements and those of the foetuses resulting in lower birth weights (Niemann and Heydenrych, 1965). In addition, the uterine capacity tends to limit the birth weights of piglets in young sows. On the other hand old sows tend to undergo a physiological deterioration and hence may not fully utilise their feed resources most efficiently in providing nutrition to the foetuses in-utero. Birth weights of the piglets out of old sows tend to be lower.
Table 2: Least squares means and standard errors for piglet birth and daily weight gain and litter birth and daily weiaht gain by parity
Parity n Piglet birth Litter birth weight Piglet Litter size Mortality weight (kg) (kg) ADG(kg/d) 1 848 1.40 ±0.009 13.65 ±0.09 0.20 ±0.010 8.68 ± 0.098 0.05 ± 0.004 2 796 1.51 ±0.010 14.82 ±0.09 0.21 ± 0.011 9.31 ± 0.100 0.04 ± 0.004 3 658 1.53 ±0.010 15.02 ±0.10 0.20 ± 0.012 10.22 ± 0.109 0.06 ± 0.004 4 612 1.49 ±0.010 14.67±0.10 0.23 ± 0.013 10.81 ± 0.112 0.07 ± 0.005 5 516 1.49 ±0.112 14.69±0.11 0.19 ± 0.013 10.91 ± 0.119 0.09 ± 0.005 6 427 1.48 ±0.012 14.57 ±0.12 0.21 ± 0.015 10.90 ± 0.131 0.08 ± 0.006 7 331 1.47 ±0.013 14.42 ±0.13 0.23 ± 0.017 10.58 ± 0.144 0.11 ± 0.006 8 231 1.43± 0.016 14.10± 0.15 0.20 ± 0.019 10.38 ± 0.169 0.11 ± 0.007 9 165 1.41 ±0.018 13.77 ±0.17 0.18 ± 0.023 9.85 ± 0.195 0.11 ± 0.009 10 132 1.42 ±0.021 13.92 ±0.20 0.18 ± 0.025 10.08 ± 0.22 0.12 ± 0.009
Litter sizes 6 through 12 piglets have significantly higher birthweights compared to those outside this range (p < .05). (see Table 3). Within this group, birth weight declines as litter size increases which is partly due to competition for nutrients. As shown in Table 4, piglets born during the summer months (October to March) are heavier than those born in winter (p < .05). This was unexpected as the winter temperatures at the PIB farm are fairly mild, averaging 160Celsius. However the impact of cold weather on productivity maybe severe if the pigs are not properly housed and provided with heating facilities. These findings are in agreement with the results from the temperate regions (Barbet al., 1991), who postulated that there is an increased rate of mobilization of body reserves by sows in summer due to a decrease in appetite induced by the high temperatures. This results in increased growth during prenatal life by 15% but a decrease in growth rate by 12% afterbirth.
Table 3. Least squares means and standard errors for the effect of litter size on average birth weight, total birth weight and mortality
Litter size n Average birth Total birth Mortality weight (kg) weight (kg) 5 321 1.37 ± 0.400 6.83 ± 2.002 0.096 ± 0.177 6 190 1.64 ± 0.259 9.86 ± 1.552 0.076 ± 0.143 7 293 1.60 ± 0.283 11.18 ± 1.982 0.080 ± 0.133 8 444 1.56 ± 0.261 12.51 ± 2.090 0.065 ± 0.117 9 568 1.50 ± 0.231 13.48 ± 2.076 0.063 ± 0.100 10 745 1.48 ± 0.226 14.77 ± 2.264 0.063 ± 0.097 11 698 1.45 ± 0.225 15.96 ±2.473 0.061 ± 0.092 12 615 1.43 ± 0.206 17.16 ± 2.470 0.057 ± 0.083 13 374 1.37 ± 0.209 17.85 ± 2.711 0.047 ± 0.072 14 259 1.34 ± 0.210 18.79 ± 2.947 0.042 ± 0.062 15 209 1.34 ± 0.202 20.12 ± 3.026 0.035 ± 0.061
Table 4. Least squares means and standard errors for piglet and litter birth weight and litter size and mortality by month of birth
Month n Piglet birth Litter birth Litter size Mortalitv of birth weight (kg) weight (kg) 1 375 1.51 ± 0.013 14.77 ± 0.124 10.56 ± 0.139 0.08 ± 0.006 2 327 1.47 ± 0.014 14.47 ± 0.132 10.21 ± 0.147 0.08 ± 0.006 3 380 1.45 ± 0.013 14.29 ± 0.122 10.20 ± 0.137 0.08 ± 0.006 4 369 1.43 ± 0.013 14.07 ± 0.123 9.95 ± 0.138 0.08 ± 0.006 5 462 1.42 ± 0.012 13.94 ± 0.118 10.07 ± 0.132 0.08 ± 0.006 6 403 1.44 ± 0.013 14.14 ± 0.123 9.87 ± 0.132 0.09 ± 0.006 7 384 1.44 ± 0.013 14.10 ± 0.124 10.02 ± 0.139 0.08 ± 0.006 8 412 1.46 ± 0.013 14.28 ± 0.120 10.30 ± 0.134 0.10 ± 0.006 9 399 1.46 ± 0.013 14.33 ± 0.119 10.29 ± 0.133 0.08 ± 0.006 10 423 1.48 ± 0.012 14.59 ± 0.116 10.15 ± 0.130 0.08 ± 0.006 11 376 1.49 ± 0.013 14.70 + 0.122 10.15 ±0.137 0.09 ± 0.006 12 406 1.49 ± 0.013 14.70 ± 0.120 10.30 ± 0.134 0.08 ± 0.006
Month of birth1=January
Piglets born heavier also had significantly higher weaning weights and higher preweaning ADG. Piglets of higher birth weight consume more milk per suckle than their lighter ones and this seems to be the primary reason why heavier piglets outgain lighter ones (Campbell and Dunkin, 1982). Big piglets are better since they have larger glycogen reserves and are likely to survive less from hypothermia or hyperthermia. Piglets born out of sows in parities 1-8 outgained those born in later parities (p < .05) owing to a combination of their heavier birth weights and higher milk production from their dams Even though breed of sow was a significant source of variation, piglets out of crossbred sows did not consistently outweigh those out of purebred sows at birth. (see Table 5).
Table 5. Least squares means and standard errors for the effect of sow breed on piglet and litter birth weight, litter size and mortality
Sow breed n Piglet birth Litter birth Litter size Mortality weight (kg) weight (kg) HR 312 1.49 ± 0.015 14.63 ± 0.140 10.11 ± 0.156 0.096 ± 0.007 L 171 1.43 ± 0.018 14.09 ± 0.174 10.45 ± 0.195 0.095 ±0.008 L,LR 393 1.44 ± 0.012 14.17 ± 0.115 10.12 ± 0.129 0.079 ± 0.006 Lr 2415 1.46 ± 0.006 14.38 ± 0.056 10.37 ± 0.060 0.075 ± 0.003 R 1231 1.47 ± 0.008 14.44 ± 0.074 9.88 ± 0.081 0.074 ± 0.004 RRL 194 1.48 ± 0.016 14.48 ± 0.157 10.09 ± 0.175 0.074 ± 0.008 R=Landrace; H=Hampshire; L = Large White
Litter size at birth
Parity, birth weight, sow breed, year of birth (p < .001) and month of birth (p < .05) were significant sources of variation for litter size at farrowing. Litters born to sows in their early (1 - 2) and late (9 and beyond) parities were smaller in size compared to those born, to sows in their mid-parities (3 - 8). This is probably because gilts and young sows produce fewer fertile ova compared to mature sows (MacPherson et al., 1977). Older sows tend to have a higher incidence of farrowing problems like dystocia, which lead to higher piglet mortalities. This has been attributed to the reduced muscle tone which sets in as the sows grow older, which results in farrowing problems (English et al., 1977).
Litters farrowed in summer (October - March) were significantly larger than those farrowed in winter (p < .05). The highest litter sizes were in those piglets farrowed in January and the lowest in June. (see Table 4). Low litter sizes in winter can be partly explained by the high mortality rates during that season. The winter months of June and September also had significantly higher mortality rates when compared to the rest of the year (p < .05). However, in temperate climates, the reverse has been found to be true. Brooks (1991) reported that under hot summer conditions, pregnant or lactating sows eat less, lose more weight and this tends to lower birth. weights and litter sizes in summer. Prior to this study there was a general belief that the same situation prevails in Zimbabwe.
As with the weight traits, purebred sows eg Landrace outperformed some crossbred sows eg Landrace x Large White (p < .05). This was not expected since hybrid sows tend to produce large litters owing to hybrid vigour. The superiority of the crossbred sow over its purebred counterparts has been well documented (Bags et al., 1992; Kuhlers et al., 1989).
Mortality
Mortality was affected by parity, litter size, month and year of birth, sow breed and average birth weight of the piglets (p < .01) and month of farrowing (p < .05). Sows in their later parities (7 and beyond), had higher mortalities because older sows are more prone to reduced muscle tone and hence farrowing problems. Lower litter sizes resulted in higher mortalities because they are usually associated with higher individual birth weights, which might lead to dystocia. Average birth weight affects mortality in that grossly underweight piglets fail to make it and hence die at birth and obese ones tend to be farrowed with difficulty and are likely to die at birth.
CONCLUSION
Contrary to widely held beliefs, productivity of pigs in Zimbabwe is higher in summer than in winter. In addition, producers in Zimbabwe tend to cull their sows at parity six but results of this study clearly show that productivity of sows is still reasonable up to the eighth parity. The advantage of using crossbred sows over purebred sows was not clearly demonstrated.
REFERENCES
Baas, T.J, Christian, L. L. and Rothschild, M. F. (1992) J. Anim. Sci. 70 : 89-98.
Barb, C. R., Estienne, M. J., Kraeling, R. R., Marple, D. N., Rampaceck, G. B., Rahe, C. H. and Sartin, J. L. (1991) Domest. Amm. Endocrinology 8: 117-121. Brooks, P. (1991) Pig Farming. 39 : 48.
Campbell, R.G. and Dunkin, A.C. (1982) Amm. Prod. 35 : 193-197.
Dufour, J. J. and Fammy, M. H. (1975) Canadian J. Anim.Sci. 55 : 9-16.
English, P. R. and Smith, W. J. (1975) The Vet. Ann. 15 : 95. Bristol , Switzerland. Fisher, R. A. (1949) The Design of experiments. Edinburgh: Oliver and Boyd Ltd.
Holness, D.H. and Chabeu£ N. (1991) Pigs. MacMillan. CTA.
Kuhlers, D. L., Jungst, S. B. and Little, J. A. (1989) J. Anim. Sci. 67: 920-927. Li, X. and Kennedy, BV. (1994) J. Anim. Sci. 72 : 450.
Maburutse, Z.A. (1992) MSc Thesis. University of Zimbabwe.
MacPherson, R. M., Hovell, F. D. and Jones, A. S. (1977) Anim. Prod. 24 : 333-343.
Niemann, P. J. and H. J. Heydenrych. (1965) Technical Communication No.41. Dept. Agric. Tech. Serices. R.S.A.
Omtvedt,I. T., Whatley (Jr.) J. A., and Willham, R. L. (1966) J. Anim. Sci. 25:372-376. SAS. (1989) SAS User's Guide: Statistics. SAS Institute., Inc., Cary, NC., USA.
Yen, H. F., Islen, G. A., Harvey, W. R. and Irvin, N. K. M. (1987) J. Anim. Sci. 64 : 1134- 1348. CROSSBREEDING FOR WEANER PRODUCTION IN PIGS
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSION
REFERENCES
Z. Maburutse and H.P.R. Tawonezvi World Vision International, P O Box 2420, Harare, Zimbabwe and Stocrop Pvt. Ltd., P O Box 694, Marondera, Zimbabwe.
SUMMARY
Analysis were carried out using 1 933 sow and 7 478 piglet records to determine breed and heterosis effects on sow fertility, piglet growth and survival and overall wearer production. The data was collected between 1982 and 1988 from purebred Landrace and its crosses with Large White and Hampshire. A total of 21 mating types or piglet genotypes were assessed. Individual sow effects were significant (p < .01) for all traits and sire effects were important (p < .01) for preweaning growth, indicating that there is potential for selection within breeds. Crossbred sows and piglets generally performed better than purebreds. It was concluded that breed differences exist for most traits contributing to weaner production and can be exploited to enhance productivity. Hampshire and Large White breeds appeared more appropriate as terminal sires while the Landrace was more appropriate as a dam breed in order to exploit both heterosis and breed complementarity.
INTRODUCTION
Breed differences are important in the genetic improvement of livestock.. The profitability of commercial pig production depends primarily on efficiency of weaner production. Thus characterization of breeds for major traits contributing to weaner production is essential for making decisions in both purebreeding and crossbreeding programmes. Such characterization will allow maximum exploitation of heterosis and additive genetic differences between breeds.
Several studies have shown large differences between pig breeds and potential for exploitation of heterosis (Jungst and Kuhlers, 1984; Yen, et al., 1987). Commercial pig production in Zimbabwe is based on the use of exotic breeds in which Large White and Landrace are predominant (Bellis, 1978). Currently farmers are advised to use Hampshire x Landrace or Duroc x Landrace as dam lines and Large White or Landrace as sire lines in order to exploit breed differences in both additive and heterotic effects. However, information on the comparative perfonnance of these crosses is limited. Therefore a study was conducted at the Pig Industry Board (PIB) to deternvne the productivity of purebred and crossbred pigs. This paper presents results from breed comparisons for preweaning traits.
MATERIALS AND METHODS
The experimental animals were part of an on-going crossbreeding programme at the PIB. The Landrace breed was used as the dam line with Landrace, Large White and Hampshire as sire breeds. Selection of boars and replacement gilts was based on pedigree records for growth rate, backfat thickness, fertility traits, structural soundness and performance test results. The management of the animals was described by Maburutse (1992).
Number of piglets born alive, number of stillbirths and litter weight at farrowing were recorded for each individual sow. Farrowing interval was derived as the differences in days between two consecutive farrowmgs from each sow. Farrowing index was calculated as:
Farrowing index = 365/farrowing intervals in days.
Live weight at birth, 21 days of age and weaning (35 days) were recorded for each piglet and those piglets which died before weaning were also recorded. In addition, average daily gain from birth to 21 days and to weaning were derived for each piglet.
A total of 1 933 sow and 7 478 piglet records from 21 genotypes were analysed using Mixed Model Least Squares and Maximum Likelihood methodology (Harvey, 1987) in which the effects of mating type (or piglet genotype) and individual dams and sires within genotypes were fitted as appropriate for each trait. All important environmental effects and interactions were also fitted where appropriate. Estimates of breed additive, maternal and heterosis effects were estimated using regression procedures (Maburutse, 1992). Because of the practice of fostering piglets in this herd, genotype means for sow traits after farrowing were derived indirectly from the least-squares means of component traits. The derived sow traits were number of piglets weaned per litter, number of piglets weaned per sow per year and weight of weaned piglets per sow per year.
RESULTS
Mating type was significant (p < .01) for litter weight at farrowing, while genotype of piglet was significant (p < .001) for birth weight. Individual sire effects were significant (p < .001) for birth weight, weaning weight and average daily gain to weaning while sow effects were significant (p < .01) for all the traits analysed.
Piglets born from backcross sows tended to survive better than piglets born from pure Landrace and F1 sows. Birth weight ranged from 1.12 to 1.77kg (p < .001). Piglets born from F1 sows were heavier at birth than the piglets born from purebred Landrace and backcross sows. Piglets born from sows mated to Hampshire boars were heavier at birth, at three weeks of age and at weaning than those mated to Large White and Landrace boars. The ranking of sire breeds on the average daily gain to weaning of progeny was Hampshire > Large White > Landrace.
Breed additive genetic and breed maternal genetic effects for sow traits were computed as deviations from purebred Landrace. Hampshire additive effects for litter weight at farrowing were significantly superior (p < .01) to Landrace, while Large White additive effects were slightly inferior to Landrace. In terms of additive effects for weight of weaned piglets per sow per year, Hampshire was significantly superior (p < .05) to Landrace while the difference between Landrace and Large White was not significant. Breed maternal effects for the three breeds were not significantly different for all traits. Individual heterosis for Large White x Harnpslure was significant (p < .05) and negative for number of piglets born alive. Individual heterosis for litter weight at farrowing was significant (p < .05) and positive for Large White x Hampshire and significant (p < .05) and negative for both Large White x Hampshire and Hampshire x Landrace. For farrowing interval, individual heterosis estimates were favourable for all crosses but was only significant (p < .05) for Large White x Landrace.
Breed additive genetic and breed maternal genetic effects for piglet traits are also computed as deviations from the purebred Landrace. For additive effects, performance of Large White and
Hampshire was significantly (p < .05) better than that of Landrace for all the piglets traits except birth weight of Large White which was not significantly better than that of pure Landrace. Interms of maternal effects Large White and Hampshire dams produced significantly (p < .05) heavier piglets at birth than Landrace darns Piglets of Landrace dams had significantly (p < 05) superior growth rates than piglets of Large White and Hampshire dams. This resulted in heavier weaning weights of piglets from Landrace dams compared with those from Large White and Hampshire dams.
Individual heterosis for Large White x Landrace was significant (p < .05) and negative for birth weight and positive (p < .001) for all the other growth traits. Individual heterosis for Large White x Hampshire was significant (p < .05) and positive for birth weight and three-week weight. Although all other estimates of individual heterosis were not significant, they were positive. Maternal heterosis was not significant (p > .05) but was favourable for most traits.
DISCUSSION
The analysis of variance indicated significant differences between sires and between dams for most of the traits. This suggests that within genotype, there are significant differences which can be exploited by selection. Therefore estimation of genetic parameters will be essential in order to determine appropriate selection criteria and expected rates of genetic improvement. Estimates of genetic parameters (heritability, genetic correlation and repeatability) are not available under Zimbabwean conditions.
Crossbred piglets and crossbred sows generally performed better than purebreds. This could be due to heterosis and possibly superior genes from the other breeds crossed with the Landrace. Similar results have been reported elsewhere (Bass et al., 1992; Kuhlers et al., 1989; Jungst and Kuhlers, 1984). The breed combinations in this study were not experimentally designed. Consequently it was not possible to estimate breed parameters by the usual methods of estimation from genotype means. For such unstructured data, the regression approach has been recommended (Dillard et al., 1980) and has gained popularity in recent years (Olson et al., 1990). The regression approach also has other advantages. Firstly, it provides a clearer way of separating component parts of performance. Secondly, it utilises information from all breed groups in estimating component parts of performance. Secondly, it utilises information from all breed groups in estimating genetic effects. Thirdly, it can be used to predict performance of various breed combinations of interest which were not included in the original experiment (Dillard et al., 1980).
In this study both Hampshire and Large White breeds were superior to Landrace for breed additive effects, while the opposite was true for breed matemal effects on piglet growth. These results suggest that both Large White and Hampshire are suitable as sire breeds while Landrace can be used as a dam breed in terminal crossbreeding systems.
Estimates of individual heterosis indicated that Large White and Landrace breeds combine well in a cross especially for preweaning growth, litter weight and farrowing interval. These results are similar to those reported by Fahmy and Holtman (1977). There was no significant improvement due to maternal heterosis, probably because of the limited number of observations for this parameter. Other workers have reported significant maternal heterosis for preweaning traits in pigs (Johnson, 1978 and 1981).
CONCLUSION
From our study, we conclude that breed differences for most traits contributing to weaner production and can be exploited to enhance productivity. The observed variation within breeds indicated the need to select within breeds in addition to crossbreeding. In planning crossbreeding systems, the direction of the cross is important in order to obtain the benefits from breed complementarity and heterosis. Hampshire and Large White breeds appear more appropriate as terminal lines for preweaning growth while the Landrace breed can be used as a dam line in terminal crossbreeding systems. Large White and Landrace appear to be the most appropriate breeds where the main objective of crossbreeding is to exploit individual heterosis. However, terminal crossbreeding is often difficult to implement because of its managerial complexity (Sheridan, 1981) and other simpler forms of crossbreeding should be considered.
REFERENCES
Bass, T.J., Christian, L.L. and Rothschild, M.F. (1992) J. Anim.Sci. 70: 89-98. Bells, D.B. (1978) Rhodesia Agric. J. Technical Handbook, No. 1.
Dillard, E.U., Rodriguez, 0. and Robinson, OW. (1980) J. Anim. Sci. 50: 653-663. Fahmy, M.H. and Holtman, W.B. (1977) Anim. Prod. 24: 261-268.
Harvey, W. R. (1987) MMLSML PC Vesion, USDA, USA.
Johnson, RK., Omtvedt, I.T. and Walters, L.E. (1978) J. Anim. Sci. 46: 69-81. Johnson, R.K. (1981) J. Anim. Sci. 52: 906-921.
Jungst, S.B. and Kuhlers, D.L. (1984) J. Anim.Sci. 59: 1140-1148.
Kuhlers, D.L. Jungst, S.B. and Little, J.A. (1989) J. Anim. Sci. 67: 920-927. Maburutse, Z. (1992) MSc Thesis. University of Zimbabwe.
Olson, T.A., Elzo, M.A., Koger, M., Butts, (Jr) WT. and Adams, E.L. (1990) J. Anim. Sci. 68: 317-323.
Sellier, P. (1976) Livestock Prod. Sci. 3: 203-226.
Sheridan, A.K. (1981) Anim. Breed. Abstr. 49:131-144.
Yen, H.F., Isler, G.A., Harvey, W.R., and Irvin, K.M. (1987) J. Anim. Sci. 64: 1340-1348. EVALUATION OF THE INDIGENOUS PIG IN ZIMBABWE
SUMMARY
INTRODUCTION
POPULATION
PRODUCTION SYSTEM
EVALUATION OF THE INDIGENOUS PIG
DISCUSSION
CONCLUSION
REFERENCES
P. N. B. Ndiwem and K. Dzama University of Zimbabwe, Department of Animal Science, Box MP 167, Mt Pleasant, Harare. Zimbabwe and University of Zimbabwe, Department of Paraclinical Veterinary Studies, Box MP 167, Mt Pleasant, Harare, Zimbabwe.
SUMMARY
Research to evaluate the Zimbabwean indigenous pig has been sporadic and inadequate. Consequently, the Zimbabwean indigenous pig has not been sufficiently characterised. Different authors have classified the indigenous pigs of Zimbabwe and the neighbouring countries differently. Genetic fingerprinting is likely to give more definitive answers about the origin of the indigenous pig and its relationship to the wild, Asian and European pigs. Improvements in the genetic constitution of indigenous pigs must be accompanied by improvements in management nutrition and health. Comparisons of indigenous and exotic pigs at the same metabolic age may be needed to get fair comparisons between these breeds.
INTRODUCTION
Southern Africa has a fairly large population of indigenous pigs. Bonsma and Joubert (1952) suggested that the so-called indigenous pig of Southern Africa was introduced by European and Chinese traders 300-400 years ago. Holness and Smith (1973) regarded the bush pig (Potamachoerus porcus) and warthog (Phacochorus aethiopicus) to be probably the only truly indigenous members of the family Suidae in Zimbabwe.
Mason and Maule (1960) classified the Southern African indigenous pig into the kolbroek ("Breech markings"), short, fat, short snout and resembling the Chinese lard pig and the Windsnyer ("Wind-Cutter"), long-nosed, razor-back. The later predominates in Zimbabwe and parts of Mozambique and Zambia (Holness, 1973, 1991). The indigenous pigs used by Holness, originating from the Mukota area of northeastern Zimbabwe. Sometimes indigenous pigs in general, have been called Mukota pigs or Zimbabwe Mukota pigs. The justification of such usage, however, must await a thorough characterisation of the Zimbabwean indigenous pigs. Genetic fingerprinting is likely to give more definitive answers about the origin of the indigenous pig and its relationship to the wild, Asian and European pigs.
