Comparative study of different sire evaluation procedures under two models using crossbred progeny

Abstract

The Friesian sires used for producing halfbreds at Jabalpur, Rahuri and Hisar under AICRP on Cattie were evaluated under 2 models. Model was of within genetic group and Model 2 was over the genetic groups. Model 2 was efficient over M 1 as R² under the formula was 72.1 % as compared to 60% of the later model. There were no significant differences among ranking, for the different methods, viz. LS, PD, SRIS
and BLUP, under hoth the models. Accuracy results indicated BLUP to he significantly lower to the other methods under both the models. The efficiency measured as SD of the constants showed SRLS to he the most efficient under both the models, The relative efficiency results showed that all the methods of evaluation under M 1 were   nearly as good as LS but under M 2 none of the methods was as efficient as LS. Rdative accuracy measured as ratio of effective progeny si7e (n) for realized accuracy to that of actual progeny , size(K) showed SRLS under M 1 and LS under M 2 to be twice as accurate. Similar relative accuracy estimates to achieve specific superiority (20%) over population mean showed BLUP to be the most accurate for both the models.