NPDES Permits in New England
Outfall Monitoring Science Advisory Panel (OMSAP) Meeting
Tuesday, March 28, 2000
10:00 AM – 3:00 PM
Woods Hole, MA
Members Present: Andy Solow, WHOI (chair); Robert Beardsley, WHOI; Bob Kenney, URI; Scott Nixon, URI; Judy Pederson, MIT/Sea Grant; Mike Shiaris, UMass Boston; and Jim Shine, Harvard School of Public Health.
Observers: Don Anderson, WHOI; David Borkman, URI; Peter Borrelli, Center for Coastal Studies; Margaret Callanan, Cape Cod Commission; Phil Clapham, NMFS; Cathy Coniaris, OMSAP staff; David Dow, NMFS; Jim Fitzpatrick, HydroQual; Mike Hickey, MADMF; Carlton Hunt, Battelle; Russell Isaac, MADEP; Bruce Keafer, WHOI; Ken Keay, MWRA; Wendy Leo, MWRA; Matt Liebman, EPA; John Lipman, Cape Cod Commission; Steve Lipman, MADEP; Ron Manfredonia, EPA; Stormy Mayo, Center for Coastal Studies; Mike Mickelson, MWRA; Jack Pearce, Marine Pollution Bulletin; Virginia Renick, MWRA; Andrea Rex, MWRA; Jack Schwartz, MADMF; Rich Signell, USGS; Ted Smayda, URI; Dave Taylor, MWRA; Heather Trulli, Battelle; Steve Tucker, Cape Cod Commission/Bays Legal Fund; and Jeff Turner, UMass Dartmouth.
Summary prepared by C. Coniaris. Post-meeting comments are included in [brackets]. All such comments have been inserted for clarification only. They do not, nor are they intended to, suggest that such insertions were part of the live meeting components and have been expressly set-off so as to avoid such inference.
SUMMARY OF ACTION ITEMS & RECOMMENDATIONS
- February 23, 2000 minutes were approved as amended.
- OMSAP recommends that EPA and MADEP approve
MWRA's food web model scope of work as having fulfilled their
NPDES permit requirement. See page 11.
- OMSAP recommends that MWRA delete the current Alexandrium cell count threshold and would like to review the new paralytic shellfish toxicity threshold currently under development by D. Anderson. See page 17.
February 2000 Minutes
C. Coniaris pointed out a correction in the draft minutes. The phytoplankton bloom [Asterionellopsis glacialis] mentioned on page 4 occurred in the fall of 1993, not 1994. OMSAP approved of the minutes as amended.
Model Evaluation Group Update
B. Beardsley stated that the Bays Eutrophication Model Evaluation Group is currently in the process of completing their final report to OMSAP. One section by Don Harleman examines the effects of the transfer of effluent from Nut Island to Deer Island on nitrate concentrations in Boston Harbor. An informational report on issues surrounding the Chesapeake Bay Model review was distributed by email. The situation for the Chesapeake Bay effort is somewhat different, but there are some parallels. He hopes to have the MEG report finalized within the month, sent to OMSAP, and discussed at the next meeting.
Chesapeake Bay Model Evaluation
S. Nixon gave a brief summary of the recent Chesapeake Bay model evaluation, of which he was a reviewer. The evaluation was requested by the Scientific and Technical Advisory Committee for the Chesapeake Bay Project, but the Chesapeake Bay Project had no desire for this review, thus the evaluation was hampered by politics. The model group was unprepared for a review and the model had not been well documented since 1994. This made review by the panel difficult. The strategy of the model evaluation was to sample three of the most important basic areas of the model: primary production, vertical density structure, and nutrient cycling/respiration in the water column and the sediments.
The $2 million/year Chesapeake Bay model has been used in a large number of scenarios by the modeling committee (composed of management agencies). The group in Annapolis conducting the modeling for the Chesapeake Bay Project did not manipulate the model well and provide the Chesapeake Bay Model Evaluation Group (CBMEG) the requested diagnostic output.
In the view of the CBMEG, the primary production was seriously undercomputed, water column stratification was over-computed, water column respiration was underestimated, benthic oxygen uptake matched well, the basic relationship between nitrogen inputs and the fixation of organic carbon was not well captured, nutrient regeneration was not well-captured, and preferential uptake of the nitrogen by the phytoplankton was misrepresented. Model results showed that Chesapeake Bay was heterotrophic whereas observations suggest that it is either near production and respiration balance or autotrophic. There were also problems comparing observational and computed data due to the model grid and field sampling designs. These were serious enough concerns that CBMEG felt that it was inappropriate to use the model in making management decisions and were quite critical in their final report. The group did not think that the model was very credible. Another problem was that the Chesapeake Bay Project believed that if the model worked well for one parameter, there was no need to investigate further. For example, the dissolved oxygen (DO) seemed to be modeled well, however, the primary production was so low that there had to have been other compensating serious errors in the model in order to produce the correct DO.
R. Signell asked if there was a feeling from the review panel that these problems could be corrected in the model. S. Nixon replied that opinions varied, but overall, CBMEG thought the problems could potentially be fixed. The model was originally developed by HydroQual and is currently maintained by the US Army Corps of Engineers. At the June 2000 OMSAP meeting, Jim Fitzpatrick clarified that HydroQual was not directly involved in the development of the Chesapeake Bay Model. HydroQual only worked on the development of the Chesapeake Bay Sediment nutrient flux model. The term "model" refers to both the hydrodynamic and water quality models because they run coupled. CBMEG also recommended that the Chesapeake Bay Project modeling committee conducts additional reviews, bring in new people, and document the model better. Until now, the push was to run as many scenarios as possible. People have bought so much into the process that they cannot criticize it. He hopes that the Massachusetts Bays Eutrophication Model is open and does not "over-sell" itself. He also hopes that reviews of the model are done with doubt and skepticism and that the model is always examined with "fresh eyes". His sense is that the MWRA modeling is in much better shape than the Chesapeake Bay modeling.
R. Signell stated that there has been an effort in the BEM/HD models to describe strengths and weaknesses, as opposed to the Chesapeake Bay modeling. S. Nixon thinks that is an important difference. With the Chesapeake Bay model there seemed to be a fear of admitting to managers and the public the limitations of the model.
R. Isaac asked whether the present or redesigned field program could answer any questions in Chesapeake Bay. S. Nixon replied yes. One criticism from the CBMEG was that the monitoring and modeling programs were not designed to interact enough. For example, there had not been a systematic comparison of the model and monitoring primary production. R. Isaac wondered if there are any models with smaller margins of uncertainties. S. Nixon does not think that there are. Another problem with the Chesapeake Bay model was that no one could provide information on initial design requirements. The group would have liked to have seen a document describing modeling goals and abilities.
S. Nixon does not think we can model complex nature on 10-year time scales, e.g. if nitrogen decreases, what will happen in 10 years. A. Solow asked whether the MEG's findings for the Massachusetts Bays Eutrophication Model will be as negative as the one for Chesapeake Bay. B. Beardsley said no, however, there will be some recommendations for model revisions.
