Risk Factors for Culling Intervals and Lifetime Performance on Commercial Breeding Farms

The objectives of this retrospective swine cohort study were 1) to characterize culling interval, and 2) to determine risk factors for culling intervals for both the served female breeding pigs and farrowed sows. Data from farm-entry to removal included 142,704 lifetime records of females on 155 Spanish farms, served between 2011 and 2016. Culling intervals for served females were defined as the number of days from first-service to culling without farrowing. Culling intervals for farrowed sows were the number of days from weaning to culling without subsequent service. Females were categorized into two groups: no-return and returned female groups. Also, two farm categories were defined on the basis of the upper 25^th percentile of the farm means of the number of pigs weaned per sow per year: high-performing farms (> 26.2 pigs) and ordinary farms (≤ 26.2 pigs). Two-level linear mixed-effects models were applied to the data. Culling intervals (± SEM) in served females and farrowed sows were 72.2 ± 0.21 and 9.4 ± 0.04 days, respectively. The ratio of served females to farrowed sows was 1.0:2.4. Prolonged culling intervals for both served females and farrowed sows were associated with low parity, reproductive failure and being fed on ordinary farms (P < 0.05). Prolonged culling intervals for served gilts were associated with being fed on ordinary farms and being returned (P < 0.05). In addition, prolonged culling intervals for served sows were associated with prolonged lactation length and weaning-to-first-service interval of 0-6 days (P < 0.05). Furthermore, served females that were culled at parities 0-3 on high-performing farms had 11.4-13.7 days shorter culling intervals than those on ordinary farms (P < 0.05). For farrowed sows, prolonged culling intervals were associated with sows having 16 or more pigs born alive, farrowing 0 stillborn piglets and being fed on ordinary farms (P < 0.05). Additionally, farrowed sows that were culled only parity 1 on high-performing farms had 4.0 days shorter culling intervals than those on ordinary farms (P < 0.05). In conclusion, to reduce culling intervals and improve farm productivity, we recommend keeping an eye on at-risk females and also reconsidering culling policy for farrowed sows, especially at low parity on ordinary farms.