Genetic associations between mid-infrared-predicted methane production and heat tolerance of production traits in dairy cattle

ABSTRACT The future direction of the dairy industry is breeding cattle that have low environmental impact and are resilient when faced with environmental perturbations. In Canada, genetic selection for reduced environmental impact is partly based on methane production predicted from mid-infrared spectroscopy data (MPMIR). While relationships between MPMIR and production have been previously reported, correlations between heat tolerance of production traits and MPMIR have not been studied. The objectives of this study were to 1) estimate genetic correlations between MPMIR and heat tolerance defined by milk yield (MY), fat yield (FY), and protein yield (PY), and 2) evaluate how these relationships change across the lactation. Data comprised 7,714,324 MY, 7,128,396 FY, 7,177,657 PY, and 1,017,436 MPMIR records collected on first parity Canadian Holsteins. A random sample of 1,000 herds (approximately 15% of all data) was selected and 6 ranges of days in milk (DIM) within the subset were generated: 5 to 305, 5 to 185, 30 to 185, 60 to 185, 90 to 185, and 120 to 185 DIM. Hourly ambient temperature and relative humidity were extracted from the NASA POWER database (https://power.larc.nasa.gov/; accessed 2025–08–05) and average temperature and humidity index (THI) values across each test-day and the 2 preceding days (d-2 to d0) were matched to each production record. Repeatability reaction norm models were fit for production traits using THI thresholds of 65 (MY), 55 (FY), and 60 (PY) and a repeatability model was fit for MPMIR. Pairwise models were run between MPMIR and each production trait across the 6 DIM windows. While magnitude varied, relationships between MPMIR and production traits were generally consistent across DIM windows. Genetic correlations between MPMIR and heat tolerance were negligible for MY (−0.01 ± 0.05 to 0.26 ± 0.69) and PY (−0.03 ± 0.04 to 0.09 ± 0.09), and lowly favorable for FY (−0.05 ± 0.07 to −0.16 ± 0.04). Correlations for MY and PY were not significant across any DIM window evaluated while correlations for FY were significant across 2 of the 6 windows (5 to 305 and 5 to 185). Results indicate little to no change in heat tolerance of MY and PY and marginal improvements in heat tolerance of FY with selection for lower methane production. However, interpretation of these results should be made with caution, as methane production was derived from a mid-infrared predicted phenotype rather than direct measurements.