A Refresher on MWRA's Food Web Model Scope of Work (FWMSOW)
C. Hunt described the FWMSOW dated May 21, 1999 ["Scope of work for a food web model to characterize the seasonal abundance for important prey species of endangered species in Massachusetts and Cape Cod Bays" is available at: http://www.mwra.state.ma.us/harbor/enquad/pdf/99-09_enquad_report.pdf ]. The FWMSOW is an incremental stepwise scope of work that examines whether the environmental conditions in Massachusetts Bay will worsen as a result of the outfall relocation. He then summarized a recent report ["A review of issues related to the development of a food web model for important prey of endangered species in Massachusetts and Cape Cod Bays is available at: http://www.mwra.state.ma.us/harbor/enquad/pdf/99-14_enquad_report.pdf ]. The purpose of the report is to make an effort to understand where we are in the system, i.e. "step 1" in the FWMSOW. Two food web modelers were asked to describe their particular modeling approaches, including advantages and shortcomings. Peter Yodzis utilizes a trophodynamic approach. This evaluates dynamic interactions between and among the species in a food web and can be used to evaluate perturbations. Robert Ulanowicz utilizes network analysis, a more static approach that is not predictive, but more effective in understanding energy flow. Both of these approaches require an understanding of the functions and roles of all species relevant to the food web model. Each of these modeling approaches requires extensive data at the correct scale in order to be effective.
Based on these reviews, the development of a food web model that attempts to link the discharge to the occurrence of right whales in Massachusetts Bay will likely be an exercise in futility because: (1) food web models are most effective when addressing measurable perturbations in a system, and such perturbations are not expected to result from outfall relocation; (2) food web models must have complete and accurate species-by-species biomass information which is generally not available; (3) uncertainty in the overall importance of the Bays to the energetics of the whales (i.e. inability to close the food web domain); and (4) food web model development at a local or habitat specific scale is unwarranted given the importance of external factors that affect the distribution of whales. The MWRA NPDES permit indicates that the scope of work has been developed for the seasonal abundance of the important prey species. According to the permit, OMSAP is to provide MWRA with comments on the scope of work. Within 90 days MWRA is required to submit to EPA and MADEP a revised scope of work, after which EPA and MADEP will determine whether implementation of the food web is warranted.
M. Shiaris asked C. Hunt to comment on Ulanowicz' model and how well it could be used here. C. Hunt thinks we can model energetics and linkages if there is enough data. Ulanowicz probably cannot perturb a compartment and then determine how the whole system changes. His model functions with a certain steady state and is interpretive rather the predictive and it can be used to gain a better understanding of the system.
A. Solow said that one reason a food web model was requested to better integrate the monitoring data. OMSAP agreed that a detailed, predictive, numerical, and quantitative model was not possible at this time. However, there may be a qualitative model that can be used for integrating these data. C. Hunt thinks a qualitative model would be a good research tool to help us focus our questions and better understand the information differently but it could not be used to determine if there has been an impact from the outfall on the right whale in Massachusetts Bay.
J. Shine pointed out that the food web model scope of work flowchart includes "is such a change likely to harm whales?" In order to answer this question, there needs to be some qualitative idea of what might be harmful to whales. C. Hunt said that the assumption is that the prey of the whales would be adversely impacted. M. Mickelson added that this qualitative model is included in a report by Jack Kelly et al. with commentary from Bob Kenney that describes the "big picture" of right whales. K. Keay added that the presentations at the OMSAP workshop in September 1999 attempted to put the individual study results in a context of the food web using diagrams showing where the individual monitoring components fit into the ecosystem.
D. Dow thinks that it would be difficult to model Phaeocystis, even qualitatively. J. Fitzpatrick said that the Dutch are modeling Phaeocystis in the North Sea. The problem with steady state food web modeling, such as the Ulanowicz model is that Phaeocystis does not appear every year and that type of food web modeling would not be predictive and answer the necessary questions. T. Smayda thinks the Ulanowicz network type of modeling is not static. J. Fitzpatrick pointed out that the Ulanowicz model will would not be capable of describing when Phaeocystis will occur.
D. Borkman added that there seems to be a 3-4 year periodicity in the Phaeocystis occurrence in the Dutch work which closely matches two of three recent outbreaks here so what seems to be developing is a large-scale, almost ocean basin farfield influence. He asked how big a model would have to be to address this. J. Fitzpatrick replied that there may be an opportunity to take a look at expanding the BEM to examine this. HydroQual will most likely explore this further but he had reservation about certain food web approaches since they do not address some spatial questions. The BEM predicts that the influence of nitrogen from the outfall in Cape Cod Bay is ~5-10% and with these low percentages, he does not believe that the model coefficients and the biomass estimates will be accurate enough and thus will not give much confidence in the predictions.
S. Nixon believes the current modeling effort has proceeded in a deliberate and thoughtful way. OMSAP's job is to weigh the probabilities and the preponderance of evidence. It would be great to learn more about these questions, but there is no evidence suggesting that it is the burden of the MWRA ratepayers. There is also no indication that there will be any major change associated with the new outfall and the federal review process does not believe there will be an impact on the whales. The default "no" in the FWMSOW flowchart requires continued monitoring and evaluation. For now, the case has been made that there is no compelling evidence to require a numerical food web model. J. Shine added that zooplankton will continue to be monitored and MWRA will develop a better way of examining the zooplankton dynamics with the current data.
C. Hunt suggested people read the report by Cabell Davis on results of his late February 1999 Video Plankton Recorder survey ["Data Report for Video Plankton Recorder Cruise R/V Peter W. Anderson, February 23-28, 1999" is available at: http://www.mwra.state.ma.us/harbor/enquad/pdf/00-03_enquad_report.htm ]. He has high resolution information on a variety of phytoplankton and zooplankton distributions throughout the entire bays. C. Hunt also pointed out that spring 2000 rapid screening samples indicate that there is a Phaeocystis bloom in the nearfield area.
Barnstable County Science Advisory Panel's Alternative to MWRA's Food Web Model Scope of Work
S. Tucker stated that the Cape Cod Commission believes that examining the right whale in the context of trophic levels is advantageous for a number of reasons. The primary producers are the creatures most likely to show an impact from the discharge of effluent. It is known there is a direct pathway from this level of primary production to the apex predator in this system, in this case the North Atlantic right whale. Collecting data pertaining to this tropic relationship and analyzing that data in terms of the model about to be presented will help to clarify any effects of the effluent discharge on these populations. Right whales exist in very low numbers and the population is showing a reproductive decline.
These facts have led us to a conclusion that a precautionary management stance is warranted when considering alterations to the right whale habitat. This is not to say that we expect the negative to be proven. Work can be done to pre-emptively illustrate that there will be absolutely no negative effects resulting from the project. It is incumbent upon us to take advantage of the tools that are at our disposal to monitor the status of the bays system as it responds to the outfall, especially during its first months and years of its operation. Tools have already been developed that help decision makers identify the likelihood and intensity of the effects of the discharge. Though the BEM does not capture some anomalous events, it is a tremendous accomplishment. However, it is not a perfect tool, and there is a degree of uncertainty in applying its outcomes in a descriptive or a predictive manner. Hopefully it will become a sound tool and an important key element for providing context to future management decisions.
