                                       
                                       



                             Office of Water
4304T
                                                               EPA-822-S-09-001
                                                                 September 2009

                                       
                                       
SUMMARY OF SCIENTIFIC VIEWS AND EPA RESPONSES
                                       
--------------------------------------------------------------------------------
Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health (2000), Technical Support Document Volume 3: Development of Site-Specific Bioaccumulation Factors
                                       
                                       
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
      Introduction
--------------------------------------------------------------------------------
      
      This document summarizes scientific views received in response to the Internet posting of the draft "Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health (2000), Technical Support Document Volume 3: Development of Site-Specific Bioaccumulation Factors". Scientific views have been organized and summarized in the attached table (Table 1) according to the input received by EPA from eight (8) commenters. One of the commenters (Federal Water Quality Coalition) submitted a White Paper from AMEC Earth & Environmental, Inc. The AMEC White Paper has been included in Table 1 as a separate set of comments. 
      
      Table 1 also includes EPA's response to each view received. These responses were drafted by Great Lakes Environmental Center (GLEC) and were subsequently revised following discussions with EPA's Bioaccumulation Work Group. 
      

   Views received from the following groups:
         o Thomas Barron, Standards Section, Div. of Water Quality Standards, Bureau of Water Standards and Facility Regulation, State of Pennsylvania (PA)
         o .Federal Water Quality  Coalition (FWQC) 
         o Central Costa Contra Sanitary District (CCCSD)
         o Sacramento Regional County Sanitation District (SRCSD)
         o Bay Area Clean Water Agencies (BACWA)
         o James J. Pletl, Ph.D., Hampton Roads Sanitation District (HRSD)

--------------------------------------------------------------------------------
Table 1. Scientific Views from Public
--------------------------------------------------------------------------------
View # 
and commenter
--------------------------------------------------------------------------------
Scientific View
--------------------------------------------------------------------------------
EPA Response
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

Commenter: Anonymous
--------------------------------------------------------------------------------
                  	
--------------------------------------------------------------------------------
                  
--------------------------------------------------------------------------------
1
--------------------------------------------------------------------------------
In the site specific baseline BAF example calculation the freely dissolved chemical concentrations of page 3-6 are greater than the total chemical concentrations given on page 3-4. Errors in the example calculations make it rather difficult when you are trying to determine if you are applying the methodology correctly.
--------------------------------------------------------------------------------
Example calculations have been reviewed and errors have been corrected. The error was of minor significance and does not affect other content or data found in the document. EPA regrets the error.
--------------------------------------------------------------------------------
 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Commenter: Thomas Barron
Standards Section, Div. of Water Quality Standards, Bureau of Water Standards & Facility Regulation , State of Pennsylvania


--------------------------------------------------------------------------------
1
We do not have comments regarding the scientific views or specific methods described in the Draft Volume 3, we do have several comments or concerns about implementation of these methods. Volume 3 provides multiple ways of calculating a site specific Bioaccumulation Factor (BAF). In all cases, a large amount of data must be generated for the calculations. This would require a major study in the basin of concern, to be conducted either by the regulatory agency, a discharger, consultant, or some other interested party.
--------------------------------------------------------------------------------
No response.
--------------------------------------------------------------------------------
2
A site specific BAF will require chemical specific monitoring, trophic level determinations, fish tissue, % lipid and water quality analyses. This would be followed by data analysis and the BAF determination. The Department currently does not have the resources to perform site specific BAF studies throughout the Commonwealth. Similarly, we are concerned that if we receive multiple requests, and associated reports for BAF determinations, that this too would require considerable staff resources to review and determine if the studies and conclusions are scientifically sound and appropriately followed the guidance. As a result, and since this is based on EPA's guidance, we will likely need to turn to EPA Region 3 and/or Headquarters for their oversight and guidance on reviewing any requests we receive in Pennsylvania.
--------------------------------------------------------------------------------
Development of a site-specific BAF does not always require a field study; in fact Section 5.3 explains how existing data for similar sites could be used/adjusted/extrapolated to other site(s) (see Method 5).
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Development of this guidance is not a requirement to implement site-specific BAFs in criteria calculations. Should a State, Territory or Tribe wish to implement BAFs, U.S. EPA is available to provide technical assistance and guidance on a case-by-case basis.  
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Commenter: Federal Water Quality Coalition
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
1
The approach used by EPA in the Draft Guidance is, in general, technically sound for derivation of site-specific BAFs for non-polar organic chemicals. To assess the specifics of the Agency's approach, we asked AMEC Earth & Environmental, Inc. ("AMEC") to perform a technical review of the Draft Guidance. The results of that evaluation are in the AMEC Report that is attached to these comments. Based on the Coalition's review and the AMEC report, we believe that there are a number of aspects of the Draft Guidance that merit revision or clarification. These aspects are as follows:

   * The Guidance should recognize that the BAF methodology does not apply to most metals.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
BAF derivation procedure for inorganic and organometallic chemicals is addressed in Section 2.1 and reference is made to Section 5.5.6 of the 2000 Human Health Methodology. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Text was added to Section 1.2 (Scope of Document) clarifying that this TSD was developed to apply to nonionic organic chemicals, and other chemicals with similar partitioning and bioaccumulation behavior. This will include the introduction sections from Section 5.5 and 5.6 of the 2000 Human Health Methodology document, to further clarify the scope of this TSD.

--------------------------------------------------------------------------------
Text was also added to Section 2 to clarify the focus of the document, stating that the BAF methodology in this TSD does not apply to most metals.
--------------------------------------------------------------------------------
2
   * The Guidance should clearly state that the BAF methodology does not apply to pharmaceuticals and personal care products.
--------------------------------------------------------------------------------
BAF derivation for ionic organic chemicals is addressed in Section 2.2 and reference is made to Section 5.5 of the 2000 Human Health Methodology. 
--------------------------------------------------------------------------------
Section 3 (p. 3-8 and 3-9) also provides examples of "other" classes of organic chemicals, for which this TSD may not apply, including surfactants, dyes and pigments. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The comment is far too broad, as there are many ingredients in pharmaceuticals and personal care products (PPCPs) that may bioaccumulate. Therefore, TSD Volume 3 applies to these chemicals. Bioaccumulation of a pollutant should be evaluated on a case-by-case basis. 
--------------------------------------------------------------------------------
3
   * The Guidance should explain that the BAF methodology does not apply to ephemeral and effluent-dominated waters.
--------------------------------------------------------------------------------
A site may include intermittent and ephemeral waters that are by nature highly variable on seasonal, annual, and inter-annual time-scales.  The nature of this variability will in large part, determine the types of fish that may be encountered in these waterbody types.  The appropriateness of a BAF developed for fish that may inhabit these waters is a very site-specific decision, and some aspects of this guidance may not be applicable to such waters.  As such, EPA recommends that a state or tribe consult with EPA prior to beginning development of a site-specific BAF for these types of waters.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
This text has been added to the document on p. 2-7.
--------------------------------------------------------------------------------
4
   * The Guidance should address the potential cost implications for agencies, regulated parties and other stakeholders to implement the field programs specified in the Draft Guidance.
--------------------------------------------------------------------------------
Factors that can affect costs of field studies are addressed in Sections 3 (fish and water sampling and analysis), and 4 (sediment sampling and analysis), although cost estimates are not presented. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
This document is strictly intended as technical guidance, should a management decision be made that development of a site-specific BAFs is warranted. Cost implications and management goals are factor-specific to a given project, and thus are considered to be best addressed by the regulatory authority and permit holder. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Language was added to section 2.5.7 that states: "Resource limitations, institutional context, and the use to which the BAFs will be put may also be important selection factors in choosing a BAF method."
--------------------------------------------------------------------------------
5
   * The Guidance should recognize that for smaller-scale projects, the approaches specified in the Draft Guidance may not be necessary or relevant.
--------------------------------------------------------------------------------
The guidance in TSD Volume 3 is intended to apply to all projects, large or small. In part, the different BAF methodologies address how institutional needs and resources can vary.
--------------------------------------------------------------------------------
6
   * The Guidance should explain more fully how the site-specific BAF approaches in the Draft Guidance can or should be utilized in the process of developing a Total Maximum Daily Load ("TMDL").
--------------------------------------------------------------------------------
See response to AMEC comment 7, below (page 13).
--------------------------------------------------------------------------------
7
   * The Guidance should explain if and how site-specific BAFs could be used in deriving remediation goals for water bodies under RCRA and CERCLA.
--------------------------------------------------------------------------------
See response to AMEC comment 9, below (page 14).
--------------------------------------------------------------------------------
8
   * The Guidance should discuss how site-specific exposure information  -  such as fish consumption rates that are lower than national rates  -  should be used in deriving site-specific BAFs
The use of site-specific fish consumption rates is addressed in Section 4.3.3.1 of EPA's Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health (2000) [hereafter known as 2000 Human Health Methodology; EPA-822-B-00-004], see page 4-25). 

