Application of New Genetic Technologies to Animal Breeding: Proceedings of the 16th Biennial Conference of the Association for the Advancement of Animal Breeding and Genetics (AAABG) 25-28 September 2005

The 16th Biennial Conference of the Association for the Advancement of Animal Breeding and Genetics (AAABG) gathers together scientists, extension workers, producers and industry personnel to review developments in the application of new technologies to animal breeding. Conference presentations include 30 invited reviews and papers, and 95 contributed papers. All papers are peer-reviewed, and cover session topics that focus on genetic evaluation systems, gene expression profiling, identification and manipulation of quantitative trait loci, progress in applied programs and advanced statistical and computing techniques.

Industry applications are discussed for improvement in production, health and reproduction of domestic livestock, aquaculture species and even crocodiles and ostriches. Institutions and industries in Australia, New Zealand, USA, South Africa, South-East Asia and Japan are represented with significant participation of major Cooperative Research Centres.

These proceedings contain the full text of all contributed papers and summaries of the invited reviews which are published separately in the Australian Journal of Experimental Agriculture.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Sep 1, 2005
• ISBN: 9780643098534

Book preview

APPLICATION AND IMPACT OF NEW GENETIC TECHNOLOGIES ON BEEF CATTLE BREEDING: A ‘REAL WORLD’ PERSPECTIVE*

E. J. Pollak

Department of Animal Science Cornell University Ithaca, NY, 14853 USA.

SUMMARY

Molecular genetics is a maturing discipline with innovations that are finding application in animal breeding. Currently, DNA tests are available for parent identification or verification, and markers tests also exist for quantitative trait loci (QTL) affecting important traits in beef cattle. The beef industry is, however, a particularly challenging industry in which to design breeding programs that fully capitalise on the potential of this technology. Hence, adoption within this industry to date has been below expectation. This paper examines several applications that are being investigated and will include discussion on issues constraining the transfer of DNA technology. An example of using DNA parentage testing for expanding the reach of selection programs into the commercial sector of the beef industry is explored in some depth, as it represents a potential high impact application. Use of molecular information in selection programs and in genetic evaluations is also discussed.

* From invited paper. The full text, including this abstract is published in Australian Journal of Experimental Agriculture 45, (7-8) in press.

GENETIC EVALUATION FOR THE BEEF INDUSTRY IN AUSTRALIA*

H.-U. Graser, B. Tier, D.J. Johnston and S. A. Barwick

Animal Genetics and Breeding Unit+, University of New England, Armidale, NSW 2351, Australia

SUMMARY

In Australia, genetic evaluation for beef cattle has been performed using an animal model with best linear unbiased prediction since 1984. The evaluation procedures have evolved from simple to more complex models and from few to a large number of traits, including traits for reproduction, growth and carcass characteristics. This paper describes in detail the current beef cattle genetic evaluation system ‘BREEDPLAN’ used for the Australian beef cattle industry, the traits analysed and underlying models, and presents a short overview of the challenges and planned developments of coming years.

* From invited paper. The full text, including this abstract is published in Australian Journal of Experimental Agriculture 45, (7-8) in press.

+ AGBU is a joint venture of NSW Department of Primary Industries and the University of New England

PRACTICAL ASPECTS OF A GENETIC EVALUATION SYSTEM USING PARENTAGE ASSIGNED FROM GENETIC MARKERS*

K.G. Dodds¹, J.A. Sise¹ and M.L. Tate²

¹AgResearch, Invermay Agricultural Centre, Private Bag 50034, Mosgiel, New Zealand.

²Ovita Ltd, PO Box 5520, Dunedin, New Zealand.

SUMMARY

Animal breeding values can be calculated when genetic markers have been used to help determine the parentage of some of the animals, but their parentage has been incompletely determined. The pedigree sampling method is one computing strategy for calculating these breeding values. This paper describes and discusses methods for dealing with a number of practical issues arise when implementing such a system for industry use. In particular, diagnostic systems for detecting inadequacies or possible errors in the genotyping systems and the recording of animal management are developed. Also, characteristics of the best-assigned pedigrees are calculated according to mating group and used to check for sires missing from these groups. The correlation between breeding values estimated from a single sampled pedigree (using parentage probabilities) and those estimated as the mean from many sampled pedigrees gives a diagnostic to indicate which estimated breeding values are more influenced by uncertainties in relationships. For the analysis of survival traits, a method to enumerate and assign likely parentage to dead offspring that have not been DNA sampled and genotyped is described. When embryo transfer technology is used, the genetic dam needs to be included as a possible dam when considering parentage. If some fixed effects that depend on the parent are missing, these can be sampled similarly to parentage, and this may improve the evaluation if certain assumptions are met. A method to provide a likely list of parents, the ‘fitted pedigree’, which is based on the most likely parents, but modified to reduce the occurrence of unlikely family sets (e.g. very large litters) is also presented. The use of these methods will enhance the practical application of DNA parenting when used in conjunction with genetic evaluation.

