Longitudinal genetic analysis for growth traits in the complete diallel cross of rainbow trout (Oncorhynchus mykiss)

•Complete diallel cross experiment of five strains was conducted in rainbow trout.•Longitudinal genetic analysis was employed to estimate changes in genetic effects.•Optimal genetic models were selected by Bayesian information criterion.•For body length, there were 11 genetic effects significantly changing with time.

A complete diallel cross involving the five strains from Bohai, Denmark, Donaldson, Norway and California and self-breeding within each strain were carried out to evaluate general combining ability, specific combining ability, reciprocal, maternal, linebred effects and heterosis for two growth traits in rainbow trout. To estimate the changes in various genetic and permanent environment effects, longitudinal genetic analysis was employed and implemented with restricted maximum likelihood (REML) method. Using Bayesian information criterion (BIC), the longitudinal genetic model with linear genetic effects and Legendre polynomial of order 2 for permanent environment effects was chosen as the optimal one for body length, whereas genetic effects were constant for body weight. Significance tests for regression effects show that for body length, there were 11 genetic effects changing with growth time. Among all cross combinations, the specific combining ability between Bohai and Donaldson was the smallest across the measuring period. No genetic effect was found to be consistently high or low between the two analyzed traits.