Reference to appropriate guidance in EPA's 2000 Human Health Methodology was added. 

Further, Section 3.3.5 states that composite sample types selected for chemical analysis should be based on the tissue types consumed by the target population of concern. 
--------------------------------------------------------------------------------
9
   * The Guidance should require that the site-specific BAF process recognize the possible impact of different fish preparation methods, which can lower the levels of some substances that are consumed from fish tissue.
--------------------------------------------------------------------------------
Section 3.3.5 states that analysis of whole fish should mirror the way the consumer prepares the fish. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Reference to appropriate guidance in the 2000 Human Health Methodology was added. Section 4.3.3.1 of the 2000 Human Health Methodology (see page 4-28) states that uncooked weights should be used in the derivation of a AWQC value.
--------------------------------------------------------------------------------
10
   * The Guidance should point out the need to derive statistical uncertainty estimates when deriving field BAFs or biota-sediment accumulation factors ("BSAFs").
--------------------------------------------------------------------------------
The TSD gives considerable attention to the uncertainty of BAFs and BSAFs, with respect to the design of site-specific sampling. The statistical uncertainty of field-derived BAFs is addressed in Sections 3.2.1 (p.3-19 thru 3-28) and Appendices 3A and 3C. For BSAFs, measurement uncertainty is addressed in Sections 4.4 (p. 4-27 thru 4-35) and Appendices 4A and 4B. Uncertainty in Method 3a is addressed in Section 5.1.1 (p.5-18), Method 3b in Section 5.1.2 (p.5-29), Method 4a in Section 5.2.1.1 (p.5-42), and Method 4b in Section 5.2.3.1 (p.5-68). 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The TSD is also flexible, and indicates that study design should take into account the specific management goals of a given project, including issues regarding uncertainty. Thus, if a decision is made that uncertainty should be addressed quantitatively, then the design should reflect that decision. The scope and type of uncertainty analysis required needs to be determined based on the specific project/water/chemical(s) being assessed. These issues should be addressed in a problem formulation/ project work planning stage.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Text was added that encourages investigators to report the uncertainty associated with field measurements.
--------------------------------------------------------------------------------
11
   * The Guidance should address variability issues more fully, including temporal variability. There is generally some lag time between the reduction of water column concentrations and the reduction in fish or sediment concentration of bioaccumulative substances. This means that temporal variability is an important issue for EPA to address in the Guidance.
--------------------------------------------------------------------------------
EPA disagrees with this comment.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Section 3.2 states that a successful sampling design procedure should consider temporal gradients in chemical concentrations.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The modeling simulation of BAF sampling designs (Section 3.2.2) addresses how temporal variability of water and biota concentrations relate to optimal field sampling designs. "Lag time" in Π addressed in Section 4.3.3.
--------------------------------------------------------------------------------
12
   EPA should recognize that there are fundamental problems in applying the BAF methodology to mercury, and should instead focus on other regulatory approaches for mercury, which are detailed in EPA's draft guidance for implementing its recommended mercury fish tissue criteria.
Section 2.1 states that guidance on determining national BAFs for inorganic and organometallic
--------------------------------------------------------------------------------
chemicals is provided in Section 5.6 of the 2000 Human Health Methodology. The statement in Section 3 (p. 3-9) will be modified to state that TSD Volume 3 does not apply to methyl mercury. 

Commenter: AMEC
(White Paper, Evaluation of "Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health (2000), Draft Technical Support Document, Volume 3 (2008): Development of Site-Specific Bioaccumulation Factors)
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
1
ISSUES RELATED TO THE IMPLEMENTATION OF THIS METHODOLOGY

Overall, the BAF Guidance (USEPA, 2008a) is technically sound and supports the development of AWQCs for non-polar organic chemicals. However, there are several issues related to the implementation of the methodology and those issues are discussed below.

Issue #1: Evaluation of Metals, Metalloids, or Organometallics

With the exception of the organometallics (e.g., methyl mercury), many of the approaches presented in USEPA (2008a) are not directly applicable to metals or metalloids. This is because these chemicals are often polar under environmental conditions (and therefore are not readily bioaccumulated), can be (in some cases) physiologically regulated (Bury et al., 2003), or have unique bioaccumulation pathways compared to the non-polar organics. 

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------





Section 2.1 states that guidance on determining national BAFs for inorganic and organometallic
--------------------------------------------------------------------------------
chemicals is provided in Section 5.6 of the 2000 Human Health Methodology. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
2
Metal Bioaccumulation

In order to apply the methodologies presented in the BAF Guidance to metals, a sound understanding the bioaccumulation potential of metals must be known. Recently, DeForest et al. (2007) reviewed several studies that measured the uptake of eight metals or metalloids (arsenic, cadmium, chromium, copper, lead, mercury, selenium, and zinc) by aquatic organisms. An inverse relationship was observed between the BAFs and the chemical concentration in the water. The authors concluded that no one BAF or tropic transfer factor (TTF) can be used to express bioaccumulation and/or trophic transfer without consideration of the exposure concentration of the metals. The authors recommend to account for the inverse relationship in environmental assessments through the use of bioaccumulation regression models in lieu of discrete values. The potential physiological basis for this was not addressed by the authors. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

Section 2.1 states that guidance on determining national BAFs for inorganic and organometallic
--------------------------------------------------------------------------------
chemicals is provided in Section 5.6 of the 2000 Human Health Methodology. 
--------------------------------------------------------------------------------
Hence, the issue of metal bioaccumulation is not addressed in TSD Volume 3.
--------------------------------------------------------------------------------
3
Metal Valence State 

Chemical speciation (i.e., varying valence states) of the same metal is also important for the assessment of bioaccumulation potential and potential human health effects of select metals (e.g., Fairbrother et al., 2007).  Speciation requires specialized sample handling, storage and analytical methods (Leermarkers et al., 2006; Kim et al., 2007), which can be cost prohibitive. 

