Objections to Both the Majority and Minority Reports

of the Citizen Advisory Committee on Critical Areas Code Updates

by

Dr. Robert N. Crittenden

March 28, 2007

274 Sturdevant Rd. Sequim WA 98382

360 582-9550

I have the following objections to the majority and minority reports which make it impossible for me to support either of them:

1. The general approach proposed in both reports may reasonably be expected to result in the opposite of what they aim to achieve because they penalize those property owners who have protected their critical areas and reward those who have degraded or eliminated them. That may be expected to promote the further degradation or destruction of critical areas and to foster anti-environmental and anti-government sentiments among the public.

2. A buffer is a government-regulated area. That regulation deprives the property owner of his or her decision-power over that area. That is a penalty. Jail is another example of that same general kind of penalty, that is, the deprivation of personal decision-power. It is unjust and contrary to our heritage for people to be penalized when they have done nothing wrong; and it is even more unjust for them to be penalized when they have done the right thing rather than the wrong one. --- The proper approach is to penalize those individuals who do the wrong thing. For example, those who have deliberately polluted water bodies. We have laws to do that and there is a lack of evidence that such acts have been a significant problem in Jefferson County.

3. The definition of "wetlands" employed in their delineation in the processes laid out in both the majority and minority reports is so broad that few members of the public would recognize many of them as wetlands nor can large parts of those areas be assumed to have the same functions and values as marshes or bogs. Consequently, the buffers and other policies designed to protect wetlands, under the misapprehension that they are marshes or bogs, are in many cases inappropriate. Furthermore, those regulations will not be likely to be accepted by the public, as they do not appear to serve any legitimate purpose for those large areas of the "wetlands" that are not marshes or bogs.

4. Department of Ecology's new wetlands rating system is primarily based on two reports by Dr. Hruby et al. These are not scientific studies but reports of the results of two meetings convened for the purpose of obtaining support for the adoption of a new wetlands rating system based on some ideas that were, at that time, being promoted by the Federal Government. However, the attendants at the first of those meetings repeatedly rejected the idea that an overall score for the value of a wetland could be obtained by combining scores from the various functions. So, in order to achieve their predetermined outcome (that is, obtaining support for the acceptance of their new wetlands rating system) the department convened a second meeting for the purpose of obtaining support for combining the scores from the various functions. I presume that they did not invite those individuals who had objected strenuously to the combining of scores during the first meeting and that, had they failed to obtain their desired outcome in the second meeting, they would have held yet another meeting. Thus, by the evidence of their own supporting literature, their rating system was not supported by the scientific community. --- The majority report the the Citizen's Advisory Committee makes some modifications to the department's new rating system that may reduce this problem, by relating buffer widths to the scores for the specific functions, but the minority report uses the Department's new rating system without modification.

5. Because, the species are not identified for which the wetlands buffers seek to provide protection and habitat, it is not possible to demonstrate that those species are worthy of protection, need protection, that the buffers will provide habitat for them, or that the habitat thus provided can reasonably be expected to meet their needs.

6. In the case of the majority report, the habitat components of that plan from the Buffer Subcommittee were not accepted when that report was adopted by the full Citizens' Advisory Committee but, the process from the Fish and Wildlife Subcommittee's report for identifying species of local significance that are in need of protection and the protection that should be provided for them, was adopted, instead, to be incorporated into the buffer subcommittee's report, replacing its habitat components. This would have satisfied the previous objection. However, after adopting that process, that change seems to have been forgotten and their final report was produced without it.

7. The Forest Practices Act or elements from it has been adopted in several places in the majority report. However, the Timber, Fish ,Wildlife (TFW) negotiations from which it originated only had representation from the timber industry, environmental groups and the Tribes and only considered commercial forestry applications. Consequently, the Forest Practices Act is in many places inappropriate for other types of applications, including residential, agricultural, business, and so on... For these reasons, the Forest Practices Act should be adhered to only in those cases where the County is required to do so.

8. There are compelling reasons to believe that the salmon populations did not decline during the recent salmon crisis, during the 1990's, due to a lack of freshwater habitat. I wrote a book about this, in 1992, before the salmon crisis had occurred, saying that the Washington Department of Fish and Wildlife, in collaboration with the Tribes and leading environmental groups, was deliberately depressing the salmon stocks. --- I personally have no doubt that that was the case, as I was working as a consultant to State government at the time and spoke with several of the principles involved about this, attended some of the public meetings at which these policies were discussed and decided upon, spoke about this to other individuals within the departments who were involved in the implementation of these programs, and know departmental employees who quit the department on account of these policies. --- However, perhaps, the best evidence (because it should be apparent to a wide segment of the public) that the decline of Washington's salmon stocks was not caused by a lack of freshwater habitat, is that the salmon populations declined in degraded rivers as well as in pristine ones. --- Another one of the major causes of their decline was the reopening of the West Vancouver Island fishery, as the result of the collapse of the Canadian-American Salmon Interception Treaty, due to the South Alaska salmon fishery's refusal to stop intercepting Frazer River sockeye. When, I asked the manager of that fishery why she allowed it, she said that she had been told to do it by one particular prominent elected official from her State. So, that, too, was deliberate and political.

   There actually was a legitimate fisheries management reason from them to have done that, although, I do not know whether or not that was actually why they did what they did. That was that they needed more data from low abundance years in order to obtain statistically significant estimates of the parameters in the Ricker spawner-recruit curve. That is an important management tool and I had (Crittenden 1994) demonstrated the high cost of that lack of statistical significance but, also, that they only needed a few more years of data to obtain statistical significance.

   What they should not have done is deliberately perpetrate the hoax that the salmon crisis was caused by a lack of freshwater habitat and use that to leverage land-use regulation. Furthermore, this is still continuing, for, at the public hearing at the Jefferson County Library on their hatchery management plan, their representatives said that they were still holding the stocks in several rivers down for a few more years to obtain more data. I asked about the Samish and Dungeness Rivers and they said, "Yes. They were doing that in both of them." However, I did not ask about the streams and rivers in Jefferson County.

   I do not suggest that we should destroy salmonid habitat but it is well to be aware that the loss of freshwater habitat was not and is not the cause of the decline of the salmon stocks. Then, the County will be less likely to be misled into taking unnecessary actions for its protection or enhancement.

9. No-touch buffers are not only not necessary but can be counter-productive. The regulation of buffers should only prohibit activities that interfere with the functions of the buffer. Prohibiting all activities is unnecessary and prohibiting activities such as the management of woody growth is counter-productive.

10. Many of the studies cited in the Department of Ecology's review of the "best available science" (That is, Washington Wetlands, Vol 1 by Diane Sheldon et al.) are not scientifically valid because of methodological problems. Their most common weakness is that they failed to control the various environmental variables that affect buffer performance. Furthermore, many of them were government reports or other types of gray literature that were never peer reviewed.

    Another problem with their best available science is that the Department of Ecology funds studies on those issues which they know will be on their agenda in the near future and they tend to provide funding to scientists that generally agree with their viewpoint. Consequently, the studies that are available when the issues arise are a biased sample.

11. Nevertheless, a considerable preponderance of studies indicates that dissolved nutrients and suspended solids can be effectively removed by buffers which have a width of ten meters or less. In some cases, buffers as narrow as two meters have been shown to adequately provide that protection. Similarly, shading, cover, and insect drop can be effectively provided by narrow buffer strips. Nevertheless, as filtration and the removal of nutrients is one of the functions of wetlands, it can be questioned whether any buffer for the removal of dissolved nutrients is needed, except in extreme cases such as beside animal feed lots. However, some kinds of pesticides and herbicides are not adequately removed even by buffers of a thousand feet or more. They should be dealt with by other means, such as prohibiting their use within an appropriate distance of a water body or wetland.

    The great width of the proposed buffers, in the majority and minority reports, which in some cases are in excess of 100 feet, is primarily due to a desire to provide wildlife habitat. It has already been stated above, that the species in need of this and their needs have not been identified. The point that I wish to make here, is to question the right of the County to require private land owners to provide wildlife habitat. --- The private landowner not only has no obligation to provide habitat for the public's animals but could lawfully exterminate many species of them if he or she desired to do so. The preservation and fostering of wildlife is a general public benefit and the cost of providing that benefit should be born by the public in general not by one sub-group of them. Thus, there appears to be no justification for any buffers of more than ten meters in width. Nevertheless, if some species can be shown to justifiably need buffers for their protection, the procedure outlined in the Fish and Wildlife Subcommittee's report would allow for those buffers to be created.

12. Perhaps, it was unintended and simply a matter of viewpoint but both the majority and minority reports contain a certain amount of featherbedding. In particular, the majority report, was primarily prepared by a biologist who does wetland delineations and related studies for private parties and it requires, in many cases, that such studies be part of the permitting process or it allows them as options; the individual who represents commercial forestry and who has studied small forestry applications, has repeatedly tried to insert the Forest Practices Act into the majority report; the minority report was prepared by environmentalists and provides funding opportunities for their class of people; and the agriculture report was prepared largely by an employee of the Conservation District and another from WSU, and it suggests a heavy reliance on the services that those two government entities provide. --- There is entirely too much self-interest in the majority and minority reports.

13. Much of what I have said above, is that many of the proposed regulations in both the majority and minority reports are unreasonable. In addition, in many cases they are also arbitrary and unduly restrictive or hurtful to the land owner. Two studies by Stanford Professor P.G. Zimbardo showed that the arbitrary, unreasonable, unpredictable and cruel application of power caused deindividuation among those upon whom it was applied. That resulted in their depersonalization, emotionalization, group-membership, and irrational and/or anti-social behavior. There were also personality changes among about a third of those wielding the power, such that they come to devise creative ways to torment those who had the misfortune to be under their power.

    His first study dealt with these phenomena in an open population, whereas, his second study, which was the more definitive, dealt with them in a simulated prison setting. However, the difference between land-use regulation in Jefferson County and his mock-prison study are not all that great, as buffers are penalties that deprive the victim of decision-power. Although, that is less severe and is in an open setting, instead of a prison, that is the same kind of penalty as a prison sentence. In addition the volunteers who participated in his study were college students and, thus, were the same class of people as enter government. A short quote from his 1982 paper goes a long ways towards explaining what he found: "The counterpart of the mastery and control [exhibited by the guards] was the depression and hopelessness witnessed in the prisoners." --- We have heard in public testimony or have personally experienced these types of behaviors among government employees, involved in land-use regulation. So, we should not fail to recognize that these phenomena are applicable to land use regulation in this County.

Nevertheless, all that these two studies really tell us, is that psychologists have studied the mechanisms by which repressive regulation and/or its unjust application can lead to the breakdown of law and government and the development of a widespread sentiment of belligerence and defiance or, worse yet, depression and lethargy, among the public.

When we consider the risks and costs associated with regulation, these risks should not be overlooked, for they quite possibly have a greater impact upon the quality and enjoyment of life even than the financial burden that regulation unavoidably imposes upon the public. In my opinion, it is highly unlikely that the doubtful gains that may be obtained by the environmental protections that are suggested in the majority and minority reports will be greater than these risks and costs.

14. During the first meetings of the Citizens' Advisory Committee, I suggested that we should take the opportunity to confirm and support the widely-held view among the public of Jefferson County, in favor of what might be called "sustainable-living." In particular, I suggested that any land use, including not only agriculture but also residential use and small businesses, should be exempt from the critical areas code regulations if the property owner followed a few simple and easily-understood rules that would provide reasonable assurance that no pollution or other environmental damage would occur. Central among these rules was the requirement that no pesticides, herbicides, fertilizers, or other chemicals would be used in gardening or farming except those that were grown or originated on the parcel. There would have to be a few other rules, such as setbacks for compost heaps, driveways, and buildings, the provision of shade along water bodies and control of stormwater runoff. --- In fact, one could simply add such an exemption to the allowed uses of buffers in either the majority or minority reports. --- This would go a long ways towards making the critical areas ordinance livable for the vast majority of the public, by allowing them a way to escape government regulation, if they do no harm. It would also confirm and support a widely held ethic in this County and, quite possibly, might result in this County's becoming a World leader in sustainable living. Unfortunately, it soon became apparent that this proposal was inconsistent with the desires for featherbedding on the part of the various interest groups represented within the Citizen's Advisory Committee. They were all opposed to a non-regulatory approach, probably, because it did not provide them with the opportunities for funding, authority, or power that they desired. Thus, we appear to have missed an opportunity to take a significant step forwards.

15. The majority and minority reports appear to be two different interpretations of the Department of Ecology's recommendations: Specifically, option three in appendix 8C of Volume two of Washington Wetlands. These two reports differ mainly in details, one being more prescriptive (which is not entirely without merit, as the law is, then, fixed and known), whereas, the other allows more for creative solutions. It should be possible to combine many parts of these two reports into a single ordinance. Nevertheless, as most of my objections apply to elements which are common to both of them, I can not endorse either of them nor is it likely that I would be able to endorse a combined version.

    The only possible exception, would be, if, the outcome included a provision that would allow individuals to avoid regulation, if they followed a set of simple rules that will insure that they would do not harm. However, "no regulation" within that path, really means no regulation and anything short of that defeats the purpose.

16. Finally, I object to the suppression of minority opinion: The Citizen's Advisory Committee voted to allow minority reports but the Planning Commission has instructed them, near the end of the process, that such reports can only be accepted from groups within the Advisory Committee, not from individuals.

Literature Cited

Crittenden, R.N. 1992. Salmon at Risk, first edn. Hargrave Publishing, Carlsborg WA.

Crittenden, R.N. 1994. Optimum Escapement Computed using the Ricker Spawner Recruit Curve. Fisheries Research. 20: 215-227.

Sheldon, D. et al. 2004. Washington Wetlands. Vol. I. Dept. of Ecology, Lacey WA.

Zimbardo, P.G. 1970. "The human choice: Individuation, reason and order versus deindividuation, impulse and chaos. in W.J. Arnold and D. Levine (eds.) Nebraska Symposium on Motivation. 1969. Univ. of Nebraska Press, Lincoln, Nebraska.

Zimbardo, P.G. et al. 1982. "The psychology of imprisonment" in J.C. Brigham and L.S. Wrightsman (eds.) Contemporary Issues in Social Psychology Brooks/Cole Publ. Co. Monterey CA 395 p.

Testimony Submitted  to

the Jefferson County  Planning Commission,

Regarding Their Draft Critical Areas Ordinance 

by

Dr. Robert N. Crittenden

June 21, 2006

Crittenden Biometrical

274 Sturdevant Road, Sequim WA 98382

360 582-9550

Below I list several issues that need to be raised during the hearings on this ordinance, in order that they may be available to be examined by the courts.

1.      The restriction imposed by the Hearings Board that, only scientific studies and issues that were mentioned during the public hearings on an ordinance can be examined by them is unjust, as a member of the public will rarely know what issues or scientific studies may be relevant, particularly when it is necessary for them to have been identified before the issue to which they apply arose.

2.      I identify here broad categories of scientific literature in order to provide a body of scientific literature upon which they can draw. That includes all articles in respected peer-reviewed scientific journals.  These journals include by are not limited to the Canadian Journal of Fisheries and Aquatic Sciences, Fisheries Research, Ecological Modeling, Journal of the American Fisheries Society, North American Journal of Fisheries Research, Journal of the American Statistical Association, and Biometrics.

3.      I also identify the review entitled Pacific Salmon Life Histories, by Groot and Margolis. By citing that book, I also identify all the scientific works that they cite, therein.

4.      I also identify all articles and technical reports published by myself.

5.      That includes a scientific article on buffers that is still in progress. ---  Articles that have not been completed yet have been cited elsewhere for inclusion as Best Available Science. An example of this abuse can be found  in the City of Bainbridge Island’s best available science review for their proposed ordinance on  marine buffers.  This practice needs to be curtailed. However, so long at it hasn’t been, I might as well include this not-yet-completed article of my own. Taking a further step, I shall also include one on wetlands that I have not begun yet and may, in fact, never begin. This way, I open the opportunity for myself to write into it anything I may want, at some future date, when the need arises.

6.      Another issue that needs to be examined is whether the State’s definition of “Best Available Science” satisfies the due process requirement. The State’s definition was provided by the Department of Trade and Community Development and appears in the Washington Administrative Code. The problem is that that definition is a lower standard than the standard for scientific validity. As the scientific standard is the strict standard of reason and one of the due process requirements is that a law, rule or ordinance be reasonable, studies that meet State’s standards for Best Available Science but do not meet the scientific standard may not provide a rational basis for a law, rule or ordinance as required by the due process requirement.  The next two items identify points where the State’s standards are lower but this is not meant to imply that these are the only such points:

7.      Government reports and reports by consultants to their clients rarely receive independent review by peers of the scientific community. As they, thus, often do not meet the standards for scientific publication they should not be regarded as scientific publications, unless they can be demonstrated to have met those same standards.

8.      The standard for a “scientist” that is established in the State’s definition of Best Available Science is a lower standard than that for a “peer of the scientific community” for the purposes of providing peer review in a respected scientific journal and is also lower than that usually found for a scientist within the academic community.

9.      Jefferson County’s proposed buffers on wetlands appear to be based on option 3 in Appendic 8C in Volume II of Washington Wetlands. The key study on which that is based in a report by Hruby et al. in 2004. However, it is merely a modification of an earlier study by Hruby et al in 1999. That earlier study was a report done by a committee, which appears to have been conducted by the Delphi and Consensus processes. These processes provide a means for directing a committee’s conclusions to a pre-determined outcome. Therefore, the results presented in their report should be considered to be dubious.

10.   That committee repeatedly reached the conclusion that the scores provided by the various elements of their scoring of the quality of a wetland could not be combined to provide an overall score for that wetland. ---  Thus, although the first committee’s results may be dubious, because of the process, if they are to be regarded as being valid, we should at least respect their repeated conclusion that the individual scores should not be combined. Nevertheless, the purpose of the second report was to combine them to provide overall scores.

11.  Furthermore, the supporting scientific literature, provided in Volume I of Washington Wetlands, was published after the recommendations made in Volume II, that they support.  Thus, Volume I bears the appearance of  providing support for a predetermined outcome, instead of the outcome being based upon the science.

Testimony on King County's 2004

Proposed Critical Areas Ordinance.

by

Dr. Robert N. Crittenden

Crittenden Biometrical

274 Sturdevant Road, Sequim WA 98382

360 582-9550

This report was prepared for the Citizens' Alliance of Property Owners.

Permission to copy and distribute this report is granted to the Citizens' Alliance of Property Owners.

Executive Summary

The proposed ordinances on clearing, grading, and stormwater would remove from use more than half of the total land area of rural King County and restrict the use of the remaining area. Even if those restrictions were removed from the proposed ordinance its proposed buffers would still remove a substantial area. However, the studies upon which the clearing grading and stormwater ordinances are based lack scientific validity and fail to meet the standards for Best Available Science; likewise, relatively few studies on buffers are both scientifically valid and applicable to the Puget Sound Region. I, therefore, recommend against the adoption of these proposed ordinances.

A few of the defects in the supporting studies are listed below:

1. The overall context and scientific framework of the Best Available Science underlying the Critical Areas Code update is based on the principles of conservation biology. Those principles include two which show that conservation biology is not science:

   A) Conservation biology assumes complexity as an initial hypothesis, in violation of Occum's Razor and

   B) Studies in conservation biology are mission oriented, instead of being value-free.

2. A key study identified as providing the scientific basis for these ordinances is by Booth, in 2000. It is not directly applicable to current conditions in King County, because its data were collected from a watershed which did not have detention ponds.

3. The modeling study establishing its applicability is unavailable for examination.

4. That study was "modeling" but was presented as "expert opinion," thus avoiding the necessity of exposing it to review, but also making it and Booth's study, which rests upon it, fail the standards of Best Available Science.

5. The conclusions about the relative roles of deforestation and impervious surfaces are unsupported, because Booth was unable to separate the impacts of these two factors. He relied on the above mentioned modeling study and others to reach his conclusions. Unfortunately, their validity remains unknown because they are unavailable for examination.

6. Two earlier studies cited by Booth, in 2000, are also inapplicable, because they are based on the same data set from the watershed which did not have detention ponds.

7. Those two studies also did not adequately describe their modeling methods to be repeatable or to allow their evaluation. That component of their papers was a hydraulic model which they used for similar purposes as the model discussed above in items 3, 4, and 5.

8. Many of the studies identified as supporting the clear, grading and stormwater ordinances are weak in statistics and quantitative analysis. For this reason they fail to demonstrate that their conclusions achieve an acceptable error-rate and also fail to meet the standards for Best Available Science.

9. The suggested thresholds for clearing and impervious surface are based on observed correlation's, but correlation does not prove causation. Thus, the conclusions do not follow logically from the results, failing the standards for Best Available Science.

10. Although several thousand studies have been done on buffers, worldwide, most of them are inapplicable to the Puget Sound region, as its soils are of glacial origin and it has a temperate marine climate. That makes it relatively unique.

11. In addition, a surprisingly high proportion of buffer studies are not scientifically valid.

12. After the inapplicable or invalid studies are removed, few remain.

13. Many of the studies supporting the proposed buffers are weak in statistics and quantitative analysis.

14. The County's Best Available Science review cites several correlational studies supporting their buffers. But, correlation does not prove causation.

15. A related problem is that data on salmonid abundance is often drawn from periods during which they were declining for reasons not related to the availability of freshwater habitat. Consequently, based on data from these periods, significant positive or negative correlation's, significant regressions, and so on... can be found between salmonid abundance and urban growth . But, these results are misleading.

16. There is no consensus in the literature regarding single buffer widths for particular functions or to accommodate multiple functions.

17. The GMA does not mandate the restoration of habitat, only its protection and preservation. Thus, buffers are not necessarily required.

18. Agricultural pollution is probably most effectively controlled through the use of best management practices, specific to the particular application, instead of buffers.

19. Prohibiting a specific list of egregious agricultural practices, which pose an unusually high risk of pollution, has been remarkably effective in Skagit County.

20. I present the case for narrow buffers based on the reviews by Desbonette et al, Chris May, and DOE.

21. A US Judge ruled, in 2004, that it is better to restrict the use of chemicals near the water than to use buffers.

22. Contrary to the statement in the County's BAS review that, no studies show that no buffers are needed, I give examples of seven studies which show this.

23. The proposed ordinances do not allow for engineering or other alternative solutions.

24. The ordinances can not be justified by a desire to "save the salmon" as the salmon crisis was caused by ocean conditions not habitat. Also, some runs are current being kept at low numbers, by WDFW, for legitimate management reasons.

25. The proposed ordinances place a disproportionate burden on rural property owners.

Introduction

My testimony focuses on the proposed buffers and requirement that a minimum of 65% of the land area of parcels in rural King County be in native vegetation and a maximum of 10% be impervious surface.

The County's review of Best Available Science states that the primary purpose of these regulations is to protect the functions and values of critical areas.

Best Available Science:

Ordinances which affect critical areas are required to consider best available science (BAS), but that is not much different from the requirements which apply to ordinances controlling other areas, as the State's BAS requirement is essentially their interpretation of the Federal due process requirement that all laws, rules, ordinances and their applications be reasonable.

"Reason" and "science" are approximately equivalent, for that which is truly supported by reason will be scientifically valid. Thus, scientific validity becomes the central issue when we consider the studies which provide the basis for the County's proposed ordinances.

However, the State's uses Best Available Science, instead. The State's standards for BAS are published in the Washington Administrative Code, WAC 365-900 through 925 and are, also, summarized in Appendix A of Volume I of the County's Best Available Science document.

BAS has six criteria:

1. The study must have passed independent peer review;

2. Its methods must be valid and adequately described for the study to be repeated;

3. Its results must have passed appropriate statistical tests and quantitative analysis;

4. Its conclusions and inferences must follow logically from the results;

5. It must be in its proper context; and

6. It must cite the relevant scientific literature.

The standards for Best Available Science differ depending upon the type of study. The highest standards apply to research and modeling. They must pass all six of the criteria. These standards are approximately equivalent to the conventional guidelines for peer review provided by respected scientific journals.¹

The criterion for Best Available Science that appropriate statistical tests be conducted, establishes the traditional scientific error rate, of 0.05%, as the "acceptable error rate"² in the context of the due process requirement. Thus, when Best Available Science is required, passing the statistical tests demonstrates that, in the State's opinion, an acceptable error rate has been achieved .

But, the criteria for Best Available Science for other types of studies, than research and modeling, have lower standards. In particular, "expert opinion" is only required to meet the fourth, fifth, and sixth criteria. Therefore, for expert opinion, as well as for those other types of studies for which lower standards apply, "Best Available Science," does not necessarily guarantee scientific validity, nor does it guarantee that the study provides reasonable support or achieves an acceptable error-rate.

Fortunately, in most of the studies which are examined here are research or modeling and, therefore, BAS and scientific validity are approximately equivalent.

Overall Context and Scientific Framework:

The overall context and scientific framework of the County's review of Best Available Science are described in their Chapters 1 and 2. They rest upon the ideas of conservation biology enunciated by Noss and Cooperrider, in 1994,³ and Noss et al., in 1996.⁴ But, the principles of conservation biology include two which directly violate the long-established principles of scientific methodology:

1) Conservation Biology Violates Occum's Razor: Their principle that ecosystems are complicated beyond human understanding, and that their highly complex nature should be adopted as the working hypothesis, directly contradicts Occum's Razor, that the simplest hypothesis should be adopted as the working hypothesis until proven otherwise. The simplest hypothesis, is that simple direct effects should be accounted for first, before interactions and more complex processes are considered.

There is a large body of successful ecological research demonstrating the effectiveness of this traditional scientific approach, whereas only a few successes support the alternative approach taken in conservation biology.

The statement of this principle from conservation biology, can be found on page 330 of Noss and Cooperrider's book. It is a quotation from Frank Egler, in 1977: "Ecosystems are not only more complex than we think, but more complex than we can think."

The relevance of this principle to the body of Best Available Science underlying the proposed ordinances became clear, in 2000, when it was identified by Dr. Chris May. He was one of the reviewers of the County's report on Best Available Science, he was one of the contributors to the studies it is based on; and he received his doctorate from the department of Civil Engineering of the University of Washington, which is where most of the central studies cited in the County's review of Best Available Science originated.

Dr. May explicitly recognized this principle as one of the central principles of conservation biology, in his study of a system of proposed salmonid refuges for Kitsap County.⁵ That study was done concurrent with King County's Best Available Science document. However, his study for Kitsap County was rejected by their Planning Commission, partly, because it rested upon that principle.

2) Research in Conservation Biology is Mission-Oriented: A second principle which Noss and Cooperrider enunciated was that, "Conservation biology ... is not value-free science. Rather it is mission-oriented." That can be found on page 89 of their book. They go on to say that the first step is to set goals and, only afterwards, are studies conducted. They serve the purpose of achieving those goals.

The principle, that conservation biology is mission-oriented, directly recognizes the non-scientific nature of conservation biology. It should serve as a warning to carefully examine the studies presented in King County's review of Best Available Science.

It should also alert the County Commissioners to the possibility that the County's employees may have goals and a goal-setting process independent of the County Council. If that is the case, hopefully, the County Council will establish control over that process so that they will determine the goals which the County's ordinances serve.

3) Furthermore, In Conservation Biology, People are Regarded as Being Part of the Ecosystem: Conservation biology is an approach to ecosystem management. As people are included as part of the ecosystem,⁶ conservation biology includes the management of people. Dr. May's, 2000, Salmonid Refugia Report for Kitsap County illustrates how the principles of ecosystem management can be applied to molding public opinion and, thus, controlling peoples' behavior.

Validity and Applicability of Studies:

There appears to be at least 3000 to 4000 studies of buffers and their related issues, worldwide. However, the vast majority of these studies are either scientifically invalid or inapplicable to the Puget Sound Region. King County's review of Best Available Science did a good job in limiting the studies they cited to ones which are applicable, but some of the key studies have flawed science.

I examined about 100 studies, at random, from the documents which were assembled for the court case by the two sides for or against the narrow agricultural buffers originally required by Skagit County. More than half of those studies suffered from some problem in their scientific methodology. ---- The most prevalent flaw was probably their failure to control for other variables, such as slope or soil type, which would be expected to influence the effectiveness of the buffers. Probably the second most common mistake was their failure to do the statistical tests necessary to establish the scientific validity of their results. A few studies even did the statical tests , showed that their results were not significant but, nevertheless, drew conclusions from them. Good sampling designs which had adequate replicates, randomization, and controls were the exception rather than the rule and serious mistakes in statistical methods were not uncommon. The vast majority of these studies are "gray literature": That is, they were the reports of consulting companies, government agencies, or were published in minor or obscure journals, engineering journals,⁷ symposia, or as chapters in books. These types of publications often do not receive independent review by peers of the scientific community. Nevertheless, a few of the flawed studies were published in respected scientific journals.

