Prediction of dry matter intake of lactating dairy cows with daily live weight and milk production measurements

In dairy farming, monitoring tools are developing rapidly and the possibility of combining available data to generate additional information raises new questions. The high throughput body weight is a new and frequent phenotype on dairy farms. Feed intake monitoring in dairy cows remains largely absent, even though it is useful for herd management. The purpose of this work was to predict the dynamics of dry matter intake with a model using live weight, milk production and ration characteristics. This model is based on the assumption that changes in body weight are closely related to dry matter intake through two components: short-term changes in digestive content related to ingestion and long-term changes in body reserves, which are linked to the cumulative energy balance of the cow since calving. The energy balance results from the difference between the energy inputs from the ingested quantities and the energy requirements related to milk production. The model was tested in a trial with 65 cows receiving the same complete ration during the first 18 weeks of lactation. Observed individual intake was predicted well by the model, with average weekly errors of 2.3 kg, although the daily errors are still high (SD = 3.31 kg). Intake prediction during the first month of lactation showed a systematic bias and further investigation will be necessary to understand this bias and correct it