POPULATION
The present pig production of 285 000 in Zimbabwe consists of 110 000 pigs on commercial farms and 175 000, predominantly indigenous pigs, in communal areas (FAO, 1993). The large population of pigs in the communal areas shows the potential contribution of these pigs to pig meat production. Africa has a small percentage (1.5%) of the world pig population of 826 million compared with 49.0, 22.9, 10.5 and 6% in Asia, Europe, North and Central America and South America, respectively (FAO, 1989).
PRODUCTION SYSTEM
The indigenous pigs, kept mostly in the communal areas of Zimbabwe, are kept free-ranging during the dry season and in simple pig houses or fold yards during the rainy season (Scherf, 1990). Their survival under unhygienic conditions would testify to their disease resistance. Mating is uncontrolled during the free ranging period (Scherf, 1990). Reproductive processes follow an annual rhythm with the peak season of birth occurring during October (Scherf, 1990). Age at first farrowing is 1-2 years and the farrowing interval is generally 1 year (Scherf, 1990). The average litter size is 7.3 to 7.9 (Scherf, 1990; Holness, 1991). Natural weaning is practised.
The pigs scavenge for food during the free-ranging period and during confinement are fed feeds such as maize, coarse maize meal, maize husks, green maize, kitchen waste, cabbage waste, pumpkins, groundnut shells, fruits, grasses and forbs (weeds from the field) and brewer's wastes ("Masese" or "Butu") (Scherf, 1990). Animals receive little water (61itres, about once per day) and no hygiene measures are generally employed (Scherf, 1990). It is clearly evident that the potential for the improvement of the management and productivity of the indigenous Zimbabwean pig is vast.
EVALUATION OF THE INDIGENOUS PIG
Research to evaluate the Zimbabwean indigenous pig has been sporadic and inadequate. Consequently, the Zimbabwean indigenous pig has not been sufficiently characterised.
Reproductive Performance
Holness and Smith, (1973, 1974) fed indigenous gilts and sows at high, medium and low planes of nutrition, defined as 100, 75 and 50% of the ARC (1967) requirements adjusted for differences in metabolic weight (Kleiber, 1961). The number of corpora lutea (indicative of ovulation rate) increased significantly (p < .01) with plane of nutrition in sows (Holness and Smith, 1973). The increase in the number of fertile embryos with increasing plane of nutrition was significant for sows (p < .01) but not for gilts (p > .05). Dietary treatment did not have a significant effect on the number of follicles in either gilts or sows (Holness and Smith, 1973).
Although the absolute weights of the anterior pituitary, thyroid, adrenal glands and left and right ovaries decreased with decreasing plane of nutrition, only the pituitary gland exhibited a significant decrease (p < .01) as a fraction of sow body mass (Holness and Smith, 1973). The effect of plane of nutrition on reproductive performance appeared to be mediated through the anterior pituitary endocrine activity.
Growth and carcass composition The livemass of the indigenous gilts fed a low plane diet was 32% and 46% of that of high plane gilts at first service and fifth pregnancy (Holness and Smith, 1974). The mean birth mass of piglets differed significantly (p < .01) between the high and low plane treatments at all parities except the third but not between the high plane and medium plane treatments (p > .05) (Holness and Smith, 1974). Total litter mass was a constant percentage of sow livemass at farrowing, irrespective of nutritional level or parity (Holness and Smith, 1974). The number of stillbirths per litter decreased while the number of foetuses produced, piglets born alive and piglets weaned increased with increasing plane of nutrition (Holness and Smith, 1974). The growth rate of piglets increased with parity but total litter mass weaned increased with parity only in the high plane treatment (Holness and Smith, 1974). The increase in piglet weaning mass with increasing piglet birth mass was significant only in the high plane treatment (Holness and Smith, 1974). In related experiments, Chigaru et al. (1981) studied feed intake, feed conversion efficiency, dressing percentage and carcass composition of indigenous and Large White pigs fed high protein and low protein diets from 8 to 32 weeks and slaughtered at different ages. Feed intake was 35% lower in indigenous than Large White pigs. The decrease in feed conversion efficiency with age was lower in Large White than indigenous pigs. In addition Large White pigs had a lower fat: protein ratio.
Breeding
Van Ercket (1993) studied the growth rate of pigs with 25%, 50% and 75% Large White blood given to communal farmers in Mangwende, Kandeya and Chivi and raised on either the traditional system or intensive system.There was no advantage to improving the genotype of Mukota pigs unless the management was simultaneously improved. Scherf,(1990) compared Mukota pigs and their crosses with Large White, Landrace or Duroc and came to a similar conclusion.
Management
The only work on management of indigenous pigs was done by Scherf (1990). Daily gain and weight gain increased with level of feeding. Young pigs were more sensitive to under-feeding than older pigs. Anthelminthic treatment had negative effects on daily gain and weight gain in young pigs and improved only weight gain for pigs 90 - 120 days. It was concluded that anthelminthic treatment was futile unless housing was also improved.
DISCUSSION
The litter size of Zimbabwean indigenous pigs (7.9) compares favourably with those reported for indigenous pigs in South Africa (7.2), Nigeria (6.5) and Ghana (6.3) (Holness, 1991). Comparisons of the highly prolific Meishan pig with European breeds indicate that high prolificacy may be achieved by increasing embryonic survival (Young et al., 1994) through more uniform embryo development (Bazer et al., 1988) and slow embryonic growth rate (Conley et al., 1992). Slow embryo development and large litter size have been associated with the d/d haplotype of the MHC locus but not the a/a or c/c haplotype. Recent attention has been focused on the leukaemia inhibitory factor (LIF) which is sensitive to oestrogen produced by the peri-elongation period blastocyst (Conley et al., 1992). Ovulation rate increases with parity in Meishan pigs but plateaus in European pigs (Christenson, 1993). Conflicting results have been obtained depending on whether ovulation rates were compared at similar ages or reproductive stages. Heat stress particularly long duration (Britt et al., 1992) has been implicated in reduction of embryonic survival.
The weakness of the work done to date in Zimbabwe is that comparisons were not made in reproductive traits with exotic breeds under similar conditions. Low heritabilities of maternal and direct genetic effects on litter size and their negative correlation have resulted in slow progress of genetic selection for litter size in exotic breeds (Roeche and Kennedy, 1993a,b; Irgang et al., 1994; Perez-Enciso et al., 1994). Litter size may improve due to increase in ovulation rate when the nutrition and management of indigenous pigs are improved but thereafter progress may be slow. The use of crosses with indigenous prolific breeds will depend on the market demand. Colcom, but not communal area markets, have discriminated against indigenous pig carcasses in the past.
Improving the nutrition and management of the indigenous pig may result in improvements in number of foetuses produced at parturition, number born alive, number of piglets weaned, mean piglet birth mass but not total litter mass as a percentage of sow body mass at farrowing (Holness and Smith, 1974).
The work of Chigaru et al. (1981) indicates the problems of making comparisons between rapidly maturing indigenous pigs and slowly maturing exotic pigs at the same chronological age. At the same chronological age these pigs will be at different stages of physiological growth. The magnitude of the differences is smaller when indigenous and exotic pigs are compared at similar stages of physiological development than at the same weight or age.
The interest in evaluating protein level arises because of the higher milk production (> 10kg/day), litter weight gains (> 2kg/day) of pigs than in the past (Hansen and Lewis, 1993; Johnston et al., 1993). The high protein diet used by Chigaru et al. (1981) was similar to that which produced maximum weight gains and feed conversion efficiencies in the studies of Hansen and Lewis (1993).
Van Eckert (1993) indicated the benefit of genetic improvement of indigenous pigs under an intensive but not a traditional system. The present genotype of indigenous pigs appears adequately suited to the production system of the communal areas.
Improvement in the feeding level of indigenous pigs improves weight gains (Scherf, 1990). Anthelminthic treatment has negative effects on young pigs and is of limited value in older pigs unless housing is improved (Scherf, 1990). The removal of helminths appeared to open the way for coccidia suggesting that a dosing program must include all major parasites.
CONCLUSION
Improvements in the genetic constitution of indigenous pigs must be accompanied by improvements in management. Comparisons of indigenous and exotic pigs at the same metabolic age may be needed to get fair comparisons between these breeds. A comprehensive study currently underway at the Unversity of Zimbabwe seeks to evaluate the digestive capacity of indigenous pigs to determine if they utilise fibrous diets better than exotic breeds and compare growth and carcass characteristics of indigenous, crossbred and exotic breeds (0, 25, 50, 75 and 100% exotic blood) at the same metabolic age.
REFERENCES
Agricultural Research Council. (1967) Pigs. No. 3. HMSO. London.
Bazer, F.W., W.W. Thatcher, F. Martinat-Botte and M. Terqui. (1988) J. Reprod. Fertil. 83: 723-728.
Bonsma, F.N. and D.M. Joubert. (1952) Fmg. S. Afr. 27: 167-170.
Britt, J.H, S.L. Whaley and V.S. Hedgepeth. (1992) J. Anim. Sci.70 (Suppl. 1): 271 (Abstr).
Chigaru, P.R.N., L. Maundura and D.H. Holness. (1981) Zimbabwe J. Agric. Res. 19: 31. Christenson, R.K. (1993) J. Amm. Sci. 71: 3060. Conley, A.J., R.K.
Christenson, S.P. Ford, R.D.
Geisert and J.J. Mason. (1992) Endocrinology 131: 896.
FAO. (1989) Production Yearbook,
FAO, Rome. FAO. (1993) Production Yearbook, FAO, Rome. Hansen, B.C. and A.J. Lewis. (1993) J. Anim. Sci. 71: 2110-2121.
Holness, D.H. (1973) Rhod. Agric. J. 73: 59-63.
Holness, D.H. (1991) Pigs. CTA. Macmillan.
Holness, D.H. and A.J. Smith (1973) Rhod. J. Agric. Res. 11: 103-112.
Holness, D.H. and A.J. Smith. (1973) Rhod. J. Agric. Res. 12: 19-25.
Holness, D.H. and A.J. Smith. (1974) Rhod. J. Agric. Res. 12: 27-43.
Igrang, R., J.A. Favero and B.W. Kennedy. (1994) J. Anim. Sci. 72: 2237-2246.
Johnston, L.J., J.E. Pettigrew and J.W. Rust. (1993) J. Anim. Sci. 71: 2151-2156.
Kleiber, M. (1961) John Wiley and Sons, NY, USA.
Liao, C.W. and T.L. Veum. (1994) J. Anim. Sci. 72:2369-2377.
Mason, l.L. and J.P. Maule. (1960) Tech. Comm. 14. Commonwealth Bureau of Animal Breeding and Genetics, Slough, England.
Perec-Enciso, M., J.L. Foulley, L. Bodin and J.P. Poivey. (1994) J. Amm. Sci. 72: 2775.
Rocche, R. and B.W. Kennedy. (1993a) J. Anim. Sci. 71: 2891-2904.
Rocche, R. and B.W. Kennedy. (1993b) J. Anim. Sci. 71: 3251-3260.
Scherf. B.D. (1990) Research report. Department of Research and Specialist Services, Ministry of Agriculture, Zimbabwe.
Van Eckert, M. (1993) PhD Thesis. Technische Universitat Berlin, Germany.
Youngs, C.R., L.K. Christenson and S.P. Ford. (1994). J. Anim. Sci. 72: 725-731. GOAT BREEDING RESEARCH AND DEVELOPMENT ACTIVITIES IN ZIMBABWE
SUMMARY
INTRODUCTION
GENETIC RESOURCE
PROGRESS IN RESEARCH
DEVELOPING BREEDING PLANS
CONCLUSION
REFERENCES
J. L. N. Sikosana Matopos Research Station, P Bag K 5137 Bulawayo, Zimbabwe
SUMMARY
Nearly all goats in Zimbabwe are kept in the communal areas. Two indigenous types have been identified. There is need for proper description and evaluation of the existing genetic resource and identifying breeds or types to be used to improve the indigenous goat. Information on goat breeding is limited but improvement is needed for optimum production of meat, milk, fibre and skins. This paper summarises the current state of goat breeding and also looks to the future for breeding strategies and research priorities.
INTRODUCTION
There are more than 2 million goats in Zimbabwe. Nearly all goats (about 98 per cent) are kept in the communal areas. The vast majority of them are indigenous (Sibanda and Sikosana, 1986). Within the agro-ecological zones, there exists variable populations of a wide variety or type of goat which has become adapted to the environemnt. Goats are used for income generation or improvements in living standards by the production of meat, milk and skins. The intention of this paper is to focus on the current status and contribution of goats and how these resources are used. Issues that are worthy of increased research and development, both to increase current output and allow further development of this species in livestock production are emphasised.
GENETIC RESOURCE
There are 107 described breeds of goats in the world (Acharya,1992). These breeds provide meat, milk, fibre and skins (pelts), meat being the most important component in the economics of goat production. In Africa the majority of goats are "indigenous" types (Wilson, 1991) and a large number are non-descript populations. Generally these types of goats are usually common in and adapted, to a specific environment. It would be incorrect to classify the two types of indigenous goats found in Zimbabwe as breeds. Selection over long periods of time has ensured adaptation to local ecological conditions. There is a need to describe and evaluate the genetic resource in terms of population size, flock size and structure, and production characteristics.
Mason and Maule (1960) have divided goats into two main groups, the long-cared and short eared. Devendra and Burns (1983) have gone further, classifying them as large, small and dwarf goats. Most of these types are widely distributed in Southern Africa. In Zimbabwe both the major types of goats are found at research institutions. Within these types there are differences which need investigating. Relationships between types need classifying through blood typing. In eastern and central areas of Zimbabwe goats (Small East African) are small, compact, hardy and have short and horizontal ears. In the south they are more heterogeneous and larger. Some have long ears, some pendulous with turned up tips, and many with an intermediate type of cars, varied hairiness and a wide range in coat colour. The smaller type of goat is normally termed the Mashona goat and the larger in the south the Matabele goat. Crossbreeding for specific characteristics should be attempted to improve productivity.
PROGRESS IN RESEARCH
The general hypothesis for research can be stated as "Small ruminant productivity (milk, fibre, meat, skins) can be increased through improved fertility, higher plane of nutrition, provision of appropriate breeds, and reduction of reproductive wastage through good management". During the past decade research has concentrated on production systems. These production studies were in combination with breeding systems but not identifying or evaluating a breed or type. In all instances performance recording was done at institutional level with few on-farm studies. On-station recording is convenient but flocks are small and management resources more readily available than on-farm.
Productivity of Matebele goats under an accelerated kidding management system (Sibanda, 1990) has been evaluated. Results showed that compared to annual breeding, accelerated frequency of breeding reduced reproductive performance of does within a semi-extensive environment (Table 1.). No similar study with the smaller type of goat (Small East African Dwarf goat) has been reported. Another study of Matebele goats in a communal management system worth noting was reported by Sibanda (1992). In this study litter size was about 1.2 lower than that of goats raised on-station. Differences are attributed to environmental factors and farmer interventions. Baffour-Awuah (1987) reported that the level of performance of large indigenous goats under a semi-extensive system (on-station) of production was comparable to those of other tropical breeds kept for meat production (Table 2).
Table 1. Reproductive performance of does kidding either three times in 2 years (accelerated kidding,) or annually (Sibanda, 1990)
Accelerated kidding Annual kidding Suppl. Non suppl. Non suppl. Fertilitya 0.6±0.03 0.5±0.32 0.9±0.05 No. parturition/doe 1.8±0.08 1.6±0.10 1.7±0.10
exposed Reproductive rateb 0.9±0.05 0.8±0.05 1.5±0.11 Kidding 330±16.8 374±18.3 364±15.0
interval(days)
Suppl. - supplemented animals Non suppl. = non- supplemented animals afertility = Number of does kidding/ number of does exposed. bReproductive rate = Number of kids born/number of does exposed.
Table 2. Overall least-square means and standard errors for growth characters of indigenous goats(Baffour-Awuah, 1987)
Body weight (kg) at: n birth 837 2.3±0.04 12 weeks 662 11.4±0.42 weaning (20 weeks) 659 16.0±0.44 18 months 596 30.4±0.63
ADG1 (from birth to 12 weeks) 662 109.0±4.8
ADG2 (from birth to weaning) 659 98.0±3.0
ADG =average daily gain (g/d).
There exists, in indigenous goats, considerable genetic variation in growth traits and other economically important characteristics for production of meat and thus progress through selective breeding should be possible. With the smaller goat there area few performance studies reported. Performance was measured from data collected from a survey under a communal management system (Ndlovu, 1988).
Prolificacy was found to be 1.30 (with a twinning rate of about 30%) lower than that of the large type of goat. Poor growth rates might be an indication of inbreeding as there was little influx of males from other areas into any of the areas surveyed. In this study survival rates of kids was low.
Environmental influences have also been studied in crossbreds (Khombe, 1985). High preweaning mortalities due to type of birth and low growth rates were observed in goats fed ad libitum. Identification of genotypes with higher growth rates is encouraged in crossbreeding studies. In nutrition studies opportunities to increase productivity of the indigenous goat in Zimbabwe, through manipulation of energy and protein nutrition under intensive feeding conditions are described (Hatendi, 1992). From this work it was concluded that selection and breeding programmes of indigenous goats for clearly identified production attributes are essential if production is to be intensified.
Characterisation of goat keeping systems in the lowveld is in progress (Chifamba and Sithole, 1992). Flock dynamics are being assessed. Such studies need to be extended to other regions and components of genetic relationships and characterisation of goats added to the exercise. Other studies in progress include crossbreeding for fibre production, the objective being to investigate the potential in respect of fibre and meat from suitable crossbreds (Sikosana and Maphosa, 1993). Weaning weight of crossbreds is about 14kg at the age of 20 weeks. Weight of fibre at first shearing, at an average age of six months, ranged from 0.2kg to 0.8kg per animal.
DEVELOPING BREEDING PLANS
The results presented so far suggest that there is scope to improve the performance of the existing goat types through breeding. Breeding strategies could use the following approaches: Open nucleus flocks
This would involve the farmers' participation during implementation. The scheme would involve selection of animals performing well within farmers' flocks. Selected superior males and females could then be tested both on-station and on-farm (with infcrior flocks) before distribution. Transfer of males between flocks would reduce inbreeding. With the advancement of the 'acceptability' of technology and planned production systems selected males could be made available through artificial insemination.
Grading flocks with a superior indigenous breed
This is similar to the approach described above. It will involve producing superior males and distributing to flocks to be improved. Inbreeding would be controlled. Breed charaterisation and standardisation are the key to develop a desirable genetic pool. This could be done by determining genetic relationships among goats. Low genetic potential usually results in unnecessary crossbreeding with exotic breeds. Animal breeders and farmers are compelled by economic forces to adopt germplasm for immediate (short-term) benefits without projecting sustainability consequences (Kiwuwa, 1990). In such circumstances results have shown improvement in productive traits at the expense of fitness traits and this could lead to disaster in the long term.
CONCLUSION
More work should be aimed at developing breeding systems for the local type of goat. Genetic variability among indigenous goats needs determination and straight and crossbred compositions with other breeds and crosses is desirable. Initial work in this sphere would concentrate on identifying strains, superior individuals, their selective improvement and determination of genetic parameters. Emphasis should be placed on meat, milk, fibre and skin production. There might be need in the fixture to establish a Central Breeding Station for continuous genetic improvement of goats. If breeding programmes are to be undertaken animals should not be bred outside the environment they are to perform. These strategies have to consider both 'traditional' and 'modern' production systems.
REFERENCES
Archarya, R. M. (1992) In Small Ruminant Production: Systems for Sustainability. ed. Lokeshwar, R.R. pp 37-45.
Baffour-Awuah, O. (1987) Proc. Zimbabwe Soc. of Anim. Prod. Livestock Res. Symposium, Harare, Zimbabwe.
Chifamba, I. K. and Sithole, L. (1992) In: Annual Report of the Division of Livestock and Pastures 1991-92. DR&SS, Ministry of Agriculture, Zimbabwe. (in press).
Devendra, C. and Burns, M. (1983) Goat Production in the Tropics (Tech. Comm. No. 19 Commonwealth Bureaux Animal Breeding Genetics).
Hatendi, P. R. (1992) In: Proceedings of the First Biennial Conference of the African Small Ruminant Research Network. eds. Rey, B., Lebbie, S.H.B. and Reynolds, L.
Khombe, C. T. (1985) Proceedings of a Conference on Small Ruminants in Africa Agriculture. eds. Wilson,R.T and Bourzat, D.
Kiwuwa, G. H. (1990) In Proceedings of the First Biennial Conference of the African Small Ruminant Research Network. eds. Rey, B., Lebbie, S.H.B. and Reynolds, L. Mason, I. L. and Maule, J. P. (1960) The Indigenous Livestock Of Eastern and Southern Africa (Tech. Comm. No. 14 Commonwealth Bureax of Animal Breeding Genetics)
Ndlovu, L.R. 1988. Proc. Goat Development Workshop, Harare, Zimbabwe.
Sibanda, L. (1992) PhD Thesis. University of Reading, United Kingdom.
Sibanda, R. (1990) MSc Thesis. University of Zimbabwe.
Sibanda, R. and Sikosana, J. L. N. (1986) mimeograph, Matopos Research Station, Bulawayo, Zimbabwe.
Sikosana, J. L. N., and Maphosa, V. (1993) In: Annual Report of the Division of Livestock and Pastures 1992-93, DR&SS. Ministry of Agriculture, Zimbabwe.
Wilson, R. Y. (1991) Animal Production Health Paper No.88. Food and Agricultural Organisation, Rome, Italy. USE OF MOLECULAR GENETIC TECHNIQUES IN LIVESTOCK: A CASE EXAMPLE IN GOATS
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
E. Bhebhe and S. K. Davis University of Zimbabwe, Department of Paraclinical Veterinary Studies, P O Box MP 167 Mt. Pleasant, Harare, Zimbabwe and Texas A&M University, Department of Animal Science, Animal Genetics Section, College Station, Texas 77843-2128, USA.
SUMMARY
Microsatellite markers were isolated by sequencing (GT)n positive pUC 19 clones containing size selected (300-900 base pair) goat genomic DNA inserts. After locating (GT)n repeat sequences longer than 18 base pairs, primer sequences were designed and synthesized to amplify the repeat region. Each microsatellite primer pair was included in PCR reactions to amplify genomic DNA from a two generation goat pedigree. This facilitated determination of compliance of the microsatellite marker with Mendelian inheritance. Second, the primer pair was used in similar reactions using DNA templates from 4-way cross goats derived from the four genetically diverse breeds East African, Galla, Anglo Nubian and Toggenburg to determine the degree of polymorphism at the microsatellite locus. Twelve microsatellite markers with polymorphic information content (PIC) values ranging from 0.34 to 0.89 were isolated and scored in 270 goats comprising 21 half-sib families. Sequences for microsatellite markers SR-CRSP-1 to SR-CRSP-10 were submitted to Genbank data library (Accession numbers: L22192, L22194, L22195, L22196, L22197, L22198, L22199, L22193, L22200 and L22201).
The microsatellite markers SR-CRSP-2 and SR-CRSP-3 were determined to be linked with a lod score of 3.14, a recombination rate of 22% and an estimated distance of 24.2 map units (mu). Linkage of SR-CRSP-2 with SR-CRSP-7 at a distance of 30.4 mu was suggested by the data, but the evidence for linkage was not conclusive (lod score = 2.28). The remainder of the microsatellite markers were determined to be unlinked.
Breedtype, type of birth and gender were significantly associated (p < .05) with birthweight, weaning weight and yearling weight. Significant associations were also found between microsatellite markers SR-CRSP-7 and birth weight (p < .04), SR-CRSP-2 and weaning weight (p < .02), and SR-CRSP-11 and yearling weight (p < .01).
