A stochastic model simulating the dairy herd on a PC

Abstract

A dynamic model simulating production and state changes in a dairy herd including young stock on a PC is presented. The model is a time stepping model with steps of one calendar week enabling seasonal management patterns. All discrete events at animal level, e.g., heat detection, conception, foetal death, sex and viability of the calf, involuntary culling, and death, are triggered stochastically. Integrated PC programs enable the user to formulate and test biological parameters and management strategies at cow level, and to compare herd scenarios for up to ten years and with ten replicated runs. An experiment with four different feeding regimes, three different culling and reproduction strategies, and two types of production constraints was simulated. It was found that the effect of culling rate on milk production and live weight gain per cow depends on the feeding regime. It was further clarified that the type of production constraint, herd size or milk quota, had a major effect on the economic ranking of feeding regimes and management strategies.