Genetic analysis of growth curve traits in Iran-Black sheep breed: Comparison of standard multivariate and structural equation models

In this study, a dataset comprising 20,328 body weight records from birth to 360 days of age was utilized to compare five growth curve models and genetic analysis of growth curve parameters in Iran-Black sheep. The data and genealogical information were collected between 1981 and 2007 from the Breeding Station of Abbasabad in the Khorasan Razavi province of northeast Iran. The performance of five statistical models including Brody, negative exponential, von Bertalanffy, Logistic, and Gompertz for describing the growth curve of the studied population was evaluated by applying the SAS software. The statistics l measures used for model comparisons were Akaike's information criterion (AIC), root mean square error (RMSE), and adjusted coefficient of determination ( ). The Brody model, exhibiting the highest  and the lowest values for both AIC and RMSE, was selected as the best model for characterizing the growth curve in this breed. Consequently, the parameters of the growth curve, including parameter A (considered as weight at maturity), B (considered as an integration constant), and K (maturation rate) were predicted by applying the Brody model. To investigate the effect of maternal components on the growth curve parameters, nine univariate animal models, including different combinations of the direct additive genetic, maternal additive genetic, maternal permanent environmental, and maternal temporary environmental effects, were fitted. Subsequently, two multivariate animal models, comprising the standard (SMM) and fully recursive (FRM) models were analyzed by using the Bayesian inference. The FRM outperformed SMM in terms of lower means square error (MSE) and higher Pearson's correlation coefficients between the actual and predicted records (r(y, )) values, indicating better goodness of fit. The posterior means for heritability of A, B, and K parameters were low but statistically significant under SMM and FRM. It may be concluded that the growth trajectory traits of Iran-Black sheep are influenced mainly by non-additive genetic and environmental effects, emphasizing the importance of considering these effects for developing the corresponding breeding strategies. The Spearman's rank correlation coefficients between the estimated breeding values for growth curve traits under SMM and FRM indicated significant re-ranking of animals, favoring FRM for genetic evaluation in Iran-Black sheep.

Genetic analysis of growth curve traits in Iran-Black sheep breed: Comparison of standard multivariate and structural equation models Keywords:

causal effect , genetic parameters , goodness of fit measures , growth trajectory , sheep