USEPA (2003b) summarized the key issues related to arsenic bioaccumulation by freshwater and marine aquatic organisms. This effort included an evaluation of valence state and the organic form of the arsenic that was reported in tissues from field and laboratory studies. The general conclusion of this evaluation was that there was insufficient field and laboratory data to clearly define relationships between arsenic levels in the water and tissue levels. Inorganic arsenic (as arsenite and arsenate, both with +3 valence states) was included as part of the analytical suite for the National Lake Fish Tissue (NLFT) Study to attempt to expand the field database for arsenic. The NLFT Study was a four year (2000-2003) nation-wide survey of lake fish tissue quality (http://www.epa.gov/waterscience/fish/study/) that focused on chemicals with bioaccumulative potential. Arsenite and arsenate were detected at a low frequency across all four years of samples.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

Section 2.1 states that guidance on determining national BAFs for inorganic and organometallic
--------------------------------------------------------------------------------
chemicals is provided in Section 5.6 of the 2000 Human Health Methodology.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Text will be added to Section 2 to clarify the focus of the document, stating that the BAF methodology in this TSD does not apply to most metals.
--------------------------------------------------------------------------------
4
Issue #2: Contaminants of Emerging Concern

One of the areas of interest since early 2000 is the potential environmental and human health impacts of "contaminants of emerging concern"  -  such as pharmaceuticals and personal care products (Daughton and Ternes, 1999; Kolpin et al., 2002). The knowledgebase is still under development for this diverse group of chemicals that have wide-ranging physiological effects and analytical requirements. A framework for evaluating the potential ecological impacts of these chemicals is under development by EPA (USEPA, 2008b).  

The development of acceptable analytical methods and toxicity testing approaches for these contaminants of emerging concern has been the primary focus to date. Many of these chemicals do not readily bioaccumulate and can be metabolized by aquatic organisms. Consequently, they should not be considered under the BAF Guidance (USEPA, 2008a).  Nonetheless, USEPA (2008a) recommends assessing the bioaccumulation potential of these chemicals if they are of interest for a given project.  

To screen for the bioaccumulation potential of a chemical of interest, USEPA's Office of Pollution Prevention and Toxics has developed the on-line Persistent, Bioaccumulative, and Toxics Estimation Tool (PBT Profiler; http://www.pbtprofiler.net/]. The PBT Profiler is a screening tool to estimate the bioaccumulation potential and toxicity for chemicals that lack experimental data. As a screening tool, it chiefly identifies whether the chemical may need further evaluation as a potential PBT chemical. Searches are performed using the chemical's name and Chemical Abstract Service (CAS) Registry number. 


This tool, or other sources (e.g., Weisbrod et al., 2007) should be considered to determine the need to apply the BAF Guidance to a project site.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Contaminants of emerging concern may include a wide range of chemicals, including those that bioaccumulate. Bioaccumulation potential should be evaluated on a case-by-case basis.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Section 3 of the TSD (p. 3-8 and 3-9) provides examples of "other" classes of organic chemicals, for which this TSD may not apply, including surfactants, dyes and pigments.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
As noted by the commenter, the PBT Profiler is a screening tool to estimate the bioaccumulation potential and toxicity for chemicals that lack experimental data. As such, it would be highly inappropriate to use this tool to calculate a site-specific BAF.
--------------------------------------------------------------------------------

EPA agrees that all measurements of bioaccumulation should be considered in the development of BAFs for a Human Health AWQC. The Weisbrod et al. 2007 reports reviews databases that contain BCF, BAF, and BSAF. These data should be evaluated for use in the methods for deriving the BAF. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
5
Issue #3: Cost Implications

An aspect that was not addressed by any of these technical guidelines was the potential cost implications for states, other regulatory parties, or regulated parties, to implement the recommended field programs to comply with the BAF Guidance. For larger scale projects, a comprehensive monitoring program to develop the chemical database can easily exceed several million dollars. Furthermore, the collection of site- or regional-specific angler survey (creel) data can also be cost-prohibitive and may not be technically feasible. In addition, the relevance of such data under fish consumption advisories can be complex.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Cost of field studies is addressed in Sections 3 (fish and water sampling and analysis) and 4 (sediment sampling and analysis), although cost estimates are not presented. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
TSD Volume 3 guidance is for site-specific BAFs, so cost varies by project (i.e. management goals, site, chemicals, etc.) It is impossible to give more specific estimates. This is an implementation and management issue, not a technical or scientific issue. See response to Federal Water Quality Coalition comment 4 above (page 5).
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
6
Issue #4: Relevance to Different Sized Projects

Section 2.4 of the BAF Guidance provides a broad definition for a site, and states that the approaches are to nearly any project. However, clearly the relevance of deriving an AWQC for a small project should be considered when assessing the need to comply with this guidance. The preferred approaches (Method 1 described in Section 2.1 of this white paper) require an understanding of the local fishery (e.g., fish migratory patterns and life histories), large numbers of field samples of multiple media, and depending upon the chemical(s) of interest can also require complex sample preparation and analytical methods.  Such approaches would be relevant only for larger-scale projects. 

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The commenter has raised implementation and resource availability issues that a project manager would address outside of this technical guidance. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The guidance in TSD Volume 3 is intended to apply to all projects, large or small. In part, the different BAF methodologies address how institutional needs and resources can vary.
--------------------------------------------------------------------------------
7
Issue #5: Development of TMDLs and Discharge Permit Limits

Development of TMDLs (and associated discharge permit limits) was identified in the BAF Guidance as a potential application of the BAF methodology. Although TMDL development is a logical application, specifics regarding this application were not presented.  

A relevant case study where the BAF methodology was applied to develop a TMDL for PCBs in the Delaware River (Fikslin et al., 2007). Prior to the release of this guidance, the Delaware River Basin Commission (DRBC) developed three TMDLs for the Delaware River. The values were 44.4 pg/L and 44.8 pg/L where fish consumption was minimal (the slight variation was due to different levels of use as a drinking water source) and 7.9 pg/L where fish consumption occurred at a higher rate. Starting in the late 1990s through the early 2000s, DRBC undertook a large scale monitoring program of fish, surface water and sediments in the estuary (Greene, 2002) and also performed a creel survey to derive fish consumption rates (Versar, 2003). A revised TMDL of 15.9 pg/L was calculated based on survey/monitoring information comparable to the methodology outlined in BAF Guidance. This revised TMDL was lower than two of the TMDLs originally developed for the river. Fikslin and co-workers (2007) did not further assess this discrepancy.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Application of BAFs in TMDL development are beyond the scope of this TSD. The example below demonstrates that TMDLs developed using methodology in this document provides values in line with previous determinations in the Delaware River.
--------------------------------------------------------------------------------
8
Issue #6: Relevance to Ephemeral and Effluent-dependent Waters

Ephemeral and effluent-dominated waters represent unique cases for the use of the BAF Guidance. Since the focus of this guidance is on the potential for fish bioaccumulation, the guidance is not particularly relevant to ephemeral water bodies, which generally do not support populations of resident fish.  In fact, such water bodies are not even mentioned in this guidance. 

