GxE

Modeling genotype-by-environment interaction effects for reproductive performance in sows across a disease challenge gradient

---

Abstract

Porcine Reproductive and Respiratory Syndrome (PRRS) is a viral disease that has significantly impacted the global pig industry. Previous studies showed that genetic parameters and estimated breeding values for reproductive performance traits vary, depending on whether they are evaluated before, during, or after a PRRS outbreak. Therefore, understanding genotype by environment interactions (GxE) for these traits is critical to maximize selection for enhanced reproductive performance under both PRRS-free and PRRS-challenged conditions. This study aimed to investigate GxE effects for reproductive performance traits across an environmental gradient (EG), including natural exposure to PRRS. Data recorded from 53,377 farrowing events, from 16,674 Large White sows, across three farms in the United States and two in Spain affected by PRRS outbreaks, were used for this study. Pedigree information was available for all animals, and genotypes for 24,329 genomic markers (after quality control) were available for 483 sires and 3,518 sows with phenotypic data. Evaluated traits included: number of piglets born alive (NBA), number of piglets born mummified (MUM), and number of piglets stillborn (NSB). Reaction norm models (RNM) were fitted by regressing phenotypic records on standardized estimates of farm-year-week (FYW) effects for NBA, which ranged from -5.23 to 1.72, with values lower than -1.64 indicating PRRS outbreak occurrence. Different RNM were fitted, modeling residual variances assuming either homogeneity or heterogeneity across the EG with up to five classes. The results showed that heterogeneous residual variance structures improved the fit for NBA and MUM, while a homogeneous residual variance model was best for NSB. The heritability estimates for the reaction norm intercept ranged from 0.04 to 0.07 for NBA, from 0.01 to 0.04 for MUM, and 0.01 for NSB. The heritability estimates for the reaction norm slope were 0.01 for NBA, from 0.004 to 0.03 for MUM, and 0.08 for NSB. The genetic correlation estimates between intercept and slope for NBA, MUM, and NSB were 0.14 ± 0.13, -0.89 ± 0.02, and -0.23 ± 0.13, respectively. The average heritability across EG for NBA, MUM, and NSB were 0.08, 0.05, and 0.09, respectively. Across traits, the genetic correlation estimates declined as the distance between EG levels increased, with low correlations when contrasting PRRS-affected weeks (FYW < -1.64) with positive FYW values, reaching their lowest values at -0.05, 0.27, and -0.81 for NBA, MUM, and NSB, respectively. There is clear evidence of GxE effects for reproductive performance traits across various levels of PRRS exposure, leading to animal reranking under PRRS-free vs. challenged conditions. These results suggest that, to maximize genetic improvement across environments, collecting data under both PRRS-free and PRRS-challenged conditions should be considered, and the RNM slope may serve as an indicator trait to select for enhanced resilience under PRRS challenge.

Keywords: 2026

How to Cite:

de Oliveira, L., Dunkelberger, J., Sevillano, C., Knol, E., Arirangan, S., Wedow, R., Tegtmeyer, M., Tuinstra, M. & Brito, L., (2026) “Modeling genotype-by-environment interaction effects for reproductive performance in sows across a disease challenge gradient”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286452. doi: https://doi.org/10.31274/wcgalp.24013

Rights: 1