Animal welfare

Quantifying social genetic variance for feed efficiency in pigs

---

Abstract

Recent advances in pig production systems allow a detailed characterization of feeding behavior traits and social interactions, which are key to improving feed efficiency and animal well-being. Accounting for social genetic effects in breeding strategies can improve the accuracy of genetic evaluations and enhance productivity. This study aimed to assess the impact of including social interactions in the genetic evaluation of feed efficiency in Iberian pigs. Data from 1,195 pigs housed in groups of 12 ± 3 animals were analyzed. Genetic parameters were estimated for feed conversion ratio (FCR). Three models were compared: a traditional additive genetic model (GM), a classical social model with a binary incidence matrix (SBM), and a social model with continuous interaction intensities (SCIM). The SBM assumed equal interaction among contemporaneous animals using a binary social incidence matrix. The SCIM incorporated continuous interaction intensities derived from automatic feeder data, quantified as the frequency of displacements between pairs of animals. Displacements were defined as feeder visits in which the interval between consecutive animals was less than 60 seconds together with a reduction in feed intake of the displaced animal. Heritability estimates for FCR were moderate (0.15 ± 0.04). The proportion of phenotypic variance explained by the social variance was higher in the SBM (0.13±0.04) compared with SCIM (0.06±0.02). The SBM showed an important reduction of pen effect variance and an increase in additive variance. Additionally, the correlation between additive and social effects was strongly positive in SBM (0.71 ± 0.19) and moderate in SCIM (0.35 ± 0.18). The higher correlation observed in SBM likely reflects confounding arising from the homogeneous structure of the social incidence matrix. Correlations between EBVs from GM and TBVs from the social models highlighted differences in selection criteria. The percentage of coincidence between animals selected by EBV from GM and TBV from social models were low (< 50%), indicating that accounting for social interaction intensities alters selection priorities. However, special attention should be paid to the accuracies of the SBV estimation before selecting by TBV, as SBV is amplified by (n-1) and may have an important influence on TBV. These preliminary results indicate that SBM may not adequately resolve collinearity between direct genetic, social genetic, and pen effects. In this regard, SCIM and SOIM provides a biologically more realistic representation of social interactions but requires larger datasets to accurately estimate social variance.

Keywords: 2026

How to Cite:

Nuñez, P., Negro, S., Casellas, J., Varona, L. & Ibáñez-Escriche, N., (2026) “Quantifying social genetic variance for feed efficiency in pigs”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286321. doi: https://doi.org/10.31274/wcgalp.23925

Rights: 1