The information obtained was farm identification, sow identification, date of farrowing, number of piglets born alive (at first and second parity), and date of weaning. The weaning to conception interval was calculated as the length of the farrowing interval minus the average (115 days) gestation length currently [10]. The response variable was the second-litter syndrome, and it was defined as the sow with the same or lower numbers of pigs born alive in parity 2 as compared to parity 1 [2]. Sows were categorized into two groups: 1 if sows had similar or lower number of pigs at the second parity than that at the first parity and 0 for sows that increased litter size at the second parity. The data from 8,592 farrowing records for 4,296 sows were analyzed using binary logistic regression procedures.

The risk factors evaluated were farm (1, 2, and 3), year of farrowing (2003�C2011), season of farrowing (Dry, rainy and windy), number of pigs born alive (��8, 9-10, 11-12, and ��13 piglets), lactation length (��17, 18�C24, and ��25 days), and weaning to conception intervals (��3, 4�C11, and ��12 days). Seasons were categorized based on temperature and rainfall in the region, during the year. All statistical analyses were carried out with the SPSS program [11]. To declare significant effects P < 0.05 values were used.3. ResultsThe overall frequency of sows with the second-litter syndrome was 55.8%, and the frequencies for farm 1, 2, and 3 were 60.4%, 52.2%, and 52.3%, respectively. The litter size means for farms 1, 2, and 3 and for parities 1 and 2 were 12.5 and 9.58, 10.9 and 8.

09, and 10.4 and 8.08 pigs, respectively.There were significant effects of farm, year of farrowing, season of farrowing, number of pigs born alive, and weaning to conception interval on the second-litter syndrome (Table 1). However, there was no effect of lactation length (or weaning age) of the sow (P > 0.05). The odds of the occurrence of the second-litter syndrome were 1.56 and 2.01 times higher for farms 2 and 3 as compared with farm 1. There was not any particular trend in the second-litter syndrome with years. The odds of the second-litter syndrome were 1.20 and 1.24 times higher for the sows farrowing during the dry and rainy seasons versus those farrowing in the windy season. Sows with large litters (>12 pigs) had higher odds (33.2) showing the second-litter syndrome than sows with small litters (<9 pigs).

Sows with shorter weaning to conception intervals (<4 and 4�C11 days) had higher odds showing a decrease in litter Batimastat size at the second parity, as compared with sows with longer weaning to conception intervals.Table 1Factors associated with the second-litter syndrome in three pig farms in Yucatan, Mexico, using a binomial logistic model.4. DiscussionThe overall frequency of sows showing the second-litter syndrome (55.8%) in the three farms, here studied, is higher than the 49.5% reported by Saito et al.