Bay Protection and Toxic Cleanup Program - (BPTCP)


RECOMMENDATIONS ON BPTCP MONITORING ACTIVITIES

STATE OF CALIFORNIA
STATE WATER RESOURCES CONTROL BOARD
REGIONAL WATER QUALITY CONTROL BOARDS
DEPARTMENT OF FISH AND GAME
SCIENTIFIC PLANNING AND REVIEW COMMITTEE:

JANUARY 1997

PREFACE

The Scientific Planning and Review Committee (SPARC) was convened by the State Water Resources Control Board's Bay Protection and Toxic Cleanup Program (BPTCP) to review the scientific aspects of the Program's monitoring activities. SPARC has held two meetings. This report summarizes the SPARC recommendations.

The SPARC recommendations have been used by the BPTCP staff to (1) improve the Statewide monitoring approach and the Program's Quality Assurance Project Plan, (2) develop better ways to effectively identify polluted sites, and (3) train the scientists employed by the Department of Fish and Game, the Regional Water Quality Control Boards and the State Water Resources Control Board to provide more informed assessments of polluted sites.

SCIENTIFIC PLANNING AND REVIEW COMMITTEE

  • Dr. Rick Swartz, Environmental Protection Agency
  • Ms. Rachel Friedman-Thomas, Washington State Department of Ecology
  • Dr. Bruce Thompson, San Francisco Estuarine Institute
  • Dr. Mel Suffet, Environmental Science and
  • Engineering Program, University of California, Los Angeles
  • Dr. John Knezovich, Health & Ecological Assessment Lawrence Livermore Laboratory
  • Dr. Don Stevens, Environmental Statistics and Aquatic Monitoring ManTech Environmental Research Services, Inc.
  • Dr. Ed Casillas, Environmental Conservation Division National Marine Fisheries Service

TABLE OF CONTENTS

LIST OF APPENDICES

APPENDIX A: Scientific Planning and Review Committee Briefing Document for the Bay Protection and Toxic Cleanup Program (March 1995)

APPENDIX B: Scientific Planning and Review Committee Briefing Document for Recommendations on the Bay Protection and Toxic Cleanup Program Monitoring Activities (May 1996)

 

EXECUTIVE SUMMARY

The Scientific Planning and Review Committee (SPARC) was established by the State Water Resources Control Board in 1994 to review the scientific aspects of the Bay Protection and Toxic Cleanup Program (BPTCP) monitoring activities. The SPARC members are independent experts representing the fields of toxicology, benthic ecology, organic and inorganic chemistry, program implementation and direction, experimental design, and statistics. This report contains the recommendations of the SPARC that were solicited at technical workshops held April 12-13, 1995 and May 15-17, 1996. This report also contains the briefing documents provided to the SPARC prior to the two workshops.


During the two meetings the SPARC made over 100 recommendations on all aspects of BPTCP monitoring. The SPARC discussed approaches for interpreting the toxicity, chemistry, and benthic data collected during the BPTCP monitoring efforts. SPARC also addressed bioaccumulation of contaminants and several Region-specific issues. While differences of opinion are shared among the members, the SPARC reached a strong consensus on the BPTCP monitoring and data interpretation approaches.


There was a strong vote of confidence by SPARC for using a triad of measures (i.e., toxicity testing, sediment chemical measures, and assessments of benthic organisms) to identify the worst toxic hot spots. There was also agreement on the criteria for identifying toxic hot spots using the triad of measures.


Overall, it was clear that the SPARC endorsed the BPTCP's approaches for monitoring and data interpretation. SPARC also encouraged the BPTCP to publish the results of the monitoring efforts in peer-reviewed scientific literature.



BAY PROTECTION AND TOXIC CLEANUP PROGRAM

SCIENTIFIC PLANNING AND REVIEW COMMITTEE

INTRODUCTION

The California Water Code established the Bay Protection and Toxic Cleanup Program (BPTCP) to protect the existing and future beneficial uses of California's bays and estuaries. The BPTCP has provided a new focus on identifying polluted and contaminated locations in California's bays and estuaries. The BPTCP has four major goals: (1) protect beneficial uses of bay and estuarine waters; (2) identify and characterize toxic hot spots; (3) plan for the prevention and control of further pollution at toxic hot spots; and (4) develop plans for remedial action at existing toxic hot spots and prevent the creation of new hot spots. The primary focus of the BPTCP has been on the identification of toxic hot spots.


The SWRCB established the SPARC in 1994. The SPARC brings together independent experts in the fields of toxicology, benthic ecology, organic and inorganic chemistry, program implementation and direction, experimental design, and statistics to review the monitoring approaches taken by the BPTCP. The committee has provided comments on the Program's monitoring approach(es), given input on the scientific merit of the approach(es) taken, and provided suggestions for monitoring improvement.


In 1995 and 1996 the Bay Protection and Toxic Cleanup Program (BPTCP) sponsored two meetings of the Scientific Planning and Review Committee (SPARC). The purpose of this report is to present the recommendations provided by the SPARC.


SCIENTIFIC PLANNING AND REVIEW COMMITTEE

APRIL 1995 RECOMMENDATIONS

Focus of the April 1995 Workshop

The workshop centered around the following key questions:


1. What is toxic?


2. How should we show association between toxicity, benthic community, etc. and chemical concentrations?


3. What is a benthic impact?


4. Should we use a probability-based sampling design (random sampling) or directed point sampling approach (i.e. based on best professional judgment)?


5. Should we use a screening and confirmation approach?


6. What biological methods should we use?


7. What chemical methods should we use?


Please refer to Appendix A for the issue papers that describe each of these issues.


Recommendations

The SPARC recommendations from the April 1995 meeting were:


Issue 1. Toxicity

1. The selection of toxic and reference sites will ultimately be a policy decision based on best available scientific approaches for determining biological response.


2. The reference envelope approach is preferred over simple comparison to laboratory controls, and there is agreement that this is the statistical approach to pursue for determining the level of toxicity suitable for meeting toxic hot spot toxicity criterion.


3. All toxicity data should be normalized to laboratory controls to account for any variation in laboratory factors or test organism condition.


4. Compare test site response to large reference envelope population from a comprehensive data base of reference site results for the protocol used.


5. Compare test site response to reference envelope population from samples collected concurrently with test samples.


6. A site is toxic if it falls below the reference envelope lower bounds for both the reference site data base and concurrent samples.


7. If a site is toxic relative to the large reference envelope population from the comprehensive database, but concurrent reference site results are low, the site should be revisited.


Selection of Reference Sites Within Each Region

Some level of pollution will always be unavoidable. However, reference sites should be selected through the following process:


1. Reference sites should not include those sites where toxicity is observed in association with pollution. Common sense and knowledge of local conditions should be used in order to avoid areas known to be disturbed or polluted.


2. Randomly sample the rest of the water body, conducting analyses of chemistry, benthic community structure, and toxicity.


3. Allow trained benthic ecologists to select the sites that have moderate to high species richness, abundant presence of amphipods or other indicator species, absence of indicators known to be characteristic of polluted sediments, and any other indicator of ecological health that can be argued convincingly.

4. Evaluate the chemistry data and narrow the sites to those that do not exceed more than one upper value of a PEL or ERM for existing chemistry guidelines.


5. Evaluate the toxicity data and eliminate only those sites that have extremely high toxicity, as determined by a qualified toxicologist, not by a priori criteria.


6. Once reference sites are chosen they are sampled along with test sites. Include the new reference site toxicity results in the reference envelope regardless of the magnitude of the toxicity response. The reference envelope toxicity result will fall where it may.


7. Compile a data base of toxicity responses from appropriately selected reference sites, and include past and current reference site data in the reference envelope. Allow the number of data points in the reference envelope to grow as more studies are completed in the area.


Issue 2. Association of Chemistry and Biological Effects

1. Causal relationships between chemistry and biological effects are desirable to provide evidence of links between pollutant concentrations and biological effects. However, correlation does not necessarily establish causality.


2. Development of spiked bioassay data could be used to unequivocally identify chemicals responsible for observed effects.


3. Simultaneous Extracted Metals and Acid Volatile Sulfides (SEM/AVS) data is essential for understanding metal effects.


4. Measurement of Total and Dissolved Organic Carbon (TOC and DOC) in the pore water is recommended to help understand organic and metal bioavailability.


5. The effect of oxidation state of the environment and of the chemical compounds should be investigated.


6. Pore water toxicity and chemistry are valuable in determining causal relationships.


7. It is recognized that sorbed pollutants may become bioavailable after ingestion and metabolism.


8. Professional judgement and knowledge of local conditions should be used to decide how best to allocate resources to determine causal relationships.


9. The Program should use all available criteria and biological measurements in assessing the relationships between chemistry and biological effects (i.e., use weight of evidence approach).


Issue 3. Benthic Impacts

No single index is defensible in a regulatory setting. A site should be characterized as "healthy", "intermediate", or "degraded" based on the best professional judgement of a qualified ecologist, using whatever methods are most appropriate to the site.


Replication of Benthic Ecological Analysis

An analysis of existing data should be conducted to determine benthic replication, keeping in mind the types of analyses that can be done with benthic data, the cost of the analysis and benefits derived. Do not replicate unless there is a clear reason to do so. Broad spatial/temporal coverage of sampling is usually preferable to replication at fewer stations/times.