This is an issue because examples exist where AWQCs have been developed for ephemeral streams without the use of the BAF Guidance. The Arizona ambient water quality criteria for ephemeral water bodies is such an example (Arizona Administrative Code, 2003). These values were designed to protect use of the "ephemeral water by animals, plants, or other organisms, excluding fish, for habitation, growth, or propagation."  This approach avoids the need to develop water quality criteria based on the BAF Guidance since it is derived based upon the potential ecological impacts to animals that may use these water bodies. The criteria are toxicity-based and were derived for organisms that spend portions of their life cycles in water (e.g., amphibians). 

Effluent-dominated waters can also be problematic within the context of the BAF guidance. Development of appropriate water quality criteria for these water bodies is an area of on-going research (e.g., Brooks et al., 2006) and is complicated by the wide variety of different hydrologic conditions of these water bodies.  

Some effluent-dominated water bodies may be exempt for evaluation if they are not included as "waters of the state."  The definition of the latter is state-specific, since it generally takes into account issues such as local climate (e.g., arid southwest compared to temperate northeast to determine how transient the water body may be) or its use (e.g., settling lagoon versus a surface water impoundment). These, and other related issues, should be considered before utilizing the BAF Guidance to assess an effluent-dominated waterway.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Decisions regarding water body use/management goals certainly play into how a WQS would be developed, and if bioaccumulation should be a factor, but that decision is made outside the scope of this technical guidance. This guidance is used once it has been determined that BAFs are necessary. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
See response to Federal Water Quality Coalition comment 3 above (pages 4-5).
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
9

Issue #7: Development of Remediation Goals for Water bodies with Contaminated Sediments

Since the BAF Guidance is related to deriving AWQCs, it is not directly applicable to the development of remediation goals for water bodies under RCRA or CECRLA programs (although it may be identified as part of the compilation of the Applicable, Relevant and Appropriate Requirements under CERCLA). However, regulators may reference the BAF Guidance as a framework for developing BAFs (or BSAFs) at contaminated sites. Since BAFs and BSAFs can be used to develop media clean-up goals, the regulator may require a responsible party into performing more extensive monitoring in order to satisfy the requirements of the BAF Guidance, unless agreement is made early in the process that the less stringent BAF derivation methods (e.g., Methods 3 through 5 discussed in Section 2.5) would be acceptable.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The issues raised by the commenter are all implementation, management, and resource issues that are outside the scope of this guidance. This guidance is used after it is has been determined that a BAF is needed for a given project.
--------------------------------------------------------------------------------
10
Issue #8: Exposure Assumptions

Generally, the BAF Guidance allows for the use of site-specific exposure information  -  principally related to fish consumption rates  -  to derive the site-specific BAFs. What is unclear is whether this would be accepted by regulatory authorities if fish consumption rates are low at a given project site.

In addition, the impact of different fish preparation methods on chemical residues is not included in the BAF Guidance. Disregarding this can result in conservative (lower) estimates of the back-calculated AWQCs.

--------------------------------------------------------------------------------
EPA has addressed these comments above. See Comments 8 and 9 from the Federal Water Quality Coalition (pages 5- 6).
--------------------------------------------------------------------------------

Commenter: Central Costa Contra Sanitary District
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
1
Interestingly, our review of the BAF data (see Attachment Nos. 1 and 2) shows that application of Method 2 to predict BAFs for PCDDs and PCDFs in the Great Lakes ecosystem was not consistent with the fundamental criterion of similar hydrophobicity for the selection and use of the reference chemical. For example, the Kow for PCB52 is 691,830 (Log Kow 5.84). This reference chemical is not appropriate for predicting BAF for any of the PCDD and PCDF congeners because the Kow is well outside the criterion Kow of +-0.5 of the Kow (of the chemical of interest.) The Kows of the chemical of interest are in the range of 3x10[6] to 630x10[6] (see Attachment 3). Similarly, DDT, PCBlO5, and PCBll8 are also not appropriate reference chemicals for predicting BAFs for all except for one or two PCDD and PCDF congeners for the same reason, as indicated above. As such, the Great Lakes ecosystem BAFs for PCDDs and PCDFs are not appropriate based on the established criteria, and if the same recommended methodology is used for national BAFs or local and regional site-specific BAFs, those BAFs would similarly be faulty.

--------------------------------------------------------------------------------
Upon review, EPA recognizes that the +-0.5 logKow reference chemical guidance (text box, p. 4-14) is overly restrictive. For example, the National BAF TSD (Volume 2) states that the reference chemical should be similar to the chemical of interest. The 2000 Human Health Methodology states that the chemical of interest and the reference chemical should have similar fugacities, i.e., ∏soc /Kow . Accordingly, EPA has changed the guidance in this TSD for consistency with the guidance in the National BAF TSD and the 2000 Human Health Methodology. This comment is no longer relevant, because of this change. 
--------------------------------------------------------------------------------
2
Further, in the Great Lakes ecosystem the USEPA also appears to use 2,3,7,8-TCDD as a reference chemical for predicting site-specific BAFs for 16 congeners (chemicals of interest) of PCDD and PCDF. The USEPA's accepted Kow for 2,3,7,8-TCDD is 10x10[6] (Log Kow 7.02 see Table 5-2 of the TSD, Volume 3, Attachment 3). Based upon the criterion (Kow) for selection of the reference chemical, the use of 2,3,7,8-TCDD is only applicable to one PCDD congener. It is not apparent that 2,3,7,8-TCDD is an appropriate reference chemical for any of the PCDFs because the USEPA does not recommend the use of the given Kows for PCDFs (see the footnote to Table 5-2, Attachment 3). Further, note that TSD Method 3(a) and 3(b) are also based upon Method 2.

Therefore, we suggest that the USEPA either:
(a) Select and recommend more appropriate reference chemicals other than those used in the Great Lakes ecosystem, which fully meet the criteria for the selection of a reference chemical laid out in the TSD on pages 4-13 through 4-15, or in case no appropriate reference chemicals are found; [or]
(b) Delete the reference of Method 2 and perhaps Method 3(a) and 3(b) as recommended methods for predicting site-specific BAFs.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA has clarified that Table 5-2 is an example offered to illustrate the methodology, and cite guidance on selection of appropriate Kow values and reference chemicals. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Suggestion (b) is inappropriate because for most chemicals one or more reference chemicals can be found.
--------------------------------------------------------------------------------
3
Note that the TSD (Table 5-2 footnote) does not recommend the use of the given Log Kow for predicting BAFs for PCDFs, and surprisingly, the USEPA does not recommend an alternative log Kow for these chemicals of interest. Therefore, the site-specific BSAFs and TCDD BEFs shown in Table 5-2 and in Table 9 of the Great Lakes ecosystem for PCDFs have no meaning in BAF methodology without corresponding Kows. The Kow of the chemical of interest is the key determinant in the BAF value as shown in the equation on page 1 of this letter.

