Characterization of Sediment Contamination in Central Bay Margin Areas

The Bay margins (i.e., mudflats and adjacent shallow areas of the Bay) are important habitats where there is high potential for wildlife to be exposed to contaminants. However, these areas are not routinely sampled due to logistical considerations. In 2015, the Regional Monitoring Program for Water Quality in San Francisco Bay (RMP) conducted a survey of margin areas in Central Bay to determine ambient concentrations of contaminants in sediment in these areas. The results generally confirmed the current conceptual model expectation that the margin sediments are often more contaminated that those in the subtidal open bay, with PCBs showing the largest differences. These differences between Central Bay margins and open water areas are likely the largest to be found in San Francisco Bay, as margins in other bay segments generally account for more of the total area and the adjacent land use is less heavily industrial.

The ambient data provide a useful baseline context against which the severity of contamination at specific sites can be compared. The baseline data will also be useful in setting targets and tracking improvements in watershed loads and their nearfield receiving waters, or for appropriate re-use or disposal of dredged sediment. These spatially unbiased data can also improve estimates of mean concentrations and contaminant inventories in margins. For example, based on data from this study, contamination in the margin areas accounts for ~20% of PCBs in Central Bay which is disproportionately high compared to the margin area (5% of Central Bay). Given a large inventory of contaminants in the open bay compared to annual loads, changes may be difficult to see at the bay scale in the short- and mid-term. However, with margins’ smaller inventories and greater proximity to likely sources, improvements in margin sediments may be easier to observe. This ambient sampling effort also identified or verified one expected area of very high PCB concentrations which had not been previously measured. However, the randomized monitoring design used for the Central Bay study was not ideal for identifying “hot spots”. Future studies with a goal of detecting hot spots should use deterministic sampling designs.