At the last OMSAP meeting, OMSAP and MWRA discussed removing the zooplankton threshold from the existing safeguards in the NPDES permit. While we appreciate that both OMSAP and MWRA recognize the importance of continued monitoring, we do not want to see the thresholds changed without the institution of a suitable substitute. After OMSAP accepted the MWRA assertion that the Acartia tonsa hypothesis was inadequate, the discussion reflected an interest in reconsidering the context in which the zooplankton data would be analyzed. At the end of the discussion, the panel's final recommendation was to recommend a shift to a system-wide approach for zooplankton analysis. Our proposed model began as a comprehensive and systemic effort but has been refined to focus on the most proximal, and arguably the most important parameters of that system, including zooplankton, and affecting right whales directly. We think that this food web model may be the right tool for the proper approach for zooplankton analysis recommended by the panel members as well as addressing our larger systemic concerns. We hope that OMSAP will agree that the further refinement and application of this proposed food web model will give us another tool to provide context for future management decisions.
S. Mayo does not believe that the MWRA monitoring will detect problems that may occur with right whales. The Bay has a profound influence on the feeding environment of the right whale, at least with respect to copepods and the caloric content available. He outlined why he believes a model that focuses on whales should be developed. It is clear to everyone that the right whale situation is beyond grave, probably nearly lost, since the reproductive rate appears to be crashing. It is not clear whether this is part of a long term cycle, regardless, the conservative assumption is that something is occurring to the species and he believes that it has something to do with feeding, e.g. low nutrition or pollutant uptake. The only other explanation would be a genetic problem. A food web model should be developed because the right whale population is in serious trouble. Cape Cod Bay should be included in the modeling effort since it has been and will continue to be an important feeding area. Cape Cod Bay is visited yearly by about 100 individuals (one-third of population), and this year seems to be no exception. Cape Cod Bay was designated a critical habitat for right whales by NMFS, due to the large proportion of the population, including calves, that visit the bay. He noted that the right whales reside in Cape Cod Bay over the winter, not just the spring (approximately December to April/May).
He agrees that there is a lot of uncertainty about the environment but it troubles him that we believe in the "no jeopardy" prediction and yet admit that we cannot develop a predictive food web model the system. Currently there are no strong indications that toxic effects are affecting the reproduction of right whales but there are some hints that there are reduced food resources and that could impact the reproduction of the animals. Ongoing work by Michael Moore and suggests that the blubber thickness of right whales is lowest in the North Atlantic population. This is significant since feeding is the main activity of right whales in Cape Cod Bay.
Based on research at the Center of Coastal Studies (CCS), we have identified a feeding threshold, i.e. the number of zooplankton at which right whales begin feeding. We originally calculated 3900 organisms per cubic meter, with a correction factor for the varied calorie content of different species, however, this calculation has been revised to about 3750 organisms per cubic meter. This threshold is based on 3000-4000 samples collected near feeding right whales. The mean density within the path of feeding right whales in Cape Cod Bay is around 26,700 organisms per cubic meter. Since we have researched the filtering abilities of right whales, these values are based on what the right whale can capture in its baleen. The maximum integrated density of zooplankton is about 250,000 organisms per cubic meter. He then showed calculations of caloric intake and a model of the density of a surface patch from southeastern Cape Cod Bay in 1999. We know that right whales congregate in certain areas where there seems to be a significant plankton resource and these areas should be carefully examined. He showed a distribution plot of right whales that corresponds to high chlorophyll values and patch distributions in the eastern and southern part of Cape Cod Bay. He considers Cape Cod Bay a very interesting system that is highly partitioned and non-uniform. However, there is a fair amount of information that a modeler could use to develop a food web model.
He does not understand what occurred in 1997 when Phaeocystis was very dense over a wide area, especially in the eastern part of the Cape Cod Bay. He has been misquoted as saying that the whales left because the Phaeocystis bloomed – he does not know if this is only a coincidence.
He showed results of their modeling in which caloric availability and capture in Cape Cod Bay was calculated. 0.69 kcal/m3 is the estimated caloric requirement for a skim-feeding right whale based on several assumptions. For most of the years, the whales are capturing 2.4 times this basal need for food, so if the calculations are correct, they are doing relatively well. He then showed different ways of looking at the maximum available and captured energy that varies on an annual basis. He has calculated 3750 organisms per cubic meter as the feeding threshold in the eastern two-thirds of Cape Cod Bay. He feels that there are not enough samples but results give an indication of what could be done from a modeling point of view.
M. Mickelson asked whether 1992 and 1997 were low zooplankton years, or whether no CCS data were available. S. Mayo replied that they are just now realizing that when the water column is mixed, the zooplankton resource may be found in either the upper or bottom few meters, however, in 1997 the zooplankton resource was low in the surface and the bottom.
S. Mayo showed his first attempt at estimating what would happen if there were changes in the taxonomy of the zooplankton in right whale feeding areas taking into consideration that the taxa are filtered differentially by the whales. The data are based on the typical mixed plankton resource, existing caloric density, and average density of zooplankton. Conceptually adjusting the zooplankton community relative to size either increased or decreased the capturability of the food per cubic meter of water. He then showed another model experiment involving a small taxonomic enrichment and a step-wise reduction of biomass to examine how sensitive the animals or their capture might be to a change in both taxonomic composition and biomass. These are examples of the types of model experiments that MWRA could conduct.
J. Fitzpatrick feels it is important to note that the area of habitat from year to year suitable for right whales varies by at least an order of magnitude. He does not see how MWRA, which currently exports ~90% of its effluent into Mass Bay is responsible, assuming a fairly constant effluent load, for all the variation. S. Mayo agreed, but thinks that assumptions need to be proven. There are plenty of examples in the news of undeveloped and untested hypotheses that have been proven wrong. It is important that we be particularly conservative about the eastern two-thirds of Cape Cod Bay when dealing with a very sensitive species on the brink of extinction.
T. Smayda pointed out that this is a very complex and nitrogen-sensitive system. He thinks S. Mayo has nicely shown much of the kinetics and data that are available. He is concerned that many of the model runs were shown for only April and October. MWRA argues that the constrained water mass below the pycnocline prevents nutrients from entering the productive pool – he believes this is not the case, and that photosynthesis does occur below the pycnocline. J. Fitzpatrick pointed out that the plots shown today were total nitrogen but that organic nitrogen, ammonia, and nitrate have also been modeled. Summer stratification is taken into account and built into the model framework. Chlorophyll was also modeled and there was little difference between pre- and post-outfall relocation.
T. Smayda noted that S. Mayo's data show that there are zooplankton thriving in bottom waters. If there is sub-pycnocline delivery of nutrients, harvestable into phytoplankton, available for grazing, and then dispersed downstream, then indeed there is carbon or trophic input into the system. In addition, Alexandrium tamarense in the coastal current is flowing past the vicinity of the diffusers and this species is capable of phototaxic movements and nutrient retrieval migrations. These nutrients can potentially restore the population and lead to added growth available for subsequent downstream transfer. BCSAP would like to see a quantitative approach such as a model and believe that the Ulanowicz approach is the appropriate one. This would be a better position for MWRA in the event of future litigation over whether or not a negative event is due to the outfall. We believe a model is necessary for MWRA, as well as ecosystem protection.
B. Beardsley asked T. Smayda about the timing of the influx of the harmful algae in the surface layer, and whether it could be modeled. T. Smayda replied that in the case of Phaeocystis, there is a Dutch model that could be modified for this area to address the timing of these blooms. However the problem with Phaeocystis is not the number of cells, but rather the size of the turbulence tolerant gelatinous colonies. When silicate is used up by the diatoms, Phaeocystis takes over and is very efficient at phosphate uptake. Thus nutrient ratios can possibly serve as a surrogate for temporal changes.
