Evaluation of automated streamwater sampling during storm events for total mercury analysis

Understanding the processes by which mercury is mobilized from soil to stream is currently limited by a lack of observations during high-flow events, when the majority of this transport occurs. An automated technique to collect stream water for unfiltered total mercury (HgT) analysis was systematically evaluated in a series of laboratory experiments. Potential sources of error investigated were 1) carry-over effects associated with sequential sampling, 2) deposition of HgT into empty bottles prior to sampling, and 3) deposition to or evasion from samples prior to retrieval. Contamination from carry-over effects was minimal (<2%) and HgT deposition to open bottles was negligible. Potentially greater errors are associated with evasive losses of HgT from uncapped samples, with higher temperatures leading to greater evasion. These evasive losses were found to take place primarily within the first eight hours. HgT associated with particulate material is much less prone to evasion than HgT in dissolved form. A field test conducted during a high-flow event confirmed unfiltered HgT concentrations sampled with an automated system were comparable to those taken manually, as the mean absolute difference between automated and manual samples (10%) was similar to the mean difference between duplicate grab samples (9%). Results from this study have demonstrated that a standard automated sampler, retrofitted with appropriately cleaned fluoropolymer tubing and glass bottles, can effectively be used for collection of streamwater during high-flow events for low-level mercury analysis.