Phenotypic correlates of melanization in two Sceloporus occidentalis (Phrynosomatidae) populations: Behavior, androgens, stress reactivity, and ectoparasites.

Mechanisms underlying production of animal coloration can affect key traits besides coloration. Melanin, and molecules regulating melanin, can directly and indirectly affect other phenotypic traits including aggression, stress-reactivity, and immune function. We studied correlation of melanization with these other traits, comparing within- and between-population differences of adult male western fence lizards, Sceloporus occidentalis. We compared one high- and one low-elevation population in California where individuals are increasingly darker at higher elevations, working during comparable periods of the breeding season at each site (first egg clutch). We measured agonistic behaviors of free-ranging males in response to staged territorial intrusions (STIs). In other sets of males we measured baseline testosterone and corticosterone levels, and hormonal-reactivity to a stress handling paradigm. We counted ectoparasite loads for all males. There were no significant associations between individual variation in melanization and individual variation in any of the variables measured. However, analysis of behavior from the STIs revealed that males in the darker high-elevation population responded with more aggressive behavior compared to males in the lighter low-elevation population. Males in the low-elevation population had significantly higher mean baseline testosterone, but the two populations did not differ in adrenal function (baseline corticosterone or corticosterone after 1-h confinement stress). Males in the darker high-elevation population had higher mean mite loads compared to males in the lighter population. This array of phenotypic differences between the two populations, and the absence of trait associations when assessing individual variation, do not parallel the patterns in other vertebrates. We describe potential differences in selective regimes that could produce these different patterns across vertebrates. These data suggest that hormonal pleiotropy does not constrain phenotypic variation.