Ecobreed - The Economic Value of a Cow as Selection Criterion
At the level of a single farm, the decision to keep or cull a cow is key to the economic success. Every culling decision is complex because the costs of rearing a heifer, the profit from selling the cow and the different performance of a producing cow compared to a replacement heifer must be taken into account. If the replacement of a cow by a heifer results in a net loss, the culling decision incurs costs. An increased longevity enables to distribute the fixed rearing costs of a replacement heifer over a longer productive lifetime. In addition, the milk yield is generally higher in older than first-lactating cows. However, if the replacement rate is too low, non-profitable cows might stay in the herd for too long while more profitable and genetically superior heifers can not enter the herd.
Different approaches have been proposed to determine the optimal point in time to replace an old cow by a young heifer. One of these approaches is based on the replacement theory. Here, the expected profit of a cow is calculated using its net present value and compared to the net present value of a replacement heifer. However, the net present value is has limitations regarding the modeling of potential genetic improvements as well as in depicting the random effects in a herd structure.
Markov chain-based models are able to accommodate a higher level of complexity including different developments of traits such as age, milk yield, fertility and potential changes in interactions between the traits. In the first step of a Markov chain-based analysis, all possible states of a cow must be defined. The transition probabilities between the different states of a cow are defined in a second step. In the last step, an optimum solution is derived from the stochastic simulation of a herd where cows are expected to traverse as many states of the model as possible.
The goal of this project is to develop a Markov chain-based model for the Swiss dairy farming industry with the additional aim to design and implement a decision support tool for dairy farmers.