A. Solow listed the three parts to this discussion. Part one: what are the effects of the outfall relocation on the environment? BCSAP has raised questions about the belief that effects of the outfall will be negligible or absent in Cape Cod Bay and whether these issues are addressed by modeling. Part two: if the physical environment changes, what would be the biological effects, and would there be any effects on right whales? With this issue, the BCSAP is proposing some kind of a food web model to explore, in advance, what the potential effects of changes in the physical environment would be on the right whales. Part three: monitoring will absolutely continue, even if there is no observed effect of the relocation on the physical or biological environment. B. Beardsley pointed out that monitoring at the boundary could improve in order to be able to better tease out inter-annual variability as well as biological or physical effects that have nothing to do with the outfall. A. Solow agreed.
T. Smayda thinks that MWRA will never commit to retrofitting the treatment plant because they are not collecting the correct types of data to show changes in the environment. A. Solow then asked T. Smayda if he had a proposal on how to improve the monitoring program to satisfy these concerns. T. Smayda replied that the BCSAP has submitted proposals in the past for increasing frequency and sampling locations during critical periods (e.g. summer which is nitrogen sensitive). They have also recommended that samples at discrete depths be examined individually due to stratification and not averaged together. S. Mayo added that there is a lot of excellent data being collected but he has always wanted to see more sampling in the critical eastern part of Cape Cod Bay where the right whales feed. A. Solow thinks these issues are related to revising the monitoring program and not necessarily the development of a food web model. S. Mayo thinks that the monitoring data should be used in something like a food web model in order to try to answer some questions because there are some important hypotheses that need to be tested. Thus he thinks the food web model fits in with the monitoring in that it provides a use for the data.
J. Pederson thinks it is difficult to determine what data would be needed in order to develop even a simplified food web model that would give us a better understanding of the system. She does not think models can give us all of the answers and S. Nixon's discussion earlier today concerns those of us who try to balance the monitoring and the modeling and use the results to better understand how accurate our predictions are. She thinks there are many factors, not necessarily the outfall, driving the zooplankton that the right whales feed on. B. Kenney would like the answers to the same questions S. Mayo is asking. However, in order to develop a predictive food web model that can determine what the effects of any perturbation are going to be on the prey of right whales, it must be able to predict changes in species composition, size composition at a meter scale spatial distribution. In addition, we need to understand exactly how nutrients affect zooplankton. If we cannot make these links, then constructing the model in the first place is not possible.
S. Mayo pointed out that the Cape Cod Commission draft information briefing includes some ideas of how the BCSAP thought a model should look. He agreed with B. Kenney in that we are all searching for the same answers. However, he thinks the finding of "no jeopardy" lacks the kinds of necessary connections mentioned by B. Kenney. There some way to make those missing connections, for example, trying to determine whether changes in nutrients levels will have an influence on the resources of the right whale. Both cannot be true: predicting there will be no effects and stating that predictions cannot be made.
B. Beardsley asked whether these questions might be answered by examining the MWRA and CCS databases. S. Mayo thinks it is possible but there is a problem with variability. We only recently noticed that high surface zooplankton concentrations coincided with low concentrations in bottom waters and vice-versa. This finding could mean that there is lot less variability than previously thought. B. Beardsley pointed out that S. Mayo has made a strong statement for better monitoring in the eastern part of Cape Cod Bay, but that could be done without a food web model. S. Mayo thinks it would be important to have something like a food web model. He hopes that there can be some way of associating the existing database as well as future data that can support the "no jeopardy" determination.
B. Beardsley thinks that when discussing data results, one can keep in mind cause and effect relationships relating to how the ecosystem works, in effect, a conceptual food web model. He would like to separate the question of whether or not to develop a food web model and whether it is necessary to revisit the monitoring program. Of those two, he would emphasize revisiting the monitoring program. S. Mayo liked this idea of scientific review of the monitoring program that at the very least included a conceptual food web model in the heads of the scientists. A. Solow thinks the best qualitative "food web model" for right whales is actually S. Mayo. He cannot imagine using a Ulanowicz type model here. It would not tell us anything about the maximum density of the prey species, composition, or patchiness. It may be able to predict total biomass over a large area but that is most likely irrelevant to the feeding of right whales since they prefer the very dense patches.
T. Smayda wondered how MWRA would defend itself in court, without even a semi-quantitative model, if there ever was a large Phaeocystis bloom that appeared to be "fed" by nutrients from the new outfall, and this coincided with the right whales avoiding the area. M. Shiaris thinks this scenario would be a good argument for rethinking the monitoring but asked where a food web model would fit into this. T. Smayda agreed that improved monitoring is absolutely essential and would be a major step in the right direction. He thinks modelers such as Ulanowicz could be brought in to suggest how the monitoring could be improved so that a numerical model would not be needed.
J. Pederson noted that we will continue to review the monitoring to make sure it is answering the relevant questions. In fact, this process of reviewing trigger levels is a part of this process of constantly reviewing and refining the monitoring. However, we need to always keep in mind the type of program we are advising.
S. Nixon observed that everyone seems to agree that the state of the art is not such that we have a credible predictive model for linking things such as nutrient inputs to the changes that are most important to the whales. In addition, the models are not going to be constrained well enough to be an effective legal tool. Thus the responsible thing to do is to monitor the system carefully in the face of such uncertainty and the refinement of the monitoring program will be an ongoing process. OMSAP's job is to advise EPA and MADEP as to what MWRA is required to do based on the best available science. He is skeptical of models, as mentioned earlier, but feels that modelers are good at modeling physics. Problems develop when they try to model chemical partitioning and biological processes. Phaeocystis is perhaps linked with the nitrogen, but we do not know for sure, and a model would not give us that answer. Even S. Mayo feels that maybe there is a connection with the whales, maybe there is not, so we cannot put it into a model, and certainly not into a Ulanowicz model in a credible way at this point. It is also worth noting that J. Fitzpatrick, who is in the business of developing models, is saying that a predictive food web model as it relates to whales cannot be developed at the present time. No one is claiming for sure that there will be no negative impacts from the new outfall. Nature surprises us all the time, but that is why there is a monitoring program. All we can do at this point is monitor as responsibly as possible, and model what we believe we can model reasonably well. He does not see any compelling argument for a food web model relating to whales. However he does believe there is value in re-examining the monitoring and exercising due diligence and operating on the preponderance of evidence.
S. Mayo asked whether, since there is uncertainty about effects of the outfall, the statement of "no jeopardy" with respect to whales also contains a level of uncertainty. B. Kenney replied that the section 7 statutes require NMFS to make a decision based on the best scientific evidence available at the time. Model results indicated that relocating the outfall to the new location will have no measureable change in nutrients in the eastern half of Cape Cod Bay where the right whales feed, therefore, the "no jeopardy" conclusion. S. Mayo thinks it is negligent to say there will be no effect or damage with respect to right whales or the rest of the system; "no jeopardy" implies that the story with the right whales is closed (OMSAP members disagreed). B. Kenney replied that in the event new evidence arises, section 7 consultations may be reopened, as is currently occurring with the Biological Opinion of the fishery management plans. S. Mayo thinks that if a food web model is not considered, there should at least be a different approach when considering the eastern margin of Cape Cod Bay.
