Mathematical prediction of ileal energy and protein digestibility in broilers using multivariate data analysis

Area of Science:

• Animal Nutrition
• Feed Science
• Digestibility Studies

Background:

• Accurate estimation of apparent ileal digestibility (AID) of energy and protein is vital for broiler nutrition.
• Existing methods rely on table values, which may not fully capture diet-specific variations.

Purpose of the Study:

• To develop and compare prediction equations for AID of energy and protein in broiler diets.
• To evaluate the accuracy of predictions based on chemical composition versus standard table values.

Main Methods:

• A mixture design created 56 diets with varying ingredients and chemical profiles.
• Apparent ileal digestibility was determined in broiler growers (21-24 days post-hatch).
• Multivariate data analysis was used to establish prediction equations based on chemical composition.

Main Results:

• Energy digestibility predictions from table values were accurate (RMSE% = 5.15), comparable to chemical composition predictions.
• Protein digestibility predictions from table values were less accurate (RMSE% = 8.21).
• Multivariate regression using chemical composition improved protein digestibility predictions (RMSE% = 5.46), with starch, crude fiber, and fat as key predictors.

Conclusions:

• Table values provide accurate estimates for broiler diet energy digestibility.
• Chemical composition analysis significantly enhances the prediction accuracy for broiler diet protein digestibility.