Issue 4. What is the most appropriate sampling design

1. During the screening phase, sampling should incorporate a stratified random design in order to provide an opportunity to find unknown toxic hot spots.


2. Confirmation phase sampling should be based on grids covering the site of concern, with random placements of stations within grid blocks.


3. Grids should be configured to match site characteristics.


4. Temporal variations should be accounted for with repeated sampling at locations at least one meter apart.


5. Spatial and temporal scales should be based on knowledge of the site.


Field Replication

6. Random sampling over suitably sized grids may be preferable to replication. There is no need to replicate unless there is a clear and defensible reason why.


7. It would be best to conduct statistical analysis of past data to determine replication needs for future work.


Issue 5. Toxic Hot spot designation (Screening and Confirmation approach)

1. A three tiered data analysis approach should be used. This would include chemical, toxicity, and benthic community analyses. Having hits in all three components of a triad analysis, would classify a site as a worst case toxic hot spot. Hits on fewer than all three would result in classification as a site of concern. All sites could be ranked in this way.


2. Under the BPTCP, the screening phase would consist of using either toxicity or benthic community analysis or chemistry or bioaccumulation data or some combination of all of these. Screening should be flexible, designed to fit the Regional Board's needs. Analysis in this phase should be done only when needed to provide sufficient information to convince the Regional Boards to list or consider the site as a priority site of concern for further action. A hit in any of these analyses would elicit concern, trigger confirmation phase monitoring under the BPTCP and/or perhaps prompt a specific Regional Board to pursue some other type of regulatory review action. It would be very important to involve potential responsible parties as early in the process as possible and coordinate studies and funding.


3. The confirmation phase should consist of toxicity and chemistry and benthic community analyses on a previously visited site of concern or wherever previous evidence indicates a site may be impacted. A confirmatory hit in all three analyses performed during this phase would classify a site as a worst case toxic hot spot. This phase could also include intensive investigations to identify causal relationships, and intensive grid sampling necessary to show gradients and spatial extent.


4. Allow for a mechanism for de-listing sites if intensive studies prove preliminary designation was in error.


5. It is important to focus on the most impacted sites for successful toxic hot spot designation and application of regulatory actions.


Issue 6. Appropriate Biological Methods

1. Use the amphipod 10 day solid phase test and the sea urchin 96 hour larval development test in pore water for screening sites.


2. Use the amphipod solid phase test, the sea urchin larval development test in pore water, and the sea urchin larval development test at the sediment water interface (SWI) for confirmation. (A sensitive chronic test, such as the 28 day protocol for Leptocheirus, or tests using resident species may also be useful for confirmation).


3. Centrifuge pore water for bioassay test. Use non-sorbing centrifuge tubes such as stainless steel, glass and/or Teflon. Frozen storage is not acceptable for biological testing.


4. Pore water dilutions are not necessary for screening, but do provide additional information for confirmation.


5. Pore water toxicity coupled with chemical analyses may be useful for establishing correlations between chemistry and biological effects.


6. Use of the Neanthes test should be discontinued because it provides no additional information beyond that provided by the amphipod and sea urchin protocol.


7. Studies should be conducted to investigate whether inhibition of embryo/larval development in pore water and solid phase (SWI) exposures can be correlated, or is associated with ecological perturbation, such as impacts on benthic community structure.

Biomarkers

1. Biomarker analyses are currently difficult to interpret in terms of ecological effects. These types of analyses should not be used for toxic hot spot designation at present.


2. Biomarker analyses may be useful in monitoring cleanup activities to determine if there is continued exposure to pollutants.

Bioaccumulation


Recruit the services of a bioaccumulation expert into SPARC and examine how bioaccumulation can be used in the BPTCP.

Issue 7. Appropriate Chemical Methods

Metals

1. Perform SEM/AVS with caution in evaluating potential for metal toxicity. This value may change over time at individual sites due to fluctuations in the concentration of AVS.


2. Use performance-based approach rather than rigid protocols.


3. Do bulk-phase metals in screening.


4. Do pore water metals when deemed necessary. It may help determine causality for confirmation and cleanup planning.


5. Preserve original samples for pore water chemistry.


6. Sediment extracts can be frozen for a year for chemical analysis. The time listed in standard methods for water and waste water should be the maximum holding time (Mel Suffet, personal communication, December 1996).

Organics

The April 1995 meeting ended before the organic chemical methods could be fully discussed. Nevertheless, similar recommendations to metal chemical methods were made. Further examination of this topic is scheduled for the next SPARC meeting.

1. The analyte list should be expanded to include Diazinon and other organophosphate pesticides


2. Use performance-based approach rather than rigid protocols.


3. Do bulk-phase organics and TOC in screening.


4. Do pore water organics to help determine causality for confirmation and cleanup planning.


5. Preserve original samples for pore water chemistry.


6. Sediment extracts can be frozen for a year for chemical analysis.


Region-specific Recommendations

Region 1

If local problems can be identified without toxicity screening then proceed to use the available resources as effectively as possible.


Bioaccumulation data may be appropriate to identify problem chemicals, biological exposure and potential sources of pollution in Region 1.


Biological effects measurements (toxicity screening or benthic community analysis) should be considered in cases where unknown toxic hot spots are present.


Region 2

Sampling should be done at a predetermined standard depth in a way to avoid mixing oxic and anoxic sediments. It would be desirable to show the effects of changes in oxidation state on toxicity and toxicity/chemistry relationships.


Use appropriate amphipod species based on knowledge of species tolerance limits to ammonia, salinity, and grain size.


Determine how to include bioaccumulation data into toxic hot spot screening.


Region 5

Pursue monitoring of pesticide degradation products.


Request that the SWRCB, Regional Boards, and Federal agency executive management agree to coordinate monitoring programs and share information from studies in the Bay-Delta. Also that the two Regional Boards pursuing BPTCP work in the Bay-Delta coordinate the planning and monitoring work.

SCIENTIFIC PLANNING AND REVIEW COMMITTEE


MAY 1996 RECOMMENDATIONS

Focus of the May Workshop

The topics discussed in the May meeting addressed the following topics:


1. Review and incorporation of the SPARC recommendations into the Statewide monitoring approach.


2. Interpretation of toxicity data collected.


3. Interpretation of the benthic community data collected.


4. Setting priorities using a weight-of-evidence approach.


5. Review of the studies of water column toxicity and chemistry in the Central Valley Region.


6. Completion of the discussion on organic chemistry methods.


7. The use of bioaccumulation monitoring techniques.


The briefing document that describes each of these issues is presented in Appendix B.


Recommendations

The workshop centered around the following key issues:


Issue 1: Determination of Significant Toxicity Relative to the Surrounding Water Body

1. There is consensus support for the reference envelope concept because it includes all sources of laboratory and field variation affecting toxicity test results.


2. Unexplained toxicity in samples from reference sites should be considered a problem if it occurred in more than 25% of reference samples, and should not be considered a problem if it occurred in less than 10%. There was no SPARC resolution on how to use the reference envelope approach if unexplained toxicity occurred in 10%-25% of reference site samples.


3. Investigation of unexplained toxicity should be focussed on identifying either: (a) pollutants that have not been considered previously, or (b) natural toxicity. Identification of either would be a significant finding consistent with program goals.


4. The synergistic effect of mixtures of chemicals found at low concentrations should be considered in any investigation of unexplained bioeffects.


5. The reference envelope should include toxicity data from many different sampling times. Temporal variability should be investigated. If temporal variance exists (i.e., if multiple sites vary concurrently), then the reference envelope equations must be revised to take this factor into account.


6. The reference envelope for toxicity could include reference sites from a broad geographical area (as big as the entire West Coast) or be limited to the local study area, depending on study objectives.


7. Statistical power should be analyzed to determine the minimum number of reference site samples necessary for appropriate use of the reference envelope method. Effects of sample size on data distribution (e.g., normality) should also be examined.


8. To determine statistical significance, study site results should be compared to both:


a. the tolerance limit derived from a reference envelope that includes previous data, and


b. results from concurrently collected local reference site sample(s).


9. Regional Boards should set reference envelope "p" values appropriate for their Regions and study objectives. The "p" is the percentile of the reference distribution used to set tolerance limits. There was SPARC consensus that this value is critical in establishing toxicity thresholds, provides an explicit means of selecting the statistical parameters relevant to study objectives, and should be established through policy decisions.


10. Guidelines for selection of "p" values include:


a. the degree of confidence that reference site samples are indicative of desired ambient water body conditions,


b. the level of degradation exhibited by reference site samples, and


c. the political or economic goals associated with designating study sites as toxic.


Low "p" values would be appropriate for situations where there is high confidence that reference sites are indicative of desired environmental conditions, and the economic or political costs related to a finding of toxicity are high. Higher "p" values are more appropriate when reference sites are assumed to represent less than optimal conditions, or when policy impacts are less severe.


11. Economic analyses could be used in conjunction with information on reference site quality and regulatory goals to help establish suitable "p" values for reference envelope calculations.


12. There may be greater uncertainty associated with the use of low "p" values. The lower the "p" value, the farther it extends into the tail of the reference population distribution, where deviations from normality are most extreme. This should be investigated as part of an examination of sample size and data distribution.


