Comparison of economically optimized culling recommendations and actual culling decisions of Finnish Ayrshire cows

Abstract

Our purpose was to compare culling recommendations obtained from an economic-optimization model with actual culling of Finnish Ayrshire cows. The dynamic-programming (DP) model we used optimizes replacement and breeding decisions to maximize the net revenues from cows currently in a herd and their potential replacements over a 5-year decision horizon. Cows were described in the model by five state variables: parity, stage of lactation, month of calving, milk-production level, and days open (pregnancy status). We performed survival analysis to study the effects of those five factors on culling and to compare the actual culling of cows in December 1993 and June 1994 with the optimized replacement recommendations for the same months and for cows in the same herds. The risk of culling increased as a cow grew older, both in the actual herds and in the DP recommendations for December. In the optimized replacements for June, however, the age of a cow did not play a significant role. A cow that had been in milk >270 days had a lower risk of culling than cows in earlier stages of lactation. When 305-day milk production increased by 100 kg, the risk of culling decreased by 4% in the actual herds and by 6 and 12% in the DP recommendations for June and December, respectively. When the days open lengthened by a month, the risk of culling was 2.0- and 1.6-times higher in the actual herds and 1.7- and 2.0-times higher in the DP recommendations for June and December, respectively. Month of calving had a different effect in the optimized recommendations compared with the real-life situation: cows calving from January to August had a lower risk of culling than cows calving in the fall in the actual herds, but the optimization model recommended heavier culling for cows calving between January and August. The DP did not account for diseases and did not allow replacements during the first 2 months of lactation and some of the observed differences could be due to this. However, the results suggested that Finnish farmers might not be taking full advantage of the seasonality in milk pricing and production to maximize the profits of their herds — even though their culling decisions are rational and in quite close agreement with the optimized recommendations.