J. Pederson encouraged everyone to read some of the earlier Outfall Monitoring Task Force documents. We have been dealing with this issue for about nine years and the right whale was factored into even the early monitoring decisions. 250 people commented on the original draft monitoring plan. Some felt there should only be nearfield sampling near the diffusers while others felt it was more important to monitor further out in the farfield. The Outfall Monitoring Program was the compromise, with both nearfield and farfield stations, that has evolved into what you see today. It was intended to take into account, as best we could given the resources, variables that will give some indication of what is occurring in the system. The idea was to try to develop a warning system to alert us when things were changing negatively.
The Outfall Monitoring Task Force concurred with the decision of "no jeopardy", based on current available information and model results, that there will not likely be a change in nutrients and thus phytoplankton and zooplankton production, that would negatively affect the food of the marine mammals. However, monitoring is necessary to assure that this prediction is true. We also hosted a nutrient workshop where we discussed worse case scenarios and how the monitoring program can be set up to address them before they ever occur. D. Dow pointed out that because of uncertainty with the "no jeopardy" decision, NMFS attached several conservation recommendations, one of which of led to the development of the Contingency Plan and associated thresholds.
S. Nixon thinks it is important to note that the data are made available to the public. Also, S. Mayo can in fact be considered a kind of modeler because he is quantitatively linking numbers together, determining budgets, food requirements and food availability, and modeling computations.
A. Solow thinks that the value of a food web model broadly construed is that it would inform the monitoring program, i.e. where to look for problems and effects. It would be great if we had a good predictive model, but he does not believe we are anywhere near that point, so the best we can do is monitor. As J. Pederson mentioned, the qualitative understanding of this system was used to design the monitoring program. He would like to hear if anyone thinks that this qualitative model has changed, and thus the monitoring program needs to be changed. However, that is separate from any decision to accept the food web model scope of work. He would like to try to have OMSAP reach closure on this, but leave questions about the monitoring program open. J. Shine thinks the first step is to make sure the monitoring program is collecting the right information in order to be able to determine whether there is a compelling reason to develop a full food web model.
T. Smayda asked whether OMSAP would reconsider a proposed monitoring program that had been prepared a few years ago, with several testable hypotheses. The proposed monitoring program was designed to attempt to nullify the hypotheses. A. Solow said he would reconsider that plan. However, he is sensitive to constantly coming back and "reinventing the wheel". J. Pederson agreed that we should look at any hypotheses that are brought before us but if we have already done so she would like to review the previous responses of the OMTF or OMSAP.
ACTION: S. Nixon moved that OMSAP recommend that EPA and MADEP accept MWRA's food web model scope of work statement, as having fulfilled the permit requirement for the scope of work. B. Beardsley seconded the motion. All OMSAP members voted in favor of this motion. A. Solow then invited the BCSAP to return with the proposal about modifications to the monitoring program to address these hypotheses and issues of concern.
S. Nixon agreed with J. Pederson in that many people have already spent a lot of time working on this problem. Changes to the monitoring program should be based on scientifically compelling arguments. There needs to be some burden of professional responsibility on those who want to come forward with a change in monitoring program. J. Pederson agreed and suggested that OMSAP perhaps develop some guidelines about how concerned parties should bring recommendations to revisions to the Outfall Monitoring Program before OMSAP, e.g. expected benefits of revisions, and background information. A. Solow agreed that we need to be careful when revising the monitoring. B. Beardsley agreed but thinks that OMSAP should continually encourage outside input.
J. Pederson described the process of developing and utilizing a monitoring program: ask questions, gather information, develop monitoring, obtain and analyze results, and determine if the questions are being answered. This process encourages people to come forward if they think the monitoring program is not doing its job. She suggested OMSAP review the book "Managing Troubling Waters" that describes this process and may help us define how we look at proposals to changes in the monitoring program. B. Beardsley wants this to remain an open process so that people can come forward with useful recommendations. He wants to avoid the Chesapeake Bay situation where there is a disconnect between the research, modeling, regulating, and advocacy communities. We should encourage imaginative ways of modifying the Outfall Monitoring Program to better look at the environment, always considering good science. A. Solow agreed. J. Pederson agreed and added that not only should it be "good science" but also follow the goals of MWRA's Outfall Monitoring Program.
S. Nixon asked how accessible the monitoring data are. M. Mickelson said that all data through 1999 are available on CD-ROM in Access format. J. Pederson pointed out that you can also request data from MWRA for a specific area and they will send it to you within a week. S. Nixon asked if the data are on the web. M. Mickelson replied no. S. Nixon pointed out that the Chesapeake Bay monitoring data are not readily available and cumbersome to use. A. Solow thinks MWRA's data are easily available. W. Leo added that MWRA focuses on providing synthesis reports on the web since that is what the majority of the people are interested in. Data are usually requested by graduate students and MWRA fills these requests on a routine basis. S. Nixon believes the interaction with the research community is an important and healthy. There should be a process for T. Smayda, for example, to obtain the phytoplankton and nutrient data for several stations without having to filter through the entire dataset. A. Rex said that all he has to do is call MWRA and we will send him whatever he needs.
T. Smayda asked why the last modeling year was 1994. C. Hunt replied that the MEG in 1995 requested 1993 and 1994 be modeled because there were interesting features in the monitoring data. MWRA stopped the modeling so the MEG could have the chance to review the results and provide recommendations on how to proceed.
Review of Thresholds: Nuisance Algae and PSP Toxicity
M. Mickelson described the three nuisance algae thresholds [see MWRA information briefing]. MWRA recommends changes to the way they calculate the thresholds for Phaeocystis, Alexandrium, and Pseudo-nitzschia and requests that they focus on a particular time of year for Phaeocystis and Alexandrium and whether the current use of non-zero years in calculations is appropriate. He would also like OMSAP to review the log-transformed Pseudo-nitzschia and Alexandrium calculations. M. Mickelson then described the various types of harmful algae.
Phaeocystis pouchetii is a nuisance species that can form dense gelatinous colonies and occurs between mid February and the end of April. The last major Phaeocystis bloom was in 1997 in Cape Cod Bay. He showed data from Cabell Davis' (WHOI) 1997 Video Plankton Recorder survey. High nitrate appears to increase Phaeocystis abundance. However, the MWRA discharge does not contain high enough nitrogen to stimulate Phaeocystis growth throughout Massachusetts Bay. Phaeocystis bloomed in 1992, 1994, and 1997. S. Mayo said that whale visitation to this area was down during those years, and M. Mickelson thinks we should examine this more carefully by examining the zooplankton.
The toxic pennate diatom Pseudo-nitzschia does not show evident patterns in seasonality for this species. One problem is that it is difficult to distinguish between the two species of Pseudo-nitzschia, P. multiseries (toxic) and P. pungens (non-toxic), and requires the use of a scanning electron microscope or molecular probes. A shellfish bed closure is initiated at 20 micrograms of domoic acid per gram of shellfish. The ratio of the two species varies and presence of the toxic P. multiseries appears to be transient.
