Identifying Early Indicators of Tail Biting in Pigs by Variable Selection Using Partial Least Squares Regression.

Area of Science:

• Animal Science
• Veterinary Medicine
• Animal Welfare

Background:

• Tail lesions in pigs are a significant welfare concern, impacting health and productivity.
• Identifying predictive factors for tail lesions is essential for proactive management in pig farming.

Purpose of the Study:

• To identify key variables that predict the prevalence of tail lesions in pigs during rearing (REA) and fattening (FAT) stages.
• To develop a data-driven approach for early detection and prevention of tail biting behaviors.

Main Methods:

• Utilized partial least squares regression models to analyze tail lesion prevalence data.
• Employed variable importance in projection (VIP) and regression coefficients (β) for variable selection.
• Collected data on tail lesions, environmental factors, and pig health over multiple batches and scoring days.

Main Results:

• Identified seven significant predictive variables in REA and six in FAT stages.
• Tail posture emerged as the most critical predictor in both rearing and fattening phases.
• Other key factors included skin lesions, suckling phase treatment index, water consumption, activity time, and exhaust air rate.

Conclusions:

• The identified variables provide a robust framework for monitoring and predicting tail lesion prevalence in pigs.
• Implementing data collection on these factors can enable timely interventions to mitigate tail biting and improve animal welfare.
• This predictive model supports evidence-based management practices in swine production.