Modelling of daily herbage intake and milk production by grazing dairy cows

Abstract

The wide range of grazing and feeding management of grazing dairy cows does not enable a simple and accurate estimate of their herbage intake. The purpose of this review is to describe the main predictive models of herbage intake by grazing dairy cows. The most recently published models take into account the characteristics of the herd, sward, and feeding and grazing conditions, and are included in decision support tools for grazing management (Pâtur’IN, Sepatou, Grazemore) or feeding management at grazing (GrazFeed, Diet-Check, INRAtion soon). The response of herbage intake to grazing conditions and supplementary feeding are simulated from several models. The accuracy of the existing models is studied from a global statistical approach comparing the actual herbage intake of experimental herds (n = 304) with the predicted values. The mean error of prediction ranges from 10 to 25 % according to the model, i.e. from 1.5 to 3.8 kg DM of herbage intake per cow per day. The low accuracy of the predictions is related first to the lack of robustness of the predictions for unusual grazing conditions. Moreover, few models enable to predict cow performance (milk production, body weight change) from the prediction of herbage and nutrient intake.