Feminization in common terns (Sterna hirunda) : relationship to persistent organic contaminants

Abstract

Concern about skewed sex ratios and female-female pairings among endangered roseate terns (Sterna dougallii) on Bird Island in Buzzards Bay, Massachusetts prompted studies with common terns from the same site as a surrogate species. Over seventy percent (11115) of male common tern embryos sampled from this site in 1993 had testes containing areas of ovarian cortical tissue (ovotestes), suggesting that terns may be affected by endocrine-disrupting contaminants. These terns are exposed to non-ortho PCBs which bind to the aryl hydrocarbon receptor (AhR), and lower chlorinated PCBs, hydroxymetabolites of which bind to the estrogen receptor. Our objectives were to document the presence of ovotestes in common tern embryos from Bird Island and Nauset, a reference site, in 1994, and to determine the relationship between environmental contaminants and ovotestes development, as well as other health-related effects. Pipping tern embryo gonads were examined histologically, and yolk sacs were extracted and analyzed for PCBs and chlorinated pesticides. Extracts also were analyzed for dioxin equivalents (TCDD-EQs) using a chick embryo hepatocyte (CEH) bioassay. Total PCBs were significantly higher in Bird Island tern embryos (mean 114, range 17- 663 ugig lipid) than Nauset (mean 35, range 8- i 78 ugig lipid); but were variable at both sites. Total PCBs were highly correlated with TCDD-EQs. Tern hepatic EROD activity was relatively insensitive to induction; only when TCDD-EQs were above 82 ± 26 ng/g lipid were EROD activities elevated. Levels of organic pesticides were below levels thought to be of toxicological significance. The percentage of male tern embryos with ovotestes at Nauset (60%) and Bird Island (78%) was high and not significantly different; ovotestes in terns from both sites ranged in severity from absent (1) to intersex (4). There was no significant relationship between ovotestes severity and any of the contaminants measured. However, the data suggested a contaminant level threshold of 100 ug/g lipid total PCBs and 30 ng/g lipid TCDD-EQs above which the formation of ovotestes in tern embryos is more likely to occur. Principal component analysis of PCB isomer patterns distinguished between Bird Island and Nauset sites, with Bird Island having relatively higher levels of lower chlorinated PCBs; however, there was no distinction between terns with ovotestes and those without. Common tern prefledglings and paired same-nest eggs were collected from Bird Island in 1995 to examine the persistence of ovotestes. Gonadal histology revealed no ovarian tissue on testes, indicating that the ovotestes do not persist after 3 weeks posthatch. Our data suggested that: 1) ovotestes could be related to contaminants; 2) a background level of ovotestes may be present, but elevated by contaminant exposure; or 3) ovotestes could be normally present in male common tern embryos at hatching. Observations of the normal presence of persistent cortical tissue in the testes of several avian species suggest that concern about feminizing effects of contaminants on bird populations as measured by ovotestes development in hatching birds may be exaggerated. Before conclusions can be made about terns from Bird Island, it is necessary to establish whether persistent ovarian cortical areas in testes are normally present. Studies with common terns from a pristine site are currently underway.