The toxic dinoflagellate Alexandrium tamarense is found from Maine to about New York City. D. Dow asked what the standard is for shellfish toxicity. D. Anderson replied that 80 micrograms in 100 grams of shellfish meat is the threshold at which an area will be closed (this is twice the detection limit). Shellfish become dangerous for human consumption at levels 5-10 higher than the threshold.
M. Mickelson then showed a map of station locations and showed where MWRA has increased plankton sampling based on OMSAP recommendations at the February 2000 meeting. 200 plankton samples are taken per year. He recommended that MWRA's threshold calculation be based on a narrower period of time during the spring. The thresholds for all three nuisance species state that "the baseline seasonal mean shall not to exceed 95th percentile". The existing threshold wording involves the nearfield, but MWRA's statistician determined that it was better to use all of the baseline nearfield and farfield data in the calculations. However, he is not certain whether the entire dataset should be used during post-discharge. Another question is what season to use. MWRA has been considering dividing in the year into three seasons as opposed to four because it would better follow the natural biological season. This would change the way the 95th percentile is calculated.
There are two different approaches when responding to threshold exceedances, "urgent" versus "deliberate". With an urgent approach, if one sample were found to be very high, further monitoring and notification would be triggered quickly. The deliberate approach continues monitoring to see if the system changes over several years. He asked OMSAP which approach was best for the nuisance algal species.
D. Anderson pointed out that the only significant differences between the MWRA versus WHOI sampling programs are station locations and sampling frequency. MWRA samples once in the farfield and in the nearfield during the potential Alexandrium interval and WHOI samples every couple of weeks, but only during the season when Alexandrium occurs within the Bay – typically April through June. MWRA and WHOI data seem to have some comparability, however, since D. Anderson focuses on only Alexandrium, he is more efficient, capturing less "zero" data. R. Signell thinks there is no clear relationship between the MWRA and WHOI datasets. MWRA samples throughout the year but does not capture the extent of blooms and WHOI samples April to June. D. Anderson replied that they do not sample during other times of the year because the cells are either absent or present in very low abundances. M. Hickey added that the WHOI program targets sampling when the abundance is expected to he high.
B. Beardsley asked why there is such a big difference in abundances between the WHOI and MWRA sampling. D. Anderson replied by emphasizing the differences in station locations, sampling timing, and frequency. He thinks the differences in sampling methods are negligible. M. Mickelson thinks it would be worth looking further into the comparability of the two studies. He then explained how the 95th percentile is calculated for Pseudo-nitzschia, Phaeocystis, and Alexandrium. Years in which these species are not measured ("zero" years) are not used in the 95th percentile calculations. A. Solow cautioned MWRA to be careful when calculating the 95th percentile. He then pointed out the apparent annual trend in the Pseudo-nitzschia data, and that there do not appear to be random annual values that fit the normal distribution. He suggested this pattern be monitored.
M. Liebman asked what the minimum number of samples are needed to calculate the 95th percentile. M. Mickelson said this relates to power or confidence intervals and he does not have that information available today. The 95th percentile is 1.6 million cells per liter for Phaeocystis, 37,000 cells/L for Pseudo-nitzschia, and 10 cells/L for Alexandrium. Using the WHOI Alexandrium data, the 95th percentile is 71 cells/L. M. Mickelson then briefly described the Maine and Massachusetts paralytic shellfish (PSP) monitoring programs. The MWRA permit requires that MWRA develop two outfall contingency simulations: chlorination failure and red tide bloom, and describe responses to both of those types of emergencies.
M. Mickelson asked OMSAP if MWRA can focus on the seasons when the Phaeocystis and Alexandrium blooms occur. Seasonality was recommended at the 1996 OMTF workshop because MWRA had presented annual averages of nuisance species. The experts considered that to be a "dilution" since there were parts of the year when the species did not occur. He then asked OMSAP if it was acceptable to use only non-zero years for Phaeocystis and Alexandrium.
B. Beardsley does not think that the MWRA monitoring is capturing Alexandrium as well as WHOI. Perhaps Alexandrium exists in small numbers throughout the year. A. Solow asked if MWRA is concerned that with their data, the threshold is calculated as 10 cells/L and with WHOI data, the threshold is 70 cells/L. J. Pederson pointed out that the problems we always have in setting Contingency Plan thresholds are dependent on the amount of monitoring and long-term data comparability. Thus, as long as monitoring remains consistent for a certain suite of parameters, even if results are of no value outside of the program, as seen here with Alexandrium, it is acceptable, as long as the program remains consistent. The Contingency Plan is designed to use information from the monitoring program in order to determine whether conditions have changed in the bay enough to require action. Thus it does not matter that the MWRA and WHOI thresholds are different because the programs are designed differently. B. Beardsley agreed but thinks there is value in trying to figure out why the two programs have such different results. J. Pederson responded that if we really wanted to monitor for only Alexandrium, we are probably measuring in the wrong place. But if we wanted to monitor for Alexandrium to see whether or not the new outfall makes a difference, then we are monitoring in the right place. M. Mickelson pointed out that the threshold is set to signal a change in the natural state of the system between pre- and post-discharge years. A. Solow thinks that even though the MWRA and WHOI datasets are somewhat different, they still detected the 1993 bloom, and that is very important. He feels J. Pederson's point is that as long as the interest is in whether there has been a change rather than absolute numbers of Alexandrium, and as long as methods and sampling remain constant, the results are acceptable.
J. Shine noted that there appears to be a lot of variability within the year and thus there could be problems using the annual mean and not accounting for the annual variability. In addition, if annual data are used to calculate thresholds, if there is a problem, it will take a whole year to notice. M. Mickelson replied that the Contingency Plan blends the "urgent" and "deliberate" threshold-response approaches. However, coordinating with WHOI and notifying MADMF is something that could be done from a single high sample. The response to a single high sample should not be the retrofitting of the treatment plant. J. Shine thinks that it would be an indication that something was wrong if one sample was found to be much higher than all samples previously collected. In addition, if we rely on annual means as benchmarks, spikes in data would be diluted out. There has to be some acute benchmark level that is not diluted out by using means and including zeros. B. Beardsley supported J. Shine's comments. M. Mickelson said that if OMSAP has any ideas for better approaches, MWRA would research them. D. Dow asked whether the data with a lot of zero years were examined using any other statistical distribution in order to figure out the confidence intervals (around the means) so the zeros could be incorporated. K. Keay replied the data were examined using other methods, but that there was no standard distribution to which the data could be fit.
M. Liebman recommended that since there currently is no warning threshold for Alexandrium, to add one based on human health, similar to the PSP data results. M. Mickelson cautioned that a warning level has implications for MWRA response that may involve the preparation of a new engineering design.
M. Liebman thinks that in the case of Alexandrium and Pseudo-nitzschia, there should be some relationship between the levels measured and an actual impairment, i.e. some type of appropriate risk-based number. M. Mickelson replied that D. Anderson is working on this type of threshold with Alexandrium and will discuss preliminary results.