INTRODUCTION The problems and economic burden caused by internal and external parasites need not be overemphasized. Current systems of parasite prevention and control put relatively little emphasis on the innate animal mechanisms that have evolved to maintain the healthy status and lean heavily on the use of therapeutic agents. The use of therapeutic agents carries several penalties. First, it tends to minimize the selective advantage of natural resistance in animals while favouring the diversification of the parasite to forms more resistant to drugs (Nicholas, 1987; Waller, 1990). Second, it poses a serious threat to human health, and drugs and drug residues are potentially detrimental to the environment. The damage caused in such cases generally becomes obvious only after considerable use. As an example, the adverse effects of Ivermectin residues on invertebrate fauna in animal dung has been well documented (Wall and Strong, 1987). Finally, the cost of drugs is high. In 1984, costs for anthelminthic drug treatment in Australia totaled $90 million and lost production attributable to parasites was assessed at $360 million (Baker et al., 1990). The emerging awareness worldwide of environmental issues, in particular consumer demand for animal products and pastures free of chemical residues, is placing immense pressure on the agricultural industry as a whole to find alternatives that would reduce the industry's contribution to environmental destruction. These economic and environmental considerations provide a strong impetus to develop alternative or complementary strategies for control of the parasites.
Classical livestock improvement through selection has relied on the biometrical evaluation of breeding values from individual performance records and from the performances of relatives (Falconer, 1989). Over the last few decades, animal breeders have been very successful in employing this approach to exploit concepts in quantitative genetics theory to breed animals with superior production potential. However, it has become apparent that production systems based on this approach may not be self sustaining, and there is evidence suggesting that general fitness has been decreasing in some breeds due to intensive production practices and "genetic stress" imposed by selection (Wu et al., 1989). Livestock breeding programs have not included selection for parasite resistance for three reasons. First, the industry felt secure because chemotherapeutic methods appeared to provide satisfactory control, although anthelminthic resistance has been observed where drugs have been used frequently and for prolonged periods of time (Donald, 1982). Second, obtaining measurements of eggs per gram of feces (EPG) and (or) whole blood packed cell volume (PCV) on sufficient numbers of animals to effect a selection program for improved resistance is impractical and may not be cost-effective. Third, considerable reservations have been expressed by parasitologists concerning the use of EPG and PCV as measures of resistance to internal parasites (Dargie, 1982; Gruner, 1991). Despite these concerns, it is now widely accepted that EPG is highly repeatable, heritable, and correlated with actual worm burden (Gray, 1991). The identification of molecular markers that explain a substantial amount of the phenotypic variation in resistance and the incorporation of such marker information in marker-assisted selection (MAS) (Soller and Beckman, 1982; Smith and Simpson, 1986) schemes would enhance response by increasing the accuracy of selection and reducing the generation interval. In addition, isolation of markers linked to a quantitative trait locus (QTL) or disease locus of interest would be the first step towards isolation of the gene(s) controlling that trait thereby facilitating manipulation of the gene(s) at the molecular level.
The innate resistance of certain breeds of sheep or goats to helminth parasites was first documented by Steward et al. (1937) for the infection of sheep by Ostertagia circumcincta. It has since been shown that such differences are under genetic control (Wetherall et al., I 991). Other studies have reported resistance to parasites in various species, including resistance to Haemonchus contortus in goats (Davis et al., 1991) and sheep (Radhakrishnan et al., 1972; Altaif and Dargie, 1978a, b: Gray et al., 1987) trypanotolerance (Murray et al., 1982; 1991), resistance to ticks in cattle (De Castro, 1991), resistance in poultry to salmonella (Bumstead and Barrow, 1988), coccidiosis (Bumstead and Millard, 1987), Marek's disease (Hansen et al., 1967) and fowl cholera (Lamont et al., 1987), and the resistance of humans to malaria (Kemp et al., 1987; Weber, 1988).
If the evidence for the involvement of heritable resistance to parasitic challenge is abundant and diverse, so are the methods that have been employed in the search for genetic markers for disease/parasite resistance. However, a disproportionately large number of these studies seek to explain resistance or susceptibility to a pathological condition solely on the basis of variability within the major histocompatibility complex. Such studies are based on the assumption that an immune response is involved in resistance to parasites and that the genetic polymorphisms of class I and class II MHC molecules in the membranes of monocytes, macrophages and B-lymphocytes restrict the ability of some animals to respond to parasite antigens (Outteridge, 1991). Interest in the MHC arose from the extraordinary polymorphism exhibited by the loci within this chromosomal segment and the observation that immune responsiveness in inbred mice is controlled by the MHC (Wetheral et al., 1991). The role of the MHC in disease associations has been reviewed by Nicholas (1987). Jennings et al. (1978) reported differences in the resistance of strains of mice to Trypanosomiasis congolenses. It was later shown that resistance to T. congolenses in mice is under the control of a single gene (Blackwell, 1988; Pinder, 1984), which turned out to be unlinked to the MHC complex (Morrison and Murray, 1979; Pinder et al, 1985). In reviewing the rodent model for resistance to a variety of parasites, Wakelm (1991) noted that the genes that exerted the strongest influences on immunity to nematodes appear to lie outside the MHC. It thus should not be too surprising that, in livestock species other than poultry, most of the associations reported between MHC type and host resistance/susceptibility to parasitic challenge are not convincing. In summarizing an extensive set of breeding experiments aimed at elucidating the number and character of genes affecting immune responsiveness, Biozzi et al. (1980) concluded that most immune responses behave like quantitative traits, ie, they are determined by the additive effects of several independent loci. In these studies, it was demonstrated that the mouse MHC accounted for only 12-26% of the interline genetic differences in antibody responses to sheep red blood cells. In discussing preliminary results of studies of goats resistant to Haemonchus contortus (based on fecal egg counts) Davis et al. (1991) noted that resistance is unlikely to be under the control of a single gene locus, and they suggest that two or more genes, perhaps with several modifiers, determine the phenotype of resistance. Teale et al. (1991) noted that since parasites are antigenically complex, it is naive to expect strong associations between resistance/susceptibility and class I or II genes of the MHC per se. In addition, studies attempting to relate parasite or disease resistance to variation within the MHC are further complicated by linkage of the MHC to non-MHC genes of possible significance in disease control, such as those encoding complement components and tumor necrosis factor (Teale et al., 1991). The limited success of the massive efforts targeting the MHC, coupled with evidence suggesting polygenic control of disease resistance traits, suggests that it may be more fruitful to search the entire genome. Furthermore, a total genomic search may be highly desirable if the experimental design and data collected can accommodate statistical testing of marker information against a battery of genetically unrelated traits. In fact this approach is now being implemented in sheep, where initial studies on parasite and disease resistance usually involved RFLPs with cDNA probes for the MHC, but recent investigations are shifting towards a more general RFLP study to provide a map of the sheep genome which can ultimately be used to locate genes for resistance to parasites and other quantitative traits (Outteridge, 1991). In light of the available information, a study employing a reverse genetics approach as opposed to a candidate gene approach was setup. The reverse genetics approach assumes all marker loci are candidate genes (Davis et al., 1991), and has been the basis for successful searches for QTLs with major effects in plants (Edwards et al.; 1987) and in humans (Talmud and Humphries, 1986; Boerwinkle and Singh, 1987).
The objectives of the study were to identify polymorphic microsatellite markers in the goat genome and to statistically evaluate associations between these markers and resistance/susceptibility to Haemonchosis and coccidial infestation, as well as to appraise marker associations with production traits. In addition, it was desirable to investigate the conservation of goat derived microsatellite primer sequences in other species, so that the goat map could be tied to the framework of other mammalian gene maps.
MATERIALS AND METHODS
Animals, animal parameters, and DNA sampling
Phenotypic data and DNA samples used in this study were obtained from a dual purpose goat breeding project established and managed under the auspices of the Small Ruminant Collaborative Research Support Program (SR-CRSP) in collaboration with the Kenyan government. The primary objective of the project is to develop a four-breed composite goat that will thrive in the Kenya smallholder production environment and possesses increased genetic potential for milk and meat production. The composite breed is being developed from two imported breeds (temperate milk breeds) and two indigenous breeds to produce an animal whose genetic constitution is equal proportions of Toggenburg, Anglo Nubian, East African and Galla. The flock generated in the process provided an ideal population for QTL analysis. Detailed descriptions of the environmental conditions, general management of the goats and data collection were provided by Ruvuna et al. (1988). All animals except the founder population have pedigree, multiple vs. single birth, dates of birth, weaning and disposal, in addition to birth, weaning, yearling and mature weights recorded. Fecal and blood samples were obtained continuously on all goats at 2 wk intervals starting at 4 months of age until one year of age.
A total of 407 blood samples were collected from the four goat breeds and their F1 and four way crosses. Blood samples were collected in 50 ml EDTA tubes, and white blood cells were separated by centrifugation. White blood cells were further purified by selectively lysing the remaining red blood cells with 0.2% ammonium chloride. Genomic DNA was extracted from the white blood cells by digesting the white blood cells with l0mg/ml proteinase K and 25 µl of 20% sodium dodecyl sulphate (SDS) at 55°C for 2 hours or 37°C overnight. DNA samples were then purified by extracting twice with an equal volume of PCI (a 25:24:1 mixture of phenol, chloroform, and isoamyl alcohol) and twice with equal volumes of chloroform. The DNA was then precipitated with two and a half volumes of cold absolute ethanol, dried at room temperature, resuspended in water to a concentration of approximately 0.5-1.0 µg/ µl and stored at 4°C.
Genotyping procedures
Microsatellite primer sequence information was obtained from size selected partial genomic libraries as follows. Goat genomic DNA was digested to completion at 37 °C overnight with Sau3A I in a 100 µl reaction mixture containing 0.1 to 0.2 micrograms/µl DNA, 1X Sau3A I buffer, O.lµg/µl BSA and 24 units of Sau3A I. Choice of restriction enzyme was based on the desire to cut the DNA as frequently as possible and the fact that the enzyme generates sticky ends complementary to those of the pUC19 polylinker digested with Bam HI.
Size selection was achieved by electrophoresis of the DNA digest against a lambda/Dra I size standard on either a 0.8% agarose or 1.5% low melt agarose horizontal minigel. The minigel was then stained with ethidium bromide and excision of DNA fragments ranging from 150 by to 800 by was carried out under uv light. DNA was recovered from the agarose gel by first crushing the plugs with a sterile glass rod followed by freeze-thawing (-80°C and 55°C, respectively) three to four times. The resulting agarose pulp was centrifuged at 13,000 rpm for 5 minutes and the liquid phase transferred to a new tube. Precipitation of DNA was achieved by addition of 1/10 volume of 2 M NaCl and 2.5 volumes of absolute ethanol. The DNA was then resuspended in 90 microliters of 10 mM Tris Cl (pH 8.3), and treated with calf intestinal phosphatase (Promega, Madison, WI) following the manufucturer's recommendations to inhibit chimera formation. The mixture was then treated with RNase at a final concentration of 100 µg/ml and the enzymes removed by extracting once each with equal volumes of PCI and chloroform. An equimolar amount of pUC 19 (New England Biolabs, Beverly, MA) DNA was digested overnight with Bam Hl (Promega, Madison, WI) following the manufacturer's recommendations. The genomic DNA and pUC19 digests were combined and coprecipitated by the addition of 1/10 volume 2 M NaCl and 2.5 volumes of cold absolute ethanol and left at - 80°C for at least 2 hours. The coprecipitate was then centrifuged at 10,000 rpm for about 10 minutes and the air dried pellet resuspended in 20 µl of water. Ligation of the genomic DNA fragments into the Bam H1 cloning site of pUC 19 polylinker was carried out at 16°C overnight using T4 DNA ligase (IBI, New Haven, CO) following the manufacturer's recommendations.
A 2 µl volume of the ligation mixture was used to transform 50 µl of DH5, competent cells (GIBCO BRL, Gaitherburg, MD) at 37°C for one hour. A 30 to 50µl volume of the transformation mixture was plated out on 150 x 15 mm LB agar plates containing ampicillin and incubated at 37°C for 12 to 14 hours. Nitrocellulose filter replicates of the bacterial colonies on the LB agar were made following a protocol similar to that of Sambrook et al. (1989). The filter bound DNAs were screened by hybridizing to synthetic (CT)15 or (GT)15 probes end labeled with 32p-ATP. Hybridizations were carried out in 5X SSC (pH 7.0) at 37°C overnight. Filter washes were carried out in 2X SCC solution and positive clones identified by exposing X-ray film to the filters for 2 hours. DNA from all positive clones was denatured to generate single stranded DNA and sequenced using sequenase® version 4.0 (United States Biochemical, Cleveland, OH) following the manufacturer's suggested protocol. When a microsatellite sequence with ten or more repeats was located, the repeat sequence plus at least fifty bases flanking both the 5' and 3' end of the repeat were read into MacVector® version 2.0 and PCR primers designed with the aid of the program.
Primers were synthesized on an Applied Biosystems 392 DNA/RNA synthesizer. The GT strand primer (hereafter, forward primer) was 5' end labeled with 32p-ATP for one hour at 37°C in a 20µl reaction mixture containing approximately 25 pM/µl of primer, 1.66 pM 32p-ATP at 3,000 mCi/µl, 10mM MgC12 and 0.4 units/µl of T4 polynucleotide kmase.
Standard polymerase chain reactions (Saiki et al., 1985, 1988; Mullins and Faloona, 1987) were carried out on a Perkin-Elmer Gene Amp® PCR 9600 (Perkin-Elmer, Norwalk, CT) orTechne MW-2 Dri-Plate® cycler (Techne Cambridge Ltd, Duxford; U.K.). The 25 µl reaction mixture contained 10 to 20 ng of genomic template, 0.2 pM end-labeled forward primer and 0.2 pM unlabeled reverse primer, 2 mM each dATP, dCTP, dGTP, and dTTP, 1.25 mM MgC12 0.5% DMSO and 0.625 units of Taq polymerase. Samples were processed through one cycle at 95°C for 1 minute (initial denaturing), followed by 35 temperature cycles at 95°C (denaturing) for 20 seconds, 48-55°C (depending on primer pair) for 30 seconds (annealing), and 30 seconds at 72°C (extension). The final elongation step was lengthened to 10 minutes. Equal volumes of the PCR product and sequencing gel loading buffer (95% formamide, 20mM EDTA, 0.05% xylene cyanol FF, 0.05% bromophenol blue) were mixed and 3µl of the mixture was electrophoresed on 6-8% denaturing polyacrylamide gels at 40-50 watts for 2 to 4 hours. Gels were dried onto filter paper supports and then processed for autoradiography. Exposure times were from 6 to 72 hrs. A total of 255 goat samples belonging to 21 sire families each with a minimum of six half sibs groups were scored for twelve microsatellite markers.
To investigate the degree of microsatellite primer sequence conservation between the caprine and other species, PCR reactions were also carried on DNA from ten unrelated cattle, ten unrelated sheep, five unrelated pigs, and two human DNA samples. In addition, the goat derived microsatellite primers were tested in several species of ruminants including Bison bison (Bison), Bos grunniens (Gaur), Odocoileus virginianus (White tailed deer), Cervis canadensis (Elk), Cervis elaphus (Red deer), and Gazella spekei (Spekes Gazelle).
Statistical analysis
Given the half-sib design, it was necessary to nest marker allele effects within a sire family effect (hereafter, sire-marker) in order to identify families segregating for QTLs of major effect. This required the identification of the paternal allele in each offspring and was achieved by comparing offspring and sire genotypes. When the sire's DNA sample was not available for genotyping, a maximum likelihood procedure was used to infer the sire's genotype from his offspring. The use of derived sire genotypes enabled the inclusion of 114 more animals in the data set. The paternal allele for offspring with the same genotype as their heterozygous sires could not be determined in most cases since a majority of the dams were not genotyped.These cases were assigned an arbitrary allele designation (99) at that particular locus and were excluded in statistics for that locus. The DNA samples for a majority of the does were not available for genotyping, hence it was not possible to determine marker-QTL phase relationships within parental animals. In the course of the analysis it was determined that sires, and hence sire-marker combinations, were not cross classified with breedtype. This made it necessary to nest sire-marker effects within breedtype. Linear statistical models were fitted to estimate and test the statistical significance of genotypic effects associated with each marker locus in separate analyses for each trait. For birth, weaning, and yearling weight dependent variables, the full model assumed in analyzing the data was:
Yijklmnp = µ + Ti + Sj + Zk + YRl + Bm + Gmn + Eijklmnp where:
th th th Yijklmnp = observation on the pth kid of the j gender, born in k season of the the l year, having the ith type of birth and mth breedtype that received the nth allele from its sire.
µ = population constant common to all observations;
Ti = the fixed effect of the ith type of birth; i = 1 or 2 (single, twin);
Sj = the fixed effect of the jth gender of kid; j = 1 or 2 (female, male);
Zk = the fixed effect of the kth season of birth; k = 1,2,3;
YRl = the fixed effect of the lth year of birth; 1= 1,...,6;
Bm = the fixed effect of mth breedtype; m = 1,...,8; (I . East African; 2. East African x Galla; 3.
Toggenburg x East African; 4. Toggenburg x Galla; 5. Anglo Nubian x East African; 6. Anglo Nubian x Galla; 7. Unplanned crosses; 8. 1/4 East African, 1/4 Galla, 1/4 Toggenburg and 1/4 Anglo Nubian);
th th Gmn = the fixed effect of the n paternal allele within the m breedtype; n=4,...,72 (depending on the number of alleles at the marker locus included in the model);
Eijklmn = random error term assumed to be normally and independently distributed with mean 0 and variance equal to 2.
Seasons were defined by the rainfall and temperature pattern, so that season 1 includes the hot and dry months of the year (December, January, February and March), season 2 includes the cold and wet months (April, May, June, October and November) and season 3 includes the cold and dry months (July, August and September). Age of dam was included in preliminary models as a fixed effect or as a covariate but was found not to be significant (P>0.05) and was thus excluded from the final model. A similar model was fitted for the characters chosen to represent parasitic burden with the following modifications. Preliminary analysis found type of birth to have no significant effect on LEPG and PCV values, hence this variable was excluded from the model. In addition, the year and season effects in this model are those for the year and season when the samples for EPG, COC and PVC determinations were obtained. The two models described above constituted the full model and the sum of squares associated with the marker effects reflect the effect of the sire and marker effects combined. By removing the marker effect from the model, the variance associated with the marker effect becomes included in the random error term. Thus, to construct the partial sum of squares for the marker effect, reduced models excluding the marker genotype were fitted. When using this approach, test of significance of the marker effect examines the reduction in the variance accounted for by the model due to exclusion of the marker effect from the model.
PIC at each marker locus was computed following Botstem et al. (1980) and direct count heterozygosity was determined as the fraction of individuals that were heterozygous at the marker locus.
n n-1
PIC = 1 - ( pi2-)-2n 1 I pi2 pi2. i=1 i=lj =i+1
Where: I = the ith marker allele of interest
n = the number of alleles at the locus and
pi and pj = are the sample frequencies of the ith and jth alleles.
Significance of each marker effect was determined using an F-statistic. The numerator mean square was obtained by eliminating the marker effect from the model and calculating the partial sum of squares due to this effect as the difference between the sum of squares for the full and reduced models, then dividing by the degrees of freedom for the effect. The denominator for the F ratio was the error mean square from the full model.
Linkage between microsatellite markers was determined using CRIMAP (Green et al., 1990).
RESULTS AND DISCUSSION
Microsatellite screening in size selected partial 2enomic libraries
The size selected librariescontained genomic inserts which could be sequenced in their entirety using a set of universal pUC primers, eliminating the need to synthesize internal primers for each clone. In all but one case, size selection resulted in genomic inserts less than 800 bp. A total of 123 positive clones were detected, 78 by hybridizing to a radioactive (GT)15 probe and the rest by hybridizing to a radioactive (CT)15 probe. Upon sequencing, 43 of the (GT)n positive clones contained a (GT)n sequence (where n _ 5) within 300bp of one of the pUC universal primers. The 35 (GT)n positive clones that failed to yield a microsatellite sequence were either false positives resulting from low stringency washes, or were clones in which the microsatellite sequence was more than 300 bases from both of the universal pUC 19 sequencing primers. Goat microsatellites identified in the course of this study were of varying complexity, ranging from uninterrupted repeat sequences (perfect repeats) to complex combinations of different repeat sequences interspersed with other nucleotides. PCR primer pairs were designed and synthesized for twenty three microsatellite sequences. The remaining 20 sequences were dropped either because the microsatellite sequence contained less than 10 repeats, or because suitable primer pairs could not be designed for the flanking sequences. The minimum number of repeats for marker locus consideration was based on the observation that,in humans, informativeness of (GT)n sequences is directly related to the number of repeats, with those sequences containing 10 or fewer repeats exhibiting very low to negligible PIC values (Weber 1989). In cases where sequences were eliminated due to failure to design an appropriate primer pair, one or both of the regions flanking the microsatellite sequence contained GC compressions making accurate determination of the sequence impossible, or one of the flanking regions contained a long single base repeat. In three cases, potential marker loci could not be considered because one of the flanking regions was adjacent to the vector cloning site and thus was too short to permit primer design.
After optimization of the PCR reactions and testing for compliance with a codommant segregation pattern in a two generation pedigree, twelve primer pairs produced discrete PCR products of the expected size based on the sequence of the clone used to generate the primer pair. The other eleven primer pairs either failed to reliably amplify or produced extremely complex banding patterns whose inheritance could not be determined in the two generation pedigree. Failure to amplify a microsatellite region in individuals other than the individual from which the library was made may be indicative of poor conservation of one or both of the microsatellite primer pairs.
The occurrence of PCR artifactual bands for each allele was evident in all the microsatellite loci, being more pronounced for some PCR primers than others. Several measures, including reducing PCR primer concentration, titrating levels of DNA template, varying Mg2+ concentrations, increasing annealing temperature and reducing the number of PCR cycles, failed to reduce the frequency or intensity of these additional bands. Despite the presence of PCR artifactual bands, it was still possible to determine the segregation pattern and genotypes of each individual animal for all twelve microsatellite loci. The occurence of PCR artifactual bands has been reported previously (Litt and Lutty 1989; Weber and May 1989; Olsen and Eckstein 1989; Weber 1990) and it has been suggested that they may be the result of incomplete primer extension by the Taq polymerase, terminal transferase activity of the enzyme or strand slippage during PCR replication (Olsen and Eckstein 1989).
In addition to revealing the targeted (GT)n repeat sequences, several other repeat types including (AT)n and (GC)n were revealed on the sequencing gels. The most interesting of the untargeted repeat sequences were long (usually greater than 10 repeats) mononucleotide repeat sequences upstream or downstream of some microsatellite sequences. Most of the (CT)n positive clones sequenced failed to yield a microsatellite sequence and since (GT)n screening was working acceptably well. no further attempts were made to resolve the problem of false (CT)n positives. However, it seems likely that increasing the stringency of the washes could eliminate many of these false positives. The number of alleles per microsatellite locus was highly variable and ranged from 4 to 12. Overall, the expectation of increasing marker informativeness with an increase in the number of alleles per marker locus was observed (r = 0.65). Unfortunately, the difficulty of allele identification increased with the number of alleles.
Microsatellite primer sequence conservation
Nine of the twelve microsatellites used in this study were tested on 186 cow DNA samples, all produced a PCR amplification product, and all but one were found to be polymorphic. The microsatellite markers were also tested in several other ruminant species. No PCR amplification products were obtained with human or porcine DNA samples and varying levels of heterozygosity were observed with the other species.