After eliminating the studies which were not scientifically valid, almost all of those which remained were inapplicable to the Puget Sound Region. --- This area is relatively unique, as it has soils of glacial origin and it is in a littoral zone with a prevailing onshore wind, which gives it a temperate marine climate with a dry season in the Summer. The only similar regions may be Western Scotland, Southern Norway and Denmark.

This climatic zone is delimited on the East by the Cascades. The prevailing onshore wind keeps temperatures mild, the air humid, and also results in a dry period usually from August through October. This climatic region, may or may not be considered to include the Frazer Valley, as the cold air from the Canadian interior will flow down it, at times during most winters; but, the Northern limit of this region is certainly not further north than the mountains along the North side of that valley, as they divide the Arctic airmass of the Canadian Interior from the marine airmass of the Lower Canadian Mainland. Much of the coastal regions of Vancouver Island are included in this climatic region, but along the coast north of there, the conditions tend to be colder.

In addition, the Puget Sound Region has soils of glacial origin. The glaciers of the last Ice Age extended as far South as Olympia. They left behind them sediments mostly composed of either gravel and cobble or of the silts and clays which precipitate from glacial milk. All of these materials were very well washed by the ice-melt during the glacial periods. That removed much of their soluble minerals and changed their soil chemistry. The result is that this region has a type of soil in terms of porosity and chemistry which is only found in a band around the Earth, at approximately the limit of the glaciers of the Ice Age.

Thus, having both a glacial origin and a temperate marine climate, makes this region is fairly unique. It extends roughly from Olympia to Vancouver BC and from the Pacific Ocean to the crest of the Cascades. Studies done in Western Scotland, Southern Norway or Denmark may also be applicable to this region, but the applicability of other studies needs to be demonstrated before they are applied.

King County's Review of Best Available Science clearly tried to limit the studies it cites to ones which are from this region, when such studies were available. However, some of the key studies they cite are not scientifically valid. I point out some of these in the following sections.

Criticism of the Studies Supporting the Proposed Clearing, Grading, and Stormwater Ordinances:

Volume 1 of King County's review of Best Available Science contains an essay by Dr. Booth,⁸ which they say reviews the studies which support their clearing, grading, and stormwater ordinances.

That study and some of the studies it cites, fail to meet the standards for Best Available Science nor are they scientifically valid. A few of their defects are as follows:

1. Not directly applicable: Booth's 2000 article is based on a study of a watershed in which ponds for the detention and infusion of stormwater were not used, at that time. Consequently, a large part of the runoff from roads, parking lots, roofs and other impervious surfaces was probably conveyed, by storm drains, directly to the streams and rivers, where it made an immediate and substantial contribution to peak flows. As these practices are now avoided in King County, the author's conclusions about the relationship between urbanization and peak flows are not directly applicable to current conditions in King County.

2. The model establishing its applicability is unavailable for examination: He attempted to bridge this critical gap in his reasoning, by relying on a modeling study by David Harley, in 2000. However, that study was cited only as "written communication." Evidently, it was not a peer-reviewed published work. Nor, is it available for examination. Consequently, its validity remains unknown. For this reason, this critical gap in Booth's logical argument remains unbridged and his paper, fails to demonstrate that his conclusions follow logically from his results. For this reason, among others, Booth's paper fails to meet the standards both of Best Available Science and scientific validity.

3. That study was "modeling" but was presented as "expert opinion": In essence, Booth presented the conclusions of Hartley's research as "expert opinion," whereas, his study was research involving modeling. ... Expert opinion has the lowest standards in BAS; whereas modeling and research must meet all the criteria and, thus, have the highest standards. Thus, as it is presented, Hartley's study does not meet the standards for Best Available Science. It also clearly fails to establish its scientific validity. Furthermore, Booth's presentation of Hartley's conclusions did not even meet the criteria for "expert opinion," as only the conclusion of his study was given, without showing that it follows from the results, nor did he establish the context or cite references.

4. The result of the unpresented model appears unlikely: In Chris May's Ph.D. Dissertation, he states that approximately 65% of the impervious surface in rural King County is roads and parking lots. As roads and parking lots are often traditionally ditched to drain directly into streams and rivers, whereas other impervious surfaces, such as roofs or driveways rarely are, roads and parking lots must be contributing disproportionately to increasing peak flows in streams and rivers. They may account for roughly 80% or 90% of the problem. Therefore, the expected conclusion, which one would anticipate from that model, would be that the problem of increased peak flows in streams and rivers can not be resolved unless quick runoff from roads and parking lots is substantially reduced. However, they gave the result, that the problem of increased peak flows in streams and rivers can not be resolved unless the problem of quick runoff from other impervious surfaces is substantially reduced. However, as they may contribute only roughly 10% or 20% of the problem, it is doubtful whether one could even observe the beneficial effects of reducing that source of quick runoff.

5. The conclusions about the relative roles of deforestation and impervious surfaces are unsupported: Similarly, Booth's study recognized that both deforestation and the creation of impervious surfaces, are results of urbanization. He clearly showed their interrelationship in his figure 12. But, he was unable to separate the impacts of these two factors. Therefore, he relied on the modeling studies by David Hartley and others to reach his conclusions. Unfortunately, the validity of these studies remains unknown. This is another critical gap in logic, causing his study to fail the standards for Best Available Science.

6. Two earlier studies Booth cites are also inapplicable for the same reason: Booth's paper is largely a recapitulation of a symposium talk presented by Booth and Jackson , in 1994, and an article, in 1977, by those same authors.⁹ They are both subject to the first criticism, presented above, that as the watershed studied did not have detention ponds, their results are not applicable to current conditions in King County.

7. Those two studies also did not adequately describe their modeling methods: That talk and paper also had a modeling component, in which the authors examined the expected runoff at various levels of urban development. But, they did not describe that model adequately nor their parameter values and methods of analysis, to repeat the modeling exercise. Consequently, it is not possible to evaluate the validity of their results which were based on that model. Thus, their talk and paper fail several of the criteria of Best Available Science.

8. Observed correlation's do not prove causation: The County's review of Best Available Science, in speaking of the 10% threshold, recognizes that it came from a correlation found by Booth, in 2000, but they state that,¹⁰ " not all correlative studies are in agreement with these findings." However, instead of doubting the validity of those findings, they use that weakness as a justification for adding the 65% threshold on clearing. But, that threshold also came from the same observed correlation's from the same studies and suffer from the same weaknesses. Furthermore, those studies were weak on statistical tests and quantitative analysis. In addition, observed correlation's do not prove causation. The most common cause of this problem with observed correlation's is the so-called "underspecification bias" ¹¹ and Booth's study is seriously underspecified. Likewise, the study by They and May, in 1997, showing the impact of a 10% threshold on impervious surface, is inconclusive as it is based on observed correlation's.

9. At least one of the reviewers of the County's BAS document was not independent: In particular, Dr. Chris May is listed as a reviewer, but one of the studies in the section he reviewed was his own dissertation, while many of the other studies in that section were done by individuals from the same department at the University of Washington. --- Independent review is necessary to assure impartial evaluation, for, even if there is no deliberate bias, individuals who have worked or studied closely together are likely to share one narrow viewpoint. ---- The details of the histories of the other experts listed in the County's Best Available Science document, are unknown to me, but this one obviously inappropriate reviewer, raises a concern over the possibility that the County's documents may have had other non-independence of reviewers as well.

In conclusion, overall, their studies which provide the basis for the clearing, grading, and storm water ordinance, and which also pervade the rest of King County's CAO update, contain enough errors, gaps, and omissions that they do not meet the standards of Best Available Science nor are they scientifically valid.

Criticism of the Studies supporting the Proposed Buffers:

1. Chris May's dissertation is not applicable to King County due to a sampling problem: Dr. Chris May's dissertation¹² bears upon both the proposed buffers and the clearing, grading and stormwater ordinance. It involved a study of selected streams in Western Washington. However, because they were selected rather than being chosen at random, the results of his study apply only to the particular streams he selected, rather than to all streams in Western Washington. --- Had he selected them at random from some population of streams, for example, from all streams in King County, then, his results would have applied to that population of streams. But, he did not do that. Consequently, his results apply only to the specific streams he studied. --- The risk inherent to the methodology he used, is that his results may incorporate a bias due to the streams he selected not being representative of streams in King County. ..... I do not know whether this methodological error was done deliberately or through naivety, but preselection of the subjects of a study may provide a means for biasing the results, so that they become "mission oriented, " that is, so that they serve a predetermined goal, instead than being value-free science.

2. Another technical problem with Dr. May's dissertation: Dr. May used stepwise regression to develop relationships among the various variables he considered, but he said that he did all his tests at alpha=0.05. He should have decreased the alpha-level to compensate for multiple comparisons. What he did violates well-known standard methods for stepwise regression. The risk in what he did is that some of the relationships he identified may not be statistically significant, but may be due solely to random variation. It is, in fact, relatively likely that this occurred. Fortunately, the main contribution of his dissertation was not the relationships he developed, but in its extensive literature review, which undoubtedly helped to lay the intellectual foundation for the programs in the Center for Urban Water Resources Management, at the Department of Civil Engineering of the University of Washington. ---- That is a relatively new program which is entering a field which, until now, has traditionally been part of aquatic biology.

3. The studies are weak in statistics and quantitative analysis: It would appear that statistics is not one of Dr. May's strengths nor, judging from the various studies I have examined from the University of Washington's Department of Civil Engineering, is it one of their's, either, and the studies supporting the grading, clearing, and stormwater ordinances are mostly their work. --- Their studies are generally empirical or descriptive; they contain few statistical tests and often contain errors in statistical methodology; their models are simple and determinate or non-stochastic; and they do not calculate the propagated errors, to establish whether their predictions are meaningful, nor do they do sensitivity analysis to examine the impacts of structural changes in their models. --- The standards for Best Available Science, require studies which have research and modeling to have statistical tests and quantitative analysis. In their absence, it is difficult, at best, to ascertain whether their results have any predictive accuracy. This defect, alone, invalidates their studies, for without these statistical tests and analysis, there is no demonstration that they show anything at all. Thus, they not only fail the standards of Best Available Science, but they have also failed to demonstrate that their conclusions achieve on an acceptable error-rate.

4. Observed correlation's do not prove causation: The County's Best Available Science review cites several correlational studies supporting their buffers. A few examples are the study by Pess et al,, in 2002, which observed a correlation between Coho abundance and land use; May et al., in 1997, found a correlation between the coho/cutthroat ration and urbanization's impacts; and Moscript and Montgomery, in 1998, found a correlation between salmonid abundance and peak discharges.

5. Salmonid Abundance Declined for other Reasons: A related problem is that data on salmonid abundance is often drawn from periods during which they were declining for reasons not related to the availability of freshwater habitat. In particular, their two most recent periods of decline are now considered, by most scientists, to have been caused by ocean conditions and secondarily by hatchery problems.

Although the preservation of freshwater habitat and the maintenance of streams is a legitimate government objective, it should be remembered that, as Robert Lohn, the Director of the National Atmospheric and Oceanic Administration (NOAA), pointed out, "Most scientists, today, believe that the salmon crisis was caused by ocean conditions, not by habitat."

However, this conclusion should also occur to any good observer, as the size of the salmon runs, throughout the Northwest, declined and subsequently largely recovered not only in degraded streams, but also in pristine ones. ---- It is generally believed that a number of factors contributed to the decline of the salmon runs, but that stream habitat was only a minor contributor.

Thus, the decline in salmon abundance during this period was not caused by problems with terrestrial habitat, but these were also periods of urban growth. Consequently, based on these data from these periods, significant positive or negative correlation's, significant regressions, and so on... can be found between salmonid abundance and urban growth or between salmonid abundance and any other variable which is related to urban growth. Thus, salmonids provide entirely misleading data when they are used as ecological indicators. This objection eliminates many of the studies cited in section 7.2.3 of the County's Review of Best Available Science. In particular, the substance of its subsection entitled, "Salmonids as Ecological Indicators and Keystone Species", is cast into doubt.

1. No Consensus on Buffers: I generally agree with the statement on page 7-22 of the County's Best Available Science Review that, "There is no consensus in the literature regarding single buffer widths for particular functions or to accommodate multiple functions." --- Skagit County found between 800 and 900 studies supporting their narrow agricultural buffers, but the parties who challenged their ordinance found two or three times that number of studies supporting wide buffers. The Growth Management Hearings Board, however, found this evidence inconclusive. The problem, as discussed above, was that very few of the studies were both valid and applicable.

2. There is No Mandate for Buffers in the GMA: After two tries, Skagit County abandoned trying to set agricultural buffer widths. They then, required no buffers whatsoever, but established a sampling program to assure that there no further degradation of habitat occurred. Their position was that the GMA did not mandate the restoration of habitat, only its protection and preservation. This ordinance was challenged, but their ordinance and interpretation of the GMA was upheld by the Growth Management Hearings Board. Their ruling was appealed, but it was upheld by the Court of Appeals. Thus, this is now established in case law.

3. BMP's are a better way to control agricultural pollution: Agriculture is already required to follow best management practices. They are generally effective and have the advantage of being specific to the particular application. For that reason they will usually be more effective than a one-size fits all buffer.

4. Prohibiting a list of egregious agricultural practices is also effective: Skagit County also prohibits a list of specific egregious agricultural practices, such as feedlots draining into streams, dairy barns hosed out into streams, and so on... A planner from that County proudly told me that this has been far more effective than they expected and that, since they adopted this ordinance early this year, they have had only two or three complaints about agricultural pollution, whereas, they previously received many each month. --- This approach appears to have been very effective.

5. Fairly narrow grassy buffers remove most pollution from nutrients and sediments: Having looked at much of the literature, it appears that grass provides one of the best vegetated buffers for removing dissolved nutrients and suspended sediment. Many studies have show that a large proportion of these types of pollutants are removed by grassy buffers even as narrow as 6 to 12 feet in width. Sixteen to Thirty foot ( 5-10 meters) grassy buffers are most often recommended as the standard.

6. Buffers are ineffective at removing certain types of chemicals. Buffers of 300 feet (100 meters) or more do poorly at removing certain types of herbicides and pesticides. The use of these types of chemicals should be restricted instead of relying on buffers.

7. The case for narrow vegetated buffers drawn from the reviews presented by Desbonnet et al.,¹³ Chris May, and the Washington Department of Ecology's (DOE) draft report on buffers. These are some of the more widely cited reviews of buffers:

• 78% of the vegetated buffer widths cited by Desbonnet et al, support buffers of 100 feet or less for pollutant removal resulting from heavy agricultural use.

• Desbonnet shows (in his figure 8) that the first thirty feet are the most effective in pollutant removal in heavy agricultural uses, and that beyond 100 feet there is virtually no increase in function.

• Desbonnet concluded that, "In general, the greater than 50% removal standard can be met with vegetated buffers about 5 meters (16 feet) wide."

• In another place he stated that, "A multiple use vegetated buffer of 5 meters (16 feet) could be considered a reasonable minimum buffer-width standard"

• 92% of temperature studies examined by Desbonnet found adequate microclimate control with buffers less than 100 feet wide. Chris May reached the same conclusion.

• 62% of studies cited by Chris May for fine sediment removal had buffers of 100 feet or less.

• May also found that a buffer of 9 meters (29 feet) was 98% effective on an 11% slope.

• 88% of studies cited by Chris May for animal waste removal had buffers of 100 feet or less.

• May found that a 4.6 meter (15 feet) buffer was 98% effective at animal waste removal at an 7% slope.

• 52% of studies cited by Chris May for wildlife habitat had buffers of 100 feet or less.

• In DOE's synthesis, the studies cited show effective sediment removal with buffers of 100 feet or less.

• DOE's summary on sediment control was that 30 foot buffers gave 85% removal.

• DOE's synthesis found 85% removal of sediment in the first 30 feet of buffer.

• In DOE's synthesis, 88% of the studies they cite showed effective removal of nutrients from animal waste by buffers of 100 feet or less.

• In DOE's synthesis, 66% of the studies cited showed that fecal coliform was effectively removed by buffers from 13 to 100 feet in width.

• in DOE's synthesis, they found that a 16 foot vegetated buffer provided 50% or greater sediment and nutrient removal and that a 100 foot buffer provided only 70% or better removal of sediments and pollutants.

• The majority of studies cited by DOE show that buffers of 16 feet or less are very effective in buffering low intensity residential uses.

• DOE states, "Castelle and Johnson (2000) note that the apparent effectiveness of small buffers in removing toxics is due to the absorption of many toxics to sediment particles. When vegetated buffers are effective as filtering sediments, they will also be effective at filtering those toxics and nutrients adhered to them."

1. A US Judge ruled, in 2004, that it is better to restrict the use of chemicals near the water than to use buffers: In January 2004, US District Judge John Coughenour signed an order prohibiting the use of certain pesticides and herbicides within 64 feet of a stream or river which contained salmon listed as threatened or endangered. He had considered the Best Available Science, and had the benefit of expert witnesses from numerous State and Federal agencies and scientists from the public and private sectors. He concluded that it was not necessary to prohibit uses, such as agriculture and residential use, but only to ban chemicals within 64 feet of the water.

2. Some studies show that no buffers are needed: The County's Review of Best Available Science goes on to state that,"However, neither does the literature indicate that buffers are not needed." Some examples of studies which show that no buffers are needed, buffers are ineffective, or fish do better when their are no buffers, are:

• Bisson, Peter A. and James R. Sedell, 1984. "Salmonid populations in streams in clearcut vs. old-growth forests in Western Washington." In Meehan, William R. et al. eds. Fish and wildlife relationships in old-growth forests, proceedings of a symposium, April 1982. American Institute of Fishery Research Biologists.

• Hall, James D. and Richard L. Lantz., 1969. "Effects of logging on the habitat of coho salmon and cuthroat trout in coastal streams. In: Northcote, T.G. ed. Symposium on salmon and trout streams. H. R. MacMillan Lectures in Fisheries, a symposium held by the University of British Columbia, in 1968. Vancouver: Institute of Fisheries, UBC.

• Meehan, William R. 1996. Influence of riparian canopy on macro-invertebrate composition and food habits of juvenile salmonids in several Oregon streams. Portland: US Forest Service, PNW Research Paper 496.

• Rafael, Martin G. Peter A. Bisson, Larence L. C. Jones, and Alex D. Foster, 2002. Congruent Management of Multiple Resources, Proceedings from the Wood Compatibility Initiative Workshop. Two papers in PNW Station GTR 563, 2002

• O'Connell, M.A. J.G. Hallett, and S.D. West 2000. Effectiveness of riparian management zones in providing habitat for wildlife. University of Washington and Washington DNR Timber Fish and Wildlife Report. TFW-LWAGI-00-001 459 p.

• Reynolds, William Gene, and Heidi, 2003. State of the Lakes Report. Everett: Snohomish County Public Works Department.

• Ward, Fruce R., Donald J.F. McCubbing, and Patrick A Slaney, 2003. "Evaluation of the addition of inorganic nutrients and stream habitat structure in the Keogh River watershed for steelhead trout and coho salmon." In Stockner, John G. ed. Nutrients in salmonid ecosystems: sustaining production and biodiversity. Proceedings of the 2001 Nutrient Conference, Eugene. Bethesda American Fisheries Society.

In fairness, I must add that this is gray literature and some of it is from other regions. But, it does demonstrate that there is a literature supporting the idea that no buffers are needed.

Other Criticisms:

The ordinances do not allow engineering or other solutions: When an engineering or other alternative is possible in many cases. The County should embrace such alternatives to take the fullest advantage of the public's energy, creativity and ingenuity. Furthermore, they should go beyond merely passively allowing this, but sponsor the development of such solutions and they should also be pursuing some of them, themselves. Several obvious possibilities include the following:

• Rainbarrels or other suitable ways for storing water captured off impervious surfaces;

• Off-channel storage ponds, with their associated drainage ditches, curtain screens, or other structures to capture water from peak flows and convey it to the ponds;

• Flood by-pass channels;

• Organic or other land management methods which increase the soil's ability to hold water.

• The use of desirable or useful non-native plants which are good at retaining water in the soil, as an alternative to native vegetation;

• and so on.....

In the first two of the above items, the County should also be actively pressuring the legislature to change the water law to allow the Department of Ecology to issue the associated water rights, so that the captured stormwater can be put to a beneficial use.

The salmon crisis was not caused by the loss of freshwater habitat: As discussed above, most scientists agree that the salmon crisis was not caused by the degradation of terrestrial habitat.

The salmon crisis is now over and the abundance of most stocks has recovered to record run sizes.

But some stocks are being deliberately kept at low numbers for management purposes. ---- In the public hearing on Washington State's hatchery plan, in Hadlock, during 2004, the representatives of the Washington Department of Fisheries and the National Marine Fisheries stated that they were keeping selected salmon runs at low levels, until they had gathered enough data to obtain statistical significance in fitting on their spawner-recruit curves. That is an appropriate management decision which will be beneficial in the long-run, but the depressed condition of those runs should not be used to justify ordinances to protect the environment.

For these reasons, ordinances to preserve stream habitat can not be motivated, as being necessary to "save the salmon." King County's documents do not often invoke that justification, but in the few places they do, it should be removed as it reduces the credibility of their case.

The proposed ordinances would place a disproportionate burden upon the rural property owners. If an equal burden, of up to 65% of the property value, were placed on the urban property owners, a great deal might be accomplished towards alleviating the impacts of stormwater runoff from those areas and it should be remembered that the most severe impacts are occurring in the urban areas.

The County's argument for not applying the regulations equally is that: ¹⁴

King County's proposed thresholds however do not apply to urban areas, (nor can it because of historical development). As such , this may represent a departure from BAS in those watersheds and basins where such application is still possible and beneficial."

Their statement contains several errors:

• The thresholds only apply to new development. Therefore, if they were equally applied to urban areas they would only regulate new development.

• They would not apply to areas where no further development is possible;

• except that they would probably be later extended to regulate redevelopment, whenever there is a change in use;

• There are still many urban areas in King County which are not fully developed , where the ordinances would still be applicable;

• There are still streams in urban areas which are not severely impacted stream and need protection;

• The application of the ordinances to urban areas would move those areas towards "urban forest." --- As illustrated by Carmel, California, urban forest is achievable through land use regulations;

• The proposed ordinances for clearing, grading, and stormwater are not based on Best Available Science;

• Thus, application of the regulations to urban areas would be possible and beneficial (if they are applicable and beneficial in rural areas, which has yet to be demonstrated).

Undue Financial Burden: By preventing the use of 65% of the rural land in King County and placing additional restrictions on the use of the remaining 35%, the proposed ordinances place an undue financial burden on the rural property owners.

Conclusion

The proposed buffers and the clearing, grading, and stormwater ordinances would remove from use more than half of the total land area of rural King County and restrict the use of the remaining area, but many of the key studies upon which they are based lack scientific validity and fail to meet the standards for Best Available Science. I, therefore, recommend against their adoption.

However, I have, at various places in this document made suggestions of what I think will be better or more effective approaches.

Dr. Robert N. Crittenden

October 17, 2004

Address to the Kitsap County Commissioners,

Monday Nov 5, 2001

by

Dr. Robert N. Crittenden

 I was asked by the Kitsap Alliance to review the County's Refugia Study. --- I will very briefly give you the highlights.

The first thing which I wish to point out is that the Refugia Study is not science, nor is it based on science. In fact, the authors, themselves, admit this when they state, in their page 4, that the principles upon which their study is based are not testable. --- Testability is essential to science, a hypothesis which is not testable can not be used in the scientific process. Thus, it is clear that the principles upon which their study is based are not scientific.

Nor, in fact, do those principles come from a scientific publication. They come from a book by Noss and Copperrider. That book is popular literature. It lays out the principles of environmentalism, as its authors see them. And that is what the Refugia Study is based on.

Let me give you another example of how the study is not science.... You will find it develops on the theme, that the environment is not only complex, but too complex for people to even comprehend. --- That flies in the face of science. Every successful scientific study in ecology or environmental science demonstrates that that is not true.

This viewpoint, which the authors of the Refugia Study have adopted, that ecology is incomprehensible complex, directly contradicts one of the central principles of science. That is that one should account for what is simple and obvious first, before postulating complications. The name of that principle is familiar to you, it is "Occam's Razor." That has been one of the central principles of science for over half a millennium. --- But, the authors of the Refugia Study adopt its antithesis as one of their central themes.

Do I need to say any more. The Refugia Study is not science! And once you realize this, you will find many examples of it throughout their study.

The second point which I wish to make is that in the middle of their study, they discuss many technical issues and make many technical errors. You will find some of them discussed in my review.

I will skip, instead, right to the bottom line. The bottom line of this study are its results. The results of the Refugia Study are their rankings of the various refuges which they have proposed. What they gave you rankings but what they did not give you were their variances or confidence limits. As a result they have told you nothing

Let me explain by giving an example..... Suppose that you wanted to know how many Chinook there are in the Dewato River. You might go to WDFW, and they might tell you that their estimate is 324 fish. --- I don't know what their estimates are, this is just a hypothetical example. --- Now you realize that their estimate is just an estimate, that is that there are not precisely 324 fish, but you suppose that their "324" means there are approximately that many, perhaps somewhere in the range of 300 to 350 fish. But, that would be your assumption, not what they told you. They did not tell you anything about their level of precision. Their confidence limits might, infact, be between 0 and 10,000 fish, which would be equivalent that their having said that they really haven't the vaguest idea how many fish there might be.

WDFW pulls this trick, all the time. It is something you need to be aware of.

But, they aren't the only people who do this: The authors of the Refugia Study did it, too. In particular: they gave you rankings of the refuges, but they did not give you any measure of the precision of their estimates. So, in fact, they told you nothing!

But, the authors have been honest about this: In several places in their study they told you that their estimates were imprecise and they, also, told you that the principles on which their study was based are not testable. It is not their fault if you were not aware of the full meaning of what they told you.

Let me conclude what I have told you, here, today, I have said three things:

• The study is not science;

• It has many technical errors;

• and, in the end, it tells you nothing.

•  

I hope that this will helpful to you and I thank you for listening to my comments.

Review of Dr. May's 2000 Kitsap Penninsula Salmonid Refugia Plan

Introduction

The background and theory sections of the Refugia Study contain a surprisingly high level of sophistication in their understanding of the biology and genetic structure and of salmon populations. This provides much of the platform from which the study commences. But,  the authors, take several unsupported steps and, also, progressively change their focus, until they reach, in the end,  a conclusion that is scientifically unsupportable and  far removed from the study's original purpose.

In particular, the study begins, in its executive summary, with a description of the Endangered Species Act (ESA) listing of salmon species in the Puget Sound Region and of how, by complying with its  4(d) rule, the County can avoid the onerous consequences of its "take" provisions.  The  Refugia Study is initially presented  as part of the County's response to those listings and as "an integral part of the recovery process." It is intended to identify high priority areas for preservation, "the last remaining areas of high quality salmonid habitat on the Kitsap Peninsula." But, it concludes by recommending, as its highest priority, the preservation of  a refuge which has the potential to support only a small population of one of the listed species and is more than half inside an Indian Reservation and is, therefore, not subject to the ESA nor to regulation by the refuge, as the refuges will be created by the County and, thus, will derive their authority from them.

They view the purpose of  refuges not  as being for the preservation and enhancement of the listed species but, as they state on their page 9,  as being the "optimum strategy" for "managing current and future human activities."  

They  identify  "adaptive management" as the process through which  these changes will occur. --- The  refuges' function is to create a regime under which  local councils will use this process, with a view towards conservation, thereby, modifying their own viewpoints and behavior and, also, implementing management for conservation on a sub-watershed or community level.

The Refugia Study initially presents itself as science or science-based, but,  it is not. Science merely provides  the platform from which they launch a political program. 

In fact, the authors, themselves, admit the study's  lack of firm scientific support when they state on  their page 4,  that although they feel that the concepts underlying their study are "consistent with generally accepted scientific theory," they are "not testable in a practical sense."

