July 14-15, 2001 (Meeting between SEI, Science Panelists, and Agencies) Meeting Summary Meeting Minutes Meeting Agenda Attendees Meeting Summary SEI Science Panel Monitoring and Adaptive Management, July 14-15, 2001 Main Points Need for a: tightly, well-designed, responsive monitoring program as the core. institutional flexibility and the authority to enter into an adaptive management framework. some kind of mechanism to determine how information will be used in the decision-making framework. formal framework that is integrated throughout the formal establishment of an 'interdisciplinary team' that periodically reviews the information and develops a set of management recommendations, as appropriate. the creation of a scientifically credible document, but a legally defensible document. 'decision tree' to determine in advance how results will be used. Need to take chances on science, to try things, as part of an overall adaptive management approach. Some monitoring is already under way that could be integrated with the monitoring needs of the project. Basic research, though difficult to justify, can be targeted at some of the biggest uncertainties, thereby providing broader understanding. In terms of scale of a monitoring program, risks need to be balanced with predictive effects. Issues of high risk and high uncertainty are the first priority of things to monitor; issues of low risk and high uncertainty are a second-order priority. 12 criteria for selecting monitoring options that can be prioritized and tiered: related to salmonids ability to detect change; statistical power ability to detect project impact (direct and indirect) ability to provide feedback to management testing of hypotheses testing of assumptions relationship to jeopardy decisions cost effective contribution to new "knowledge" relevance to risk (relationship to the conceptual model) scale of monitoring effort and statistical power to risk Clarifying the questionnaire: Five years is the minimum amount of time for monitoring. Ideally, an adaptive management framework is pursued over the lifetime of the project Listed fish and shallow-water habitats, from the upstream to the downstream end of the project, are the main monitoring priorities. As the most sensitive areas, have the highest monitoring priority Five issues of concern for shallow-water peripheral habitat monitoring: monitoring of emergent vegetation fish use and growth (there's already existing research in this area} habitat opportunity and quality (tie in, of course with the CORIE program) wild fish or hatchery fish in the periphery food as a component of habitat Meeting Goals To discuss general concepts of monitoring and adaptive management To gain understanding of the relationship between risk and uncertainty and monitoring and adaptive management. To discuss the application of monitoring criteria to specific issues. To review responses of the three panel members to key questions on the survey. Goal Attainment The group spent quite a bit of time on general concepts. A Panelist provided Project Managers with a reference list for further reading. All parties acknowledged the need for monitoring and adaptive management; the challenge lies in working out the details. The group discussed the following risk/uncertainty matrix as a means of determining monitoring priorities.   Low Risk High Risk Low Uncertainty Lowest priority for monitoring Shouldn't do it High Uncertainty Moderate priority Highest priority for monitoring In general, the group did not want to focus on specifics, but much discussion concerned the regulatory agencies' priority for monitoring shallow-water riverine and estuarine areas, and the difficulty of monitoring in those areas for effects of the project. Twelve criteria for selecting monitoring options were established. The Corps pointed out that some monitoring of these areas is already under way, and that project monitoring could be integrated with these existing efforts. The three panel members present clarified some of their responses on the questionnaire. Project managers agreed to give SEI a list of their questions regarding further clarification from other panel members. Meeting Highlights Science Panel members, Steve Bartell, Don Boesch, and Martin Cody, were invited to participate in a special meeting with the Project Managers and others to discuss concepts of monitoring and adaptive management. General Monitoring and Adaptive Management Concepts Boesch emphasized a tightly, well-designed, responsive monitoring program as the core. Bartell emphasized the need for institutional flexibility and the authority to enter into an adaptive management framework. He also urged that there be some kind of mechanism to determine how information will be used in the decision-making framework. Larson expressed concern about the ability to connect changes in the estuary with the project, given all the activities, unrelated to dredging, that are going on in the estuary. Perry questioned whether a monitoring program could be designed that would be able to identify anything in the changing system that could be tied to the project Boesch responded that monitoring would not occur in 'a black box.' "You wouldn't just measure the salmon populations. You would monitor how does the deepening affect the properties -- water flow, depth -- all the ones we've talked about, as well as the dynamics of the salmon populations. And if you can't draw a line of evidence to suggest how that activity could've affected the salmon, or if there are other observations that suggest there are other factors affecting salmon, I think it begins to answer that question." Bartell