Published online by Cambridge University Press: 18 August 2016
Segregation analyses were applied to data from an experimental sheep flock to investigate the presence of a major gene affecting litter size. The data set contained 14 years of litter size data, with up to five parities per ewe, from Cheviot sheep carrying the putative Thoka fecundity gene from Icelandic sheep. Segregation analyses were performed using a Markov chain Monte Carlo method implemented using Gibbs sampling. Uniform priors were initially used for estimating variance components, the gene effect and fixed effects in the data. Genotypes in the base generation were assumed known based on the use of the imported Icelandic donor semen from the founder rams. The use of alternative priors (naïve and inverse-gamma distributions) for the variance components did not significantly affect the results, demonstrating the data to be sufficiently powerful for the analyses used. Segregation analyses detected a major gene for litter size in the Thoka Cheviot flock increasing litter size by 0·70 lambs per ewe lambing for a single copy of the gene. When the analysis was repeated without fixing the genotypes in the base population, the analyses predicted a different genotype than that previously used for one of the founder rams and suggested the major gene to be segregating in the Cheviot founder animals prior to the introduction of the Thoka rams. A liability threshold analysis was also applied to the data. As identified in other studies, the threshold analysis overestimated the heritability, but the estimated major gene effect was not significantly different from other analyses. The results confirm the segregation of the Thoka gene in a Cheviot flock and highlight the statistical method as a useful tool for identifying carrier animals to be used for future matings.