Comparison of different lactation curve models in Sahiwal cattle up to fourth parity using monthly test day milk yields

Abstract

In present study data of 6391 monthly test day milk yield (MTDMY) records of first lactation pertaining to 691, 4048 MTDMY records of second lactation pertaining to 454, 3013 MTDMY records of third lactation pertaining to 340 and 2280 MTDMY records of fourth lactation pertaining to 249 Sahiwal cows, spread over a period of 51 years (1961- 2011), maintained at NDRI, Karnal were used to compare different lactation curve models. The five different models viz. inverse polynomial function, Wilmink’s exponential function, gamma type function, quadratic cum log function and mixed log function were fitted on MTDMY by statistical package SAS version 9. The relative efficiency of the models was determined on the basis of the higher value of coefficient of determination (R²-value) and lesser value of root mean square error (RMSE). The R² values ranged from 72.96% (0.154 kg) for Wilmink’s function in first lactation to 99.98% (0.0.043 kg) for inverse polynomial function in third lactation with RMSE values in parentheses. The R² value for different lactation curve models showed that the inverse polynomial function gave best fit; whereas Wilmink’s function was found to be least fit in Sahiwal cattle. Hence it can be concluded that inverse polynomial function is the model of choice for estimating the trend of milk production in Sahiwal cattle.