Estimation of variance components for milk yield of sahiwal cattle using repeatability animal models

Abstract

Estimates of variance components and genetic parameters for lactation milk yield (LMY) were obtained using single-trait 4 different repeatability animal models based on algorithm of derivative free restricted maximum likelihood (DFREML). These repeatability animal models include; the simple animal model (AM I), animal model with sire×herd interaction effect (AM II), animal model with maternal effect (AM III) and animal model with sire×herd interaction effect and maternal effect (AM IV). The data included 3226 first five lactation records of 1223 cows sired by 45 sires and calved between 1968 and 2006. Using AM I, AM II, AM III and AM IV the direct estimates of heritabilities for lactation milk yield were 0.434, 0.359, 0.357 and 0.315, respectively. Sire×herd interaction explained 8.8% and 8.2% of total variation in LMY using AM II and AM IV, respectively whereas maternal effects explained 7.2% and 4.4% of total variation in LMY using AM III and AM IV, respectively. The results indicated that along with direct genetic effects, both sire by herd interaction effects and maternal effects should be included in a selection programme for lactation milk yield. The impact of interaction effects was more significant and consistent. Although, all the animal models had almost equal efficiency (> 99%) and accuracy (Log L value) but AM IV helped in proper partitioning of additive genetic variance and prevented the inflation of direct estimate of heritability by keeping almost fixed magnitude of error variance. Cows at Karnal herd had highest LMY (2078.89 kg) whereas cows at Luknow herd had lowest LMY (1441.83 kg).