Trace elements partition between skin and vertebrae in the demersal shark Squalus megalops: implications for tissue selection in elasmobranch biomonitoring.

Sharks are used as sentinels of marine trace-element contamination. Here we compare two rarely examined matrices, skin and vertebrae, sampled from the same individuals. Seventeen elements were determined by ICP-OES in 72 skin and 72 vertebral samples from shortnose spurdogs (Squalus megalops) taken as discards by the artisanal line fishery of the Canary Islands, central-east Atlantic (2008-2009); individuals provided both tissues, and all comparisons between tissues were made within individual. The element profiles differed clearly between the two tissues (PERMANOVA with permutations restricted within individual, pseudo-F (1,69) = 40.5, partial R2 = 0.37, p < 0.001). Sixteen of the seventeen elements were more concentrated in the skin, manganese being the only exception, and the contrast was largest for magnesium (R2 = 0.82; 5560 ± 2077 against 1529 ± 945 mg/kg dry weight), zinc, copper, boron and vanadium. Cadmium, cobalt, molybdenum, nickel, lead and vanadium lay below the limit of quantification in most vertebral samples and are reported at screening level only. Concentrations decreased with body length in both tissues, and more steeply in the vertebra (Spearman rho to -0.72), a pattern consistent with growth dilution; the higher values recorded in males disappeared once length was accounted for (multivariate R2 from 0.27 to 0.006) and reflect the larger size attained by females. The two tissues were positively but only moderately correlated (rho = 0.24-0.64). Skin is therefore the more concentrated matrix and can be sampled without sacrificing the animal, which makes it the more practical basis for trace-element biomonitoring in this data-poor, heavily exploited species, whereas the vertebra carries a lower, growth-averaged signal. Body size must be accounted for whenever elemental burdens are compared.