A. A. Swan

Within- and across-flock genetic relationships for breech flystrike resistance indicator traits.

Flystrike is a major cost for the Australian sheep industry. The industry is currently implementing selection strategies for flystrike resistance to reduce the need for breech flystrike prevention. The following indicator traits are used to select for breech flystrike resistance: wool cover, skin wrinkle on the body and breech, scouring (dags) and wool colour. The aims of this study were to estimate genetic correlations between these indicator traits and production traits using the Sheep Genetics database, to distinguish between within- and across-flock genetic relationships, and to quantify responses to selection using indexes that include breech wrinkle as a proxy trait for flystrike resistance. Breech flystrike indicator traits are all heritable; however, there are significant antagonisms between wrinkle score and some production traits, primarily fleece weight and fibre diameter. Thus, simultaneous improvement in both flystrike resistance and production will be most efficient when index selection is used. Our results show that, depending on the level of emphasis placed on breech wrinkle in the index, reductions in breech wrinkle score of 0.4–0.9 units can be achieved over a 10-year period. As across-flock relationships are generally stronger than within-flock relationships, breeders will be able to take advantage of this additional variation, depending on the relative merit of their flocks. Therefore, ram breeders should combine within-flock selection with across-flock selection where possible. Sheep Genetics released early breech wrinkle Australian Sheep Breeding Values in September 2009 to assist Merino breeders in making faster progress towards reducing breech wrinkle by using flock selection.

Genetic variation within and between subpopulations of the Australian Merino breed

Genetic variation within and between Australian Merino subpopulations was estimated from a large breeding nucleus in which up to 8500 progeny from over 300 sires were recorded at eight sites across Australia. Subpopulations were defined as genetic groups using the Westell–Quaas model in which base animals with unknown pedigree were allocated to groups based on their flock of origin if there were sufficient ‘expressions’ for the flock, or to one of four broad sheep-type groups otherwise (Ultra/Superfine, Fine/Fine-medium, Medium/Strong, or unknown). Linear models including genetic groups and additive genetic breeding values as random effects were used to estimate variance components for 12 traits: yearling greasy and clean fleece weight (ygfw and ycfw), yearling mean and coefficient of variation of fibre diameter (yfd and ydcv), yearling staple length and staple strength (ysl and yss), yearling fibre curvature (ycuv), yearling body wrinkle (ybdwr), post-weaning weight (pwt), muscle (pemd) and fat depth (pfat), and post-weaning worm egg count (pwec). For the majority of traits, the genetic group variance ranged from approximately equal to two times larger than the additive genetic (within group) variance. The exceptions were pfat and ydcv where the genetic group to additive variance ratios were 0.58 and 0.22, respectively, and pwec and yss where there was no variation between genetic groups. Genetic group correlations between traits were generally the same sign as corresponding additive genetic correlations, but were stronger in magnitude (either more positive or more negative). These large differences between genetic groups have long been exploited by Merino ram breeders, to the extent that the animals in the present study represent a significantly admixed population of the founding groups. The relativities observed between genetic group and additive genetic variance components in this study can be used to refine the models used to estimate breeding values for the Australian Merino industry.