13. The reference envelope approach is strongly tied to an assumption of normality of the underlying data distribution, and that distribution should be checked as a matter of routine. Any suggestion of strong departure from a bell-shaped or triangular distribution (e.g., skewness, multiple modes, or a flat distribution) should be cause to use the reference envelope approach results with caution. If the reference envelope approach produces tolerance limits that are counter to best professional judgment, the following steps should be taken:


a. Check the data distribution, transform data if necessary.


b. Consider switching test protocols (Criteria for protocol rejection should be established).


c. Check that reference sites were selected appropriately.


d. Check if the "p" value is appropriate. This may involve re-evaluation of reference sites, program goals, and/or policy considerations.


e. If unexplained reference site toxicity exists, investigate it. Do not use a statistical test based on reference site data that are poorly understood.


Issue 2: Selection of Reference Sites

1. Do not consider nickel in evaluating reference site chemical pollution. However, use common sense in cases with highly elevated nickel concentrations.


2. While evaluation of SEM - AVS (simultaneously extracted metals minus acid volatile sulfide) is useful in evaluating potential for metal toxicity in reference samples, this value may change over time at individual sites due to fluctuations in the concentration of AVS. In addition, generalizations regarding AVS effects on bioavailability may not apply to all toxic metals. The issue of whether or not AVS - SEM should be used in reference site selection was not resolved by SPARC at this meeting.


3. Effects Range-Median (ERM) and Probable Effects Level (PEL) values are very similar. The lower of the two should be used in screening concentrations of individual chemicals in reference site selection.


4. For reference site selection, a Total DDT concentration of 100 mg/g TOC was suggested as a cutoff value, based on toxicity studies.


5. For reference site selection, use the sum of ERM quotients that totals less than 5. This value was supported by data from numerous studies described at the meeting by Ed Long. However, all available data and criteria (including EPA EqP and lowest AET) should be evaluated, especially in cases of unexplained toxicity.


6. Benthic community data should not be the sole basis for reference site selection because:


a. benthic community impacts can be hard to measure and/or interpret,


b. the community may have adapted to pollutants, and


c. relatively healthy benthic communities can exist in surface layers above polluted strata.


7. There was no resolution on the use of toxicity data in reference site selection. Contrasting issues of unexplained toxicity and potential for subjective data screening could not be resolved by the entire committee.


8. H2S and NH3 at reference sites:


a. Use toxicity test species that can tolerate reference site concentrations.


b. Use exposure systems that can minimize reference site concentrations (e.g., Sediment Water Interface tests).


c. H2S and NH3 are less of an issue with amphipods than with embryos or larvae exposed in pore water tests.

d. The program should use written guidelines for rejecting reference sample toxicity data when H2S or NH3 are above threshold values for test species.


Issue 3: Proposed Tiered Comparison to Determine Significant Toxicity

Significant toxicity relative to the surrounding water body should be determined by comparing the test sample result to:


1. a tolerance limit calculated from a "universal" reference distribution, and/or a tolerance limit calculated from a "local" reference distribution,


2. results from one or more concurrently collected local reference site sample(s), and


3. 80% of the laboratory control survival.


Significant toxicity would be indicated if the sample result was below the tolerance limit selected for the study (either "universal" or "local" or both, above), and significantly lower than the result from a concurrently collected reference site sample (using a one-tailed t-test), and the sample mean survival was less than 80% of the laboratory control mean. (A "universal" reference distribution refers to one derived from sites from a broad geographical area, such as the entire West Coast of the United States.)


The first comparison [to the reference envelope tolerance limit(s)] accounts for all sources of laboratory and field variation affecting toxicity test results. The second comparison addresses the possibility of a unique toxicity event occurring in the water body at the time of sampling. The third comparison precludes a determination of toxicity when a statistically significant difference is smaller than generally believed to be biologically relevant.


The following should be considered in selecting local versus universal reference populations:


a. The "universal" envelope should be used if local reference site sample results fall within the "universal" reference envelope.


b. In "cleaner" areas or Regions, the local reference envelope should take precedence over the "universal".


c. In areas where local reference samples are more toxic than "universal" reference samples, Regional Board staff should select the reference distribution appropriate to meet study objectives.


Issue 4: Central Valley Monitoring

1. Consider measuring selenium.


2. Mercury is likely to become bioavailable in areas where high residence time allows methylation.


3. Mercury source tracking, Ceriodaphnia toxicity studies, and TIEs were well done. Suggestions for obtaining additional evidence for pesticide effects:


a. Benthic communities should be evaluated and linked to toxicity.


b. Water column community effects should be linked to toxicity.


c. Investigate effects on Salmonid prey species and larval fish.


d. Investigate sediment toxicity tests with flow-through site water.


e. Model hydraulic system inputs and flow to further demonstrate fate.


4. EPA staff working with the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) should be made aware of pesticide data to allow better coordination of management programs.


5. Coordinate Delta toxicity studies with California Endangered Species Act studies.


Issue 5: Organic Chemistry Issues

1. The SPARC supports the modification of current BPTCP organic analytical procedures to allow additional analytes to be measured from a single extraction, thereby expanding the analyte list in a cost effective way.

2. Additional analytes of concern that the program should consider measuring include:


a. Cholinesterase inhibitors, such as the organo-phosphates diazinon and chlorpyrifos, and the carbamate carbofuran. BPTCP currently looks for chlorpyrifos but not the others (e.g., carbamates (methomyl) are used heavily in Elkhorn Slough). Organo-phosphates are important in Regions 5 and 2, and probably elsewhere.


c. Triazines (Atrazine in particular). Both Atrazine and Simazine are used in California. These are highly phytotoxic compounds.


d. Higher molecular weight polynuclear aromatic hydrocarbons (HMW PAHs) may be appropriate to add, though consideration should be given to determining the best HMW PAHs to add.


e. Nonylphenolic surfactants are estrogenic compounds which appear to have synergistic effects at low concentrations, and bioaccumulate. Analytical methods are poorly defined but these compounds may come through our current methods.


f. Alachlor and pthalates.


3. Sample matrix is important. As a guideline, for compounds with a low to moderate Log Octanol/Water Partition Coefficient (Log Kow), it would be more useful to analyze for diazinon in water, pore water, and tissue rather than in sediment. Whereas for moderate to high Log Kow, it would be best to measure the sediment and tissue rather than the aqueous phase.


4. PAH fingerprinting can be added to BPTCP analyses for minimal cost. All PAH signatures are not created equal. Rather than comparing the sum of 26 compounds in samples with different PAH profiles, the BPTCP should develop an index to describe a sample's PAH signature so that samples can be "typed" prior to statistical comparison.


5. Samples exhibiting bioeffects without concomitant elevated concentrations of measured chemicals (that may be related to unexplained toxicity) should be investigated to identify the source and nature of the toxicological agent in these cases.


6. For analysis of water samples, samples must be filtered using glass fiber filters. Plastic in filters actively binds organics. Total Organic Carbon (TOC), Dissolved Organic Carbon (DOC), and Total Suspended Solids (TSS) measurements should be taken on these samples in order to provide a normalizing index for analytical results. There remains an unresolved argument in the literature about filters vs centrifugation for sample analysis, but Dr. Suffet has found filtration to work well.


7. All chemistry data should continue to be reported in units of dry weight, along with normalizing factors like TOC and AVS, if possible.


Issue 6: Bioaccumulation

1. Bioaccumulation data and related health advisories should be used to identify chemicals of concern in a study area. The concentrations of those chemicals in test sediments should be given added consideration in the designation or ranking of sites.


2. A large area (e.g., an entire bay) can be considered an area of concern based on tissue contamination. In such cases, source control would be the preferred cleanup option, as activities such as sediment removal may be impractical.


3. Salmon should be considered for use in bioaccumulation studies.


4. Using models to back-calculate tissue concentrations affecting human and ecosystem health from sediment concentrations can lead to estimates of very low chemical concentrations of concern in sediments. However, the effects of bio-accumulating chemicals, and hot spot designation based on those chemicals, should not be totally dismissed because of low concentrations in sediments.


5. Persistence is not the only issue to consider when evaluating bioaccumulation information. Events of limited duration may still affect ecosystem and human health.


6. Fish (and other organism) tissue burdens in the Sacramento/San Joaquin River Delta should be investigated. The contamination observed in previous studies warrants an evaluation of potential risks to human and ecosystem health.


Issue 7: Benthic Community Analyses

1. Choice of indicator species used in BPTCP/EMAP Southern California Coastal Lagoons and San Diego Bay studies was appropriate. There was very little overlap in the presence of positive and negative indicator species.


2. Indicator species selection should be specific to study area. Indicator species should be selected prior to sample analysis, and should include species whose distributions are not limited by natural sediment characteristics likely to be found at study sites (such as grain size, TOC, etc.).


3. The following parameters should be measured (or sampled and preserved) in situ to assist with interpretation of benthic community analyses: grain size, salinity and concentrations of dissolved oxygen, ammonia, hydrogen sulfide, and TOC.

4. Numerical scaling of the benthic index should be re-evaluated and discussed with interested SPARC members and program staff.


5. The cutoff point indicating community degradation should not be chosen arbitrarily. Samples ranked between 1 and 2 on the present index should be individually re-evaluated to determine "degraded" status.


