Early lifetime performance parameters affecting selection and reproductive success in gilts

The sow replacement rate in Australia is 56.1%, with the average parity at which a sow is culled, currently sitting at 4.1 (Australian Pork Limited 2013). There are several key reasons for premature sow turnover, with failure to express pubertal oestrus and poor reproductive performance during the early parities considered a major cause for removal. Most on farm selection criteria for replacement gilts focus on gilt attributes at selection into the breeding herd. However, including criteria from early lifetime performance parameters such as birthweight, weaning weight, and pre-weaning growth may aid in selecting gilts with a higher probability of reproductive success (Knauer 2016). The hypothesis of this study was gilts that are born heavier and which do not have any growth deficits during the pre-weaning or post-weaning period will have a higher probability of (1) being selected into the breeding herd, and (2) displaying pubertal oestrus, resulting in at least one successful mating.

From January 2014 until March 2015, individual weights at birth and d 21 were recorded on 10 480 multiplier gilts (Large White × Landrace, PrimeGro™ Genetics, Corowa, NSW, Australia) born at the genetic supply unit of a large commercial pig producer located in southern New South Wales, Australia. As a matter of routine recording, the date of birth, parity and gestation length of the dam, as well the number of total piglets born in the litter, were known for each individual gilt. Post weaning weights were recorded 2 weeks after weaning on a subset of 3288 gilts. Of the 10 480 gilts included in the project, 8852 (84.5%) gilts were selected to enter the breeding herd and were sent to five different sites on the one farm. A subset of selected gilts had data for weight (n = 7446) and P2 backfat (n = 3399) at selection. Of the gilts selected, 7612 (72.6%) were mated and 6870 (65.5%) farrowed at least once. The significance of early weights and development for the probabilities (0/1) of a gilt being selected (SEL), mated (MATE) and successfully farrowing at least one litter (FARR), was investigated using stepwise logistic regression (PROC LOGISTIC, SAS v9.4, SAS Institute Inc., Cary, NC, USA). Apart from year-quarter of birth (season) and dam parity group (four levels: Parities 1, 2, 3–5, 5+), early in life explanatory variables submitted to the procedure included deciles (allocated within birth year-quarter) and were all treated as class effects. Within each explanatory variable, gilts with missing records were allocated to a separate class. Site (five levels) was also submitted for MATE and FARR. Only factors significant at P < 0.05 were included in the final models for each trait. The significance of the difference between each factor level and the reference decile was assessed via the odds-ratio. The sixth decile was the reference level for each explanatory variable. The corresponding probability of gilts being selected, mated and farrowing for deciles significantly different to Decile 6 were back-calculated using the corresponding odds and the odds-ratios.

Season of birth was the most significant factor contributing to SEL (P < 0.0001). After season, gilts from the lowest decile for birthweight were 5% less likely to be selected than gilts in Decile 6 (P < 0.001). Gilts in the lowest 10% for 21 days weight and post-weaning gain had a 14% lower probability for SEL compared to Decile 6 (P < 0.0001). The probability of SEL was also reduced by 6% for gilts in the lowest 30% for weaning age (P < 0.05). Season and site were the most significant factors affecting both MATE and FARR (P < 0.0001). Relative to the reference level, gilts in the lowest 10% for pre-weaning gain, the lowest 20% for weight at selection, and the lowest 30% for P2 backfat at selection also had a reduced MATE, with corresponding reductions of 9%, 7% and 10%, respectively (P < 0.01). Gilts in the four lowest deciles for pre-weaning gain had a 10% reduction in the probability of FARR, whereas gilts in the highest 20% for post-weaning gain, or the top 30% for P2 backfat at selection, had an increased probability of FARR (+10% and +8%) relative to the reference level. The association between P2 backfat and FARR was generally linear. Being in the low deciles for birthweight, 21 days weight and post-wean gain were detrimental for surviving to selection or meeting the minimum weight requirement at selection age. Once selected, being in the lower deciles for pre-weaning gain, selection weight and P2 backfat at selection decreased the probability of being mated and successfully farrowing a litter whereas, being in the higher deciles for post-wean gain and P2 backfat at selection significantly increased a gilts’ probability of success.