Note from Doug Steding:

This post came about when Randi Wexler from Integral Consulting Inc. sent me an email saying they were working on a comment to one of my previous fish consumption posts. Integral has been tracking this issue closely for a number of years, and I was excited to be able to offer Integral’s perspective as a guest blog post. This is a collaborative effort on their part, and includes authors Ellen Ebert, Priscilla Tomlinson, and Randi Wexler.

One of my hopes when I started this blog a few years back was that it would become a forum for discussing emerging science, law, and policy issues in the Pacific Northwest. I appreciate Randi, Ellen, and Priscilla taking the time to write this post and hope that it isn’t the last from the folks at Integral Consulting Inc. I think you’ll find their perspective on Ecology’s rulemaking informative:

Doug raises important issues concerning the attainability of water quality standards (WQSs) based on highly protective fish consumption rates (FCRs). The meeting of Ecology’s Delegates’ Table for the WQS rule-making process on February 10 addressed issues of attainability, particularly related to arsenic, mercury, and PCBs, chemicals that present unique challenges for regulators and dischargers alike. We at Integral Consulting have identified some approaches that we believe should be considered during Ecology’s rule-making process to ensure that the WQS eventually established will be scientifically based and attainable while still ensuring protection of human health.

The selection of the most appropriate FCR to be used in developing WQSs for the state of Washington is a critical component of the rulemaking process. There are a wide range of fish consumption studies available that have targeted different populations of the state and have provided estimated FCRs for a variety of species and populations using variable data collection methodologies. The availability of these multiple and sometimes conflicting FCR estimates greatly complicates the FCR selection process. There are a number of ways that this process might be simplified.

First, it is important to compare apples to apples when selecting a default FCR to be used in deriving WQSs and then selecting fish species to be used in measuring compliance of persistent and bioaccumulative compounds with the resulting standards. This is because there are differences among species in terms of both their desirability and availability as food sources and their tendency to take up and bioaccumulate persistent compounds. For example, species such as white sucker, which are often selected for biomonitoring purposes, are bottom-feeding, resident species that have a high potential to take up sediment-bound constituents but are rarely consumed. At the same time, while anadromous species (e.g., salmon) tend to be consumed at higher rates than are bottom-dwelling species, their tendency to take up persistent compounds from localized sources is considerably reduced compared with resident species that have protracted exposure to localized concentrations of contaminants. Thus, it is not appropriate to base a default FCR on higher levels of consumption of anadromous species when deriving WQSs if compliance with those standards will be assessed by measuring chemical concentrations in resident species of fish.

One possible way to address this discrepancy would be to establish separate WQSs for anadromous species and resident species, using appropriate FCRs for each type of fish. Compliance with these separate WQSs could be measured based on the appropriate species for the water bodies of interest. For landlocked water bodies or water bodies that would not be expected to support substantial populations of anadromous species of edible size, the WQS based on resident species could be applied and samples of resident species collected to measure compliance. For water bodies where anadromous species are anticipated to be present, both resident and anadromous species could be collected and compared with their respective WQS to assess compliance.

It is also important to consider FCRs and target risks simultaneously. Federal guidance on WQSs provides for a range of target risk levels depending on the specific population or subpopulation considered. Specifically, federal guidance states that a WQS can be based on either a 1E-05 or 1E-06 risk for the general population, as long as identified subpopulations are not at a risk greater than 1E-04. At the February 10 Delegates’ Meeting, Ken Johnson of Weyerhaeuser proposed using probabilistic risk assessment to establish WQSs and then linking target cancer risks and noncancer hazards to population estimates as follows: 1E-5 risk at the 50th percentile of the FCR distribution for the entire state population, including subpopulations with higher consumption rates; 1E-4 risk at the 99th percentile; and hazard quotient of 1 for the 90th percentile. A similar probabilistic approach was used by the Florida Department of Environmental Protection in setting WQSs. Probabilistic risk assessment provides a more detailed, accurate picture of risk than does the traditional point estimate approach.

The probabilistic approach is consistent with EPA’s guidance, as long as the distribution of the FCR for the general population includes the range of FCRs that have been estimated for tribal and other subpopulations in the state. The Maine Department of Environmental Protection has used a variation on this approach by coupling an FCR of 138 g/day, based on a specific subpopulation of the state, with a target risk of 1E-4 when setting the WQS for arsenic, compared to 32.4 g/day and 1E-6 for other chemicals for the general population of the state.

This is similar to Ecology’s proposed approach for setting the WQS for PCBs, which recognizes the ubiquitous environmental concentrations of these chemicals that render it impractical to achieve risk-based concentrations at 1E-6 or even 1E-5 risk. Following the approach used by the Washington Department of Health, Ecology proposes to base the WQS for PCBs on a noncancer hazard of 1, which is associated with a cancer risk of 4E-5. While higher than the target risk used for other chemicals, this cancer risk is within EPA’s target risk range.

In developing a representative range of fish consumption rates for the general population to include various subpopulations, it is critical for regulatory actions to be based on peer-reviewed survey methods and data analyses that demonstrate the technical strength necessary to support decisions affecting health and economic resources. At the same time, it is recognized that the Pacific Northwest Tribes, which have collected multiple data sets on tribal consumption, are concerned about preserving the privacy of the tribal data. One solution to this potential conflict could be the introduction of a neutral third party with appropriate technical credentials, who could evaluate the technical strengths and comparability of all available data for Washington State, including the tribal data, to develop an appropriate distribution of fish consumption state-wide, without compromising that privacy.

Alternatively, high quality FCR data could be collected through a state-wide survey, similar to the effort that is currently being conducted by the Idaho Department of Environmental Quality, with assistance from EPA in collecting tribal data. Such a survey effort has been proposed for Washington State by Boeing. The survey, which could be designed to collect fish consumption data for all populations and subpopulations in the state using a consistent methodology, would allow Washington to develop comprehensive FCR distributions for total fish, resident fish, and anadromous fish from which percentiles could be selected, as appropriate, as the bases for deriving WQSs.

Ecology’s goal is to publish a draft of the WQS by the beginning of March, so stay tuned for further developments.