Runaway sexual selection without genetic correlations: social environments and flexible mate choice initiate and enhance the Fisher process

Abstract

Female mating preferences are often flexible, reflecting the social environment in which they are expressed. Associated indirect genetic effects (IGEs) can affect the rate and direction of evolutionary change, but sexual selection models do not capture these dynamics. We incorporate IGEs into quantitative genetic models to explore how variation in social environments and mate choice flexibility influence Fisherian sexual selection. The importance of IGEs is that runaway sexual selection can occur in the absence of a genetic correlation between male traits and female preferences. Social influences can facilitate the initiation of the runaway process and increase the rate of trait elaboration. Incorporating costs to choice do not alter the main findings. Our model provides testable predictions: (1) genetic covariances between male traits and female preferences may not exist, (2) social flexibility in female choice will be common in populations experiencing strong sexual selection, (3) variation in social environments should be associated with rapid sexual trait divergence, and (4) secondary sexual traits will be more elaborate than previously predicted. Allowing feedback from the social environment resolves discrepancies between theoretical predictions and empirical data, such as why indirect selection on female preferences, theoretically weak, might be sufficient for preferences to become elaborated.

Figure 1

The interaction coefficient Ψ. The genotypic range of a male trait that influences preferences, , is portrayed on the x-axis. The y-axis shows female preference, p, defined as the difference between the male trait value she chooses and the average male trait, . The inherent preference of females is the y-intercept given by a_p (solid circle). We illustrate females with inherent preferences for greater than average male trait values, which should be prevalent under directional and open-ended preferences. The change in acceptable trait values (dashed line) is dictated by Ψ. (a) Ψ= 0. Female preference does not change across the range of male phenotypes in the social environment. (b) Ψ > 0. Social interactions with larger male traits increase female preferences. (c) Ψ < 0. Social experience decreases female preference. The light shaded areas above the dashed lines indicate the male trait values, relative to the average, that a female will accept given her inherent preference and the social environment experienced. Dark shading in (c) indicates conditions in which females accept below average males.

Figure 2

The effect of IGEs on the likelihood of runaway sexual selection. The solid circle represents a population that has been displaced slightly from a line of equilibrium with positive slope α, indicated by the heavy line. The dashed lines indicate the slope of the population's evolutionary trajectory as derived in the main text and indicated above the graphs. Both graphs show a situation in which covariance between trait t and preference p is nonexistent. The direction of evolution is indicated by small arrows. Traits lacking IGEs (i.e., when Ψ= 0) are shown in (a). The male trait evolves to a stable optimum due to the action of natural selection, and female preferences do not change except by drift. (b) The social environment provided by male traits affects expression of female preference. If Ψ exceeds α, the population evolves via runaway along a slope and in a direction indicated by the dashed lines. Once a covariance develops, runaway is enhanced.