Feed intake and feeding behaviour traits of gestating sows are associated with undesirable outcomes

Abstract

<p>This study hypothesised that feed intake and feeding behaviour traits recorded during gestation are genetically and/or phenotypically associated with poor reproductive performance or undesirable outcomes, including unanticipated removals. Data from 2,847 pedigreed sows recorded for 3,939 gestation events (Farm A; commercial, 90.5% F1 sows) and 540 sows over one gestation (Farm B; nucleus, six maternal and terminal lines) were used. The reproductive traits were numbers of piglets born in total (TB), born alive (NBA), stillborn (SB) and weaned (NWEAN). The outcome traits were failure at farrowing (FFAIL), presence of any (SBLIT) or excessive (SBFAIL) stillborn piglets, lactation failure (LFAIL), and sow removals within 35 (REMW) or 60 (REM60) days post farrowing. Feed intake and feeding behaviour traits were derived from data recorded using electronic sow feeding (ESF) systems, averaged across or within specific periods of gestation. Lower feed intake was detrimentally associated with increased incidences of FFAIL, SBLIT, LFAIL, REMW and REM60. Undesirable outcomes were more likely for sows that missed meals or had low (< 1 kg) intake days, or that spent less time eating. Outcome traits were less heritable in Farm A than in Farm B, with the exceptions of SBLIT (0.07 ± 0.02 and 0.10 ± 0.08), which were similar in both farms. Heritability estimates were low to moderate for FFAIL (0.20 ± 0.10) and SBFAIL (0.12 ± 0.08) in Farm B and negligible in Farm A. Heritability estimates for lactation and removal traits were negligible. More time spent eating and more feeding events per day were genetically (r_A < 0.23) correlated with larger litter size (NBA and TB), longer gestation length, but also increased removals (REMW) in Farm A (all r_A < 0.20). In Farm B, sows with low feed intake or missed meals genetically had shorter gestation (r_A: -0.26 to 0.51), lower litter size (r_A: 0.62 ± 0.43) and more stillborn piglets (r_A: 0.40 to 0.54). This study demonstrated that the information obtained from the ESF can be used to identify sows with a higher risk of poor reproductive performance or undesirable outcomes. The phenotypic relationships between feed intake and feeding behaviour traits with outcomes were similar for both production systems (Farms A and B), demonstrating suitability of generating useful phenotypes from ESF across two populations. There was weak evidence that selection for more time spent eating and a higher appetite could result in improved farrowing and lactation outcomes and reduced removals for sows and their piglets.</p>