Therefore, we suggest that the USEPA either:
(a) Identify appropriate alternate Log Kow for PCDFs, or
(b) Delete the reference of Log Kow for PCDFs in Table 5-2.
Otherwise, states and tribes may use these non-recommended Log Kow for derivation of BAFs for PCDFs as has already been done by the USEPA for the Great Lakes ecosystem. This inappropriate use of the Log Kows in the derivation of BAFs may consequently result in faulty AWQCs.

--------------------------------------------------------------------------------
EPA has changed the footnote on Table 5-2 to "Use of the Kow values in this table does not imply approval by EPA", or an equivalent statement.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
4
Additionally, it appears that the Log Kows for PCDDs are mainly based upon USEPA's own data (Table 5.2). Our review of the literature shows that the Log Kows for the same congeners are substantially different from the USEPA's selected Kows. Since the Log Kows are the key factors in determining the BAFs, we suggest that the USEPA resolve the differences between Table 5.2 Log Kows and that published in the literature before adopting and using these Log Kow values in BAF methodology.

--------------------------------------------------------------------------------
We cannot judge the merit of this comment without information regarding the specific Kow values identified in the commenter's literature search. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA has reviewed Section 5.2.3.2 of the TSD to ensure that the recommendation is made to search the literature for chemical-specific Kow data and select the most appropriate value of Kow based on specific guidance.
--------------------------------------------------------------------------------

Commenter: Sacramento Regional County Sanitation District
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
1
The most important comment offered by SRCSD addresses the concept of the use of BAFs in the EPA regulatory program. Through its experience with a mercury TMDL, SRCSD has determined that the use of BAFs to create water column objectives for a bioaccumulative pollutant is (a) unnecessary for NPDES permitting and TMDL implementation, (b) confusing, in that it places emphasis on control of water column concentrations rather than on mass load management in the attainment of fish tissue criteria, and (c) a hindrance to the development of either offset or water recycling solutions to bioaccumulative pollutant management by NPDES-permitted point sources. SRCSD points out that EPA's own "Draft Guidance for Implementing the January 2001 Methyl mercury Water Quality Criterion" recognizes that permits do not need concentration limits. At a minimum, the TSD should contain clear statements that the use of BAFs is discretionary and a decision that should be made with consideration of implementation consequences. Additionally, SRCSD requests EPA to address this point at a fundamental level in its Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health.

--------------------------------------------------------------------------------
This comment falls outside the scope of the TSD.
--------------------------------------------------------------------------------

A statement was added (p. 1-4) that State and Tribal decision makers also retain the discretion to use EPA's national BAFs, or scientifically defensible methodologies, including those discussed in this guidance and others, to develop site-specific BAFs, as appropriate."
--------------------------------------------------------------------------------
A reference was made to EPA's Implementation Guidance for mercury.
--------------------------------------------------------------------------------


--------------------------------------------------------------------------------
2
Attachment A - General Comments by SRCSD on the Use of BAFs 
       The driving force behind BAFs at the national level is the argument that they are needed to allow ease of NPDES permit writing. Yet EPA's own "Draft Guidance for Implementing the January 2001 Methyl mercury Water Quality Criterion" (http://www.epa.gov/waterscience/criteria/methylmercury/guidance-fs.html) recognizes that permits do not need concentration limits. Alternative regulatory tools that could be applied to any bioaccumulative pollutant include: 
             Use Attainability Analyses for cases where naturally-occurring or other conditions prevent attainment of fish consumption use 
          Establish a mass load limit either at existing effluent levels or consistent with a TMDL 
             Permit a discharge when it can be demonstrated that other pollutant source reduction efforts will offset the discharge's contribution 
             Ensure that the pretreatment program protects a publicly-owned treatment work's effluent from contributing to excursions of the fish tissue criterion 
             Express NPDES permitted water quality-based effluent limits (WQBELs) in narrative form, requiring Pollutant Minimization Programs 
             Require regular monitoring using low-level detection limit methods to trigger enhanced efforts 
       TMDLs for impaired water bodies can provide the flexibility to derive appropriate site-specific goals and associated load reduction requirements without the burden of additional regulatory constraints such as concentration limits. Evidence of the success of this flexibility is the fact that implementation of the San Francisco Bay mercury TMDL has not been impeded by the lack of a water column objective. 
       Water column concentration limits will not provide a better means of beneficial use protection than the fish tissue objectives. A focus on concentrations in water complicates the issue and does not offer any gains in management. Fish tissue criteria keep the management actions focused on broad-based watershed solutions while concentration limits focus only on point discharges. The TSD should point out that there are several disadvantages to employing concentration criteria in addition to fish tissue criteria, including: 
                Focusing on concentration criteria presents an obstacle to trading. The likely parties to participate will be dischargers operating under NPDES permits. If such dischargers are required to install additional treatment to meet concentration limits, the need for trading will be diminished or eliminated and the financial resources for trading will be similarly reduced. 
               Concentrations of many bioaccumulative compounds in water are highly variable in space and time. Testing for a consistent local relationship would require specialized sampling, long-term monitoring, and dedicated funding. A site-specific dataset would, in the end, result in relationships based on a pair of mean concentrations (fish versus water column) that may still not predict how changes in discharge concentrations would change fish tissue concentrations. 
             A variance procedure would be needed if water column methyl mercury objectives are promulgated and used to set NPDES effluent limits for wastewater treatment facilities unable to meet concentration limits through source control or offsets. The feasibility of potential changes to treatment processes to meet mercury as well as other water quality limits must be evaluated. 
       Fish tissue levels represent the cumulative effects of exposure and consumption over their lifetimes. Yet it is common to calculate BAFs based on concurrent measurements. The temporal variability in measured tissue-to-water ratios makes BAFs a moving target, which may be largely ignored by those using this guidance. 

    SRCSD does not believe that site-specific BAFs should be used to translate tissue residue values to water column values as part of a standard setting process. This volume of the TSD should clarify that this is not a recommended approach. SRCSD believes that even the site-specific "field measured" BAF approach may be inadequate for bioaccumulative pollutants such as methyl mercury, because of the largely untested and unproven nature of the hypothesis that the accumulation of methyl mercury by aquatic organisms varies in linear proportion (see definition of BAF on page 1-7) to the methyl mercury concentration in the water column. In mercury bioaccumulation studies typically cited (e.g., Brumbaugh et al. 2001 "A National Pilot Study of Mercury Contamination of Aquatic Ecosystems Along Multiple Gradients: Bioaccumulation in Fish." Biological Science Report, USGS/BRD/BSR-2001-0009), results are based on a pooling of data from different watersheds and water bodies to assert a water-to-tissue relationship. Yet, such local, site-specific relationships are the key to effective mercury management. If water column levels are to be used as variables to control fish tissue concentrations, regional relationships are inappropriate to regulate water column levels on a site-specific basis. 