Testability is the essence of science: A theory which is not testable can not be used in the scientific process to advance  understanding. Thus, the concepts upon which the Refugia Study is based, being  untestable theories, are not part of science. Nor, does the study attempt to advance scientific understanding. What it does is use or misuse selected ideas from science for other purposes.  

Once you have gotten, thus,  beyond the  illusion that the study is science or science-based, its mask falls away and you begin to see its true nature. ---  It is  a statement and implementation  of the political doctrine associated with the salmon recovery effort.

The authors' summarize that viewpoint  in the concluding sentences of  their introduction/theory section. In particular, they say, on their page 9, that the approach of  "ecosystem management"  should also be considered. But, that is an understatement as it appears to be the approach they adopted. They describe it  as follows:

This strategy involves managing ecosystems for the protection of native biodiversity and assumes  that the protection of ecosystem structure and function will result in the protection of all species including those already listed as threatened as endangered. Under this strategy, humans are recognized as a integral part of the ecosystem and the success of conservation efforts is highly dependent on an understanding of human influences on ecological processes and landscape patterns. Therefore, ecosystem management on the scale of refugia is not just scientific theory; rather it is an integrated program of scientific, socio-political, cultural, and economic values designed to achieve long-term ecosystem conservation.

Under this approach, it is not strictly necessary that a refuge even provide habitat for the listed species, as its primary  impact is intended to be on human behavior.  Should that be effective,  its impact may be far more widespread  than merely the strict confines of the refuge.

 This helps to explain why the author's would recommend, as their highest priority, the creation of  a refuge which s unlikely to ever provide significant habitat to either of the listed species.

Biology

The logical development of their position is not always explicit, but the major steps in their biological argument appear to include the  following points:

• There is a salmon crisis;

• It is due to habitat loss,  largely  caused by rapid urbanization;

• The problem is that although salmon adapt to a ever-changing environment, the rate of change and frequency of disturbances has exceeded their ability to adapt;

• Refuges can provide an important component of the restoration program by protecting core populations which can later become foci for colonization or re-colonization;

• These refuges need to be on a landscape scale;

• These refuges should, ideally be large, complex, and distributed over the region.

• Technological solutions rarely work;

• There is much about the ecosystem that we do not know and can never know;

• But, if we return the environment to a condition which emulates nature, in structure and function, the salmon will probably prosper;

The term "salmon crisis," usually refers to the widespread decline in salmon abundance throughout the Northwest States, beginning in the late 1970's and reaching a low-point in the mid-1990's. Enough of the overall pattern in salmon abundances can be explained by ocean harvest, to provide a brief chronology of it.  

Immediately following the Boldt I decision, the Tribes began taking their half of the salmon harvest. But, due to the Washington Department of Fish and Wildlife's having  artificially inflated their return estimates in the years immediately prior to the decision, the Indians actually appear to have taken approximately 80% of the allowable catch, although through no fault of their own.   The resulting heavy over fishing caused the runs to decline sharply. During this early period, also, a number of mistakes were made by parties who were new to harvest management.  But, after a few years, these problems were sorted out and a  workable scheme for co-management was established. 

But, the Chinook stocks continued to decline. A National Marine Fisheries Service Study conclusively showed that this was primarily caused by the Canadian harvest West of Vancouver Island. Therefore, when the United States and Canada renegotiated their treaty regulating salmon interception, they curtailed that fishery. This had its expected effect and the Chinook runs recovered. 

However, that only lasted a few years because, beginning in 1991, stronger ocean currents caused the Frazer River Sockeye to make landfall further north when they returned  in their spawning migration, across the North Pacific from their feeding grounds on the Arctic fronts. This caused them to make landfall north of the Alaskan/Canadian Border. The Alaskan fishermen caught many of them. The Alaska Department of Fish and Game refused to curtail this interception. The result was that the Canadian/American Salmon Treaty collapsed, in 1992.  The West Vancouver Island Fishery, then, resumed catching Washington Chinook  and their abundance, once  again, declined.  

This ended only three years ago, when Canada unilaterally closed the West Vancouver Island Fishery. In the years which followed, large and even record  runs returned to Washington State.

But, these were only  a few of the leading events in the pattern of harvests. The success of the salmon runs over these years was, also, impacted by changes in the ocean currents and upwelling; sport, Indian, and all-citizen commercial harvests;  poor hatchery practices;  egg takes by the Department of Fish and Wildlife; the growth of predator populations;  the progressive loss of  habitat; and a long list of other contributing factors. 

The Refugia Study correctly states that habitat was only one of several factors influencing the decline of the salmon runs. 

What is generally not appreciated is that the vast majority of Western Washington is still pristine forest. --- People travel down roads and highways through the most heavily developed and rapidly developing areas. Most of them never see the rest of Western Washington, because it is relatively inaccessible.  But, according to a study by Palmisano et al in 1993, over 90% of the original salmon habitat in Western Washington was still intact and functionally in its original condition.

However, this conclusion does not apply to Eastern Washington. There, he found that over 70% of salmon habitat had been lost, not due to the major dams on the Columbia River, but due to the many small  water diversion dams on its tributaries. They mostly  lacked upstream fish passage facilities. 

Perhaps, the clearest demonstration that habitat loss was not a major cause of the salmon crisis in Western Washington is the fact that both the decline and the recovery of the runs occurred throughout the entire region. If it had been caused by habitat loss, it would have been limited to the impacted rivers. But, that was not the case: both pristine and impacted rivers saw the decline and the subsequent recovery. 

So, with regard to the authors' first point, of whether there is a salmon crisis, the answer appears to be that there was, but that it is over, at least, for the moment.

On their second point, of whether it was caused by habitat loss, the answer appears to be that that was not a major factor overall in Western Washington. 

Nevertheless, it was an important factor for stocks in  impacted rivers. As these are the rivers most people see the most, there is a false perception among the public that habitat loss is the leading cause for the decline of the salmon runs.

The author's third assertion was that the rapid rate of urbanization, rapid change in the landscape, and trend towards the simplification of salmonid habitat  has exceeded their ability to adapt, and that this was  the cause of their decline. ---  This is obviously not true, if habitat loss was not an important causal factor.  

Nevertheless, disturbances do occur, so the authors' concept of providing habitat complexity  which would allow the fish to escape  from local disturbances is a valid point.   

Likewise the refuge concept  is consistent with the genetic structure and life history patterns of the salmonids and could be an effective program. If properly implemented, it might make  a significant contribution to assuring the long-term sustainability of salmon by guaranteeing that there will be habitat for them. 

There, also, can be little doubt that the program would be more effective if there were more refuges which were larger, more complex, and  more widely distributed. But, there obviously are tradeoffs and other considerations. 

In addition, the Kitsap Peninsula does not have any large river which could support a large core population of Chinook salmon. However, the importance of the Kitsap Peninsula may be larger when it comes to Chum.  Still, on the West side of Hood Canal there is pristine habitat for Chum which is already protected by the National Park. So, in the larger perspective of the overall health of salmon in the  Puget Sound Region, there should be some  doubt as to whether  the  creation of  refuges in  Kitsap County is  an optimum use of funds, effort, and good will.  

Programs for the preservation of water quality and quantity and  habitat often grossly exaggerate the width of setback that is necessary to avoid adverse impacts. For example, in the Refugia Study, the authors suggest, in some cases,  extending the boundaries of the refuges across the landscape to those of the drainage basin. The need to do this can probably be refuted on physical grounds in many specific cases, simply by considering the scales involved. 

 For example, in  Clallam County, a public committee, the Ground Water Work Group, in 1998, examined the question of whether all the new private wells allowed within a half mile of the Dungeness River under the Growth Management Act could impact instream flows. By doing a mass balance, they demonstrated  that domestic water uses were insignificant.  For something to have an impact, it needed to be very close to the river.

However,  in the Refugia Study, their desire to extend the boundaries across the landscape may not have been motivated by a  concern about the physical environment, so much as by a desire to impact more people and draw them into their socio-political program. 

The nearshore marine environment is not as strongly impacted by  land-uses along its shorelines as streams are. In particular, shading is rarely important to water temperature, insect and leaf fall are not major sources of food and nutrients and,  freshwater runoff is not the source of the water body.  Consequently the biological needs for setbacks are different and in many cases  less for the protection of  marine habitat than freshwater habitat. Brackish waters are, of course, yet another issue. But, the position taken in the study that, because the authors don't know much about the needs of the marine environment, the regulations protecting it should be more restrictive, appears to be poorly justified.    

Their axiomatic position, that technological solutions rarely work, also, is poorly supported. Remote site incubators are an example of a technological solution which is elegant in its simplicity  and clearly works. Various hatcheries are, also, currently experimenting with innovative ways of raising salmon, which resolve many of their earlier difficulties. For  example, the Dungeness River Fish Hatchery is currently testing a program involving a natural stream setting, surreptitious feeding, and no protection from predators. --- It is difficult to see why such technological advances, should be ruled out, when there is every reason to believe that they will resolve some current problems.  

 The authors'  express the sentiment, on their page 9,  that "Not only are ecosystems more complex than we think, they are more complex than we can think."  This is something to which I must strongly object.  

 It has been my experience,  that if I carefully observe the primary biological processes which are occurring in nature, write  equations which faithfully represented them, and carefully collected data for a designed experiment;  those processes have almost invariably been found to account for the vast majority of the observed variation.  

With regard to salmon, in particular, during the early 1990's I had the opportunity to analyze what was, then, the largest and longest time series of observations on the abundance of one of the most studied stocks of Sockeye salmon. I was able to account for approximately 90% of the total year-to-year variation in run size in terms of processes which had been observed in the field, for that particular population (Crittenden 1994). ---  I believe that if the scientific approach of observation and  quantitative analysis  is carefully adhered to, it will generally yield similarly successful results. Thus, my experience is that ecological phenomena generally can be largely accounted for in terms of the major observable processes. 

The accepted approach in science, dating back to William Occam, in the fourteenth century, is  to account for what is simple and observable,  before postulating complications.  Thus, the authors' preemptive assertion that ecological systems are of  unfathomable complexity, is not only unfounded but fundamentally inconsistent with established scientific practice, as it has been for over half a millennium.  

Socio-Politics

The authors of the Refugia Study have turned the belief that ecological phenomena are too complex to understand to a constructive use. --- They conclude, that for this reason, these systems should be managed simply to emulate nature and that management decisions should be made using the approach  known as "adaptive management".  

That approach usually follows the following steps: 

1. observe the present conditions; 

2. brainstorm or consult authorities  to develop management alternatives; 

3. pick one; 

4. see what changes occur; 

5.  repeat the process, beginning at step 2.  

This approach  is fully compatible with committees and group decision making; it is something which any citizens' council can do as it requires little or no expertise; but what is probably  most important is that group decisions can be readily influenced  by the individuals the committee members turn to for suggestions of  possible management alternatives or for funding. 

Once such a group makes a decision, particularly if it is a difficult one reached unanimously, the group members will often acquire a personal commitment to it. This is what gives the individuals higher up in the organizational ladder or those  who control funding the ability to mold their opinions.   

 This is the heart of the political program for salmon recovery. It is an engine for causing and directing change in public behavior. 

The effectiveness of these methods depends on the committee members  using the right-brained processes of group discussion and group decision making. ---  This is, yet, another importance of the belief enunciated by the authors, that ecological systems are beyond understanding. Some viewpoint of this general nature has to be promulgated in order to suppress the use of left-brained, logical, either/or, decision making  by the committee members.  If this is not suppressed, adaptive management will be ineffective in changing the beliefs and values of the committee members. 

In order to create a high level of commitment of the committee and their decisions, each committee member has to have a high personal stake in the outcome and has o have made a compromise during the process. 

Governor Gary Locke gave a good brief summary of this change from the traditional method of decision making to the new one in an address he gave to the Seattle Area Chamber of Commerce Leadership Conference held in Vancouver, BC, October 9, 1998. 

It comes back to our ability to compromise our own narrow interests for the benefit of the common good. And, it comes back to our ability to put aside the old ways of thinking and either/or choices and focus, instead, on achieving the balance which moves us forward. .....  the effort to restore wild salmon will make history in bigger ways than we usually realize.... The stakes are very high. 

This quotation is open to interpretation. It could be interpreted as having an innocuous meaning; but when you are fully aware that the salmon crisis is being used to leverage us into accepting this new form of decision making  and of governance, its meaning becomes abundantly clear. 

This statement by the Governor, also, helps to confirm that the innovations contained in the Refugia Study are not just the ideas of  a small isolated group, but are part of a wider program. 

The Watershed Councils and Water Resource Inventory Areas are already in place and use these methods. But, they operate over broad regions. The refuges, in contrast, are planned to be on a sub-watershed scale and will impact communities and individuals more directly.  

The weaknesses of these new methods  include, "groupthink", "deindividuation", and "extremity shift."  The term "groupthink" was coined by Janis in 1971 and 1972 describe the tendency of group members to adhere to the group norm. There is often intense pressure inside the group for the adherence of the group members to the group's views. With their increased sense of group membership comes a tendency for increased anonymity and deindividuation together with a decreased sense of personal accountability. Groups, also, have a tendency to move progressively towards more-and-more extreme positions. This is called "extremity shift." The result is that group decisions tend to be different from and often more radical than those made by individuals.

As these methods are what the Refugia Study aims at introducing, it would be advisable to become cognizant of them and their various strengths and weaknesses.    

The appointed councils in the  Methow and Dungeness Valleys were pilot projects for the application of this approach to the regulation of environmental issues. With, now, more than a decade of experience with them,  their long-term impact on these communities is reasonably clear: They have engendered suspicion, resentment,  and opposition.  These attitudes are focused against these committees, their process, all the policies they have developed, and  all the agencies who implement them.   Although, these attitudes are more prevalent among the political right, they can, also, be observed across virtually the entire community.  And County Commissioners who implemented and supported them have been turned out of office, in elections in which these issues were among the leading issues. 

At the same time, the Clallam Conservation District is popular and enjoys strong and loyal support, again, across virtually the entire community. They have followed almost diametrically the opposite course. Specifically, they are governed by an elected board, follow the conventional process, and provide technical information and assistance on a  voluntary basis.  

Both the councils and the district provide essentially the same function: that is conservation. But, the contrast between how they are received by the public could hardly be more different. And, this has had an impact on  their ability to achieve their missions. 

Technical Issues 

There were several inconsistencies in the authors' method for ranking refuges. 

• They did not follow the rules they enunciated for assigning weights;

• Although they write about the variables they used, they say nothing about the variables which were available to them, but they did not  use; and

• Both of their measures of rank, that is both GSI-based and fish-based measures, are heavily influenced by the GSI data. The fish-based and biological data appear to have been downplayed; and

• The delineation of the refuges was largely qualitative.

As a  result of these factors, the authors probably had sufficient flexibility  to predetermine which refuges they would assign  high priorities to. Thus, their use of an analytical process  is no  guarantee of  the impartiality of  their selection procedure. 

As the authors did not do a parameter and structural sensitivity analysis of their ranking method, nor any analysis of the various other uncertainties involved; we do not know whether the rank orders they present have any significance. 

The practical importance of these technical criticisms is, that both the rankings they assigned the various refuges  and their boundaries are not beyond question, as indeed, the authors tell us in several different places in their report.

It might be appropriate to consider the proposed Little Boston Creek Refuge, as a case-in-point, as the author's gave it their highest ranking. ---  It is most peculiar, as it includes more than one drainage basin but only limited habitat for the listed species. But, what would probably be the most important  factor in the practical operation of that refuge,  is the fact that  more than half of it is inside an Indian Reservation. That area is not subject to the ESA nor will  it be subject to control by the refuge, as they will derive their authority from the County. Nevertheless,  the Tribe and local Treaty Council will almost certainly be granted seats on the refuge's committee. Then, because they are government entities with significant financial resources, fisheries management responsibilities, and employees who they can assign to the process,  it should be fully anticipated that they will soon dominate it and, in all likelihood, become the refuge's  lead entity. At that point, the Tribe will have effectively gained the power to regulate land-use in those adjoining lands  which are outside the boundaries of their reservation but inside the refuge.

This is not mere speculation, but was the experience with the pilot watershed council in the Dungeness Valley. Today, the local tribe is its lead entity and has, thereby, significantly increased their power over land use decisions in the watershed.

The situation this creates is known as  "arbitrary government". The definition of this phrase, as it was meant at the time this country was founded, comes from John Winthrop, then Deputy Governor of the Commonwealth of Massachusetts, in 1644:

Arbitrary government is where a people have men set over them without their choice or allowance: who have power to govern them and judge their causes without a rule.

The non-tribal-members who have, thus, come under the power of the refuge will have  become subject to arbitrary government because they can not vote on who sits on  the Tribal Council nor, if past experience is indicative, will they have any influence on who sits on the remainder of the refuge's committee either, and according to the Refugia Study that committee will base their decisions on the principles of "Ecosystems Management,"  which are indeterminate. They may even include cultural and aesthetic values. 

A significant number of the complaints enumerated in the Declaration of Independence were about the abuses of arbitrary government. Thus, it is clearly not within the authority of the County nor the National Marine Fisheries Service to impose it, yet this will almost certainly be the  outcome of this particular refuge, if it is created.  

It was the flexibility of the methods used in the Refugia Study which allowed them to set up this situation. As to whether that was a deliberate power grab or not, is more than I know. But, this case-in-point serves as a warning that there is a  need for the critical reexamination of each refuge which is seriously contemplated as well as of the overall Refugia Program.  

Conclusion

The Refugia Study is not a scientific report; it is primarily political in nature.  

As a political statement, although I do not agree with the authors' position, I can not help but admire its thorough internal consistency. Even their writing style shows a right-brained construction, instead of the rigid lock-step that one expects of  the logical development of ideas in a scientific report.  Overall, it is an elegant exposition, in theory, practice, and presentation of the political doctrine underlying the salmon recovery effort.  

Literature Cited

Crittenden R.N. 1994. "A model for the processes regulating recruitment in a sockeye salmon stock." Ecological Modelling, 71: 69-84

Crittenden, R.N. and Ground Water Work Group 1998. Expected water demand for all future domestic wells along the Dungeness River. Clallam County, Dept. of Community Development. also available online at www.hargravepublishing.com.

 Janis, I.L. 1971. "Groupthink" Psychology Today, Nov. 1971 p43.

 Janis, I.L. 1972 "Victims of Groupthink" Houghlin Mifflin, Boston. 

Palmisano et al in 1993. The impact of environmental and management factors on Washington's wild anadromous salmon and trout . Washington Forest Protection Association and Sate of Washington Department of Natural Resources. Olympia, 371 p.

Winthrop, John 1644. "Arbitrary government described and the government of the Massachusetts vindicated from that aspersion" pp 90-112 in American Historical Documents,  The Harvard Classics. FF Collier & Sons, NY.

Critique of Kitsap Peninsula's 2514 Process,

WRIA 15

Dr. Robert N. Crittenden

Crittenden Biometrical

(360) 582-9550

Sept. 7, 2002

Executive Summary

• The "optimum" instream flows provided by IFIM¹⁵ often are greater than the natural flows during the low-flow period.

• Setting such unattainable instream flows effectively transfers the power over land use from the County and Cities to those parties who determine the instream flows.

• The two Tribes who are party to the attachment to the MOU stated their objectives in that attachment: These included at a minimum, the aforementioned "optimum" flows.

• Their conditions for participation in the process are not binding on the other members, but those tribes were granted seats at the table, with a veto power. --- Thus, they have clearly stated their conditions, at the outset, and are in the position to force the process to meet their goals or terminate.

• In addition, IFIM has a qualitative element, specifically the choice of the species, life history stages, and composition of the fish in the stream or river. This gives the people doing the IFIM assessment considerable latitude to select a desired or predetermined "optimum" instream flow.

• This qualitative element of the study would normally be determined by the consultants. The two tribes who are parties to the attachment, appear to have recognized their importance, as one of their stated conditions is to require their consent for the selection of the consultants. This further increases their influence over the outcomes of the process.

• Another item in that attachment, attempts to block any later renegotiation of the instream flows, except where it serves the Tribes' commercial fishing needs or any other unspecified Tribal need. For example, were they to increase the size of their commercial fishery or buy land for a housing development, they could adjust the instream flows accordingly.

• The reassessment of the instream flows to meet their needs, will be done not by the Planning Unit but by those Tribes, DOE and WDFW. Thus, these parties will control the instream flows.

• Many of the items in the attachment are directed at exempting the two tribes from any restrictions arising from the process.

• Another item asserts Winters' Doctrine: that is, that any grant by the Federal Government to a Tribe carries with it an implied grant of all that is necessary to achieve the primary purpose of the grant. For example, a grant of land to be added to a reservation for housing or other development might carry with it an allocation of unappropriated water.

• The two Tribes explicitly do not restrict the implied grants under Winter's Doctrine only to water.

• We see a loose application of Winter's Doctrine, in their claiming a right to a seat a the table to set flows and manage water and land use in any watershed in which they have a history of habitat management activities, not just those within their jurisdiction. --- Thus, when the jurisdictional authority granted them the right to do habitat restoration or management projects off reservation, the two tribes assert that that carried with it an implied grant to be in the 2514 process for those watersheds, with rights equal to theirs.

• This opens the door, should this same principle be applied later, for their claiming a right to manage water and land use throughout the Peninsula, simply because they sat at the table in the 2514 process, or because the outcome of that process included water to meet the needs of their commercial fishery or other unspecified needs.

• The Tribal representatives have verbally stated that they have no intentions beyond those few watersheds in which they have habitat related activities, but an item in their attachment, specifically says that no statement they make during the negotiations is binding on them.

• The Tribes special status creates difficulties in making agreements with them: In particular, no penalty for non-compliance can be enforced against them.

• Consequently, neither the MOU nor its attachment are binding on them. But, reciprocally the attachment says that neither are the conditions binding on the other parties. However, there is a danger that through changes in the law they may potential become binding.

• Nevertheless, despite their disavowal, several of the items in the attachment treat themselves as if they were binding. For example: Item 8.13 says that the 2514 process, thus, created will continue in perpetuity, not withstanding any future legislation.

• In light of these considerations, it would be advisable to reconsider the MOU and its attachment to determine whether it might potentially create a bad process, possibly even in perpetuity.

   

Introduction

The purpose of this report is to point out several problems with the 2514 process for Water Resource inventory Area 15 (WRIA 15) which may lead to serious difficulties if they are not recognized and remedied early in the process.

Optimum Flows:

The seatholders in a 2514 process characteristically do not understand how IFIM works until after the studies are done, if indeed they ever understand it. Consequently, they often do not realize that the "optimum" flows are not necessarily the "scientific" recommendation.

As the "optimum" flows recommended by IFIM characteristically are larger than the natural flows during the low flow period, they can not be achieved most years. Therefore, setting such high minimum instream flows will eliminate virtually all future water rights. In the next stage of the 2514 process, measures will be taken to augment instream flows. Then, it may be expected that the holders of existing water rights may be pressured to surrender some of their water. Furthermore, if hydraulic continuity can be proven, and there undoubtedly will be an attempt to do so, existing and future wells will also be at risk.

Without water rights, productive agricultural uses of the land will be severely limited and building permits can not be issued. Thus, unless these problems are recognized and corrected, the instream flow rules will probably strongly impact the orderly pattern of land use and development throughout the Peninsula.

This is what happened to the Methow Valley. That is probably the best known example of these problems. But, the same thing has, also, occurred in other river basins, throughout Washington.

The Squaxin Island and Skokomish Tribes¹⁶, who are parties to the attachment to the MOU, appear to have understood the importance of the "optimum" flows, and have clearly stated their objectives. Specifically, item 8.8 in the attachment states:

8.8 Tribal claims to instream flows include at a minimum: minimum flows necessary to provide optimum habitat conditions for salmon.

If we go back to the MOU to see whether this is binding upon the process, we find the following item:

8.0 Tribal Conditions for Participation: The attached Squaxin Island and Skokomish Tribes Statement of Conditions for participating in watershed planning under ESHB 2514 are acknowledged by all the parties of this agreement. These conditions outline those Tribes conditions for joining this planning effort, but they are not binding on the other parties to this agreement or the Planning Unit.

Thus, this might appear to be innocuous. But, these Tribes were accepted as members of the expanded initiating governments and as such have veto power over the decisions of the Planning Unit. --- As they entered the unit under the conditions given in the attachment, it should be assumed that they will use their veto power if these are not met. --- The practical result of this is to require the minimum instream flows to be set at the "optimum" flow or higher: That would usually be equivalent to requiring that they be set higher than the average natural flows during the low flow period.

Qualitative Element in IFIM:

Due to a qualitative element in IFIM, the optimum flow is potentially subject to manipulation. This gives it flexibility to meet different objectives or allow creative solutions.

However, the qualitative element can, also, readily be hidden behind a facade of science. Consequently, IFIM has far too often has been used as a tool for driving negotiations to some politically determined objective. This is likely to happen whenever, the bulk of the seatholders at the table, do not recognize its flexibility, but accept the recommended instream flows, set according to the objectives of some other party, as if they were what science recommends.

Most of the initiating governments were probably unaware of the qualitative element in IFIM, or even that IFIM would be used, when they formed the MOU; but the two Tribes who are parties to the attachment to it appear to have understood how these studies are done. --- We see this in item 8.9 of that attachment, which requires their consent for the selection of consultants.

Unless the Planning Unit becomes aware of the importance of this qualitative step in IFIM and make these decisions themselves, the consultants are the individuals who would normally do this. Thus, the two tribes by requiring control over the selection of the consultants, gained control over this critical step.

Lock-in on Instream Flows:

Item 8.6 of the attachment, attempts to prevent any reassessment of instream flows once they are set, except where that serves their commercial fishing needs and any other unspecified Tribal needs; and item 8.7 exempts any readjustments of instream flows for those purposes from the 2514 process.

8.6 The Tribes will not agree to re-negotiation of instream or base flows that have been set by rule or agreement in the watershed, except to the degree that modifications are necessary to meet the requirements of the fish resource and the reserved rights of the Tribe. Further, the Tribes will not agree to re-negotiation of closures to further appropriation by Department of Ecology rule or determination in any stream reaches in the watershed.

8.7 Assessment of instream flow, for purposes of the state or Tribes, will be conducted jointly by the Department of Ecology with the affected Tribes, and in cooperation with the Department of Fish and Wildlife.

Thus, they create a vehicle by which they and the Department of Ecology in cooperation with the Department of Fish and Wildlife can readjust any instream flow which is set, for any stream over which the Tribes can claim a fishing right or any other type of right.

If these two items have any force in law, or come to have such force through some future legislation or court ruling, then the setting of instream flows by the Planning Unit becomes irrelevant, for those two Tribes together with the Departments Ecology and Fish and Wildlife can change them at will.

Bearing in mind that control of water translates to control of the land, these two items would transfer power over land use regulation from the Counties and Cities to these other government entities.

If that is their objective, then their requiring that the minimum instream flows be at least the "optimum" flows makes sense, for setting such a high instream flow would terminate many existing and all future water rights and end development, except at their pleasure.

In many places in the attachment to the MOU, the Tribes repeatedly state that no outcome of the process is binding upon them nor restricts them in any way, unless they agree in writing. --- They were very careful to exempt themselves from any control by the Planning Unit or any rule or plan that arises from the process.

Winters' Doctrine:

Item 8.1 contains a thinly veiled reference to Winters' Doctrine. That legal doctrine is that when a grant is made by the Federal Government to a Tribe for a purpose, there goes with it an implied grant of all that is necessary to achieve the primary purpose for which the grant was made. That particularly applies to implied grants of unappropriated water to serve the primary stated purposes of grants of land to Tribal reservations.

Winters' doctrine comes for a decision made by judge McKenna in 1908¹⁷ regarding an Indian Reservation at Fort Belknap, Montana. That Tribe had traded a larger amount of range land for a smaller amount of arable land, in order to become farmers. However, their water rights later came into question. Judge McKenna's ruling was that there had been an implied grant of the water necessary for them to farm their land.