* From invited paper. The full text, including this abstract is published in Australian Journal of Experimental Agriculture 45, (7-8) in press.

IMPLEMENTING GENETIC EVALUATION IN THE NEW ZEALAND DEER INDUSTRY

J.A. Archer,¹, J.F. Ward¹, S-A.N. Newman¹, G.J. Cruickshank²and A.J. Pearse³

¹AgResearch Ltd, Invermay Agricultural Centre, PB 50034, Mosgiel, Otago, New Zealand

² Genequest, RD 12 Havelock North, New Zealand.

³Deer Industry New Zealand, PO Box 10-702, Wellington, New Zealand

SUMMARY

Genetic evaluation is a tool now attracting growing use by the New Zealand deer industry. Following a pilot genetic evaluation system, a fully commercial genetic evaluation system has been implemented on an industry wide and industry supported basis. Selection of a model for implementation considered, among other criteria, the need for competition in some areas contrasted by the demand for the national good of an industry-backed single entity in other areas. Issues of operation scale in the deer industry also influenced the outcome. The result is a genetic evaluation system operating commercially and available to red deer breeders from 2005. This outcome, together with recent initiatives to develop data suitable to implement across-herd analysis, will have a significant impact on genetic progress in farmed deer. The potential for improvement of production efficiency in the deer industry is large. Objective identification and use of superior genetics combined with improved management systems will have an important role to play in aligning venison production with market demand.

Keywords: Deer, genetic evaluation, breeding values, sire referencing

INTRODUCTION

The New Zealand farmed deer industry was established in the 1970’s from a large and diverse resource of feral deer. In 30 years of deer farming significant gains have been made via importation of new deer genotypes and selection within the population of farmed deer. A breeding structure has evolved, with approximately 40 to 50 recognised breeding herds servicing a farmed deer population estimated to be 0.9 million hinds (MAF, 2005). The seedstock herds provide approximately 20 to 30% of stags required by the deer industry annually.

Until recently, most selection emphasis in seedstock herds was placed on antler traits for velvet production. This was based on approximately 50% of industry revenue being from velvet income in the early days of the industry with very high per animal returns. Also, velvet antler production is heavily dependant on genetic potential and commercial antler traits are generally highly heritable producing strong genetic responses to mass selection. The emphasis in selection for velvet is generally seen in the genetic trends calculated for most seedstock herds. Emphasis on antler traits has been further promoted as interest in producing trophy antler heads for hunting operations has grown steadily in the last decade.

However, the core nature of the deer industry has changed over time. The industry revenue split is now approximately 77% from venison, 12% velvet and 8% hides and leather and 4% co-products (Deer Industry New Zealand, 2005). This change has had some impact on breeding objectives, with growth traits receiving more emphasis in some breeding herds. Selection for traits of moderate heritability (such as growth) has brought with it a demand for more objective means of evaluating genetics rather than phenotypic performance alone. Also, more sheep and beef farmers have diversified into deer bringing to the industry a greater appreciation of genetic tools used in other industries. These factors have generated strong support within the deer industry for developing a national genetic evaluation system, reflected in a survey conducted by DEEResearch Ltd where genetic evaluation ranked as a high research priority,.

DEVELOPMENT OF GENETIC EVALUATION FOR DEER

Pilot genetic service. In response to demand from a small number of stud herds, AgResearch offered a pilot breeding value service for deer, beginning in 1998. The service was restricted to within-herd analysis as historical breeding structures among deer seedstock herds produced only limited linkage. Traits analysed were growth and velvet, using a modular approach. The service was developed with funding from the Foundation of Research, Science and Technology, and most operating costs were recovered from breeders using the service.

From an initial 3 herds, the service grew to a point where in 2004 breeding value analyses were conducted for 15 herds (of approximately 35 recognised elite herds breeding red deer of English and European origin nationally). Several of these herds conducted breeding value analyses for the first time in 2004, while some of the herds previously using the service began twice-yearly analyses.

Commercial operation of genetic evaluation. The ready adoption of the pilot evaluation service demonstrated that the wider deer industry, and deer breeders in particular, were willing to use and pay for genetic evaluation. To establish an enduring viable genetic evaluation service for the deer industry, the pilot scheme needed to evolve to a fully commercial service with cost recovery of all operating costs. Drivers behind this development included the need for industry to have input into genetic evaluation systems, and to better utilise resources by separating development and operational functions.

Different models for the delivery of genetic evaluation services were explored. Several principles guided the choice of model. Firstly, the operation of a genetic evaluation service should be on a commercial basis, with financial support from the deer industry only for industry-good functions where market conflict or compromise exists. Secondly, while competition encourages efficiency, there are significant benefits from the development of a single national database and genetic evaluation system. A structure which combined industry-supported single entity structures for functions where a monopoly is beneficial, with potential for competition in other areas (e.g. DNA parentage services) was sought. In reality, there is no legislative ability within the industry to create a monopolistic structure or prevent others from providing competing services. A further consideration was the limitations imposed by the relatively small scale of the deer industry, which meant that setting up a full service was unlikely to be viable without taking advantage of technology and business structures developed by other industries.

