Parámetros genéticos en la varianza residual de variables de comportamiento en toros de lidia

Abstract

The objectives were to analyze the presence of genetic variance in the residual variance of behavior notes in fighting bulls. From a livestock ranch, the conditional residuals were analyzed (RCON; e = y - E[y|a] = y - Xβ - Za) and marginal (RMAR; ξ = y - E[y] = y - Xβ), from the notes of horse tienta (n, 2369), foot tienta (n, 1432), horse bullfight (n, 792) and foot bullfighter (n, 820); in two databases: BD1, logarithmic transformation; BD2, logarithmic transformation eliminating extreme data (x ±2σ). With the MTDFREML software, a multivariate analysis (four notes) was performed within the residual type and for each database; heritability (h²) and genetic correlations (rg) were calculated through variables. The h2 ranged between 0.14 and 0.96, with an average value of 0.58; h2 of RMAR in BD2 presented above estimates (> 0.78). These results show genetic variability in the residual variance and present the possibility of response to selection (environmental channeling) to reduce environmental variance. The selection for residual variance can improve the response to the concepts of resilience, plasticity, robustness and rusticity. The rg were positive, in the range of 0.01 to 0.99 and an average value of 0.79; with these rg there may be a correlated response in the residuals of the four notes.