Influence of hormonal protocols on genomic prediction of sexual precocity in Nelore heifers

Abstract

Reproductive efficiency is a key determinant of profitability in beef cattle systems, and hormonal proto- cols are widely used to improve reproductive performance in tropical breeds. The objective of this study was to assess the effect of hormonal stimulation on the estimation of variance components and genomic prediction ability of the probability of calving before 30 months of age (PC30) of Nelore heifers. The data- base contained information from 31 554 heifers of the National Association of Breeders and Researchers, Brazil. Among them, 22 105 heifers were not genotyped, and 9 449 were genotyped. These animals were categorised based on mating type (MT), with and without hormonal stimulation. For the estimation of genetic parameters and variance components in the complete database, two models were used: one with the inclusion of MT and one without. A single-step genomic BLUP animal threshold model was used. The Linear Regression (LR) validation method was employed to assess model performance. The dataset was divided into training subsets with 31 554 animals with phenotype records and validation subsets with 1 184 genotyped animals with phenotype records, but without progeny records. The subsets were further stratified by mating type. The heritability (h2) estimated for PC30 was of low magnitude and similar for both models (0.175 without MT and 0.18 with MT). Only the group of stimulated heifers with the inclu- sion of MT in the model exhibited a higher partial genomic breeding value (GEBV) than the complete GEBV. For LR estimators, Spearman correlation was not influenced by the inclusion of MT in the model. In animals with hormonal stimulation, a correlation of 0.95 was observed, whereas without hormonal stimulation, it was 0.96. Accuracy was similar across mating types (MT), and the inclusion of MT in the model did not alter the estimation (animals without hormonal stimulation: 0.36 without MT and 0.37 with MT; animals with hormonal stimulation: 0.38 with and without MT). The bias (± SD) for stim- ulated animals (−0.09 ± 0.05 without MT and 0.04 ± 0.05 with MT) compared to non-stimulated animals (−0.01 ± 0.05 without MT and −0.02 ± 0.06 with MT). In conclusion, the estimation of variance compo- nents and h2 for early pregnancy before 30 months is not influenced using hormonal stimulation and/ or the inclusion of mating type in the evaluation model. However, there is a bias effect when considering hormonal stimulation in genomic prediction for PC30.