Inclusion of inbreeding coefficients into genetic evaluation models for Jersey and Brown Swiss cattle in Mexico

The inclusion of the inbreeding coefficient (F) into the models to predict breeding values (BV) for Jersey (J) and Brown Swiss (BS) Mexican registered dairy cattle was studied. The conventional model including the fixed effect of herd-year-season, and the genetic additive and permanent environmental random effects was compared to an alternative model that additionally included F. Traits analyzed in J were the yields per lactation of milk (MY), fat (FY) and protein (PY), and the percentages of fat (%F) and protein (%P) in milk. In BS only MY was analyzed. The analyses were done using an animal model and the MTDFREML program; pedigree files included 21026 and 101861 animals in J and BS. Several criteria to compare results from the alternative models were used. The most notorious change was for the BV for %P in J; the BV for the other traits they were similar in both breeds. Coincidence among the top 100 sires from the two models ranged from 90 to 98 %. Regression coefficients of BY and their accuracies from the conventional model on the BV and their accuracies from the alternative model were around 1.0 in both breeds. The estimates of the genetic variance and h(2) obtained with both models remained essentially the same for all traits in both breeds. The results suggest that the inclusion of F into the models to evaluate genetically Jersey and Brown Swiss purebred Mexican dairy cattle is not necessary. However, it is required to perform this evaluation when the average for F is higher than it is today.