A field study on serum vitellogenin levels in male Reeves' pond turtles (Chinemys reevesii) from estrogen-contaminated sites and a reference site

To ascertain whether wild male turtles were influenced by environmental estrogens, we examined serum vitellogenin (VTG) levels of male Reeves' pond turtles (Chinemys reevesii) collected from four study sites (A-D) in Kyoto, Japan. Sites A-C, which were impacted by domestic or industrial wastewater and effluents from sewage treatment plants, were chosen as contaminated sites, and site D was intended as a reference site. This contaminated/reference site characterization was confirmed by measuring estrogenic activities of the water samples collected at each site for over a year. Serum VTG levels in the turtles were quantified by an enzyme-linked immunosorbent assay established previously. Estrogenic activities of the water samples were measured using a previously validated yeast two-hybrid assay and expressed as the estradiol-17β equivalent. Estrogenic activity was observed at high levels at sites A-C, but was almost undetected at site D throughout the sampling period: the mean and range were 0.74 (< 0.07-2.1), 0.52 (0.17-1.6), 1.7 (< 0.07-7.3), and 0.07 (< 0.07-0.62) ng/l at sites A-D, respectively. Significant differences were found only in site D versus sites A, B, and C. Therefore, site D and sites A-C were confirmed to be a reference site and contaminated ones, respectively. Overall, 320 male turtles were captured and examined. The majority of the turtles showed normal VTG values (0.10-0.74 μg/ml). Although only five turtles from sites A-C showed unusually high VTG values (1.1-5.9 μg/ml, nearly one order of magnitude higher than normal values but much lower than values in adult females), there was no significant difference in the incidence of these high values between sites A-C and site D. Moreover, among the five turtles, one turtle was captured again 2 months later, but its VTG value dropped to the normal level. The unusually high VTG values may therefore be transient elevation caused by incidental and/or individually specific agents. Excluding the unusually high values, the mean serum VTG (accompanied with the range) was 0.22 (0.10-0.74), 0.27 (0.11-0.62), 0.27 (0.17-0.68), and 0.23 (0.10-0.57) μg/ml at sites A-D, respectively. There were no significant differences in the mean VTG values among sites A-D. Although it should be clarified how and why the unusually high VTG values occurred at sites A-C, our results suggested that wild male C. reevesii turtles would not be significantly affected by xenobiotic estrogens at environmentally relevant levels in terms of serum VTG elevation.