Incorporating genomic information in evaluations for farmed deer in New Zealand

The New Zealand deer industry has made notable genetic progress in the last decades. Initially based around velvet antler and venison, deer are now selected based on several traits including carcass composition, reproduction, and disease resistance within DEERSelect, the industry performance recording system in New Zealand. Due to its low cost and manageable logistics for deer, the genotyping-by-sequencing (GBS) technology was chosen to replace microsatellites for parentage assignment and has allowed a national genetic evaluation across flocks since 2015. Genomic information, however, is not yet fully exploited as evaluations currently only use a traditional pedigree-based, best linear unbiased prediction (BLUP) approach, to estimate the genetic merit of an animal. To assess the benefits of using genomic information in the evaluations, here we compare BLUP, genomic BLUP (GBLUP) and single-step genomic BLUP (SSGBLUP) evaluations for several production traits in the NZ deer industry. Using forward-validation, we estimate the prediction accuracy and bias for these three approaches in 19,863 red animals born between 2018 and 2020. We show that regardless of the approach, GBLUP or SSGBLUP, incorporating genomic information explicitly improves prediction accuracy and reduces bias. Across all traits, we estimate gains in accuracy of 16% for GBLUP and 18% for SSGBLUP on average for red deer. We therefore recommend the incorporation of genomic information in the evaluations performed by DEERSelect and propose a computational pipeline to support the medium to long-term growth of this dynamic livestock industry in New Zealand.