Originally, Winter's Doctrine was applied only to implied grants of unappropriated water to achieve the purposes of grants of land to Tribal reservations, but during the last approximately four decades, it was much more broadly applied, to many other types of Federal grants of land. However, most recently the courts have restricted its application to only the primary purpose of Federal grants. ¹⁸

The statement in item 8.1 of the attachment to the MOU which refers to Winters' Doctrine is:

"Further, the Tribes reserve their judicial, statutory, and equitable rights to any and all waters to meet the changing needs of its tribal members in fulfilling the purposes ¹⁹ of the reservation. "

Winters' Doctrine was also pivotal in the Boldt II decision --- As you may recall, Judge Boldt had granted half the salmon catch to the Tribes.²⁰ In the second half of that case, Judge Orrick ruled, based on Winters' Doctrine, that there was an implied grant of water and habitat necessary to maintain the salmon stocks. Furthermore, that as the fishing right was a reserved right which had existed before statehood, these rights to water and habitat superseded the rights of the citizens of the State. But, these water and habitat rights only come into force if the Tribes can no longer obtain a modest income from fishing for salmon.²¹

Judge Orrick had assumed that water and freshwater habitat were the factors limiting the abundance of salmon. However, according to Bob Lohn, Director of the National Marine Fisheries Service for the Northwest Region, "Most credible scientists, today, believe that the salmon crisis was caused by ocean conditions, not habitat."²² --- The most readily observable evidence in support of his position is that salmon abundance's declined in virtually all rivers throughout the Northwest, not merely those which had suffered habitat losses. Likewise, the recovery occurred everywhere, rather than being localized. --- Thus, the applicability of the premise underlying the Boldt II decision, is today widely questioned.

Notwithstanding, the applicability or inapplicability of that key underlying assumption, the water and habitat rights granted in the Boldt II decision were overturned by the ninth circuit court, in 1982.²³ Later, there was a conflicting ruling on another case before the ninth circuit, and the issue was taken to the Supreme Court. They vacated the challenge to the 1982 ruling.²⁴

However, some Tribes would like to resurrect the Boldt II decision or apply Winters' Doctrine to other purposes, and they, also, would like to interpret the applicability of Winter's Doctrine broadly. -- They appear to have introduced it implicitly, interpreted broadly, into their attachment to the MOA.

Nevertheless, the applicability of Winter's Doctrine has been restricted and Boldt II was overturned in part, it remains applicable in its narrow sense: That is the implied grant of unappropriated water necessary to achieve the primary purposes of a grant of land by the Federal Government to an Indian Reservation. This, may prove to be important.

For example, consider for the moment that all the instream flows are set at unattainable levels. That would block the construction of any new housing or development on the Peninsula. However, one of the Tribes could buy land and then get the Federal Government to make that land part of their reservation for the purpose of building housing or other development. The Tribe could then decrease the minimum instream flow in some stream, thereby releasing the necessary water from appropriation. Consequently, by the implied grant, that water would become theirs. --- According to the attachement to the MOA, this reassessment of instream flows will be done in collaboration with the DOE and WDFW. --- Thus, these government entities would acquire power over land use on the Peninsula.

Grant of Jurisdiction?

What item 8.1 of the attachment does is reserve any unspecified implied rights under Winters' Doctrine which they might wish to claim. This should serve as a warning that any grant to these Tribes, whether it be of water, habitat, the right to sit at the table, ... virtually anything, should be carefully considered to determine whether there might be an implied grant of some kind.

Item 8.5 provides an example of an assertion resembling Winters' Doctrine or of that doctrine construed very broadly: In it the two tribes claim rights in the 2514 process equal to those of an initiating government for any watershed in which they had a history of habitat management activities:

Effective watershed planning cannot take place without the full participation, as an Initiating Government, of all Tribes with a legal interest or history of fish habitat management activities in the affected watershed. To achieve this, the Initiating Government will invite such Tribes to participate as Initiating Governments in determining the scope of planning, composition of the Planning Unit, and all other authorities of such governments.

Thus, when the legal jurisdictional authority allowed them to conduct habitat management activities off-reservation, that provided the basis for this claim to an implied grant of authority equal to their own within the 2514 process for those watersheds.

There are two qualifications here: 1) Their inclusion at the table was mandated by 2514, because their reservation lies within the WRIA, not by Winters' Doctrine; and 2) The initial grant was from a County or City not the Federal Government.

Nevertheless, item 8.5 illustrates that because of the special status of the Tribes, special care needs to be taken whenever anything is granted to them, because, a law may later be passed or a court ruling delivered which will convert that into an implied grant of something else which was never intended as part of the initial action.

One must wonder, whether, the Tribes' inclusion at the table in the 2514 process does not constitute "management activities" for the entire Kitsap Peninsula, which might be used, at some future date, to justify their assuming management authority over that entire region. That would be relevant if the vehicle created by items 8.6, 8.7 and 8.8 discussed above, came into practical use.

Is the MOA and its Attachment Binding?

I am told that the Tribes have denied, verbally, having any interest beyond the few watersheds in which they have an active involvement.²⁵ But one should bear in mind Item 8.10 of the attachment in which they claim that no verbal agreement or statement is binding upon them:

No discussion, plan, proposal, agreement, offer of compromise, proposed agreement, concession, statement, data, material or document, whether oral, written or in electronic or other format, made by or on behalf of the affected Tribe in furtherance of the planning process envisioned by this agreement shall be admissible in any legal or administrative proceeding, regardless of whether the Tribe is a party to that proceeding.

Looking at the broader issue, there are special difficulties to forming agreements with the Tribes.... The problems are their special status and the ambiguity of their status:

The options on how to regard them seem to include 1) a branch of Federal Government, 2) a local jurisdictional body; 3) a minor foreign nation which is under the protection of a more powerful nation; 4) an Indian Tribe, which is a unique category unto itself, or 5) an indigenous people.

The Centennial Accord, between the State of Washington and the Tribes, clarifies this issue to some degree as it says that the State will treat them as sovereign nations. --- That is the policy of Washington State. Presumably it also applies to all the departments and divisions of the State, including the cities and counties.

The interpretation of their status as sovereign nations comes from the ruling of US Supreme Ct. Justice Marshall in 1832. ²⁶ It dealt with the Cherokees, who had been a nation and signed a treaty with England during our colonial period. This may be the status in which the tribes are usually regarded as being, when they are regarded as sovereign nations under treaties with the US.

A difficulty arises in applying Marshall's opinion to many of our local tribes, as many of those which signed the Steven's Treaties were not nations. --- The Steven's Treaties refer to the Yakima as a nation; the Dwamish, Squeamish, and; Makah as tribes; but the rest as bands. Many of the latter had no established form of government; Stevens had to appoint leaders or representatives for them in order for them to sign the treaties. As the two tribes which are party to the attachment appear to have been bands, they may never have had a prior existence as a sovereign nation. --- The Centennial Accord did not say that they were or are sovereign nations, only that the State would treat them as such.

If they are sovereign nations, the State and the various divisions of it may not have the power to make agreements with them as the U.S. Constitution grants that power only to the Senate; but the real problem is that law suits under State and Federal Law can not be brought against foreign nations nor sovereign Indian Tribes , whichever way you wish to interpret their status.

Their status surfaces in items 8.1 through 8.4 of the attachment, where they claim to be a Tribe subject to a Treaty; and in item 8.17 where they claim sovereign immunity:

8.17 Nothing herein shall be a waiver of Tribal sovereign immunity for any litigation including but not limited to any general stream adjudication.

And in item 8.13 where they claim that the agreement is not subject to legislation:

8.13 The parties agree the consensus decision-making process defined under this agreement will remain in force not withstanding any future legislation.

In this last item, one might assume that they are claiming to be above State legislation, but should they claim status as an indigenous people under the U.N. Treaty on Indigenous Peoples or some similar international treaty, it might also be taken to be an assertion that they are above Federal Legislation. ²⁷

That raises a question: "What kind of agreement is an memorandum of understanding (MOU) with a Tribe, and how does that differ from a memorandum of agreement (MOA) or a contract?" As I have heard a Tribal representative emphatically and repeatedly referring to an MOA as an MOU, I incline to the opinion that there is some significant legal difference between them.²⁸

Although I can not claim to be an authority on this, I have generally assumed that an MOA is a contract between parties; but an MOU is merely a written statement that the parties are of like mind on some issue. The text of the MOU and its attachment is not informative on this as it treats itself and the attachment as if they were binding agreements in some places, while in others they explicitly state that they are not.

Contracts have certain required parts. One of those parts is the penalty for non-compliance. This poses a difficulty for any contract with a Tribe, as they can be expected to assert Tribal immunity the moment you try to sue them for non-compliance. That is, they will assert their special status as a sovereign nation or Indian Tribe not subject to our laws.

Consequently, if a contract must have all its parts to be binding but no contract with a Tribe can have a penalty for non-compliance, no binding contract can be formed with a Tribe.

Reciprocally, item 8.5 of the MOU states

These conditions <in the attachment> outline those Tribes conditions for joining this planning effort, but they are not binding on the other parties to this agreement or the Planning Unit.

If nothing is binding upon the Tribes and the attachment is not binding upon the other members, then all that remains, if anything remains at all, is the body of the MOU. It may be binding upon the other members. --- Overall, It appears to be a doubtful agreement, possibly of no more weight than a written statement that the various parties are of like mind, to attempt to reach negotiated agreements through the 2514 process.

However, there are parts of the attachment which treat it as a binding agreement. Also, it must always be born in mind that future legislation or a court ruling might change what is not binding, today, into something which is binding. --- Item 8.13 is of particular concern, in this regard, as it asserts that the 2514 process, thus, created is not only binding but will continue in perpetuity:

8.13 The parties agree the consensus decision-making process defined under this agreement will remain in force not withstanding any future legislation.

In conclusion to this section, the MOU and its attachment may or may not currently be binding. However, they contain some potentially ominous sections which are entirely too open to judicial interpretation or future changes in law. Therefore, devoting some attention, at the outset, to replacing those sections which are unacceptable or could be misconstrued or modified by outside action, with explicit mutually acceptable agreements, would give the process a higher likelihood of resulting in a good rather than a bad outcome.

Best Available Science:

As flawed methods of assessment of instream flows have been and continue to be widely used throughout Washington State, the issue of best available science becomes relevant.--- Failure to base the 2514 process on valid science creates a problem as the minimum instream flows which will be set by the 2514 process and the watershed plan, which will come afterwards, will create obligations for the cities and counties to implement them in ordinances. But, as such policies will necessarily regulate critical areas for salmon and other fish species, according to the sections of the Growth Management Act, in the RCW, which deal with critical areas, the counties and cities must base their ordinances which deal with those areas upon best available science. Thus, should the planning unit base their decisions on instream flows upon studies which are not best available science, they could potentially create obligations for the cities and counties which they could not implement without violating the Growth Management Act. It is, therefore, necessary, by implication, that the instream flow assessments for the 2514 process be scientifically valid.

Literature Cited

Bovee, K.D. et al. 1982. A Guide to Stream Habitat Analysis Using the Instream Flow Incremental Methodology. Instream flow Information Paper #12 U.S. Fish and Wildlife Service, Fort Collins, Colorado.

Getches, D.H. and C.H. and Wilkinsen 1986. Federal Indian Law, Cases and Materials. Second Edition. American Case Book Series. West Publishing, St. Paul Minnesota.

Review of Kitsap County's

Second Draft Critical Areas Ordinance

by

Dr. Robert N. Crittenden

June 2005

Crittenden Biometrical

274 Sturdevant Road, Sequim WA 98382

360 582-9550

Prepared for the Kitsap Alliance of Property Owners

© copyright 2005. All rights reserved by Dr. Robert N. Crittenden

Permission to make copies is granted to the Kitsap Alliance of Property Owners

Section 1: Executive Summary

This report raises serious concerns about the validity of Department of Ecology's (DOE's) recommended regulations, studies and assessment method for wetlands.

1. Kitsap County's wetlands buffers generally follow one of the recommended options that DOE published in Volume 2 of Wetlands in Washington State. Specifically they follow option 3 in their Appendix 8C.

2. The key study upon which those recommendations are based, for Western Washington, is the report published by Hruby et al., in 1999.

3. That report presents a new assessment method for Wetlands in Western Washington.

4. Their new method is based on a mechanistic model. That is a decision-making model instead of being one of the conventional approaches for environmental modeling.

5. In developing that new method, they made decisions on which factors should be considered and their relative importance. These decisions are imbedded within their assessment method.

6. Thus, they become imposed, implicitly, upon any jurisdiction which accepts their assessment method. This deprives those jurisdictions of the opportunity to consider these key underlying issues.

7. The local jurisdictions should explicitly address these issues and not accept the new assessment method without considering the possibility of amending it to incorporate local views, concerns and conditions.

8. Their new assessment method should be relatively easy to amend to reflect different views and values. This can be done by changing the weights assigned to each of the variables in their assessment method.

9. If that is done, the County should still be able to rely on DOE for the training and certification of wetlands specialists, as only the computations would differ, not the field measurements, although additional variables could be measured if they are desired.

10. Changing the weighting in the assessment method will change many of the County's required buffer widths, because they reflect the current weighting.

11. Hruby's 1999 report is not a scientific study but reports the conclusions reached by a committee.

12. The objective of that committee was to adapt a national rating system for wetlands to local conditions.

13. They used group processes to reach their decisions: In particular, these appear to have included the Delphi and consensus processes.

14. Those processes can be controlled so that they produce a desired outcome, DOE has expertise in managing those group processes, and they organized and managed the meeting of this committee.

15. Having a familiarity with these methods conducted by DOE, I have little doubt that they obtained the results they desired.

16. The committee field tested their new assessment method, but only to determine whether its results were generally compatible with the old method, not to evaluate its scientific validity.

17. Their new assessment method is partially based on qualitative judgments instead of value-free science.

18. Their new assessment method is also partially based on variables that are correlated with the environmental processes or qualitative values that they considered to be of interest. But, correlation does not prove the existence of a causal relationship between the variables measured and those they predict from them. Thus, their use of these variables lacks scientific support.

19. DOE presented the supporting Best Available Science (BAS) for their assessment method and recommended regulations in Volume 1 of Wetlands in Washington State.

20. Their publication of the supporting BAS followed their publication of their assessment method and recommended management options, rather than the BAS being assembled first and the assessment method and management recommendations being based upon it.

21. Although they do not state it explicitly, it appears that this reversal of the order of publication may have occurred because they were following the principles of conservation biology. In particular, Noss and Cooperrider, who laid out these principles, in 1994, recommended that policy goals be established, first, and the science that supports them should be found, subsequently. When this approach is followed, the goals are not based on the science, but the "science" on the goals.

2: Wetlands

2.1 DOE's Wetlands Studies

The Washington State Department of Ecology (DOE) recently completed several coordinated major works on wetlands science, assessment and regulation:

Dyanne Sheldon et al. March 2005. Wetlands in Washington. Volume 1: A Synthesis of the Science Washington State Department of Ecology, Lacey WA. --- This is their review of the Best Available Science for Wetlands.

Teri Granger et al. April 2005. Wetlands in Washington, Volume 2: Guidance for protecting and Managing Wetlands. Washington State Department of Ecology, Lacey WA. --- This lays out their recommendations for the regulation of wetlands, particularly in setting buffer widths. Kitsap County is essentially following option 3 in Appendic 8-C. Jefferson County will probably also do this.

Thomas Hruby et al. August 2004. Washinton State Wetland Rating System for Western Washington (Revised).Washington State Department of Ecology, Lacey WA. --- This lays out DOE's new rating system for wetlands, based upon the study done by Hruby et al. in 1999.

Thomas Hruby et al. 1999. Washington State Wetlands Assessment Project for Western Washington. Washington State Department of Ecology, Lacey WA. #99-115 --- This is the key study upon which the others rest.

All four of these works are available online, at ecology's website, www.ecy.wa.gov/programs/sea/pubs. and free printed copies of the first three are available from Ecology upon request.

2.2: The New Assessment Method is a Decision Model

In his 1999 report, Hruby presented a new assessment method for wetlands in Western Washington. It is a "mechanistic model" and he stated that models of that type are decision-making models. That is, they are a standardized or synthetic decision-processes.

This should be contrasted against the two leading conventional forms of mathematical modeling of environmental or ecological processes. These are empirical modeling and realistic modeling. In empirical modeling, an equation or equations are fit to a set of data and are then used to make predictions, whereas, in realistic modeling the form of the equations is derived from the processes they model. Modeling, is conventionally done using either one or the other of these two approaches, depending upon which is appropriate for the particular circumstances.

But, the new assessment method is neither of these. It is a decision-process.

Technically, what that they did. instead of using a mathematical modeling approach, was to use a Baysian-like approach or, perhaps, it would be better to call it a "pre-Baysian" approach. --- In the early or initial stages of a Baysian approach, the decisions are often based upon qualitative judgments and values instead of on hard scientific evidence. That is the stage their method is at and Hruby explicitly stated in several places in his 1999 report that the new assessment method is partially based on qualitative judgments. For example he says on page 15 of his study, in describing his approach of mechanistic modeling, that:

Mechanistic Models are Decision-Making models ... Decision-making models represent "the acquisition and merging of subjective, expert knowledge ... Often several persons with varying backgrounds are to be taken into the analysis, e.g., engineers, ecologists, economists, managers, and politicians" (Varis et al. 1994). Each variable in a model represents a decision criterion used to establish a level of performance, rather than an independent variable that estimates the rate of an environmental process. These decision criteria are bused on the judgments and experience of the Assessment Teams and on the research that has been done to date.

By "performance" he appears to have meant the social or political performance of the assessment method, which he contrasted against the rate of an environmental process. So, when they field-tested their new assessment method, they examined its predictions relative to the old method, presumably, as a similarity would be desirable from a political and social perspective because that would maintain continuity with existing regulations, instead of evaluating its accuracy in estimating the environmental functions and values of wetlands. Thus, the new assessment method was focused and guided by social and political concerns, not by science nor by a concern for protecting or restoring the environment.

2.2 It has Decisions Imbedded Within It

Their new assessment method has values and judgments imbedded within it and imposes them on whoever uses the method. For example, the committee who constructed it made decisions on which functions and values should be considered and the merit of each of them relative to the others.

For a simple example, filtration and wildlife habitat are two such variables but, wildlife habitat often requires much larger buffers than filtration. So, what weight should each of these factors receive? ²⁹--- Let F stand for the buffer width needed for filtration, H for its width for habitat, W(f) for the statistical weight for filtration, W(h) for the statistical weight for habitat, and B for the recommended buffer width; then if the recommended buffer width is obtained as a weighted mean, it follows that B={W(f) F + W(h) H} / {W(f)+ W(h)}. Thus, the recommended value for the buffer width will depend upon the relative value of the two W's. --- That is what they have done: They have set the weights. By doing this, they have determined the relative importance of these two factors. This has deprived the counties and cities of the opportunity to make the decisions on these important issues.

In particular, they decided to consider habitat for amphibians, including tree frogs and salamanders. Some of these species migrate long distances from the water in which they breed. For some species, that may be as much as half a mile. In contrast, ten meters of buffer may be sufficient to remove more than half of the dissolved nitrogen and phosphorous. So, by giving these two factors equal weight, the recommended buffer becomes effectively determined by the habitat needs of tree frogs and salamanders. At the same time, the value of human uses of the land, which is another variable which might have been considered, but wasn't, received a weight of zero.

Thus, the issue of which factors should be considered and their relative importance were decided by the committee who formulated the new assessment method. Some of the decisions they reached, such as preferring tree frogs and salamanders over land owners, are not ones that a public process would ever be likely to reach. Therefore, this might be seen as a subtle way for removing the decision-power over those issues from the local jurisdictions in order to achieve goals which would otherwise be unachievable.

2.3 This Can be Corrected

The relative values they imposed upon the assessment method can be determined by doing a sensitivity analysis of the equations in their assessment method. Those equations are simple enough that this can probably be done analytically. However, if that is not possible, computer methods are also available. The point is that they made some set of assumptions and each of them should be critically reconsidered to determine whether it is appropriate for Kitsap County. If they are not, a different set of weights can be used, instead. By changing the weights, the County can readily incorporate local views, needs and concerns into the assessment method in place of the decisions which were implicitly imposed upon them.

However, the DOE representative who presented this new assessment method to the Jefferson County Planning Commission, suggested that Jefferson County could change the suggested buffer widths for the various categories, but that they should not change the assessment method. --- Perhaps, she understood the role the assessment method plays in imposing DOE's views and objectives upon the Counties.

If the County chooses to modify the equation used in the assessment method, it should, nevertheless, still be possible for them to rely on DOE's training and certification of wetlands specialists, as only the computation would differ. The field data collected would be the same, although, addition variables could also be measured if they were desired.

However, if the County does not have the time, manpower or technical ability to develop a new equation for the assessment method, they could facilitate it's being developed at some future date, by explicitly giving that equation in their code, instead of adopting it by reference from DOE. At the very least, the County should be more specific where it adopts DOE's assessment method by reference. In particular, the County Code should include the publication dates for the documents describing that method, so that it is clear that the County Code refers to that specific method, rather than whatever DOE's current wetlands assessment method may be, should they change their method at some time in the future. But,. if the County explicitly gives the assessment method in its ordinance, they might also insert a neutral set of weights into it and define their values as a separate item in their code. That would make it easy to change them at some later date.

However, changing the weighting in the assessment method will raise many other issues as that will probably necessitate changing many, if not all, of the buffer widths in the County Code, for they appear to reflect the weighting which is implicit to the new assessment method. For example, their assessment method places a high value or heavy weight on wildlife habitat relative to filtration, but if this situation were reversed, the recommended buffer widths would be markedly reduced.

2.4 Hruby et al., 1999, Presents the Results of a Committee Meeting.

Hruby's 1999 publication was not a scientific study. It was a report of the results reached by a committee.

Their objective in meeting was to adapt a national system for wetlands to local conditions. In particular, he stated on page 7 that "The highest categories (that is, classes) for wetlands in a region are defined nationally (Table 1). Subclasses for each of these classes are defined regionally by experts within the area. The wetland experts in each region can, therefore, tailor the subclasses to address differences in the performance of functions by different wetland types in their region."

They gathered together a group of wetlands scientists to do this. The principle event was a four-day workshop. It was conducted by group processes and, in particular, it appears to have been an application of the consensus and Delphi processes. That is, the scientists were broken into several sub-committees each responsible for a section of the overall project. There was also a central committee of appointed members which gave them their initial direction and afterwards assembled their results. Each of these sub-committees was conducted by the consensus process and had a facilitator to direct and coordinate their efforts. --- The Delphi process was developed by Rand, during the 1950's, and they said that by applying this method they could obtain any desired result from a committee of scientists, whereas the consensus process originated in Elton Mayo's study of industrial psychology. Its purpose was to induce workers to more readily accept top-down control. --- Those of us who have gained a familiarity with the watershed and water resource inventory area (WRIA) processes are well aware of DOE's expertise in the application of the consensus and Delphi processes and their ability to direct these groups' decisions towards whatever outcome they desire.

Nevertheless, whether you believe that DOE deliberately influenced the outcome of that committee towards the one they desired or not, it is clear that Hruby's 1999 report was not a report of the results of a scientific study but gives the results of a committee meeting, instead.

2.5 What the New Assessment Method Measures

Although, they did field test their new assessment method on 122 wetlands in Western Washington,³⁰ their objective in doing this was to demonstrate that the new assessment method produced wetland classifications which were similar to those provided by the current method.

They did not address questions such as whether their assessment method is reasonable, whether it accurately estimated the level of activity of the various functions and values it purports to assess, nor how much protection which might be needed. In fact, Hruby states on page 21 of his 1999 report that the, "assessment methods do not reflect the degree to which different functions or subclasses should be protected." And on page 15 of that report, he states that the assessment methods "rely on easily observed characteristics that are correlated with the actual environmental processes... When it is not feasible to use a variable because it cannot be rapidly assessed, it is sometimes possible to use an indicator as a surrogate for that variable. Indicators are easily observed characteristics that are correlated with quantitative or qualitative observations of an environmental variable."

That latter statement raises several concerns. The most alarming of these is his "qualitative observations of an environmental variable." Science should be value-free, instead. But, no less alarming is his statement that these qualitative variables are not being measured directly, but some other variable that is correlated to them is measured, instead. The problem with that, is that correlation does not prove the existence of a causal relationship between the measured variable and the other variable it is used to predict. Thus, the substance of what he said is that they were in some cases measuring variables which are not necessarily causally related to the qualitative variables they are trying to predict. --- That is not a valid scientific method.

2.6 DOE's General Approach

One of the reasons for the County's not meeting the deadline, of December 2004, for completing their CAO update may have been that they were waiting for DOE to publish Volume 2 of Washington Wetlands. That was not published until April of 2005 and the final draft of Volume 1, which presents the supporting Best Available Science was completed a month later.

Notice that that sequence of events is backwards from what it should be: That is, they completed the suggested regulations first and the review of the supporting science afterwards. They should have completed the review first and then based their regulations upon it. Furthermore, the key study is their new assessment method and that was completed during 1999, five to six years before the supporting science was found.

The explanation for this peculiar sequence of publication dates appears to be that DOE has been following the strategy of conservation biology, as it was enunciated by Noss and Cooperrider, in their 1996 book, Saving Nature's Legacy.³¹ The principle in question, here, is their recommendation that the science should be found that supports whatever program they desire, rather than basing the program upon the science. In particular, they say on their page 89, "Conservation biology,..., is not value-free science. Rather it is mission-oriented,... Goal-setting must be the first step in the conservation process, preceding scientific, technical, and political questions."³² DOE's use of group processes fits well with this general approach because, by using them, a committee of scientists can be induced to provide the desired "scientific" support or, as in this case, the desired assessment method.³³

If I had only the rather slim evidence provided by just DOE's sequence of publications, I would not have suggested that DOE did this. However, I know from several other sources that this approach has, over recent years, been followed by various parts of DOE. However, rather than belaboring the point, I will merely suggest this as a possible explanation for the peculiar order of their publications.

2.7 Conclusions

Under critical examination, the fabric of DOE's suggested regulations and its supporting "science" falls apart. The key flaw is the study by Hruby et al., in 1999. It is not science but reports the results of a committee meeting and a socio-political process. Unfortunately, basing studies on political doctrine and social processes is not a reliable way to do science. So, it is not all that surprising to find that that study is not scientifically valid and contains serious errors both in scientific methodology and substance. As DOE's three other wetlands studies are based upon that study and their proposed regulation are, too, they do not have a valid scientific basis.

These wetlands studies from DOE also provide a significant part of the basis for the other types of buffers proposed by the County in their second draft Critical Areas Ordinance. That is, its their buffers on streams, lakes and marine shorelines.

Consequently, a thorough reconsideration of the underlying studies is needed and that should be expected to result in corresponding major changes to the County's draft ordinance.

------Footnotes----

Critique of Best Available Science for

Pierce County's Proposed Regulations of Marine Shorelines, Title 18E

by

Dr. Robert N. Crittenden

274 Sturdevant Road, Sequim WA 98382

360 582-9550

Executive Summary

The County's Best Available Science report treats the littoral zone as if it were a type of riparian zone. But, the riparian and littoral zones differ in some, but not all, of their functions and values: Specifically,

• Filtration is important in both zones;

• The functions and values of drainage, differ dramatically between the two zones;

• Large woody debris is a mixed blessing in both zones;

• Shading is an important determinate of water temperate in streams, but rarely in the marine environment; and

• Insect fall from nearshore vegetation provides an important food source to salmonids in streams and rivers, but only to some salmonids in marine waters.

Consequently, riparian setbacks, buffers and regulations can not be expected to achieve their intended purposes when they are applied to the littoral zone.

Introduction

The County's review of Best Available Science describes the littoral zone, that is the land areas contiguous to marine shorelines, as the "marine riparian" zone. That newly-coined phrase is an oxymoron as "riparian" comes from the Latin word for "river" and has traditionally has been used in the ecological literature to refer to the shorelines of streams and rivers or, more generally, to freshwater shorelines as distinguished from marine shorelines.

However, the phrase encapsulates a new viewpoint which is emerging in contemporary land use regulations, that there are many similarities between the littoral and riparian environments. For example, two recent publications which express this viewpoint include one by Levings and Jamison, in 2001, and another by Brennan and Culverwell, in press.

If their viewpoint were correct, streamside buffers and regulations could be applied to the littoral zone. However, there are significant differences between these two environments. This report briefly examines their respective functions and values to show their similarities and differences.

Filtration

Filtration is one of the important functions and values of both riparian and littoral environments. Vegetation or engineering solutions based on best management practices (BNP's) can remove silts, fine sediments, and various pollutants. These are known to be harmful life in streams and rivers, smothering or poisoning the biota, and they do likewise in the nearshore marine environments. This is a point of similarity where similar regulations are applicable to both environments. A reasonable buffer and/or use of BNP's are appropriate.

