estimate residual feed intake (rfi) in iranian dairy cows using multitrait stochastic regression framework

In the dairy industry, nutritional efficiency is typically assessed using the residual feed intake (rfi) method. the traditional application of this method relies on linear regression, which inherently overlooks how the components of rfi change over time, leading to potential inaccuracies in the findings. by employing a multitrait stochastic regression framework, the relationships were explored between milk production, live weight, intake of dry matter (dmi), and body condition score (bcs) throughout the lactation period. furthermore, at each measurement point, an animal effect was estimated for intake through a matrix regression analysis based on the variance covariance matrix and the animal effects of the three predictor traits. by comparing this predicted effect with the actual intake effect, an estimate was derived for the rfi. the model was evaluated using historical data collected from the iranian national breeding centre from 2008 to 2023, encompassing 1,852 lactations from 870 cows. the analysis revealed strong positive correlations between various animal effects, particularly for milk production and intake of dry matter (dmi), as well as between body weight and dmi. these correlations peaked around mid-lactation and remained stable over time, averaging around 0.4 for body weight and bcs. additionally, the correlations for milk and weight, dmi and bcs, as well as milk and bcs, showed a gradual decline as lactation progressed. on the legendre polynomial coefficient scale, the correlations were measured with high precision, indicated by an average standard error of 0.04, with minimum and maximum values of 0.02 and 0.05, respectively. the estimated animal effect for feed intake consistently demonstrated a strong correlation with milk production and, for most of the lactation period, also with body weight. however, the correlation with bcs was only moderate and turned negative during the latter half of lactation. the relationship between the average rfi throughout lactation and rfi at individual time points was consistently positive, exceeding 0.5, with peak correlations observed mid-lactation, exceeding 0.9. this suggests a robust and reliable model for understanding the dynamics of nutritional efficiency and its relationship with milk production and other physiological parameters in dairy cows.