Abstract

Modelling is being used to compare and contrast farm management strategies. A Whole Farm Model (WFM) has been developed that links models for cow metabolism (‘Molly’) and pasture growth driven by climate, with management policies. It has been evaluated against trial data and the challenge is to now use it for prediction. This requires that cows be initializedso that their production is determined by their environment and genotype. Livestock Improvement Corporation has developed a system of quantifying genotype through production values (or PV's). The aim of this study was to assess the ability of the WFM to predict Strain Trial cow production using PVmilk. The relationship between PVmilk and milk yield and composition was determined by allowing a parameter within ‘Molly’ to adjust so that predicted matched actual production for 3 Strain herds and one Jersey herd. WFM was then initialized using observed PVmilk to set genotype of each cow in 8 other Strain herds and farmlets were simulated for 2003/04 season. Predictions were considered useful if the prediction errors for milk yield and milksolids (629 L and 61 kg MS respectively) were lower than the observed standarddeviations (825 L and 66 kg MS respectively) i.e. meaning model predictions of observed values were more accurate than simply using the mean as an estimate of observed values. The consistent under-prediction of milk yield and milksolids require further adjustments to the initialization of the model. PVmilk has the potential to set genotype in ‘Molly’, within the WFM.

AL, Vaughan, AW Greer, RW McAnulty, and AR Sykes

Proceedings of the New Zealand Society of Animal Production, Volume 66, Napier, 83-87, 2006