Issue 8: Weight of Evidence Approach

1. BPTCP should evaluate all three legs of the triad (chemistry, toxicity and benthic community analysis) to most effectively use the Weight of Evidence Approach. In the San Diego study, samples missing one leg of the triad should not be ranked as if there were no effect for that analysis. Missing data should be obtained before ranking all sites together, especially in cases where available data suggests possible degradation.


2. Weight of Evidence could be quantified using an approach similar to Chapman/Long's Ratio to Reference. However, it is informative to present each site with numerical values for each leg of the triad. These values could be either the data values from each analysis (such as percent survival for the toxicity tests), or the rank or percentage relative to other sites studied. These values should not be summed, but each leg should be presented individually. This was suggested in addition to color coding on maps, so that color would indicate hot spot status and numerical values would give a sense of the degree of impact.


3. The legs of triad should be applied independently and should not be expected to agree. Information from one type of analysis should not be disregarded because of different information from another type of analysis. Such cases should be evaluated individually to tease out useful information and supporting evidence.


4. It is not necessary to have two toxicity hits; toxicity, chemistry and benthic ecology should be treated equally.


5. Consider a sampling design that allows samples for all triad analyses to be taken from a single sediment grab. This allows synoptic sampling for all analyses, even if benthic or chemistry samples are archived, and could make sampling more economical.


6. High priority stations are sufficiently confirmed by the BPTCP weight of evidence approach to be considered for the next level of Regional Board or responsible party investigations. Moderate priority stations, and stations for which not all triad data are available, still need additional evidence from BPTCP triad approach prior to follow-up by Regional Board or responsible party investigations.


7. Adjacent stations should be evaluated together to look for similar chemistry and bioeffects. A number of closely spaced sites exhibiting impacts and pollution from similar chemicals may qualify as an area of concern.


8. Confirmation should include consideration of spatial extent. Sites should be characterized by at least three stations.


9. The following points should be considered in using chemistry data in ranking sites:


a. Do not use nickel at all (unless concentrations are extremely high) because there is little confidence in the available sediment guidelines.


b. Use MacDonald's Palos Verdes data for DDT.


c. Use both single chemical ERM quotients and quotient averages.


d. Use the average of ERM or PEL quotients in applying the weight of evidence approach, as opposed to the sum of the quotient. This provides a natural cutoff point where averages exceeding 1 indicate elevated chemistry. This number should be used as a guide along with best professional judgment.


e. Subdivide chemicals into groups likely to have additive effects to better estimate combined effects. For example, low molecular weight PAHs are likely to be additive in their biological effects.


f. Even though the effects of many different chemicals are not always additive, combinations of chemicals are still likely to produce increased effects. ERMs and PELs do work empirically and should be used.


10. It was suggested that the BPTCP examine Washington State's algorithms for combining data to establish weight of evidence.


11. Weight of evidence assessments should always include graphical evaluation of the data.


12. The reference envelope approach has been applied to benthic community data and chemistry data (by Bob Smith). There was no consensus on whether this approach should be used by the BPTCP.


Issue 9: Toxicity Identification Evaluations (TIEs)

1. TIE of sediment pore water should be conducted if it furthers study objectives. TIE is especially important in establishing causal relationships.


2. The TIE approach may provide additional information to guide chemical analysis. There was general agreement that Region 5's investigation of pesticide toxicity supported the power of this approach.


3. For sediments, focus on pore water for TIEs, but realize that removing interstitial water from the sediment matrix may alter the physical availability of analytes. Sorption onto system components may effectively alter the characteristics of the sample and the outcome of the TIE. Removal of pore water from the sediment could be considered one step in the TIE process.


4. A non-filtered pore water treatment should be included in the TIE process. Total suspended solids and dissolved organic carbon are important in determining bioavailablity. These should be measured, although measuring TSS in pore water may be difficult.


5. Chemical analysis should be used as part of the TIE process to verify the compounds identified. Chemicals should be measured at the beginning and end of the TIE toxicity exposures to verify stability.

6. Be aware that there are multiple contaminants everywhere, which may confound the ability to remove toxicity in a TIE. Cumulative effects make it difficult to establish cause/effect relationships.


7. Be aware that TIE procedures may not always provide clear answers, and do not eliminate consideration of a site of concern solely on the basis of the inability of a TIE to identify responsible compounds.


MAJOR SUMMARY RECOMMENDATIONS OF THE

SCIENTIFIC PLANNING AND REVIEW COMMITTEE


Major SPARC Recommendations (from the 1995 meeting)

1. Base program decisions on defensible science to provide common ground for all participants and interested parties.


2. Prepare workplans in advance to allow adequate scientific review, efficient allocation of funds, and timely reporting.


3. Use a carefully considered weight-of-evidence approach to accomplish program goals.


4. Include a bioaccumulation expert on the SPARC and examine how bioaccumulation can be used in the BPTCP. Thought should be given to reconciling the two different aspects of toxic hot spot designation: human health risk vs. observed ecological effects.


5. Food web models are not sophisticated enough to allow development of sediment quality criteria based on fish tissue concentrations. The mobility of most fish species limits utility for designation of toxic hot spots on a reasonable scale.


6. Site specific investigations are necessary for toxic hot spot designations. Focus immediately on sites most likely to be successfully designated as a toxic hot spot.


7. Regional Boards must have authority and take responsibility for the planning of work in their respective regions. Local knowledge should be used to focus on the most relevant sites and analyses.


8. In designating toxic hot spots, follow a three-tiered approach: (1) carry out a flexible screening phase using any analysis of the triad or bioaccumulation technique; (2) a confirmation phase using all triad analyses (and); (3)intensive site specific studies demonstrating spatial extent, and causal relationships between pollutants and observed biological effects. It is very important to bring the potential responsible parties into the process as early as possible. Potential responsible parties, and other appropriate entities, should be brought into the process to cooperate in the funding and execution of post-confirmation studies.


9. Confirmation and intensive cleanup studies should use a stratified random sampling design, with grids of suitable size to cover the area of concern. Field replication of all measures (toxicity, chemistry, benthic community structure, and bioaccumulation) should only be used when there is a clear and valid reason. Bioaccumulation studies should be focussed on contaminants in tissues of fish or other organisms.


10. Statistical significance of toxicity should be determined based on a comparison to a reference envelope.


11. Benthic community degradation should not be based on a single index. A single community index is too easily discredited. Benthic community degradation should be based on convincing evidence determined on a site specific basis by a qualified ecologist.


12. Performance-based chemistry should be used.


13. Pore water toxicity, concurrent chemistry and spiked assays may be useful to determine associations between pollutants and biological effects. Correlations are not nearly as convincing in demonstrating associations. The presence of multiple pollutants may complicate interpretation of toxicity test results. A TIE approach would also provide evidence of cause-effects relationships but should be used judiciously because of cost.


14. SEM/AVS are recommended for all samples.


15. Statewide and site-specific chemical objectives should be pursued.


16. Bioavailability concerns complicate interpretation of solid-phase sediment toxicity testing in evaluating the relationships between pollutant and biological effects.

17. Solid-phase sediment toxicity testing is useful for sediment quality assessment and toxic hot spot designation.


Major SPARC Recommendations (from the 1996 meeting)

1. The triad approach now used by the BPTCP is appropriate for identifying the most and least impacted sites, allowing the program to achieve its major goals.


2. BPTCP data collected to date allows for a scientifically defensible ranking of high priority sites. If further study, as part of confirmation or remediation, shows a site to be less of a problem than originally indicated, the site's status can be changed as part of the process. The data is currently sufficient to justify regulatory actions.


3. The State and Regional Boards should be actively cooperating with potential responsible parties to develop funding and study designs for the next level of investigation at sites identified by the BPTCP as sites of concern.


4. Moderately impacted sites should not be disregarded, especially if there are a number of moderately impacted sites in close proximity. Some action, such as source control, may be necessary even if there is not a single high priority station.


5. Sites that have significant toxicity, high chemistry, or a degraded benthic community, but are missing a leg of the triad, should be resampled to complete all three analyses. Information from sites of concern with only two legs of the triad measured should not be compared to sites with all triad components measured until the missing data are collected. Priority should not be downgraded (for sites with two legs of the triad measured) because of missing data.


6. "Other deleterious substances" (ODS), such as hydrogen sulfide, low dissolved oxygen, etc. that are likely to have resulted from human inputs should be considered as chemicals of concern.


7. The BPTCP provides a model for identifying problem sites that other states may wish to follow. SPARC encouraged the program to support publication of objectives, criteria, methods and results in the peer-reviewed literature to make them more widely accepted and available.

CONTRIBUTORS TO THE SUMMARY DOCUMENT

  • John Hunt University of California, Santa Cruz
  • Gita Kapahi State Water Resources Control Board
  • Brian Anderson University of California, Santa Cruz
  • Rusty Fairey San Jose State University
  • John Newman University of California, Santa Cruz
  • Fred LaCaro State Water Resources Control Board
  • Max Puckett Department of Fish and Game
  • Mark Stephenson Department of Fish and Game
  • Craig J. Wilson State Water Resources Control Board

A P P E N D I X A


Scientific Planning and Review Committee

Briefing Document for the Bay Protection and Toxic Cleanup Program


March 1995

 

A P P E N D I X B


Scientific Planning and Review Committee

Briefing Document for Recommendations

on the Bay Protection and Toxic Cleanup Program

Monitoring Activities


May 1996

STATE OF CALIFORNIA
STATE WATER RESOURCES CONTROL BOARD
REGIONAL WATER QUALITY CONTROL BOARDS
DEPARTMENT OF FISH AND GAME
SCIENTIFIC PLANNING AND REVIEW COMMITTEE:

RECOMMENDATIONS ON
THE BAY PROTECTION AND TOXIC CLEANUP PROGRAM
MONITORING ACTIVITIES

JANUARY 1997

PREFACE

The Scientific Planning and Review Committee (SPARC) was convened by the State Water Resources Control Board's Bay Protection and Toxic Cleanup Program (BPTCP) to review the scientific aspects of the Program's monitoring activities. SPARC has held two meetings. This report summarizes the SPARC recommendations.

