Genetic background of heat tolerance in maternal pig lines based on reaction norm models for reproductive performance traits

Understanding the genetic background of heat tolerance in maternal pig lines is important since reproductive performance is directly affected by increased thermal load. In this context, the main objective of this study was to evaluate how alternative environmental gradients (EGs) and critical periods affect genetic parameter estimates, environmental sensitivity, genomic estimated breeding values, and animal ranking for heat tolerance based on reproductive traits and reaction norm models. Records from 24,970 Landrace (LR) and 30,754 Large White (LW) animals were analyzed, including 17,141 and 27,419 genotyped individuals, respectively. Traits evaluated included number of piglets born alive (NBA), stillborn (SB) and mummified (NM) piglets, litter birth weight (LBW), average piglet birth weight (APBW), number of piglets weaned (NW), and litter weaning weight (LWW). Eight environmental gradients were evaluated (T2M, T2M_MAX, T2M_DEW, THI_MEAN, THI_MAX, THI_P95, DH, and HrsAbove) over 7-, 15-, and 30-day periods before trait measurement. Heritability estimates ranged from 0.02 (NM) to 0.50 (APBW) in LR and 0.02 (NW) to 0.36 (APBW) in LW. Slope additive genetic variance ratios (indicator of environmental sensitivity) ranged from 0.003 (NW) to 0.078 (LBW). Genetic correlations between intercept and slope varied in magnitude and direction across traits and breeds such as negative for LWW (-0.25) in LR and APBW (-0.18) in LW, indicating greater environmental sensitivity among animals with higher genetic merit for these two traits, and positive for LBW and NW in LR and for NW (0.67), LBW (0.46), and LWW (0.16) in LW. Genetic correlations across EG levels were high for most traits (0.88-0.99 in LR; 0.80-0.99 in LW), but lower minimum values for LBW (0.41) and NW (0.46) in LR and LBW (0.38) and LWW (0.11) in LW indicated greater genotype-by-environment interaction (G×E). G×E was also reflected in re-ranking: between low and high EG levels, the overlap among top-ranked animals decreased from 78% to 54% for NW in LR, and from 71% to 39% for LBW and from 74% to 27% for LWW in LW. The choice of EG influenced most analyses, and the strongest G×E effects were captured by DH-based gradients, with additional evidence for THI and T2M_DEW, especially for 7- and 15-day periods. These results indicate that thermal gradients should be incorporated into genetic evaluations to identify animals with greater robustness or adaptation to thermal load, especially for LBW, NW, and LWW.