    SRCSD supports the TSD's recommendation to collect a sufficiently representative number of local fish tissue data to better understand levels in various species and trophic levels for compliance determinations (page 3-20). Use of fish tissue criteria is a non-traditional approach in the development of WQBELs in NPDES permitting and TMDLs. Fish tissue criteria would be the best surrogate for beneficial use protection because they better reflect the risk to human (and wildlife) consumers of fish, are easier to measure, and avoid the need for calculating and using BAFs. 
    SRCSD also supports the TSD's recommendation to collect low detection-limit data to better understand water column concentrations in effluents and ambient waters as the foundation for watershed-based approaches to mercury management (page 3-56). This TSD should caution against biased statistical analysis (i.e. the selection of data that supports a relationship and the exclusion of data that does not). But for regulation, the focus should be on fish, not on paired fish and water data. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
These comments involve two different issues: AWQC and TMDLs. TMDLs are developed after it has been demonstrated that a water body is impaired, which is usually the result of the water body not meeting an AWQC. Hence, a BAF is still required to establish the AWQC (i.e., a TMDL development program does not stop EPA from developing AWQC). 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The TSD describes how to select a BAF. All of the comments under the bullet "Water column concentration limits will not provide a better means of beneficial use protection than the fish tissue objectives " are outside of the scope of the document.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
This comment ("Fish tissue levels represent the cumulative effects ...") is addressed in Section 3.2.2; see FWQC comment 11 and response, above (pages 7- 8).
--------------------------------------------------------------------------------

This comment ("SRCSD does not believe ...") relates primarily to pollutant-specific issues (e.g. methyl mercury and selenium), and would be addressed in any contaminant-specific AWQC and accompanying implementation guidance. EPA recognizes this fact, and added references to methyl mercury and selenium as examples where EPA has addressed such issues in other guidance, as appropriate.

--------------------------------------------------------------------------------
Section 3 (p. 3-9) will be modified to state that this TSD does not apply to methyl mercury.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Section 3.2 (p.3-17 thru 3-28) offers guidance on sampling designs to achieve low bias in field-derived BAFs; this is also addressed in detail in Appendix 3A.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

Commenter: Bay Area Clean Water Agencies
--------------------------------------------------------------------------------

1
Although the BAF methodology appears to be scientifically superior to the BCF methodology, it also has limitations and raises concerns about the usefulness of its application. The use of the BAF methodology to develop AWQC appears to result in AWQC values that are substantially lower than the current criteria developed using the BCF approach. The current criteria for constituents like dioxin and PCB is below accurate analytical detection, and measuring attainment is not accurately quantifiable. Lowering the AWQC further only exacerbates the detection issue. Developing criteria further below analytical detection limits does not appear to have any practical value.

--------------------------------------------------------------------------------
EPA agrees that the use of the BAF methodology to develop AWQC may result in AWQC values that are substantially lower than the current criteria developed using the BCF approach, because usually BAF > BCF. However, this approach is consistent with the current state of the science. Analytical methods continue to improve, such that previously-undetectable concentrations have become detectable. If there is specific concern regarding analytical detection issues, the concern should be addressed prior to developing any BAF.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Developing criteria below analytical detection limits has the practical value of informing the public regarding scientific understanding of public health safety. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
.
2
BACWA has the following concerns with the proposed methodology:
1. the application of BAFs for ambient water quality criterion (AWQC) development,
2. the high degree of variability of bioaccumulation factors (BAFs) derived from the different proposed methods, and
3. the determination of effluent compliance with AWQC utilizing BAFs.

The data in the TSD shows that BAFs vary +- 5-times depending upon the method used for their derivation (see Figures 4-2, 5-6, 5-7, 5-8, 5-11, and Table 5C-2). In fact, the TSD states (page 2-19), "Large differences in individual site-specific BAFs for a given species or trophic level (e.g. greater than a factor of 10) should be investigated ... " This statement implies that a BAF variability up to a factor of 10 is acceptable. The TSD further states (page 2-20), "In some cases the uncertainty associated with very limited BAF data from a 'more preferred' method (field-measured BAFs) may be offset by the greater quality and diversity of data that are available from otherwise 'less preferred' method (extrapolated or predicted BAFs)."

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA disagrees with BACWA's interpretation of this text. The "factor of 10" is presented as an example (hence the "e.g."). EPA does not state that BAF variability up to a factor of 10 is acceptable, nor does it agree with this implication. BAF variability is an issue that would be addressed during development of an AWQC, and any uncertainty "acceptability" threshold is a case-specific management decision. 
3
The fact is, the "more preferred" methods are generally more labor intensive and costly, and thus less data is usually available compared to "less preferred" methods. Therefore, BACWA is concerned that the USEPA, States, and Tribes most likely will tend to use "less preferred" methods to derive BAFs. Consequently, the BAFs derived using these methods may be up to 10 times greater than actual BAFs (field-measured).

--------------------------------------------------------------------------------
The purpose of developing the guidance in this TSD was to `show the way' to derive site-specific BAFs as an alternative to national BAFs. The latter are "less preferred", as put forth in the 2000 Human Health Methodology. A site-specific BAF is superior. If an interested party does not prefer using a national BAF, or one of the other site-specific BAF methodologies, then their option is to derive a field-measured BAF. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
If a specific project does not have the resources to develop a site-specific BAF, then the project managers should assess that issue and factor it into their decisions about the amount of uncertainty they are willing to accept.
4
BACWA recognizes that the high variability might be inherent in different methods. However, BACWA believes that this substantial variability, in particular the potential positive bias (up to 10-times) in BAFs depending upon the methods, may result in substantially more stringent AWQC, which in turn may result in over-regulation and ineffective use of resources without much water quality improvements. Therefore, BACWA recommends that the USEPA considers one of the following:

a. Delete or de-emphasize the reference of all proposed BAF derivation methods other than the site-specific field-measured BAF method; or
   

b. If the use of other BAF methods is deemed acceptable, other than the site specific field-measured method, a specific uncertainty factor should be applied (up to a factor of 10) in either the AWQC development methodology or the determination of effluent compliance.

For example, under this scenario, a given discharger's effluent will be deemed in compliance with effluent or AWQC limitations unless the effluent or ambient water concentration exceeds the AWQC by a factor of 10. The incorporation of BAF derivation method uncertainty into either the AWQC development methodology or effluent compliance determination procedure will eliminate false positive violations and will assure that resources are not wasted in terms of unnecessary treatment process upgrades or permit violation penalties.

--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
"Potential positive bias" is conjecture not supported by data. The commenter is suggesting that BAFs should be intentionally biased.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA disagrees with (a): the 2000 Human Health Methodology is clear on this subject  -  site-specific data are always preferred, followed by regional, state and national last.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA also disagrees with (b): There is no basis for "a specific uncertainty factor" to be applied to BAFs. This consideration totally depends on the data in hand, resources, and the level of acceptable uncertainty given the project goals.
5
Further, the BAFs vary not only depending upon the method of derivation, as discussed above, but also depending upon the time of derivation using the same "more preferred" method. The attached figure (see Attachment 1 at end of this document) shows total PCB concentration in fish tissue, sediment, and water column for San Francisco Bay for 1997, 2000, and 2006. The fish species is White Croaker, a trophic level 4 fish. The total PCB consists of a total sum of 41-50 congeners. The data is from the San Francisco Bay Regional Monitoring Program (RMP) annual reports prepared by the San Francisco Estuary Institute (SFEl.org).

Annual averages across all RMP sampling stations were used for both sediment and water column. Samples were collected approximately three times a year. There were about 25 sampling stations throughout the Bay. Fish samples were collected approximately once per year. The fish tissue concentrations are an average of 9-24 individual samples per year.