Genetic importance of fat and eye muscle depth in Merino breeding programs

Australian Merino breeders have traditionally selected animals for breeding predominately on the basis of wool characteristics. Over recent decades, an increasing proportion of Merino breeders are interested in producing a ewe that can be used for prime-lamb production, but that also performs well for wool characteristics. Correlations between ultrasound carcass traits and other traits such as wool, internal parasite resistance and reproduction traits, are not very well known. The aims of this study were three-fold: (1) to estimate the genetic relationships between ultrasound carcass traits and wool, internal parasite resistance and reproduction traits, (2) to determine the value of recording ultrasound carcass traits in Merino breeding programs, and (3) to evaluate the impact of improving ewe genetic merit for fatness on their reproduction performance. Ultrasound fat and eye muscle depth had small to moderate genetic correlations with most traits, with positive correlations observed for bodyweight, fibre diameter, fibre curvature and reproduction, and negative correlations observed for fleece weight, fibre diameter coefficient of variation, worm egg count and breech wrinkle. As expected on the basis of these genetic correlations, estimated breeding values for fat depth of ewes had a positive association with their observed reproduction performance, but the effect explained only minimal variation in reproductive performance, and was extremely variable among flocks and years. A range of measurement scenarios was investigated for three standard MERINOSELECT indexes. Measuring fat and eye muscle depth resulted in 3%, 4% and 21% additional economic index gain for the fine, medium and dual purpose indexes, respectively, whereas measuring reproduction traits directly resulted in 17%, 27% and 45% additional gain in the economic index. Dual purpose index gains benefited more from measuring ultrasound carcass traits as it is the only index with a direct economic value placed on carcass traits. Measuring fat and eye muscle depth also resulted in a greater reduction in worm egg count. The results indicated that desirable genetic progress can be made in wool, ultrasound carcass, internal parasite resistance and number of lambs born and weaned simultaneously using multiple trait selection to account for the mix of favourable and unfavourable correlations between these traits. These results also demonstrated that the best method to maximise economic gain is to measure as many traits (or closely correlated traits) in the breeding objective as possible.

Integration of genomic information into beef cattle and sheep genetic evaluations in Australia

Genomic information has the potential to change the way beef cattle and sheep are selected and to substantially increase genetic gains. Ideally, genomic data will be used in combination with pedigree and phenotypic data to increase the accuracy of estimated breeding values (EBVs) and selection indexes. The first example of this in Australia was the integration of four markers for tenderness into beef cattle breeding values. Subsequently, the availability of high-density single nucleotide polymorphism (SNP) panels has made selection using genomic information possible, while at the same time creating significant challenges for genetic evaluation with regard to both data management and statistical modelling. Reference populations have been established in both the beef cattle and sheep industries, in which an extensive range of phenotypes have been collected and animals genotyped mainly using 50K SNP panels. From this information, genomic predictions of breeding value have been developed, albeit with varying levels of accuracy. These predictions have been incorporated into routine genetic evaluations using three approaches and trial results are now available to breeders. In the first, genomic predictions have been included in genetic evaluation models as additional traits. The challenges with this method have been the construction of consistent genetic covariance matrices, and a significant increase in computing time. The second approach has been to use a selection index procedure to blend genomic predictions with existing EBVs. This method has been shown to produce very similar results, and has the advantage of being simple to implement and fast to operate, although consistent genetic covariance matrices are still required. Third, in sheep a single-step analysis combining a genomic relationship matrix with a standard pedigree-based relationship matrix has been used to estimate breeding values for carcass and eating-quality traits. It is likely that this procedure or one similar will be incorporated into routine evaluations in the near future. While significant progress has been made in implementing methods of integrating genomic information in both beef and sheep evaluations in Australia, the major challenges for the future will be to continue to collect the phenotypes needed to derive accurate genomic predictions, and in managing much larger volumes of genomic data as the number of animals genotyped and the density of markers increase.

Genetic parameters for liveweight, wool and worm resistance traits in multi-breed Australian meat sheep. 1. Description of traits, fixed effects, variance components and their ratios

Sheep breeders in Australia that focus on lamb production simultaneously breed sheep that have higher growth rate, improved carcass quality and are resistant to internal parasites. The objective of this study was to estimate genetic parameters for 11 traits recorded in Australian meat sheep, covering liveweight, carcass and internal parasite resistance traits. As the population of meat sheep in this database have become increasingly crossbred this study also investigates the genetic variation within and between breeds. The data comprised 1 046 298 animals from 149 Poll Dorset, 17 Suffolk, 24 Texel and 118 White Suffolk flocks. The results are averages of analyses of 10 datasets constructed by randomly sampling 25% of these flocks. There was considerable genetic variation in all traits analysed: the lowest heritabilities (0.12) were found for weaning weight and the highest heritabilities (0.31–0.32) for eye muscle depth. There were also significant differences between breeds for most traits, which breeders appear to be utilising through crossbreeding. Direct heterosis effects were small and only significant for the liveweight traits ranging from 2% to 3.4% of the phenotypic means. Maternal heterosis was not significant for any trait studied. The inclusion of heterosis effects in the model did not significantly influence the estimated genetic parameters. The results from this study have been used to review the genetic parameters used in the LAMBPLAN routine genetic evaluations conducted by Sheep Genetics.

