Assessment of nonlethal methods for predicting muscle tissue mercury concentrations in coastal marine fishes.

Caudal fin clips and dorsolateral scales were analyzed as a potential nonlethal approach for predicting muscle tissue mercury (Hg) concentrations in marine fish. Target fish were collected from the Narragansett Bay (Rhode Island, USA) and included black sea bass Centropristis striata [n = 54, 14-55 cm total length (TL)], bluefish Pomatomus saltatrix (n = 113, 31-73 cm TL), striped bass Morone saxatilis (n = 40, 34-102 cm TL), summer flounder Paralichthys dentatus (n = 64, 18-55 cm TL), and tautog Tautoga onitis (n = 102, 27-61 cm TL). For all fish species, Hg concentrations were greatest in muscle tissue [mean muscle Hg = 0.47-1.18 mg/kg dry weight (dw)] followed by fin clips (0.03-0.09 mg/kg dw) and scales (0.01-0.07 mg/kg dw). The coefficient of determination (R (2)) derived from power regressions of intraspecies muscle Hg against fin and scale Hg ranged between 0.35 and 0.78 (mean R (2) = 0.57) and 0.14-0.37 (mean R (2) = 0.30), respectively. The inclusion of fish body size interaction effects in the regression models improved the predictive ability of fins (R (2) = 0.63-0.80; mean = 0.71) and scales (R (2) = 0.33-0.71; mean = 0.53). According to the high level of uncertainty within the regression models (R (2) values) and confidence interval widths, scale analysis was deemed an ineffective tool for estimating muscle tissue Hg concentrations in the target species. In contrast, the examination of fin clips as predictors of muscle Hg had value as a cursory screening tool; however, this method should not be the foundation for developing human consumption advisories. It is also noteworthy that the efficacy of these nonlethal techniques was highly variable across fishes and likely depends on species-specific life-history characteristics.