D. Anderson described progress on developing a method of deriving a threshold based on shellfish toxicity in the bay, based on the 30-year MA Division of Marine Fisheries dataset. He believes that the Alexandrium population that causes toxicity from Maine to Massachusetts Bay is largely derived from populations from the western coast of Maine that are transported to the south in a coastal current formed by outflow from the Kennebec and Androscoggin Rivers. Most years there is toxicity in southern Maine and northern Massachusetts but cells seldom enter the bay in sufficient numbers to bloom substantially and cause toxicity in the bay. In a sense there is a filtered delivery, i.e. most of the cells do not enter the bay and instead move along Stellwagen Bank and out to sea, such that the pattern of toxicity in the bay is sporadic. Alexandrium cells are dividing perhaps once every two days during the time of year when they are in this area. If they were stimulated by nutrients from the outfall, any effects would be expected to be seen downstream because of the mesoscale circulation through the bay.
R. Signell pointed out that USGS has been measuring currents for several years and has found that there is no preferred direction during any particular season near the outfall. Water that does make it out further east eventually joins the prevailing current. D. Anderson said that water movement in the nearfield is complicated, but at least the Alexandrium data suggest that the cells move towards the south, and grow slowly as they proceed along the South Shore.
A. Solow asked whether Phaeocystis is transported in the same direction. D. Anderson replied that we do not know but would expect that it is coming from the north and that there is not a large in situ population within the bay. T. Smayda believes that both are true – there are introduced species as well as local cohorts. A. Solow thinks this is an argument for measuring both north and south of the outfall. D. Anderson agreed and reminded the group that MWRA recently added two new plankton stations north of the outfall.
D. Anderson showed shellfish monitoring data in micrograms (ug) of saxitoxin per 100 g of shellfish meat. There was an interval in the 1980's with relatively high and frequent toxicity in the Bay. Since the large bloom in 1993, there has been almost no toxicity in subsequent years. Overall, there is very large inter-annual variability measured at the state MADMF's shellfish monitoring stations. Shellfish are great integrators of what has occurred in the water column during the past week. Thus zeros values in shellfish are better indicators that there are very few cells, if any. A single discrete seawater sample, such as those collected in the MWRA cruises, is only an indication of what is there at that point and that time.
D. Anderson feels we should worry if there is a large bloom after the outfall goes on-line. There needs to be an assessment of pre-outfall, long-term variability. His group was funded by Sea Grant to work with WHOI's Marine Policy Center to suggest thresholds based on the MADMF PSP toxin monitoring data. He then showed the MADMF stations that have been sampled since the first large bloom in this region in 1972. The team is trying to see if there is a useful pattern in these toxicity events, e.g. in bloom progression or magnitude. Data from the 1993 bloom suggests that as the bloom moves from north to south, toxicity levels decline. The same decrease was seen in the WHOI Alexandrium cell count data from several years.
D. Anderson then described how they are working on developing a threshold based on PSP. He showed a data plot of PSP data from Cohasset, Scituate, Marshfield, Plymouth, and Sandwich. Using the baseline data, conditional probabilities of toxicity at each station can be calculated based on observations of toxicity at "upstream" stations. D. Anderson's team is attempting to identify spatial and temporal patterns during the 30 years of monitoring to determine whether the pattern of post-relocation toxicity is significantly different. Most of the effort thus far has been aimed at identifying the pattern, later they will focus on the magnitude.
One preliminary conclusion is that the occurrence of significant toxicity in the northern stations appears to be a predictor of the occurrence of toxicity in the south, but this is dependent upon the size of the upstream blooms. Small blooms in the north are not often seen within the bay. On the other hand, large blooms and high toxicity north of Cape Ann frequently results in toxicity within the bay. There is also an overall tendency for the southern blooms to occur later than the blooms in the north, consistent with the transport theory. This suggests that post-relocation occurrence of toxicity at the seven southern stations without a prior bloom at the northern stations would be indicative of an outfall effect since there is no record of toxicity within the Bay without a bloom further north. Preliminary analysis also suggests that the magnitude of these blooms decreases southward. If this pattern holds, then another indicator of an outfall effect could be high toxicity at a southern station and low toxicity at the northern stations.
D. Anderson cautioned that these promising results and inferences are still preliminary. He feels very uncomfortable with the current MWRA cell count threshold. The PSP threshold being developed is based on 30 years' worth of data as opposed to seven years of spotty plankton monitoring data, making it more statistically defensible. However, this approach is not possible for Phaeocystis and Pseudo-nitzschia. For Alexandrium, thresholds based on the monitoring data should be reevaluated.
J. Fitzpatrick asked if he thought that the MWRA sampling program for the Alexandrium is insufficient. D. Anderson thinks it is insufficient to characterize the blooms of Alexandrium in Massachusetts Bay and to reveal significant change. However he also thinks that his WHOI program also does not sample frequently enough during the spring due to budget constraints, and there will be years with no WHOI sampling at all, again due to funding. However, he thinks the MWRA Phaeocystis and Pseudo-nitzschia monitoring can be useful.
R. Isaac asked why only data from 1980 to 1999 were used. D. Anderson replied that this is still a work in progress. C. Hunt thinks this effort makes a lot of sense but asked what if there is a more localized input, e.g. local groundwater or the Plymouth discharge influencing Alexandrium.
D. Anderson believes that the history of the data may help with this, unless another new source begins providing nutrient input, e.g. a new outfall in Plymouth. We are looking for a statistical difference above the baseline. However, one benefit of using this data is that the baseline period contains 30 years. He pointed out that the cells are often more abundant along the South Shore and this may be related to the nutrients being flushed out of Boston Harbor.
D. Dow asked whether storms affect the occurrence of Alexandrium and PSP in inshore shellfish. D. Anderson replied that they have not examined this carefully in Massachusetts Bay. They have examined how storms affect the rivers in Maine and it is clear that toxicity is related to heavy springtime rainfall and snow melt. OMSAP should decide whether or not nuisance species thresholds should be based on cell counts. At least for Alexandrium, the alternative may be a PSP toxicity threshold.
A. Solow asked D. Anderson if he had any ideas on how the monitoring program could be better used to shed some light on some of these unanswered questions. D. Anderson thinks since we do not have the resources to sample as many stations as we would like, one alternative is to use the rapid analysis cell counts that provide results in just a few days. Based on those observations, we can establish a trigger or threshold for the different harmful species that leads to further actions. For example, if a threshold were triggered by the rapid cell counts, a targeted cruise could be mounted to examine what is occurring "upstream" as a way of making the monitoring program for the nuisance species more proactive. For this to be effective, the turn-around time has to be quite fast for the rapid cell counts that are done in the MWRA program.
J. Turner described the differences between the toxic (P. multiseries) and non-toxic (P. pungens) species of Pseudo-nitzschia. P. multiseries produces domoic acid which can cause brain damage. Both look the same under light microscopy and in order to differentiate, shellfish tissue needs to be analyzed for domoic acid using HPLC or electron microscopy to count the number of intercostal poriods on the inner surface of the valve. P. pungens' poriods are 1-2 per row and P. multiseries are 3-4 in a row. In addition, there are molecular probes being developed by Don Anderson, Steve Bates, and Chris Scholin. P. multiseries and P. pungens do co-occur and the unofficial Canadian threshold at which high levels of toxicity begins to occur in shellfish is half a million cells per liter (total Pseudo-nitzschia). Don Anderson pointed out that the New Zealand monitoring program uses a value of 100,000 Pseudo-nitzschia cells per liter as a threshold to initiate shellfish flesh testing. This is quite a low value, and would be exceeded on many occasions in the Bay. The reason New Zealand has adopted such a conservative approach is that their shellfish industry is large and thus vulnerable to toxicity outbreaks.
