Monitoring mercury in the loggerhead sea turtle, Caretta caretta.

The validity of using blood samples and keratinized scutes for nonlethal routine monitoring of mercury (Hg) in loggerhead sea turtles, Caretta caretta, is evaluated in the context of how effectively these matrixes predict internal tissue Hg burdens and the different temporal scales of exposure they represent. Total Hg (THg) was measured in blood and scutes collected from live captures (n = 34) and liver, kidney, muscle, spinal cord, blood, and scutes collected from freshly stranded loggerhead turtles (n = 6) along the coast of the southeastern United States. Linear regressions between monitoring compartments and internal tissues from stranded animals were all statistically significant (r2 > 0.805, p < 0.015) but varied in their utility as a predictive tool depending on which tissues were paired. Blood was an effective predictor of THg in muscle (r2 = 0.988, p < 0.0001) and spinal cord (r2 = 0.988, p < 0.0001), while scute was the most accurate predictor of THg in liver (r2 = 0.948, p = 0.0010). The strength of the relationship between tissues types is believed to reflect the similarity in the temporal scales they represent and the variability in the fraction of methylmercury present. The stability of Hg in the scute matrix makes this tissue preferable for approximating long-term exposure, while blood Hg levels can be affected by recent changes in Hg intake. THg levels in blood and scutes from live captures were highly correlated (linear regression r2 = 0.926, p < 0.0001) and increased significantly with body mass (r2 = 0.173, p = 0.016 and r2 = 0.187, p = 0.012 respectively), further supporting thatthere is a component reflecting long-term accumulation of Hg in these matrixes. We also present a novel technique using the residuals from the blood-scute regression as an index of recent exposure (IRE). The interpretation of this value is derived from the comparison between the most recent Hg intake (which contributes to the Hg measured in the blood) relative to the average past intake (which is recorded in the scute). A stepwise multiple regression revealed a significant positive relationship between the IRE and the proximity of the capture site to the nearest major industrial river mouth (p = 0.0102). This suggests that there is an elevation of bioavailable Hg in nearshore habitats where terrestrial influences and anthropogenic impacts are high. Seasonal foraging site fidelity and the variability in environmental Hg may explain the high intraspecific variability and occasional highly contaminated turtle seen in this and previous studies.