Comparison of alternative models for the joint analysis of skin lesion counts and aggression traits in group-housed pigs

ANGARITA, BELCY K.; CANTET, RODOLFO J C; CARLY OMALLEY; SIEGFORD, JANICE M.; ERNST, CATHERINE W.; STEIBEL, JUAN P.

Mineapolis, Minesota

Simposio; Pig Welfare Symposium; 2019

Mixing pigs into new social groups may induce aggressive interactions that result in severe skin lesions on the bodies of animals if pigs cannot escape or retreat to avoid aggression. The genetic relationship between fighting and skin lesions is often studied using bivariate animal models. More recently, social genetic effect models (SGE) have emerged as an alternative. SGE model the direct effect of the animal?s genotype on its own phenotype and the indirect effect of the animal?s phenotype on the group-mates phenotypes. The goal of this study was to compare bivariate models of aggression and lesion count (LC) traits to SGE models of LC that are parameterized as a function of intensity of damaging social interactions such as reciprocal fights and attacks (ISGE). The dataset included 792 grow-finish pigs from 81 litters, housed in 59 pens (10 to 15 pigs per pen). Trained observers counted skin lesions in front, middle and rear regions and recorded the duration of attacks and reciprocal fights between pairs of animals. Bivariate animal models (BAM) were fit to pairs of behavioral and lesion count traits. Alternatively, ISGE models used dyadic aggression data to parametrize genetic social effects for LC, which represent the effect of each animal genotype on the lesion counts of its group-mates. BAM estimated a positive genetic correlation between duration of reciprocal fight and LC (rg=0.72 to rg=0.89, P0.1). Consistently, the genetic correlation between direct and social effects of LC was the highest with ISGE based on reciprocal fights (rg=0.87 to rg=0.92,P