saw the ultimate determinant as the likelihood of local extinction of listed fish stocks. Boesch acknowledged that some changes might take a long time to manifest themselves. He felt that the effects of channel deepening can probably be determined in the first five years. Other changes are more challenging, and may need to be integrated over time with yet other changes occurring in the system. He cautioned the group, however, not to wait 30 years, but to take chances, to try things, as part of an overall adaptive management approach. Larson stated that the Corps has already been doing monitoring that either fits into an existing adaptive management framework or that could be developed as part of an adaptive management framework. Hicks qualified that once the decision is made to go ahead with dredging, it could not be stopped unless some significant change were detected in the system that could be connected to the project. Bartell clarified that NMFS is willing to walk down the road of adaptive management, and that the Corps is similarly willing to go down that road, with the understanding that the feedback mechanism can only be under taken in the context of avoid and minimize, and mitigate. Boesch added that the group also needs to consider some sort of threshold -- a 'train brake.' Young said the train brake is called 'jeopardy.' Boesch further clarified that while planning frameworks are essential, it's impossible to anticipate everything (" We never thought we'd be tearing down dams..."). Under the current planning framework, the Corps is assuming that if the project gets the green light, it will do everything it can do to lessen the impact. NMFS is saying that, if at some point in the process it sees jeopardy, then its responsibility is to declare jeopardy and 'stop the presses.' Boesch didn't see the two positions are incompatible. He urged the development of an adaptive framework to address the Corps' concerns about maintaining the project once the green light is given and NMFS' concerns about the ultimate question of jeopardy. Developing a Framework for Addressing Uncertainty Boesch and Bartell advocated for a formal framework that is integrated throughout. "If you're really serious about having a monitoring program, and given the complexity of the system and all the time we've spent on this, you've got to build in a structure." Tortorici suggested the formal establishment of an 'interdisciplinary team' that periodically reviews the information and develops a set of management recommendations, as appropriate. She also stressed the importance of developing not only a scientifically credible document, but a legally defensible document. Hicks said any interdisciplinary effort should include all activities occurring in the river and estuary. Bartell agreed that wetlands restoration needs to be part of the picture. Boesch pointed out that cost would be a factor to consider. Boesch laid out some objectives the group might want to consider when designing a monitoring program: 1) trends 2) goals 3) resources (e.g., harvests) 4) precaution 5) models (develop information that we can apply) 6) research (context for testing hypotheses) 7) public information (most programs could do a better job) 8) emergencies. Research Mishaga expressed concern about the unpredictability of basic research. Boesch clarified that research as applied to this project is 'strategic applied research,' not basic research, which is curiosity driven. Nevertheless, he added that there is an important interplay between basic research and strategic applied research. Basic research provides important understandings, which in turn, help agencies manage the system more effectively. He agreed that group's focus is on monitoring for specific purposes, but with the possibility that there may be some strategic applied research that's necessary to help understand the monitoring results. Cody stated that there's a continuum between monitoring and research. On the short term, compliance monitoring will be essential. On the long term, research that gets to the heart of the biggest uncertainties is in his opinion more useful, but the most difficult to justify. Mishaga agreed, but said there has to be a separation between the whole continuum of monitoring and adaptive management and the effects of the project. Cody acknowledged that the biggest problem will be separating project effects from all other effects. Scope and Scale Courtney asked what should be the appropriate scope and scale is of the monitoring program. He pointed out that NMFS and FWS are interested in scaling the monitoring program to risks whereas the Corps prefers to scale it to predicted effects. Marsh recommended building the monitoring program on existing fish and channel dredging monitoring programs. Larson saw Corps fish monitoring as the environmental baseline. Boesch suggested the need for an inventory and assessment of ongoing estuarine monitoring programs. Risk and Uncertainty Boesch introduced the concept of the '2x2 contingency.' The first priority of things to monitor is where there is high risk and high uncertainty; second-order priority is with low risk and high uncertainty. He pointed out that many things associated with the project (e.g., toxicity, biological issues) fall into the latter category. The group established 12 criteria for selecting monitoring options: related to salmonids ability to detect change; statistical power ability to detect project impact (direct and indirect) ability to provide feedback to management testing of hypotheses testing of assumptions relationship to jeopardy decisions cost effective contribution to