Linkage between markers and marker QTL associations
The microsatellite markers CRSP- SR2 and SR-CRSP-3 were determined to be linked with a recombination rate of 22%, a lod score of 3.14 and a map distance of 24.2 cM (Kosambi). SR- CRSP-2 was also found to be located 30.4 cM from SR-CRSP-7, but the evidence of linkage was not conclusive (lod score = 2.28). The rest of the microsatellite markers were determined to be unlinked. Significant associations were found between microsatellite markers SR-CRSP- 7 and birth weight (p < .04), SR-CRSP-2 and weaning weight (p < .02), and SR-CRSP-l I and yearling weight (p < .01). No significant associations (p < .05) were detected between any of the marker loci and measures of parasitic burden.
CONCLUSION
PCR based marker systems are rapidly replacing Southern blot-based systems in molecular genetic studies. However, the need to know the nucleotide sequences flanking the region of interest makes initial development of these markers time consuming and expensive. Microsatellite markers were chosen for this study because of their co-dominant mode of inheritance, the high degree of polymorphism, extremely high repeatability of the genotyping results, and the high conservation of these markers across a wide variety of ruminant species more than compensates for the extra money and time involved in their isolation. The co- dominant mode of inheritance allowed for determination of the paternally inherited allele in all individuals except those offspring with the same genotype as the sire. Given the high degree of polymorphism observed for microsatellite markers, the number of individuals with the same genotype as the sire will be relatively small. Analysis of marker-QTL associations within sire families is advocated since this approach eliminates erroneous marker-QTL associations resulting from markers being confounded with sirefamily effects. Furthermore, the linkage relationship between the alternate marker alleles and the alternate alleles for the QTL of interest is not necessarily the same in progeny from different sires. Hence, models which are otherwise well parameterised but cross-classify genotypes with family effects are less likely to find significant associations even when they exist. The high degree of conservation of these markers across species not only makes comparative gene mapping relatively easy, but also lowers the overall cost of marker development. The occurrence of 'shadow bands' presented a challenge in genotype determinations for some of the marker loci, however this did not preclude genotype determination. It is likely that as more researchers adopt microsatellite genetic markers for their investigations, minor modifications to PCR protocols, reaction conditions and primer design strategy will provide ways to reduce the prevalence of shadow bands. Indeed, five different factors that can lead to 'shadow bands' have recently been identified by Litt et al. (1993) and possible remedies presented.
The identification of two, and possibly three linked microsatellite markers produced the first linkage group reported in this species and will contribute to marker saturation of the goat genome. In addition the identification of microsatellite markers associated with birth weight, weaning weight and yearling weight traits should stimulate a more extensive search for microsatellite markers in this species. However it must be noted that the associations reported between markers and QTLs in this study constitute strictly statistical associations whose validation is outside the scope of this study. The achievement of more definitive conclusions will require larger families and a better sampling scheme.
REFERENCES Altaif K.I. and Dargie J.D. (1978a) Research in Veterinary Science 24: 391-393.
Altaif K.1. and Dargie J.D. (1978b) Parasitology 77: 177-190.
Baker R. L. Watson T.G., Bisset S.A. and Vlassoff, A. (1990) Proceedings of the Australian Association for Animal Breeding and Genetics 8: 173-178
Biozzi G., Siqueira M., Stiffel C., Ibanez O.M., Mouton D. and Ferreira V.C.A. (1980) Progress in immunology Vol. IV, pp. 432-457. Academic Press, New York.
Blackwell J.M. (1988) Genetics of Resistance to Bacterial and Parasitic Infections. Taylor and Francis, London.
Boerwinkle E. and Singh C.F. (1987) Ann. Hum. Genet. 51: 211.
Botstein D., White R.L., Skolnik M. and Davis R.W. (1980) American Journal of Human Genetics 32: 314.
Bumstead N. and Burrow P.A. (1988) British Poultry Science 29: 521-30.
Bumstead N. and Millard B.M. (1987) British Poultry Science 28: 705-16.
Dargie J.D. (1982) In: Animal Models in Parasitology. (ed. G.D. Owen). McMillan Press Ltd, London.
Davis S.K., Taylor J.F., and Ruvuna F. (1991) In: Biotechnology in Kenya: Proceedings of the national conference on plant and animal biotechnology. (ed, by E.M. Mailu, J.O. Mugah and P.O. Fungoh). Initiative publishers, Nairobi.
De Castro J.J. (1991) In: Breeding for Disease Resistance in Farm Animals (ed. by J.B. Owen and R.F.E. Axford), pp. 244-262. C.A.B. International, Wallingford.
Donald A.D. (1982) Proc. Resistance of Parasites to Anthelmintics Commission of European Communities Vol. IV, pp. 15. Cent. Vet. Inst., Lelystad.
Edwards M.D., Stuber C.W. and Wendel J.F. (1987) Genetics 116: 113-119.
Falconer D. S. (1989) (3rd ed.). Longman scientific and Technical, New York.
Gray G.D., Presson B.L., Albers G.A.A., LE Jambre L.F., Piper L.R. and Barker J.S.F. (1987) In: Merino Improvement Programs in Australia (ed. by B.J. McGuirk), pp. 365-369. Wool Research and Development Corporation, Melbourne.
Gray G.D. (1991) In: Breeding for Disease Resistance in Farm Animals (ed. by J.B. Owen and R.F.E. Axford), pp. 139-161. C.A.B. International, Wallingford.
Gruner L. (1991) In: Breeding for Disease Resistance in Farm Animals (ed. by J.B. Owen and F.E. Axford), pp. 187-202. C.A.B. International, Wallingford.
Hansen H.P., van Zandt N. and Law G.R.J. (1967) Poultry Science 46: 1268-1290.
Jennings F.W., Whitelaw D.D., Holmes P.H. and Urquhart G.M. (1978) Research in Veterinary Science 25: 399-412.
Kemp DJ, Coppel L.R. and Anders F.R. (1987) Annu. Rev. Microbiol. 41: 181-208.
Lamont S.J., Bolin C. and Cheville N. (1987) Immunogenetics 25: 284-289.
Litt M. J. and M. Luty. (1989) American Journal of Human Genetics 44: 397-401. Morrison W.I. and Murray M. (1979) Experimental Parasitology 48: 364-374.
Murray M., Morris W.I. and Whitelaw D.D. (1982) In: Advances in Parasitology Vol. 211. (ed. by Baker J.R. and R. Muller), pp. 1-68. Academic Press, London.
Murray M., Stear M.J., Trail J.C.M., d'leteren D.G.M., Agyemang K. and Dwinger R.H. (1991) In: Breeding for Disease Resistance in Farm Animals (eds. J.B. Owen and R.F.E. Axford), pp. 203-223. C.A.B. International, Wallingford.
Nicholas F.W. (1987) Veterinary Genetics. Clarendon Press, Oxford.
Nicholas F.W. (1987) Veterinary Genetics. Clarendon Press, Oxford.
Outteridgc P.M. (1991) In: Breeding for Disease Resistance in Sheep (eds. G.D. Gray and R. R. Woolaston), pp. I 1-18. Wool Research and Development Corporation. Melbourne.
Pinder M., Fumoux F. and Roelants G.E. (1985) Progress in Leucocyte Biology 3: 495-500.
Pinder M. (1984) Infection and immunity 57: 185-194.
Radhakrishnan G.V., Bradley R.E. and Loggins R.E. (1972) American Journal of Veterinary Science 33: 817-823.
Ruvuna F., Cartwright T.C., Blackburn H., Okeyo M. and Chema S. (1988) J. Agric. Sci. Camb. 111: 363-368.
Smith C. and Simpson S.P. (1986) J. Anim. Breed. Genet. 103: 205-217.
Soller M. and Beckmamn J.S. (1982) Proc. of the 2nd World Congress on Genetics Applied to Livestock Production Vol. VI, pp. 396-404. Editorial Garsi, Madrid.
Steward M.A., Miller R.F and Douglas JR. (1937) Journal of Agricultural Research 55, 923- 930.
Talmud P.J. and Humphries S. (1986) Lancet ii, 104: 43-60.
Teale J.A., Kemp J.S. and Morrison I.W. (1991) In: Breeding for Disease Resistance in Farm Animals (ed. by J.B. Owen and R.F.E. Axford), pp. 86-99. C.A.B. International, Wallingford.
Wakelin D. (1991) In: Breeding for Disease Resistance in Farm Animals (ed. by J.B. Owen and R.F.E. Axford), pp. 54-72. C.A.B. International, Wallingford.
Wall R. and Strong M. (1987) Australian Veterinary Journal 64: 9-50.
Waller P.J. (1990) Parasitology Today 6, 127-129.
Weber J. L. (1988) Exp. Parasitol. 66: 143-170.
Wetherall J.D., Groth D.M. and Karlsson L.J.E. (1991) In: Breeding for Disease Resistance in Sheep (ed. by G.D. Gray and R.R. Woolaston), pp. 109-114. Wool Research and Development Corporation, Melbourne.
Wu M.C., Shanks R.D. and Lewin H.A. (1989) Proc. of the National Academy of Sciences of the USA 86: 993. PHENOTYPIC DIFFERENCES IN DIGESTIVE PARAMETERS BETWEEN LOCAL AND DORPER SHEEP
SUMMARY
INTRODUCTION
MATERIAL AND METHODS
RESULTS AND DISCUSSION
REFERENCES
B. Mwenya, M. Blummel and C. Wollny University of Malawi, Bunda College of Agriculture, P 0 Box 219 Lilongwe, Malawi.
SUMMARY
The intake, the selection and the digestibility of low quality roughage diets were studied in Malawian local and high grade Dorper intact male sheep. Twenty animals were kept in metabolic cages facilitating total urine and feces collection. Treatment consisted of 3 different levels of maize stover (1, 1.5 and 2 times the maintenance requirements) which were supplemented by urea nitrogen as a block. In a subsequent experiment the amount of stover offered was twice the maintenance requirement supplemented with bean pods and poultry manure. The maize stover intake (DMI-MS) increased significantly from 30 to 52.lg/kg/LW 0.75/d and from 33 to 54.1 g/kg/LW 0.75/d in the local and Dorper sheep respectively when the amount offered was raised from 1 to 1.5 of the maintenance requirement. No further increase in DMI-MS was observed when stover was offered to provide twice maintenance. There was no significant (p > .05) difference between the two breeds in DMI-MS. The in vivo digestibilities were consistently (p < .001) higher by about 7 percentage units in the Dorper sheep. The differences are probably not due to a higher selective capacity in the Dorper but may be related to longer rumen retention times in the Dorper. Individual animal variations in intake and digestibility showed some consistency over all five treatments. It may be feasible to identify recordable traits, to estimate genetic parameters and to select animals in order to establish breeding programmes aimed at the more cfficient utilization of crop residues.
INTRODUCTION
Scarcity of land in smallholder farming necessitates for optimized crop-animal integrated systems (Gartner, 1984), giving an important role to crop residues. There has been considerable effort to improve the inherently poor quality of crop residues by chemical upgrading (Sundstol et al., 1984; Preston, 1985). However Greenhalgh (1984) concluded that in many situations chemical upgrading will be superceded by breeding more nutritious straws and by improving their harvesting and supplementation. It has been pointed out (Owen et al., 1989; Chuzaemi et al., 1994) that increased animal production from crop residues could be achieved by exploiting the selective eating behaviour of animals. Animal species and breeds differ in this characteristic (Ledger et al., 1970: Moran et al., 1979) and consistent animal differences have been reported in Indonesian fat tail sheep (de Jong and Van Bruchem, 1993). Small ruminants like sheep and goats are superior in their selective capacity to cattle, whereas larger ruminants have higher fibre digesting capacities (Van Soest, 1994). However, animal differences have been reported for fibre digestibilites in dairy cattle, mainly due to different out flow rates from the rumen (Orskov et al., 1988).
In this study, the crop residue used was maize stover, the most common crop residue in Malawi. The stover was fed to indigenous Malawi and to Dorper sheep, since both breeds are of particular interest to Malawi smallholders. The objective of the study was to examine the variation between such breeds and between individual animals for traits that may contribute to a breeding goal to select animals for more efficient crop residue utilization.
MATERIAL AND METHODS
Laboratory and in vitro analysis
Fifty stalks of maize stover (hybrid variety MH 41) were randomly collected and ground to a 1 mm particle size. Nitrogen and NDF analysis were done according to AOAC (1975) and Van Soest and Robertson (1985). In vitro gas production was measured according to Bmmel and Orskov (1993) and the estimation of the metabolizable energy (ME) was based on the gas volumes and the relationships of Steingass and Menke (1988). The same procedures were followed in the analysis of urea blocks, bean pods and poultry manure.
In vivo experiments
The animals were penned in metabolic cages facilitating urine and faces collection. A total of five treatments (T1 to T5) were examined as follows:
T1-T3: Maize stover was offered to provide 1, 1.5 and 2 times maintenance requirement which was assumed to be 0.440 MJ/kg0-75. A urea lick was given to each treatment. T4 & T5: Maize stover was offered at twice maintenance level together with bean pods (T4) and poultry manure (TS).
An attempt was made to provide at least 1-1.2% of nitrogen in all treatments. Where those requirements were not covered by the feed offered ie T4, the difference was supplied by fertilizer urea nitrogen dissolved in water and sprinkled over the feed.
A minimum of 6 intact male animals per group were used in treatments 1-3 and a minimum of 8 animals in treatments 4 and 5. The local sheep were randomly purchased from village farmers and the Dorper sheep were from breeding stations within the country and were seven eights or more Dorper. Average body weight was 21.0 (±4.7) and 27.2 (±5.2) kg for the locals and Dorpers respectively. The animals were adjusted to the diet for about 3 weeks and then feces and urine were collected for 7 to 10 days. Stover was offered in T1-T3 three times daily, the refusals being removed each time before new stover was given. After weighing the refusals were stored to await fractionation and further analysis at the end of the treatment periods. Urea licks were offered twice a day for 30 minutes in the morning and late afternoon (T1-T3). Bean pods and poultry manure were offered twice daily (T4 and T5).
Statistical analysis
The general linear model procedures of SAS (1988) were used for the statistical evaluation taking into account the possible effects by breed and treatment.
RESULTS AND DISCUSSION The findings reported here are of a preliminary nature but nevertheless support the hypothesis that maize stover could be utilized more efficiently by providing selected animals with a surplus of stover. Stover intake increased from about 30 to 50 g/kg/LW0.75/d when the amount offered was increased from about 90 to 130 g/kgLW0.75/d (Figure 1), allowing the refusals to increase from about 60 to 80 g/kgLW0.75/d. These findings are in general agreement with results reported by Owen et al. (1989) and de Jong and Bruchem (1993) for barley straw and rice straw respectively. No further increase in consumption was observed when 175 g/kgLW0-75/d of stover were offered so giving approximately 120 g/kg/LW0.75/d of refusals. Even though there was no .significant difference (p > .05) in the stover intake between the indigenous and the Dorper sheep, the latter extracted more digestible energy from the stover than the locals (Figure 2). It seems unlikely that variations in the digestibility coefficients were due to a higher selective capacity of the Dorper breed, since the botanical fractionation of the refusals did not reveal any significant compositional differences between the two breeds. However the botanical fractionation is not necessarily conclusive about the nutritive quality. Maize stems which were by far the greatest part of the refusals (> 87%) are not uniform, the pith being of a higher quality than the outer part and some selection may have occurred in chopped and split stems. In vitro analysis is currently under way to investigate possible differences in the nutritive quality value of the refusals from the breeds. The higher dry matter digestibilities in the Dorper sheep were consistent over all five treatments being statistically significant (p < .05), except in treatment four where a dislike of bean pods, which had a higher in vitro digestibility than the stover, was observed in the Dorper.
Higher digestibilites on the same feed are thought to be achieved by longer rumen retention times (Orskov et al., 1988), this hypothesis is being tested using particulate (chromium mordanted fiber) and liquid (Co-EDTA) markers. The analysis is still ongoing and no results can be reported yet. There is however some indirect evidence to support this hypothesis. It can be seen from the results presented in Figure 3, that animal and breed differences in the digestibility coefficients were reasonably consistent over all the five treatments. Interestingly the sheep with the higher digestibilities had relatively lower intakes (Figure 4). Higher digestibilities are probably the result of longer retention times which might influence adversely voluntary feed intake (Thornton and Minson, 1973).
It may be feasible to select species, breeds and finally individual animals which may display more efficient utilization of crop residues. Subsequently, a selection experiment could be conducted to evaluate possible genetic gain. Assuming high heritabilities for relevant traits on crop residue utilization, effective breeding strategies such as nucleus schemes may be employed for the fast dissemination of improved animals. Animal breeders and nutritionists could cooperate to provide improved crop residues to animals more capable of efficient utilization.
REFERENCES
Association of Official Analytical Chemists. (1975) 12th edition. AOAC, Washington, D.C. Blummel, M. and Orskov, E. R. (1993) Anim. Feed. Sci. & Technol. 40: 109-I 19.
Chuzaemi, S, Sulastri R. D., Williams, B. A., Lammers-Wienhoven, T.S.C.W., Chien, X.B., Blummel, M., Zemmelink, G., Prawirokusumu, S., and Van Bruchcm, J. (1995) Anim. Feed
Sci. Technol. (in press). de Jong, R. and van Bruchem, J. (1992) Final Report. Commission of the European Community. Gartner, J. A. (1984): A Systems Approach, FAO, Rome.
Greenhalgh, J. D. F. (1984) In F.M.C Gilchrist and R.I. Mackie (cds), Science press, Pretoria. Ledger, H. P., Rogerson, A. and Freeman, G. H. (1970) Anim. Prod. 12: 425-431. Moran, J. B., Norton, B. W. and Nolan, N. (1979) Aust. J.Agric. Res. 30: 333-340. Orskov, E. R., Oiwang, 1., Reid, G. W. (1988) Anim. Prod. 47: 47-51.
Owen, E., Wahed, R. A., Alimon, R. and El-Naiem, W. (I 989) The Fourth Annual Workshop, Mankon Station, Bamenda, Cameroon.
Preston., T. R., Kossils, VL., Goodwin, J. and Reed. S.B. (1985) FAO Animal Production and Health Paper 50, FAO, Rome.
SAS Inc. (1988) SAS/STAT Version 6.03.SAS Inc. Cary, N. C., USA.
Thornton, R. F., and Minson, D. J. (1973) Austr. J. Agric. Res. 24: 889-98.
Van Soest, J. P. and Robertson, J. B. (1985) A laboratory manual for Animal Science 612. Cornell
University, NY, USA.
Van Soest, P. J. (1994) 0 and B Books Inc-Corvallis, Oregon, USA. THE ROLE OF INDIGENOUS CATTLE BREEDS: ADAPTIVE AND PRODUCTION TRAITS
SUMMARY
INTRODUCTION
ADAPTIVE AND GROWTH TRAITS
CRITICAL ATTRIBUTES OF INDIGENOUS BREEDS IN SOUTHERN AFRICA
IMPLICATIONS IN BREED UTILISATION AND IMPROVEMENT PROGRAMS
CONSERVATION OF INDIGENOUS LIVESTOCK GENETIC RESOURCE
REFERENCES
F. J. C. Swanepoel and L. Setshwaelo Department of Animal Science, University of the Orange Free State, Bloemfontein, 9300, Republic of South Africa and Eastern and Southern African Management Institute, P O Box 3030, Arusha, Tanzania.
SUMMARY
The key environmental factors that influence overall livestock productivity fall within three major areas; the physical environment; biological diversity and the socio-economic environment. The degree to which these factors impact on productivity will differ between the agro-ecological regions and between production systems. The level of production achieved by a particular genotype in harsh environments depends on the contribution and expression of many different traits which may be partitioned into those directly involved with production, and those involved with adaptation. Indigenous breeds in Southern Africa are heat tolerant, are tolerant to some diseases and parasites and are adapted to poor feed resources. While Southern Africa may not be in any immediate danger of loosing breeds, a cohesive national system is required to develop a rational approach to conservation, monitor vital changes in genetic resources and co-ordinate measures and strategies for preservation of breeds deemed at risk.
INTRODUCTION
Livestock production systems are dynamic and continue to evolve over time in response to market and environmental changes. In tropical areas, land and climate are some of the major determinants of the production systems, as potentially high productive ecological-regions are used mainly for crop production leaving the less productive and marginal for livestock production where keeping animals becomes the most efficient and economic way of utilising these areas. Cattle breeds found in these eco-regions have therefore evolved over a millennium where nature has produced genetic constitutions with behavioral, physiological and immunological mechanisms that provide them with the capacity to survive and reproduce despite the various harsh environmental challenges they have to face. These breeds are therefore less prone to physiological breakdown because of environmental stress. Although their potential for production is lower than that of temperate breeds when compared in more favourable production environments, under stressful tropical, arid to semi-arid conditions the reverse has been observed. These breeds have exhibited higher levels of longevity, fitness and overall lifetime productivity than the temperate breeds.
Increasing demands for livestock products due to high population increase and the rise in incomes and living standards in Southern Africa make it absolutely essential that productivity of livestock be increased substantially to meet this demand. To achieve this enormous task, producers have to work within an unfavourable production environment.
Some of the key environmental factors that influence overall livestock productivity fall within three major areas; the physical environment; biological diversity and the socio-economic environment. The degree to which these factors impact on productivity will differ between the agro-ecological regions and between production systems. They will also always deten'nine the types of livestock and genetic constitutions suitable for use in these stress systems. Thus in broad terms they define the production environment.
The Physical Environment
This is a composite of a range of biotic and abiotic factors which interact closely to prescribe the level of production that can be achieved by individual cattle breeds. These factors which include temperatures, diseases, pests and nutrition are, in the arid and semi arid areas, most harsh and unfavourable for production. The nature and degree of their effect on cattle performance and productivity levels will also differ between the agro-ecological regions (the lowlands and the highlands) and cattle genotypes.
These factors, because of their profound influence on the ability of the individual breeds to optimally express their production potential, will cause significant variations in breed performance levels between locations. When designing and planning breeding and genetic improvement programs, these critical factors have to be taken into account. Ideally, a number of technical interventions, such as use of drugs to control diseases and pests and supplementary feeding of animals, could be employed to sufficiently counter or minimise some of the stresses. However, given the socio-economic conditions under which the majority of cattle operations are executed, these opportunities are less feasible.
The Biological Diversity
Genetic constitutions of livestock and the rest of the ecosystems which have evolved under the impact of specific biotic and abiotic stresses found in the different regions, make up what is referred to here as the biological diversity. The indigenous types of cattle found in Southern Africa have therefore developed appropriate mechanisms that ensure their continued survival in the different agro-ecological areas found in the region. These genotypes possess an inherent capacity to sufficiently tolerate the stresses imposed by the physical environment and considerable resilience to persistent droughts that occur in the region. Because of these qualities, they will continue to have a comparative advantage over introduced high potential breeds and have a very crucial role in all the efforts meant to increase livestock productivity through breeding. Careful planning and rationalisation of the breeding strategies will be required to ensure optimal use of their adaptive qualities with due concern given to conserving them for sustained use.
The Socio-economic Environment
This feature refers to the social, financial, technical and infrustructural capacity of individuals or a group of producers to adopt and effectively implement technological innovations within their own cattle production operations. The potential impact of technological innovations depends to a great extent on availability of resources to purchase the necessary inputs and level op development in technical expertise. While commercial cattle producers to some degree, may be in a relatively better position to adopt and effectively utilise some of the livestock production technologies, the capacity of the small producers is limited by lack of technical expertise and resource constraints.
ADAPTIVE AND GROWTH TRAITS
The level of production achieved by a particular genotype in harsh environments depends on the contribution and expression of many different traits which may be partitioned into those directly involved with production, eg food intake and digestive and metabolic efficiency, measured in the absences of environmental stress and those involved with adaptation, eg low maintenance requirement, heat tolerance and parasite and disease resistance (Seifert, 1984).
