A simple statistical model to estimate precision of 300-day milk and fat production for dairy cows

Abstract

In this paper we use a statistical model to examine the relative precision of lactation milk yield and milk fat yield, under various production recording regimes with different metering and sampling precision, recording frequencies combined with empirical daily variations in milk yield and fat composition. The analysis shows that if traditional monthly production recording was performed with devices of International Committee for Animal Recording (ICAR) precision, the resultant coefficient of variation (CV) for lactation yields of milk and fat would be 2.22 and 4.81%, respectively. These values are determined primarily by measurement frequency coupled with the natural daily biological variations in animal milk yield and fat composition. Meter and sampler precision play a relatively minor role. For example, our model shows that if the devices had double the CV prescribed by ICAR, the milk and fat yield CV would increase by only 0.20 and 0.19%, respectively. These are minute losses of precision. If the ICAR precision standards for meters and samples were relaxed, it could encourage development of innovative metering and sampling technologies that could provide much more cost-effective and convenient ways for farmers to conduct production recording.