Animal welfare

Genetic Evaluation of Sow Reproductive Traits with Selection Based on Free Farrowing Data: Implications for Current and Future Breeding Program

---

Abstract

While it is common to use farrowing crates for sows in many countries, there is increased pressure to move to a free farrowing system for animal welfare reasons. Crated systems aim to reduce piglet crushing, but impose restrictions on the sow, including full expression of maternal behaviour. The aim of this analysis was to compare reproductive data from sows kept in free farrowing pens (FF, free from farrowing to weaning) or farrowing crates (FC) to investigate the variability in genetic merit across these environments. Data was obtained from a Large White (LW) and a Landrace (LR) based line, located in Canada, Norway, Germany and Brazil. Data was available across multiple parities for 8,342 LW sows, and 17,309 LR sows, where 44.1% and 89.5% of the animals were in FF for LW and LR, respectively. Genetic parameters were estimated with ASReml-R for the traits total number born (TNB), (natural log of) number of stillborn piglets (STB), and (natural log of) number of dead piglets in the lactation period (ND), within farrowing system (FF or FC), and correlated between systems. Trait collection was the same across management systems. Heritability estimates for all traits were similar across farrowing systems. For LW, heritability estimates were 0.10 (±0.01) for TNB, 0.10-0.14 (±0.02) for STB (with the higher values in FF), and 0.07 (±0.02) for ND. For LR, heritability estimates were 0.14 (±0.01-0.02) for TNB, 0.07 (±0.02) for STB, and 0.07-0.09 (±0.01-0.02) for ND (with the higher values in FC). Within-trait correlations (i.e. across farrowing systems) for LW were 0.87 (±0.13) for TNB, 0.81 (±0.12) for STB, and 0.63 (±0.20) for ND. Similar, but slightly higher correlations were found for LR, with correlations of 0.99 (±0.20) for TNB, 0.83 (±0.27) for STB, and 0.69 (±0.30) for ND. The selection response was always highest when the trait under selection was the same as in the system used. Thus selection should be based on data from the same system as the breeding objective. While the heritability estimates and genetic variances were similar across systems, the lower genetic correlation for ND indicates that this is not the same trait in FF and FC systems. This has implications for the breeding program design if the breeding goal is aiming at a future with free farrowing systems. In that case it will be beneficial for maximising genetic progress to collect data in production systems with a free farrowing system.

Keywords: 2026

How to Cite:

Iversen, M., Langmoen-Kosaka, E., Olsen, D., Reimer, J. & Vargovic, L., (2026) “Genetic Evaluation of Sow Reproductive Traits with Selection Based on Free Farrowing Data: Implications for Current and Future Breeding Program”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2285410. doi: https://doi.org/10.31274/wcgalp.23690

Rights: 1