Multiple-trait QTL mapping and genomic prediction for wool traits in sheep

BackgroundThe application of genomic selection to sheep breeding could lead to substantial increases in profitability of wool production due to the availability of accurate breeding values from single nucleotide polymorphism (SNP) data. Several key traits determine the value of wool and influence a sheep’s susceptibility to fleece rot and fly strike. Our aim was to predict genomic estimated breeding values (GEBV) and to compare three methods of combining information across traits to map polymorphisms that affect these traits.MethodsGEBV for 5726 Merino and Merino crossbred sheep were calculated using BayesR and genomic best linear unbiased prediction (GBLUP) with real and imputed 510,174 SNPs for 22 traits (at yearling and adult ages) including wool production and quality, and breech conformation traits that are associated with susceptibility to fly strike. Accuracies of these GEBV were assessed using fivefold cross-validation. We also devised and compared three approximate multi-trait analyses to map pleiotropic quantitative trait loci (QTL): a multi-trait genome-wide association study and two multi-trait methods that use the output from BayesR analyses. One BayesR method used local GEBV for each trait, while the other used the posterior probabilities that a SNP had an effect on each trait.ResultsBayesR and GBLUP resulted in similar average GEBV accuracies across traits (~0.22). BayesR accuracies were highest for wool yield and fibre diameter (>0.40) and lowest for skin quality and dag score (<0.10). Generally, accuracy was higher for traits with larger reference populations and higher heritability. In total, the three multi-trait analyses identified 206 putative QTL, of which 20 were common to the three analyses. The two BayesR multi-trait approaches mapped QTL in a more defined manner than the multi-trait GWAS. We identified genes with known effects on hair growth (i.e. FGF5, STAT3, KRT86, and ALX4) near SNPs with pleiotropic effects on wool traits.ConclusionsThe mean accuracy of genomic prediction across wool traits was around 0.22. The three multi-trait analyses identified 206 putative QTL across the ovine genome. Detailed phenotypic information helped to identify likely candidate genes.

Genetic parameters for liveweight, wool and worm resistance traits in multi-breed Australian meat sheep. 2. Genetic relationships between traits

Australian sheep breeders aim to simultaneously breed sheep that have high growth rate, superior carcass quality and are resistant to internal parasites. The objective of this study was to estimate genetic and phenotypic relationships between 11 traits recorded in Australian meat sheep, covering liveweight, carcass and internal parasite resistance traits. The data comprised 1 046 298 animals from 149 Poll Dorset, 17 Suffolk, 24 Texel and 118 White Suffolk flocks. Within all trait groups, the different age expressions were moderately to highly correlated. The genetic correlations between liveweight with both fat and eye muscle depth were generally negative ranging between –0.10 and –0.42. Fat and eye muscle depth were moderately correlated (0.27–0.59). Results indicate that selection for an increase in liveweight will have a negative effect on fat and eye muscle depth. The negative correlations of ultrasound scan traits and worm egg count indicates that animals with high genetic merit for fat and eye muscle depth are less prone to infection from internal parasites. The results from this study have been used to review the genetic correlations used in the LAMBPLAN routine genetic evaluations conducted by Sheep Genetics.

Genomic prediction of weight and wool traits in a multi-breed sheep population

The objective of this study was to predict the accuracy of genomic prediction for 26 traits, including weight, muscle, fat, and wool quantity and quality traits, in Australian sheep based on a large, multi-breed reference population. The reference population consisted of two research flocks, with the main breeds being Merino, Border Leicester (BL), Poll Dorset (PD), and White Suffolk (WS). The genomic estimated breeding value (GEBV) was based on GBLUP (genomic best linear unbiased prediction), applying a genomic relationship matrix calculated from the 50K Ovine SNP chip marker genotypes. The accuracy of GEBV was evaluated as the Pearson correlation coefficient between GEBV and accurate estimated breeding value based on progeny records in a set of genotyped industry animals. The accuracies of weight traits were relatively low to moderate in PD and WS breeds (0.11–0.27) and moderate to relatively high in BL and Merino (0.25–0.63). The accuracy of muscle and fat traits was moderate to relatively high across all breeds (between 0.21 and 0.55). The accuracy of GEBV of yearling and adult wool traits in Merino was, on average, high (0.33–0.75). The results showed the accuracy of genomic prediction depends on trait heritability and the effective size of the reference population, whereas the observed GEBV accuracies were more related to the breed proportions in the multi-breed reference population. No extra gain in within-breed GEBV accuracy was observed based on across breed information. More investigations are required to determine the precise effect of across-breed information on within-breed genomic prediction.

