Residual correlations, and not individual properties, determine a nest defense boldness syndrome

Abstract

Different behavioral traits often covary, forming a behavioral syndrome. It is poorly known whether this covariance occurs on the between-individual level and what its selective consequences are. We used repeated measures (N = 562 observation events) of individual tawny owl Strix aluco females (N = 237) to study the integrated effects of seasonal timing of reproduction and clutch size on boldness displayed during defense of their clutch, in relation to plumage coloration, including local recruit production as a selective force on these traits. Using a Bayesian multivariate mixed model, we quantified the covariances between these traits on phenotypic, residual, and between-individual level and used a structural equation modeling approach to test the significance of presumed causal relationship between these traits in an a priori hypothesized path. On the phenotypic level, boldness was determined through early timing of breeding and larger clutch size, and early breeding increased recruitment probability. However, this relationship was entirely due to residual covariances and was not present on the between-individual level. The low individual-level correlations did not constrain the capacity of the population to respond to evolution as quantified by average autonomy (a metric summarizing evolutionary constraint on multiple traits). In the tawny owl, the association between early breeding and bold behavior, which is favored by selection, is solely due to extrinsic, nonheritable factors. We conclude that phenotypic evidence is insufficient to demonstrate syndrome covariance. © 2014 The Author.