2025
Factors Associated with Sow Efficiency
Linking lactation feeding patterns, litter performance, and sow characteristics to sow efficiency metrics.
Researchers: Elly Kirwa Rafael da Rosa Ulgium, Ana Paula Mellagi, and Gustavo Silva of VDPAM, Iowa State University; and Beau Peterson, Caleb Grohmann, and Matt Frizzo of Carthage Innovative Swine Solutions, LLC
In US breeding herds, data collection is widespread; yet too often, this information remains fragmented across software systems.
While producers rely on reports for sow records, task management, and performance summaries, many overlook a critical opportunity: harnessing advanced analytics to maximize reproductive efficiency, profitability, and sow longevity.
The solution is strategic data analysis. By pinpointing key performance drivers, producers and veterinarians can refine management practices to shorten wean-to-estrus intervals, boost breeding success, and improve farrowing rates - directly translating to herd productivity and lifetime performance. The objective of this study was to assess individual sow factors associated with sow efficiency, defined as a wean-to-estrus interval (days), subsequent farrowing success, and subsequent total born.
Methods
Experimental data was sourced from six lactation trials conducted between 2021 and 2022 on a commercial sow farm.
To ensure consistency and limit variability in the study, animals were housed in the same lactation rooms, had the same genetics, and had no Porcine Reproductive and Respiratory Syndrome virus (PRRSV) or Porcine Epidemic Diarrhea virus (PEDV) health challenges.
The final dataset had 4,300 observations, including productivity performance, daily lactation feed intake, sow and litter weights, caliper measurements, and subsequent performance.
Generalized linear regression models were built using the R program to identify factors with each outcome. A total of 23 predictors were evaluated in each model, and trial ID was included as a random effect to account for the potential variations across different trials.
The model-building process involved a manual stepwise forward selection approach, where interactions and confounders were tested based on biological relevance. Predictors with p-values below 0.05 were retained in the final model to ensure statistical significance. Pairwise comparison was tested using t-tests with Tukey-Kramer adjustment, considering p-values <0.05 as statistically significant.
Results
Risk factors associated with wean-to-estrus interval (WEI) included parity (P <0.001), piglets after cross-fostering (P = 0.01), ADFI in the first three days of lactation (P = 0.01), and lactation ADFI (P = 0.019), with the farrowing season as a confounder (P = 0.03), as shown in Figure 1.
Figure 1. Risk factors associated with wean-to-estrus interval (WEI). The plot presents the effects of parity, lactation average daily feed intake (ADFI) for the first seven days, piglets after cross-fostering, and lactation ADFI intake pattern on wean-to-estrus interval (in days). The bars represent the mean WEI for each category, with error bars indicating the 95% upper confidence intervals. The letters above the bars indicate significant differences among groups. Categories that share at least one letter are not significantly different from each other, whereas those with distinct letters are statistically different (P <0.05).
Sows with more than 15 piglets after cross-fostering had a 1.3-day increase in WEI (P = 0.01). Lactation ADFI <10 lbs. in the first three days was associated with a one-day increase in WEI (P <0.001).
Figure 2. Risk factors associated with subsequent farrowing rate. The plot presents the effects of stillborn, lactation average daily feed intake (ADFI) the first seven days, piglets after cross-fostering, and body weight change on the subsequent farrowing rate. The bars represent the mean subsequent farrowing rate for each category, with error bars indicating the 95% upper confidence intervals. The letters above the bars indicate significant differences among groups. Categories that share at least one letter are not significantly different from each other, whereas those with distinct letters are statistically different (P <0.05).
For subsequent farrowing, risk factors included litter size (P = 0.02), stillborn rate (P = 0.01), ADFI during the first week of lactation (P = 0.01), and body weight change (P = 0.01), as shown in Figure 2.
Sows with at least one stillborn showed a 7% decrease in subsequent farrowing (P = 0.01), while those with more than 15 piglets after cross-fostering had a 12% decrease in farrowing rate (P = 0.02).
Figure 3. Risk factors associated with subsequent performance. The plot presents the effects of parity, previous litter size, piglet birth weight, percentage of stillborn, and sow calliper change on subsequent litter (total born). The bars represent the mean total piglets born in subsequent farrowing for each category, with error bars indicating the 95% upper confidence intervals. The letters above the bars indicate significant differences among groups. Categories that share at least one letter are not significantly different from each other, whereas those with distinct letters are statistically different (P <0.05).
Risk factors associated with subsequent total born included parity (P <0.0001), previous litter size (P = 0.01), piglet birthweight (P = 0.01), caliper change (P = 0.04), percentage of stillborn (P = 0.01), and interaction of sow body weight change and litter wean weight (P = 0.002), as shown in Figure 3.
Sows that previously farrowed more than 14 piglets had, on average, one additional piglet in the subsequent litter (P = 0.01), whereas those with lower birth weights produced two more piglets. More than 5% of the stillborn (of the total born) were associated with a decrease of two pigs in subsequent farrowing (P = 0.01).
Sows that gained at least one unit of caliper during lactation had two more piglets in subsequent litters vs. those that lost one unit of caliper.
Conclusion
This study identified relevant predictors of sow efficiency and performance.
Overall, these predictors provide actionable insights for optimizing sow management, enabling targeted nutritional interventions and reproductive strategies to improve longevity, farrowing rates, and overall herd productivity.
Actionable Insights
To optimize efficiency, prioritize early lactation feed intake, ensuring sows hit at least 10 lbs./day in the first three to seven days.
Monitor cross-fostering practices to avoid overloading sows, as larger litters compromise both WEI and farrowing success.
Furthermore, the tracking of caliper changes during lactation to identify sows needing targeted nutritional support and reducing stillbirth risks is essential, going forward.
Dr. Elly Kirwa
Dr. Elly Kirwa is a PhD student at Iowa State University, joining the ISU Field Epidemiology team in 2023. His current research focuses on leveraging data-driven approaches to enhance swine production, specifically through the development of prediction algorithms for sow efficiency and gilt retention.