Blue Mussels as Indicators of Stormwater Pollution in Nearshore Marine Habitats in Puget Sound: Proposed Revised Statement of Hypothesis

This report summarizes the feasibility of applying a probabilistic random sampling design for monitoring the status and trends of toxic contaminants in blue mussels (Mytilus spp.) in Puget Sound, and proposes an alternate design and study questions. The original intent of the Puget Sound Stormwater Work Group (SWG) was to apply a probabilistic study design to compare contaminant loads in mussels from urban growth areas (UGAs) to those from NOAA’s Mussel Watch program. The Mussel Watch program, which has been conducted for decades, involves collection of samples thought to represent ambient¹ conditions. The SWG recommended the Mussel Watch program in Puget Sound be expanded to improve our understanding of stormwater runoff-related issues in Urban Growth Areas (UGAs). Recently, the Washington Department of Fish and Wildlife (DFW) was tasked with (1), evaluating whether there were sufficient mussel populations in Puget Sound to support a such a monitoring program, (2) estimating how many sample sites would be required for a statistically rigorous UGA/non-UGA comparison and (3) determining whether a statistically valid design could be achieved by combining existing NOAA Mussel Watch stations with 30 to 50 new stations, in a probabilistic stratified random sampling design.

The SWG recommended the funding for these samples come from local jurisdictions through their National Pollutant Discharge Elimination System (NPDES) permit monitoring requirements. Sampling would begin in either December 2015 or December 2017 to coincide with NOAA Mussel Watch sampling timeframes and the SWG-recommended schedule for implementing the broader regional stormwater monitoring program.

A desktop census evaluation of mussels in Puget Sound (Appendix A) indicated a high likelihood that Puget Sound mussel populations are widespread and abundant enough to support a moderate monitoring program. However, mussels were not ubiquitous along all Puget Sound shorelines and uncertainty existed for some key areas, which cannot be resolved without ground-truthing.

Results from power analyses using existing recent data from NOAA’s Mussel Watch program indicated sample sizes needed to compare UGA and non-UGA conditions were too large to be feasible (Appendix B). We do not anticipate having enough funds to collect hundreds of samples. To conduct a simple t-test comparison, 100 to 220 stations in total would be required, depending on the contaminant of interest, to detect contaminant differences between UGA and non-UGA mussels. These results highlight the high variability of contamination in mussels within UGAs, which is thought to be related to the wide range of land-use types occurring within those areas. Land use types within Puget Sound UGAs range from highly impacted (developed), to relatively undeveloped (green space).

In addition to NOAA Mussel Watch, two additional efforts have developed a network of Mussel Watch sites in Puget Sound. A previous DFW project proposed a further extended Mussel Watch network program in Puget Sound. These projects have demonstrated the utility of citizen scientist volunteers to save costs and engage communities in monitoring and recovery efforts. We recommend that the SWG study coordinate with and complement these efforts.

Given these findings, the limitations of contributions from NPDES permittees, and other current constraints for developing a status and trends monitoring program for mussels in Puget Sound, DFW recommends an alternate study design to address the following questions:

1. What is the geographic extent and magnitude (tissue concentration) of chemical contamination in mussels across a gradient of land-use in UGAs?
2. What land-use characteristics (factors) are most highly correlated with contamination?
3. Can the relative abundance of contaminants (i.e., fingerprint patterns) in mussels be used to help identify sources of contamination?
4. How are these conditions changing through time?

A study design addressing these questions would apply limited resources to establish mussel stations systematically across the full range of nearshore land-use conditions, from highly urbanized to relatively undeveloped (Figure 2). Its aim would be to select mussel locations that represent as wide a range of stormwater-influenced conditions across UGA and non-UGA shorelines as possible, allowing for exploration and identification of stormwater-mussel contaminant patterns and adjustment of sample locations in future years.

¹ the intent behind NOAA’s study design is to represent ambient or average conditions, away from urbanized shorelines