The SPARC recommendations have been used by the BPTCP staff to (1) improve the Statewide monitoring approach and the Program's Quality Assurance Project Plan, (2) develop better ways to effectively identify polluted sites, and (3) train the scientists employed by the Department of Fish and Game, the Regional Water Quality Control Boards and the State Water Resources Control Board to provide more informed assessments of polluted sites.

SCIENTIFIC PLANNING AND REVIEW COMMITTEE

Dr. Rick Swartz, Environmental Protection Agency

Ms. Rachel Friedman-Thomas, Washington State Department of Ecology

Dr. Bruce Thompson, San Francisco Estuarine Institute

Dr. Mel Suffet, Environmental Science and

Engineering Program, University of California, Los Angeles

Dr. John Knezovich, Health & Ecological Assessment Lawrence Livermore Laboratory

Dr. Don Stevens, Environmental Statistics and Aquatic Monitoring ManTech Environmental Research Services, Inc.

Dr. Ed Casillas, Environmental Conservation Division National Marine Fisheries Service

 

TABLE OF CONTENTS


LIST OF APPENDICES

APPENDIX A: Scientific Planning and Review Committee Briefing Document for the Bay Protection and Toxic Cleanup Program (March 1995)


APPENDIX B: Scientific Planning and Review Committee Briefing Document for Recommendations on the Bay Protection and Toxic Cleanup Program Monitoring Activities (May 1996)


EXECUTIVE SUMMARY

The Scientific Planning and Review Committee (SPARC) was established by the State Water Resources Control Board in 1994 to review the scientific aspects of the Bay Protection and Toxic Cleanup Program (BPTCP) monitoring activities. The SPARC members are independent experts representing the fields of toxicology, benthic ecology, organic and inorganic chemistry, program implementation and direction, experimental design, and statistics. This report contains the recommendations of the SPARC that were solicited at technical workshops held April 12-13, 1995 and May 15-17, 1996. This report also contains the briefing documents provided to the SPARC prior to the two workshops.


During the two meetings the SPARC made over 100 recommendations on all aspects of BPTCP monitoring. The SPARC discussed approaches for interpreting the toxicity, chemistry, and benthic data collected during the BPTCP monitoring efforts. SPARC also addressed bioaccumulation of contaminants and several Region-specific issues. While differences of opinion are shared among the members, the SPARC reached a strong consensus on the BPTCP monitoring and data interpretation approaches.


There was a strong vote of confidence by SPARC for using a triad of measures (i.e., toxicity testing, sediment chemical measures, and assessments of benthic organisms) to identify the worst toxic hot spots. There was also agreement on the criteria for identifying toxic hot spots using the triad of measures.


Overall, it was clear that the SPARC endorsed the BPTCP's approaches for monitoring and data interpretation. SPARC also encouraged the BPTCP to publish the results of the monitoring efforts in peer-reviewed scientific literature.



BAY PROTECTION AND TOXIC CLEANUP PROGRAM

SCIENTIFIC PLANNING AND REVIEW COMMITTEE


INTRODUCTION


The California Water Code established the Bay Protection and Toxic Cleanup Program (BPTCP) to protect the existing and future beneficial uses of California's bays and estuaries. The BPTCP has provided a new focus on identifying polluted and contaminated locations in California's bays and estuaries. The BPTCP has four major goals: (1) protect beneficial uses of bay and estuarine waters; (2) identify and characterize toxic hot spots; (3) plan for the prevention and control of further pollution at toxic hot spots; and (4) develop plans for remedial action at existing toxic hot spots and prevent the creation of new hot spots. The primary focus of the BPTCP has been on the identification of toxic hot spots.


The SWRCB established the SPARC in 1994. The SPARC brings together independent experts in the fields of toxicology, benthic ecology, organic and inorganic chemistry, program implementation and direction, experimental design, and statistics to review the monitoring approaches taken by the BPTCP. The committee has provided comments on the Program's monitoring approach(es), given input on the scientific merit of the approach(es) taken, and provided suggestions for monitoring improvement.


In 1995 and 1996 the Bay Protection and Toxic Cleanup Program (BPTCP) sponsored two meetings of the Scientific Planning and Review Committee (SPARC). The purpose of this report is to present the recommendations provided by the SPARC.


SCIENTIFIC PLANNING AND REVIEW COMMITTEE


APRIL 1995 RECOMMENDATIONS

Focus of the April 1995 Workshop

The workshop centered around the following key questions:


1. What is toxic?


2. How should we show association between toxicity, benthic community, etc. and chemical concentrations?


3. What is a benthic impact?


4. Should we use a probability-based sampling design (random sampling) or directed point sampling approach (i.e. based on best professional judgment)?


5. Should we use a screening and confirmation approach?


6. What biological methods should we use?


7. What chemical methods should we use?


Please refer to Appendix A for the issue papers that describe each of these issues.


Recommendations

The SPARC recommendations from the April 1995 meeting were:


Issue 1. Toxicity

1. The selection of toxic and reference sites will ultimately be a policy decision based on best available scientific approaches for determining biological response.


2. The reference envelope approach is preferred over simple comparison to laboratory controls, and there is agreement that this is the statistical approach to pursue for determining the level of toxicity suitable for meeting toxic hot spot toxicity criterion.


3. All toxicity data should be normalized to laboratory controls to account for any variation in laboratory factors or test organism condition.


4. Compare test site response to large reference envelope population from a comprehensive data base of reference site results for the protocol used.


5. Compare test site response to reference envelope population from samples collected concurrently with test samples.


6. A site is toxic if it falls below the reference envelope lower bounds for both the reference site data base and concurrent samples.


7. If a site is toxic relative to the large reference envelope population from the comprehensive database, but concurrent reference site results are low, the site should be revisited.


Selection of Reference Sites Within Each Region

Some level of pollution will always be unavoidable. However, reference sites should be selected through the following process:


1. Reference sites should not include those sites where toxicity is observed in association with pollution. Common sense and knowledge of local conditions should be used in order to avoid areas known to be disturbed or polluted.


2. Randomly sample the rest of the water body, conducting analyses of chemistry, benthic community structure, and toxicity.


3. Allow trained benthic ecologists to select the sites that have moderate to high species richness, abundant presence of amphipods or other indicator species, absence of indicators known to be characteristic of polluted sediments, and any other indicator of ecological health that can be argued convincingly.

4. Evaluate the chemistry data and narrow the sites to those that do not exceed more than one upper value of a PEL or ERM for existing chemistry guidelines.


5. Evaluate the toxicity data and eliminate only those sites that have extremely high toxicity, as determined by a qualified toxicologist, not by a priori criteria.


6. Once reference sites are chosen they are sampled along with test sites. Include the new reference site toxicity results in the reference envelope regardless of the magnitude of the toxicity response. The reference envelope toxicity result will fall where it may.


7. Compile a data base of toxicity responses from appropriately selected reference sites, and include past and current reference site data in the reference envelope. Allow the number of data points in the reference envelope to grow as more studies are completed in the area.


Issue 2. Association of Chemistry and Biological Effects

1. Causal relationships between chemistry and biological effects are desirable to provide evidence of links between pollutant concentrations and biological effects. However, correlation does not necessarily establish causality.


2. Development of spiked bioassay data could be used to unequivocally identify chemicals responsible for observed effects.


3. Simultaneous Extracted Metals and Acid Volatile Sulfides (SEM/AVS) data is essential for understanding metal effects.


4. Measurement of Total and Dissolved Organic Carbon (TOC and DOC) in the pore water is recommended to help understand organic and metal bioavailability.


5. The effect of oxidation state of the environment and of the chemical compounds should be investigated.


6. Pore water toxicity and chemistry are valuable in determining causal relationships.


7. It is recognized that sorbed pollutants may become bioavailable after ingestion and metabolism.


8. Professional judgement and knowledge of local conditions should be used to decide how best to allocate resources to determine causal relationships.


9. The Program should use all available criteria and biological measurements in assessing the relationships between chemistry and biological effects (i.e., use weight of evidence approach).


Issue 3. Benthic Impacts

No single index is defensible in a regulatory setting. A site should be characterized as "healthy", "intermediate", or "degraded" based on the best professional judgement of a qualified ecologist, using whatever methods are most appropriate to the site.


Replication of Benthic Ecological Analysis

An analysis of existing data should be conducted to determine benthic replication, keeping in mind the types of analyses that can be done with benthic data, the cost of the analysis and benefits derived. Do not replicate unless there is a clear reason to do so. Broad spatial/temporal coverage of sampling is usually preferable to replication at fewer stations/times.


