ChileanJAR

The objective of the present work was to test a new mathematical expression destined to improve the interpretation of somatic cell counts and milk production levels. A multiple model of the lactation curve it was compared to the classical Wood model separately for each of the cell ranges (RC). A refined data base was used taken from a control system of five dairy herds, which represent a total of 1200 Holstein-Friesian dairy cows. The milk production curve was adjusted, separately for adult cows (AC) and first lactation cows (FLC), via the equation y(t) = a t^b e^-ct. The multiple model proposed was based in the previous equation incorporating the variable RC in the form of y(t, RC) = a t^b e^{-c t + d RC + f RC2}, where y (t, RC) = milk production in the t^th week for an AC or FLC with a cell range RC. The goodness of fit for individual models varied between 0.36 and 0.74 for FLC and 0.77 and 0.93 for AC. The multiple model with coefficients of determination of 0.59 and 0.89, for AC and FLC respectively, was found simpler and applicable. With respect to total estimated milk production, this multiple model permitted the inference that in both animal groups the difference between the lowest and the highest range is in the order of 25-28%. It is concluded that the multiple model proposed, seems to be simpler to understand and apply than the traditional Wood model.