Frisch (1981) and Vercoe and Frisch (1982) consider growth rate and food intake in the absence of stress as the growth potential of an animal. In the absence of stress voluntary food intake is closely correlated to growth and is highest in Hereford x Shorthorn (Bos torlrus), lowest in Bos indices and intermediate in Bos indicus x Bos taurus.
Stress depresses growth rate primarily through the depression of food intake, but also by affecting digestion and metabolism (Vercoe and Frisch, 1982). In stressful environments animals which are more resistant to stress therefore have higher growth rates and also lower mortalities, and are consequently more productive (Hetzel and Seifert, 1986)..
Frisch and Vercoe (1982) also postulated that selection for growth rate under continued stress would lead to an animal with high adaptation but low growth potential, similar to the Brahman and/or Afrikaner. Alternatively selection under continued low stress would lead to a genotype with high growth potential but low adaptation similar to the Bos taurus breeds.
Bos indicus cattle evolved through natural selection in a stressful environment, including a low plane of nutrition. Selection pressure would therefore have increased gene frequencies for adaptive traits. Because of the low plane of nutrition selection pressure would have operated against growth potential and in favour of a low maintenance requirement (Frisch and Vercoe, 1982). In the Brahman growth potential and adaptation would .therefore appear to be genetically negatively correlated.
British cattle on the other hand have been selected for growth in a relatively stress-free environment on relatively high planes of nutrition. Gene frequencies for growth potential would therefore be increased but those for adaptive traits remain unchanged. If gene frequencies for adaptation in the initial populations were low they would remain low. In addition the increased heat production from high food intake, metabolic rate and the digestive process could be an added advantage in selection for growth potential in a cool climate. In temperate breeds, Hereford and Shorthorn for example, growth potential and adaptation would therefore appear to be genetically negatively correlated.
Selection of British cattle for growth in a stressful environment will exert pressure on those traits limiting growth the most, in this case adaptive traits (Frisch, 1981). Consider the conflict between the need for high heat tolerance and the high heat load associated with high food intake and metabolism. If the gene frequencies for heat tolerance traits, are low and gene frequencies for growth potential high, selection for growth in the presence of heat may initially operate to enhance heat tolerance and against high growth potential. However, when heat tolerance traits improve to the point where they are no longer the most limiting factor, selection would again be directed towards growth potential. The initial change in traits due to selection will be highly dependent on the gene frequencies present in the populations. Thus, provided adaptation is adequate, selection for growth under continued stress, but on a high plane of nutrition, should increase only growth potential. However, it is possible that as growth potential increases with an accommodating increase in heat load, selection pressure may be directed towards further improvements in heat tolerant- traits.
This is adequate practical and genetic reasons for the establishment of productive genotypes from crossbred populations and for the utilisation of indigenous genotypes.
CRITICAL ATTRIBUTES OF INDIGENOUS BREEDS IN SOUTHERN AFRICA
Heat Tolerance
Exceptional challenges faced by livestock in arid and semi-arid environments are numerous, but heat stress is one of the major challenges that animals have to deal with for a longer period of the year. High ambient temperatures outside the thermo - neutral zone cause significant changes in physiological processes including feed intake, due mostly to the direct effects of thermal stress. Differences between breeds in their ability to tolerate high ambient temperatures have been reported in several studies involving Bos indicus and Bos taurus cattle. Bonsma (1949) reported higher heat tolerance in the Afrikaner cattle compared to European breeds. Studies with other tropical breeds have revealed similar observations. Numerous experiments have been conducted to explain this phenomena. Baker et. al. (1994) have attributed this to breed size (mature size) and body weight; the rate of heat dissipation per unit weight has been found to decrease with increase in temperature. This implies that at high ambient temperatures, larger breeds would accumulate in their body more total heat (core body temperature) resulting in higher rectal temperatures than in smaller breeds. Body conformation, coat type and colour, skin thickness and pigmentation, and metabolic rates (and associated heat of production) have also been attributed to heat tolerance of the tropical breeds.
Because of their specialised functions like reproduction, pregnancy and lactation, breeding cows have to efficiently regulate the body temperature. This feature alone is very important when choosing breeds for use as damline under tropical conditions. The relatively low efficiency of thermoregulation in cows of Bos taurus breeds suggests that production would be significantly impaired because of high heat loads. As a result, the comfort zone for these cows may be at much lower ambient temperatures than for Bos indicus types. High rectal temperatures of 40°C in Bos taurus cross cows at about 33°C ambient temperature against 40°C rectal temperatures in Bos indicus types at 39 - 40°C ambient temperatures have been reported. Negative correlations between pregnancy rate and rectal temperatures in cows have also been observed (Vercoe et. al., 1992). These results have significant implications on cow breed types- and reproductive management in the tropics. Most beef herds in the region are bred or breed during the rainy period when range condition is good. Ironically, this is also the period when temperatures are highest. This combination of management practice and climate for lactating, open cows may have a negative effect on reproductive performance of Bos taurus type cows due to increased heat stress. Cows of Bos taurus types in the tropics are therefore more likely to suffer setbacks in reproduction than indigenous types due to thermal stress.
Adaptation to Feed Resources
High daily temperatures and soils in the arid regions influence the type, quality and quantity of forages available for livestock. Seasonal variations in quantity and quality of forage in the rangelands have significant implications on breed's or genotype's suitability to utilise these otherwise unproductive range resources. Differences have been observed between breeds in their ability to select high quality feed from the rangelands. Indigenous cattle have an advantage because of their high selectivity of feed in the rangeland. Again differences observed in performance between tropical and temperate cattle breeds reared in these pastures have been associated, among other things, with their ability to utilize better the poor feed resources. Superiority of the large, faster growing temperate breeds is reduced when nutrition is not high enough to support their high growth rates. The smaller size of the indigenous breeds with lower growth rates and level of milk production (the high potential for milk production has been associated with increased cow maintenance requirements) have also been considered as adaptive attributes to these feed resources. The amount of nutrients required to maintain an animal to undertake day to day physiological processes necessary for survival (such as tissue building and repair) depends among others on body size. The larger the size the higher the requirements. The small size of the indigenous breeds is therefore a result of natural synchronization of the genotype with available feed resources. In these marginal production areas where feed resources are limiting, both in quantity and quality, the large sized breeds are more likely to suffer reproductive impairments than the smaller breeds. Use of large breeds in straight breeding systems, have been found to have no advantage over the smaller breeds in production levels, due to high maintenance requirements. Indigenous breeds as damlines under these sub-optimal conditions are therefore most ideal. Simulation studies demonstrated that where nutrition is limiting, the breeds with lower production potential were more efficient in total herd offtake per unit of energy used.
Disease and Parasite Tolerance
Tick-borne diseases are a major constraint to livestock productivity in Southern Africa. Genetic resistance to some of the diseases reduces substantially, expenditure on disease and parasite control. Quite often the small producers cannot afford to purchase drugs required to control ticks and prophylactic treatment of tick-borne diseases. This problem is further aggravated by development of resistance to some accaricides compounds by ticks which consequently increases the cost of production and hence the retail price of the accancides.
It has been observed that most Bos indicus breeds consistently have lower levels of tick infestations compared to other breeds in the same environment. This ability has been found to be heritable and varies between breed types. It is manifested as the ability of the animal to respond to tick infestations immunologically (Latif, 1992) and the ability of the animal against attaching ticks, keeping the tick populations low. Early observations on breed resistance to ticks were made by Bonsma in the 1940s, which showed that Afrikaner cattle carried fewer ticks .than the imported temperate breeds and had lower mortalities (6% vs 60%) due to heartwater. This led to further research by other scientists on several indigenous breeds in Africa. This phenomena has been reported in some of the breeds found in Southern Africa; the Afrikaner, Bonsmara, Nguni and Nkone. This attribute could be exploited more effectively to reduce the chemical control of ticks in livestock and possible effects on the environment. Small holder operations where cash follow problems are more pronounced depend on natural resistance to ticks and tick-borne diseases, in the indigenous animals. The association between the ticks and the animals has developed a state of endemic stability where parasites are transmitted constantly between livestock and ticks with no clinical signs of the diseases (Latif, 1992). Where crosses and high grade animals with less resistance have been introduced in smallholder operations the problems of tick-borne disease have become significant due to low tolerance in these animals.
Strict control measures are difficult. However, over the past eight to ten decades, control has been largely through acaricide applications and successful eradication of the East Coast Fever in South Africa was achieved by 1954. In commercial enterprises, a breakdown of the dipping routine could result in enormous loses of farm level. However, the major problem faced by the livestock industry is development of resistance by ticks to several acaricide compounds, putting to question the issue of sustainability of this control method, both in economic (increasing the cost of acaricide) and biological terms. Environmental concerns over pollution by dips also make it necessary that the livestock industry looks more seriously at opportunities for integrated control of ticks and tick-borne diseases which include use of genetic resistance and tolerance in livestock.
Disease tolerance has been reported in several of the indigenous cattle as the ability of the breeds to develop tolerance against certain diseases and vector born diseases. Although the mode of tolerance has not been well established in most of the breeds and specific infections, research on trypanotolerance by International Livestock Research Institute (ILRI) has indicated that the genetic tolerance of the N'dama cattle is due to their ability to control paracetemea and anaemia. When little or no control measures for intestinal parasites are practiced, and the level of worm challenge is high, it has been observed that Bos indicus cattle would have lower worm burdens. This has been attributed to the ability of these animals to eject the parasite, thus reducing the level of worm infestation. More research still needs to be done on the inheritance of tolerance in tropical livestock breeds.
IMPLICATIONS IN BREED UTILISATION AND IMPROVEMENT PROGRAMS
The economic state of the majority of the cattle enterprises is such that the production environment cannot be altered sufficiently to suit the high potential temperate breeds and their crosses. As a result, breeders have to word within the limits of the environment to optimise production. Beef cattle production in particular, to a great extent depends on the use op adapted indigenous breeds as the basis for all breed improvement programs and production of commercial stock. Hence it is important to develop and adopt breeding strategies that will not only optimise production in the short run but also ensure continued access and utilisation of these genetic resources.
Southern Africa is well endowed with an abundance of livestock genetic diversity, which provides a unique opportunity to develop more comprehensive breeding programs that can economically increase production under a range of production environments. Together-with the improved types (Afrikaner, Bonsmara, Drakensberg, etc), several strains of the Sanga type cattle have also been identified (Nguni, Shangaan, Venda, Tswana, etc). Significant breed by environment interactions are possible because of the diverse nature of the production systems found in the region. Genotype x environment interactions have been observed where genetically diverse genotypes have been used in different environments. In this regard, the superior performance of Bos indicus over Bos taurus types of cattle has been demonstrated in less favourable environments, and vice versa. Also genetic gains made through selection in one production environment may not necessarily be effective in the other. Genetic manipulation using the existing variability within and between breeds, is still a powerful tool for increasing livestock productivity. Breeding programs should therefore aim at strategic use of both indigenous and exotic breeds to increase production levels under specific production environments. Efficient use of breed comparative advantage and hence complimentary between the breeds may help to offset antagonism between production traits.
Breed Productivity Levels
Performance differences observed between breeds within a given production environment have largely been associated with the breeds' ability to adapt effectively to the physical environment. Performance values of indigenous and exotic cattle breeds have been obtained in a number of breed evaluation and performance testing programs, from both on-farm and experimental stations. Although performance data is not presented in this paper, the information obtained has revealed marked superiority of indigenous breeds under sub-optimal production environments (which are more of a norm within the region).
The indigenous cattle breeds because of their adaptive qualities have always played an important role in the food production systems, by providing meat, milk, draught power and manure, particularly in the smallholder mixed farming system in the region. However, early efforts to increase production levels were bases mainly on breeding strategies that encouraged progressive replacement of local breeds with imported ones, mostly from the temperate regions. Although for a long time there was little interest in the local genotypes, some important initiatives were undertaken by breeders and livestock producers to select within the local breeds to improve performance levels. They also used these breed types in combination with imported breeds to develop some other genotypes suitable for use under local environments. Breeds such as the Afrikaner, Drakensberg, Bonsmara and Tuli; which now play a pivotal role in beef production, are successful outcomes of these credible initiatives. Nonetheless, until a little over a decade ago, for the majority of indigenous cattle breeds in Southern Africa, the attitudes were not so positive. Uncontrolled crossing of these breeds continued on a large scale with little concern over the rate at which genetic erosion was taking placing in these populations, and the likely possibility of their eventual extinction. It is good to report that we have come a long way, attitudes have now changed for the better.
CONSERVATION OF INDIGENOUS LIVESTOCK GENETIC RESOURCE
Not long ago, efforts to increase production levels to meet the increasing demands for meat and milk encouraged replacement of the local breeds with the high potential breeds imported mainly from the temperate regions. At the time little was done to protect the indigenous breeds from eventually disappearing due to outcrossing. Although genetic variation is still believed to be high among these cattle populations, extensive uncontrolled crossbreeding with imported breeds may have rendered some vulnerable as a result of genetic erosion or dilution. A closer assessment is required to determine numbers and identify those populations which may possibly be at risk, in order to take appropriate action before it is too late. Statistics showing current population numbers of the different genotypes is necessary to ascertain the status of these populations. Vulnerability of animal populations is usually determined by the number of available breeding females.
Status Number of breeding females Normal More than 10,000 Insecure 5 - 10,000 Vulnerable 1 - 5,000 Endangered 100 - 1000 Critical Less than 100
The Food and Agricultural Organisation of the United Nations, in collaboration with the (ILRI) are presently working with individual countries to identify and compile breed population statistics.
Extinction may not by a concern for major well established breeds such as the Afrikaner, which are already recognised for their high production potential. For these breeds, institutional machineries responsible for their conservation, improvement, promotion and dissemination are well in place. However, intensive selection may narrow their genetic base and result in the loss of some important genes. Very rarely have conservation programs been initiated purely for preserving genetic variability and endangered breeds. One of the earliest and most notable conservation undertakings reported in the region was implemented by the Government of South Africa around 1932, to collect and preserve the Nguni cattle in a wildlife park, Natal Parks. Later on, through the recommendations of a committee appointed in 1950 to look at the status of the indigenous cattle in South Africa, another conservation project for the then endangered Nguni cattle was initiated in what became known as the Bartlow Combine Project. Conservation of the Nguni cattle and subsequent interest in the breed's genetic qualities, which led to constitution of a breeders' society is indeed a true success story for all concerned with animal genetic resources. The government's initiative to preserve a breed which at the time was seen as having no economic utility is a learning lesson for all of us.
While Southern Africa, hopefully, may not be in any immediate danger of loosing breeds, a cohesive national system is required to develop a rational approach to conservation, monitor vital changes in genetic resources and co-ordinate measures and strategies for preservation of breeds deemed at risk. Such a system will also have the authority to commission reports on current status op populations and mobilise support for conservation activities. The South African Studbook and Livestock Improvement Association may serve as such a co-ordinating mechanism.
REFERENCES
Baker, R.L., and Rege, J.E.O. (1994) Proc. 5th World Congress on Genet. Appl. Livestock Prod.
Guelph, Canada.
Bonsma, J.C. (1949) J. Agric. Sci. 39: 204 - 212.
Frisch, J.E. (1981) J. Agric. Sci. 96: 23 - 28.
Frisch, J.E. and Vercoe, J.E. (1982) Proc. 2nd World Congress on Genet. Appl. Livestock Prod.
Madrid, Spain.
Hetzel, D.J.S. and Seifert, G.W. (1986) Proc.3rd World Congress on Genet. Appl. to Livestock
Prod. Lincoln, Nebraska, USA.
Latif, A. A. (1992) Proc. Workshop on Future of Livestock Industries in East and Southern
Africa. Edited by J.A. Kategile and S. Mubi 119 - 122.
Seifert, G.W. (1984) Proc. 2nd World. Congr. Sheep Beef Cattle Breed. Pretoria, Republic of
South Africa.
Swanepoel, F.J.C., Seifert, G.W. and Lubout, P.C. (1992) Proc. 10th Austr. Assoc. Anim. Breed and Genet., North Rockhampton, Australia.
Vercoe, J.E., and Frisch, J.E. (1992) Austr. J. Anim. Sci. 5: 401 - 409. LINE X ENVIRONMENT INTERACTION FOR FERTILITY IN AFRIKANER CATTLE
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
REFERENCES
M L Beffa Matopos Research Station, P Bag K 5137, Bulawayo, Zimbabwe
SUMMARY
Two closed lines of 100 grade Afrikaner cows each were established in 1957 acid subjected to different nutritional and management regimes, termed environments. The supplemented (S) line was offered protein-rich concentrates at a rate of 0.5 kg per head per day and early mated from December to March, while the non-supplemented (NS) line was late mated from February to May. Selection within line was based on weight of bulls at weaning and weight of heifers at mating. In 1976 the number of cows per line was increased to 150 and two equal sublines were created. One subline remained in the environment of origin and the remaining sublines were interchanged between environments. A line x environment interaction was shown for weaning rate due to the relatively low perforniance (-9.9_2.4 percentage units) of the S subline in the NS environment compared with the NS controls, whereas the line difference in the S environment was 4.3_2.4 percentage units in favour of the NS subline. Cattle from the S line were consistently heavier (1-3%) than their NS counterparts. However, effciencies (calf weight relative to cow weight) were similar. Environmental differences were large mainly reflecting calving season. Results indicate a need to match mature size to the environment for optimum fertility.
INTRODUCTION
Beef production in Zimbabwe is primarily based on the natural range, where nutrition in the dry season, particulary of the breeding female, is limited in both quantity and quality (Elliott and Croft, 1958) and is probably the major cause of the low fertility (60%) experienced in the beef industry.
Emphasis is placed on selecting animals that are heavier at certain ages, for example at weaning, in an attempt obtain higher growth rates. Marketable weights are therefore reached at relatively younger ages. However, growth rate and mature size are positively correlated and there appears to be a negative relation between mature size of the breeding female and fertility. This relation is poorly understood, especially the point at which the size of the breeding cow limits her fertility within an environment.
This paper presents results from a long term selection trial for weaning weight of grade Afrikaner cattle reared in two environments. In the initial stages of the study animals remained in their environment of origin, but since 1976 there has been a system of interchange.
MATERIALS AND METHODS
Matopos Research Station is situated in southwest Zimbabwe, an area primarily suited to semiextensive forms of land utilisation (Vincent and Thomas, 1961). The climate is characterised by wide fluctuations in the quantity and distribution of rainfall within and across season. Mean recorded rainfall is 609mm with a range of 257 to 1376mm. Rainfall normally occurs between November and April and is followed by a long dry season (Ward, et al., 1979).
Two closed lines of 100 grade Afrikaner cows each were established in 1957 from a common gene pool and subjected to different nutritional and management regimes, hereafter termed environments. In the supplemented line (S) cows were offered supplements during the dry season (9kg silage and l kg cottonseed meal per head per day) and were mated to calve early (October to December) relative to the expected onset of the rains. In the non-supplemented line (NS) cows were mated to two months later than those in the S line. A fixed mating season of 90 days and four single-sired herds of 25 cows each was used within each line.
Selection for weight within line was to establish genotypes adapted to their respective environments and was based on weight at weaning for bulls and weight prior to mating (three years) for replacement heifers.
In 1976 the number of breeding females in each line was increased to 150 and two equal sublines were created. One subline Within each environment remained as a control and the remaining sublines were interchanged between environments. The supplementation regime in the S environment was altered so that animals were offered 0.5kg of protein-rich concentrates per head per day during the dry season. Bulls were selected from the control sublines and used within line across environments. Within each subline cows were mated in five single-sired herds, where four bulls were replaced annually with one being retained as a repeat mate. Replacement heifers were generated within subline and entered their respective groups at 27 months of age. Cows were culled for poor production (infertility and low calf weaning weight). Culling and replacement rates were equal across sublines.
Data from animals born in their new environments post-crossover (1980 onwards) were analysed. Weaning rate, relative to the cows exposed to the bull, was measured as zero (no calf) or one (calf weaned) trait. Other traits measured were calf weight at birth, peak wet season weight (taken in May) for weaners and yearlings and weight of heifers at two years of age (December). Cow peak weight was also measured at the end of the wet season (May) and an efficiency index was calculated as weight of calf in May as a percent of peak cow weight.
Data were analysed using General Models Procedure in SAS (1988). Dependant variables were treated as fixed and these included: line (S and NS); environment (S and NS); year of birth, age and previous lactation status (dry or suckling) of the cow; and sex of calf. Since year of birth, age and previous lactation status of the cow were confounded, these effects were combined to form a cell model. Also, age of cow was reclassified to three, four, five to ten and greater than ten years of age to reduce the number of levels with these grouped Year-Age of Dam-Previous Lactation Status effect. Cells with fewer than two observations were deleted, accounting for four observations at most. Orthogonal contrasts were used to test for line differences within environment.
RESULTS
The overall weaning rate for the Afrikaner cows in this experiment was low (Table 1) and the rates in the NS sublines were higher than those of their S counterparts in both environments. A line x environment interaction was shown (p < .05) for weaning rate which was due to the poor performance of the S subline reared in the NS environment compared with the NS controls (-9.9_2.4 percentage units), whereas the line difference in the S environment was only 4.3_2.4 percentage units in favour of the NS subline (Table 1).
A line x environment interaction was also shown for birth weight and appeared to be due to the depressed birth weight of calves from S line cows in the NS environment. The birth weight of calves from the NS line were similar across environment.
Differences between lines for cow and calf weights at different times of the year were small, although, cows from the S line were consistently heavier (1-3%) than their NS counterparts (Table 1). However, there were no differences between lines in terms of efficiency (calf weight relative to cow weight; Table 1). Environmental differences were large, mainly reflecting calving season and hence age of calf and were in favour of the S environment.
Table 1. Weaning rate (%) and cow and calf weights (kg) of two lines of Afrikaner cattle reared in two environments
Line Environment Character Number Mean s.d. cliff' s.e. diflT' S.e. Weaning rate Z 3099 59.7 49.1 Birth weight 2002 31.0 4.8 Calf May weight3 1840 153 33.0 2 1.2 35 1.8 18-month weight 1421 267 35.2 5 1.6 27 1.6 24-month weights 734 267 42.1 2 2.2 36 2.2 Cow peak weighty 3398 464 62.5 12 1.3 17 1.3 Efficiency' 1840 32.2 6.9 -0.6 0.3 6.2 0.3
1. Difference: supplemented mean less non-supplemented mean; line differences for weaning rate and birth weight not given due to presence of line x environment interaction.
2. Expressed as calves weaned relative to cows exposed to the bull. 3. At average ages of 180 and 140 days for the supplemented and non-supplemented environments. 4. Weight in May at average ages of 540 and 500 days for the supplemented and non- supplemented environments. 5. Females only: Weight in December at average ages of 27 and 24 months for the supplemented and non-supplemented environments. 6. Measured in May. 7. Calf May weight as a percent of cow weight.
DISCUSSION
Low levels of fertility of the Afrikaner have been reported in other studies (Moyo, 1990) and have been ascribed to the protracted length of the post-partum anoestrus period in this breed (Hotness et al., 1980). The poor performance of the S line in the NS environment may have arisen as a result of selection in an environment where originally the level of nutrition was very high during the dry season (9kg silage and lkg cottonseed meal), which sustained "larger" cattle and higher levels of production. After the introduction of the crossover phase and a reduction in dry season supplementation (0.5kg concentrates) in the S environment, these levels of performance were no longer sustainable by the larger cattle. Performance levels were reduced further when S line cattle were reared in an NS environment. Although line differences for mature weight were small (15kg for peak cow weight), these represent half a condition score.
