Dynamic probabilistic simulation of dairy herd management practices 2. Comparison of strategies in order to change a herd's calving pattern.

A dynamic probabilistic simulation model is further extended and used for a comparison of different strategies in order to change the calving pattern of a herd. The Markov chain approach is used to simulate herd dynamics. Strategies to change the calving pattern focusing on the farm's intake of replacement heifers, allowing a certain variation in age at first calving, are compared. A method has been developed which allows the tuning of the available replacement heifers to the desired heifer calving pattern, using linear programming. In the basic analysis a spring calving herd is changed into an autumn calving herd. The difference in gross margin per cow per year between the starting and the desired situation is Dfl. 115. The strategy that allows the largest variation in age at calving is fastest in changing the calving pattern. It takes 9 years to realise the desired herd calving pattern, while the desired heifer calving pattern is reached after 2 years. This strategy is also the most profitable one. When considering a period of 10 years, this strategy on average yields Dfl. 105 per cow per year. For a strategy that does not allow changes in the initial age at calving, the increase is only Dfl. 6 per cow per year after 10 years, while in the previous years the costs of changing exceed even the benefits. An additional measure which does not allow cows to be inseminated in certain months during the first few years, shows not to be economically attractive