Using random regression models to optimize selection for yield, persistency and calving interval in Philippine dairy buffaloes

Abstract

Eigenvalue (EV) decomposition of the genetic coefficient matrix from random regression analysis of milk test day yields in Philippine dairy buffaloes provided independent variables for use in selection. The 1st, 2nd and 3rd EVs explain 80.3%, 18.8% and 0.2% of the total genetic variance, respectively. Selection on the 2nd eigenvalue resulted in higher response in the latter half of lactation, thus better persistency. There was a 5% reduction in milk yield (MY), 5.6% higher net economic gain and improvement in persistency from 0.62 to 0.79 with an optimal index relative to selection on total lactation. There was also a high genetic correlation between calving interval (CI) and the 2nd EV for MY. As CI is mostly correlated with MY in late lactation, this suggests CI is mainly associated with lactation length and may not be a true measure of fertility.