The model adopted sees genetic evaluation delivered to the New Zealand deer industry via a partnership between AgResearch and Sheep Improvement Ltd (SIL), a subsidiary of Meat & Wool New Zealand. Under this partnership arrangement, SIL provide the database and operational capability, while AgResearch provides genetic evaluation software. The link between the database and genetic evaluation software is a fully automated link. Modification of the SIL database infrastructure to extend its capability to deer was funded by Deer Industry New Zealand and the New Zealand Deer Farmers Association (NZDFA) in 2004 as an industry-good function.

The service is retailed to deer breeders via the SIL bureau structure (Geenty 2000). The bureau accepts data from the breeder and uploads it via the internet to the database. When a genetic evaluation report is requested, the database sends a data file to the genetic software, which analyses the data and uploads the outputs back to the database (Newman et al. 2000). The report is formatted and delivered to the breeder by the bureau. The number of bureaus servicing deer breeders is currently restricted to two. This allows competition for the service while also giving sufficient volume of business to make bureaus viable. Both bureaus are also existing sheep bureaus, further reducing the overheads required for training and support.

International experience has shown the importance of industry involvement in developing genetic evaluation services (Rickards, 1997). An important feature of the service is the establishment of a Genetic Evaluation Steering Committee to facilitate industry input. This committee is established by Deer Industry New Zealand and the NZDFA and includes breeder and commercial farmer representatives, deer industry organisation representatives, technical expertise from AgResearch and SIL and an independent chair. The role of the committee is to determine policy and set standards with regards to genetic evaluation in the New Zealand deer industry, and provides the opportunity to take a uniform national approach to issues and avoid fragmentation. The committee’s role does not extend to commercial governance of the SIL/AgResearch service. However, the committee governs the structure of genetic evaluation services and could seek to replace SIL/AgResearch with other providers if necessary. Also, additional services could be provided by other providers in the future at the discretion of the committee – an example could be the addition of other trait group modules or mate selection services.

The genetic evaluation service will also extend to providing a data link with DNA laboratories providing parentage services in deer. Deer breeders are significant users of DNA parentage technology due to the difficulties of obtaining parentage information in deer by other means. Linkage between DNA parentage and genetic evaluation databases generates significant synergies, as DNA laboratories require information on mating and calving groups which is stored on genetic evaluation databases, while genetic evaluation utilises the outcomes of DNA pedigree matches. Database linkage will result in significant savings in data handling, but will not compromise the potential for competition between DNA laboratories.

TRAITS AND MODELS

The genetic evaluation system initially offers two modules, for growth and velvet antler production. The growth module utilises data on weaning weight, autumn weight, 12 and 15 month weights and mature weights. Weaning weight and autumn weight are analysed using a direct plus maternal genetic effect model, while mature weights are analysed with a repeatability model, fitting a permanent environmental effect. Fixed effects include contemporary group, age of dam and birth date or estimated conception date fitted within herd-year-sex where data is available. Breeding values are reported for weaning weight (direct and maternal), 12-month weight and mature weight.

The velvet module utilises data on velvet antler weight (cut at the optimal time) as a 2 year old, and at older ages (3 years plus) as well as 12-month liveweight. A direct additive genetic model is fitted with a permanent environmental effect included for velvet production at older ages. Fixed effects include year of birth x contemporary group. Breeding values for mature velvet weight are reported.

FUTURE DEVELOPMENT

Future developments are likely to encompass new economically relevant traits and expantion of the number of herds participating in across-herd analysis. Modules will be refined, with genetic parameters re-estimated when more data are available and new models tested to improve the accuracy of evaluation.

Traits. The new genetic evaluation system as implemented offers modules for growth (direct and maternal) and velvet production. However, while these traits are the current priority, several other traits are commercially important and the evaluation system will be expanded to accommodate these traits as demand and use builds. It is likely that a modular approach taken to the current system will continue to be used, with sub-indices developed for trait groups. Priority areas are likely to include reproductive success, seasonality (early calving), carcass traits and temperament. Evaluation of trophy merit may also be incorporated, as the trophy market while small is a significant and growing part of the industry. At this point of time little or nothing is known about the genetic control of many of these traits, and so new research will be required before evaluation of these traits can be considered.

Across-herd analysis. Expansion of genetic linkage to support a national across-herd genetic analysis is necessary to obtain maximum benefit from the genetic evaluation system. A major limitation has been breeding structures used in the industry, which have a low level of sire linkage between breeding herds. In the past artificial reproductive technologies were not widely used, although during the last 5 years the use of artificial insemination (AI) has increased, due principally to the development of transcervical AI techniques in red deer with high success rates. An attempt in the early nineties to progress genetic evaluation by establishing a sire referencing program failed due to competing commercial interests.

