Hormonal mediation of alternative strategies: Integrating variation in antipredator responses, polymorphism, and learning

In response to stressful events such as a predator encounter, vertebrates rapidly increase plasma corticosterone levels, yet the immediate behavioral consequences, subsequent preparative effects, and geographical variation of this response are understudied. I examined the antipredator behavioral effects of corticosterone in males of two lizard species. In response to a caged predator (collared lizard, Crotaphytus nebrius) in outdoor enclosures, tree lizards (Urosaurus ornatus) with exogenously elevated corticosterone respond more quickly and hide longer than control lizards. Field measures are consistent with these findings: males with higher endogenous corticosterone are warier in staged encounters with this predator. Inhibiting corticosterone elevation during an encounter with a novel attacker also impairs immediate escape responses, and limits learning during subsequent encounters in eastern fence lizards (Sceloporus undulatus). Males with pharmaceutically-blocked acute corticosterone responses do not alter escape behavior during repeated encounters with the attacker (toy car) but unblocked lizards progressively increase flight-initiation distance and decrease hiding duration. Exogenously restoring corticosterone levels in blocked lizards maintains aversive learning at control levels. A parallel experiment with free-ranging males corroborates these results. Thus, my research demonstrates that elevated circulating corticosterone enhances immediate antipredator behaviors and also mediates learning and remembering novel predators in wild lizards. Given these presumably adaptive effects of the hormonal stress response, I examined the geographical variation in functional integration of several morphological and physiological traits related to social and antipredator tactics in tree lizards. Morphology associated with social tactics (throat color) and those presumed for crypsis (dorsal pattern) were correlated with each other in males, but the pattern of which trait combinations were most abundant differed among five sites. There was, however, a site-independent association between stress-induced corticosterone levels and the dorsal pattern in males. Despite some levels of phenotypic trait integration, no particular substrate type surrounding perch locations was associated with any specific trait complex. In fact, specific substrate types within each microhabitat associated independently with the morphological and hormonal traits. Given that fitness of organisms depends on the functional integration of many traits, my study illustrates that although geographical variations in habitat quality and stability may have a strong role in predicting phenotypes, selection for trait complexes are less clear. Furthermore, the fact that I did not detect integration among these phenotypic traits in females suggests that alternative behavioral strategies associated with these traits in males may be a strong target of selection.