Determination of Trace Amounts of Mercury in Fish Tissues: Results of a North American Check Sample Study

Abstract

Abstract Three homogenized fish samples were distributed to 29 laboratories for analysis for mercury content. The majority (19) of the laboratories used acid digestion followed by atomic absorption analysis to measure the mercury content. Six laboratories used neutron activation; while 2 used dithizone methods and 2 used pyrolysis followed by flameless atomic absorption. Results from 4 laboratories were grossly aberrant or incomplete. Graphic analysis of the results on 2 samples showed that most laboratories tended to get either high or low results with both samples and that several had more consistent results with low fat samples than with high fat samples. Comparison of the combined results by flameless atomic absorption spectroscopy with those obtained by the flame method (t-test) showed that the latter gave just significantly lower results but much poorer precision. Differences between the combined results by flameless atomic absorption spectroscopy and those by neutron activation were probably insignificant.