Fisherian and "good genes" benefits of mate choice: how (not) to distinguish between them

Good genes models of mate choice are commonly tested by examining whether attractive males sire offspring with improved survival. If offspring do not survive better (or indeed survive less well), but instead inherit the attractiveness of their father, results are typically interpreted to support the Fisherian process, which allows the evolution of preferences for arbitrary traits. Here, I show that the above view is mistaken. Because of life-history trade-offs, an attractive male may perform less well in other components of fitness. A female obtains a "good genes" benefit whenever males show heritable variation in quality, even if high-quality males invest so much in sexual advertisement that attractiveness has no positive correlation with any other life-history trait than male mating success itself. Therefore, a negative correlation between attractiveness and viability does not falsify good genes, if mating with a high-quality male results on average in superior offspring performance (mating success of sons included). The heritable "good genes" benefit can be sustained even if sexually antagonistic genes cause female offspring sired by high-quality males to survive and reproduce less well. Neglecting the component of male mating success from measurements of fitness returns from sons and daughters will bias the advantage of mating with a high-quality male downwards. This result may partly account for the rather weak "good genes" effects found in a recent meta-analysis.