Comparison of threshold models forestimation of genetic parameters of litter size in Makuie sheep

Genetic parameters for litter size trait in Makuie sheep were estimated using Bayesian method and different models. For this purpose, six threshold models were fitted. Different models with direct genetic effects, maternal and permanent environmental effects were implemented. Effects after significance test (P<0.05) by logistic procedure of SAS software were included in these models (herd and year of lambing). Variance components of litter size trait were estimated by Thrgibbs1f90 software. The most appropriate model was determined based on deviance information criterion (DIC). Model 4 was recognized as the best model for the genetic analysis of litter size trait in Makuie sheep. The meanof the posterior distribution of estimated additive and maternal heritability for best model (model 4) were 0.03% (0.004) and 0.109(0.017), respectively. Estimated genetic parameters of litter size in this studyshowed that for genetic gain in this trait more attention should be paid to maternal effect and improve environmental condition.