Drainage

Increased peak stormwater discharges are recognized as being the single most important impact which urbanization has upon streams and rivers. A review of this was presented by Booth, in 2002.

Although his report, as well as several of the studies he cites, do not meet the standards of Best Available Science, due to an their adequate description of their modeling and lack of statistical tests, there can be little doubt in the overall result they present: that is, that urbanization results in a more rapid hydraulic response of the watersheds, and that, if it becomes too severe, destabilizes the streambeds resulting in the severe degradation of habitat quality in those streams and rivers.

The primary cause of the increased peak flows is the impervious surfaces of roads and parking lots. Many of them drain directly into streams and rivers. In addition, according to May, in 1996, they account for roughly 65% of the impervious surface in rural King County.

Regulations to slow the hydraulic response of urbanized watersheds to storm events are appropriate. These may include ponds for the detention of stormwater, other engineering solutions, and restrictions on clearing, grading, and impervious surface. To be effective these must apply to roads and parking lots, as they are the primary cause of the problem, but it might also be appropriate to extend them to other land uses.

However, they do not apply to land areas which do not drain into streams or rivers. In particular, they do not apply to the littoral zone, which drains directly into saltwater.

Different processes occur in the littoral zone, relating to drainage.... For example, the stability of bluffs along the marine shorelines is highly dependent on the amount of water retained in the soil. Achieving adequate drainage, to maintain their stability, becomes an important issue.

Therefore, different regulations are needed on drainage of riparian zones than for littoral zones, as they serve different functions and values.

Large Woody Debris

Large Woody Debris (LWD) are often considered beneficial to streams and rivers, as they provide habitat, pools and environmental complexity. However, they are not without hazards, as they can cause log jams which can lead to an avulsion event (that is, a sudden change in the course of the river channel) or pose a risk to recreational users of the stream or river.

Likewise, LWD's contribute to the beach wrack, which provides habitat for many species, and they can protect the shorelines from wave erosion. However, a poorly placed log on the shore can reflect wave energy, disrupting the soft-substrate benthic community in front of it, and floating LWD's pose a risk to navigation.

There is increasing awareness that LWD's may be good or harmful, and that there may be an associated risk and liability. --- LWD's need to be carefully placed and adequately anchored. --- This undoubtedly, also, applies to the marine environment. Thus, LWD's may originate in either the riparian or littoral zones, but in both cases, they are not always benefitial.

Shade and Microclimate

Shading streams and rivers is one of the important functions and values of riparian vegetation. It reduces the solar loading, helping to keep the water cool. Adult salmon also prefer to migrate through shaded reaches as that helps them avoid predators, and likewise juveniles preferentially utilize shaded areas, particularly areas with overhanging roots or vegetation.

In contrast, shading has little impact on the temperature of marine waters, except in very limited areas. Exceptions are beaches, tide pools, and shallow water regions very close to a Northern shoreline.

It is possible, that shade may improve the survival of smelt eggs or of other species which inhabit such areas. However, that impact is limited to only those few areas and species and is not applicable to the majority of marine shorelines. Shading is, thus, of considerablly less importance to marine environments than to freshwater ones.

The most important impacts on the microclimate go the other way: In particular, the marine waters tend to keep the littoral zone cool, rather than shading keeping the waters cool.

Insect Fall

Insects which fall from riparian vegetation are recognized as being a major component of the diet of salmonids in freshwater. In contrast, up until relatively recently, insects have not often been found in the gut contents of the juvenile stages of salmonids living in the shallow nearshore marine environment.

One exception is those salmonids which inhabit salt marshes. Insects make up a large part of the fauna of those environments and are a large part of the diet of salmonids which live in salt marshes.

Salmonids learn their feeding behavior. Therefore, they tend to eat whatever is available in whatever environment they inhabit. --- When they live in marine environments, they eat marine species, but when they live where there are insects they are found in their diet.

The increased frequency of observation of insects in the diets of juvenile salmonids in marine waters may be due to a change in the sampling methods and a change in the particular environments sampled. For example, one recent study was conducted in a narrow inlet overhung by trees. Predictably, insects were found in the gut contents of the juvenile salmonids which lived there.

Similarly, several of recent studies conducted on Puget Sound have used beach seines to capture the juvenile salmonids. --- There is a group of salmonids who live near the edge of water, moving in and out with the tide. They generally have insects in their gut contents and that is the case whether they are near the high tide line or in an ealgrass bed during a lower low tide.

The older studies, in contrast, used other kinds of nets and did not preferentially capture this group of juvenile salmonids which seems to focus on the edge of the water, where insects abound.

Thus, this difference between the older studies and more recent studies, possibly reflects a change in methods. --- The new studies demonstrate that some salmonids eat insects, while they are in saltwater. But, they have not established that insects are a major part of juvenile salmonid diets in the marine environment, overall. In contrast, insects are the dominant component of their diet in freshwater. Thus, insect fall is an important function and value in the riparian zone but not in the littoral zone.

Conclusions

Although there are some similarities in the functions and values of the two zones, including their providing filtration and a source of large woody debris, the two zones differ in their most important functions and values: specifically, this includes drainage and shading, but they also differ in the importance of insect fall. Because of these differences, it is not appropriate to treat the littoral zone as a special type of riparian zone nor is it approriate to apply setbacks, buffers and regulations designed for riparian zones to the littoral zone, for they can not be expected to protect the functions and values of the littoral environment where they differ.

Literature Cited

Booth, Derek P. 2004. Forest Cover, Impervious Surface Area, and the Mitigation of Urbanization Impacts in King County, Washington. 18 p. Published as an appendix to, Best Available Science, Vol. 1. Critical Areas, Stormwater, and Clearing and Grading Proposed Ordinances. King County.

Brennan, Jim, and Hilary Culverwell, in press. Marine Riparian: An Assessment of Riparian Functions in Marine Ecosystems. Washington Sea Grant. 41 p.

Levings, Colin and Glen Jamieson 2001. Marine and Estuarine Riparian Habitats and Their Role in Coastal Ecosystems, Pacific Region. Research Document 2001/109, Fisheries and Oceans, Canada, Science Branch.41 p.

May, Chris W. 1996. Dissertation. Civil Engineering, Univ. of Washington. 383p.

Review of Bainbridge Island's Draft Shoreline Management Master Program and the Related Summary of Best Available Science

by

Dr. Robert N. Crittenden

November 2002

Crittenden Biometrical

P.O. Box 157 Carlsborg WA 98324

(360) 582-9550

This report was prepared for  Bainbridge Concerned Citizens.

Cite as, R. N. Crittenden 2002. Review of Bainbridge Island's Draft Shoreline Management Master Program and the Related Summary of Best Available Science, Crittenden Biometrical, P.O. Box 157 Carlsborg WA 98234. 15 pages.

Executive Summary

The draft shorelines ordinance and the associated summary of best available science have several fatal flaws such that the draft ordinance would be invalid or unimplementable if it were adopted as it is. Some of their more serious defects are as follows:

• The Summary of Best Available Science does not identify which results from which scientific studies support each specific article in the draft ordinance.

• The Summary of Best Available Science does not go through the mechanics of demonstrating that the various works it presents as best available science meet the standards for best available science established by rule in the Washington Administrative Code.

• Some of the works presented in the Summary as best available science do not meet those standards.

• The failure to provide a rational justification for the draft ordinance would appear to violate due process, thereby, opening the possibility of litigation under the Federal Civil Rights Act, were the draft Ordinance to be implemented.

• The failure to show a rational basis for the draft Ordinance also opens the door to litigation for violations of due process under the present Shorelines Ordinance, as it is very similar.

• A refusal to issue permits for drainage or structural improvements necessary to stabilize a bluff which results in the risk of the loss of a home through the mass failure or catastrophic erosion of the bluff may give the City employees who are responsible for issuing those permits personal and criminal liability, as the risk to the home necessarily involves the recklessly endanger of its occupants. Should a fatality occur as the result of reckless endangerment, the crime involved is first degree manslaughter. --- This is not mere speculation as the Attorney General's Office wrote a letter to a State employee stating their intention to prosecute him for reckless endangerment if he did not issue permits, in a similar case elsewhere where a failure to issue a permit resulted in a risk to homes and lives.

Overall, my criticisms fall into two categories: 1) The City has not satisfied the requirements for best available science; and 2) The course the draft Ordinance maps out may not be within the City's legitimate authority.

The City should consider abandoning their present vision of the future and looking into alternatives.

Introduction

After 1996, when Bainbridge Island's original shoreline ordinance was enacted, the State Legislature added the requirement that city and county ordinances which deal with critical areas must have a scientific basis and those studies must be explicitly identified. As many of the areas regulated by the Shorelines Ordinances are critical areas, this requirement is applicable to the update of the City's Shorelines Ordinance drafted in 2002.

When the Legislature added this requirement, they also authorized the Office of Community Development to define and set standards for "best available science" and to set those definitions and standards into rule in the Washington Administrative Code (WAC). They can be found in WAC 365-195-900 through 925. Those standards are reiterated below to make it clear exactly what they are.

Criteria for Best Available Science:

Section 365-195-905 of the WAC specifies that "best available science" has the following elements:

1. Peer review. The information has been critically reviewed by other persons who are qualified scientific experts in that scientific discipline. The criticism of the peer reviewers has been addressed by the proponents of the information. Publication in a refereed scientific journal usually indicates that the information has been appropriately peer-reviewed. <Earlier in that section of the WAC further clarifies what is meant by "expert," as follows:> "Whether a person is a qualified scientific expert with expertise appropriate to the relevant critical areas is determined by the person's professional credentials and/or certification, any advanced degrees earned in the pertinent scientific discipline from a recognized university, the number of years of experience in the pertinent scientific discipline, recognized leadership in the discipline of interest, formal training in the specific area of expertise, and field and/or laboratory experience with evidence of the ability to produce peer-reviewed publications or other professional literature. No one factor is determinative in deciding whether a person is a qualified scientific expert."

2. Methods Clearly Stated. The methods that were used to obtain the information are clearly stated and able to be replicated. The methods are standardized in the pertinent scientific discipline or, if not, the methods have been appropriately peer-reviewed to assure their reliability and validity.

3. Logical conclusions and reasonable inferences. The conclusions presented are based on reasonable assumptions supported by other studies and consistent with the general theory underlying the assumptions. The conclusions are logically and reasonably derived from the assumptions and supported by the data presented. Any gaps in information and inconsistencies with other pertinent scientific information are adequately explained.

4. Quantitative analysis is appropriate. The data have been analyzed using appropriate statistical or quantitative methods.

5. In Context. The information is placed in proper context. The assumptions, analytical techniques, data, and conclusions are appropriately framed with respect to the prevailing body of pertinent scientific knowledge.

6. References. The assumptions, analytical techniques, and conclusions are well referenced with citations to relevant, credible literature and other pertinent existing information.

.

.

.

Information derived from one of the following sources may be considered scientific information if the source possesses the characteristics in Table 1. A county or city may consider information to be scientifically valid if the source possesses the characteristics listed in (a) of this subsection. The information found in Table 1 provides a general indication of the characteristics of a valid scientific process typically associated with scientific information.

Table 1

Source

Peer Review

Methods

Logical conclusions and reasonable inferences

Quantitative Analysis

Context

References

Research

X

X

X

X

X

X

Monitoring

X

X

Y

X

X

Inventory

X

X

Y

X

X

Survey

X

X

Y

X

X

Modeling

X

X

X

X

X

X

Assessment

X

X

X

X

Synthesis

X

X

X

X

X

Expert Opinion

X

X

X

X = characteristic must be present for information derived to be considered scientifically valid and reliable

Y = presence of characteristic strengthens scientific validity and reliability of information derived, but is not essential to ensure scientific validity and reliability

<The WAC also provides the following clarification of the various types of sources:>

1. Research. Research data collected and analyzed as part of a controlled experiment (or other appropriate methodology) to test a specific hypothesis.

2. Monitoring. Monitoring data collected periodically over time to determine a resource trend or evaluate a management program.

3. Inventory. Inventory data collected from an entire population or population segment (e.g., individuals in a plant or animal species) or an entire ecosystem or ecosystem segment (e.g., the species in a particular wetland).

4. Survey. Survey data collected from a statistical sample from a population or ecosystem.

5. Modeling. Mathematical or symbolic simulation or representation of a natural system. Models generally are used to understand and explain occurrences that cannot be directly observed.

6. Assessment. Inspection and evaluation of site-specific information by a qualified scientific expert. An assessment may or may not involve collection of new data.

7. Synthesis. comprehensive review and explanation of pertinent literature and other relevant existing knowledge by a qualified scientific expert.

8. Expert Opinion. Statement of a qualified scientific expert based on his or her best professional judgment and experience in the pertinent scientific discipline. The opinion may or may not be based on site-specific information.

Summary of Best Available Science :

In applying these standards to an ordinance, the first step is to explicitly identify which results in which scientific studies support each of its articles. As the criteria for best available science differ for each of the eight different types of studies identified in the WAC, it is necessary to first determine which type each study may be. Then, it can be examined to see if it meets the corresponding criteria.

That is the process which is laid out in the WAC and RCW. It is intended to result in a clear statement of the scientific basis of the ordinances and if properly applied it can provide an adequate systematic approach for determining the quality of scientific research.

However, you will find nothing of the kind in the Bainbridge Island Nearshore Assessment: Summary of Best Available Science.

That report is a general review covering a broad range of topics which bear upon the ecology of the shorelines of Bainbridge Island; it is presented in a non-technical manner probably suitable for students at the seventh to twelfth grade level; and although it may prove to be a useful educational tool, it does not identify which results from which scientific studies support each article of the Draft Shorelines Ordinance nor does it go through the mechanics of demonstrating that each scientific work it cites meets the criteria for best available science.

As a listing of the supporting best available science is required of any city or county ordinance which regulates critical areas, but such a document is lacking in this case, the draft Shorelines Ordinance, would be a nullity, were it adopted, today. It is, therefore, premature for the City to bring the Draft Ordinance to public hearings, before the science upon which it is based has been identified and examined to determine if it meets the standards for "best available science."

Furthermore, many of the studies cited within the Summary of Best Available Science appear not meet the criteria for "best available science."

It cites approximately 200 works, of which only approximately 10% are scientific articles published in respected scientific journals. Most of the remaining works, the bulk of those they cite, are "gray literature." They mostly consist of reports by government agencies, reports by consulting companies to their clients, reviews, books, and personal communications, together with a smattering of other types of publications.

Some of the weaknesses of gray literature are that the researchers doing the work may not have been scientists, the work may not have met scientific standards, and it may not have received independent review by qualified peers of the scientific community. Thus, gray literature are of unknown quality. --- Although some of it may be good work, all of it must be critically examined.

Most reports by government agencies fall into the category of gray literature as they characteristically do in-house review instead of independent peer review. Also, the results they present may reflect departmental positions and policies. In addition, the quality of government publications varies substantially among the various agencies: Some are comparable to scientific journals while others are unreliable. Thus, publication by a government agency can not be regarded as equivalent to scientific publication.

Nor are scientific publications above question. --- Some traditionally reliable elements of the scientific and academic communities have become highly politicized and are no longer reliable. This is part of the reason why the Legislature chose to let the government bodies at the local level make the determinations on what may be valid science. Although they may often lack scientific expertise, they are the level of government closest to the people, where their decisions will have an impact. --- Today, the rule on "science" is that all "science" must be critically examined.

An example, of the "best available science" cited in the Summary of Best Available Science is provided by section 9 of chapter IV of the Summary, entitled " Marine Riparian Zones." I have selected that section because, possibly more than any other section, it speaks to the issue of the ecological importance of Native Vegetation Zones to the nearshore marine environment. Those zones are one of the central elements of the draft Ordinance and will be the topic the November 14, 2002, public hearing or workshop on it.

That section begins by giving two citations to define their terminology and then enters into a brief discussion of the importance of the interactions between the terrestrial and water elements of riparian zones. ⁱ Their first substantive statement is, "Although marine riparian zones have not been subject to the same level of scientific investigation, increasing evidence suggests that riparian zones serve similar functions regardless of the salinity of the water bodies..." The supporting "science" which they cite to support this statement is a work done, in 1994, by Desbonnet et al. It turns out not to be a piece of scientific research, but a review and bibliography; and it was not published in a respected refereed scientific journal but by Rhode Island Sea Grant. Next, that section of the Shorelines Assessment gives a list of all the impacts which they feel alterations of the natural riparian zones may cause, but they cite no scientific support. Finally, at the end of that section, they cite two works which apparently speak to the difficulty of regulating these environments. The first of these is a report to a government task force, while the second apparently has not even been completed, as it is listed as being "in preparation."

Thus, although that section addresses one of the key issues in the Ordinance, it appears to provide no supporting " best available science," whatsoever. The literature they do cite, where they cite any at all, addresses peripheral issues, does not meet the standards of "best available science," or both. --- These problems are symptomatic of the Summary of Best Available Science: Much of the literature they cite appears not to meet the criteria for "best available science."

Although the Summary of Best Available Science covers a broad range of topics, it has missed a few key issues. One is the role of vegetation in establishing interflow; another is the need for technological solutions such as drainage and bulkheads to maintain the integrity of banks and bluffs until native vegetation can reestablish these ecological functions; and a third is technological alternatives to revetments, bulkheads, and seawalls which absorb wave energy instead of reflecting it.

Regarding the first, the roots of vegetation, rot out after the plant dies, leaving a tunnel which can provide drainage. Rapidly-growing short-lived trees such as alders may play a particularly important role in this. Yet they still live around 80 years before they die, so it may take a century or several centuries to fully reestablish the network of subterranean channels which provide effective drainage through interflow.

Until such drainage is fully reestablished, technological solutions may be needed to provide drainage and mechanical support, for if there is a mass failure, the process of reforestation will have to begin over again from the beginning, adding many years to the length of time needed to reestablish these ecological functions and values.

A vertical seawall can reflect virtually all of the wave energy. That would effectively double the wave energy disrupting the benthic community in front of the seawall. Bulkheads and revetments have similar problems if they are in the surf zone. Both the Draft Ordinance and the Summary of Best Available Science ignore the existence of technological solutions to this problem. --- Designs for alternatives which absorb wave energy, other than just "soft armoring," exist and there is a literature on them. Their use should be encouraged in Bainbridge Island's working harbors where boat wakes are unavoidable.

The Summary includes lists of the functions of the various elements of the ecological system. These appear to represent an attempt to list all the functions, some of which are relevant to the objectives of the ordinance and others which are not. Although, they are to be commended for having provided such lists, the substance of the requirement of best available science is that a cohesive rational argument supporting the ordinances be presented which justifies them based on valid scientific knowledge. In that process, the functions are the mechanisms occurring in nature upon which those logical arguments should be based. That is, they provide the model showing why the ordinances may achieve their objectives. Simply providing a shotgun assortment of mechanisms does not serve that purpose.

Best Available Science, Takings, Due Process, and the Present Shorelines Ordinance

Most of the public regard regulations which take part of their property or their right to use it as a "taking" and feel that they are due compensation. However, the standards established by Washington's Counts are that a "taking" occurs only when 100% of the use of 100% a parcel of property is taken for a public use. Thus, in few if any cases would Bainbridge Island's Shorelines Ordinance result in a "taking." ⁱⁱ

What is involved is not a "taking" but an application of the police power for the enforcement of public health and safety. That is how environmental regulations are regarded.

But, the police power can not be indiscriminately applied. It must:

• Be justifiable in terms of serving a legitimate government purpose. --- Best Available Science is a clarification of this: Regulations to promote the health of the environment must be scientifically justifiable.

• Mitigation or regulation must be related to the impact; and

• Mitigation or regulation must be proportionate to the impact.

This puts best available science into its proper context. It is a clarification of the above requirements for due process. According to the review presented in Washington Practiceⁱⁱⁱ, "the affected person must seek damages under the Civil Rights Act, 42 U.S.C.A., 1983.,... ...<but that> denial of due process is actionable only if it is invidious or irrational... or arbitrary."

Best available science avoids these possibilities. --- If the environmental regulation is scientifically justifiable, then it is supported by reason and careful observation: It is not irrational.

The definition of the second key word in that passage, "arbitrary", in the current edition of Black's Law Dictionary is to regulate or govern without an underlying principle. --- Thus, we find the requirement that the regulation be based on explicitly stated functions, values, scientifically established principles and knowledge and serves a know objective guarantees that it is not arbitrary.

And a regulation is "invidious" if it is unduly oppressive, which would be the case if the required action or mitigation were out of proportion to the impact or unrelated to it. --- Thus, again, the requirement that the regulations be rationally justified assures that they are not invidious.

Thus, the requirement for best available science can be seen as a clarification of the requirement for due process. That requirement for due process predates Bainbridge Island's current Shorelines Ordinance, so it, too, is subject to the requirement that the ordinances have a rational basis.

Now, that the City and their contractors have attempted to show the scientific basis for their draft Ordinance, but have failed, and as that draft is not very different from the current ordinance, they have, thereby, provided evidence that neither the current ordinance nor the draft have a rational or scientific basis. Then, it follows that both may violate due process and both are cast into question.

Principle Impact of the Ordinance.

One of the major thrusts of the Draft Ordinance appears to be to prevent the property owners from installing adequate drainage or physical structures which would stabilize the bluffs and banks. The expected outcome will be their erosion in the form of gullies and mass failures. This, in turn, will have two impacts:

The first will be to provide a source of sands and gravels to sustain the beaches. There is no doubt that this will be the result. But, the train of argument gets weaker from there to their legitimate environmental objective of sustaining salmonids and other desirable marine species. The argument seems to be that native vegetation zones will sustain the beaches through erosion, that will sustain the eelgrass beds, and they are invaluable as salmonid habitat. Also, the beaches provide spawning grounds for baitfish, which are prey for older salmonids. Thus, the train of argument goes from preventing the property owners from stabilizing their bluffs and banks to sustaining salmonid populations.

The onus is not upon me to provide a counter argument, but upon the City or their contractors to prove that these processes are critical to salmonid survival and that their ordinances will advance that objective. Nevertheless, a few of the weaknesses in their argument are: 1) There is a lack of evidence that salmonid abundance is limited by the availability of eelgrass beds or by survival through those life history stages. Life history patterns vary widely among the various salmonid stocks and species, but for many stocks the critical periods in their life history are their smolt migration and, later, their survival in the ocean and their survival of fishing pressure. 2) Although young salmonids are known to utilize eelgrass beds, they also utilize other habitats. So, if the eelgrass were replaced by other types of communities, as the beaches are starved of their supply of sands and gravels, salmon may survive comparably well in those other habitats, when they arise. 3) If the pre-Colombian natural condition involved forestation of the banks and bluffs, and that stabilized them, then the beaches in their pre-Colombian condition were starved of sands and gravels; whereas, alternatively, if forestation of the banks and bluffs does not effectively stabilize them, what is the purpose of the natural vegetation zone? Could it be to prevent the stabilization of the banks and bluffs or is it purely matter a of aesthetics? 4) The list of weakness goes on and on... The cause of this is that neither the draft Ordinance nor the Summary advance a clear and cohesive line of argument. They seem to leave that to the reader.

The second impact of preventing the property owners from installing effective drainage or structures to support the banks and bluffs will be to threaten their homes. The only possible course of action, the planners who drafted the ordinance may have reasoned, would be for the property owners to move their homes back from the bluffs. The time horizon they give in the draft Ordinance for making such land use changes for the more than two-thousand parcels involved is 20 years. So, obviously they must feel that the circumstances the Ordinance will create will provide a strong motivation to make those changes.

Personal and Criminal Liability

But, when those homes are threatened, the lives of the residents within them are also put at risk. --- Mass failures can occur with little warning beyond the obvious signs which prompted the property owners to apply for permits to install drainage or supporting structures. When those permits are denied and the bluff fails in the middle of the night, depositing the home in a mud slide into the sea, the residents in that home have little chance of survival. Thus, a failure to issue the necessary permits places the residents lives at risk.

In a very similar situation, the Attorney General's office wrote a letter to a Washington Department of Fish and Wildlife employee stating that they would prosecute him for reckless endangerment if he did not issue a hydraulic permit which would allow several property owners to repair a dike which protected their homes from the Dungeness River. Thus, it is clear that in the opinion of the Attorney General's Office failure to issue a permit which is necessary for a property owner to protect his home and the lives of his family can constitute reckless endangerment.

According to the Revised Code of Washington, when a death results from reckless endangerment, the offense becomes first degree manslaughter.

In addition, if the train of scientific argument from denying permits to enhanced salmonid populations has weaknesses, and it has many, there may be grounds for litigation for violation of due process. Likewise, the same applies to the arguments supporting native vegetation zones as their purpose was not unambiguously stated nor the scientific support provided.

The result is that should the City's employees who are responsible for issuing permits, under the Draft Shorelines Ordinance or possibly even under the existing ordinance, not issue those permits, thereby, placing homes and lives at risk, they may acquire personal and criminal liability.

Consequently, the expected outcome of denying permits may be quite different from the property owners moving their homes back from the edge of the bluffs. It may result in a change of personnel at the City, instead.

If the employees acquire a personal liability, the City probably acquires liability, too. Infact, it may already have acquired some.

Alternative Vision

The course which was taken by the City of Sequim was to allow variances or the purchase the property in cases where their ordinances resulted in regulatory takings. The same approach could be followed here, by allowing variances for drainage or structures or purchasing the property in cases where denying the permits would constitute a criminal act. But, then, the main thrust of the Ordinance would be altered.

My general impression is that the course mapped out by the draft Ordinance can not achieve its intended objective, which appears to be to create a buffer of native vegetation around the entire coastline of the Island and also to remove all artificial structures from that zone. I have doubts whether the Ordinance is intended to serve any other objective, such as preserving salmonid stocks, and there is at this point a lack of evidence that they would accomplish that, even if that were their objective. --- When you have satisfied yourself of the impossibility of the present course, that is will not achieve its objective whichever it may be, you will seriously consider the alternatives.

The alternative involves embracing the changing environment which will result from stabilizing the banks and bluffs and residential development along the shorelines, but planning for it, so that it moves towards a better rather than a worse outcome. What the future may hold, may be a littoral zone with homes surrounded by smaller trees, lawns and shrubs above a coastline of exposed claystone, with a few small beaches and the larger bays with bottoms of sand and gravel. Although, whatever results will be different from the present conditions, that does not necessarily mean that it has to be ecologically or environmental non-productive nor aesthetically unpleasing.

Testimony on the WRIA 18 Watershed Management Plan

by

Dr. Robert N. Crittenden

Crittenden Biometrical

274 Sturdevant Road, Sequim WA 98382

360 582-9550

Oct. 12, 2004

Executive Summary:

Tetra Tech's, 2004, study of the impact which the buildout in domestic wells will have on the Dungeness River has flaws in its methodology and assumptions such that it is not scientifically valid nor does it meet the criteria for Best Available Science.

A corresponding study done by Clallam County's Groundwater Work Group, completed in 1999. is attached. The principle result of that study is that the impact of the full buildout of domestic wells would be so small as be unmeasureable. Two other important results of that study were that if more than a third of the new wells went into a confined aquifer or if more than one-twentieth of the buildout involved the conversion of irrigated agriculture to residential uses, the instream flows would be enhanced rather than diminished.

Finally, I address the question of whether the County Commissioners have the right to delegate authority to the WRIA. --- I had, originally, thought that they did not, but after reading American Jurisprudence on the delegation of authority, my conclusion is that the Commissioners would have that power, provided that three criteria were met:

1. The Watershed Management Plan is narrowly and explicitly defined;

2. The authority of the County Commissioners who follow them in office is not diminished; and

3. The civil rights of the residents of the community are protected.

The plan, in its present form, does not meet these criteria. Therefore, I recommend that it should be remanded to the WRIA for revision.

Introduction:

I testified, on September 21st, on my concerns about the unscientific nature of the studies underlying the WRIA 18 Watershed Management Plan. I appreciate the additional time which the County Commissioners have allowed the public, to examine it. ---- I wish to add several additional issues to my previous testimony.

Tetra Tech's Buildout Study:

The groundwater study completed by Tetra Tech, in 2004, evaluating the impacts which full buildout in domestic wells may have on the Dungeness River, makes several assumptions which are invalid and, also, fails to do the statistical tests which are necessary to establish its scientific validity.

