Runaway evolution from male-male competition

Abstract

Wondrously elaborate weapons and displays that appear to be counter to ecological optima are widespread features of male contests for mates across the animal kingdom. To understand how such diverse traits evolve, here we develop a quantitative genetic model of sexual selection for a male signaling trait that mediates aggression in male-male contests and show that an honest indicator of aggression can generate selection on itself by altering the social environment. This can cause selection to accelerate as the trait is elaborated, leading to runaway evolution. Thus, an evolving source of selection provided by the social environment is the fundamental unifying feature of runaway sexual selection driven by either male-male competition or female mate choice. However, a key difference is that runaway driven by male-male competition requires signal honesty. Our model identifies simple conditions that provide clear, testable predictions for empirical studies using standard quantitative genetic methods.

Keywords: aggression; honest signals; indirect genetic effects; male-male competition; quantitative genetics; runaway evolution; sexual selection; social evolution; weapons.

FIGURE 1

Evolution of a male signal, body size, and aggression in response to male‐male competition. Panels (a) and (b) show evolutionary trajectories for each trait over 200 generations, and panels (c) and (d) show predicted lines of equilibria (heavy line) and their stability (gray arrows). In all panels, all three traits have the same genetic variance ($G=1$), benefit ($b_A=0.2$) and cost of aggression ($c_A=0.05$), fitness cost deriving from the threat of a male's opponent ($c_B=0.2$), cost of signal size ($c_s=0.05$; with naturally selected optimum ${\theta }_S=0$), and a responsiveness of aggression to body size (${\delta }_{\mathit{AS}}=0.4$). The line of equilibria is calculated from Equation 9a using these values. In panels (a) and (c), signal size is weakly correlated with body size ($G_{\mathit{SB}}=0.4$), while in panels (b) and (d), the two traits are more strongly correlated ($G_{\mathit{SB}}=0.8$). When the genetic correlation between signal size and body size is weak, all three traits reach equilibria (a), with equilibrium aggression predicted solely by costs and benefits. Signal size and body size reach a point on the predicted line of equilibrium (c) that differs depending on their starting values. When the genetic correlation is strong, aggression still reaches an equilibrium, but signal size and body size run away together (b), overshooting the predicted line of equilibria (d). As in Fisherian selection from female mate choice (Lande, 1981), male‐male competition can drive traits to runaway elaboration or extinction when the line of equilibria is unstable (d)

FIGURE 2

Stronger dependence of male aggressive behavior leads to more elaborate traits at equilibrium. Panel (a) illustrates a relatively weak influence of opponent signal on male aggression (${\delta }_{\mathit{AS}}=0.4$), while panel (b) illustrates a stronger influence (${\delta }_{\mathit{AS}}=0.8$). In each panel, we use starting values for traits relevant to the highly sexually dimorphic earwig Labidura riparia, which uses its forceps as a signaling trait and is shown to the right of each panel (drawing modified from Lucas, 1920). Other parameters are the same as Figure 1a. When the influence of opponent signal is weak (a), both body size and signal show a moderate evolutionary increase. When the influence is stronger (b), both body size and signal increase more, but the final signal size is much larger relative to body size. The highly elaborate elongate forceps in panel (b) may be found in other earwig species like Forcipula gariazzi