The body weight of the dairy cow II. Modeling individual voluntary food intake based on body weight and milk production

Abstract

Voluntary food intake of the dairy cow is an important variable in the dairy herd operation, We are unable to measure it individually in the conventional dairy operation when cows are kept in groups or grazing. This prevents using voluntary food intake as a parameter in the management decision-making process for the individual cow. Existing formulas that calculate dry matter intake (DMI) from ration and performance variables (see Mertens 1985), are not applicable for estimating DMI of an individual cow for on-line decision making such as adjusting ration density by computerized self feeders (Maltz et al. 1991, 1992Maltz and Metz 1994). In this work we calculated the dairy cows' DMI on a daily and individual basis, employing a non-conventional approach involving only body weight (BW) and milk yield (MY). These two variables are measurable on-line daily for individual cows. Daily MY, BW and DMI data of 24 cows during 147 days of lactation were obtained from the IMAG-DLO research farm in Duiven, the Netherlands, as described by Devir et al. (1996). A multivariable non-linear regression model was developed (Makidridakis et al. 1983) using MY, BW and interrelations between them as independent variables and DMI as the dependent variable. The steps of selecting variables and model were: (1) Visual analysis of plotted daily BW, MY and DMI data of all cows separately; conclusions were withdrawn regarding similar patterns in all cows. (b) Statistical determination of the variables' daily fluctuations. (c) Factors' selection and filtration using Cross-Correlation analysis. (d) Choosing and evaluating the model. (e) Running the model with statistical analysis of the filtered factors, using SPSS statistical software, especially the function Cross-Correlation and Step-Wise software. The model wad calibrated on 8, 12, 16, and 20 randomly selected cows and its validity was checked each time on the rest of the cows. No significant improvement in DMI prediction was found beyond 16 cows' calibration. The model predicted DMI efficiently in terms of actual values, timing of maximal DMI and its patters (Fig. 2). It was concluded that the selected variables are strongly correlated with DMI and that Polynomial Fitting of daily model calculated DMI (Fig. 3) enables to use it as a decision-making aid in calculating individual cows ration energy density (Maltz and Metz 1994). More model testing is needed regarding different rations and cows before this model can be used as a decision-making aid.