Using this data, the field-measured BAFs for PCB for 1997, 2000, and 2006 for trophic level 4 fish for San Francisco Bay are approximately 288,000, 310,000, and 1,000,000. Obviously, unlike bioconcentration factor (BCF) based AWQC (1980 methodology), the BAF based AWQC (proposed methodology) is a moving target. As is evident from the field data, the sediment and water column PCB concentrations have decreased by two-thirds in the last decade. Yet, there has not been a corresponding decrease in fish tissue concentrations over this period. Therefore, it appears that there may be a lag period of over 10 years between the decrease in sediment/water column and the fish tissue PCB concentration (for White Croaker).

BACWA believes that the same may be true for other similar chemicals of concern, like dioxin, etc. Note that the observed increase in fish tissue concentrations in 2006 is not due to increased lipid content of the fish, as one might suspect. The lipid content was actually higher in 1997 than in 2006.

--------------------------------------------------------------------------------
There is no reason for EPA to investigate the derivation of site-specific BAFs for San Francisco Bay, as part of this guidance. The summary data offered by BCWA do not include sufficient information to determine representativeness, accuracy, or uncertainty of these data. Furthermore, total PCB concentrations reflect a mixture of chemicals, which violates the TSD guidance requiring BAFs be derived from chemical-specific measurements. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Modeling simulation of BAF sampling designs (Section 3.2.2) addresses how temporal variability of water and biota concentrations relates to optimal field sampling designs. "Lag time" in Π is addressed in Section 4.3.3.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA clearly acknowledges that BAFs and BSAFs can (and will) change with different environmental exposure conditions. This knowledge supports the position that in ecosystems with changing conditions, BAFs should be evaluated on a periodic basis. States are required to reevaluate AWQC periodically, at least once every three years, which provides such an opportunity.
6
In the current AWQC development methodology (1980 methodology), the BCF is a fixed value because it is laboratory derived. For example, the BCF for PCBs is 31200. This value has been used for almost 30 years and is still valid today. On the other hand, the above San Francisco Bay field data suggests the BAF value varies depending upon time of derivation (sampling). Therefore, a BAF value determined today may not be valid 5 years from now. Thus, the BAF value and the resulting AWQC for a given chemical of interest determined using the TSD proposed guidelines may be substantially under or overprotective in the next 5-10 years. This is especially true for the national BAF values that the USEPA has proposed to establish. The national BAF values would most likely be based upon a limited temporal data (1-2 years) and may not be representative of future conditions. BACWA is concerned about the BAFs temporal variability and believes that the TSD has not fully addressed this issue.

--------------------------------------------------------------------------------
EPA disagrees with the commenter's logic: Simply because a BCF value "has been used for almost 30 years", it does not necessarily follow that it "is still valid today" (or for that matter, that it ever was valid).
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA has already acknowledged that temporal variability is an issue with developing a site-specific BAF, and EPA agrees that temporal variability may also affect how "valid" a national BAF might be over time for a given application. Again, that concern would necessarily be evaluated on a case-by-case basis.
7
BACWA strongly recommends that USEPA cannot adopt this methodology without providing further guidelines on appropriately handling the BAF temporal variability, so that the adopted BAF for a given species or trophic level remain valid throughout the period of its application in water quality regulations.

--------------------------------------------------------------------------------
EPA has addressed time variability in a number of places in TSD Volume 3 (see FWQC comment 11 and response, p.7-8 above). One could argue that anything related to science and data quality may change over time, which in this case could lead to an AWQC becoming outdated to some degree. EPA does not consider AWQC to be "invalid" after a specific period of time. EPA considers this to be a contaminant-specific data issue.

Commenter: James J. Pletl, PhD 
Hampton Roads Sanitation District
--------------------------------------------------------------------------------

1
HRSD encourages EP A to develop equally sound guidance for BAF development for bioaccumulated metals, particularly since mercury impairments are so prevalent nationwide. Additional comments are detailed below: 

Section 2.1 states that guidance on determining national BAFs for inorganic and organometallic
--------------------------------------------------------------------------------
chemicals is provided in Section 5.6 of the 2000 Human Health Methodology.
2
Data Quality Objectives and uncertainty: 
    Though the guidance is technically sound, HRSD does have concerns with the lack of specific information regarding Data Quality Objectives (DQOs) and uncertainty. EPA states in section 1.2 that national BAFs "can be applied to achieve an acceptable degree of accuracy", but EPA does not define the degree of accuracy that is acceptable and how this was determined. Further, EPA needs to define the level of accuracy that can be attained using the approaches of this guidance. 
  
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
The acceptable degree of accuracy depends on the management goals, the site, the chemical(s), and other related factors. This determination is integral with the problem formulation and DQO/QAPP process before sampling.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
TSD Volume 3 references current DQO guidance documents.
3
EPA states that it recognizes that more guidance is necessary to ensure that BAFs are accurate and defensible, but this additional guidance should have been written before site specific approaches were developed. The additional guidance would likely take the form of DQOs and MQOs; and according to EPA's own guidance these objectives are selected before studies take place. 
EPA has revised this statement (Section 1.3) to clarify it's meaning: "EPA believes that the additional guidance provided in this TSD will help to ensure that site-specific bioaccumulation factors are accurate and defensible, whether they are determined directly by field measurement or indirectly by estimation methods."
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Additional quality assurance guidance, in the form of DQOs and MQOs, is beyond the scope of this document because this document is not a study. 
4
    Though the guidance references EPA's DQO process in section 3.1, EPA does not follow the DQO process in this document. Since the quantitative and qualitative objectives of the approaches were not established first, EPA has not and can not determine whether any of these approaches meet the uses of the data. EPA must establish the DQOs and MQOs for these values as well as any others that can be used in a regulatory context. 
  
--------------------------------------------------------------------------------
Since preparing this document is not a field study, EPA did not follow a DQO process. We do refer to appropriate DQO guidance in the TSD.
5
     In section 2.5.8, EPA maintains that BAFs should be reviewed to assess sufficiency, quality, variability and overall uncertainty, but the guidance does not address how this should be done and what is acceptable for each measure. The guidance indicates that Sections 3 and 4 address quality assurance, but a system and expectations for QA have not been provided. 
  
--------------------------------------------------------------------------------
Again, this is not a field study. Expectations for QA must be provided for specific studies.
6
    EPA asserts that national BAFs "are based upon the highest-quality data"; however EPA has not outlined the objectives defining data quality. As stated above, EPA must first determine the quality and quantity of data required to conduct a study before a study can be carried out and this guidance does not address this requirement. Section 2.5.7 states that "what constitutes high quality data" is addressed in Sections 3 and 4, but this is not the case. These sections discuss the quantity of data to be used in relation to variability, but this does not address the accuracy, representativeness or comparability of resulting BAFs. 
  
--------------------------------------------------------------------------------
The commenter is incorrect. The quality assurance elements that define "high-quality" data are defined in Sections 3.3 (biota), 3.4 (water) and 4.5 (sediment). 
7
    EPA presents a number of options for calculating a site specific BAF, but does not address the uncertainty associated with each option. Section 2.5.8 states that each approach must be compared in relation to uncertainty using a weight of evidence approach, but the guidance does not address how to do this. EP A should provide users with information on the relative uncertainty for each of the options and, as each approach will incur different costs, specifics on the relative costs associated with the case studies so that users can make informed choices about how to proceed with developing a site specific BAF. 
  
