Abstract
BACKGROUND: Ear and tail biting are behaviours in pigs that cause both welfare problems and financial losses. Data collection of behaviour is difficult at the large scale needed for breeding. The damage inflicted on victims can, however, serve as a proxy for animal breeding. Here, we analysed tail and ear damage scores on their original scale, binary scale, and summed versions of these damage traits to investigate which trait definition is best for genetic selection. Using data from six purebred lines (33,329 animals in total) we aimed to (1) estimate genetic parameters for ear and tail damage using direct genetic models, (2) estimate the genetic correlation between tail and ear damage, (3) compare different trait definitions and their impact on accuracy, dispersion, and bias of estimated breeding values (EBV), and (4) compare expected responses to selection for each trait definition. RESULTS: The heritability of the damage traits ranged from 0.04 to 0.06. Ear and tail damage were moderately correlated (0.41-0.45), meaning that the genetic propensity of being a victim is a different trait for tail versus ear biting. Estimates of the accuracy of the EBV for the traits with a five-fold cross-validation and the linear regression method based on pedigree relationships ranged from 0.27 to 0.57, the dispersion from 0.91 to 1.18, and the bias was negligible. With a selected proportion of 5%, genetic progress of ~ 0.20-0.78 genetic standard deviations per generation can be reached, depending on the trait. It was trait dependent whether direct or indirect selection yielded the most response. CONCLUSIONS: Ear and tail damage are heritable traits and are moderately positively correlated. The EBV for the evaluated traits related to ear and tail damage showed moderate accuracies, minor dispersion, and no bias. We hypothesize that from a welfare perspective, ear and tail damage on the original scale are the relevant breeding goal traits. For ear damage on the original scale, the highest response to selection can be expected when selecting on the trait itself, whereas for tail damage on the original scale, selection on summed damage showed the highest gain. Results from this study show that genetic improvement of the direct genetic effect of ear and tail damage is possible.