Development of equations, based on milk intake, to predict starter feed intake of preweaned dairy calves.

There is a lack of studies that provide models or equations capable of predicting starter feed intake (SFI) for milk-fed dairy calves. Therefore, a multi-study analysis was conducted to identify variables that influence SFI, and to develop equations to predict SFI in milk-fed dairy calves up to 64 days of age. The database was composed of individual data of 176 calves from eight experiments, totaling 6426 daily observations of intake. The information collected from the studies were: birth BW (kg), SFI (kg/day), fluid milk or milk replacer intake (MI; l/day), sex (male or female), breed (Holstein or Holstein×Gyr crossbred) and age (days). Correlations between SFI and the quantitative variables MI, birth BW, metabolic birth BW, fat intake, CP intake, metabolizable energy intake, and age were calculated. Subsequently, data were graphed, and based on a visual appraisal of the pattern of the data, an exponential function was chosen. Data were evaluated using a meta-analysis approach to estimate fixed and random effects of the experiments using nonlinear mixed coefficient statistical models. A negative correlation between SFI and MI was observed (r=-0.39), but age was positively correlated with SFI (r=0.66). No effect of liquid feed source (milk or milk replacer) was observed in developing the equation. Two equations, significantly different for all parameters, were fit to predict SFI for calves that consume less than 5 (SFI5) l/day of milk or milk replacer: ${\rm SFI}{{,\lt,5}} {\equals}0.1839{{,\pm,0.0581}} {\times}{\rm MI}{\times}{\rm exp}^{{\left( {\left( {0.0333_{{,\pm,0.0021 }} {\minus}0.0040_{{,\pm,0.0011}} {\times}{\rm MI}} \right){\times}\left( {{\rm A}{\minus}{\rm }\left( {0.8302_{{,\pm,0.5092}} {\plus}6.0332_{{,\pm,0.3583}} {\times}{\rm MI}} \right)} \right)} \right)}} {\minus}\left( {0.12{\times}{\rm MI}} \right)$ ; ${\rm SFI}{{,\gt,5}} {\equals}0.1225{{,\pm,0.0005 }} {\times}{\rm MI}{\times}{\rm exp}^{{\left( {\left( {0.0217_{{,\pm,0.0006 }} {\minus}0.0015_{{,\pm,0.0001}} {\times}{\rm MI}} \right){\times}\left( {{\rm A}{\minus}\left( {3.5382_{{,\pm,1.3140 }} {\plus}1.9508_{{,\pm,0.1710}} {\times}{\rm MI}} \right)} \right)} \right)}} {\minus}\left( {0.12{\times}{\rm MI}} \right)$ where MI is the milk or milk replacer intake (l/day) and A the age (days). Cross-validation and bootstrap analyses demonstrated that these equations had high accuracy and moderate precision. In conclusion, the use of milk or milk replacer as liquid feed did not affect SFI, or development of SFI over time, which increased exponentially with calf age. Because SFI of calves receiving more than 5 l/day of milk/milk replacer had a different pattern over time than those receiving <5 l/day, separate prediction equations are recommended.