Issue 4. What is the most appropriate sampling design

1. During the screening phase, sampling should incorporate a stratified random design in order to provide an opportunity to find unknown toxic hot spots.


2. Confirmation phase sampling should be based on grids covering the site of concern, with random placements of stations within grid blocks.


3. Grids should be configured to match site characteristics.


4. Temporal variations should be accounted for with repeated sampling at locations at least one meter apart.


5. Spatial and temporal scales should be based on knowledge of the site.


Field Replication

6. Random sampling over suitably sized grids may be preferable to replication. There is no need to replicate unless there is a clear and defensible reason why.


7. It would be best to conduct statistical analysis of past data to determine replication needs for future work.


Issue 5. Toxic Hot spot designation (Screening and Confirmation approach)

1. A three tiered data analysis approach should be used. This would include chemical, toxicity, and benthic community analyses. Having hits in all three components of a triad analysis, would classify a site as a worst case toxic hot spot. Hits on fewer than all three would result in classification as a site of concern. All sites could be ranked in this way.


2. Under the BPTCP, the screening phase would consist of using either toxicity or benthic community analysis or chemistry or bioaccumulation data or some combination of all of these. Screening should be flexible, designed to fit the Regional Board's needs. Analysis in this phase should be done only when needed to provide sufficient information to convince the Regional Boards to list or consider the site as a priority site of concern for further action. A hit in any of these analyses would elicit concern, trigger confirmation phase monitoring under the BPTCP and/or perhaps prompt a specific Regional Board to pursue some other type of regulatory review action. It would be very important to involve potential responsible parties as early in the process as possible and coordinate studies and funding.


3. The confirmation phase should consist of toxicity and chemistry and benthic community analyses on a previously visited site of concern or wherever previous evidence indicates a site may be impacted. A confirmatory hit in all three analyses performed during this phase would classify a site as a worst case toxic hot spot. This phase could also include intensive investigations to identify causal relationships, and intensive grid sampling necessary to show gradients and spatial extent.


4. Allow for a mechanism for de-listing sites if intensive studies prove preliminary designation was in error.


5. It is important to focus on the most impacted sites for successful toxic hot spot designation and application of regulatory actions.


Issue 6. Appropriate Biological Methods

1. Use the amphipod 10 day solid phase test and the sea urchin 96 hour larval development test in pore water for screening sites.


2. Use the amphipod solid phase test, the sea urchin larval development test in pore water, and the sea urchin larval development test at the sediment water interface (SWI) for confirmation. (A sensitive chronic test, such as the 28 day protocol for Leptocheirus, or tests using resident species may also be useful for confirmation).


3. Centrifuge pore water for bioassay test. Use non-sorbing centrifuge tubes such as stainless steel, glass and/or Teflon. Frozen storage is not acceptable for biological testing.


4. Pore water dilutions are not necessary for screening, but do provide additional information for confirmation.


5. Pore water toxicity coupled with chemical analyses may be useful for establishing correlations between chemistry and biological effects.


6. Use of the Neanthes test should be discontinued because it provides no additional information beyond that provided by the amphipod and sea urchin protocol.


7. Studies should be conducted to investigate whether inhibition of embryo/larval development in pore water and solid phase (SWI) exposures can be correlated, or is associated with ecological perturbation, such as impacts on benthic community structure.

Biomarkers

1. Biomarker analyses are currently difficult to interpret in terms of ecological effects. These types of analyses should not be used for toxic hot spot designation at present.


2. Biomarker analyses may be useful in monitoring cleanup activities to determine if there is continued exposure to pollutants.

Bioaccumulation


Recruit the services of a bioaccumulation expert into SPARC and examine how bioaccumulation can be used in the BPTCP.

Issue 7. Appropriate Chemical Methods

Metals

1. Perform SEM/AVS with caution in evaluating potential for metal toxicity. This value may change over time at individual sites due to fluctuations in the concentration of AVS.


2. Use performance-based approach rather than rigid protocols.


3. Do bulk-phase metals in screening.


4. Do pore water metals when deemed necessary. It may help determine causality for confirmation and cleanup planning.


5. Preserve original samples for pore water chemistry.


6. Sediment extracts can be frozen for a year for chemical analysis. The time listed in standard methods for water and waste water should be the maximum holding time (Mel Suffet, personal communication, December 1996).

Organics

The April 1995 meeting ended before the organic chemical methods could be fully discussed. Nevertheless, similar recommendations to metal chemical methods were made. Further examination of this topic is scheduled for the next SPARC meeting.

1. The analyte list should be expanded to include Diazinon and other organophosphate pesticides


2. Use performance-based approach rather than rigid protocols.


3. Do bulk-phase organics and TOC in screening.


4. Do pore water organics to help determine causality for confirmation and cleanup planning.


5. Preserve original samples for pore water chemistry.


6. Sediment extracts can be frozen for a year for chemical analysis.


Region-specific Recommendations

Region 1

If local problems can be identified without toxicity screening then proceed to use the available resources as effectively as possible.


Bioaccumulation data may be appropriate to identify problem chemicals, biological exposure and potential sources of pollution in Region 1.


Biological effects measurements (toxicity screening or benthic community analysis) should be considered in cases where unknown toxic hot spots are present.


Region 2

Sampling should be done at a predetermined standard depth in a way to avoid mixing oxic and anoxic sediments. It would be desirable to show the effects of changes in oxidation state on toxicity and toxicity/chemistry relationships.


Use appropriate amphipod species based on knowledge of species tolerance limits to ammonia, salinity, and grain size.


Determine how to include bioaccumulation data into toxic hot spot screening.


Region 5

Pursue monitoring of pesticide degradation products.


Request that the SWRCB, Regional Boards, and Federal agency executive management agree to coordinate monitoring programs and share information from studies in the Bay-Delta. Also that the two Regional Boards pursuing BPTCP work in the Bay-Delta coordinate the planning and monitoring work.

SCIENTIFIC PLANNING AND REVIEW COMMITTEE


MAY 1996 RECOMMENDATIONS

Focus of the May Workshop

The topics discussed in the May meeting addressed the following topics:


1. Review and incorporation of the SPARC recommendations into the Statewide monitoring approach.


2. Interpretation of toxicity data collected.


3. Interpretation of the benthic community data collected.


4. Setting priorities using a weight-of-evidence approach.


5. Review of the studies of water column toxicity and chemistry in the Central Valley Region.


6. Completion of the discussion on organic chemistry methods.


7. The use of bioaccumulation monitoring techniques.


The briefing document that describes each of these issues is presented in Appendix B.


Recommendations

The workshop centered around the following key issues:


Issue 1: Determination of Significant Toxicity Relative to the Surrounding Water Body

1. There is consensus support for the reference envelope concept because it includes all sources of laboratory and field variation affecting toxicity test results.


2. Unexplained toxicity in samples from reference sites should be considered a problem if it occurred in more than 25% of reference samples, and should not be considered a problem if it occurred in less than 10%. There was no SPARC resolution on how to use the reference envelope approach if unexplained toxicity occurred in 10%-25% of reference site samples.


3. Investigation of unexplained toxicity should be focussed on identifying either: (a) pollutants that have not been considered previously, or (b) natural toxicity. Identification of either would be a significant finding consistent with program goals.


4. The synergistic effect of mixtures of chemicals found at low concentrations should be considered in any investigation of unexplained bioeffects.


5. The reference envelope should include toxicity data from many different sampling times. Temporal variability should be investigated. If temporal variance exists (i.e., if multiple sites vary concurrently), then the reference envelope equations must be revised to take this factor into account.


6. The reference envelope for toxicity could include reference sites from a broad geographical area (as big as the entire West Coast) or be limited to the local study area, depending on study objectives.


7. Statistical power should be analyzed to determine the minimum number of reference site samples necessary for appropriate use of the reference envelope method. Effects of sample size on data distribution (e.g., normality) should also be examined.


8. To determine statistical significance, study site results should be compared to both:


a. the tolerance limit derived from a reference envelope that includes previous data, and


b. results from concurrently collected local reference site sample(s).


9. Regional Boards should set reference envelope "p" values appropriate for their Regions and study objectives. The "p" is the percentile of the reference distribution used to set tolerance limits. There was SPARC consensus that this value is critical in establishing toxicity thresholds, provides an explicit means of selecting the statistical parameters relevant to study objectives, and should be established through policy decisions.


10. Guidelines for selection of "p" values include:


a. the degree of confidence that reference site samples are indicative of desired ambient water body conditions,


b. the level of degradation exhibited by reference site samples, and


c. the political or economic goals associated with designating study sites as toxic.


Low "p" values would be appropriate for situations where there is high confidence that reference sites are indicative of desired environmental conditions, and the economic or political costs related to a finding of toxicity are high. Higher "p" values are more appropriate when reference sites are assumed to represent less than optimal conditions, or when policy impacts are less severe.


11. Economic analyses could be used in conjunction with information on reference site quality and regulatory goals to help establish suitable "p" values for reference envelope calculations.


12. There may be greater uncertainty associated with the use of low "p" values. The lower the "p" value, the farther it extends into the tail of the reference population distribution, where deviations from normality are most extreme. This should be investigated as part of an examination of sample size and data distribution.


