Evaluation of Two Equations for Prediction of Digestible Energy in Mixed Feeds and Diets for Horses.
Authors: Martínez Marín Andrés Luis, Valle Emanuela, Bergero Domenico, Requena Francisco, Forte Claudio, Schiavone Achille
Journal: Animals : an open access journal from MDPI
Summary
# Editorial Summary Predicting digestible energy (DE) from feed composition is valuable for nutritionists planning rations without costly in vivo digestibility trials, yet existing prediction equations have known limitations. Martínez Marín and colleagues evaluated two commonly-used DE equations against a dataset of 32 mixed feeds and complete diets with measured in vivo digestibility values, assessing how well each equation predicted DE from proximate analysis alone. Both equations showed reasonable correlation with measured values (R² = 0.89 and 0.87), though the first performed marginally better with lower prediction error (183 versus 217 kcal/kg DM); however, both exhibited systematic bias, consistently overestimating DE in lower-quality forages and underestimating it in more digestible feeds—a bias directly linked to crude fibre content. For practitioners, these findings suggest that whilst these equations remain useful screening tools for feed evaluation, they warrant particular caution when dealing with extreme feed types (very coarse hay or highly digestible concentrates), and developers should refine how crude fibre's interaction with other nutrients affects overall digestibility rather than treating it as an independent variable.
Read the full abstract on PubMed
Practical Takeaways
- •Current predictive equations for digestible energy in horse feeds have acceptable accuracy but systematically over- or under-estimate energy values depending on feed type; practitioners should account for crude fiber content when interpreting results
- •These equations offer a practical alternative to expensive in vivo digestibility trials but should be validated against measured values, particularly for high-fiber or high-energy feeds used in your operation
- •Consider that feeds high in crude fiber may be undervalued by these equations, potentially affecting ration formulation and energy calculations for performance or maintenance diets
Key Findings
- •Both predictive equations performed similarly well with R² values of 0.89 and 0.87 for estimating digestible energy in horse feeds
- •Linear bias in both equations resulted in overestimation of digestible energy in low-energy feeds and underestimation in high-energy feeds
- •Crude fiber content was significantly related to prediction bias and accuracy, suggesting it is a key variable affecting digestibility calculations
- •Prediction errors ranged from 183-217 kcal/kg dry matter, with both equations showing concordance correlation coefficients of 0.86-0.91