Sexual selection's impacts on ecological specialization: an experimental test

Abstract

In many species, individuals specialize on different resources, thereby reducing competition. Such ecological specialization can promote the evolution of alternative ecomorphs-distinct phenotypes adapted for particular resources. Elucidating whether and how this process is influenced by sexual selection is crucial for understanding how ecological specialization promotes the evolution of novel traits and, potentially, speciation between ecomorphs. We evaluated the population-level effects of sexual selection (as mediated by mate choice) on ecological specialization in spadefoot toad tadpoles that express alternative ecomorphs. We manipulated whether sexual selection was present or reversed by mating females to their preferred versus non-preferred males, respectively. We then exposed their tadpoles to resource competition in experimental mesocosms. The resulting distribution of ecomorphs was similar between treatments, but sexual selection generated poorer trait integration in, and lower fitness of, the more specialized carnivore morph. Moreover, disruptive and directional natural selection were weaker in the sexual selection present treatment. Nevertheless, this effect on disruptive selection was smaller than previously documented effects of ecological opportunity and competitor density. Thus, sexual selection can inhibit adaptation to resource competition and thereby hinder ecological specialization, particularly when females obtain fitness benefits from mate choice that offset the cost of producing competitively inferior offspring.

Keywords: competition; disruptive selection; mate choice; resource polymorphism; speciation.

Figure 1.

Distribution of combined measure of morphology for mesocosms where sexual selection was present (SS+) versus mesocosms where sexual selection was reversed (SS−). Higher values of morphological index are more carnivore-like, whereas lower values are more omnivore-like.

Figure 2.

Spline analysis (with 95% CIs) showing fitness minimum for intermediate phenotypes in both SS+ and SS− treatments for two fitness proxies: (a) SVL and (b) GS. Although both treatments experienced disruptive selection, the SS− treatment exhibited significantly stronger disruptive selection. Linear selection favouring the carnivore ecomorph was also significantly stronger in the SS− treatment, with SS− carnivores obtaining higher fitness than SS+ carnivores.

Figure 3.

Comparison of the correlation coefficients between each pair of morphological traits in the SS+ versus the SS− treatment. Greater integration of these traits (as indicated by a stronger correlation between them) is favoured for more successful resource acquisition (arrows indicate direction of correlation that is favoured). Error bars show 95% CIs. Asterisk indicates significantly different correlation between treatments. Trait abbreviations: OH: size of the orbitohyoideus muscle (larger OH favoured in carnivores); MP: mouthparts score (higher MP favoured in carnivores); DR: number of denticle rows (fewer DR favoured in carnivores); GC: number of gut coils (fewer GC favoured in carnivores).

Figure 4.

Comparison of the effects (Cohen's d ± 95% CIs) of manipulating conspecific density, ecological opportunity and sexual selection on the degree of bimodality, trophic divergence and strength of disruptive selection (for the fitness proxy SVL) in experimental mesocosms. Effect sizes represent the magnitude and direction of change in the response variable in response to the presence of (i.e. sexual selection), or increase in (i.e. conspecific density and ecological opportunity) the manipulated treatments.