Regression-Based Precision Feeding Models for Improving Productive and Reproductive Performance in Large White and Landrace Pigs

Aims: To develop precision feed intake-based swine nutrition models for Large White and Landrace pigs associated with physiological stage to enhance feed efficiency, reproductive performance, and overall cost-efficient production. Study Design: Descriptive statistics were used to characterize performance indicators, while correlation and multiple regression analyses were applied to formulate precision feeding models and assess their predictive reliability. Place and Duration of Study: International Training Center on Pig Husbandry, Lipa City, Batangas, Philippines from January 1, 2022 to June 25, 2025. Methodology: Productive and reproductive performance data of Large White (407 individual records) and Landrace (361 individual records) pigs were collected and analyzed according to physiological stage: breeder boars, sows, and grower-finishers. Variables included, actual feed intake, average daily gain (ADG), feed conversion ratio (FCR), actual feed intake, lactation length, litter size, birth weight, and mortality rate. Results: Findings indicated that ITCPH implemented semi-automated feeding system, along with breed-specific feed formulations and the incorporation of alternative feed resources, contributed significantly to improved production. Large White pigs demonstrated higher ADG and extended lactation periods, while Landrace pigs exhibited superior reproductive metrics, including heavier birth weights, increased litter sizes, and reduced mortality rate. Both breeds achieved favorable feed conversion ratios, performing better than the national standard. For breeder boars, Model 3 which includes ADG and FCR have the highest predictive reliability with 95.81% for Large White and 91.83% for Landrace, indicating that integrating growth rate and feed efficiency provides a more robust and biologically sound. Among sows, Model 2, which incorporated reproductive traits achieved 100% predictive reliability for both breeds allowing accurate prediction of sow productivity. In grower-finishers, Model 3 (birth, weaning, nursery, finisher weight, ADG and FCR), was more reliable for Landrace pigs (98.58% reliability) which reflects the breed’s stronger response when growth and feed efficiency indicators were integrated, whereas Model 1 (birth, weaning, nursery, finisher weight), based on simpler growth indicators, was more effective for Large White pigs (90.68% reliability; p = 0.0134). Conclusion: Developed precision feeding models provide a basis for implementing breed- and stage-specific feeding strategies in commercial Large White and Landrace swine production. Adoption of these models can enhance feed efficiency, optimize productive and reproductive performance, and promote cost-efficient pig production.