Empirical Models for Predicting the Dry Matter Yield of Grass Silage Swards Using Plant Tissue Analyses

Abstract

Quantifying spatial variability in forage grass yield within individual fields is hampered by the lack of accurate yield monitoring equipment. Here, it is shown how dry matter (DM) yield of silage swards can be predicted on the basis of their mineral composition. This empirical method of predicting yield enables diagnoses of sward nutrient status to be made simultaneously from the tissue test information, and provides a unique opportunity for identifying the nutritional and non-nutritional factors responsible for variability in sward productivity at sub-field scales. Maps of sward DM yield at first, second and third cut silage stages in 1999, and at first cut silage stage in 2000, on a large (7.9 ha) grassland field were produced using two different yield models: one model for first cut and a separate model for second and third cuts. The maps indicated that DM production varied considerably across the field, particularly at first cut, but that the pattern of yield variability at this cut was consistent from 1999 to 2000. The results of the plant tissue tests suggested that N deficiency had been responsible for limiting DM production on the lower yielding parts of the field.