Herd management procedures and factors associated with low farrowing rate of female pigs in Japanese commercial herds

The objective of the present study was to compare management procedures and production factors between low-farrowing-rate herds (LFR herds) and the remaining herds (Non-LFR herds). The questionnaires were sent to the producers of 115 herds that use the same recording system. The questionnaire requested information about management procedures in 2008: (i) daily frequencies of estrus detection: once or twice a day; and (ii) the timing of first insemination. Data from 93 completed questionnaires (80.9%) were coordinated with the reproductive data of individual female pigs from the recording system. The data included 78,321 service records from 37,777 sows and gilts. Herds were classified into two groups on the basis of the lower 25th percentile of farrowing rate: LFR herds (76.5% or lower) and Non-LFR herds (76.6% or higher). At the herd level, a two-sample t-test, was used to compare the surveyed management procedures between the two herd groups. At the individual level, two-level mixed-effects models were applied, by using a herd at the level two and an individual record at the level one to determine associations between low farrowing rate and management procedures or production factors in gilts and sows. Gilt and sow models were separately constructed. Means (+/- SEM) of farrowing rate in LFR herds and Non-LFR herds were 71.3 +/- 0.92 and 85.5 +/- 0.54%, respectively. The lower farrowing rates of gilts and sows in LFR herds were associated with once-daily estrus detection, late timing of first insemination and single mating (P < 0.05). In LFR herds that detected estrus only once a day, the farrowing rate decreased by 10.5% in first-serviced gilts and by 4.2% in reserviced sows compared with twice daily estrus detection (P < 0.05). However, there was no such association in Non-LFR herds (P > 0.05). The LFR herds had higher percentages of single-mated gilts and sows than Non-LFR herds (P < 0.05). Fewer LFR herds than Non-LFR herds performed first insemination immediately after first estrus detection for gilts or by 6-12 h for sows (P < 0.05). In order to improve the farrowing rate in LFR herds, we recommend detecting estrus twice a day and performing first insemination earlier after first estrus detection; immediately for gilts and by 6-12 h for sows. Additionally, increasing the percentage of multiple inseminations can effectively improve the farrowing rate. (C) 2012 Elsevier B.V. All rights reserved.