Field‐related quality management system for grass silage production

A model for the prediction of the silage quality was tested in several large farms in northeast Germany. The model uses data and information on the fresh grass and the application of silage additives at ensiling to evaluate the “ensilability.” During clamp filling, the model analyses information on the filling performance, activity of compacting machines and hermitic covering of the clamp and evaluates the “ensiling technique.” Both “ensilability” and “ensiling technique” are assessed as “good,” “medium” or “bad” and pooled in “silage evaluation grades” 1–5. In the present model version, the predicted silage net energy lactation (NEL) and crude protein (CP) content is between 3% (for the best grade 1) and 15% (for the weakest grade 5) lower than the fresh grass NEL and CP content. In our experiments, all the grass silage production systems were characterized by “good” “ensilability” and “ensiling technique.” Under those circumstances, the experimental results indicated a closer numerical relation between fresh grass and measured silage NEL and CP content than between predicted silage and measured silage NEL and CP content. The results reveal that the model prediction calculation should be changed compared to the present version. If “ensilability” and “ensiling technique” are evaluated as “good,” the predicted silage NEL and CP content should be forecasted in a range between 0% and 3% decrease compared to fresh grass.