Possibly the most important flaw in their study is that they assumed that the groundwater recharge from agricultural irrigation will be reduced by 10%, to account for the reductions in agricultural irrigation, caused by the conversion of agricultural land to residential uses, but, they failed to make a corresponding reduction in agricultural water withdrawals from the river. As the total rate of water use for agriculture is very much larger than the total rate of domestic use this mistake may be the primary factor causing the model's results. It invalidates the model.

In addition, as agricultural irrigation water is predominately drawn directly from the river, through the irrigation ditches, whereas many of the domestic wells will be at locations remote from the river, agricultural irrigation will have a greater impact on the river for each gallon used than domestic wells will have.

Furthermore, Their assumption of a ten-percent reduction in agricultural uses, does not correspond to full buildout in domestic wells, as full buildout occurs when all land available for that use is put to that use. In that case, the only remaining farmland would be that to which the development rights have been sold or for which domestic use is prohibited for some reason.

Their model uses the structure and parameter values from their 2003 study, but there were a few important changes. The most important of these, may be that they appear to have assumed in the second model that the aquatard which separates the upper unconfined aquifer (the first or top layer) from the first confined aquifer (the third layer) does not form a continuous barrier between those two water-bearing layers. They state this on their page 8. But, in the earlier study, on their page 2-2, they state that, "The upper confining bed is generally about 80 ft thick and consists of silt and clay deposits. It contains locally discontinuous layers of water bearing sand and gravel and ranges in thickness from 5 to more than 175 ft." Thus, their assumption in the second model was not supported by their earlier model.

That assumption is, also, inconsistent with the requirements which the Department of Ecology currently places upon the well drillers, that they drill into the third layer, because it is a confined aquifer. (Personal Communication with a Oasis Drilling.)

This apparently flawed assumption in their later model results in its predicting a greater movement of water between the first and third layers, than may exist. There also appears to be a similar problem with the model's treatment of the movement of water between the third and fifth layers.

Their discussion of their results and particularly their presentation to the public during the October 6th hearing, failed to adequately point out that the large impacts which the model predicts are very localized impacts at the sites of the wells for the municipal water supply systems for Sequim and Carlsborg, but are not found generally over the valley. The drawdown over most of the Valley is relatively small. In addition much of this predicted impact may be attributed to reduced agricultural recharge.

The model also predicts a largish impact directly beneath the City of Sequim. This is probably due to reduced recharge caused by the high proportion of impervious surface. However, contemporary building standards and the use of infusion ponds for roads and parking lots should greatly reduce that impact. This is not adequately accounted for in the model.

Their overall prediction that domestic wells will reduce the instream flow in the Dungeness River by a few cfs, is probably an overestimate, as stated above. But, even if it were not, that small amount they predict is insignificant compared to the WRIA's recommendation of 175 cfs during the low flow period, which occurs during September and October. Such a small impact on instream flows may not even be measurable using the current methods for the measurement of river discharge.

Their study, examining the impact of the buildout in wells, also, presents no statistical tests, confidence intervals, nor other measures of dispersion for their predictions. Consequently, those predictions are meaningless for it remains unknown whether they are relatively accurate, inaccurate, or show nothing at all. Thus, they failed to establish the scientific validity of their results.

These tests or measures of dispersion also a requirement for modeling studies if they are to qualify as "Best Available Science" (see Appendix 1 for the criteria.). The faulty assumptions in the model are other points where their study fails to meet those criteria.

The County is placed at a difficulty when the Watershed Management Plan is not based on Best Available Science, as is the case here, because the County may acquire an obligation to implement the plan by forming ordinances which necessarily affect critical areas, but as the plan is not based on Best Available Science, neither will those ordinances be. That creates a conflict with the Growth Management Act which requires that ordinances affecting critical areas be based on Best Available Science. --- But, perhaps this is for the best, as the result may be that the County would have no obligation to implement the plan.

The Groundwater Work Group's Study:

That report is attached.

It follows a very different approach from the above models:

• It is a stochastic model, instead of a deterministic model;

• It is a mass-balance model, instead of a groundwater model,

• It is deliberately simple in form, because the processes modeled were strong enough that further refinement was not necessary to achieve statistical significance.³⁴ This is in marked contrast against the large and heavily parameterized model which Tetra Tech produced.

• Its results were statistically tested and are highly significant, whereas Tetra Tech did not do statistical tests on their results and their model's heavy parameterization raises doubts as to whether such tests would have been statistical significance had they conducted them and had they also corrected the errors in their model's form.

• It models only the region within a half-mile of the Dungeness River, instead of the entire Dungeness Valley.

This model was initially constructed by myself and was then discussed and modified, for approximately two years, by the County's Groundwater Work Group. It was submitted to the County in 1999. Both of Tetra Tech's models are of more recent origin.

A few of the results of the Groundwater Work Group's study are similar to those of the corresponding model by Tetra Tech. For example, both models indicate that the impact which domestic wells will have on the Dungeness River will be relatively small compared to the recommended flows, during the low flow period.

However, the results of these two models differ on their other results. In particular, in contrast to Tetra Tech's results, the Groundwater Work Group's model indicated that:

• Wells drilled into the confined aquifer will contribute to raising the water table and, therefore, may increase instream flow; and

• The conversion of agricultural land to residential use will produce a large savings in total water use.

The difference between the results of the two models can mostly be attributed to the mistakes in Tetra Tech's model, in how they modeled reductions in agricultural irrigation and the flow of groundwater from the first to the third layers and from the third to the fifth layers.

The Groundwater Work Group's study has different weaknesses. These are that it has not been peer reviewed and it lacks references. --- These problems could be easily overcome, as they are not problems with the study itself, but with its presentation.

Delegation of Power:

Adopting the Watershed Management Plan involves more than just a delegation of authority. It is a transfer of power. --- By accepting the Watershed Management Plan, the County Commissioners transfer their legislative power to form plans to the WRIA and, then, subordinate themselves to them. Thereafter, their authority over land use and water is limited to implementing the WRIA's plans.

At first, I was of the opinion that the Commissioners could not lawfully do this. But, after reading American Jurisprudence on the delegation of power, I reached the conclusion that it might be possible, under certain conditions. These are:

• The Watershed Management Plan is narrowly and explicitly defined;

• The authority of the County Commissioners who follow them in office is not diminished; and

• The civil rights of the residents of the community are adequately protected.

When power is lawfully delegated, that traditionally involves the assigning to some other government body the power to fill in the small details of some plan. That plan has to be fairly narrowly defined. --- The current form of the Watershed Management Plan certainly does not meet that criterion. It will have to be substantially tightened up and made concrete and explicit.

The plan may have to contain an option, so that future County Commissioners can opt out of it. Alternatively, it could simply advise and commit itself to never creating any obligations for the County. Something along those lines would probably be necessary to avoid reducing the authority of future County Commissioners.

However, protecting the civil rights of the residents of Eastern Clallam County, is my principle concern. The issues at question generally are violations of due process. Due process is guaranteed by the fifth and fourteenth ammendments and involves far more than just how meetings are conducted.

A few places where the current process used by WRIA 18 may infringe on due process are:

• The seatholders in WRIA 18 are not elected, but are self-nominating Thus, the public have no choice over who governs them.

• Representation is not proportionate (that is, if it were representation)

• The WRIA is free to govern without a rule. --- Because, House Bill 2514 granted the seatholders immunity to prosecution for their actions within the 2514 process, there is no penalty for violations and, thus, there is no rule.

• Some of their decisions have been irrational. This can be seen in their basing their decisions on invalid studies. --- Science is the strict standard of rationality and "Best Available Science" is the State's interpretation of the due process requirement that all laws, rules, ordinances and their implementation be rational. But, many of the studies which provide the foundation for the plan are neither scientifically valid nor meet the standards of Best Available Science.

• Parts of their plan are unrelated to the problem its seeks to protect against. Two examples follow:

• The watershed management plan, in various places, states that the purpose for protecting instream flows is to "save the salmon." However, the salmon currently have record runs. In addition, most scientists today believe that the salmon crisis was caused by ocean conditions, not by a lack of freshwater habitat, and that a secondary causal factor was hatchery problems. Therefore, it follows that establishing large instream flows can not be expected to materially improve the strength of hte salmon runs. Furthermore, during the public hearing on Washington State's hatchery program, in response to my question, they acknowledged that they are currently keeping the runs in many streams and rivers depressed in order to obtain more data points on the spawner-recruit curve. That is precisely what they should be doing, but other parties should not use the low run levels in those streams and rivers to justify various types of environmental restoration.

• The "optimum" flow predicted by the instream flow incremental methodology (IFIM) is simply the amount of discharge which maximizes the amount of fish habitat in the stream or river. In the West, where there is a dry season, during which the stream or river shrinks down to a much smaller size than its channel can hold, the natural discharge of that stream or river is unrelated to the optimum discharge, for that is determined by the shape and full capacity of the channel. For example, in the Samish River the optimum from IFIM is a little greater than ten-times the median natural flow during the low flow period. Thus, the optimum from IFIM, is unrelated to the natural conditions of the river during the low flow period. That will be true for most rivers in the West, because we have a marine climate, with a dry season. Nevertheless, IFIM is the basis on which WRIA18 recommended the instream flows for the Dungeness River during its low flow period.

• Parts of the plan are out of proportion to the harm it seeks to guard against. For example, the increased water demand due to full buildout in the Dungeness Valley is small relative to the recommended instream flows, yet to acheive this insignificant water savings, the plan will regulate all new wells and quite possibly prohibit most new development.

• and so on ...

The point is that WRIA 18's process is demonstratably "arbitrary,³⁵ irrational and invidious ³⁶.

According to Washington Practice, Volume 24, in its chapter on takings, Demonstrating any one of those characteristics is sufficient to establish that a violation of due process has occured. However, there is more to due process than just that, it basically encompases all that is contained in the concept of fairness.

Unfortunately, it is difficult to see how the civil rights of the public can be adequately protected, as long as HB 2514 continues to guarantee immunity for violations which occur within the WRIA's. Nevertheless, perhaps there is some way to do it. Therefore, I recommend that the plan be remanded to the WRIA to provide them with the opportunity to make the necessary changes.

Literature Cited

Crittenden, R,N. and Groundwater Workgroup 1999. Expected Water Demand from all Future Impacts Domestic Wells along the Dungeness River. Department of Community Development, Clallam County, WA. 8 p.

Tetra Tech F.W. INC. 2003. Dungeness Water Users Association Comprehensive Water Conservation Plan: Groundwater Model Report. Tetra Tech 121pp NE 195th St. Suite 200. Bothel WA 98011

Tetra Tech F.W. INC. 2004 Report: Dungeness Groundwater Modeling Evaluation of Full Buildout in the Dungeness River Area. Tetra Tech, 12100 NE 195th St. Suite 200, Bothel WA 98011. 61p.

Appendix 1: Best Available Science

Section 365-195-905 et sec. of the Washington Administrative Code (WAC) specifies that "best available science" has the following elements:

1. Peer review. The information has been critically reviewed by other persons who are qualified scientific experts in that scientific discipline. The criticism of the peer reviewers has been addressed by the proponents of the information. Publication in a refereed scientific journal usually indicates that the information has been appropriately peer-reviewed. <Earlier in that section of the WAC further clarifies what is meant by "expert," as follows:> "Whether a person is a qualified scientific expert with expertise appropriate to the relevant critical areas is determined by the person's professional credentials and/or certification, any advanced degrees earned in the pertinent scientific discipline from a recognized university, the number of years of experience in the pertinent scientific discipline, recognized leadership in the discipline of interest, formal training in the specific area of expertise, and field and/or laboratory experience with evidence of the ability to produce peer-reviewed publications or other professional literature. No one factor is determinative in deciding whether a person is a qualified scientific expert."

2. Methods Clearly Stated. The methods that were used to obtain the information are clearly stated and able to be replicated. The methods are standardized in the pertinent scientific discipline or, if not, the methods have been appropriately peer-reviewed to assure their reliability and validity.

3. Logical conclusions and reasonable inferences. The conclusions presented are based on reasonable assumptions supported by other studies and consistent with the general theory underlying the assumptions. The conclusions are logically and reasonably derived from the assumptions and supported by the data presented. Any gaps in information and inconsistencies with other pertinent scientific information are adequately explained.

4. Quantitative analysis is appropriate. The data have been analyzed using appropriate statistical or quantitative methods.

5. In Context. The information is placed in proper context. The assumptions, analytical techniques, data, and conclusions are appropriately framed with respect to the prevailing body of pertinent scientific knowledge.

6. References. The assumptions, analytical techniques, and conclusions are well referenced with citations to relevant, credible literature and other pertinent existing information.

...

Information derived from one of the following sources may be considered scientific information if the source possesses the characteristics in Table 1. A county or city may consider information to be scientifically valid if the source possesses the characteristics listed in (a) of this subsection. The information found in Table 1 provides a general indication of the characteristics of a valid scientific process typically associated with scientific information.

Table 1

Source

Peer Review

Methods

Logical conclusions and reasonable inferences

Quantitative Analysis

Context

References

Research

X

X

X

X

X

X

Monitoring

X

X

Y

X

X

Inventory

X

X

Y

X

X

Survey

X

X

Y

X

X

Modeling

X

X

X

X

X

X

Assessment

X

X

X

X

Synthesis

X

X

X

X

X

Expert Opinion

X

X

X

X = characteristic must be present for information derived to be considered scientifically valid and reliable

Y = presence of characteristic strengthens scientific validity and reliability of information derived, but is not essential to ensure scientific validity and reliability

The WAC also provides the following clarification of the various types of studies:

1. Research. Research data collected and analyzed as part of a controlled experiment (or other appropriate methodology) to test a specific hypothesis.

2. Monitoring. Monitoring data collected periodically over time to determine a resource trend or evaluate a management program.

3. Inventory. Inventory data collected from an entire population or population segment (e.g., individuals in a plant or animal species) or an entire ecosystem or ecosystem segment (e.g., the species in a particular wetland).

4. Survey. Survey data collected from a statistical sample from a population or ecosystem.

5. Modeling. Mathematical or symbolic simulation or representation of a natural system. Models generally are used to understand and explain occurrences that cannot be directly observed.

6. Assessment. Inspection and evaluation of site-specific information by a qualified scientific expert. An assessment may or may not involve collection of new data.

7. Synthesis. comprehensive review and explanation of pertinent literature and other relevant existing knowledge by a qualified scientific expert.

8. Expert Opinion. Statement of a qualified scientific expert based on his or her best professional judgment and experience in the pertinent scientific discipline. The opinion may or may not be based on site-specific information.

Two Attachments:

1. Crittenden, R,N. and Groundwater Workgroup 1999. Expected Water Demand from all Future Impacts Domestic Wells along the Dungeness River.

2. Testimony on the WRIA 18 Watershed Management Plan by Robert N. Crittenden Sept. 21, 2004.

Testimony on the WRIA 18 Watershed Management Plan

by

Dr. Robert N. Crittenden

Sept. 21, 2004

I recommend against the County's accepting the Watershed Management Plan because. it does not have a valid scientific basis, is not based on best available science, will create a conflict with the Growth Management Act, can not be expected to achieve its intended objectives, nor is there a mandate for some of its stated objectives.

Before proceeding further, I should say that I am not only a resident and property owner in WRIA 18, but that I am also a scientist with expertise relevant to these issues.³⁷

The first of the flawed studies which I will point out, was the Department of Ecology's, 1997, recommendations for minimum instream flows in WRIA 18's small drainages. These streams include Johnson, Jimmycomelately, Chicken Coop and Dean Creeks. DOE's study used Swift's toe-width method to arrive at their recommendations. Unfortunately, Swift's original studies, in 1976 and 1979, in which he developed the toe-width method, had two fatal flaws. Both are technical errors in statistical methodology. --- The first is that he developed his method using stepwise regression, but failed to discount his alpha-level to account for multiple comparisons. This violates standard methods and can be expected to lead to erroneous results.³⁸ His second mistake was in how he selected streams for his study: He chose representative streams and representative reaches within those streams, rather than selecting them at random. Consequently, his results, if his methodology had been otherwise valid, would have applied only to those specific reaches of those specific streams rather than to all streams in the Puget Sound Region.³⁹ Furthermore, all of the streams and rivers he sampled were larger than those to which his method is being applied in Clallam County.⁴⁰ Because of these mistakes, Swift's toe-width method is not scientifically valid and, consequently, neither are the Department of Ecology's recommendations for the minimum instream flows in small streams in Eastern Clallam County.

In addition, this method was probably also used in other small streams in WRIA 18, wherever an minimum instream flow was desired, but relatively little data on the stream was available. Whereever it was used, it is inapplicable.

Similarly, the minimum instream flow recommendations for the Dungeness River were made using the instream flow incremental methodology (IFIM), but that methodology contains three flaws which invalidate it for this type of application. The first is that IFIM was not intended for this kind of application. Bovee made it clear, in his definitive 1982 review of this methodology, that it was intended for the assessment of small changes not for large ones nor for long-term planning. The problem is that IFIM does not model the movement of the sediment-load of the stream or river (that is the movement of the sands and gravels in the streambed or riverbed.) Consequently, it is only applicable when the changes resulting from the perspective development or management plan are either small enough or occur over a sufficiently short time period that there is no significant change in the stream or river's pattern of erosion, sediment transport, and deposition. The WRIA 18 watershed management plan involves large changes and long-term planning. Consequently, IFIM is not appropriate for this type of application. The second flaw in IFIM is that it involves qualitative input of the species and species composition of the fish whose habitat it models. This qualitative element in the IFIM allows the individuals who input these values to achieve any desired prediction of the optimum instream flow, within a fair wide range of possible outcomes. Most of the seatholders of the WRIA's are probably not aware that this had been done. The third flaw of IFIM is that the "optimum" is only optimum in a purely mathematical sense; it is not necessarily the best nor the desired outcome in an ecological sense nor in terms of human values. Because of these three flaws, the State of Oregon recognized the invalidity of IFIM for setting minimum instream flows and discontinued its use for this purpose.⁴¹

A third key study underlying the WRIA 18 Watershed Management Plan is Blake Thomas et al's 1999, groundwater study for the Dungeness Valley. It is neither scientifically valid nor best available science. Its fatal flaw is that he assumes that the substrate at the southern boundary of his groundwater model consists of a thin layer of soil overlying impervious bedrock. Unfortunately, that bedrock is not impervious, but is fractured basalt and is a relatively good conductor of water. If you wish to verify this for yourselves, you can observe this bedrock at the DNR quarry off Norris Drive, which is a private road off of Sturdevant Road. This quarry is located almost exactly on the southern boundary of his model. The fractures in the basalt are clearly evident, and it is also clear the water flows through them and does so relatively easily. In addition, recent studies have shown that the Dungeness River loses a significant part of its total discharge, in the reach just below the confluence of the Dungeness and Greywolf Rivers, where it flows over fractured basalt. Thus, a critical assumption in Blake Thomas's groundwater study was in error. Consequently, his study is not scientifically valid, nor does it meet the criteria of best available science.

The WRIA 18 Watershed Management Plan recommends that in some, unspecified, parcels, a maximum of 7% impervious surface will be allowed. The studies supporting this recommendation are not listed. However, it is probably based on the work of Dr. Booth of the Department of Civil Engineering at the University of Washington. There is one study by him, presented as an appendix to King County's Best Available Science Review, Volume 1, for their current Critical Areas Code Update, and a second study by Booth and Jackson in a engineering journal. Both are based on the same initial study. --- The single largest impact which urbanization has on streams is increased peak flows. This can result in the severe degradation of streams. Dr. Booth did a study of this phenomenon in a watershed near Issaquah, Washington. He examined how the percentage of development was related to increases in the peak flows and the degradation of the streams. However, his data were collected before ponds for the retention and infusion of stormwater were in common use. He used a hydraulic model to overcome this data gap. His conclusions were that the threshold where severe stream degradation began was at about 10% impervious surface and 65% native vegetation, and that the streams could not be effectively protected by best management practices for roads and parking lots alone. Unfortunately, that hydraulic model is unrepeatable as he failed, in both of his publications, to report his methods of modeling. Consequently, his studies and not scientifically valid nor do they meet the criteria for best available science.

Furthermore, the results which he obtained are unlikely as, according to Dr. Chris May's Ph. D. Dissertation, 65% of the impervious surface in rural King County are roads and parking lots. As they are often ditched directly to streams, whereas other rural imperious surfaces rarely are, it would appear likely that as much as 80% or 90% of the problem of increased peak flows may be due to roads and parking lots. In light of this, the restrictions on impervious surface and clearing are not likely to have much effect, nor are they capable of achieving their intended purpose. That will be achieved primarily by improvements in how stormwater from roads and parking lots are treated.

The Watershed Management Plan will create obligations for the County to create ordinances to implement it, and then to proceed to their application. However, the County can not lawfully do this as the plan, being based on studies which fail to meet to standards for best available science, violates the requirement, in the Growth Management Act, that ordinances regulating critical areas be based on best available science. Thus, the Watershed Plan is not implementible.

Clallam County's Groundwater Workgroup developed a model for the expected impact which full buildout in domestic wells within a half-mile of the Dungeness River would have on instream flows. That is the region throughout which domestic wells might be expected to have an impact. That study was a simple mass-balance model with the refinement of a stochastic component so that confidence limits could the set. Its results were that full buildout is expected to have no measurable effect on instream flows, but that if more than 5% of the parcels converted irrigated agricultural land to residential uses or if more than a third of the domestic wells went into the first confined aquifer, the net impact of full buildout would be positive rather than negative. Consequently, the restrictions on domestic wells which the WRIA 18 Watershed Management Plan seeks to impose, can not be expected to achieve their intended purpose.

Similarly, restoration of the salmon is one of the principle objectives in the WRIA 18 Watershed Management Plan. But, as Robert Lohn, the Northwest Director of National Marine Fisheries Service, pointed out, "Most scientists today agree that the salmon crisis was caused by ocean conditions, not by habitat." ( He meant by "ocean conditions" both harvest and upwelling.) ---- However, it is also generally recognized among fisheries biologists that mismanagement of hatcheries was another important contributing factor. In contrast, freshwater habitat was demonstrably not a major factor. This can be readily seen in the fact that, during the salmon crisis, runs declined in both degraded and pristine streams. Consequently, the WRIA 18 Watershed Management Plan is unlikely to achieve its intended objective of salmonid restoration.

Furthermore, the Growth Management Act mandates the protection and preservation of the functions and values of critical areas and salmonid habitat, it does not authorize their restoration to some previous pre-development level. In the absence of a mandate, the application of the Watershed Management Plan for environmental restoration meets one of the tests for a "taking" Consequently, the County must condemn and buy the properties or uses for restoration, instead of applying the police power for the preservation of public health and safety (which includes the health of the environment). Thus, the Watershed Management Plan commit the County to substantial financial burden.

In conclusion, as the WRIA 18 Watershed Management Plan does not have a valid basis, conflicts with the Growth Management Act, and can not be expected to achieve its intended objectives. I, recommend against its acceptance.

Literature Cited

Bovee, K.D. 1982. A guide to stream habitat analysis using instream flow incremental methodology. Instream Flow Papers 12. USFWS Fort Collins, Colorado FWS/OSD-82/26

Crittenden, R,N. and Clallam County Groundwater Workgroup 1999. Expected Impacts of Full Buildout in Domestic Wells along the Dungeness River.

Swift et al. 1979. Estimation of stream discharges preferred by steelhead trout for spawning and rearing in Western Washington. US Geological Survey Open-File Report 75-155. Tacoma WA

Thomas, Blakemore, Layna Goodman and Therresa Olson 1999. Hydrologic assessment of the Sequim-Dungeness Area. Clallam County, Washington. USGS Water Resources Investigations Report. 99-4048. 165 pages.

Wampler, P. and J. Hiss 1991. Fish habitat analysis for the Dungeness River using the instream flow incremental methodology. USDI Fish and Wildlife Service, Western Washington Fisheries Resource Office, Olympia WA.

Washington Department of Ecology 1997. Flow recommendations for WRIA 18 small drainages.

Expected Water Demand from all Future Domestic Wells along the Dungeness River

by

Robert N. Crittenden and Ground Water Work Group

( The Ground Water Work Group was a citizens' committee formed by Clallam County to examine groundwater issues in the Dungeness Watershed.)

Abstract

The additional exempt wells which would result from completing all of the development allowed under the Growth Management Act within approximately a half--mile of the Dungeness River would be expected to decrease instream flows by approximately 1.1 ± 0.5 cfs. Such a small change in instream flows is less than can be detected by the measurement methods currently used. But, that is a maximum which applies only if all future wells were drilled into the unconfined upper aquafer and all land converted to homesites was previously not irrigated. --- As the proportion of wells into the confined aquifers is increased, the cumulative impact of domestic wells decreases rapidly. In particular, it would be zero if approximately thirty percent of them were drilled into a confined aquafer but, if the proportion were larger than that, the instream flows would be augmented rather than decreased. In addition, each conversion of a five acre parcel of irrigated farmland into a homesite results in a water savings adequate to supply approximately 21 homesites. Thus, if at least five percent of the land converted were previously irrigated farmland, there would be a net savings. In conclusion, the addition of all future domestic wells is expected to have no measurable negative impact.

Introduction

The purpose of this report is to consider the impact which allowable future growth in the number of private exempt wells would have on instream flows in the Dungeness River.

Methods

The approach taken was to estimate the number of such wells and then multiply that by the average rate of consumptive water use. Various corrections were then applied to account for other factors which either increase or decrease the consumptive use. (See Appendix.) An attempt was made to make all assumptions conservative. That is, where in doubt, to err on the side of overestimating the reduction in instream flows. The strategy employed was to use a simple approach and approximate parameter values, but to determine the uncertainty which would result from that imprecision. The feeling of the groundwater group is that the conclusions reached show such strong effects that a higher level of precision is not necessary. Specifically, a Monte Carlo simulation was done to account for the uncertainty in the various parameters. The parameters were assumed to be normally distributed, except for those which are proportions. The latter were simulated using normal distributions with their tails recoded to zero or one. Five--hundred runs were done at each set of parameter values examined. Region Considered: The region considered is that within approximately ahalf--mile of the Dungeness River. At any location more distant from the river than that, it seems unlikely that the small withdrawal rate from an exempt residential well could have any material effect upon the river level. In addition a half--mile is approximately the width of the geological region throughout which the river--bed materials extend. Throughout that region, the river has cut into the geological strata, and then replaced them with river sediments. That is, therefore, the region throughout which there may be some direct connection between the ground water and the river. It is slightly wider in the lower reaches of the river. Buildout: The region modeled was delineated on a county map, generally following sector lines. That allowed parcels and wells to be identified by township, range, sector, and forty acre block in the county's data base. The buildout, that is one minus the ratio of the existing wells to the maximum allowable number, was estimated by randomly sampling 40's in the approximately mile--wide swath along the river and recording the number of logged wells and parcels.

Results

Forty--eight 40's were sampled to estimate the buildout. The mean buildout is 0.5325, with a standard error of 0.0569. Thus, it would appear that slightly less than half of the potential number of wells are already in place. The buildout was approximately uniformly distributed between 0.0 and 1.0, except that a disproportionate number of 40's were either built out completely or not at all. The following parameters were used in the principle set of runs to estimate the impact on instream flows:

Parameter values used in the principle set of runs

Parameter

Mean

Standard Deviation

Total number of parcels

3250

125

Buildout

0.53

0.0569

Gal. per person per day

90

5

Persons per household

2.5

0.25

Proportion consumptive use

0.25

0.1

Irrigation rate (gal. per day)

2000

500

Proportion consumptive use in irrigation

0.33

0.3

Proportion of homes which irrigate

0.33

0.3

Proportion of wells with a riverwater source

0.75

0.13

Proportion of wells which are active

0.9

0.05

Proportion shallow wells

1

0.05

The predicted decrease in the instream flow is 1.09 cfs with a standard deviation of 0.508 cfs. The predictions were approximately normally distributed, but with a slight right skew The system's response to variation in the proportion of deep wells is rapid and reaches zero at about 30 percent deep wells.