Furthermore it is postulated that the S line cattle had larger frames and, although heavier than the NS counterparts, they were in poorer body condition. This could also explain the depressed birth weight of calves from S line cows reared in the NS environment. Similar results have been reported in Australia, where cows selected for high estimated breeding value for fertility tended to be lighter (5%) and progeny grew at slightly lower rates (Hetzel et al., 1989).
These results further support the hypothesis that care should be taken to ensure that mature size of a breeding female must be sustainable in a particular environment if optimum fertility is to be maintained.
ACKNOWLEDGEMENTS
Efforts of past and present staff, paxticulary those employed at Lucydale section, for care and maintenance of animals and collection of data is gratefully acknowledged.
REFERENCES
Elliott, R. C. and Croft., A. G. (1958) Rhod. Agnc. J. 55: 40-49.
Hetzel, D. J. S., Mackinnon, M. J. Dixon, R. and Entwistle, K. W. (1989) Anim. Prod. 49: 73- 81. Holness, D. H., Hale, D.H. and Hopley, J.D.H. (1980) Zim. J. Agric. Res. 18: 3-I 1.
Moyo, S. (1990) ILCA Report, submitted to the Ministry of Agriculture, Zimbabwe. December 1990 130 pp.
SAS. (1988) SAS/STAT User's Guide (Release 6.03). SAS Inst. Inc., Cary, NC.
Vincent V. and Thomas, R. G. (1961) Part 1. Government Printer, Salisbury.
Ward, H. K., Richardson, F. D., Denny, R. P. and Dye, P. T. (1979) Rhod. Agric. J. 76: 5-18. SELECTION FOR IMPROVED REPRODUCTIVE EFFICIENCY IN BEEF CATTLE UNDER TROPICAL AND SUB-TROPICAL ENVIRONMENTS
SELECTION FOR IMPROVED REPRODUCTIVE EFFICIENCY IN BEEF CATTLE
UNDER TROPICAL AND SUB-TROPICAL ENVIRONMENTS
SUMMARY
INTRODUCTION
FEMALE FERTILITY
MALE FERTILITY
RELATIONSHIP BETWEEN MALE AND FEMALE FERTILITY
CONCLUSION
REFERENCES
J. Hoogenboezem and F. J. C. Swanepoel Department of Agriculture, University of Zululand, P Bag X1001, Kwa DlangesN\Ta 3886 Republic of South Africa and Department of Animal Science, University of the Orange Free State, P 0 Box 339, Bloemfontein 9300, Republic of South Africa.
SUMMARY
Reproductive ability is the primary source of all benefits derived from livestock. Indigenous cattle breeds, possessing high gene frequencies for adaptation, play a particular important role in livestock production systems in the tropics. However, it is important to match the cow genotype to available feed resources. A favourable correlation exist between scrotal circumference and puberty in young bulls, as well as between scrotal circumference of bulls and age at puberty of their female progeny. Scrotal circumference is also favourably correlated to days to calving, calving rate and calving interval. Cow fertility can be genetically improved by indirect selection on bull fertility, especially using scrotal circumference as an indicator trait.
INTRODUCTION
Reproductive ability is the primary source of all benefits derived from livestock. However, the relative importance under certain circumstances can be lower. In extended dry periods, and especially during droughts the probability of lactating cows dying is higher than for non- lactating cows. Cows which calve regularly remain in relatively poor condition, and rarely have the opportunity to gain weight during severe droughts, therefore highly fertile cows are "high risk". Non-lactating cows function as insurance and therefore they enhance economic survival. Fertility of beef cattle in Southern Africa is characteristically low mainly as a result of low quantity nutrition during the dry season and limited intake of feed during the hot and humid summer months. Due to the low heritability of fertility traits, it is important to match the cow genotype to available feed resources. In this regard, indigenous cattle breeds, possessing high gene frequencies for adaptation, play a particularly important role in livestock production systems in the tropics.
Fertility might be considered as two traits: inherent fertility and expressed fertility. Inherent fertility refers to the genetic potential for reproductive performance and is not directly measurable. Genes that affect overall physiological and endocrine functions may control inherent fertility and account for the generally favourable genetic relationships of measures of early reproductive fitness with growth, milk and overall productivity. On the other hand, expressed fertility can be measured by age at puberty, quality and quantity of spermatozoa, and conception rate. Expressed fertility is dependent upon the external environment and the environment (additional stress) created by the animal's potential for growth, size or milk production. Under nutritional stress environments the relationship of productivity with expressed fertility may be antagonistic even though the relationship between inherent fertility and productivity may be favourable. Thus, the observed genetic improvement in fertility traits may be less than expected even though the expected genetic progress is being made in inherent fertility.
One might think of improved inherent fertility as a protective mechanism. Under low stress conditions there may be little difference in expressed fertility between herds of moderate versus superior inherent fertility. However , under stress conditions (tropics and sub-tropics) herds with superior inherent fertility may have acceptable expressed fertility compared to those with lower potential. Herds with superior inherent fertility offer more flexibility in a given environment to increase growth and/or milk without sacrificing expressed fertility.
FEMALE FERTILITY
Cow fertility can be measured in a number of ways, though all traits are interrelated (Meyer et al., 1990). These traits can be defined as follows:
calving success whether a cow produces a calf or not at each calving opportunity; number of calves total number of calves over a cow's lifetime; calving rate number of calves per calving opportunity; days to calving number of days between the beginning of joining and of calving; calving interval days between day of birth of successive calves.
Prospects for genetic improvement of fertility have generally been considered to be limited because of the low estimates of heritability in Bos taurus cattle in temperate areas (Dearborn et al., 1993). In these regions, generally with a controlled calving season, most cows calve annually and days to calving best identifies these cows with high inherent fertility. However under African conditions a lower proportion of cows calve annually and mating and calving periods are often extended to the whole year. Thus calving success or number of calves per lifetime may be more useful indicators of fertility under these circumstances.
Heritability estimates for tropical and sub-tropical breeds have been reported to be moderate. (Table 1).
Table 1. Heritability estimates for female reproduction traits for beef cattle under tropical and sub-tropical environments Calving traits Reference Heritability Calving success Seebeck (1973) 25 Turner (1982) 44 Mackinnon et al. (1990) 20 Calving rate Meyer et al. (1990) 17 Number of calves Meyer et al. (1990) 36
Furthermore, there is greater phenotypic variability due to lower population means in these herds. Thus, the total amount of genetic variation may be high which emphasise that progress through selection is possible. Hetzel and Mackinnon (1989) have demonstrated that selection for fertility can be effective. Also the repeatability of cow fertility is moderate (Seifert et al., 1980) and continued culling of cows which fail to calve will increase herd fertility, both phenotypically and genetically (Seifert and Rudder, 1975).
MALE FERTILITY
Reported heritability estimates for bull fertility ranged from 0-0.17 and was highest in purebred Bos indicus cattle (Mackinon et al., 1990; Meyer et al., 1990). Swanepoel et al. (1986) concluded that scrotal circumference is a more reliable parameter for sexual development in Bos indicus than in Bos taurus breeds. Heritability for scrotal circumference is higher, averaging 0.31 (Davis, 1993) and as high as 0.68 (Coulter and Foote, 1979). All the studies indicate that scrotal circumference is a moderate to highly heritable trait (approximately 0.50) and that selection should be very effective in changing scrotal circumference. However Zebu and Zebu-cross bulls in the tropics are more variable in their realised reproduction, than Bos taurus breeds in temperate areas, making evaluation of bull fertility more important in the tropics, than in temperate areas (Davis, 1993). For testosterone concentration heritability estimates were also high, averaging 0,52 (Mackinnon et. al., 1991). Thus greater genetic variation in male reproduction traits is exposed by using measures more closely related to physiological processes associated with reproduction. However, selection for testosterone concentration is expensive and only justifiable under special circumstances. The use of scrotal circumference has the advantage of being able to be measured at a young age and is highly correlated to sperm production (Hoogenboezem and Swanepoel, 1995).
RELATIONSHIP BETWEEN MALE AND FEMALE FERTILITY
Possibilities for genetic improvement of cow fertility is more limited than for bull fertility because h2 estimates for female traits are low to moderate whereas those for male traits are moderate to high hz; selection for cow fertility can only be practiced in females, usually at lower selection intensities; and in bulls one can select for scrotal circumference relatively early in life, whereas in the female only later in life because, to date, no reliable measure of fertility prior to mating exist in heifers. These limitations would be largely overcome if male fertility traits are genetically correlated to female fertility. However, culling for reproductive failure will improve herd fertility (Seifert and Rudder, 1975).
Negative genetic correlations (thus favourable) between scrotal circumference and age at first oestrus (-0.39) and standardised age at first oestrus (-0.19) were reported by Morris et al. (1992). Toele and Robinson (1988), reported correlations of scrotal circumference with days to calving of 0.35, with calving rate of 0,62, with age at puberty of -0.55 and with calving interval of 0.66. These very strong relationships, together with a correlation of 0.98 between scrotal circumference and age at puberty in heifers indicate that these two traits are essentially the same.
From these results it is clear that cow and bull fertility are favourably genetically correlated and that cow fertility can therefore genetically be improved by indirect selection on bull fertility. Generally, selection for increased testicle size would lead to improvement in cow reproduction, particularly calving rate, age at puberty and calving interval. Furthermore, the fact that the correlations (Toele and Robinson, 1985) are equally reliable when scrotal circumference was measured at 7 months or 12 months, again stressed the benefit of early selection in male animals. This is substantiated by Swanepoel et al. (1992) with their conclusion that male traits such as scrotal circumference can be used to speed up genetic progress in cow fertility.
CONCLUSION
Reproductive ability is the primary source of all benefits derived from livestock. Indigenous cattle breeds, possessing high gene frequencies for adaptation, play a particular important role in livestock production systems in the tropics. However, it is important to match the cow genotype to available feed resources. Cow fertility can be measured in terms of calving success, number of calves per cow per lifetime, calving rate, days to calving and calving interval. Prospects of genetic improvement of fertility in Bos taurus breeds are considered to be limited, because of the low estimates of heritabilities. However, heritability estimates for tropical and sub-tropical .adapted breeds have been reported to be moderate.
Scrotal circumference is a moderate to highly heritable trait. Therefore selection should be effective in changing scrotal circumference. The use of scrotal circumference as a selection trait has the advantage of being able to be measured at a young age and it is highly correlated to sperm production. A favourable correlation exist between scrotal circumference and puberty in young bulls, as well as between scrotal circumference of bulls and age at puberty of their female progeny. Scrotal circumference is also favourably correlated to days to calving, calving rate and calving interval. Cow fertility can be genetically improved by indirect selection on bull fertility, especially using scrotal circumference as an indicator trait. .
REFERENCES
Coulter, G.H. and Foote, R.H. (1979) Theriogenology 11: 297. Davis, G.P. (1993) Austr. J. Agric. Res. 44: 179.
Deabom, D.D., Koch, R.M., Cundiff, L.V., Gregory, K.E. and Dickerson, G.E. (1973) J. Anim. Sci. 36: 1032.
Hetzel, D.J.S. and Mackinnon, M.J. (1989) Proc.Jst Nat. Conf of the Beef Improvement Ass. Armidale, Australia.
Hoogcnboczem, J.M and Swanepoel, FIC. (1995) Proc. Austr. Soc. Anim. Prod. (in press).
Mackinnon, M.J., Taylor, J.F. and Hetzel, D.V.S. (1990) J. Anim. Sci. 68: 1208
Mackinnon, M.J., Corbet, N.J., Meyer, K., Burrow, H.M., Bryan, R.P. and Hetzel, D.J.S. (1991) Livest. Prod. Sci., 29: 297.
Meyer, K., Hammond, K., Parnell, P.F., Mackinnon, M.J. and Sivarajasingam, S. (1990) Livest. Prod. Sci. 25: 15.
Morris, C.A., Baker, R.L. and Cullen, N.G., (1992) Livest. Prod. Sci. 31: 221
Seebeck, R.M. (1973) J. Agric.Sci. 81: 253-258.
Seifert, G.W., Bean, K.G. and Christensen, H.R. (1980) Proc. Austr. Soc. Anim. Prod., 13:62.
Seifert, G.W. and Rudder, T.H. (1975) Proc. 23rdEaster School in Agricultural Science. Eds: Brester, W.H. and Swan, H., Nottingham, Butterworths, London., UK.
Swanepoel, F.J.C., Seifert, G.W. and Lubout, P.C. (1992) Proc. 10th Austr. Ass. Anim. Breed and Gen. North Rockhampton.
Swanepoel, FIC., Venter, H.A.W., Van Zyl, J.G.E. and Heyns, H. (1986) Proc. 3rd World Congr. Genet. Appl. Livestock Prod. Lincoln, Nebraska, USA.
Toele, V. D. and Robinson, OW. (1985) J. Anim. Sci. 60: 89-100.
Turner, H.G. (1982) Anim. Prod. 35: 401- 412. EVALUATION OF BREEDS FOR BEEF PRODUCTION IN ZIMBABWE
SUMMARY
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSION
REFERENCES
S. Moyo Matopos Research Station, P Bag K 5137, Bulawayo, Zimbabwe.
SUMMARY
Results from a breed evaluation study where Zimbabwe's indigenous cattle (Mashona, Nkone and Tuli) were compared with the other beef breeds were assessed. The indigenous cattle were considerably more productive than exotic types mainly due to their high reproductive rats. Progeny with Brahman breeding were the heaviest at weaning and at 18 months. Feedlot performance and carcass traits indicated that progeny with Bos taurus genes were superior to their Sanga and Zebu counterparts.
INTRODUCTION
Breed characterisation for the major traits of economic importance will play an important role in beef production strategies. This paper presents results from a breed evaluation study in Zimbabwe from which utilisation strategies are proposed to address this symposium's theme of improving livestock production through animal breeding and genetics.
MATERIALS AND METHODS
The performance of a wide range of purebred and crossbred cow genotypes has been recorded at Matopos Research station since 1974. The station is situated in the semi arid regions of Zimbabwe where the average annual rainfall is 609 mm and most of it falls between October and March but both amount and distribution are highly variable. The project began with the purchase of foundation heifers and bulls throughout the country to generate pure and crossbred genotypes which were subsequently evaluated as breeding cows. The cow genotypes in the studv comprised Mashona, Tuli, Nkone (indigenous), Afrikaner, all these being Sanga type, Brahman (Zebu), Charolais and Sussex (Bos taurus) and several resultant crosses amongst them. These were mated annually from 1979 to the following seven terminal sire breeds; Aberdeen Angus, Afrikaner, Brahman, Hereford, Holstein - Friesian and Simmental. Management practices were the same for all genotypes. Single sire mating in a fixed breeding season starting mid-December to mid-March was practised. No selection of females was undertaken. All animals were weighed monthly. Calves were weaned at an average age of eight months, after which they grazed on rangeland, grouped in separate sexes, until 18 months of age when they entered a feedlot for fattening prior to slaughter at an average age of 22 months. Animals were rotationally grazed on rangeland at a stocking rate of approximately five hectares per livestock unit (500 kg). In the dry season, cows were offered a winter supplement. From calving until regrowth of grass started cows were given 2 kg per day of maize grain. A phosphate salt lick was available at all times, together with bone meal.
The General Linear Models GLM procedure of Statistical Analysis Systems (SAS, 1987) was used in the analyses. Other than the few analyses of calf growth and carcass traits made specifically to compare calf genotypes, reproductive performance, growth traits and productivity indices were analysed as traits of the dam.
RESULTS
The main traits of economic importance that contribute to an efficient and profitable enterprise in a beef production system include: fertility; survival and growth of progeny; and carcass quality. In this study reproductive performance was estimated using calving rate. Among the purebred cows the most fertile cows were Mashona, followed by Tull, Nkone, Brahman and Charolais while Sussex and Afrikaner were the least fertile (Table 1). Charolais X Brahman cows were the most fertile among the crossbreeds. There were favourable heterosis estimates for cow reproductive performance in most crosses (Table 2). In particular, highly significant (p < .01) estimates for calving rate were obtained for Afrikaner crosses with the Sussex (8.7°/,) and Charolais breeds (7.8%).
Table 1. Comparative productivity of beef breeds evaluated in Zimbabwe (1978 - 1989) (For crossbred cows sire breed listed first)
Cow No Cow Calving Weaning 18 Productivity indices genotype of weight, rate, % weight, months cows kg kg weight, kg
per cow mateda per unit weight cowb Weaning 18 Weaning 18 months months Mashona 253 351 77 176 261 153 235 206 346 (M) Nkone(N) 147 388 71 188 279 160 246 203 335 Tuli(T) 349 400 72 187 275 160 240 193 318 Afrikaner 447 398 59 189 273 138 208 169 281 (A) Brahman 106 440 71 207 299 166 248 197 311 (B) Sussex (S) 134 418 59 180 275 139 222 162 281 Charolais 116 480 67 188 276 160 241 172 284 (C) AM, MA 249 380 69 184 267 152 233 196 322 AN, NA 216 389 67 192 278 156 233 195 322 AB, BA 257 436 65 205 293 162 238 192 300 AS, SA 470 439 68 199 288 165 247 189 306 CA 119 480 72 209 295 175 255 187 302 SB 147 432 75 202 291 168 250 196 315 CB 197 478 77 208 298 183 268 200 313
CS 140 460 67 193 283 164 248 181 298 a) Weight of calf per cow exposed to the bull b) Weight of calfper 100 kg live weight .73 of cow exposed to the bull
Table 2. Estimates of hcterosis (and s.e.) for cow reproductive performance and production traits
a Trait F1 genotype Calving Annual Production of rate, % Weaner calf per unit 18 month calf per unit weight of cow weight of cow Afrikaner (A) crosses AM 0.9 (2.46) 7.8 (2.46) * 10.7 (5.21). AN 1.8 (2.68) * 9.6 (3.06) * 12.9 (5.87) AB 1.0 (2.63) 9.6 (3.03) * 10.6 (5.86) AS 8.7 (2.04) ** 23.9 (4.01) ** 28.5 (4.68) ** Charolais (C) crosses CA 7.8 (3.42) * 15.7 (5.23) ** 17.3 (7.32) ** CB 6.7 (3.7) * 14.3 (4.92) ** 16.1 (6.84) CS 2.4 (2.34) 16.9 (5.86) ** 12.7 (7.31) CM -8.4 (2.07) * 8.3 (3.01) 3.7 (1.24) CN 1.3 (0.0) 7.0 (2.23) 0.2 (0.13) Sussex (S) crosses SM 5.8 (1.90) * 13.6 (4.23) ** 13.9 (4.56) SN 4.3 (1.09) * 9.9 (3.13) 11.5 (3.84) SB 9.3 (2.47) ** 16.6 (5.03) ** 21.4 (7.35) a First letter represents sire breed; M = Mashona; N = Nkone; B = Brahman
*p<.05
**p < .01
Calves from Brahman, F1 Charolais X Brahman, Charolais X Afrikaner, and reciprocals Afrikaner X Brahman and Brahman X Afrikaner (pooled) were the heaviest at weaning and at 18 months, while calves from Mashona cows were the lightest at both stages (Table 1). Results on the production of weaned and 18 month calf per cow per year showed that the most productive cows were F1 Charolais X Brahman, Charolais X Afrikaner, Sussex X Brahman and purebred Brahman. Productivity per 100 kg metabolic body weight of cow per year was highest in Mashona cows followed by F1 Charolais X Brahman.
The other indigenous cattle breeds (Nkone and Tuli) also ranked highly in productivity indices (Table 1). There was a favourable and significant heterosis estimate in F Afrikaner X Sussex 1 for production of weaned and 18 month calf per 100 kg metabolic body weight of cow (Table 2).
Results of daily feed intake, daily gains and feed conversion ratio showed that within feed level calf genotypes with Tuli, Afrikaner and Mashona genes a poorer performance than their Bos taurus counterparts especially, Hereford,. Aberdeen Angus and Simmental (Table 3). Within each feed level, the lightest. and shortest carcasses were from purebred Afrikaner, Tuli and their crosses with Mashona, while Brahman and Charolais crosses had the heaviest and longest carcasses. The lighter and shorter carcasses tended to have poorer fleshing indices. There was no distinct pattern in the dressing percentage of the different genotypes (Table 4). Subcutaneous fat (as indicated by fat over the muscle longissimus area) was higher in carcasses of Aberdeen Angus and Hereford crosses than in that of other genotypes. The greatest eye muscle areas were from Simmental and Aberdeen Angus crosses and some Tuli crosses while purebred Afrikaner and some Mashona crosses had the smallest areas. There were small differences between breeds in the degree of marbling (Table 4).
Table 3. Feedlot performance of steers (s.e.) of different genotypes allocated 1140 kg of high energy diet (air fresh and estimated dry matter 90 per cent)
senotypea Days in Feed Average daily Feed pens intake, kg gain, g/day conversion ratio, kg/kg AA 112 (1.6) 10.3 (0.17) 1022 (88) 10.1 (0.80) AM 111 (1.5) 10.3 (0.16) 1212 (84) 8.5 (0.77) BM 100 (1.9) 11.5 (0.20) 1127 (109) 10.2 (0.92) TM 115 (2.2) 9.9 (0.24) 1012 (121) 9,8 (1.11) HM 95 (3.3) 10.4 (0.36) 1630 (181) 7.4 (1.66) AaM 107 (2.4) 10.6 (0.26) 1626 (156) 6.5 (1.43) FM 98 (2.4) 11.9 (0.26) 1201 (130) 9.9 (1.20) Sim 103 (1.8) 11.1 (0.20) 1437 (100) 7.7 (0.92) TC 114 (2.0) 10.0 (0.22) 1314 (108) 7.6 (0.99) AaC 111 (3.3) 11.3 (0.36) 1899 (220) 5.4 (2.02) aA = Afrikaner; M = Mashona; T = Tuli; B = Brahman; F = Holstein-Friesian; H = Hereford; Aa = Aberdeen Angus, C = Charolais; SIM = Simmental
Table 4. Carcass characteristics (s.e.) of steers of different genotypes slaughtered after pen fattening
Genotype Hot carcass Carcass Dressing Fat over Muscle weight, kg length, cm % muscle longissimus longissimus area, cm2 area, mm AA 225 (7.9) 119 (1.1) 56 (0.8) 9.8 (1.51) 64.1 (3.00) AM 239 (7.6) 118 (1.1) 57 (0.8) 12.9 (1.62) 65.8.(3.21) BM 239 (9.3) 120 (l.3) 57 (0.9) 13.8 (1.63) 62.9 (4.25) TM 218 (10.8) 115 (1.5) 57 (l.l) 12.9 (1.79) 66.9.(4.09) HM 260 (16.5) 119 (2.3) 58 (1.7) 15.9 (2.74) 66.7 (5.44) AaM 251 (14.0) 119 (1.9) 55 (1.4) 14.8 (2.66) 73.9 (8.83) FM 236 (12.0) 121 (1.6) 55 (1.2) 7.6 (1.98) 59.5 (6.55) Sim 258 (9.3) 123 (1.3) 54 (0.9) 10.2 (1.59) 70.8 (4.33) TC 254 (9.8) 120 (1.4) 56 (2.4) 8.7 (2.41) 77.2 (3.59) AaC 283 (19.9) 120 (2.7) 56 (2.0) 20.5 (4.56) 74.9 (8.96) aA = Afrikaner; M = Mashona; T = Tuli; B = Brahman; F = Holstein-Friesian; H = Hereford; Aa = Aberdeen Angus; C = Charolais; SIM = Simmental
DISCUSSION
The potential of increasing both the biological and economic efficiency of a beef cattle enterprise can be enhanced by high reproductive rates through the reduction of breeding herd costs per meat animal marketed (Dickerson, 1970; Bourdon and Brinks, 1987). The poor performance of the Afrikaner and its crosses in this study is consistent with other studies of this breed (Holness, Hale and Hoplcy 1980; APRU, 1981; Hetzel, 1988). Hcyns (1992) has singled out high inbreeding levels, deliberate selection against fertility, and lactation anoestrus as some of the reasons for the low fertility in the Afrikaner cow. Seen against the low fertility of the purebred Afrikaner, the favourable heterosis responses in Afrikaner crosses indicate the potential advantage in using these crosses over the purebred Afrikaner for better reproductive performance. hlowever, the hcterosis estimate for calving rate in Afrikaner crosses while significant, may be of(sct by the cost of maintaining the less efficiently reproductive parents (such as Sussex and Afrikaner) to generate the cross when there are other more efficient genotypes in the national cow population to choose from.