In 2003 a sire reference scheme was commenced by AgResearch at Invermay Agricultural Centre, with funding from Foundation of Research Science and Technology. The scheme facilitates the development of genetic linkage between herds and implements across-herd analysis. The first across-herd analysis of growth data occurred in December 2004 using the pilot genetic evaluation system and future data will be analysed under the new SIL/AgResearch system to assist development of a national evaluation.

While public funding has been used to support sire referencing, this is not a sustainable long-term situation, and the deer industry will need to assess whether an on-going scheme is necessary, and provide support if required. In time breeding structures are likely to evolve to provide sufficient genetic linkage without formal design. This will occur partly in response to the information from across-herd breeding values as breeders will seek to source the identified superior sires.

REFERENCES

Deer Industry New Zealand (2005) In Deer Industry New Zealand Annual Report 2003-2004, Deer Industry New Zealand, Wellington.

Geenty, K.G. (2000) Proc. NZ Soc. Anim. Prod. 60:180.

MAF (2005) In Situation and Outlook for New Zealand Agriculture & Forestry January 2005, Ministry of Agriculture and Fisheries, Wellington.

Newman, S.A., Dodds, K.G., Clarke, J.N., Garick, D.J. and McEwan, J.C. (2000) Proc. NZ Soc. Anim. Prod.60:195.

Rickards, P.A. (1997) Proc. Assoc. Advmt. Anim. Breed. Genet.12: 693.

THE ALLIANCE CENTRAL PROGENY TEST (CPT): AN EVALUATION OF SHEEP MEAT GENETICS IN NEW ZEALAND

A.W. Campbell¹, N.B. Jopson², N. J. McLean¹, K. Knowler¹, M. Behrent³, T. Wilson³, G. Cruickshank⁴, C.M. Logan⁵, P.D. Muir⁶, and J.C. McEwan¹

¹AgResearch, Invermay Agricultural Centre, Private Bag 50034, Mosgiel, New Zealand

²Abacus Biotech Limited, PO Box 5585, Dunedin, New Zealand

³Alliance Group Limited, 51 Don St, Invercargill, New Zealand

⁴Sheep Improvement Limited, PO Box 12146, Ahuriri, Napier, New Zealand (current address

Genequest Limited, RD 12, Havelock North, New Zealand)

⁵Lincoln University, PO Box 84, Lincoln University, New Zealand

⁶On-Farm Research Ltd, PO Box 1142, Hastings, New Zealand

SUMMARY

The Alliance Central Progeny Test (CPT) was established in 2001 with the aims of improving the carcass dollar value of sires under the Alliance grading system and establishing genetic links among sire referencing schemes. Over the past two seasons, 52 sires from both terminal and dual purpose breeds have been evaluated for meat genetics at the AgResearch Woodlands farm and in 2003 the Lincoln University Ashley Dene farm. All progeny were slaughtered through two Alliance plants in slaughters at monthly intervals at a threshold liveweight of 34kg. Carcass measurements included cut yield (measured by both ViaScan grading and commercial primal cut bone-outs) eye muscle area, GR, meat pH, meat and fat colour and days to reach an 18kg carcass weight. These measurements were used to calculate three key indices; days to kill, meat value and a combined meat and growth index. Results showed significant variation among sires for all key traits and that no single ram or breed dominated the top of the indices or breeding values. Public availability of findings has resulted in increased interest from commercial farmers in using performance recorded data for ram selection.

Keywords: Ovis aries, progeny test, meat evaluation, breeding values

INTRODUCTION

Central progeny test (CPT) schemes have been running in other countries for many years (Fogarty et al. 2002; Bibé et al. 2002). In New Zealand, the standard form of genetic evaluation has been across-flock analysis within a breed or breed group prior to the Alliance CPT. A number of these groups have been operating for over 30 years. However, genetic links between sire reference groups within a breed, or across breeds have been poor or non-existent. In 2001, the Alliance CPT was established with the twofold aim of providing information to improve the value of carcasses under the current and future Alliance grading systems, and to provide across-breed genetic linkages so that ram breeders have better information to select rams to meet their own breeding objectives.

MATERIALS AND METHODS

As at the end of 2004, two full cycles of slaughter progeny evaluations have been completed. For the first cycle, all animals were grazed at AgResearch’s Woodland Research Farm (WDL), while the second cycle included animals at Lincoln University’s Ashley Dene Pastoral Systems Research Farm (AD). Over the two cycles, a total of 52 sires and 1009 progeny have been evaluated for meat and growth traits, comprising 32 terminal sire and 20 dual purpose sires from 18 breeds. There were four link sires across years and three link sires across the two sites for the second cycle. Rams were nominated by various breeding groups either for artificial insemination (AI) or for natural mating. The ewe flock comprised either Coopworth ewes (WDL and AD) with a small number of ¼ East Friesian x Coopworth ewes (18% of ewes at WDL)

Farm management. All ewes were synchronised using CIDRs, and approximately half were mated using AI. The remainder were naturally single sire mated over a period of one week. Sufficient ewes were allocated to each sire to achieve a minimum of 20 lambs (accuracy of selection of 0.79 for a trait with a heritability of 0.30) based on expected success rates at each of the sites. A total of 46 of the 52 sires had at least 20 progeny evaluated (range 13 to 59).

