The way an animal perceives and reacts behaviorally to its environment is mediated largely by its physiology. For example, in vertebrates, glucocorticoids are released in response to a perceived stressor and help to keep the animal in a heightened state of awareness to enable an appropriate response to the perceived threat. It is also possible for organisms to react in non-adaptive ways to novel stressors. For example, an organism could perceive novel species, such as humans, as threats, when in fact the humans are engaging in recreational activity. This incorrect assessment could lead to wasted energy (particularly detrimental in poor-quality habitats) or the desertion of quality habitat, often to species that react less strongly to the presence of humans. It could potentially even lead to chronic elevation of glucocorticoids, reducing ability to mount a stress response to actual life-threatening stressors. Additionally, chronically elevated glucocorticoid levels may affect the fitness of the offspring of stressed individuals.

These physiological responses can alter fitness sub-lethally. Such sub-lethal effects could reduce population fitness more than lethal impacts because more individuals might experience acute stress that interferes with reproduction. Thus, understanding organismal physiology and how it is modified in response to novel stressors, including habitat loss, human recreational activities, or even invasive species, is a key way to examine individual and population-level reactions to anthropogenic activities.

Using the widespread freshwater turtle, Chrysemys picta, as a model system, I identified responses to anthropogenic stressors and consequences of transgenerational transmission of stress effects. The research presented here quantifies these stress responses across several levels of biological organization: measures of multiple traits within individuals as well as repeated measures over time, comparisons of different populations to understand how the stress response varies across an urban-rural gradient, and degrees of individual differences in physiology and behavior in offspring exposed to varying levels of simulated maternal stress at oviposition to understand the consequences of heightened maternal stress on fitness.

In the first study, I measured an important biomarker of the stress response, plasma corticosterone (CORT), in two populations of C. picta that differed primarily in the level of human recreational activity to which they were exposed during the reproductive season. Individuals from both populations exhibited similar levels of circulating CORT despite drastic differences in the number of humans they encountered while performing critical behaviors such as basking, mating, and nesting. In the second study, I measured differences in an important behavioral measure, flight initiation distance, between these same two populations of C. picta. Turtles from the population exposed to few humans had longer flight initiation distances than the population exposed to thousands of humans engaged in common recreational activities in the area. That is, turtles less familiar with humans did not allow researchers to approach as closely as did turtles that encountered humans on a regular basis. In the final study, I topically applied CORT to recently laid C. picta eggs as a proxy for three levels (low, medium, and high) of maternal stress. I measured important markers of hatchling fitness, such as survivorship to hatch, size, and mass, performance in righting trials (to replicate predation events), and dispersal ability. Elevated levels of CORT at oviposition decreased probability of embryos surviving to hatch. Furthermore, hatchlings from CORT-treated eggs did not right themselves as often as their conspecifics, but were quicker when they did. However, these behavioral differences did not translate into differential survivorship in the release experiment. This body of work provides insight into how freshwater turtles may react to increasingly human-modified environments.