Analysis and experimentation on ecosystems

{\textless}p{\textgreater}Anticipating the genetic and phenotypic changes induced by natural or artificial selection requires reliable estimates of trait evolvabilities (genetic variances and covariances). However, whether or not multivariate quantitative genetics models are able to predict precisely the evolution of traits of interest, especially fitness-related, life-history traits, remains an open empirical question. Here, we assessed to what extent the response to bivariate artificial selection on both body size and maturity in the medaka Oryzias latipes, a model fish species, fits the theoretical predictions. Three populations were selected for divergent body size while maintaining a constant selection pressure against late maturity. The observed evolutionary trends did not match the predictions from a bivariate quantitative genetics "animal" model. The most parsimonious model identified environmental, but not genetic, covariances between both traits, which cannot explain why body size did not evolve in the line selected for a smaller body length. We investigated alternative mechanisms (including genetic drift, inbreeding depression, natural selection, scaling or genetic asymmetry issues, and undetected genetic correlations) but could not attribute the deviation from theory to any single explanation. Overall, these results question the ability of multivariate quantitative models to provide valid and operational predictions of the evolutionary response to multivariate selection on complex traits.{\textless}/p{\textgreater}