Mathematical models for the prediction of quantitative traits of different crosses of Holstein-Friesian cows in different farming conditions

Abstract

Background: To meet up the milk demand of growing population, farmers have to be adopted crossbred cattle and among various foreign breeds Holstein-Frisian crossbred became popular in terms of production results and high performance than other cattle genotype. For better decision making and proper management, use of mathematical models to predict values from partially or incompletely recorded data can be a great option. Predicted values from mathematical models may reduce the confusion for calculating the yield. Therefore, this study was conducted to determine suitable and easier mathematical model by comparing their predicted values. This study also targeted to identify better cross of Holstein by visualizing the differences in their production level. Methods: Recorded data of single lactation milk yield of two crossbred of Holstein-Frisian (Holstein-Friesian (HF)× Sahiwal (SL) and Holstein-Friesian (HF)× Local (L)) were used to calculate the mean value of daily average milk yield, lactation yield, peak yield and live weight of cattle. Four equational models (Linear, Exponential, Polynomial and logarithm) were used to determine their predicted value and based on fit statistics (R2 and CV) these models were evaluated. Result: The daily average milk yield and lactation yield was higher in HF×SL in farm A. Variation between two genotypes and among farms were observed. Live weight gain was greater in Sahiwal cross with Holstein-Friesian. Better production level was noticed in their 3rd lactation at four farming conditions. To determine the suitable model fit statistics were used and based on that polynomial model had shown great result and was considered as fit model.