Female assessment: cheap tricks or costly calculations?

Abstract

Both commentaries raise the issue of the female capacity to assess complex cognitive traits. We did not focus on this in our review, but we broadly agree with their point: The costs of female assessment are an important constraint on the evolution of complex male displays. Indeed, a key feature of Miller’s (2000) ‘‘mating mind’’ hypothesis is that cognition for male displays and cognition for female assessment coevolve, influencing brain structure in both sexes. Whether females can afford to assess complex displays is an important empirical question. Madden et al. (2011) suggest that female bowerbirds have found a way around this problem by using male traits that may be cognitively demanding to produce but are nonetheless cheap to assess. Riebel (2011), surveying recent evidence from songbirds, argues that female preferences involve substantial learning and that costs may exert an influence during development, affecting female judgments of male song later in life. Costs for females are therefore a clear constraint, but interestingly evidence suggests that females are often willing to invest substantial time and energy in assessing males. Studies monitoring the mate-search behavior of females have shown that they typically make repeated visits to a series of males before selecting one as a mate (reviewed by Luttbeg 1996; Uy et al. 2001). This seems unnecessary if male quality is readily observable through fixed morphological characters, so perhaps indicates that females are assessing more dynamic traits, such as dietary (e.g., carotenoid) coloration or behavioral displays (Sullivan 1994). Madden et al. (2011) and Riebel (2011) prompt us to ask whether variation in female assessment behavior might be related to variation in female cognition. For instance, do females that revisit males numerous times before mating differ in their cognitive abilities from those that mate after one brief visit? Across species, do females tend to have larger brains when mate choice is based on complex behavioral displays, as opposed to fixed morphological traits? Madden et al. (2011) also note that complex cognitive ‘‘extended phenotypes’’ are susceptible to sabotage by other males and argue that this generates mate-choice errors and so reduces their value to females. We query this point. The ability of certain males to cheat or usurp the displays of other males might itself be attractive to females, whether because of direct benefits (resources stolen from other males) or indirect benefits (inherited cheating ability in offspring). Indeed, such ‘‘underhand’’ tactics might be precisely the kind of complex cognitive trait females should pay attention to. Clearly, this is an area ripe for more research, not only in bowerbirds but also with regards to food-caching and cache-pilfering corvids (Grodzinski and Clayton 2010). Finally, the suggestion of Madden et al. (2011) that both males and females contribute to brain size differences between bowerbird species leads to an interesting question concerning convergent evolution. Among songbirds, large relative brain size seems to have evolved independently in the bowerbirds (Ptilonorhynchidae), lyrebirds (Menuridae), and corvids (Corvidae) (Figure 1). In contrast to the many studies on cognition in the Corvidae, work on the other 2 families has mostly concentrated on their extreme sexually selected traits. We suggest that future research on the interaction between sexual