J. Turner recently analyzed August 1998 samples from Boston Harbor. Stations F24 had values at the fluorescence maximum of ~445,000 cells per liter and the highest value at F31 was ~650,000 cells per liter in the surface waters. He examined the Pseudo-nitzschia samples under scanning electron microscope (SEM) and P. multiseries was found in the samples. However, determining the ratio of P. multiseries to P. pungens using SEM is very time-consuming. M. Liebman asked which species of Pseudo-nitzschia dominate during blooms. J. Turner replied that P. multiseries dominated the Canadian bloom, however we cannot generalize this finding for other areas. R. Isaac asked whether it would be quicker to run HPLC than trying to count the poroids on a recently collected sample. J. Turner replied that the sampling protocol is not geared towards preparing samples for reliable and detectable HPLC measurements. A larger volume of sample would have to be screened in order to detect domoic acid using this method. M. Mickelson added that perhaps the molecular probes under development will be helpful in the future.
D. Dow asked whether there have been any studies on what nutrient ratios favor P. multiseries verses P. pungens. J. Turner replied that Steve Bates has found that starving P. multiseries for silicate will make them more toxic.
D. Anderson asked M. Hickey what MADMF does with respect to monitoring for amnesic shellfish poisoning (ASP) toxins. M. Hickey replied that they have monitored for ASP periodically as part of the phytoplankton monitoring program, but that it is not routine.
M. Mickelson pointed out that MWRA notifies MADMF once rapid samples are analyzed. He asked OMSAP for feedback on how MWRA is doing regarding nuisance algal species monitoring. For example, should D. Anderson continue to pursue the PSP threshold development, and should the Alexandrium cell count threshold be kept. A. Solow thinks the current Alexandrium cell count threshold should be deleted.
J. Schwartz stated that MADMF is not in a position to comment on changes to the Alexandrium threshold until they receive feedback from their shellfish team. M. Mickelson said that if there is a high level of Alexandrium detected, MWRA would still have to go through the procedures outlined in the outfall contingency simulation. He asked if MWRA may still proceed with this without an Alexandrium threshold. G. Renick replied that the simulation would have to be conducted in response to what the permit requires but there could certainly be a recommendation from OMSAP that the cell count-based threshold be removed. However, MWRA does not have a mechanism on how to be relieved from the permit requirement. J. Schwartz would like to know if the simulation and existence of a threshold could be decoupled. M. Mickelson thinks they may be decoupled.
ACTION: S. Nixon put forth a motion. He believes that there is convincing evidence that Alexandrium is extremely variable and patchy in terms of occurrence. It is unclear whether it is more important to document abundances in areas where it is found infrequently or where it has never been found. Given all of these uncertainties, and the fact that there is a better, more integrated measure, being developed, OMSAP's recommendation should be to drop the current Alexandrium cell count threshold and wait to evaluate the new PSP threshold being developed by D. Anderson that uses the long term shellfish monitoring. He does not believe that there is value in the current Alexandrium cell count threshold and moved that OMSAP recommend that MWRA delete this threshold. All OMSAP members voted in favor of this recommendation.
M. Mickelson asked how OMSAP felt about the Phaeocystis and Pseudo-nitzschia thresholds. He believes that MWRA is sampling Pseudo-nitzschia well and there is a pretty good understanding of its baseline distributions. He asked whether or not MWRA should continue to work towards this confirmation procedure at high levels.
J. Schwartz asked whether MWRA will recommend revisions, and re-visitation with the group that developed the draft plan, of the draft outfall contingency simulation plan that includes the Alexandrium threshold based on the discussions from today's meeting. M. Mickelson replied yes, when the improved toxicity threshold is developed, the draft plan should be revised.
A. Solow asked whether the simulation can be drafted, i.e. procedures in the event of an emergency, without having the threshold explicitly included. K. Keay replied that the permit requires MWRA to simulate the two emergency situations. During the development of the draft simulation plan, MWRA learned of the provisional nature of the PSP incidence threshold in the permit and thus made the decision to instead use the Alexandrium cell count concentration threshold. Any revisions to thresholds in the draft plan should require revisitation by the group that developed the draft plan.
M. Mickelson asked OMSAP whether MWRA should continue with the log-normal 95th percentile with the Pseudo-nitzschia threshold. A. Solow did not want to vote on this yet but felt that MWRA should keep on the same track and report to OMSAP again in the future. M. Mickelson asked OMSAP whether MWRA should continue calculating thresholds using the three-season year approach – the only difference is using the log normal. OMSAP thought this approach was appropriate.
J. Pederson pointed out that a recent Phaeocystis blooms began in Narragansett and Buzzards Bays and southeastern Cape Cod Bay and never proceeded further north. She wondered if this was factored in the calculations. C. Hunt replied that if a high count of Phaeocystis is measured, then MWRA will try to interpret where the bloom came from. T. Smayda added that Phaeocystis blooms on a regional scale, blooms occur synchronously over 300-400 km of coast. J. Pederson wondered how a threshold can be developed relating Phaeocystis to the outfall.
M. Mickelson asked if it would be possible to look at whether a bloom is locally higher than expected but on a regional basis. Since MWRA is not monitoring the entire region, the monitoring must be compared to baseline which leads to calculating the 95th percentile taking into account zero-observance years. A. Solow thinks that would not be difficult to do. M. Mickelson asked if MWRA can focus on the season when a nuisance species occurs. A. Solow asked whether an indication of impact could be observance of a bloom during a season when it is not normally observed. M. Mickelson replied that MWRA will work on this.
S. Nixon asked T. Smayda about the observance in the North Sea that one of the main indicators of change was an increase in duration of a Phaeocystis bloom. T. Smayda said that was correct and that the duration of blooms is increasing in this region. There appears to be a threshold with respect to the availability of the silica. Once the diatoms take up the silica, there is still excess nitrogen and phosphorus allowing Phaeocystis to out-compete other plankton species. Light may also be an important factor. M. Mickelson thinks MWRA should look further into this threshold and attempt to consider those things, perhaps examining seasons when Alexandrium is not normally present and it could be an indication of an outfall change if it becomes abundant during the other seasons.
D. Anderson said that the Phaeocystis threshold used in some parts of Europe is one million cells per liter based on two scientific papers. He suggested we poll our international colleagues to see what the levels of concern are in other parts of the world. S. Nixon thinks that would be very reasonable. J. Shine pointed out that would be a benchmark of ecological concern rather than a statistical threshold based on monitoring results.
T. Smayda wondered whether there is a corporate memory within the system that remembers changes to the Outfall Monitoring Program to recognize what is a refinement, an improvement, an enhancement, and what may be a dilution. The Cape Cod Commission believes that over time there has been a dilution of expectation and effort consistent with the objectives of the program. A. Solow replied Judy Pederson is OMSAP's "historian". C. Hunt added that there are a number of MWRA reports that have addressed concerns as well as OMTF/OMSAP minutes.
- OMSAP/IAAC/PIAC membership lists
- February 2000 OMSAP minutes
- Cape Cod Commission food web model information briefing
- MWRA nuisance algal bloom information briefing
- Copies of MWRA Presentation Transparencies