Effect of genetic groups on estimates of additive genetic variance.

This study examined the effect of genetic grouping on REML estimates of additive genetic variance with an animal model with selected base populations. A simulated population of 40 animals (20 males and 20 females) was followed under selection or random mating conditions for 10 generations. Each population was replicated 20 or 50 times. Genetic grouping reduced estimates of additive genetic variance in populations with selected base animals, whereas grouping had little effect on the estimate in unselected populations. The effect of genetic grouping varied according to the quantity and kind of information that was missing (percentage of deleted data and pattern of deletion). When genetic grouping was completely random, the estimates were unaffected. Because including genetic groups in the model for analysis affected the estimates of additive genetic variance, the question should be considered of what value or values for heritability should be used for genetic evaluation when grouping is used to account for prior selection.