13. The reference envelope approach is strongly tied to an assumption of normality of the underlying data distribution, and that distribution should be checked as a matter of routine. Any suggestion of strong departure from a bell-shaped or triangular distribution (e.g., skewness, multiple modes, or a flat distribution) should be cause to use the reference envelope approach results with caution. If the reference envelope approach produces tolerance limits that are counter to best professional judgment, the following steps should be taken:


a. Check the data distribution, transform data if necessary.


b. Consider switching test protocols (Criteria for protocol rejection should be established).


c. Check that reference sites were selected appropriately.


d. Check if the "p" value is appropriate. This may involve re-evaluation of reference sites, program goals, and/or policy considerations.


e. If unexplained reference site toxicity exists, investigate it. Do not use a statistical test based on reference site data that are poorly understood.


Issue 2: Selection of Reference Sites

1. Do not consider nickel in evaluating reference site chemical pollution. However, use common sense in cases with highly elevated nickel concentrations.


2. While evaluation of SEM - AVS (simultaneously extracted metals minus acid volatile sulfide) is useful in evaluating potential for metal toxicity in reference samples, this value may change over time at individual sites due to fluctuations in the concentration of AVS. In addition, generalizations regarding AVS effects on bioavailability may not apply to all toxic metals. The issue of whether or not AVS - SEM should be used in reference site selection was not resolved by SPARC at this meeting.


3. Effects Range-Median (ERM) and Probable Effects Level (PEL) values are very similar. The lower of the two should be used in screening concentrations of individual chemicals in reference site selection.


4. For reference site selection, a Total DDT concentration of 100 mg/g TOC was suggested as a cutoff value, based on toxicity studies.


5. For reference site selection, use the sum of ERM quotients that totals less than 5. This value was supported by data from numerous studies described at the meeting by Ed Long. However, all available data and criteria (including EPA EqP and lowest AET) should be evaluated, especially in cases of unexplained toxicity.


6. Benthic community data should not be the sole basis for reference site selection because:


a. benthic community impacts can be hard to measure and/or interpret,


b. the community may have adapted to pollutants, and


c. relatively healthy benthic communities can exist in surface layers above polluted strata.


7. There was no resolution on the use of toxicity data in reference site selection. Contrasting issues of unexplained toxicity and potential for subjective data screening could not be resolved by the entire committee.


8. H2S and NH3 at reference sites:


a. Use toxicity test species that can tolerate reference site concentrations.


b. Use exposure systems that can minimize reference site concentrations (e.g., Sediment Water Interface tests).


c. H2S and NH3 are less of an issue with amphipods than with embryos or larvae exposed in pore water tests.

d. The program should use written guidelines for rejecting reference sample toxicity data when H2S or NH3 are above threshold values for test species.


Issue 3: Proposed Tiered Comparison to Determine Significant Toxicity


Significant toxicity relative to the surrounding water body should be determined by comparing the test sample result to:


1. a tolerance limit calculated from a "universal" reference distribution, and/or a tolerance limit calculated from a "local" reference distribution,


2. results from one or more concurrently collected local reference site sample(s), and


3. 80% of the laboratory control survival.


Significant toxicity would be indicated if the sample result was below the tolerance limit selected for the study (either "universal" or "local" or both, above), and significantly lower than the result from a concurrently collected reference site sample (using a one-tailed t-test), and the sample mean survival was less than 80% of the laboratory control mean. (A "universal" reference distribution refers to one derived from sites from a broad geographical area, such as the entire West Coast of the United States.)


The first comparison [to the reference envelope tolerance limit(s)] accounts for all sources of laboratory and field variation affecting toxicity test results. The second comparison addresses the possibility of a unique toxicity event occurring in the water body at the time of sampling. The third comparison precludes a determination of toxicity when a statistically significant difference is smaller than generally believed to be biologically relevant.


The following should be considered in selecting local versus universal reference populations:


a. The "universal" envelope should be used if local reference site sample results fall within the "universal" reference envelope.


b. In "cleaner" areas or Regions, the local reference envelope should take precedence over the "universal".


c. In areas where local reference samples are more toxic than "universal" reference samples, Regional Board staff should select the reference distribution appropriate to meet study objectives.


Issue 4: Central Valley Monitoring

1. Consider measuring selenium.


2. Mercury is likely to become bioavailable in areas where high residence time allows methylation.


3. Mercury source tracking, Ceriodaphnia toxicity studies, and TIEs were well done. Suggestions for obtaining additional evidence for pesticide effects:


a. Benthic communities should be evaluated and linked to toxicity.


b. Water column community effects should be linked to toxicity.


c. Investigate effects on Salmonid prey species and larval fish.


d. Investigate sediment toxicity tests with flow-through site water.


e. Model hydraulic system inputs and flow to further demonstrate fate.


4. EPA staff working with the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) should be made aware of pesticide data to allow better coordination of management programs.


5. Coordinate Delta toxicity studies with California Endangered Species Act studies.


Issue 5: Organic Chemistry Issues

1. The SPARC supports the modification of current BPTCP organic analytical procedures to allow additional analytes to be measured from a single extraction, thereby expanding the analyte list in a cost effective way.

2. Additional analytes of concern that the program should consider measuring include:


a. Cholinesterase inhibitors, such as the organo-phosphates diazinon and chlorpyrifos, and the carbamate carbofuran. BPTCP currently looks for chlorpyrifos but not the others (e.g., carbamates (methomyl) are used heavily in Elkhorn Slough). Organo-phosphates are important in Regions 5 and 2, and probably elsewhere.


c. Triazines (Atrazine in particular). Both Atrazine and Simazine are used in California. These are highly phytotoxic compounds.


d. Higher molecular weight polynuclear aromatic hydrocarbons (HMW PAHs) may be appropriate to add, though consideration should be given to determining the best HMW PAHs to add.


e. Nonylphenolic surfactants are estrogenic compounds which appear to have synergistic effects at low concentrations, and bioaccumulate. Analytical methods are poorly defined but these compounds may come through our current methods.


f. Alachlor and pthalates.


3. Sample matrix is important. As a guideline, for compounds with a low to moderate Log Octanol/Water Partition Coefficient (Log Kow), it would be more useful to analyze for diazinon in water, pore water, and tissue rather than in sediment. Whereas for moderate to high Log Kow, it would be best to measure the sediment and tissue rather than the aqueous phase.


4. PAH fingerprinting can be added to BPTCP analyses for minimal cost. All PAH signatures are not created equal. Rather than comparing the sum of 26 compounds in samples with different PAH profiles, the BPTCP should develop an index to describe a sample's PAH signature so that samples can be "typed" prior to statistical comparison.


5. Samples exhibiting bioeffects without concomitant elevated concentrations of measured chemicals (that may be related to unexplained toxicity) should be investigated to identify the source and nature of the toxicological agent in these cases.


6. For analysis of water samples, samples must be filtered using glass fiber filters. Plastic in filters actively binds organics. Total Organic Carbon (TOC), Dissolved Organic Carbon (DOC), and Total Suspended Solids (TSS) measurements should be taken on these samples in order to provide a normalizing index for analytical results. There remains an unresolved argument in the literature about filters vs centrifugation for sample analysis, but Dr. Suffet has found filtration to work well.


7. All chemistry data should continue to be reported in units of dry weight, along with normalizing factors like TOC and AVS, if possible.


Issue 6: Bioaccumulation

1. Bioaccumulation data and related health advisories should be used to identify chemicals of concern in a study area. The concentrations of those chemicals in test sediments should be given added consideration in the designation or ranking of sites.


2. A large area (e.g., an entire bay) can be considered an area of concern based on tissue contamination. In such cases, source control would be the preferred cleanup option, as activities such as sediment removal may be impractical.


3. Salmon should be considered for use in bioaccumulation studies.


4. Using models to back-calculate tissue concentrations affecting human and ecosystem health from sediment concentrations can lead to estimates of very low chemical concentrations of concern in sediments. However, the effects of bio-accumulating chemicals, and hot spot designation based on those chemicals, should not be totally dismissed because of low concentrations in sediments.


5. Persistence is not the only issue to consider when evaluating bioaccumulation information. Events of limited duration may still affect ecosystem and human health.


6. Fish (and other organism) tissue burdens in the Sacramento/San Joaquin River Delta should be investigated. The contamination observed in previous studies warrants an evaluation of potential risks to human and ecosystem health.


Issue 7: Benthic Community Analyses

1. Choice of indicator species used in BPTCP/EMAP Southern California Coastal Lagoons and San Diego Bay studies was appropriate. There was very little overlap in the presence of positive and negative indicator species.


2. Indicator species selection should be specific to study area. Indicator species should be selected prior to sample analysis, and should include species whose distributions are not limited by natural sediment characteristics likely to be found at study sites (such as grain size, TOC, etc.).


3. The following parameters should be measured (or sampled and preserved) in situ to assist with interpretation of benthic community analyses: grain size, salinity and concentrations of dissolved oxygen, ammonia, hydrogen sulfide, and TOC.

4. Numerical scaling of the benthic index should be re-evaluated and discussed with interested SPARC members and program staff.


5. The cutoff point indicating community degradation should not be chosen arbitrarily. Samples ranked between 1 and 2 on the present index should be individually re-evaluated to determine "degraded" status.