Discussion

The system parameters are discussed one--by--one: Proportion of Wells in Active Use: Not all wells will be active at any given time. --- Some of the people who regularly reside in the study area will be away from home some days during the year, and there are also vacation homes which are characteristically unoccupied most of the year. Total Domestic Water Use per Person and Number of People per Household: The estimate of the total domestic water use rate is based on 80--100 gallons per person per day. That is a widely used figure for total water use, including some watering of plants, in urban areas. It would, therefore, be expected to overestimate purely household use. There is assumed to be fewer than three persons per household, which is what the 1990 census indicated for Clallam County.

Total Irrigation Water Use per Household: Each household is currently allowed to irrigate one--half acre. The rate of irrigation required for grass is one acre--inch per acre per week (Sunset Western Garden Book) and two acre--inches per acre every ten days for alfalfa. That is 1940--2715 gal per day. These are both crops with a high water demand. Therefore, 2000 gal per day should be sufficient for most irrigation needs for a half--acre.

Proportion of Homes Irrigating: 100 parcels with homes on them were selected at random and sampled during the late summer to see whether they were irrigating. Almost exactly one--out--of--three were irrigating, and many of those who did water were not doing much more than occasionally sprinkling a patch of lawn. Lush gardens, showing regular watering, were relatively rare, approximately one--in--ten or less. It should also be noted that some of those who watered may have been doing it with water from one of the irrigation ditch systems. No attempt was made to determine if the water came from a well.

Non--Consumptive Versus Consumptive Uses: Household uses are considered to be primarily non--consumptive as a septic tank and drainage field return much that water to the ground water in the unconfined upper water table with little loss. However, the watering of lawns and gardens and most other non--household uses of water are mostly consumptive. That is, part of that water permanently leaves the system. During the late Summer, when low instream flows are a concern, evaporation and evapo--transpirations rates are high. Consumptive uses are assumed to be around 25 percent for household uses and 90 percent of the non--household uses.

Proportion of Ground Water from the River: The river is not the only source of water. A great deal of rain falls directly on each parcel and the ground is quite porous, so much of that rainfall contributes to the groundwater. Assume, for the moment, that the groundwater is a static reservoir, and that 50 percent of the rainfall goes into ground water, which may not be a bad approximation as most of the rain falls during the cool season; then, over the year, a 5 acre parcel which gets 12 inches of rainfall per year, will contribute 108,900 cubic feet per year: (that is 5 acres* 1.0 foot * 0.5 * 43560 square feet per acre) to the groundwater. That is enough to supply 2231 gallons per day, every day of the year. That is approximately equal to the total needs of a household which irrigates and far exceeds purely household needs. However, it is less than the 5000 gal exemption or a household. But, that is just at 12 inches of rainfall per year, which is the minimum in the Sequim--Dungeness region.There is considerable microclimatic variation in the region, with total rainfall generally increasing with elevation. Some parts of the region receive as much as three--times that amount of rainfall. Since most of the study area is zoned for five--acre parcels, and the average parcel size is between three and four acres, obviously, on the average, each parcel is serving as a net water source. Therefore, it is certainly possible that many individuals may be drawing from their shallow wells the water which fell as rain on their own land, or that of their immediate neighbors. To this direct recharge should be added the water which fell as rain on the uplands behind the inhabited area. Direct recharge almost certainly contributes to the groundwater in the region of this study. So, as one considers wells progressively more distant from the river, the probability increases that underground flow at those locations may be dominated by direct recharge rather than being related to the river. The water which would be drawn from those wells may not have come from the river nor be destined for it.

Proportion Shallow Wells: Water from deep wells, that is wells which are drilled into the confined aquifers, was treated as being added to thesystem. The reason for this is that the first aquatard impedes the direct flow of water from the river to the aquifer. Thus, the water which comes to a deep well primarily flowed along the aquifer from a distant recharge area, rather than by a local source. Furthermore, in much of the region of this study, the level of the hydraulic head for the confined aquifers is above the confining layer. Therefore, in those areas the net flow is out of those aquifers and no recharge is occurring there. But, having no information about where the recharge areas for the aquifer may be, we must assume they are equally likely to occur at any location. Since the vast majority of the aquifer and its aquatard are distant from the river, stretching across the entire plain and underlying much of the uplands, a much higher probability must be assigned to those locations, in aggregate, than to the narrow swath along the river. Thus, the best assumption, the assumption which has the highest probability of being correct, is that most of the water in the confined aquifer is not coming from the river.Furthermore, even if it were, there must necessarily be a time--delay to allow the water to flow from the recharge area to the well. As instream flows are only a concern during a critical period of only one or two months during the year, the larger probability (about 6:1 to 12:1) is that it was not withdrawn during a low flow period. Thus, the water in the confined aquifers most likely came from a distant source, outside of the region of concern, either spatially or temporally. The water from deep wells ultimately ends up in the groundwater in the upper unconfined aquifer and contributes to raising its water table, except for the consumptive portion of the water--use which must be subtracted from the water taken from deep wells. Thus, deep wells can benefit instream flows. The system's response to the proportion of wells which go into the unconfined aquifer was examined throughout the range of values for which there is a predicted net decrease in the instream flows. However, the proportion of deep wells, among the new wells now being drilled, is usually fairly substantial in practice, because it promises a higher water quality in the long--run. Therefore, it may be outside of this range, particularly, along those reaches of the river which have more expensive homes. In those areas, the buildout of exempt wells would be expected tobenefit instream flows rather than reducing them. Conversion from Irrigated Farmland: Every five--acre parcel which is converted from irrigated farmland to residential use, saves approximately 19,100 gallons per day. That represents the change from 5 acres irrigated at 2000 gal per day per half--acre to a 5 acre residential parcel with household use and only a half--acre irrigated at that same rate. But, on the average, only about a third of the houses irrigate. Although, only a small fraction of the land converted to homesites was originally irrigated farmland, if only five percent of it were, buildout would result in a net water savings. Of course, the previous irrigation rate as farmland depended upon the type of crop grown. But, the overall concept remains true that the land in its previous use may have had a water demand and some types of irrigated farming require a great deal of water. So, the assumption used in this model, that there previously had been no water demand, is extremely conservative.

Conclusion

The principle conclusion, is that the total impact of the buildout of exempt wells on the instream flows is negligible. At most, it would be about one cfs. However, if there were a significant proportion of deep wells or of land converted from irrigated farmland, there would probably be a net increase in instream flows. Therefore, it follows, that there are probably more productive uses for effort, funds, and good--will than expending them on restricting exempt wells.

Appendix 1: Mass Balance Equation

Q = ( C1 C2 C9 C10 (C6 C8 (C7 C11- (1- C7) (1- C11)) + C1 C2 C3 C4 C9

C10 (C5 C11 -(1- C5) (1- C11))/ ((7.48)(86400))

The first term in the numerator gives the predicted reduction in instream flows due to irrigation and the second term gives the reduction due to household uses. The denominator converts gallons per day to cubic feet per second.

Q = predicted reduction of instream flow in cfs

C1 = total number of parcels

C2 = buildout

C3 = gal total per person per day for household use

C4 = persons per household

C5 = proportion consumptive household use

C6 = gal for irrigation per household per day

C7 = proportion consumptive non--household use

C8 = proportion of homes which irrigate

C9 = prob river water

C10 = proportion of wells which are active

C11 = proportion shallow wells

7.48 gal per cubic foot

86400 seconds per day

Skagit County Commissioners

700 S. 2nd Room 202

Mount Vernon WA 98273

August 8, 2002

Regarding: The 2514 process to set instream flows in the Samish River Basin.

Dear County Commissioners;

I am writing separately from the Unaffiliated Caucus, because, although I contributed to their letter, I have not seen its final form. --- I, therefore, do not know what it may or may not contain.

Nevertheless, like them, I support the County Commissioners' terminating the 2514 process for setting minimum instream flows in the Samish River Basin.

I support this for two reasons: First, because the instream flow study, which is the key piece of research upon which the negotiations should be based, is not scientifically valid, leaving an inadequate basis upon which to make those decisions; and second, because there are many problems with the process due to difficulties in the MOA and groundrules, which can not be satisfactorily resolved except by terminating it.

A more detailed statement of my position is contained in my two reports, 1) A Review of the Instream Flow Assessment- Samish Sub-Basins Lower and Upper Skagit Watershed Plan and 2) The Unaffiliated Caucus's Objections to the New Groundrules.

I apologize for burdening you with additional copies of these reports, I include them in this letter to make it a complete statement.

Sincerely

Dr. Robert N. Crittenden

Alternate Representative

Unaffiliated Caucus

The Unaffiliated Caucus's Objections to the New Groundrules:

1) They disenfranchise the public -- The planning team was direct democracy while it operated under consensus. By "consensus" we mean unanimous consent among all the seatholders. Although direct democracy has some problems, it retains some legitimacy, as the public, can be involved in the process and have some control over its outcomes. In particular, the caucuses' veto-power provided the public with a way to block any outcome to which they did not consent.

That was how the planning team operated until the May 2002 meeting. But, the new groundrules, which were imposed at that meeting, removed the caucuses' veto-power. Under the new groundrules, the influence of the caucuses was reduced to majority vote among the caucuses, and the caucuses were carefully balanced so that is unlikely that the public would have much impact. In addition, two of the caucuses represent government entities instead of the public. That further dilutes the influence of the people upon the process and its outcomes. Under the new groundrules, the power in the planning team is held by the seatholders who represent the initiating governments.

Some decisions, such as those on process, are made by the steering committee, on which the caucuses are not represented at all and, thus, there is no public representation, at all. On these issues their disenfranchisement is complete.

We object to these reductions in the power and influence of the people. We do not consent to them.

2) They lack legitimacy. --- By the State Constitution, all power resides in the people and is derived from them. ---Under representative government, the legitimacy of government is clear, as the officials in each government body are elected by the people who reside within their jurisdiction.

Direct democracy, as was used by the planning team prior to the new groundrules' being imposed, does not have as clear a legitimacy as representative government, for under it, representation is unequal over the community, that is the number of people represented by each caucus is unequal. However, the veto-power of the caucuses largely compensated for this inequality, as the core issue is not equality of representation but equality of power, and the veto-power, being an absolute power which allows a single caucus to block any outcome to which they do not consent, grants each caucus power equal to that of the remainder of the Planning Team. Thus, the disproportionate representation is not as serious an issue under the consensus process as it is under majority rule.

In addition, the presence of an unaffiliated caucus provides representation to all those who are not in one of the other caucuses. Thus, if this system is perfectly operated, every member of the community would be represented with some equality in terms of power over the process and its outcomes. Thus, the veto-power lends the consensus process legitimacy.

Considering how few people actively participate in the political process of electing representatives under our traditional approach of majority rule,the consensus process is roughly comparable, and in some cases may compare favorably, in terms of providing proportionate distribution of power among the members of the community.

But, one weakness of direct democracy, which you can readily see here, is the ease with which it withers away, leaving a dictatorial or arbitrary form of government in its place.

The new groundrules, remove the veto from the caucuses and replace it with majority voting. Thus, that very factor which gives the consensus process its proportionate distribution of power was eliminated. The new process, a hybrid voting mechanism between the initiating governments and the caucuses, is extremely disproportionate in terms of both power and representation. For this reason, it loses its legitimacy.

Furthermore, it is not just "disproportionate" in the traditional sense of an unequal distribution of power among members of the public within the jurisdiction, but involves an extremely unequal distribution of power between the people within the jurisdiction and government entities, including some which are not even located within the jurisdiction.

Returning to the original statement, from the State Constitution, that all power resides in the people and is derived from them, this new process, established by the new groundrules, has no legitimacy because it is not based upon the people, but rests primarily upon government entities.

In conclusion, as the public has little impact on the process or its outcomes, the process has correspondingly little legitimacy.

3) The process which the new groundrules create resembles "arbitrary government." The abuses of this type government were listed in several places in the Declaration of Independence where it gave justifications for the American Revolution. As this is one of the three pieces of organic law upon which our nation is founded, it is probably not within the power of the State nor any of its divisions to establish arbitrary government.

The definition of arbitrary Government which was current at the time of the Declaration of Independence comes from John Winthrop, the Vice-Governor of the Commonwealth of Massachusetts, in 1644. He said that arbitrary government is, "where people have governors set over them without their choice or allowance, who have power to govern them and to judge their causes without a rule." (His article is reprinted in American Historical Documents, The Harvard Classics. Charles W. Eliot (ed.))

None of the seatholders at the table who still hold the veto-power were elected by the residents of the Samish Watershed and the caucuses are virtually powerless. Thus, the first requirement in the definition of arbitrary government is met.

If you consider the process, you will realize that there are few constraints on what the planning team can base their decisions upon. There is the requirement to use best available scientific information, but we will see in the case of Duke Engineering's report on instream flows whether that requirement will be observed or not. They might even consider cultural values in reaching their decisions. Thus, the second requirement in the definition is met, that they govern without a rule.

Fortunately, the planning team does not have any judicial function, therefore, the last part of the definition is irrelevant.

But, I think you can see the similarity between arbitrary government and this process. --- It certainly tends in that direction.

The essence of this complaint, is that under this new process, the residents of the community are effectively being dictated to by a committee, but have no control over its composition or membership and the committee has few constraints on how they make their decisions. In addition, they are not required to follow due process.

4) The purpose of the 2514 process was to allow local planning. That is exactly the opposite of what is happening under the new groundrules, as they disenfranchise the residents of the WRIA and their interest groups (the caucuses), empower government bodies, some not from within that WRIA, and transfer power from the legitimate jurisdictional body, the County, to these other entities.

5) If the planning team is not terminated before it establishes an instream flow, it will continue indefinitely regulating land use and natural resources in the watershed. Thus, unless this process is terminated, the planning team, under its new groundrules will effectively become the form of local government regulating these issues, in perpetuity. Furthermore, if these issues are construed broadly, they may include most activities within the WRIA.

This change in our form of government is of far greater importance than the issue of instream flows.

Governor Gary Locke acknowledged this in speaking to the Seattle Area Chamber of Commerce Leadership Conference, in Vancouver B.C. on October 9, 1998, when he said: "It comes back to our ability to compromise our own narrow interests for the benefit of the common good, and it comes back to our ability to set aside either/or thinking and focus instead on the balance which leads us forward..... The effort to restore wild salmon runs will make history in bigger ways than we usually realize. ... The stakes are very high."

His statement could be construed several different ways, but I think you are familiar enough with this process to realize that he was talking about the change which is occurring all over the State by the creation of planning teams like the one here. And part of that transition is the changes in their groundrules.

The issue of instream flows, justified to restore the salmon, is not the central issue. That is merely a lever to get us to implement and accept this new form of governance. That is why process and the groundrules are the primary concern of the unaffiliated caucus.

6) The disenfranchisement of the residents of the community violates their civil rights: Specifically their rights to elect who governs them, to proportionate representation, and to due process.

No one has the right to make an agreement which compromises the civil rights of another individual, and any such agreement is null and void.

Thus, even if the new groundrules were passed, they are a nullity.

7) The new groundrules are consistent with the decision process described in ESHB 2514, but conflict with the 1998 MOA authorizing Tribal participation. In particular, the new groundrules say that there will be two full attempts to reach consensus, but failing to achieve it, a hybrid voting process will be used, with consensus among the initiating governments and majority vote among the caucuses; ESHB 2514 says the department (Ecology) will attempt to achieve consensus but the final vote on the instream flow will use that hybrid voting process; however, the MOA says that decisions will be made by consensus, where "consensus" means unanimous consent.

To quote from the MOA's section II, Definitions: "Consensus means by unanimous consent."

And its section IV B says,

"Consensus Decision-Making. The Tribe represents a particular unique interest yet forms a minority of all those governments participating in the assessment and planning. In order to protect the effectiveness of the Tribe's participation and to promote a more thorough and cooperative method of determining what the current water resource situation is in each water resource inventory area in furtherance of improved water resource management, see RCW 90.82.005, the parties agree that unless they unanimously stipulate otherwise for a specific situation or type of action, all decisions, including interim decisions, of the parties shall be made by consensus."

Robert Wheeler, pointed out that the term "parties" refers to the initiating governments, who were signatories to the MOA. I agree, as it is consistently used to mean that throughout the document.

But, I do not agree with his conclusion that the paragraph just quoted mandates the use of the hybrid voting procedure by the planning team. Although it says, "all decisions... of the parties shall be made by consensus," the MOA deals with creating the planning team to make decisions on instream flows, most of the decisions of the "parties," dealing with those issues be decisions of the planning team.

The MOA gives the following definition of the planning team which it refers to as a "planning unit" in its section II Definitions: "Planning Unit refers to the initiating governments and those entities, agencies, organizations, or individuals who have accepted an invitation by the initiating governments to join in the watershed assessment process described by ESHB 2514."

Therefore, for the decisions made by the initiating governments as part of the planning unit, to be "consensus," unanimous agreement must be achieved by the planning unit as a whole. --- Thus, I interpret these sections of the MOA to mean that all decisions must by reached by unanimous consent.

Larry Wassermann said, during the July Meeting of the Planning Team that he had participated in the writing of the MOA, and that it was not their intention to require that decisions be reached by consensus.

Nevertheless, it is likely that it was initially interpreted it to mean consensus, as the planning process operated by consensus throughout its first four years.

I take Mr. Wassermann's statement as indicating that some parties had the intention from the very beginning to initially present this process as consensus, but later to reinterpret the MOA and change the decision process to the hybrid procedure. --- Indeed, we see evidence that several parties may have tried to do this, independently, without complete coordination as, first, they apparently attempted to make the shift to it during November 1999 and January 2000, and two and a half years later, under a different facilitator, they attempted to do it that, again, apparently unaware of the earlier attempt.

Regarding the conflict between the groundrules, RCW, and MOA on which decision process will be used, the one in the MOA rules because it is the highest standard. --- In particular, the process in the groundrules meets the standards of the RCW and the process set by the MOA (that is full consensus) also meet those standards because they are a higher standard than the hybrid voting procedure. But, the process set by the groundrules do not meet the standards in the MOA.

The standards in the MOA are required in the 2514 process because the MOA is required by 2514 to initiate and authorize that the process. One must take the entire MOA, not pick and choose which of its clauses to recognize. So, if the clauses initiating and authorizing the process are accepted, so must be all the others. Thus, if the MOA requires full consensus, as I suggest, then that is what must be used.

As you can see in the paragraph quoted above, the MOA provides a mechanism by which the initiating governments can change the decision process. Specifically, the steering committee agrees to make the change by unanimous consent of that committee. Thus, they could have imposed this on the planning team at any time. But, doing so would be flagrantly unpolitic. It appears that they are trying to achieve the consent of the caucuses for this change. But, we strongly object, do NOT consent, and will not consent. --- We may not be able to stop them, but we are not going to bend over to help them do it.

8) There is no evidence that the groundrules were ever passed, but there is evidence that they were not passed. --- Although the summaries from November 1999 and January 2000 say that the groundrules were discussed and passed, and the groundrules in question, then, are the same as now, the summaries were always inaccurate and they are not legal documents. If the summaries had been entered into the public record by submitting them to the County Recorder as legal documents, they would provide legal records of what had transpired. But, as it is, they are only pieces of paper. The difference between such public records and the Planning Team's summaries is that it is a serious crime to knowingly make a false or misleading statement in a legal document, or to alter one. But, it is not a crime to put a false or misleading statement into the summaries or to alter them, or to modify the tape recordings of the meetings.

It is our belief that, sometimes, more was accomplished by making false or misleading statements in the summaries, than occurred in the meetings. That includes the supposed "passing" of the groundrules.

Tom Solberg was present at the November1999 and January 2000 meetings. He recalls the discussion of the groundrules, but does not recall their ever having been formally voted on, and believes that he would have remembered it had it occurred. Thus, the direct evidence the Unaffiliated Caucus has indicates that the groundrules were never passed.

We find it strange that after having supposedly passed the new groundrules in January 2000, the facilitator and planning team waited approximately two-and-a-half years before implementing them, in May 2002.

The "final draft groundrules" were discussed at the April 2002 meeting. We came to the May 2002 meeting prepared to further discuss them and make changes to them. We were then told by the facilitator, Robert Wheeler, that they had been adopted at the previous meeting. No one in our caucus remembered that's happening and Robert Wheeler admitted that there had not been a formal vote on them but that he had had the "feeling" that there had been consensus on adopting them.--- He put that in the summary and, thereafter, treated them as having been adopted.

At the next meeting, we questioned whether the new draft groundrules had been adopted. --- To our surprise, the facilitator and several of the representatives of the initiating governments, now said that they had been adopted not two months earlier, but roughly two-and-a-half years before, in January 2000. I requested that we continue the discussion at the next meeting, so that I could check the old summaries and verify that those older groundrules were the same as their current ones.

It was mentioned in the summary from that meeting that I would be allowed to do so, but under the assumption that I accepted the new groundrules. However, that condition was not mentioned, at the time, so obviously I never consented to it.

Considering that the new groundrules remove the caucuses' veto-power and disenfranchise the public, it is unlikely that the caucuses would ever have voted to pass them. I conclude that if they were passed, it was done by deceit, but most likely, it was never done at all, but was written into the summaries.

In conclusion, we believe that the groundrules were never accepted. But, we do not consent to them, even if they were adopted.

9) The new ground rules prevent us from introducing new scientific studies. We also object to this. This has become an issue because their studies have now been unmasked as junk science and no alternative can be presented.

10) The groundrules would impose a gag order on the planning team members, preventing them from going to the press. This violates their freedom of speech and can not, therefore, be required of team members as a condition for their participation in the 2514 process. As this clause of the groundrules is clearly illegal, and they have no severalty clause, the entire ground rules are null and void, even if they were passed.

Robert Crittenden

Alternate Representative

Unaffiliated Caucus

360 582-9550

A Review of the

Instream Flow Assessment -

Samish Sub-Basins

Lower and Upper Skagit

Watershed Plan

Reviewer:

Dr. Robert N. Crittenden

June 2002

Crittenden Biometrical

P.O. Box 157, Carlsborg WA 98324.

Prepared for the Unaffiliated Caucus of the Skagit River Planning Team; submitted to that planning team on June 26, 2002; revised June 30; and submitted to the Skagit County Commissioners July 1, 2002

Summary

The instream flow assessment is a modeling study. As such, it must meet all six of the criteria set out in the Washington Administrative Code to qualify as best available science.  These are:

1. Peer review: It is not clear that it has been peer reviewed as one reviewer was not independent, it is unknown to me whether the second one was an expert on the relevant topics nor is his review available, and the authors have not yet had the opportunity to respond to this review.

2. Methods: Not only do their methods contain flaws, but some of the standard methods they employed contain flaws.

   1.  Their failure to provide measures of accuracy or to evaluate the propagation of errors can potentially lead to undetected highly imprecise or biased final predictions. How severe these problems may be is unknown until they are evaluated. 

   2. Their use of representative sampling instead of random sampling leads to three problems which occur throughout the study. It is difficult to see how to overcome them as they arise from  their field data having not been properly sampled. These problems are :

      1. Whereas random sampling avoids introducing the biases and preconceptions of the researcher into the data set, representative sampling is based on his or her perception of what is "representative" and deliberately introduces it into the data set. 

      2. They can not validly apply the various relationships and parameter estimates they obtained from those data to the streams over-all, only to the reaches sampled. But they did that anyway. This gives rise to the context problems listed in 5 below. 

      3. They can not estimate the size of the random error component and, therefore, can not compute measures of dispersion such as variances, standard errors, or confidence limits; compute tests of the significance for their estimates; nor evaluate the propagation of errors.

   3. In several places in the study, they said that they adjusted data to improve the fit to their curves.

   4. It is sometimes difficult to understand how they took their samples and did their computations. That fails the criterion of repeatability of methods.

   5. The relationships know as the "toe-width method," which provided the basis for their small stream study, were improperly computed by Swift, in his original study. --- They were never statistically significant.

3. Logical Conclusions and Reasonable Inferences: 

   1. In light of their problems in methodology, they are not yet at the stage to draw any conclusions or inferences. 

   2. An unsupported assumption, they seem to make is that the number of fish is related to the amount of habit, whereas most credible scientists, today, believe that the salmon crisis was caused by ocean conditions, including both upwelling and fishing, not freshwater habitat.

4. Quantitative Analysis: Because of the absence of measures of dispersion, such as variances, standard errors, or confidence limits none, of the model's results have been demonstrated to be significant. --- A roughly equivalent statement is that the study tells us nothing. --- Another related problem is that the propagation of errors was not assessed.

5. Context: The study uses methods or applies results out of the context to which they are appropriate:

   1. The author's use of representative sampling and then applying the results to sites which were not sampled, is a context problem.

   2. The toe width method, which provides the basis for their small streams study, suffers from this same problem, as Swift used representative sampling both of streams and of reaches within streams. The relationships he developed, had they been scientifically valid, would have applied only to those particular reaches which he sampled, but the authors misapplied them to other streams, not from within that group.

   3. The standard errors presented on the authors' page 24 are out of context as estimates of the size of the error component are applicable only to the group of reaches Swift studied in his original toe-width study, not to the authors' study. In addition, the variation in the original study did not estimate natural variation, as Swift used representative sampling instead of random sampling. Furthermore, even if they had been properly sampled in the original study and were applicable, the authors miscalculated them, as they included the random component from the original toe-width study but neglected the variation in the parameter estimates from their study.

   4. The Instream Flow Incremental Methodology (IFIM)  is  used for a purpose for which it was never intended. It was specifically not intended to predict the impacts of long-term policies and is probably incapable of reliably doing so because it has no means for predicting the long-term impacts of management policies on the movements of the bedload. Thus, the IFIM is not applicable for the purposes of the instream flow study nor for providing a basis for setting the minimum instream flows, as these are long-term issues.

6. References: The study does a marginal job of citing references. It has two weaknesses: 

   1. When "expert opinion" is cited, the "experts" did not give references for the science supporting their opinion, and thus fail to meet the standards of "best available science" for "expert opinion"; and 

   2. It is not clear how rigorous the peer review was of some of the agency reports cited, as government agencies characteristically do only in-house review and standards vary among agencies. Thus, some of the reports they cite may not be "best available science."

For these reasons, the study is not best available science nor valid science. 

In addition, some of its problems can not be fixed except by redoing the study in a different way. Examples are 2b and 5a,b, and d.   

It should not be used as a basis for decision making by the Skagit Watershed Planning Team.

Introduction

This is an independent peer review of the Instream Flow Assessment --- Samish Sub-Basins Lower and Upper Skagit Watershed Plan, prepared by Duke Engineering for the Skagit Council of Governments, in March 2002.

It examines their report from two perspectives:

1. To determine if it is " best available science" as defined in the Washington Administrative Code (WAC) 365-195-900 through 925; and

2. To determine if it is valid science.

Their study was intended to provide part of the body of scientific information upon which the Skagit Watershed Planning Team will base their negotiations to establish minimum instream flows for the Samish River Basin. The importance of their report's being " best available science" is that the memoranda of agreement which created and authorized the Planning Team under HB2514 require that they make decisions based on "best available science."

Specifically, section V(E) of the memorandum of agreement among the initiating governments entitled, "Memorandum of Agreement Regarding Tribal Participation in Washington State Water Resources and Watershed Management Programs, specifies the following:

E. Methodologies Used/Data Collection: All methodologies used and data collection shall meet sound, scientific standards, be supported by independent review, and provide an acceptable level of certainty. The parties agree that the methodologies and data collection shall:

1. use the best available scientific information;

2. follow approaches commonly accepted by the scientific community;

3. be agreed upon by the parties; and

4. provide for peer review of its accuracy and adequacy when requested by one of the initiating governments.

These memoranda and House Bill 2514, establish the boundaries within which the planning process must remain. That Act further states that, the individuals who occupy the seats at the negotiating table and the entities they represent are protected from being sued over the outcome of the process. But, by implication, should the planning team fail to remain within the process's established boundaries, for example were they to base their decisions on studies which are not best available science, they might thus acquire liability for whatever negotiated outcomes they might reach.

In addition, the minimum instream flows and the watershed plan, which will come later, will create obligations for the cities and counties to implement those policies. But , as they regulate critical areas, according to the sections of the Growth Management Act, in the RCW, which deal with critical areas, the counties must base their ordinances which deal with those areas upon best available science. Thus, should the planning team base their decisions on studies which are not best available science, they could potentially create obligations for the cities and counties which they could not implement without violating the Growth Management Act.

For these reasons, it is important to determine whether the Instream Flow Assessment reviewed here satisfies the criteria for best available science. In the process of examining whether it meets those criteria, the question of whether it is valid science is also systematically examined.