Single bearing ewes were separated from twin and triplet bearing ewes prior to lambing. At WDL the smallest lambs in a triplet set were mothered onto single bearing ewes, while at AD triplets were not separated. At AD ram lambs were kept entire, while at WDL ram lambs were made into cryptorchids. Liveweights of all lambs were recorded at birth and docking. Lambs were weaned in early to mid December at 12-14 weeks of age. Lambs were drafted for slaughter at monthly intervals with the first draft at weaning, allowing for a maximum of four drafts. A liveweight target of at least 34 kg was set to draft lambs, with all remaining lambs going for slaughter at the last draft.

Slaughter procedure. Lambs were processed over the same period for each slaughter (commencing at 9 am). On the day of slaughter, data were collected for yield grades (ViaScan and the Meat Board National Grading system), hot carcass weight, GR on both sides of the carcass and a visual carcass conformation score. ViaScan was used to predict weights of lean in the hindleg, loin and shoulder. The following day (24 hours post slaughter) carcass pH and fat colour were measured in the chiller prior to boning out. In the boning room, measurements included: cold carcass weight, eye muscle area (EMA) and eye muscle colour (after allowing 30 minutes for the cut surface to bloom).

Data analysis. BLUP breeding values were estimated across breeds for the sires using a sire model, as the CPT sires were assumed to be unrelated. The variance components used were estimated from the CPT data itself. Components were estimated using an animal model applying a Restricted Maximum Likelihood procedure with ASREML (Gilmour et al. 1999). The direct genetic effects were fitted as a random effect in all models. Breed was not fitted as there were insufficient sires in each breed to be considered representative. The maternal genetic effects were fitted as a random effect for weaning weight (the direct and maternal genetic effects were assumed to be uncorrelated). Weaning weight was fitted in a model including terms for age of dam, birth rearing rank and sex and birth day deviation as a covariate. Weight of lean in the three primal cuts and other carcass linear or area measurements were analysed in a model that included sex and slaughter mob as fixed effects and cold carcass weight as a covariate. Meat quality measures (e.g. pH, meat and fat colour) were analysed fitting sex and slaughter mob as fixed effects. Differences between sites and years were handled by scaling the raw values so that the variances were equalised. EBVs were estimated for weaning weight, carcass weight, dressing percent, days to slaughter, eye muscle area, GR, ViaScan shoulder, loin and hindleg lean weight, meat pH, meat colour and fat colour. Economic indices were determined for meat value (the sum of ViaScan shoulder, hindleg and loin EBV weighted at 1:2:4), days to slaughter (economic value -$0.15/day), and a combined meat and growth index weighted 1:1.22 for the meat and growth indices, respectively. Genetic and phenotypic correlations between traits were estimated, but additional sires are needed to improve the initial estimates.

RESULTS

A summary of trait data is presented in Table 1. The average liveweight at weaning for slaughter progeny across years and sites was 31.4 kg. Carcass measurements were adjusted to an average carcass weight of 18.0kg, which required an average of 129 days to achieve.

Table 1. A summary of trait data collected from the Alliance Central Progeny test progeny

Of all of the traits, weaning weight, EMA, GR and fat colour were by far the most variable (CV% of 14.0, 11.7, 39.7 and 18.3%, respectively). Only minimal variation (<2%) was observed in DTS and meat pH. The heritabilities estimated for the data were mostly moderate with only loin lean being less than 0.20. Dressing percent, EMA, meat colour and fat colour all had high heritability estimates (>0.45), with EMA in particular being very high at 0.65.

DISCUSSION

In general, the heritabilities calculated from the data are somewhat higher than published estimates for comparable traits (Waldron et al. 1992; Fogarty et al. 2003). This was expected as the number of sires evaluated is still relatively low and estimates are likely to be biased upwards by between breed genetic variance. An effect of the high heritability estimates will be that the spread in breeding values would also be inflated, but this will not have changed the rankings between sires. EMA in particular was very much greater than the value of 0.31 reported by Waldron et al. (1992), although similar to the 0.59 reported by Bidé et al. (2002). The most likely explanation is the confounding of breed with sire in the analysis for a trait that is key to terminal sire selection, and often not in the selection objective for dual purpose breeds. Interestingly, the heritability for EMA was markedly greater than for loin lean which was not found by Waldron et al. (1992), possibly because of the likely presence of a major gene for EMA which the grading equipment may not be fully recognising.

Examination of the breeding values revealed that that there was considerable variation amongst sires, and that no one sire or breed dominated across all traits or indices. In presenting the results to industry and farmer groups, the importance of selecting rams appropriate to a farming environment and breeding objective was emphasised.

