Using mathematical models for lactation curve of milk and fat production in Iranian Holstein cows

An appropriate mathematical model is required to fit into lactation curves to study the environmental and biological variables that affect milk production. This study examined the suitability of 5 mathematical models (linear and nonlinear form of Wood, Wilmink, Guo and Swalve and inverse polynomial function of Nelder) for lactation curve traits (milk, fat yield and fat content) of 305-d milk yield of Iranian Holsteins. Models were compared based on R2, RMSE, and MSPE criteria by using NLIN and REG procedures in SAS. Using GLM procedure in SAS, factors of herd, calving year, calving season, season of production, age at calving, age at test and days in milk were significant on daily milk yield and the fat content during production period (P < 0.01). The study revealed that the linear form of Wood model and Wilmink function were the best for describing milk, fat yield and fat content during 305-d of lactation.