425
Efficiency of Random Regression model over conventional univariate animal model for estimation of breeding values for first lactation 305-day milk yields in Mehsana buffaloes
Analysis was done on 71,536 test-day milk yield records (TD) of 7,378 Mehsana buffaloes sired. Random Regression model (M-1) was used for estimating sire breeding values (BVs) for first lactation 305-day milk yields (305-DMY) using Legendre polynomials of order three with year-season of calving and herd-year-month of record as fixed effects, age at first calving and days in milk as co-variables; animal and permanent environmental effects as random. For model M-2, 305-DMY were predicted on TD and their BVs were estimated. M-1 yielded daily heritability ranging from 0.11 (day 305) to 0.27 (day 117). Maximum genetic variability was realized at around 3.5 to 4 months. Reliability of BVs for 305-DMY from M-1 were on an average 42.30% higher compared to M-2. Rank correlation of BVs was 76.81%. Hence, it is proposed to use Random Regression model for genetic evaluations in Mehsana buffalo.
Keywords:
Buffalo
Breeding value
Model