The superior reproductive performance of the indigenous Sanga cattle in this study was consistent with other beef breed evaluation studies in the region (APRU, 1981; Scholtz, 1988; Dioniso and Systrad, 1990; Vilakati, 1990). This is thought to be mainly due to tlwir adaptational characteristics to the tropical environments. Natural selection has resulted in breeds that are able to survive and reproduce under these harsh environments. Thus, the Bos taurus breeds which are unadapted to an environment where feed and water arc limited arc likely to suffer more reproductive failures when compared to the indigenous cattle. The relatively high fertility of the Brahman in this study conflicts with most reports world-wide (Cartwright, 1980; Franke, 1980; Turner, 1980) and the region where it was found to be intermediate in South Africa, but relatively poor in Swaziland and Botswana when compared to indigenous Sanga and Bos taurus breeds (APRU, 1981; Scholtz, 1988; Vilakati, 1990).
Growth performance of calves is a function of the combined effects of additive maternal, additive direct genetic, and heterosis effects. Calves from purebred Brahman and some Brahman and Charolais crosses were superior to other genotypes on the basis of wean, nf,JI ;gird 18 month weights. These results are consistent with the known positive correlation between early (eg weaning) weight with subsequent (postweaning) weight. Choice of a breed will be influenced by interdependent components of a beef production system. It is therefore important that measures of efficiency and profitability of the system should consider all aspects of the production cycle simultaneously instead of putting the emphasis on isolated traits.
Cow productivity has been expressed in the form of indices, with weight of calf as the main output (Newman and Deland, 1991; Morris, Baker, Cullen, Hickey and Wilson, 1993). Indices that express weight of calf per unit weight of cow exposed have been proposed as a fairer comparison between cows with different fertilities and of different sizes (Scholtz, 1988; Tawonezvi, Ward, Trail and Light 1988, Van Zyl, Schoeman and Coertze 1 992; Meaker, 1993). In a further attempt to refine the indices and include maintenance requirements, productivity has been expressed as weight of calf per unit cow metabolic body weight (Roux and Scholtz, 1984). Despite the lower weaner and 18 month weights of their progeny, indigenous Sanga cattle were among the most productive genotypes. The high reproductive rates, smaller mature cow sizes and the low maintenance needs on rangeland are some of the factors responsible for the high overall productivity of indigenous breeds.
These results emphasise the role of indigenous breeds as dam lines under ranching conditions where natural rangeland is the major resource for production. The use of a dam line of a smaller body size, is advantageous in reducing the cow maintenance costs due to lower feed requirements.
Some genotypes might play a better role in a system as crossbred cows, This is the case with the Afrikaner, Sussex and Charolais breeds. These results indicate that the potential for increasing cow productivity through exploitation of heterosis in crossbred cows can be considerable, especially in breeds of diverse genetic origin. The present findings are consistent with the theoretical expectation of greater heterosis in more genetically diverse breed types and has been reported in other studies (Peacock, Koger, Olson and Crockett, 1981; Cundiff', Gregory, Koch and Dickerson, 1986; Kress, Doornbos and Anderson, 1990; Schoeman, van Zy) and de Wet, 1993).
Direct comparisons of the results on feedlot performance and carcass char;tctcristics with those published in literature are limited because breed type, diet and the criteria of evaluation differ widely across studies. The higher intakes or appetite of the Bos taurus crosses compared to that of Sanga or Zebu crosses could be due to inherent characteristics related to nutritional adaptation (Moore, Essig and Smithson, 1975; Hunter and Siebert, 1985). Sanga and Zebu type cattle originate in environments where natural selection was on rangeland and thus, are better adapted to a highly variable (quality and quantity) tropical feed supply than their Bos taurus counterparts which are better adapted to intensive systems of finishing. Feed conversion ratios indicated the superiority of Bos taurus animals in the pens. This was not surprising since they ate more and grew faster than their Bos indicus counterparts.
Results of carcass analyses indicated that calf genotypes with Bos taurus genes were superior to Tuli, Mashona, Afrikaner, Brahman, and their crosses. Fuller evaluation of carcasses could have been obtained if this work had been extended to estimate carcass composition since ranking of breeds has been shown to change due to differences in maturation patterns (Urick, Macneil and Reynolds, 1991). The weakness of the grading system further complicated the interpretation of the results since it consistently favoured the larger framed breeds. In addition the larger genotypes had an added advantage due to the system of finishing (feedlot - high energy diet). Further research to compare these breeds under different systems of finishing and at various composition end points is warranted.
CONCLUSION
The results of the present study have shown major differences among breeds for traits of economic importance. It is clear that no single breed is suitable for all important components of the beef production cycle. The indigenous Sanga dams have a potential role to play as dam lines, but feedlot finishing and carcass characteristics favoured calves with exotic Bos taurus genes. This introduces genetic antagonisms which have been observed and discussed in other similar breed evaluation studies. There is a need for appropriate breed utilisation strategies to achieve trade-offs between genetic antagonisms. The information in this study is valuable to each of the three sectors of this country's livestock industry. The small holder farmer who has a low input system of production would benefit from using the indigenous Sanga type animal which is predominately found in communal areas. There would be no benefit in bringing in exotic breeds until the level of management and nutrition has been improved. Secondly, the small holder farmer's priority is not meat production, but draught power, manure and milk production rank very highly. An extension message on the potential role of indigenous breeds should be strengthened.
The small and large scale commercial or resettlement farmer could also choose tile indigenous breeds but, has room for strategic management of the trade-offs between genetic antagonisms through crossbreeding systems. In these systems the indigenous breeds are reconnnended for use as dam lines. The choice of sire breed will depend on whether the system is rangeland based or has facilities to pen fatten. This system will complement the growth potential from the sire breed with the high fertility and low maintenance requirement in the indigenous cow. Further more it is necessary that the feed base should sustain the high growth rates in the cross. Availability of purebred stock for crossbreeding will also influence the choice of breed or genotype.
The final choice of breed will depend on management and feeding systems. It is important to always remember that the primary objective of a farming operation is to improve profitability. Thus, data analysis to examine whether the small-bodied indigenous breeds are more economical than exotics under range conditions, ie taking into account actual costs of inputs and product prices are necessary. The high productivity of the indigenous breeds has been attributed to their adaptability. There is a need therefore to quantify the adaptive traits of the various breeds (eg disease resistance, water economy and grazing behaviour) in order to best harness these from the superior breeds.
REFERENCES
Animal Production Research Unit. (1981) Ministry of Agriculture, Gaberone, Botswana. Bourdon, R.M. and Brinks, J.S. (1987) J. Anim. Sci. 65 (4):956 - 962. Cartwright, T.C. (1980) J. Alum. Sci. 50 : 1206 - 1214.
Cundiff, L.V., Gregory, K.E., Koch, R.M., and Dickerson, G.E. (1986) Proc. 3rd World Congr. Gen. Appl. Livestock Prod. Lincoln, Nebraska.
Dckerson, G.E. (1970) J. Anim. Sci. 30 : 849 - 859.
Dioniso, A.C. and Systrad, 0. (1990) Livest. Prod. Sci. 24 : 29 - 36.
Franke, D.E. (1980) J. Amm. Sci. 50 (6): 1206 - 1214.
Hetzel, D.J.S. (1988) Anim. Breed. Abstr. 56 (4): 243 - 255.
Heyns, H. (1992) DSc. Thesis, University of Fort Hare, Republic of South Africa.
Holness, D.H., Hale, D.H., and Hopley, J.D.H. (1980) Zimbabwe J. Agric. Res. 18 : 3 - 11.
Hunter, R.A. and Siebert, B.D. (1985) British J. Nutrition. 53 : 637 - 648. .
Kress, D.D., Doornbos, D.E., and Anderson, D.C. (1990) J. Anim. Sci. 68 (1) 54 - 63.
Meaker, M.J. (1993) MSc Thesis. University of Pretoria, Republic of South Africa. Moore, R.L., Essig, H.W., and Smithson, L.J. (1975) J. Amm. Sci. 41 (1) : 203 -207.
Morris, C.A., Baker, R.L., Cullen, N.G., Hickey, S.M., and Wilson, J.A. (1993) Amm. Prod. 57: 29-36.
Newman, S., and Deland, M.P. (1991) Austr. J. Exp. Agric. 31 : 293 - 300.
Peacock, F.M., Koger, M., Olson, T.A., and Crockett, J.R. (1981) J. Anim. Sci. 52 : 1007 - 1013. Roux, C.Z. and Scholtz, M.M. (1984) Proc. 2nd World Congr. Sheep and Beef cattle breeds, Pretoria, Republic of South Africa.
Schoeman, S.J., Van Zyl, J.G.E., and De Wet, Rensia. (1993) S. Afr. J. Amm. Sci. 23 : (3/4) 6166.
Scholtz, M.M. (1988) Proc. 3rd World Congr. Sheep and Beef cattle breeds. Pretoria, Republic of South Africa.
Statistical Analysis Systems, (1987) SAS/STAT. Version 6 Edition Cary, NC/SAS Institute Inc., USA.
Tawonezvi, H.P.R., Ward, H.K., Trail, J.C.M., and Light, D. (1988) Anim. Prod. 47 : 351 -359.
Turner, J.W. (1980) J. Amm. Sci. 50:1201 - 1205.
Urick, J.J., MacNeil, M.D., and Reynolds, W.L. (1991) J. Anim. Sci. 69 (2):490 - 497.
Van Zyl, J.G.E., Schoeman, S.J., and Coertze, R.J. (1992) S. Afr. J. Amm. Sci. 22 (5) 166 - 169.
Vilakati, D.D. (1990) Ministry of Agriculture and Cooperatives, Mbabane, Swaziland. LIVESTOCK IMPROVEMENT IN SOUTH AFRICA: PERFORMANCE DRIVEN
SUMMARY
INTRODUCTION
ACKNOWLEDGEMENTS
J. P. Campher S A Stud Book and Livestock Improvement Association, P 0 Box 270, Bloemfontein, Republic of South Africa.
SUMMARY
The South African Stud Book and Livestock Improvement Association was founded in 1905. Since then, it has generated the vision of establishing and developing superior farm animals as a permanent possession and renewable resource for the peoples of South Africa. It is an independent nongovernmental organisation. The considerable progress achieved by the South Afi-ican livestock producer may further be ascribed to the concerted effort which several institutions and organisations have made over the years in the development of the renowned South African Livestock Industry. These formal structures include S A Stud Book's Veedata division with its Integrated Computer System, the Livestock Breeders' Societies, the Livestock Improvement Schemes managed by the Agricultural Research Council, the three Marketing Boards involved with animal agriculture, the AI industry, the Registrar of Livestock Improvement with the Livestock Improvement Act he administers, and the Advisory Board for Animal Production. These organisations and institutions have all formed a powerful base from which the South Afiican Livestock Producer derives a unique service. The mission of these role players in animal production is to create a well organised and scientifically structured corporate Livestock Breeding and Improvement facility for the Afiican continent.
INTRODUCTION
Mankind has survived and prospered on the fruits of this planet of which livestock contributed by far the more varied and interesting food and clothing resources. In his quest for breeding superior animals, man as early as Jacob has manipulated breeding patterns of sheep to further his interests. However, it was only towards the end of the 18th century that the first formally kept studbook, that for Thoroughbred horses, was founded. It has been estimated that should the history of the universe be compared to a period of one single year, our first attempts at formal breeding strategies started in the last second before nudnight on the 31st day of December. Indeed a very short time, to achieve so much.
Since the 18th century, formal breeding strategies and studbooks world-wide have come a long way and in breeding numerous new breeds to supply in the needs of the producer and the consumer. Also in Southern Africa today we have a rich heritage of species and breeds of livestock.
The South African Stud Book and Livestock Improvement Association, known internationally as S A Stud Book and whom I represent here today, has since its founding in 1905 generated the vision of establishing and developing superior farm animals as a permanent possession and renewable resource for the peoples of South Africa. S A Stud Book as an independent non-governmental organisation and is well known beyond South Africa's borders and thus bears ample proof of our success. S A Stud Book's pedigree certificates have become the hallmark of purebred stock in the subregion.
For most breeds, active incorporated livestock breeders' societies have been formed and are in terms of the Livestock Improvement Act of 1977, associated with S A Stud Book as its members. South Africa has to date recognised 33 cattle breeds in a population of approximately 10 million head. Beef breeds form the major number and may be grouped in the following classification.
Sanga/Zebu breeds Synthetic breeds European breeds Afrikaner Bonsmara Brown Swiss Boran Beefmaster Charolais Brahman Brangus Deutches Rotvieh Nguni Drakensberger Gelbvich Tull Romagnola Limousin British breeds Sanganer Pinzgaucr Aberdeen Angus Santa Gertrudis Simmentaler Dexter Kerry Simbra Galloway Hereford Red Poll Shorthorn South Devon Sussex Dairy breeds comprise mainly of the following: Ayrshire Guernsey Holstein Friesland Jersey S A Dairy Swiss Of more than eighteen sheep breeds found in South Africa the following are most common: Composite breeds Indigenous, Pelt producing and other breeds Afrino Blackhead Persian Dormer Damara Dorper Karakul Vandor Wool producing breeds Wool - Mutton breeds Döhne Merino Corriedale Merino. Ile de France Merino Landsheep S A Mutton Merino Mutton - Wool breeds Border Leicester; Hampshire and the Suffolk. Goat breeds include milk and meat types of which the following are commonly found: Meat breeds Milk breeds Boergoat Saanen Indigenousgoat British Alpine Toggenberger Mohair breed Angora Pig breeds of commercial significance in South Africa are the: S A Large White S A Landrace Duroc Hampshire Chester White
A number of new composite pig breeds to be utilised in terminal crosses have also been introduced into South Africa. We have even developed our own terminal line, known as the Robuster.
Some twenty six horse breeds are also bred in South Africa and our equestrian industry is rated amongst the best in the world. The performance of many South African race horses in stud breeding overseas and also the achievements of our show jumpers prove the point. Recently, the ostrich industry also formally entered the fold of livestock improvement and its associated structures by forming a breeders' society and developing an own performance testing scheme.
All of these species and breeds are bred to be well adapted to the African environment and can play a major role in other countries on the Continent and overseas. Proven genetic material of veterinary sound registered stud animals is always available and has already played a significant role in establishing a booming export industry. The world recently took notice of our achievements and many embryo collecting centres have been established all over South Africa. The many achievements of our AI industry (Taurus) are also well known in the subregion.
The considerable progress achieved by the South African livestock producer may further be ascribed to the concerted effort which several institutions and organisations have made over the years in the development of the renowned South African Livestock Industry. These formal structures include S A Stud Book's Veedata division with its Integrated Computer System, the Livestock Breeders' Societies, the Livestock Improvement Schemes managed by the Agricultural Research Council, the three Marketing Boards involved with animal agriculture, the AI industry, the Registrar of Livestock Improvement with the Livestock Improvement Act he administers, and the Advisory Board for Animal Production. These organisations and institutions have all formed a powerful base from which the South African Livestock Producer derives a unique service. The mission of these role players in animal production is to create a well organised and scientifically structured corporate Livestock Breeding and Improvement facility for the African continent.
By doing so they are able to identify superior performing animals through S A Stud Book's data processing facility, Veedata and with the up to date scientific backing of the five Livestock Improvement Schemes, advance the exchange of stud stock, semen and embryos of genetically sound and healthy animals, co-ordinate all organisational matters pertaining to livestock improvement, represent breeders and their interests in corporate and legislative organisations, facilitate communication with international bodies such as the International Committee for Animal Recording (ICAR) and the International Livestock Research Institute (ILRI), and importantly, ensure the conservation and promotion of valuable indigenous breeds on a continental level, thereby retaining Africa's genetic biodiversity.
Animals registered with S A Stud Book and/or participating in the Livestock Improvement Schemes follow a logical and user friendly set of steps to arrive at a trusted product of genetic and functional superiority.
The first prerequisite of livestock improvement is a tamper proof marking system, followed by registration of the birth and parentage of the animal. In the case of stud animals the appropriate breeders' society performs a key function of verification and subsequent inspection for breed purity and functional soundness.
S A Stud Book, which is centrally situated in Bloemfontein, through its Veedata division, operates a powerful Integrated Computer System which integrates the certified pedigree of an animal with its' performance data obtained from one of the five Livestock Improvement Schemes presently operating in South Africa. This integrated computer system, or ICS, is currently being redesigned and down-sized from the main frame computer by Qdata Consulting, an independent software developing company. The new ICS will hopefully come into full operation during 1996. It is also envisaged that the future ICS will be able to communicate with every farmer, livestock breeders' society, regional office and even foreign herd book society wishing to do so. Encouraging talks have recently been held with representatives of the herd book organisations in Zimbabwe and Namibia.
This high quality independent professional service is available to any stud industry on the African Continent. The high tech infrastructure need not be duplicated and the integration of data from related animals across the continent will vastly benefit the animal industry of Africa. An united database consisting of the many inter-related animals across the Continent will facilitate the calculation of reliable BLUP breeding values for all animals participating in performance testing schemes.
Performance testing is a scientific tool with which the breeder can make timely decisions on the performance and breeding values predicted for his animals. The first form of livestock improvement testing in South Africa was an egg laying competition which started in 1914 in Potchefstroom.
The National Dairy Animal Performance and Progeny Testing Scheme, as it is now known, was initiated in 1917 involving officials of the Department of Agriculture in all aspects of the scheme, from the collecting and analysing of milk samples, to processing production figures. Over the years the scheme evolved into a streamlined operation. Participants themselves collect samples and record production over a 24 hour period at five-weekly intervals. Milk samples and records are forwarded via the nitrogen and semen transport network of the AI Co-operative, Taurus, to the Central Milk Testing Laboratory at Irene. Butterfat, protein, lactose and somatic cell counts are returned within a week enabling timely managerial decisions by the producer. Summaries are given at five-weekly intervals and the Official Certificate of Performance, Reproduction and Classification is furnished immediately on completion of the cow's lactation. With this integrated system, dairy production has made significant progress and approximately 130 000 lactations are officially tested annually. Credibility of these figures are ensured by annual control testing of all herds on the scheme and are thus recognised by ICAR, of which South Africa is a full member.
Apart from the very positive impact which performance recording has on dairy herd management, the key to genetic progress lies in selecting animals according to their predicted breeding values. Breeding values are derived from the animal model in use throughout the world using the Best Linear Unbiased Prediction, or BLUP technique. BLUP enables correction of confounding influences such as environment, thus allowing comparison within and between herds in different years. These BLUP breeding values are currently calculated bi-annually by the so-called BLUP Working Group of the Animal Research Council at Irene.
The present National Beef Cattle Performance and Progeny Testing Scheme started in 1959. Officials of the Department in those days visited prospective participants with portable scales and the scheme came into being with 30 participants from the northern cattle areas of South Africa. Currently more than 1 300 breeders participate in the scheme. The first Central Bull Testing Centre was commissioned in 1964 with a capacity of 90 bulls at Irene. Currently 9 central testing centres, including one at Omatjenne in Namibia, with a total capacity of close to 3 000 bulls, operate countrywide. Of these, 4 are nonGovernmental centres. On farm bull growth testing, known as Phase D testing has grown to more than 9 000 bulls participating annually. Cow performance, a direct spin-off of performance recording, currently shows a calving percentage averaging more than 80 percent. At this level, cost effective beef production comes into its own. Carcass evaluation, the latest development of the scheme, is undertaken for those breeds taking performance testing from conception to consumption of the product. Again, BLUP analyses are undertaken for most of the breeds and figures are published as they become available. Sire summaries and complete breed analyses were recently published for the five major breeds. Should these BLUP analyses in future include animals from our neighbouring countries, socalled trait leaders may be identified in these countries. Thus encouraging the exchange of genetic material and the prevention of a narrow genetic base.
The National Pig Performance and Progeny Testing Scheme provides for purebred pigs and their progeny. Stringent standards are laid down for acknowledged stud breeders and breeding companies. On-farm testing forms the pivot of the scheme and provides for index selection, estimated breeding values, as well as a Rand value index. Approximately 5 000 pigs are tested annually. Centralised pig testing, incorporating 3 provincial centres, evaluate economically important traits of more than 1 000 pigs annually. An independent selection panel annually evaluates the genetic progress in participating herds using an advanced computer programme.
Small stock breeders have access to two performance schemes. The National Sheep and Goat Performance Testing Scheme has participation from 22 breeder groups. The National Wooled Sheep Performance and Progeny Testing Scheme focuses strongly on wool quality with a detailed analysis of productivity and reproductivity of the ewe. BLUP analyses of about 50 new rams are evaluated annually. Breeding values for the different characteristics of economic merit have evolved as the only tool with which genetic trend can reliably be predicted.
Artificial insemination saw its beginning more than half a century back, when it was practised on sheep and horses. Some ten years later, because of disease affecting reproduction it became common practice in dairy herds. Soon the important value of rapid proliferation of genetically desirable animals was recognised. The AI industry in South Africa presently consists of the Taurus Livestock Improvement Co-operative and several other private AI and embryo transfer centres. All have to be registered in ternis of South African legislation. Taurus, as it is commonly known, has intimately involved itself With livestock improvement through a service encompassing AI training, semen and nitrogen delivery and also providing the collection and transport network for the milk recording scheme. A comprehensive analytical service assists participating breeders to make informed and timely managerial decisions.
The Registrar of Livestock Improvement in the Department of Agriculture is responsible for all regulatory aspects concerning the AI industry and related professions, the importation and exportation of genetic material, the recognition of breeds, maintaining the certificates of incorporation of breeders' societies and S A Stud Book and the administration of the Advisory Board for Animal Production.
Livestock in South Africa, because of the dedication and experience of our breeders, have evolved strains of international breeds and new breeds which are truly adapted to the African environment. These animals can hold their own in producing a product which is in demand world-wide. South Africa invites you to share in this prosperity and become performance driven. Now is the time for the next generation.
ACKNOWLEDGEMENTS
I wish to thank the animal scientists at the Irene Animal Production Institute for helping me to compile this paper. BEEF PERFORMANCE RECORDING IN ZIMBABWE: THE WAY AHEAD
SUMMARY
INTRODUCTION
DERIVATION OF PRODUCTION INDICES UNDER THE BPTS
PROBLEMS FACING THE BPTS
THE WAY FORWARD
REFERENCES
C. T. Khombe and H. P. R. Tawonezvi Ministry of National Affairs, Employment Creation and Cooperatives, P Bag 7762, Causeway, Zimbabwe and STOCROP (Pvt) Ltd, P 0 Box 694, Marondera, Zimbabwe.
SUMMARY
The methods of calculating the 205 and 550 day indices that are the selection criteria of the Beef Performance Testing Scheme (BPTS) are discussed. Several weaknesses of the current system have been observed. These include use of inappropriate correction factors for age-of- dam at calving and sex of the calf. The scheme also fails to account for the effects of selection, selective mating and the preferential treatment of highly valued animals. Furthermore maternal effects are not considered although the scheme puts emphasis on selection for high weaning weights. Suggestions are made to use the Reduced Animal Model to estimate best linear unbiased prediction (BLUP) breeding values for both 205 and 550 day weights. It is recommended that the BPTS should be modified to use Restricted Maximum Likelihood procedures fitting a Reduced Animal Model, because it yields solutions that are unbiased.