The Alliance CPT has now moved into a stage of maternal evaluation (mating 2004 and mating 2005) where ewe lambs produced from dual purpose sires have been retained. Standard maternal measurements relating to reproduction, wool growth and lamb survival will be collected as well as faecal egg count, first oestrus and gestation length. Carcass meat evaluation will be continued for all terminal sire progeny and the male progeny from dual purpose sires as part of the same programme.

SUMMARY

The programme has had several outcomes: the first being that the detailed recording of carcass traits in processing plants, including the results from ViaScan for individual cuts has highlighted that any existing NZ evaluation system needs to address the value of cuts rather than carcass based indices. Secondly, sire referencing groups and breeds are now becoming better linked genetically, thereby allowing national across breed evaluations to be undertaken. There have also been sire reference groups developed in breeds where little sire referencing has been done in the past. Finally, results for the top performing rams are publicly available which has led to an increase in interest from commercial farmers using information from performance recording systems to aid their ram selection decisions.

ACKOWLEDGEMENTS

This work was funded by The Alliance Group Ltd. The authors thank G. Greer, L. Hewitson, R. Wheeler, N. Wood, G. Anderson, J. Dunnett, N. Jay, M. Colling, S. Duncan, A. Pont, S. Duncan and N. Lake for technical assistance. We also thank the many breeders who have made rams available for the Alliance CPT.

REFERENCES

Bibé, B., Brunel, J.C., Bourdillon, Y., Laradoux, D., Gordy, M.H., Weisbecker, J.L. and Bouix, J. (2002) Proc. 7th World Congr. Gen. Appl. Livest. Prod. 31: 335.

Fogarty, N.M., Cummins, L., Gaunt, G., Hocking-Edwards, J.E. and Edwards, N.J. (2002) Proc. 7th World Congr. Gen. Appl. Livest. Prod. 29: 449

Fogarty, N.M., Safari, E., Taylor, P.J. and Murray, W. (2003) Aust. J. Agric. Res. 54:715.

Gilmour, A.R., Cullis, B.R., Welham, S.J. and Thompson, R. (1999) ASREML Reference Manual, NSW Agriculture Biometric Bulletin No. 3.

Waldron, D.F., Clarke, J.N., Rae, A.L., Kirton, A.H. and Bennett, G.L. (1992) N.Z. J. Agric. Res. 35:287.

GENETIC RELATIONSHIPS BETWEEN CARCASS QUALITY AND WOOL PRODUCTION TRAITS IN AUSTRALIAN MERINO RAMS

J.C. Greeff, G. Cox, L. Butler, and M. Dowling.

Australian Sheep Industry CRC and Department of Agriculture of Western Australia, 10 Dore Street, Katanning 6317 Australia

SUMMARY

The heritability of carcass quality traits and the genetic correlations between wool production and these carcass quality traits were estimated on 1216 Merino rams. The results showed that carcass quality traits are moderately heritable but pH appears to be lowly heritable. Yellowness and redness of L. dorsi had very low heritability estimate. Moderate negative genetic relationships were found between the fat traits and CFW that were independent of body or carcass weight. Keywords: Merino, carcass traits, wool, genetic correlations

INTRODUCTION

The increasing importance of meat production has focused attention on the Merino because this breed contributes about two thirds of the genes of all meat sheep in Australia.. The Sheep CRC has therefore initiated a research project to develop an understanding of the genetic factors affecting wool and meat traits. Fogarty et al. (2003) and Greeff et al. (2003) have estimated the heritability of meat and carcass quality traits in Merino rams. They found moderate genetic variation in a range of meat traits and concluded that the heritability estimates in Merinos are very similar to those of other meat breeds. They also published genetic correlations between wool and meat traits but the results of Greeff et al (2003) differs from that of Fogarty et al (2003) in pH, CFW and fat measurements. Since then additional data have come available and this study updates the genetic correlations of Greeff et al (2003) in a Western Australian environment.

MATERIAL AND METHODS

Animals. Data was collected from the Katanning Merino resource flocks managed by the Department of Agriculture of Western Australia. This flock consists of four selection lines i.e. meat, staple strength, wool and a control line. Carcass records on 1216 progeny from 112 sires and 4261 ultrasound muscle and subcutaneous fat and wool records from 126 sires were available. The animals were born from 1999 to 2002. They were weighed (LWT) at 16 months of age and scanned for subcutaneous fat thickness (Scanfat) and eye muscle (Longissimus dorsi) depth at the C site (Scanmus) on the 12th rib, 45 mm off the midline using an Aloka 500 ultrasound machine. The week following scanning they were shorn. Greasy fleece weights were recorded and midside wool samples collected and tested for clean yield (Yield) (to obtain clean fleece weight (CFW)), fibre diameter (FD), coefficient of variation of FD (CV), staple strength (SS) and fibre curvature (Curva).

