Description and evaluation of the Farmax Dairy Pro decision support model

Abstract Decision support models have been developed to assist management in dairy systems. This paper describes Farmax Dairy Pro (a pastoral grazing model of a dairy farm) and presents an evaluation of it using two independent farmlet studies carried out in Hamilton and Palmerston North, New Zealand with spring-calving dairy cows. Farmax Dairy Pro predicted, to a high degree of accuracy, mean annual yields (per cow and per hectare) for milk, fat, protein and milksolids (MS; fat + protein) and mean annual concentrations of MS. Monthly predictions were predicted with less accuracy than whole lactation values, but still with moderate degrees of accuracy compared with other comparable models. The general trajectory over time of yield and MS concentration was predicted well for all datasets, but in some instances the model over or under predicted the degree of variation between months. The trajectory of body condition score over time was reliably simulated in early lactation but with some discrepancies in late lactation. The model was then used to determine if it was possible to achieve 1750 kg MS/cow per ha using forages grown within the milking area for the Hamilton study. Managerial changes represented in the model, which included earlier calving dates, use of a chicory crop and additional intakes of pasture in summer, predicted increases in performance of 50–190 kg MS/ha, still at least 81 kg MS/ha short of the target level of production. Farmax Dairy Pro can be used to predict animal, farm and financial performance for different management scenarios.