Dataset: HerringHg

Fish were frozen immediately and processed in the laboratory within three months of capture. Total length (mm), wet mass (g), and sex were recorded for each fish and otoliths were extracted for age estimation.

Whole fish were thawed, homogenized individually, and a subsample of approximately 15 g of homogenate per fish was refrozen until the mercury analysis. Fish from both sampled areas in the Gulf of Maine were selected for analysis to cover the observed range of sizes. Total mercury was measured in accordance with U.S. EPA Method 7473 using a Milestone DMA-80 Direct Mercury Analyzer (Milestone Srl, Bergamo, Italy) at Clarkson University (Potsdam, NY). Detailed methods are described in Zananski et al. (2011).

For quality control, we used organic chicken breast meat (Chicken G1, Chicken G2, and Chicken W1) and wild-caught Alaskan sockeye salmon (Salmon G1, Salmon G2, and Salmon W1). G1 samples were run through the meat grinder before herring were ground up and G2 samples were run through the grinder after the herring were ground up. The W1 samples were whole samples (i.e., not run through the meat grinder). If G1, G2 and W1 for each of chicken and salmon were similar that would indicate no contamination during processing. If G1 and G2 are equally higher than W1 then the process adds Hg (from the grinder), and if G2 is higher than G1 then the grinder was not thoroughly cleaned likely resulting in some contamination from previous samples. 

Controls

We used an approximate 10% rule to flag potential outliers. So if we had estimates of 10, 10, and 15 we excluded 15, but if we had 10, 13, and 15 we did not because we did not see a sufficient trend to justify exclusion.