Corticosterone stress response in tree swallows nesting near polychlorinated biphenyl- and dioxin-contaminated rivers.

We assayed baseline and stress-induced corticosterone concentrations from adult female and nestling tree swallows, Tachycineta bicolor, from New England, U.S.A., sites with different levels of contamination with polychlorinated biphenyls (PCBs) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Corticosterone was compared over 2 years from a highly contaminated PCB site along the Housatonic River (Berkshire County, MA, USA), a midrange contaminated site upstream, and a reference site. Adult females (n=29), sampled only in 2003, showed an inverted-U association with PCBs, with higher stress-induced corticosterone with midrange contamination than at the high-contamination site. In nestlings, stress-induced corticosterone was highest for the highly contaminated site compared with the other sites in 2003 (n=53, 29 nests), with no difference among sites in 2004 (n=93, 27 nests). In 2004, we began testing mechanisms underlying these changes in nestlings at the high- and low-PCB sites. Corticosterone response to dexamethasone injection (used to test negative feedback) was not different between sites, but stress-induced corticosterone was reduced at the contaminated site after adrenocorticotropin hormone injection (used to test adrenal responsiveness), suggesting an inhibited ability to mount a stress response. We also compared nestlings from a stretch of the Woonasquatucket River, Rhode Island, U.S.A., heavily contaminated with TCDD (n=80, 43 nests) with nestlings from an upstream site that had lower levels of TCDD and the Berkshire County reference site. Although there were no stress-induced differences, baseline corticosterone was lower at the higher TCDD site than at the reference site. Altogether these findings suggest that tree swallows chronically exposed to high PCB and TCDD levels exhibit altered baseline and stress-induced corticosterone responses, but the patterns of alteration might not be predictable.