Identification of climate-resilient Merino sheep using satellite images

Abstract

This study aimed to evaluate the potential use of data from Landsat 5 TM, 7 ETM+, and 8 OLI and meteorology SILO databases to characterise variation in environmental conditions across farms and identify resilient sheep with a low response in performance to changes in the temperature-humidity index (THI) and normalized difference vegetation index (NDVI). A total of 44,848 Merino sheep from 27 farms across Australia were used in this study. The dataset included sheep with complete pedigree and measurements for weaning weight (WWT) and post-weaning weight (PWT). The average NDVI and THI values during the 9 months prior to the phenotypic measurement were used in a linear reaction norm (RN) model with heterogeneous residual variances. The results revealed genotype by environment (GxE) interaction for WWT and PWT between extreme environments with reranking of sires' estimated breeding values along the NDVI gradient. Higher heritability and genetic variances were estimated in favourable environments. Accounting for GxE interactions could lead to a more accurate selection of resilient sheep to changes in climatic and vegetation variables in Australia, and existing environmental data is enabling for this purpose.