Assessing comparability and uncertainty of analytical methods for methylated mercury species in seawater.

BACKGROUND:The relative distribution and importance of monomethylmercury (MMHg) and dimethylmercury (DMHg) in seawater is still under debate. A lack of comparability between measurements at sub-picomolar levels hampered the further understanding of the biogeochemical Hg cycle. To overcome this, we assessed the relative standard measurement uncertainties (Uex,r) for direct measurements of MMHg and DMHg by species-specific isotope dilution ICP-MS and cryo-focusing GC-ICP-MS at femtomolar concentrations. Furthermore, Uex,r was determined for the indirect determination of DMHg (DMHgcalc = MeHg - MMHg) and MeHg (MeHgcalc = MMHg + DMHg) to compare the two methodologies. RESULTS:Expanded Uex,r (confidence interval of 95%) for cryo-focusing GC-ICP-MS was 14.4 (<50 fM) and 14.2% (>50 fM) and for SS-ID GC-ICP-MS 5.6 (<50 fM) and 3.7% (>50 fM). For concentrations above 50 fM, Uex,r for DMHgcalc was always lower than for direct measurements (14.2%). For MeHgcalc, on the other hand, Uex,r was always higher for concentrations above 115 fM (range: 3.7-13.9%) than for direct measurements (3.7%). We evaluated the comparability of directly measured and calculated DMHg and MeHg concentrations based on Hg speciation measurements for two vertical profiles in the Mediterranean Sea. We show that directly measured and indirectly determined DMHg and MeHg concentrations yield comparable results. SIGNIFICANCE:Our results validate the application of the indirect method for the determination of DMHg if a direct measurement method with a low Uex,r such as isotope dilution is used for MMHg and MeHg measurements. The validation of the indirect measurement approach opens new possibilities to generate more precise and accurate DMHg data in the global ocean.