A cross-sectional study for predicting tail biting risk in pig farms using classification and regression tree...

Area of Science:

• Animal Science
• Veterinary Medicine
• Agricultural Engineering

Background:

• Tail biting in pigs is a persistent welfare and economic issue requiring effective on-farm interventions.
• Existing methods for tail biting prevention can be challenging to implement effectively across diverse farming conditions.

Purpose of the Study:

• To introduce Classification and Regression Tree (CRT) methodologies for developing a tool to prevent acute tail biting lesions in pigs.
• To identify critical risk factors associated with tail biting on a farm-specific level.

Main Methods:

• Utilized data from 60 commercial heavy pig farms, including on-farm visits and farmer interviews.
• Collected comprehensive data on farm management, herd health, biosecurity, climate, feeding, and production traits.
• Applied Classification and Regression Tree (CRT) analyses to identify predictive variables for tail biting.

Main Results:

• CRT analysis demonstrated 86.7% sensitivity for predicting acute tail biting lesions.
• A correlation of 0.7 was found between observed and predicted tail biting prevalence using Regression Tree.
• Identified five critical predictors: stocking density, ammonia levels, pigs per stockman, floor type, and feed supply timeliness.

Conclusions:

• CRT analysis is valuable for managing tail biting risk factors on a farm-specific basis.
• The developed model offers practical applications for supporting tail biting prevention strategies, particularly for high-risk pig subgroups.
• Empowers farmers and veterinarians to assess and manage predisposing variables to reduce tail biting lesions.