new "knowledge" (physical or other measurements that help interpretation) relevance to risk (relationship to the conceptual model) scale of monitoring effort and statistical power to risk Young suggested 'tiering' some, and Tortorici agreed, adding that "there's nothing at this point that I would take off the table. I think these are all essential elements for a proper monitoring program and an adaptive management scheme." Larson said the list seems adequate. "I wouldn't have any problem if all those were met." Bartell suggested putting 'testing of hypotheses' and 'testing of assumptions' into a second tier. Bartell cautioned the agencies to use as much statistical power as possible. "However, for some variables, particularly if they don't appear to be important in the overall equation for salmonids, you might be willing to be less certain about your results and correspondingly, devote fewer resources and worry less about statistical power. Another way to say that is you're willing to make a decision on a less powerful statistic." A long discussion ensued on cost effectiveness Tortorici expressed concern over designing a monitoring program to fit into a pre-determined dollar figure, as opposed to designing a monitoring program and then scaling it to fiscal realities. "I just want us to have a monitoring program that's based on priorities in terms of what we need to get answers. Then cost it out, and if we need to scale it up or down, okay, but at least let's have it based on some solid scientific underpinnings." Clarifying the Questionnaire Duration of Monitoring. Tortorici expressed concern over the adequacy of a five-year monitoring program, given the variability in the system. "From purely a logistical and technical standpoint, I just wonder whether in five years we're going to have enough information to be able to judge the impact -- any impact.... there are matters of practicality that could drive this to more than five years, and then if we're not seeing any change, whatever that change is, just cut it off." Cody agreed that there may be an adverse impact from the project that is extremely difficult to detect, especially in a five-year time frame. Each panel member then explained the reasoning behind his response. Boesch said he felt that a monitoring program could be designed to show the effects of channel deepening over a five-year period. But then, at the five-year point, people would presumably judge whether subsequent monitoring would be necessary. Cody similarly said five years, but he saw that as a minimum, the longevity of the fish and slow reaction times in the habitat. "I think most of the participants would be disappointed if, after five years, you could not identify some useful product from the monitoring. And if that useful product is forthcoming then there would likely be a decision to collect more of this useful stuff. But if you can't identify a useful product after five years, I think you should look somewhere else." Bartell was alone among panelists in selecting a 50-year response. He assumed that an adaptive management framework would be pursued over the lifetime of the project. "It may be that as you go along, collect, and evaluate the information, there are things that you might drop out, or de-emphasize, or re-focus the monitoring to make it as efficient and productive as possible." What to Monitor. Bartell said he was interested in monitoring fish because "they're in the high uncertainty, high risk box. That's the main justification, regardless of the fact that it's hard to do. If you're simply measuring the things you know how to measure, and that are economical to measure, and they don't provide any information into the decision process, to me that's a waste of money, unless you can establish a very convincing relationship between salinity and the fish variables." Cody added that fish were a main concern because the physical models "were fairly convincing" and that the potential impacts of the project on those factors "were pretty well understood." Boesch affirmed that it would be difficult to be against measuring fish "since this whole thing is about risk to fish." What to Measure About Fish. The consensus among the three panelists was for monitoring in shallow-water habitats, where the fish fall into the category of high risk and high uncertainty. Cody qualified his response by saying that he would concentrate on the major biological variables of which habitat and habitat quality have the potential to grow and sustain juvenile salmonids. Boesch said that what he had in mind was the wetland and emergent wetland islands and the associated shallow-water environments that Bottom et al. have identified as being particularly important to juvenile salmonids. Larson pointed out that shallow-water habitats are only important to two or three of the ESUs. The other ones, which are yearling fish, are probably not using them at all. "We keep talking about juvenile salmon as this giant category, but as far as peripheral areas go, it's only sub-yearling fall chinook, maybe some coho, and ...chum. We don't know about sea run cutthroat." Boesch pointed out that the shallow-water areas are the ones that are assumed to be more sensitive to changes. Larson concurred, and then asked if that meant that it's not worth monitoring for the other species, the yearling transients. Courtney clarified that it means that other areas are simply a lower priority. Larson said the Corps is already monitoring fish in peripheral areas. Boesch urged the Corps to build on depth and velocity as a measure of suitable habitat. Hatchery Fish. There was some question as to whether