Effects of habitat on physiology and infection in aquatic turtles

To illustrate the importance and usefulness of physiologic and health data for conservation, I used slider (Trachemys scripta), painted (Chrysemys picta), and snapping (Chelydra serpentina) turtles that were collected from ponds throughout east and southeast Illinois. At each pond, I measured 48 habitat variables that included both physical pond characteristics and the land use surrounding ponds at multiple spatial scales. The relationships between and among habitat characteristics and patterns of glucocorticosteroids, heterophil/lymphocyte (H/L) ratios, parasitism, and infection with the zoonotic bacterium, Salmonella were examined.;Habitat characteristics strongly influenced slider turtle corticosterone levels in the first two years of our study. In 2004, baseline corticosterone levels and the adrenocortical response of 46 turtles varied more extensively between the two ponds sampled than with any host characteristic measured. When 78 turtles were sampled from additional ponds in 2005, baseline corticosterone levels were higher in ponds that lacked vegetation, and increased with the percentage of agriculture surrounding ponds. The fact that habitat characteristics in 2005 increased turtle baseline corticosterone levels but were not generally associated with decreased adrenocortical responsiveness may suggest that they act as chronic stressors without impairing hypothalamic-pituitary-adrenal (HPA) function.;Unlike 2004 and 2005, habitat characteristics did not influence baseline corticosterone levels or the adrenocortical responses of 46 turtles in 2006. Nevertheless, turtles had elevated baseline corticosterone levels and impaired adrenocortical responsiveness to adrenocorticotropic hormone (ACTH) injections that year, possibly due to a decline in regional habitat quality and a corresponding production of corticosterone at maximal or near maximal physiological limits. H/L ratios were positively correlated with the percentage of water surrounding (within 125 m) ponds. Since both turtle abundance in ponds and average body condition declined from 2004/2005 to 2006, these data may suggest that turtles with higher body conditions emigrated from ponds when additional water bodies were available while turtles with lower body condition, higher baseline corticosterone levels, higher H/L ratios, and lower adrenocortical responsiveness remained in the ponds.;Turtle parasitism was also altered by habitat characteristics. When five turtle species were pooled, leech intensity was higher in more turbid ponds. When we used a more restrictive model that examined slider, painted, and snapping turtle data separately, we did not identify a relationship between turbidity and leech parasitism. The prevalence and/or abundance of almost all parasite species in slider turtles increased significantly with the percentage of development and water surrounding ponds within 5000 and 10,000 m, however. Slider turtles in ponds with high human use also had a higher abundance of two of three gastrointestinal parasites that required intermediate hosts than individuals in ponds with low human use. Other habitat characteristics, such as the percentage of agriculture surrounding ponds, whether ponds were natural or manmade, and basking site availability, also significantly affected some parasites; however, the direction of the relationships varied among parasite species. Gastrointestinal parasite richness, measured as the average number of species identified in each pond, did not vary significantly among ponds. Although my data suggested that some parasites could impact host fitness, no parasites appeared to affect turtle population stability in the timeframe of this study.;Finally, Salmonella infections in slider turtles were not significantly affected by the host or habitat characteristics measured in this study. No turtles out of 100 were shedding Salmonella. Dissections of 73 turtles, however, revealed that turtles in the region had a moderate prevalence (11%) of Salmonella infection. (Abstract shortened by UMI.)...