Criteria for Best Available Science:

Section 365-195-905 of the WAC defines "best available science" as follows:

The characteristics generally to be expected in a valid scientific process are as follows:

1. Peer review. The information has been critically reviewed by other persons who are qualified scientific experts in that scientific discipline. The criticism of the peer reviewers has been addressed by the proponents of the information. Publication in a refereed scientific journal usually indicates that the information has been appropriately peer-reviewed. <Earlier in that section of the WAC further clarifies what is meant by "expert," as follows:> "Whether a person is a qualified scientific expert with expertise appropriate to the relevant critical areas is determined by the person's professional credentials and/or certification, any advanced degrees earned in the pertinent scientific discipline from a recognized university, the number of years of experience in the pertinent scientific discipline, recognized leadership in the discipline of interest, formal training in the specific area o f expertise, and field and/or laboratory experience with evidence of the ability to produce peer-reviewed publications or other professional literature. No one factor is determinative in deciding whether a person is a qualified scientific expert."

2. Methods. The methods that were used to obtain the information are clearly stated and able to be replicated. The methods are standardized in the pertinent scientific discipline or, if not, the methods have been appropriately peer-reviewed to assure their reliability and validity.

3. Logical conclusions and reasonable inferences. The conclusions presented are based on reasonable assumptions supported by other studies and consistent with the general theory underlying the assumptions. The conclusions are logically and reasonably derived from the assumptions and supported by the data presented. Any gaps in information and inconsistencies with other pertinent scientific information are adequately explained.

4. Quantitative analysis. The data have been analyzed using appropriate statistical or quantitative methods.

5. Context. The information is placed in proper context. The assumptions, analytical techniques, data, and conclusions are appropriately framed with respect to the prevailing body of pertinent scientific knowledge.

References. The assumptions, analytical techniques, and conclusions are well referenced with citations to relevant, credible literature and other pertinent existing information.

.

.

Information derived from one of the following sources may be considered scientific information if the source possesses the characteristics in Table 1. A county or city may consider information to be scientifically valid if the source possesses the characteristics listed in (a) of this subsection. The information found in Table 1 provides a general indication of the characteristics of a valid scientific process typically associated with common sources of scientific information.

Table 1

Source

Peer Review

Methods

Logical conclusions and reasonable inferences

Quantitative Analysis

Context

References

Research

X

X

X

X

X

X

Monitoring

X

X

Y

X

X

Inventory

X

X

Y

X

X

Survey

X

X

Y

X

X

Modeling

X

X

X

X

X

X

Assessment

X

X

X

X

Synthesis

X

X

X

X

X

Expert Opinion

X

X

X

X = characteristic must be present for information derived to be considered scientifically valid and reliable

Y = presence of characteristic strengthens scientific validity and reliability of information derived, but is not essential to ensure scientific validity and reliability

<The WAC also provides the following clarification of the various types of sources:>

1. Research. Research data collected and analyzed as part of a controlled experiment (or other appropriate methodology) to test a specific hypothesis.

2. Monitoring. Monitoring data collected periodically over time to determine a resource trend or evaluate a management program.

3. Inventory. Inventory data collected from an entire population or population segment (e.g., individuals in a plant or animal species) or an entire ecosystem or ecosystem segment (e.g., the species in a particular wetland).

4. Survey. Survey data collected from a statistical sample from a population or ecosystem.

5. Modeling. Mathematical or symbolic simulation or representation of a natural system. Models generally are used to understand and explain occurrences that cannot be directly observed.

6. Assessment. Inspection and evaluation of site-specific information by a qualified scientific expert. An assessment may or may not involve collection of new data.

7. Synthesis. comprehensive review and explanation of pertinent literature and other relevant existing knowledge by a qualified scientific expert.

8. Expert Opinion. Statement of a qualified scientific expert based on his or her best professional judgment and experience in the pertinent scientific discipline. The opinion may or may not be based on site-specific information.

The Study is "Modeling":

To apply these criteria, it is necessary to determine which type of study the Instream Flow Assessment may be. Its authors named it an "assessment," but that is only one aspect of it.

They say on their page 1-2, regarding the IFIM study which is the topic of their chapter one, that: "The objective of the IFIM study is to model the relationship between surface water discharge and physical habitat for salmonids in the Samish River, Friday Creek and Silver Creek. " Chapter one constitutes the bulk of their study, and that chapter is, indeed, modeling as the authors state.

Their small stream study, which is their chapter two, involves collecting data from the small streams and applying Swift's toe width method to estimate their discharge. This estimation involves the application of a model. Thus, chapter two also involves modeling.

The objective of their wetland study, which is their chapter three, is as they state on their page 3-1, "to address the relationship between water availability (wetland discharge) and rearing habitat for Coho and Chinook salmon and steelhead trout juveniles in run-of-the-river wetlands within the Samish Basin." That is, they developed a model.

Thus, if the overall study is to be put in a single category, it is relatively clear that it should be regarded as "modeling."

It follows that the study must meet all the criteria listed in Table 1, if it is to be regarded as "best available science."

Peer Review:

The facilitating company for the planning team, Triangle Associates, listed Dr. Hal Beecher, of the Washington Department of Fish and Wildlife (WDFW), as having provided peer review of the instream flow study. He has a PhD and is professionally involved with instream flow for the department, so he bears the appearance of being qualified to be regarded as an "expert." 

However, throughout the study, the authors specifically mentioned having consulted him about the study's design and even of his having "approved" parts of it. In particular, you will find him mentioned as approving the sampling design, on pages 1-13, as approving adjustments to data on page 1-17, as approving methods on page 2-4, and as approving the models used on page 1-29. Thus, he was clearly not "independent," but seems to have filled a role not unlike that of the study's " senior scientist" , providing oversight. 

The legal requirement for independent review, comes from the phrase "other persons" in the first sentence of the definition of "peer review" given in the WAC, and quoted above. It is also the convention in scientific review. 

The authors of the study are to be commended for the care they took to consult experts in designing and conducting their research. But, those same experts can not, then, provide peer review, as they would be reviewing a study which was, in part, guided by their own recommendations. 

You will find his review in Appendix 1-K, together with several other communications and the authors' responses to them. 

The other reviewer listed by Triangle Associates in their email was a water specialist working for the County. His review and the responses to it are not in Appendix 1-K, nor have I obtained copies of them. So, I can not say what criticisms he may have made and whether the authors adequately responded to them.

In his case, I can not say whether he is an expert on the relevant topics: In particular, he does not use the title "Dr." so he evidently does not have a PhD. But, he may have a master's degree. If that is the case, he probably has only limited experience with scientific publication; and If he has only a bachelor's degree, he probably does not have any. His job experience evidently involves water and may, therefore be relevant. Whatever other qualifications he may have are unknown to me. --- I leave it as an open question, which needs to be answered, as to whether he qualifies as an "expert" for the purposes of providing peer review, and if so, on what topics. Another key question is whether he was involved, in any way, in the study, its design, or its authorization.

With regard to this review, my claim to be an "expert" is as follows: I have a masters degree in marine biology from the University of the Pacific, specializing in theoretical ecology and modeling, a Ph.D. in Fisheries From the University of Washington, specializing in various topics including statistics and fisheries statistics; I did a Postdoctoral fellowship at Simon Frazer University on salmon population dynamics; I am clearly a member of the scientific community as I am a member of the scientific society, Sigma Xi, to which admission to membership is based on the excellence of published research, done after the Ph.D.; I have published scientific papers in respected, refereed, international scientific journals on various topics including, but not limited to, salmon population dynamics, statistical estimators used in salmon and fisheries management, modeling, open channel hydraulics, and the microclimatology of small streams; I have also written numerous technical reports; and my first professional job, almost 30 years ago, now, involved the use of a predecessor to the IFIM to assess salmonid habitat in a small stream and the modeling of that stream's physical environment. --- Thus, in my opinion, I meet the criteria to be considered an "expert" in most of the topics covered by the instream flow assessment study.

I am "independent" as I had no involvement in the study being reviewed. I was not even aware of its existence prior to its being presented at the June, 2002, meeting of the Planning Team.

Thus, this critical review meets the criteria to be regarded as "peer review."

Considering the three reviews which have been done, it is not clear, that the instream flow study has met the requirement for peer review, as Dr. Beecher was not independent of the study; the second reviewer may or may not qualify as an expert and it is, at present, unknown what his criticisms were and whether the authors adequately responded to them; and although this review is an independent peer review, the authors have not yet had the opportunity to respond to it.

The substantive issue, is that irrespective of the whether the first two reviewers qualify as independent experts or not, they failed to criticize several flaws, which invalidate the study if they are not corrected.

Methods:

Some of the study's most serious flaws are found in their methods.

The aufthors are engineers and they appear to have followed the conventional engineering approach of using standardized methods. This is specifically allowed by the WAC to satisfy the methods criterion. But, herein, lies a weakness in the rule, as it fails to recognize a difference between the disciplines of engineering and the biological and environmental sciences.

Engineers are trained to use standard methods. This is not unreasonable in the highly determinate world of the physical sciences. There, the underlying relationships are often known very exactly and the data and estimates are characteristically quite precise, often with confidence intervals of no more than 5 or 10%. In contrast, in biology, the underlying relationships are often unknown, the random error component large, and data and estimates with  confidence intervals of plus or minus 30% may be regarded as being "good."

This helps to explain the difference in training and orientation between engineers and biological scientists. The former characteristically use standard methods and have a determinate viewpoint, while the latter may have considerable training in statistics and tailor-make methods of specific applications to better address the unknown in a highly stochastic realm.

The problem is that the State has specifically allowed both the engineering and scientific approaches without providing any guideline as to where each is appropriate, or even a clue that such a difference exists.

In addition, the engineering approach has been widely used among municipal and county governments almost to the exclusion of the alternative. When it comes to local governments' traditional concerns of providing roads, water, sewers, and other services, the engineering approaches is appropriate. But, times have changed. Now counties and cities are required to develop critical areas ordinances, instream flow rules, riparian policies and manage natural resources. That is the other extreme where the stochastic element becomes significant if not dominant.

Understand that I am not criticizing the authors. They are engineers, they have meticulously applied the engineering approach of using standard methods, an approach specifically allowed by the WAC, and supported by the "experts" they consulted. They have done a good and workman-like job, within the standards of their field.

Unfortunately, far too many biologists went into biology because they have poor math skills. This aggravates the problem because they develop, apply, and misapply methods, without much understanding. As a result, some of their methods err, yet some of them have nevertheless been accepted by natural resource agencies as their "standard methods." Then, agency biologists recommend them to engineers, who expect  standardized methods and unsuspectingly apply them. --- The instream flow study has run afoul of this problem.

The instream flow study addresses some physical issues, such as the rating curve for the flow gauge on the river. In these applications, data and relationships are characteristically relatively accurate.

Velocity-discharge relationships are generally not as good, but still remain relatively precise for larger rivers. The log-log regression which the authors used is an appropriate method to estimate it. --- When a river can be modeled as a sheet of water flowing over an infinite plane, Mannings Roughness Formula can be derived from first principles and the velocity-discharge relationship follows a -5/3 power law. That can be fit using a log-log regression. But, traditionally a negative squareroot relationship is used instead. That works relatively well, too.

However, when one considers smaller streams, the shear stresses from the banks or from individual boulders or debris begin to dominate the hydraulics. Then, fitting these curves becomes more problematic. The hydrodynamics of small streams is a difficult topic and it becomes important to estimate the accuracy of the estimates because they may be relatively imprecise.

Likewise, the habitat preferences of the fish tend to have low accuracy. --- One has now moved into the stochastic realm where it becomes necessary to obtain some measure of precision.

The imprecision in the individual components tend to propagate into the rest of the model, affecting the precision of its final predictions. They can also cause biases. This occur whenever an imprecise estimate is used in a nonlinear equation. The worst case is when the confidence intervals for some estimate cross a discontinuity in the model's response surface, causing, for example, a division by zero, then, the confidence intervals for the final predictions from the model will characteristically include either plus or minus infinity and the bias in the predictions may also be large.

In addition, some sophisticated statistical problems can also arise, such as the error-in-covariates problem. But, that is far beyond the level of sophistication in the study. 

The authors have failed to evaluate the propagated errors. This casts the results of the study into doubt.

Another problem is that some of the standard methods they used are flawed and as they are not experts on statistics, they did not detect the flaws nor tailor-make methods to overcome them.

In particular, throughout the study, the authors used representative sampling instead of random sampling. You will find this mentioned on pages 1-13 to 1-17, 1-25, 2-2, and possibly elsewhere as well.

If one randomly samples some set of things and, then, develops an empirical relationship or parameter estimates from those data, they can be applied to any individual thing drawn from within that set. But, if representative samples are taken instead, then the relationship or estimates can only be applied to those specific things sampled.

For example, Swift's toe width method, which is one of the central methods used in their small streams study (their chapter two), was originally developed using a representative sample of streams from the Puget Sound Region. He, also, used representative sampling of the reaches he examined within each of the selected streams. He, then, used stepwise regression to develop several empirical relationships. Those relationships are only applicable to the specific reaches within the specific streams he sampled, not as the authors incorrectly state, to all streams in the Puget Sound Region.

Furthermore, Swift neglected to discount his alpha-level for multiple comparisons. --- For each relationship he developed, he discussed having done at least nineteen comparisons, that is individual statistical tests, searching for one significant relationship. He may have done more tests than that, but those were all he mentioned.

He appears to have done all of these tests at a 95% confidence level. That is an alpha-level of 0.05. The meaning of this criteria, is that if he had done twenty tests, at that level, he would have expected one of them to be significant, simply by chance alone. That one-out-of-twenty would be a spurious relationship, which passed the test but, infact, resulted from no underlying relationship, only random variation.

He did at least nineteen tests at that level and found one which passed the test. That is approximately what would be expected by chance alone.

To overcome this statistical problem, it is conventional to decrease the alpha-level for the individual tests, that is to use a higher standard in the individual tests, so that the entire procedure, overall, has a combined alpha-level of 0.05. --- In particular, if you did twenty tests, you would decrease the alpha-level for each individual test, to 0.05/20 =0.0025: That is, you would do each individual test, of the twenty tests, at a 99.75% confidence level. Then, the final result would have a significance level of 95%.

That was what Swift should have done, but did not do. --- Looking at the results and data which he presented, it appears that his final relationships are not statistically significant at the 95% confidence level, because they do not pass the higher standard for the individual tests.

In addition, he did not present all of his data. Consequently, his stastical procedure can not be repeated using the proper alpha-levels. --- As his computations can not be checked, this, also, fails the standard of repeatability.

Thus, Swift's toe-width method is an example of a flawed standard method. --- It used an inappropriate sampling method, an invalid methodology for developing a relationship, and his final results appear not to be significant. --- The authors' small streams study, their chapter two, is a application of this flawed method. Therefore, their study is not scientifically valid. That, in turn, invalidates the overall study, as it is an integral component of it.

Just as in Swift's method, much of the field data in the instream flow study were representative samples instead of random samples. The parameter estimates and relationships obtained from these field data can not, therefore, be validly applied to the streams and rivers overall, only to those specific reaches selected. But, they nevertheless, applied them to the rivers overall to obtain their overall estimates. Consequently, they are not scientifically valid.

A second problem is that, as the samples were representative rather than random, they do not contain the random error component in nature. Therefore, they can not be used to estimate the size of that error component. The result is that variances, standard errors, confidence limits, and other measures of dispersion can not be estimated with respect to the natural system, nor the river as a whole, nor can statistical tests be conducted on their results. 

Furthermore, representative sampling, by definition incorporates the researcher's perception of what constitutes a "representative" sample, instead of determining what is representative by observing nature. This opens the door to the researcher's either consciously or subconsciously biasing data, and through it, their results. ---- This is not science!

It is difficult to see how to overcome these problems, as they arise from their field data having  been taken improperly.

Here I must admit that their descriptions of their sampling methods were not clear enough to make it absolutely certain what precisely they did. It is possible that they may have sampled at regular intervals, with a random starting point, or some other similar design. ---  Herein is another problem: their description of their methods is not sufficiently clear to repeat the experiment. That, fails the methods criterion of repeatability.

A related problem is that their descriptions of their methods of computation are statistically naive enough that they sometimes raise doubts as to whether they really did what they said they did.  One example is their use of "mean error", instead of "root mean square error" or "standard error." Did they actually compute the mean error? --- Although there are some circumstances where that might be appropriate, it is unlikely that that is what they did, as they were discussing regression models and one of the constraints on regression models is that the errors sum to zero. Thus, the mean error is always zero.

One can guess at what they did, as they evidently did not do what they said they did. But, it should not be necessary for the reader to guess at what methods they employed, the authors are supposed to explicitly describe them. Furthermore, if they made this mistake, and have this level of naievity in statistics, how can one assume that they used the other conventional methods for computing the various other statistically methods they employed?

Their technique for adjusting data to improve the fit of their curves is alarming. You will find it described on pages 1-24 and 1-25 for their flow data and on page 1-17 for habitat data. While I understand that the exclusion of outliers is appropriate under certain circumstances, that methodology needs to be applied sparingly under strict criteria. Also, I understand that some of their flow data needed to be corrected to account for the mechanical performance of a pigmy flow meter. However, that should have been done systematically, not just on points which were identified as outliers. Had they not reported their adjusting their data to improve the fit, it would be fraud. But, as it is, it is just poor "science."

On the whole, their methods could use a thorough scrutiny. But,  these few comments are adequate to reveal enough serious problems with their methods to cast the results of the study into doubt.

Logical Conclusions and Reasonable Inferences:

In light of their methodological problems, it is premature for them to draw any conclusions or inferences.

This criterion is where unsupported assumptions and logical gaps should be listed. There are some.

One is their assumption, which pervades the effort that, fish abundance is related to the availability of habitat. Not only is it not demonstrated that at the present levels of flow, habitat, and fish that there is any such link, but, speaking in general, Robert Lohn, the Northwest Director of National Marine Fisheries, pointed out that, "most credible scientists, today, believe that the salmon crisis was caused by ocean conditions, not habitat." If that is true, and there are strong reasons to believe that it may be, then there may not be a causal link between instream flow and salmonid abundance at the present levels of flow, habitat and abundance. Yet that assumption pervades the effort and the authors have failed to support their position.

Another gap in the study is that it makes no assumptions about the population dynamics of the salmonid stocks modeled.  In the original IFIM report, its author, Dr. Bovee, discusses this. But, it would be difficult to do likewise here as the population dynamics of salmonids are generally unknown. --- The life history patterns of the individual stocks of the various salmonid species have the potential to differ widely and each different life history may have different controlling processes and bottlenecks. --- Still, some general conclusions might have been drawn, such as there being fewer of the older life history stages and probably, also, some estimates may be  available on roughly what the optimum spawning escapement might be.  The authors need to recognize these gaps and  either fill them or explain why it is not possible to do so.

Quantitative Analysis:

I have already mentioned the need for the analysis of the propagation of errors.

I will, dwell longer on the general absence of measures of dispersion: that is variances, standard errors, or confidence limits.

Estimates alone, without these measures of dispersion, tell you nothing unless you can assume that the are relatively accurate. In contrast, when they are not accurate, you must know both the estimate and how much faith to place in it.

Let me clarify this with an example.... Suppose you went to WDFW and asked them how many Chinook adults returned to the Samish River this year, and they told you, "1103." You know that they did not mean exactly 1103, but that that it was an estimate. You might suppose that the confidence interval was roughly plus or minus 10%, that is roughly between 1000 and 1200 fish. But, that was what you assumed, not what you were told. The confidence interval could have been between 0 and 100,000 fish, which would be equivalent to the statement that, "We have only the vaguest idea how many there may have been." So, what did their 1103 mean --- The bottom line is that, if you are given an estimate but no measure of its precision, you have been told nothing.

WDFW pulls this trick all the time. It is something more people should be aware of.

This is also one of the problems with the Instream Flow Study. They have not computed variances or confidence limits. Indeed, they can't because of the problems with their sampling method. Yet, we have every reason to believe that their estimates and predictions are not highly precise. So, in the final analysis, their study has told us nothing.

These measures of dispersion are the basis of statistical testing, so an equivalent statement is that without measures of dispersion, the final estimates and relationships provided by the study have not been demonstrated to be significant.

The standard errors found on pages 2-24 apparently were computed using Swift's toe width relationship. Those are the standard errors to his model, when it was computed using the estimated parameters for the small streams in the instream flow study. The error component represents the contribution from Swift's data set, when he originally fitted the model, it does not incorporate the random error component arising from the estimation of the parameters fitted in the instream flow study. That is to say, the standard errors presented on page 24 are out of context.

Context:

The problem with using representative samples instead of random samples is  a context problem when the estimates and relationships developed based on those data are applied out of context. That occurred when they were applied to the entire stream or river instead of just to the sites sampled.

As already mentioned, above, the toe width method is out of context for the same general reason.

The extrapolation of an empirical relationship beyond that data upon which it is based, is risky as there is no evidence that that relationship continues to be valid in the region of the extrapolation. Yet, you will find on page 1-26 that the authors state that, "The goal of the modeling effort is to be able to model predicted habitat at flows from 40 percent of the low flow calibration measurement and 2.5 times the high flow calibration measurement..."

The IFIM  is, also, out of its proper context. In fact, Dr. Bovee, the author of the IFIM  said, with regard to its use in Washington State, that it is being employed for purposes for which it was never intended. .... This is clear from his description of the philosophical basis of the method on page 2 of his 1982 report. Specifically, he says:

The first and probably most important principle is that implementation of an instream flow regime is inseparable from water management. Therefore the IFIM should be thought of as a water management tool. It is not intended to be an ecosystem model.

But, the IFIM appears to be being used in the study more as an ecosystem model than as a tool.

He went on to say,

The second principle is that the method is not intended to generate a single solution, but to predict the impacts of different alternatives. .... Therefore the user must embrace the philosophy of incrementalism and iterative problem solving before the methodology can be used to its full advantage.

It  most certainly is not being used for incrementalism, that is to predict the impacts of each of many small progressive changes. It is being used  to provide the basis for  long-term planning. 

Bovee recognized the importance of the impact which the hydraulic response of the watershed and land use policies have on the bedload of the stream or river. This must be expected to change its cross-section and, therefore, change the relationship between discharge and salmonid habitat, which is what the IFIM predicts. But, the IFIM has no means for predicting these changes. Therefore, it can only be validly applied to the prediction of  incremental changes are each small enough not to significantly alter the pattern of deposition and erosion.

Allocating the minimum instream flow is a long-term planning issue, creating a regime which may exist for half of a century or longer. Over that time period, it is anticipated that there will be significant development in the drainage basin. For example, one of the principle impacts of development is on the hydraulic response of the watershed. That is why stormwater planning is done. That is important because through changes in the hydraulic response of the watershed, development can have a strong impact on the pattern of erosion and deposition in its streams and rivers. 

Another change which should be anticipated are policies causing or influencing large woody debris (LWD's). These also affect bedload movement and the cross sections of streams and rivers. For example, they not only cause pools, but they cause logjams diverting the river and  they  tend to plug up side channels, leading to a single-channel configuration. A single channel will be wider, deeper, faster, and consequently will have a wider meander belt. Thus, an increase in LWD's will often result in the river's leaving its original river bed and cutting a new channel, with all its destructive consequences.  But, what is relevant here, is that that it will alter the river's cross-section and change the relationship between discharge and habitat. 

If buffers prohibit or limit the management of woody growth, they can provide a source of LWD's and lead to the above consequences.  

As we may expect all of these changes to occur over the long-term, clearly any discharge-habitat relationship developed, now, or at one point in time, will not be valid over the long term. IFIM, therefore, can not provide a basis for long-term planning. And setting the minimum instream flow is a long-term planning issue.

Dr. Bovee also stated a third principle for the IFIM.....

The third principle is that the objectives of any application must be rigidly defined. It is quite possible for two different identical applications of the methodology to result in vastly different solutions, due solely to the objectives of the analysts. For example, two groups may have as their objective, "the design of a flow regime to maintain a fishery at a minimally acceptable level." To one group this really means. "To maximize fish habitat within the constraints of the available water supply." To the other group, the same objective means," To maximize out-of-channel water use without eliminating the fishery."

His last principle says that we should guard against assuming that the study mandates one particular outcome. 

References:

The study does a marginal job of citing the research. 

One weakness is their citing "experts", by name and address, who provided "expert opinion." The opinions they provided are supposed to meet the standards of best available science for "expert opinion." But, they don't meet that standard because their opinions do not include references to the science upon which they are based.  Thus, they never connect to any science, they remain only opinion. 

Another weakness is the government reports they cite. The problem is that government agencies usually do in-house review instead of independent peer review. The quality of their in-house review process varies substantially among the various government agencies. 

In the best case, the publications of a few agencies are indistinguishable in their quality from scientific publications.  

In the worst case, they are unacceptable. State natural resources departments are the principle offenders but you will also find this among some Federal agencies. One of their problems is that, with the exception of a small percentage of their "scientific" staff, their employees may have quantitative skills at no more than a high school or freshman college level. --- I once met an individual who could not do junior high school level  mathematics, but who had risen through the ranks  of one of our Federal natural resources agencies to the level of scientist. ---  The level of competence within the agencies shows in the quality of their publications. 

The rigor of their in-house review process also varies. In some agencies, "review" means no more than getting the person in the next office to read it over. And when the agency is promoting some political program or agenda, that may bias their publications. 

The bottom line is that publication by a government agency  is not necessarily equivalent to scientific publication. Government publications need to be critically reviewed before they can be accepted as science.  

But, that is not to suggest that scientific publications or academia are above question. In fact, one of the things which began the issue of best available science was that many of the traditional scientific sources could no longer be trusted. Today, it is "buyer beware" for all science, but agency reports more than traditional scientific journals, and some agencies more than others.

One of the problems with the references in the Instream Flow Assessment, is that its authors seem to have  uncritically accepted government reports as scientific publications. The unfortunate result was that they based their work on  some flawed studies.

Conclusion

The study appears to not be best available science nor scientifically valid. Its problems probably can not be corrected within the few months remaining for the planning team to set instream flows. That is, if they can be overcome at all, which seems unlikely considering the sampling problems, the problems with the toe-width method, and the context problems with both the IFIM and the toe-width method.

Therefore, the Planning Team should not use the instream flow assessment as a basis for decision making.

Dr. Robert N. Crittenden

Alternate Representative

Unaffiliated Caucus

Addendum to the Critical Review

This addendum summarizes the statements Dr. Hal Beecher made during the July 2002 meeting of the Planning Team, which are relevant to my review of the instream flow study. As you may recall, he was the scientist from the Washington Department of Fish and Wildlife who had overseen that study.

1. He confirmed that "representative sampling" had been used in that study: That is that representative samples were taken instead of random samples. Although it appeared that this was what they had done, it is nice that he confirmed it. He also defended the practice saying, that had they taken random samples, the cost of the study would have been excessive. ---- As I pointed out in my review, this sampling method leads to several serious problems which are irreconcilable without taking new data.

2. He recognized that the IFIM methodology does not model bed load movement and therefore can not predict the changes in the crossectional shape of the stream or river which result from large changes in the discharge or hydraulic response of the river system. For this reason, IFIM's predictions of the relationship between discharge and habitat are not applicable to such large changes.

3. He went on to explain that the cap which they set on withdrawals, of no more than 10% of the median monthly flow, was due to this inability of the IFIM to model bedload transport.

4. During the discussion which followed his presentation, Larry Wasserman, the representative for the Tribes, said that three individuals had worked together to select the species, life history stages, and composition for each month. If I recall correctly, they represented the Tribes, Department of Ecology, and the Department of Fish and Wildlife, but I may be mistaken. He admitted that their procedure was qualitative. --- As these parameters can substantially alter the model's final predictions, this amounts to an admission that the outcomes were at least partially user-selectd.

5. Dr. Beecher did not discuss the toe width method nor chapter three of Duke Engineering's study. In fact, most of his discussion was not on that study at all, but on the IFIM.

6. After Dr. Beecher had finished his presentation, the discussion turned briefly to setting instream flows. Robert Wheeler, the facilitator presented the study's predictions of the optimum monthly instream flows as if they were the model's recommendations and we should accept them. This directly contradicts Dr. Bovee's intended use for the IFIM, as he stated that it is intended to provide predictions on the effects of alternative discharges over a range of values, but does not mandate any particular one.

Footnotes