The relationship between dewlap size and performance changes with age and sex in a Green Anole (Anolis carolinensis) lizard population

Abstract

Ornaments are believed to signal an individual's ability to reproduce successfully and/or survive. Since an individual's fitness is often influenced by multiple traits (e.g. number of copulations, ability to acquire nest sites or to escape predators), which are difficult to quantify simultaneously, we examine performance traits (bite force, jumping performance) believed to be relevant to an individual's fitness. Specifically, we ask whether variation in dewlap size is related to variation in body size, bite force and jumping ability in the lizard Anolis carolinensis. Our results show that dewlap size is correlated with jumping capacity across all individuals, whereas the relationships between dewlap size, body size and bite force differ depending on sex/age class. We argue that selection against relatively large dewlaps at the transition between small mature and large mature males might be responsible for the lack of a relationship within large males. The absence or presence of a correlation between dewlap size and bite force, on the other hand, might be explained by differences in behaviour, such as territory establishment, anti-predator tactics, and/or mate choice. Our work thus suggests that selective forces influencing the evolution of ornaments may operate differently on different sexes and life-history stages.

Appendix

Directional selection, quantity i, is calculated using Eq. 1; stabilizing (if negative) and disruptive selection (if positive), both quantity j, are calculated using Eq.2 (Endler 1986).

$$i = \frac{{\bar x_{\rm a} - \bar x_{\rm b} }}{{\sqrt {v_{\rm b} } }}\nonumber$$

(1)

$$j = \frac{{v_{\rm a} - v_{\rm b} }}{{v_{\rm b} }}$$

(2)

where$\bar x_{\rm a}$ is the mean value in trait × after selection; $\bar x_{\bf b}$ the mean value in trait × before selection; v _a the variance in trait × after selection; v _b the variance in trait × before selection.

The statistical significance of directional selection, i, is tested using Eq. 3; the statistical significance of stabilizing or disruptive selection, j, is tested using Eq. 4 (Endler 1986).

$$t_{(n - 2)} = \frac{{\bar x_{\rm a} - \bar x_{\rm b} }}{{\sqrt {n[(N_{\rm b} - 1)v_{\rm b} + (N_{\rm a} - 1)v_{\rm a} ]/(n - 2)N_{\rm b} N_{\rm a} } }}\nonumber$$

(3)

$$F_{(N_{\rm b} ,N_{\rm a} )} = \frac{{v_{\rm b} }}{{v_{\rm a} }}\quad {\rm or}\quad F_{\left( {N_{\rm a} ,N_{\rm b} } \right)} = \frac{{v_{\rm a} }}{{v_{\rm b} }}\quad ({\rm whichever}\;{\rm is}\;{\rm larger})\nonumber$$

(4)

where N _a is the sample size after selection; N _b the sample size before selection; and N=N _a+N _b.