Issue 8: Weight of Evidence Approach

1. BPTCP should evaluate all three legs of the triad (chemistry, toxicity and benthic community analysis) to most effectively use the Weight of Evidence Approach. In the San Diego study, samples missing one leg of the triad should not be ranked as if there were no effect for that analysis. Missing data should be obtained before ranking all sites together, especially in cases where available data suggests possible degradation.


2. Weight of Evidence could be quantified using an approach similar to Chapman/Long's Ratio to Reference. However, it is informative to present each site with numerical values for each leg of the triad. These values could be either the data values from each analysis (such as percent survival for the toxicity tests), or the rank or percentage relative to other sites studied. These values should not be summed, but each leg should be presented individually. This was suggested in addition to color coding on maps, so that color would indicate hot spot status and numerical values would give a sense of the degree of impact.


3. The legs of triad should be applied independently and should not be expected to agree. Information from one type of analysis should not be disregarded because of different information from another type of analysis. Such cases should be evaluated individually to tease out useful information and supporting evidence.


4. It is not necessary to have two toxicity hits; toxicity, chemistry and benthic ecology should be treated equally.


5. Consider a sampling design that allows samples for all triad analyses to be taken from a single sediment grab. This allows synoptic sampling for all analyses, even if benthic or chemistry samples are archived, and could make sampling more economical.


6. High priority stations are sufficiently confirmed by the BPTCP weight of evidence approach to be considered for the next level of Regional Board or responsible party investigations. Moderate priority stations, and stations for which not all triad data are available, still need additional evidence from BPTCP triad approach prior to follow-up by Regional Board or responsible party investigations.


7. Adjacent stations should be evaluated together to look for similar chemistry and bioeffects. A number of closely spaced sites exhibiting impacts and pollution from similar chemicals may qualify as an area of concern.


8. Confirmation should include consideration of spatial extent. Sites should be characterized by at least three stations.


9. The following points should be considered in using chemistry data in ranking sites:


a. Do not use nickel at all (unless concentrations are extremely high) because there is little confidence in the available sediment guidelines.


b. Use MacDonald's Palos Verdes data for DDT.


c. Use both single chemical ERM quotients and quotient averages.


d. Use the average of ERM or PEL quotients in applying the weight of evidence approach, as opposed to the sum of the quotient. This provides a natural cutoff point where averages exceeding 1 indicate elevated chemistry. This number should be used as a guide along with best professional judgment.


e. Subdivide chemicals into groups likely to have additive effects to better estimate combined effects. For example, low molecular weight PAHs are likely to be additive in their biological effects.


f. Even though the effects of many different chemicals are not always additive, combinations of chemicals are still likely to produce increased effects. ERMs and PELs do work empirically and should be used.


10. It was suggested that the BPTCP examine Washington State's algorithms for combining data to establish weight of evidence.


11. Weight of evidence assessments should always include graphical evaluation of the data.


12. The reference envelope approach has been applied to benthic community data and chemistry data (by Bob Smith). There was no consensus on whether this approach should be used by the BPTCP.


Issue 9: Toxicity Identification Evaluations (TIEs)

1. TIE of sediment pore water should be conducted if it furthers study objectives. TIE is especially important in establishing causal relationships.


2. The TIE approach may provide additional information to guide chemical analysis. There was general agreement that Region 5's investigation of pesticide toxicity supported the power of this approach.


3. For sediments, focus on pore water for TIEs, but realize that removing interstitial water from the sediment matrix may alter the physical availability of analytes. Sorption onto system components may effectively alter the characteristics of the sample and the outcome of the TIE. Removal of pore water from the sediment could be considered one step in the TIE process.


4. A non-filtered pore water treatment should be included in the TIE process. Total suspended solids and dissolved organic carbon are important in determining bioavailablity. These should be measured, although measuring TSS in pore water may be difficult.


5. Chemical analysis should be used as part of the TIE process to verify the compounds identified. Chemicals should be measured at the beginning and end of the TIE toxicity exposures to verify stability.

6. Be aware that there are multiple contaminants everywhere, which may confound the ability to remove toxicity in a TIE. Cumulative effects make it difficult to establish cause/effect relationships.


7. Be aware that TIE procedures may not always provide clear answers, and do not eliminate consideration of a site of concern solely on the basis of the inability of a TIE to identify responsible compounds.


MAJOR SUMMARY RECOMMENDATIONS OF THE

SCIENTIFIC PLANNING AND REVIEW COMMITTEE


Major SPARC Recommendations (from the 1995 meeting)

1. Base program decisions on defensible science to provide common ground for all participants and interested parties.


2. Prepare workplans in advance to allow adequate scientific review, efficient allocation of funds, and timely reporting.


3. Use a carefully considered weight-of-evidence approach to accomplish program goals.


4. Include a bioaccumulation expert on the SPARC and examine how bioaccumulation can be used in the BPTCP. Thought should be given to reconciling the two different aspects of toxic hot spot designation: human health risk vs. observed ecological effects.


5. Food web models are not sophisticated enough to allow development of sediment quality criteria based on fish tissue concentrations. The mobility of most fish species limits utility for designation of toxic hot spots on a reasonable scale.


6. Site specific investigations are necessary for toxic hot spot designations. Focus immediately on sites most likely to be successfully designated as a toxic hot spot.


7. Regional Boards must have authority and take responsibility for the planning of work in their respective regions. Local knowledge should be used to focus on the most relevant sites and analyses.


8. In designating toxic hot spots, follow a three-tiered approach: (1) carry out a flexible screening phase using any analysis of the triad or bioaccumulation technique; (2) a confirmation phase using all triad analyses (and); (3)intensive site specific studies demonstrating spatial extent, and causal relationships between pollutants and observed biological effects. It is very important to bring the potential responsible parties into the process as early as possible. Potential responsible parties, and other appropriate entities, should be brought into the process to cooperate in the funding and execution of post-confirmation studies.


9. Confirmation and intensive cleanup studies should use a stratified random sampling design, with grids of suitable size to cover the area of concern. Field replication of all measures (toxicity, chemistry, benthic community structure, and bioaccumulation) should only be used when there is a clear and valid reason. Bioaccumulation studies should be focussed on contaminants in tissues of fish or other organisms.


10. Statistical significance of toxicity should be determined based on a comparison to a reference envelope.


11. Benthic community degradation should not be based on a single index. A single community index is too easily discredited. Benthic community degradation should be based on convincing evidence determined on a site specific basis by a qualified ecologist.


12. Performance-based chemistry should be used.


13. Pore water toxicity, concurrent chemistry and spiked assays may be useful to determine associations between pollutants and biological effects. Correlations are not nearly as convincing in demonstrating associations. The presence of multiple pollutants may complicate interpretation of toxicity test results. A TIE approach would also provide evidence of cause-effects relationships but should be used judiciously because of cost.


14. SEM/AVS are recommended for all samples.


15. Statewide and site-specific chemical objectives should be pursued.


16. Bioavailability concerns complicate interpretation of solid-phase sediment toxicity testing in evaluating the relationships between pollutant and biological effects.

17. Solid-phase sediment toxicity testing is useful for sediment quality assessment and toxic hot spot designation.


Major SPARC Recommendations (from the 1996 meeting)

1. The triad approach now used by the BPTCP is appropriate for identifying the most and least impacted sites, allowing the program to achieve its major goals.


2. BPTCP data collected to date allows for a scientifically defensible ranking of high priority sites. If further study, as part of confirmation or remediation, shows a site to be less of a problem than originally indicated, the site's status can be changed as part of the process. The data is currently sufficient to justify regulatory actions.


3. The State and Regional Boards should be actively cooperating with potential responsible parties to develop funding and study designs for the next level of investigation at sites identified by the BPTCP as sites of concern.


4. Moderately impacted sites should not be disregarded, especially if there are a number of moderately impacted sites in close proximity. Some action, such as source control, may be necessary even if there is not a single high priority station.


5. Sites that have significant toxicity, high chemistry, or a degraded benthic community, but are missing a leg of the triad, should be resampled to complete all three analyses. Information from sites of concern with only two legs of the triad measured should not be compared to sites with all triad components measured until the missing data are collected. Priority should not be downgraded (for sites with two legs of the triad measured) because of missing data.


6. "Other deleterious substances" (ODS), such as hydrogen sulfide, low dissolved oxygen, etc. that are likely to have resulted from human inputs should be considered as chemicals of concern.


7. The BPTCP provides a model for identifying problem sites that other states may wish to follow. SPARC encouraged the program to support publication of objectives, criteria, methods and results in the peer-reviewed literature to make them more widely accepted and available.


CONTRIBUTORS TO THE SUMMARY DOCUMENT

John Hunt University of California, Santa Cruz

Gita Kapahi State Water Resources Control Board

Brian Anderson University of California, Santa Cruz

Rusty Fairey San Jose State University

John Newman University of California, Santa Cruz

Fred LaCaro State Water Resources Control Board

Max Puckett Department of Fish and Game

Mark Stephenson Department of Fish and Game

Craig J. Wilson State Water Resources Control Board

A P P E N D I X A


Scientific Planning and Review Committee

Briefing Document for the

Bay Protection and Toxic Cleanup Program


March 1995

A P P E N D I X B


Scientific Planning and Review Committee

Briefing Document for Recommendations

on the Bay Protection and Toxic Cleanup Program

Monitoring Activities


May 1996