Modeling the relationship between reproductive performance and net-revenue in dairy herds

A dynamic stochastic simulation model to study the relationship between reproductive performance and net-revenue in dairy herds is described. Herds included cows, heifers and calves described by individual characteristics and state variables. The latter were updated using time steps of one week in conjunction with calculations of milk revenue, sales revenue, feed costs, breeding costs, additonal costs and net-revenue. Discrete events at the animal level, e.g. involuntary culling, estrous detection and conception were triggered stochastically. Voluntary culling as modeled using the future profitability of cows and potential replacements as determined by dynamic programming. Management and production system were defined by decision variables, e.g. estrous detection rate (EDRT), impregnation factor (IF) and rebreeding policy (RBP). Increases in net-revenue (NR) due to a 1% improvement in EDRT ranged from Cdn. $−2.9 at an EDRT of 75% to Cdn. $22.5 at an EDRT of 35% and were influenced by RBP. Hence, highest increase in NR due to improvement of EDRT can be expected at low reproductive efficiency.