A static model for estimating energy content of compound feeds in a dynamic feed evaluation system

摘要

The objective of the study was to develop a staticempirical model for the estimation of net energy contentof compound feeds in a dynamic feeding system usingnet energy for lactation at 20 kg of dry matter intake/d(NEL20) values calculated by the Nordic Feed EvaluationSystem (NorFor) model. In the NorFor system,NEL20 is a standardized value used to describe netenergy content of feeds. The static model would allowprediction of the net energy value of compound feedswithout access to the input data needed for the dynamicmodels. Our hypothesis was that NEL20 values ofcompound feeds can be predicted using organic matterdigestibility (in vitro) and chemical components of thecompound feeds as input variables. For this, 75 compoundfeeds and their 108 associated ingredients werecollected across Scandinavia for model development.The proposed best model for prediction of compoundfeed NEL20 included crude fat, neutral detergent fiber,digestible organic matter measured in vitro, and crudeprotein (urea corrected) as independent variables. Lackof additivity of chemical components between valuesanalyzed directly in the compound feed and values calculatedby the weighted sum of ingredients was detectedas the main source of error in the model, emphasizingthe importance of accurate chemical analysis and samplingpractices. Results from practical use of the modelshow that it may be a valuable tool that could be usedby several actors in the feeding sector using the NorForsystem. Feed manufacturers could use it to monitor thenet energy content in their final product, and farmerscould use it to check the net energy content of thepurchased compound feed. However, validation of thismodel against an independent set of samples is lackingin this study and its prediction performance should befurther evaluated. The model will need recalibrationif the feed parameters used in the dynamic model forthe estimation of reference values change, as this wouldnot be reflected in the predicted values of the createdmodel.