hatchery fish, wild fish, or both use peripheral areas. Boesch said that if only hatchery fish use the periphery, then it would be hard to claim that any impacts to shallow-water areas were affecting endangered units. The issue of hatchery fish is an important question, but it is larger than the project, and because of that, Boesch gave it a low priority on the questionnaire. In contrast, Bartell rated it high. "I was concerned that it would be critical to have information as to how the project might alter the ways in which hatchery and wild fish ... differentially use the system as a result of the perturbation. It seems to me that perhaps one of the contributions to the continued degradation of the listed stocks is the hatchery fish. If I were going to do the experiment, I'd turn off the hatcheries. Obviously you can't do that." Courtney summarized five issues of concern for shallow-water peripheral habitat monitoring: monitoring of emergent vegetation fish use and growth (there's already existing research in this area} habitat opportunity and quality (tie in, of course with the CORIE program) wild fish or hatchery fish in the periphery food as a component of habitat Hicks expressed a need for some sort of prioritization of the different factors as they relate to risk in the periphery. "When you look at the survey, most of the respondents believe that there will be a very low direct impact due to the proposed action. A very low physical impact to the channel, to the main navigation piece. So now we're looking at the periphery, and it becomes an order of magnitude more uncertain, in my mind, on how physical change in the 600-foot-wide channel is going to affect something in a four-mile cross section. We keep moving farther and farther away to the part that's harder and harder to measure. The risk becomes higher and the uncertainties are even higher. I don't get how you discern that, even if you had unlimited resources and unlimited time. If you told me the problem was with food sources, I'd say let's grow more organisms and forget the study. Let's plant more plants in the estuary and forget the research component." Cody took issue with growing more food and abandoning the research, viewing that as even more risky. "... if you start growing corophrium, for example, there could be an unanticipated side effect of the corophrium out-competing other food sources that are important later in the season... I think what we're suggesting, from what we know right now, monitoring the food source is a good idea in these shallow-water, peripheral areas." Courtney emphasized that the best professional judgment from the panel is the peripheral areas. Tortorici stated that NMFS' highest monitoring priority is in the peripheral areas. "My initial reaction is that we're in the ballpark of where we need to be. That doesn't mean that it wouldn't need to be thought through some more." Young added, that from his agency's perspective, he wanted to make sure that the context of these peripheral, shallow-water habitats is from the upstream to the downstream end of the project. "So we're talking riverine and estuarine habitat." Hicks wanted to re-visit the list of 12 monitoring priorities. She wanted to know what is the hypothesis to be tested in the group's expressed interest to monitor things like emergent vegetation, fish use, and habitat opportunity? "Are we testing whether the channel deepening will have a change to the emergent vegetation in the shallow-water habitat, fish use, etc. And go down like that? Is that the hypothesis that we're agreeing we're testing? And then go to point number 2 [fish use] and have a discussion on the ability to detect change, if that is indeed the hypothesis that we're trying to test." Courtney: It was suggested that the interdisciplinary team, talked about earlier in the morning, should develop a 'decision tree' to determine in advance how results will be used. Boesch and Cody concurred. Cody: "We [the panel] also did not subscribe to the idea of just going out and collecting information. We talked about duration, collecting data such that in five years time there should be a recognizable use of that data, and if not, then perhaps the situation should be re-evaluated. We're not collecting information that impinges upon delta-lambda. What should we collect instead? What would be a more reasonable approach? So this is something that is not an open-ended drain for your dollars, which we all want to avoid. But it has to be useful, it has to be contributory, it has to have assessment points along the way." Boesch talked about four different outcomes: There is no effect on salmonids and there's ample correlative information that you haven't missed it. There is no effect on salmonids and there's not ample correlative information to know that you've missed it. There's an effect, but there's a lot of correlative information that suggests that something else was going on. You have lots of other information which helps you interpret it. There's an effect and it's real. He acknowledged that all the players have risk. "You simply have to do the best jobs you can to minimize those [risks], but you can never eliminate them." Tortorici seconded that the challenge is to "make sure we've done the best job that we can to try to get to that end to the extent that we can... To do the best job that we can and to be flexible enough as the data start rolling in that if we see opportunities to modify it further, to get a better linkage to the project. decision tree..." The group concluded that the monitoring program would need to address issues of toxicity and bathymetry in peripheral areas.