Evaluation of feeding behaviour traits to predict efficiency traits in pigs using partial least square regression

Area of Science:

• Animal Science
• Nutritional Physiology
• Precision Livestock Farming

Background:

• Improving efficiency traits like protein efficiency (PE), digestible energy efficiency (EnE), and lipid gain (LipG) is crucial for reducing environmental impact and production costs in pig farming.
• These traits are challenging to measure directly, often requiring animal slaughter, and are influenced by genetics, diet, and individual feeding behavior.

Purpose of the Study:

• To assess the potential of using individual feeding behavior traits to predict key efficiency traits (PE, EnE, LipG) in pigs fed a protein-restricted diet.
• To evaluate the accuracy of these predictions using partial least square (PLS) regression.

Main Methods:

• 587 Swiss Large White pigs were fed a protein-reduced diet ad libitum from 22.5 to 106.6 kg body weight.
• Individual feed intake and carcass composition (via dual-energy X-ray absorptiometry) were monitored.
• Feeding behavior (e.g., daily intake, meal frequency, feeding rate) and efficiency traits (PE, EnE, LipG) were recorded and analyzed using PLS regression.

Main Results:

• Partial least square regression models explained 34.64% of variation in PE, 55.42% in EnE, and 82.68% in LipG.
• Concordance correlation coefficients indicated good prediction accuracy for EnE (0.70) and LipG (0.90), but lower accuracy for PE (0.50).

Conclusions:

• Individual feeding behavior traits are valuable predictors of energy efficiency and lipid gain in pigs under dietary protein restriction.
• While feeding behavior shows promise for predicting EnE and LipG, its predictive power for PE is less robust in this context.