Genetic selection for increased mean and reduced variance of twinning rate in Belclare ewes. (16th June 2015)

• Record Type: Journal Article
• Authors: Cottle, D.J.; Gilmour, A.R.; Pabiou, T.; Amer, P.R.; Fahey, A.G.
• Is Part Of: Journal of animal breeding and genetics. Volume 133:Number 2(2016)
• Page Start: 126
• Page End: 137
• Publication Date: 2015-06-16
• Subjects: Permanent environmental effect -- REML -- selection index -- sire model -- Threshold model -- twinning -- variance component; Livestock -- Breeding -- Periodicals; Livestock -- Genetics -- Periodicals; 636.0820
• DOI: 10.1111/jbg.12173
• Languages: English
• ISSNs: 0931-2668

Abstract

Summary: It is sometimes possible to breed for more uniform individuals by selecting animals with a greater tendency to be less variable, that is, those with a smaller environmental variance. This approach has been applied to reproduction traits in various animal species. We have evaluated fecundity in the Irish Belclare sheep breed by analyses of flocks with differing average litter size (number of lambs per ewe per year, NLB) and have estimated the genetic variance in environmental variance of lambing traits using double hierarchical generalized linear models (DHGLM). The data set comprised of 9470 litter size records from 4407 ewes collected in 56 flocks. The percentage of pedigreed lambing ewes with singles, twins and triplets was 30, 54 and 14%, respectively, in 2013 and has been relatively constant for the last 15 years. The variance of NLB increases with the mean in this data; the correlation of mean and standard deviation across sires is 0.50. The breeding goal is to increase the mean NLB without unduly increasing the incidence of triplets and higher litter sizes. The heritability estimates for lambing traits were NLB, 0.09; triplet occurrence (TRI) 0.07; and twin occurrence (TWN), 0.02. The highest and lowest twinning flocks differed by 23% (75% versus 52%) in the proportion of ewes lambing twins. Fitting bivariate sire models to NLB and the residual from the NLB model using a double hierarchical generalized linear model (DHGLM) model found a strong genetic correlation (0.88 ± 0.07) between the sire effect for the magnitude of the residual (VE ) and sire effects for NLB, confirming the general observation that increased average litter size is associated with increased variability in litter size. We propose a threshold model that may help breeders with low litter size increase the percentage of twin bearers without unduly increasing the percentage of ewes bearing triplets in Belclare sheep.