Only rams born from 2000 to 2002 were slaughtered. As it is well known that shearing can result in a loss of body weight, the rams were placed on a pre-slaughter feeding regime prior to slaughter. The animals slaughtered in 2001 were fed a commercial pelleted diet of 10.8 MJ/kg and 15.5 per cent digestible protein for two weeks prior to slaughter. In spite of this feeding regime, these rams lost 4 kg body weight from 63 kg live weight just after shearing to 59 kg by slaughter over the 6 weeks period. The animals that were slaughtered in 2002 and in 2003 were fed lupins and barley for 5 weeks prior to slaughter which resulted in them maintaining body weight.

A day prior to slaughter the rams were transported to a commercial abattoir. In 2001 and 2002 the rams were slaughtered in an abattoir 120 kilometers from the research farm but in 2003 they were slaughtered in an abattoir 60 kilometers from the research farm. Hot carcass weight (HCWT) and tissue depth at the GR site, 110mm from the midline at the 12th rib (FatGR) were measured immediately after slaughter. The carcasses were electrically stimulated and stored in chillers at 4 degrees Celsius overnight and measured approximately 24 hours post slaughter. The rams slaughtered in 2003 were not electrically stimulated but were left for 48 hours in the chillers at 4°C to allow ultimate pH to reach its maximum level (Pethick, pers. comm.). Subcutaneous fat at the C-site (FatC), eye muscle depth (EMD) and width (EMW) were measured using a ruler. Eye muscle area (AREA) was the product of EMD and EMW. In 2002 these measurements were taken on the 13th rib instead of the normal 12th rib. This deviation from the standard site was a requirement by the abattoir not to damage the carcasses in order to comply with their clients’ specifications. Meat pH was measured with a WTW pH 330 Pocket pH mV meter. Muscle colour (relative lightness (L), relative redness (a) and relative yellowness (b)) was measured at the same site with a Minolta Chroma meter CR300.

Statistical analysis. Scanfat was regressed against FatC and Scanmus against EMD using Genstat (2003) to determine the accuracy of the scanned measurements while the genetic analysis were performed with ASREML (Gilmour, 2002) and different models were fitted for the different traits. All models included selection line, year of birth, birth type, age of dam, and all first order interactions as fixed factors. Day of lambing was used as a linear covariate for HCWT and animal and permanent maternal environmental effects were fitted as random effects. Permanent maternal environmental variation was not significant for HCWT or shearing weight in this study and was therefore left out of the model. The number of animals in the numerator relationship matrix was 4763. LWT was used as a linear covariate in the model for the wool traits, while HCWT was used as a covariate for EMD, EMW, FatGR, FatC, pH, L, a and b colour measurements to estimate genetic and environmental variances. Bivariate analyses were carried out to estimate the genetic correlations between the traits. The final models only included significant factors. These covariates were included in the models to determine the genetic relationships between the meat and wool traits independent of HCWT or LWT.

RESULTS AND DISCUSSION

Average LWT at scanning was 58.8 kg. Selection line, year of birth and birth type had a significant effect (P<0.05) on HCWT, while year of birth had a significant (<0.01) effect on pH. Year of birth and birth type had a significant effect on Scanmus and Scanfat when LWT at scanning was included as a linear covariate. Year of birth and birth type also had a significant effect on EMD and on FatGR using HCWT as a covariate. Relative redness (a) was not affected by any of the fixed environmental factors or HCWT. HCWT had a significant (P<0.001) positive relationship with EMD, EMW, FatC and FatGR but a negative relationship with lightness (L) and yellowness (b).

Means, standard deviations (SD) and heritability estimates (± se) of the traits are shown in Table 1. It is clear that the carcasses were quite lean at scanning and at slaughter but the fat traits showed quite a large amount of phenotypic variation, similar to that of Fogarty et al. (2003).

Table 1. Means, standard deviation, coefficient of variation and heritability of carcass traits in Merino rams.

All the traits in this study were low to moderately heritable (Table 1). In most cases the heritability estimates of the carcass traits were reasonably similar to that of Fogarty et al. (2003) and Clarke et al. (2003). The meat colour traits a and b exhibited a lot of variation in this study but they had low heritability values. The heritability estimate of pH was lower (0.18 vs 0.27) than reported by Fogarty et al. (2001). As the opportunity to change traits genetically depends on the heritability and the phenotypic variation, it appears that there is not a lot of scope to change meat colour and pH of rams.

Table 2. Slopes of Scanfat against FatC and Scanmus against eye muscle depth (EMD) on the carcass measured in 2001, 2002 and 2003 on Merino rams at hogget age

Table 2 shows that in most years Scanfat and Scanmus explained a relatively small proportion of the variation of the same trait on the carcass. It varies from year to year which can be explained by year effects, different scanners and the length of time between scanning and slaughtering. However, the results indicate that it is difficult to get an accurate assessment of fat cover and eye muscle depth using ultrasound scanning methods in Merino sheep.

The genetic correlations between wool and carcass quality traits are indicated in Table 3. Most estimates have relatively large standard errors especially the colour traits. There was a negative relationship between CFW and AREA and between CFW and the fat traits, while FD was moderately negatively correlated with EMA. CV and Yield was both negatively correlated with both fat traits while curvature was moderately positively correlated with FatC and AREA.