Pregnancy scanning can be used as a source of data for genetic evaluation of reproductive traits of ewes

Reproductive traits generated from mothering up lambs to ewes (n = 59 603 records) were compared with data resulting from pregnancy scanning (n = 46 663 records), to examine the consistency between the two data sources for deriving specific reproductive traits and to estimate genetic parameters. The reproductive traits considered were fertility (FERT: 0/1) of ewes joined, total litter size (LSIZE: lambs born), the number of lambs surviving at weaning (LSIZEW: lambs weaned) and the percentage of lambs surviving (LSURV = LSIZEW/LSIZE) for ewes that lambed, along with the composite traits number of lambs born (NLB) and number weaned (NLW) for ewes joined. Corresponding trait values were derived from pregnancy scan data (FERT_S, LSIZE_S and NLB_S) for comparison, and were classified as inconsistent if the trait values did not match from scanning and lambing records. Data were obtained from four flocks, representing different time frames, locations, management and breeds or bloodlines. Each flock recorded scan data separately from lambing outcomes. Genetic parameters were estimated separately within each flock. Average levels of inconsistency between scan- and lambing-data values varied between 4.6% and 14.8% across flocks, tending to be highest (9.1–18.5%) for litter size of ewes scanned with multiple fetuses, and lowest (0.29–7.3%) for assignment of fertility. Inconsistencies did not have a significant impact on estimates of trait heritabilities, suggesting recording errors were independent of genetic merit. In three flocks, the genetic correlations (ra) between comparable traits derived from the different data sources were not different from unity (ra ≥ 0.99) even when phenotypic correlations (rp) were lower (rp ≥ 0.84). In the flock with the highest inconsistency rate between data sources, the range in ra varied between 0.60 (fertility) and 1.0 (litter size). Therefore, pregnancy scan data can be directly substituted for reproductive traits traditionally based on lambing data, but attention should be paid to ensuring accuracy of the data sources used. Scan data also provide no information on lamb-survival outcomes after birth, so does not constitute complete data on reproductive outcomes. Genetic evaluation systems might also benefit from fine tuning for scale-induced effects (due to litter size) on parameters to improve the accuracy of across flock prediction of breeding values for reproductive traits.

Genetic and phenotypic parameters for reproduction, production and bodyweight traits in Australian fine-wool Merino sheep

The present study estimated phenotypic and genetic relationships between wool production, reproduction and bodyweight traits in Australian fine-wool Merino sheep. The data for the study originated from the CSIRO Fine Wool Project, Armidale, Australia. Data on wool characteristics, measured at ~10 and 22 months of age, bodyweight and several reproduction traits across consecutive lambing opportunities were analysed. The genetic correlations were moderately negative between fibre diameter measured as yearling and adult, and lamb survival (rg = –0.34 ± 0.15 and rg = –0.28 ± 0.14 respectively) and total number of lambs weaned (rg = –0.32 ± 0.21 and rg = –0.40 ± 0.21 respectively). The genetic correlations of yearling and adult greasy and clean fleece weights with number of lambs weaned and fecundity showed moderately to highly negative relationships and a moderately negative correlation with the number of fetuses at pregnancy scanning. Phenotypic correlations between reproduction and wool production traits were estimated to be zero, with the exception of bodyweight showing low to moderate positive phenotypic correlations with total number of lambs born and weaned. Genetic variances were generally low for the reproduction traits and resulted in low heritability estimates (from h2 = 0.03 ± 0.01 to h2 = 0.12 ± 0.13), with the exception of total number of lambs born (h2 = 0.25 ± 0.03). The study indicated that parameter estimation and trait definition of lifetime reproduction records require careful consideration and more work in this area is required.