INTRODUCTION
Performance recording of beef herds in the commercial farming areas of Zimbabwe was started in the 1960s and is currently managed by the Department of Agricultural Technical and Extension Services (AGRITEX). The Beef Performance Testing Scheme (BPTS) involves collection by the farmer of cow and calf records which are processed by AGRITEX to provide the farmer with adjusted values and ratios of weights at 205 and 550 days of age. The 205 day weight is used as a measure of calf growth to weaning and is used in dam selection, while the 550 day weight is the selection criterion for replacement stock. The former assumption is made under the premise that cows that wean heavy calves have good mothering abilities.
DERIVATION OF PRODUCTION INDICES UNDER THE BPTS
Correction factors The BPTS corrects for the environmental effects of age-of-dam at calving and sex of the calf. The current adjustments of calf weaning weight in Zimbabwe, shown on Table 1, are multiplicative and were obtained from the United States (United States Beef Improvement Federation, 1972). These factors are unlikely to be applicable under Zimbabwean conditions (Buvanendran, 1990; Machaya and Tawonezvi, 1992). Furthermore these adjustments ignore the importance of other environmental influences like previous lactation status of the dam and month of birth (Machaya and Tawonezvi, 1992). The latter authors have derived correction factors under local conditions using 6365 and l l 068 Mashona calve records, respectively, and concluded that the corrections for age-of-dam were inappropriate due to an inappropriate age classification.
Productivity indices
The 205INDEX is derived as follows:
205INDEX = [(G'*205) + BWT] * CFsex * CFdam where G' = Gain/days = wnwt - bwt/wnag, BWT is the birth weight, wnag is the weaning age, wnwt is the weaning weight, Cfsex is the correction factor for sex and Cfdam is the correction factor for age of dam.
The 205ratio is derived as 205INDEX/Corrected sample mean where the corrected sample mean is the sample mean of corrected values within each sex.
The 550INDEX is derived as follows:
550INDX = YRLNGWT - WNWT/DAYS * 345 + 205INDEX where yrlngwt is the 550 day weight.
The 505Ratio is derived from the 505INDEX in essentially the same way as the 205ratio.
Table 1. Multiplicative correction factors for weaning weight
Variable Curre Source Machaya & TaNvonezvi Nt1 Buvanendran (1992) (1990) Age of dam (years) 2 1.17 3 1.10 1.11 1.11 4 1.05 1.06 1.07 5 1.00 1.00 1.04 6 1.00 1.00 1.00 7 1.00 1.00 1.00 8 1.00 1.00 1.00 9 1.00 1.00 1.00 10 1.00 1.00 1.00 >11 1.05 1.03 1.05 Sex Females 1.10 1.09 1.06 Castrates 1.05 1.00 1.00 Bulls 1.00 1.00 1.00 Month of birth March 1.04 April 1.04 May 1.00 June 1.00 July 1.00 1.00 August 1.00 1.00 September 1.00 1.00 October . 1.00 1.00 November 1.07 1.04 December 1.07 1.01
1United States Beef Improvement Federation (1972).
PROBLEMS FACING THE BPTS
The selection criteria that are used by the BPTS, both the 205 and 550 day ratios, have the following deficiencies. Firstly, almost all the herds participating in the BPTS do not measure the birth weight. Consequently each breed is supposed to have a fixed birth weight (e.g the birth weight of Mashona cattle is set at 28kg). This is a serious shortcoming of the index since animals that have light birth weights and heavy weaning weights can not be distinguished from those with heavy birth weights and similar weaning weights. The measurement of birth weight or any weight taken within the first week of life of the calf should be incorporated in any selection program for growth in beef cattle.
Secondly, the present method used in Zimbabwe of adjusting for age by the average daily gain results in serious bias in adjusted weights of calves whose ages deviate greatly from the standard age of 205 days. Machaya and Tawonezvi (1992) argued that although the growth of calves from birth to weaning was significantly linear, with a non significant quadratic term, the current methods of adjustment resulted in an over adjustment in favour of the younger calves at weaning. Buvanendran (1990) suggested the use of a linear regression coefficient of weight on age.
Thirdly, as was mentioned earlier, the correction factors that are currently in use are inappropriate and need urgent revision. However, the more elaborate correction factors of
Machaya and Tawonezvi (1992) are likely to be only applicable to Mashona cattle and other small framed breeds. Dzama et al. (1995) have shown that although the current adjustment factors arc inappropriate for the Mashona, they are appropriate for the Hereford and Sussex breeds. The latter authors have argued that cattle of similar type, to the Hereford and Sussex breeds, are likely to have been used to derive the currently used correction factors. There is a need to derive correction factors that are applicable to a large number of genotypes that are available in Zimbabwe.
Fourthly, the ranking of animals by either their 205 or 550day ratios is not efficient since the derivation of the indices does not account for the selection occurring within the herds, relationships between animals within and between herds, selective mating occurring within herds and preferential treatment of highly valued animals. Thus the selection of replacement bulls using their offspring means, as is currently practiced, does not consider the genetic merit of their mates (the dams) and selective mating that occurs when high merit bulls are mated to high merit cows. Such breeding values are biased and do not rank animals efficiently on their genetic merit.
Finally, both indices do not account for the presence of maternal genetic and environmental effects on both 205 and 550 day weights. Maternal genetic effects are any influences of the dam on her offspring, apart from the genes she transmits in the ovum, which although environmental in terms of their effect on the genotype(s) of the offspring, are due to the genotype of the dam (Willham, 1972). Maternal genetic effects become a nuisance if they are negatively correlated with direct genetic effects, as is the case in selection programmes that use weaning weight as the selection criterion. Under such circumstances optimum response is attained when the direct and maternal effects, together with their correlation, are considered in the development of the selection index (Baker, 1980). In beef cattle maternal genetic effects have considerable effects on both birth weight and pre-weaning gain. There is some inconsistency in the literature about the age at which maternal effects become inconsequential. There evidence to suggest their influence on post weaning growth traits (Meyer, 1992).
THE WAY FORWARD
The future of the BPTS in its current state is uncertain. Commercial farmers are becoming computer literate and are making use of performance recording and breeding value predicting software packages that are now widely available. These software packages are cheaper and avail the breeding values promptly without experiencing the delays and the numerous bugs that plague the current system. The number and size of registered herds is also declining. It is very unfortunate that this scheme does not include the majority herds in the small holder farms.
The need to rejuvenate the BPTS can not be overemphasized. Producers need to link their herds to allow -the comparison of animals across herds and hence facilitate the exchange of breeding animals. Most herds are too small to run within herd selection programmes and could benefit from the development of a national recording scheme.
The methods of selecting beef cattle in Zimbabwe need to be revised in the light of the new evaluation methods that are already in use in other countries. Animal models, so called because they include a random effect representing the additive genetic merit or breeding value of each animal, are now the widely used state of the art procedures of genetic evaluation. A discussion of animal models is beyond the scope of this paper. It should suffice to state that a Reduced Animal Model (RAM) would be appropriate under our present conditions of limited computer power. The RAM uses ancestral data to evaluate the genetic potential of animals and is used to predict the likely performance of potential progeny from a proposed mating. An animal's performance and that of its progeny are included into the model, as they become available. This accumulation of information increases the reliability of the estimates of breeding values as less ancestral and more individual and progeny data are used. Inclusion of genetic ties between animals using a relationship matrix allows the RAM to account for selection, selective mating and connects animals across herds. The infrastructure of sire evaluation that is currently in existence in Zimbabwe can easily be modified to use Mixed Model Equations and provide Best Linear Unbiased Predictions (BLUPs) breeding values. The handicap of using these methods might be the lack of algorithms that can lessen the computing burden solving the numerous equations that is inlicrent -in these models. The correction of fixed effects could reduce the load of work that the computers have to carry out. Researchers at both the University of Zimbabwe and the Department of Research and Specialist Services should endeavour to assist the staff in AGRITEX to apply these methods to the BPTS. The dairy industry has recently adopted these methods and their experiences. hardware and software could be used to also benefit the beef industry.
REFERENCES
Baker, R.L. (1980) Proceedings of the New Zealand Society for Animal Production, 40:285- 303.
Buvanendran, V. (1990) Journal of Agric. Science, Cambridge, 114:35-40.
Dzama, K., Walter, J.P., Ruvuna, F. and Taylor, J.F. (1995) Animal Production (submitted).
Machaya, A.C. and Tawonezvi, H.P.R. (1992) Zimbabwe Journal of Agric. Res. 30:13-20.
Meyer, K. (1992) Livestock Production Science, 31:179-204.
United States Beef Improvement Federation (1972) North Carolina State University, U.S.A.
Willham, R.L. (1972) Journal of Animal Science, 35:1288-1293. RECENT DEVELOPMENTS IN LIVESTOCK BREEDING IN SOUTH AFRICA
SUMMARY
INTRODUCTION
DIRECT AND MATERNAL VARIANCE COMPONENTS
GENOTYPE X ENVIRONMENT INTERACTION
BREEDING OBJECTIVES
THRESHOLD TRAITS
PRACTICAL APPLICATION
CONCLUSION
REFERENCES
G. J. Erasmus and J. B. van Wyk Department of Animal Science, University of the Orange Free State, Box 339, Bloemfontein 9300, Republic of South Africa.
SUMMARY
South African animal geneticists are committed to utilising the recent advances in mathematical and computer technology in an effort to enhance breeding progress. Many local breeds have been genetically analysed and described. The availability of large, reliable data- sets on the centralised data-base of the Department of Agriculture has led to many useful studies, some of which are discussed. The decrease in the maternal component of variance with age and its unimportance in sheep fleece traits has been clearly demonstrated as has the effectiveness of selection on EBV's. Maternal breeding values can be utilised in the selection of dam lines or in low-input production systems. Investigations into the genetic improvement of categorical traits such as reproduction and survival rate, utilising a threshold model have been undertaken. Deleterious correlated responses to selection and total efficiency is being investigated. The need for sound breeding objectives when advanced technology is used, especially when natural resources are .limited, is stressed. The application of modern techniques in the local livestock industry is discussed. Practical application generally lags behind research, but in some cases in the dairy industry, application has outstripped local research efforts.
INTRODUCTION
Interest in the genetic improvement of livestock in South Africa has escalated over the past years. This stems mainly from the increasing comprehension that it is more cost effective to improve genotypes than most environments. The economic and financial realities facing livestock production have led to a conception of changing genotypes to suit environments rather than changing environments to suit unadapted genotypes. Artificial environments have not only become economically prohibitive, but also highly unstable and a risk to food security. Marked genetic changes in many livestock breeds have resulted, as well as a growing interest in indigenous breeds.
The application of state-of-the-art methodology for estimation of co-variance components and breeding values by a growing group of local scientists has made a valuable contribution in providing guidelines for genetic improvement and has opened exciting future possibilities. However, much still needs to be done. The National Livestock Improvement Schemes of the Agricultural Research Council provides the mechanism for the genetic improvement of livestock in South Africa together with the S.A. Stud Book and Livestock Improvement Association. They collect and provide the data necessary for research and subsequent analyses. The centralised large data base is the stud industries' most important asset and it is unfortunate that some breed administrators do not realize this. The recent provision of between stud breeding values and genetic profiles of individual studs for a number of breeds should, however, serve as suffcient incentive to all serious breeders to collectively pool data for a meaningful analysis.
The availability of relevant data has been the prime incentive for genetic studies in South Africa, leading to some solutions, but invariably creating new problems. The aim of this paper is to report on some of the recent results obtained, to discuss some problems unveiled and to speculate on some possible future developments. The literature cited is incomplete and the reader is referred to recent issues of the South African Journal of Animal Science for greater support of statements made.
DIRECT AND MATERNAL VARIANCE COMPONENTS
The ease with which direct and maternal additive variance components, together with their covariance, and making provision for permanent maternal environmental effects, can be estimated, has led to several such studies on local breeds and large herds or flocks. It has been clearly illustrated how the maternal component decreases and the direct component increases with age (Snyman et al., 1995). The number of wool follicles a sheep possesses is presumably determined before birth, indicating a large possible intra-uteral effect of the dam. The additive maternal effect on fleece traits has, however, been found to be negligible in Merino and Afrino sheep (Olivier et al., 1994; Snyman et al., 1995). For early growth traits in sheep and cattle, the maternal component is important. The negative co-variance mostly found between direct and maternal components decrease total heritability and poses the question whether selection should be on direct or maternal breeding values. The arbitrary weighting of the two to create a new breeding value is a rather crude solution. Selection on maternal breeding values seems obvious in identified dam lines for terminal cross-breeding and in breeds where maternal ability has deteriorated due to selection for other traits. Selection with emphasis on direct breeding values is probably the correct strategy for improving sire and so- called "pure" breeds.
GENOTYPE X ENVIRONMENT INTERACTION
Although still largely undocumented, a clear pattern supporting "Falconer's paradox" seems to be emerging. Selection in more than one environment seems to be ineffective in improving performance in any of the environments. Where a breed is found in a variety of environments, sire x herd or area interactions will have to be included in an animal evaluation model as indicated by Neser et al. (1995) for Bonsmara cattle. Sires with mediocre overall breeding values particularly, can change quite drastically in ranking when evaluated in different areas. The same was found in the dairy industry where the problem of heterogeneous variances was also addressed (Delport et al., 1994).
BREEDING OBJECTIVES
Traits vary in importance in accordance with the prevailing environment and production system. Reproduction and survival rate are, however, important in any environment or production system and are of paramount importance in determining biological and economic efficiency. There is a growing awareness of the necessity to take cognisance of this fact in any improvement programme.
The genetic improvement of production traits is seldom straightforward when a holistic approach to "improvement" is applied. This is so because of the almost inevitable deleterious correlated responses obtained, which is especially important when, as so often happens, reproduction, survival and efficiency is adversely affected. This has even led to suggestions of a moratorium on selection for traits such as growth and higher fibre production.
Genetic correlations can be high, but are never unity and when caused partly by linked genes, can even change. This implies, theoretically at least, that unfavourable genetic correlations could make selection for overall merit (production, reproduction, survival and efficiency) difficult, but not impossible. Genetic improvement in overall desirability should be possible even under adverse conditions provided the most adapted genotype is selected initially and selection is carried out under identical conditions. The most modern methodology should be used since the sensible utilization of genetic diversity in livestock can only benefit mankind. The emphasis should be on sensible utilization and hence much more research is needed, but the methodology does exist. It is perhaps a sad feature of science that the emphasis is normally on perfecting methodology and not on how it can best be implemented.
The only trait that cannot be measured is efficiency of feed conversion under natural conditions. Much work has been done in South Africa to address this problem of maximizing herd cfciency. In this context, Roux and Scholtz (1992) state: "Reliable indications of possible gains in efficiency of production are only possible on the basis of scientific theory. Without such a separation of the plausible from the probable, the danger exists that animal breeding may become an infinite regress of experiments, necessarily involving large numbers, in a vain attempt to accommodate or exploit biological variation". Roux (1992) argues that at present, there are only two ways of appreciably increasing herd feed or cost efficiency. These are increasing fertility and viability and fcedcrbreeder dimorphism.
It has been shown that the use of the Kleiber ratio as selection criterion can be effective in increasing growth without an increase in birth or mature body weight. This has led to the suggestion of expressing milk production in dairy cattle as a function of mature cow weight. Two further niggling questions remain in the dairy industry. Firstly, should more than first lactation records be included in evaluation and selection, especially in view of preferential treatment and a high erosion rate of roughly 30 percent for every subsequent generation? Secondly, should emphasis be on percentage or kilograms fat and protein in selection programmes?
Animal breeding presently finds itself in what can be termed the hi-tech computer era. Rapid genetic change is possible, utilizing modern methodology as illustrated by Olivier et al. (1995), for Merino sheep. This emphasizes the need for clear breeding goals that will be beneficial and not lead to rapid disaster.
THRESHOLD TRAITS
The two most important traits, via. reproduction and survival rate are generally considered to be "threshold traits". These traits are polygenic with an unobserved underlying continuous distribution, but an observed discreet or categorical expression. Other similar traits include calving case, disease resistance and type scores. Best Linear Unbiased Prediction (BLUP) is not. optimal for these traits, but a method which can be regarded as an extension of BLUP to a type of nonlinear problem can be used. It is based on the threshold concept (Wright, 1934) and employs a Bayesian procedure for statistical inference. A general approach proposed by Gianola and. Foullcy (1983) was employed by Konstantinov et al. (1994) to evaluate Dormer sires for reproduction and survival traits. It was concluded that the threshold model can effectively be applied for routine evaluation. Moderate heritability estimates obtained on the underlying scale show that some selection progress is possible. These findings were supported by Rust et al. (1995) for Afrikaner cattle.
The use of a threshold model has opened exciting possibilities for studying and evaluating traits connected with adaptation. An important added advantage of this methodology is that probability distributions for progeny of sires across "fixed effects" can be calculated. It is envisaged that it will soon form part of national sire evaluation of beef breeds.
PRACTICAL APPLICATION
It can safely be said that the dairy breeds lead the way in the practical application of modern technology. This is probably so because of international exposure and competition. In fact, the S.A. Holstein Friesland Association employs a highly sophisticated multiple trait animal model for evaluation of linear type traits while the relative importance of these traits under local conditions is yet to be determined. On the other hand, while sheep breeding research in South Africa is internationally acclaimed, local sheep breeders are reluctant to utilize the acquired knowledge to their advantage.
CONCLUSION
Animal breeding in South Africa has gradually progressed from being a mysterious art to becoming a logical science. The use of modern statistical procedures has greatly enhanced both research and practical application.
REFERENCES
Delport, G. J., Loubser, L.F.B., Rautenbach, L. and Erasmus, G.J. (1994) Proc. Open Session Interbull Ann. Meet., Ottowa, Canada.
Gianola, D. and Foulley, J.L. (1983) Genet. Sel. Evol. 15: 506.
Neser, F.W.C., Konstantinov, K.V. and Erasmus, G.J. 1995. Proc. 34rd SASAS Congr., Bloemfontein.
Konstantinov, K.V., Erasmus, G.J. and Van Wyk, J.B. (1994) S. Afr. J. Anim. Sci. 24(4), 119
Olivier, JI, Erasmus, G.J., Van Wyk, J.B. and Konstantinov, K.V. (1994) S. Aft. J. Anim. Sci. 24(4), 122.
Olivier, J.J., Erasmus, G.J., Van Wyk, J.B. and Konstantinov, K.V. (1995) S. Afr. J. Anim. Sci. (in press).
Roux, C.Z. (1992) S. Aft. J. Anim. Sci. 22: 11.
Roux, C.Z. and Scholtz, M.M. (1992) S. Afr. J. Anim. Sci. 22: 16.
Rust, Tina, Van Der Westhuizen, J. and Konstantinov, K.V. (1995) Proc. 34th SASAS. Congr.. Bloemfontein. Snyman, M.A., Erasmus, G.J., Van Wyk,.J.B. and Olivier, J.J. (1995) Livest. Prod. Sci. (in press)
Wright, S. (1934) Genetics 19: 506. CLOSING REMARKS
RELEVANCY OF RESEARCH
CONDUCT OF RESEARCH
CAPACITY TO DELIVER
ESEARCH AND OTHER ISSUES WHICH NEED TO BE ADDRESSED
FOLLOW-UP
S. Sibanda University of Zimbabwe, Department of Animal Science, P 0 Box MP 167, Mount Pleasant, Harare, Zimbabwe.
RELEVANCY OF RESEARCH
Any research undertaken must be problem driven rather than led by equipment and competency projections. Research should be guided by consumer and market requirements, and producer needs to meet market requirements as well as personal preferences and to remain in production. The scientific community must make collaboratory efforts to come up with strategic research that is relevant.
Problem identification
In defining problems that need to be addressed through research, farmer participation cannot be overemphasised. The use of indigenous knowledge is vital in clearly defining underlying problems which the scientific community might not be aware of. When producers participate in research activities its easier for scientists to identify opportunities that farmers might not be aware of
Breeding Policies and Strategies
It is necessary for scientists to get together and initiate the process of producing breeding policies (objectives) for different species. This will depend on what we and other stakeholders (ie scientists, farmers and the market) perceive as the desirable long-term goal. Such policies, once formed would then dictate what strategies should be adopted to achieve the long-term goals. This can be achieved by: i. Formation of an Animal Breeders Association (ABA) ii. Exchange of information through networking and exchange visits iii. ABA should have an affiliate or satellite association at a national level iv. Identification through the ABA, common areas where resources may be pooled in order to address the problem regionally v. Involvement of other disciplines to tackle policy issues. CONDUCT OF RESEARCH
Location
This will depend on type of research. Wherever possible, it is necessary to do research on- farm. This cheaper and likely to be more problem led.
Availability and use of appropriate tools
Our scientists should be aware of what tools are available to address research problems. However, where ideal tools are not available for a number of reason, particularly inadequate financial provision, alternative approaches must be considered: ie our research should not be equipment driven. The available tools include: i) Biotechnology ii) Data handling and analytical techniques iii) Hardware (equipment) and software.
Funding
There is a need to reduce dependency on foreign funding. Local resources must be mobilised. This can be achieved through networking and institute cost sharing arrangements (ICSA). ICSA can be achieved by inter-institutional and regional collaboration. Future funding should be sought not only from the government but from industry (producers and processors) as well.
CAPACITY TO DELIVER
Adequacy of current staffing levels, qualifications and disciplines should be assessed. The areas that need more personnel must be identified. The training institutions and programmes must be evaluated on whether they are attracting the and producing the right calibre and number of graduate students. Appropriateness of the existing training programmes can be addressed regionally through the ABA.
ESEARCH AND OTHER ISSUES WHICH NEED TO BE ADDRESSED
From the two days of presentations, it is clear that research addressing the following areas must be undertake. Full cooperation and collaboration among the Scientists in the region is vital. Issues such as funding and tools/equipment to use must be clearly tackled.
Genetic conservation and biodiversity i.Characterisation of indigenous animals ii.Identification of endangered species iii.Development of strategies that can be used for conservation of animal genetic resources and utilisation iv.How best to fund these activities
Studies on adaptive traits
The following areas of concern should be addressed inorder to improve adaptive traits: i.Disease resistance; with the background of drug resistance, environmental concerns and consumer preferences ii.Water economy in the drought endemic region iii:Utilisation of poor quality feedstuffs iv.Heat tolerance
The following tools can be used to carry out the above studies i.Biotechnology, where appropriate ii.Conventional breeding tools (BLUP) iii.A combination
Genotype by environment interaction
The following areas need to be addressed: i.What is the best strategy for addressing changes in the environment? For example should we use of adapted breeds or change the environment? ii.How can we best address the antagonism between adaptive traits and productive traits? iii.How can we include future changes in the environment in genetic improvement programmes?
Market requirements
Comprehensive market studies should be undertaken. Specifically, the following areas must be addressed. i.Current market intelligence. Collaboration with socio-economists is essential ii.Projections of future market (local and export) requirements ii.Use of farmer participatory approaches to evaluate subsistence needs iv.Educating the market about undesirable consumption patterns
Cost of Genetic improvement
The cost of genetic improvement must be evaluated. This can be done in the following manner: i.Technology assessment in dollar terms ii.Loss in biodiversity iii.Loss in adaptive traits iv.Other environmental and hidden costs.
FOLLOW-UP
It is important to take advantage of this momentum and contacts established here and think of concrete measures to action some of the issues discussed here in the past two days. With the All Africa Animal Production Conference in South Africa next year, this meeting may be used as a stepping stone for initiating a number of activities to be reviewed at that conference.