Table 3. Heritability estimates of wool traits and genetic correlations (rg ± SE) between wool traits and carcass quality traits in Merino rams.

CONCLUSION

This study indicates that ultrasonically scanned fat and muscle depth are not good indicators of fat cover and eye muscle depth in live Merino sheep. Scanfat and Scanmus independent of live weight, and carcass and meat quality traits, independent of hot carcass weight, are heritable traits in Merino sheep which implies that genetic improvement through selection should be possible. Selection for increased CFW will result in less muscle and less subcutaneous fat, selection for decreased FD will result in more muscular animals, while selection for increased SS will result in fatter animals.

REFERENCES

Clarke, B.E., Brown, D.J. and Ball, A.J. (2003) Proc. Assoc Advmt. Anim. Breed. Genet. 15:326

Devine, C.E. and Chrystall, B.B. (1988) Dark Cutting in Cattle and Sheep, Proc. of an Australian

workshop. Eds: S.U. Fabiansson, W.R. Shothouse and R.D. Warner.

Fogarty, N.M. Safari, E. Taylor, P.J. and Murray, W. (2003) Aust. J. Agric. Res. 54:715.

Gardner, G.E., Kennedy, L., Milton, J.T.B. and Pethick, D.W. (1999). Aust. J. Agric. Res. 50:175.

Genstat (2003) VSN International Wilkinson House, Oxford, UK.

Gilmour, A.R., Cullis, B.R., Welham, S.J. and Thompson, R. (1999). ASREML manual. NSW Agriculture Biometric Bulletin No.3.

Greeff, J.C., Davidson, R, and Skerrit, J.W. (2003) Proc. Assoc Advmt. Anim. Breed. Genet. 15:330.

MERINO STRAINS WITH HIGH WOOL PRODUCTION HAVE LOWER LIFETIME REPRODUCTION RATES

J.C. Greeff

Australian Sheep Industry CRC

Western Australian Department of Agriculture, 10 Dore St, Katanning WA 6317

SUMMARY

Previous studies have shown a negative genetic correlation between fleece weight and total weight of lamb weaned over three lambing opportunities. This study examined the reproductive performance of six different Merino strains in a Mediterranean environmental. Hogget clean fleece weight had a poor relationship with lamb output but when adjusted for body weight, it had a good relationship across strains except for the fine wool strain. The results indicate that strains that have higher wool production per unit body weight at hogget age will have a lower reproduction rate over their lifetime. This implies that strong selection for wool production will affect reproductive rate negatively. If strong emphasis is also placed on reducing fibre diameter, then this may exacerbate the negative effect. This may be related to a lower level of total fat reserves found in high wool producing sheep. Keywords: reproduction, strains, wool, body weight

INTRODUCTION

Herselman et al. (1998) have shown that wool production potential, defined as the proportion of wool relative to body weight, is phenotypically negatively correlated with total weight of lamb weaned over the first three lambing opportunities in three different flocks in different environments. Ingham and Ponzoni (2002) and Cloete et al. (2004) reported negative genetic correlations (from -0.41 to -0.05) between clean fleece weight and number of lambs weaned per ewes joined in Merino sheep. However, Cloete et al. (2002) have found a positive genetic correlation between total weight of lamb weaned over three lambing opportunities and hogget clean fleece weight but a negative genetic correlation of -0.24 when clean fleece weight was expressed as a proportion of hogget body weight. These studies imply that selection for an increase in fleece weight whilst maintaining body weight or where the genetic gain in body weight is relatively slower than the gain in fleece weight, will result in a reduction in reproduction efficiency in subsequent generations. With current high meat prices, this is an undesirable trend and this paper focuses on the performance of six different Merino strains to determine whether this may have happened in a Western Australian Mediterranean environment.

MATERIALS AND METHODS

The strains available for this study were Peppin, Collinsville, Bungaree and an Australian Merino Society (AMS) strain. Samples of the first three mentioned strains were sourced in 1981 from four representative studs per strain as described by Lewer et al (1993). The AMS strain was sourced from four representative studs in 1986. In addition, a fine wool strain and a heavy cutting medium wool strain were obtained from CSIRO in 1992.

The complete data set, of which a subset was used for the analysis, comprise 4838 matings records over five years. The animals were born and raised from 1993 to 1995 on the Department of Agriculture of Western Australia’s research station at Katanning, Western Australia. Single sire matings were carried out and full pedigrees were recorded on all animals. Complete records on the reproductive performance of each ewe were collected. Survival rate and weaning weights of lambs were also recorded, whilst clean fleece weight (CFW), body weight (BWT) and fibre diameter (FD) were collected at hogget age. No selection was practised within each flock and representative sires were obtained from the original ram sources. Ewe replacements were selected on a random basis but animals with black wool or with anatomical abnormalities were culled. The same ewe age structure was maintained across strains during the experimental period as far as