--------------------------------------------------------------------------------
The commenter is incorrect. The hierarchy of site-specific BAF methods is consistent with the degree of relative uncertainty that they can be expected to produce. Methods based on site-specific measurements are preferable to methods based on prediction or extrapolation. This is explained in Section 2.5.7 and Table 2-1. Of course, the collection of high-quality data from the site is the key determinant of BAF uncertainty.
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
In addition, results are presented in various sections of this TSD that show the relative uncertainty of specific BAF methods.
8
    If analytical methods can not meet the quality objectives of a study either different methods must be used or the quality objectives must be changed and, hence, the use of the data will change. For example, if quality objectives are established for a BAF to develop a water quality criterion, a TMDL and permit limits, and these objectives can not be met with current analytical methods, then the resulting limits can not be implemented in the same way as when objectives are met. 
  
--------------------------------------------------------------------------------
EPA disagrees with this comment. DQOs are a site-specific determination. Furthermore, we do not believe that using more sensitive techniques will change the use of the data.
9
 Use of uncensored data: 
    Section 3 states that uncensored data should be used in the calculation of BAFs. This is completely contradictory to the recommendations of EPA's Federal Advisory Committee on Detection and Quantitation. This FAC recommended that only values exceeding the respective quantitation limit (QL) be used to make regulatory decisions. Therefore, BAFs should not be based on data below the QL. It was not recommended that data users substitute a default value to any measurement less than the minimum detection limit (MDL). Such approaches will bias the results, resulting in inaccurate estimates of the BAF and possibly violating a BAF study's MQO for accuracy. 
  
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Since determining a BAF is not a regulatory decision, BAFs can be based on data below the QL without violating the FAC guidance. Page 3-9 provides broader guidance regarding treatment of low-end censored data than is cited in the comment. In the same paragraph, EPA also cites appropriate statistical methods for dealing with this issue.
--------------------------------------------------------------------------------

EPA has revised this section to soften the "should only be calculated from concentration data if fewer than
--------------------------------------------------------------------------------
20% of the reported values are nondetects" recommendation, since this is more of a "rule of thumb" than a standard practice.
10
 More specific Guidance or clarification needed: 
    In section 3.1, the text should address timing of water quality and sediment samples with the period within which a species may be fished and consumed by humans. For example, water and oyster tissue samples collected in May should not be relevant if oysters can not be harvested that month or for several more months. Similarly soft shell blue crab tissue concentrations in the Mid-Atlantic region should not be compared to water concentrations in the fall, for example, because the soft shell condition only occurs in the spring in these waters. 
  
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
This comment is addressed in "Key Study Design Questions" 4 (Biota) and 5 (Temporal Variability). In both sections, guidance is offered that investigators should appropriately consider consumption patterns.
11
Section 3.1 discusses extrapolation of individual BAFs in large-scale sites and that these BAFs must be representative, but the guidance does not define what "representative" means or to what extent BAFs can be extrapolated. EPA must provide more specific guidance on these expectations for users to successfully use this document. 
EPA defines representative in this TSD. Page 3-11 states, "The most important aspect of conducting a successful field study is collecting `representative' samples of biota and water. In general, samples will be most representative when the measured concentrations are reflective of long term average concentrations for the chemical in biota and exposure media."
--------------------------------------------------------------------------------

12
 
      In section 3.3.3 and 3.3.4, the guidance states that biota sampling should be conducted near point and nonpoint source discharges and "pollutant sinks", and instructs studies to collect organisms just before they migrate out of the site. These approaches may result in a bias to the BAF calculation. A site specific BAF should be representative of the entire site, spatially and temporally, rather than specific to one or two characteristics of that site. EPA's guidance must be clear in the goal of deriving unbiased BAFs. 

--------------------------------------------------------------------------------
EPA will edit page 3-37 to remove "areas near chemical contamination sources", so the paragraph begins "Biota sampling should be conducted in frequently fished areas, possibly including the following locations".
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
EPA will modify the italicized text to:
--------------------------------------------------------------------------------
Areas near point source discharges such as 
--------------------------------------------------------------------------------
and:
--------------------------------------------------------------------------------
Areas receiving nonpoint source inputs such as
13
      
      EPA recognizes in section 3.3.5 that a small percentage of the human population may consume tissues that naturally concentrate contaminants (decapod hepatopancreas) and that this should be addressed. The guidance should also instruct studies on how this should be addressed. It is likely that consumption, by weight, for such tissues is much less than that of other more commonly consumed tissues and that this must weigh in the calculations of BAFs. 
      
   *      Glossary - the BAFi definition does not indicate whether the concentration in water is total or dissolved, it is only inferred by other definitions. This should be clarified.
     
--------------------------------------------------------------------------------
EPA has added a reference to the appropriate section (4.3.3.1) of the 2000 Human Health Methodology for consumption studies. 
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
BAF is first introduced in the glossary without considering how the biota and water concentrations are defined. The BAFi definition is then followed by definitions for baseline and total BAFs.EPA considers this presentation to be both clear and understandable.
14
      Section 3.4.3 refers to the use of a dye study to estimate chemical concentration gradients at sampling locations. This may be useful, however the document should stress that the chemical and physical properties of the dye selected must be similar to that of the chemical being studied for conclusions regarding concentration to be reliable. 
  
--------------------------------------------------------------------------------
EPA has deleted the sentence referring to dye studies because dye studies represent stream flow, not concentration gradients. 
15
   Terminology 
      
      Section 2.3 - EPA should define what a "reasonable' estimate of bioaccumulation is and "long-term conditions" that are used to describe steady state. 
      
  
--------------------------------------------------------------------------------
EPA has replaced "reasonable" with "best". "Long-term" is parenthetically equated to "steady-state", which is defined in the glossary.
16
      EPA also uses the terms "detect" and "quantified" interchangeably (see 3.4.1), which is incorrect. 
--------------------------------------------------------------------------------
EPA has replaced "detect" with "measure" in Section 3.4.1.
17
      Section 3.3.6 and 3.4.6 - EPA uses the terms "replicate" and "duplicate" interchangeably; however they are not synonyms. Replicates represent splits of a single sample; duplicates represent completely independent samples collected under identical conditions. The guidance should be adjusted accordingly. 
--------------------------------------------------------------------------------
EPA has removed the parenthetical "duplicates" in Sections 3.3.6 and 3.4.6.


                             Commenter: B. Sachau
15 Elm St., Florham Park, NJ
                                       
1

Is EPA unbiased enough to make water quality recommendations. This agency has allowed every profiteer who came before it for the last 17 years to produce and profiteer from toxic chemicals, and this same agency has accepted what those profiteers told the agency as to the safety of those toxic chemicals, resulting in American citizens carrying around in their body over 200 toxic chemicals today. Every little baby has those toxic chemicals in their body. Is an agency like that qualified to make safe water quality regulations, when it is obviously in the pocket of the toxic chemical industry. It doesn't look like it to me. 

I think we need to disband EPA and start fresh. 
--------------------------------------------------------------------------------
These comments have no technical merit. No response is necessary.
 Attachment 1 from Bay Area Clean Water Agencies
