Full opinion text
OPINION AND ORDER CRABB, Chief Judge. This is a civil action in which plaintiffs seek definition of their treaty-based, off-reservation usufructuary rights to hunt, fish, and gather, and to enjoin defendants from interfering with the exercise of those rights. The plaintiffs are political successors in interest to the bands of the Lake Superior Chippewa who were parties to the treaties of 1837 and 1842. They are recognized by the Secretary of the Interior, and each has a government organized pursuant to the provisions of the Indian Reorganization Act, 25 U.S.C. secs. 461 et seq. Defendants are the State of Wisconsin; the Board of the Wisconsin Department of Natural Resources; Carroll Besadny, Secretary of the Department of Natural Resources; and James Huntoon and George Meyer, administrators in the Department of Natural Resources. All of the defendants are engaged in the promulgation and enforcement of the open water fishing laws and regulations of the State of Wisconsin. In prior proceedings, the Court of Appeals for the Seventh Circuit held that plaintiffs retained their off-reservation usu-fructuary rights to hunt, fish, and gather within the “ceded territory” (that portion of northern Wisconsin ceded by plaintiffs to the United States in treaties executed in 1837 and 1842), and this court has determined the nature and scope of those rights and the basis for state regulation of the rights. The prior determinations by this court made up Phase I, or the declaratory phase, of the case. The present proceedings are part of Phase II, the regulatory phase. To set the stage for what follows, I will reiterate certain relevant principles governing this case that have been established in previous decisions of this court and of the Court of Appeals for the Seventh Circuit. Plaintiffs hold usufructuary rights on the lands they ceded to the United States. Lac Courte Oreilles Band of Lake Superior Chippewa Indians v. Voigt, 700 F.2d 341 (7th Cir.) (LCO I) cert. denied, 464 U.S. 805, 104 S.Ct. 53, 78 L.Ed.2d 72 (1983). These rights include the rights to hunt, fish, and gather all the forms of animal life and vegetation identified in Lac Courte Oreilles Band of Lake Superior Chippewa Indians v. State of Wisconsin, 653 F.Supp. 1420, 1426-1428 (W.D.Wis.1987) (LCO III), using all of the methods of harvesting employed in treaty times and developed since then. Id. at 1435. Plaintiffs may trade and sell to non-Indians the fruits of the exercise of their usufructuary rights. Id. Plaintiffs may exercise their usufructuary rights throughout the ceded territory, except on those portions that are privately owned as of the times of the contemplated or actual attempted exercise of the rights. LCO I, 700 F.2d at 365 n. 14; LCO III, 653 F.Supp. at 1435. Plaintiffs enjoy greater rights to hunt, fish, and gather within the ceded territory than do non-Indians; plaintiffs’ rights are paramount. LCO III at 1429. Plaintiffs’ rights are not exclusive; they hold them in common with non-Indians, United States v. Bouchard, 464 F.Supp. 1316, 1338 (W.D.Wis.1978), rev’d on other grounds sub. nom. Lac Courte Oreilles Band of Lake Superior Chippewa Indians v. Voigt, 700 F.2d 341. It was plaintiffs’ understanding at the time of the 1837 and 1842 treaties that In the absence of a lawful removal order or in the absence of fresh agreement on the part of the Chippewa, the presence of non-Indian settlers would not require the Chippewa to forego in any degree that level of hunting, fishing, and gathering, and that level of trading necessary to provide them a moderate living off the land and from the waters in and abutting the ceded territory and throughout that territory. LCO III, 653 F.Supp. at 1432. Plaintiffs have the right to harvest within the ceded territory so much of the natural resources as is necessary to provide them with a modest standard of living. Id. at 1434-1435. Plaintiffs’ modest living needs could not be met from the presently available harvest within the ceded territory even if plaintiffs were physically capable of gathering, processing, and disposing of that harvest. Lac Courte Oreilles Band of Lake Superior Chippewa Indians v. State of Wisconsin, 686 F.Supp. 226 (W.D.Wis.1988) (LCO V). The State of Wisconsin may regulate the exercise of the plaintiff tribes’ usufructu-ary rights only in the interest of conservation and public health and safety and only upon a showing that the regulation is reasonable and necessary for either conservation or health and safety and that it does not discriminate against the Indian harvest. Lac Courte Oreilles Band of Lake Superior Chippewa Indians v. State of Wisconsin, 668 F.Supp. 1233 (W.D.Wis.1987) (LCO IV). The issue now before the court is the extent, if any, to which the State of Wisconsin may regulate plaintiffs’ exercise of their usufructuary right to harvest walleye and muskellunge within the ceded territory. For the reasons that follow, I conclude that defendants have failed to show that plaintiffs are incapable of regulating their members’ off-reservation walleye and muskellunge harvest, but that defendants have shown that plaintiffs can conduct that harvest without harm to the walleye and muskellunge only if plaintiffs enact and implement certain conservation-based measures and procedures as set forth in this opinion. If plaintiffs do this, then defendants cannot regulate tribal members’ rights and activities in harvesting walleye and muskellunge within the ceded territory, except to the extent agreed to by the parties in their pretrial stipulations. However, if any tribe fails to enact a regulatory plan that conforms to this opinion, or withdraws from the regulatory plan after it is enacted, or if any tribe proves unable to enforce general compliance with the plan by its members, or if tribal regulation of off-reservation harvesting of walleye and muskellunge proves ineffective for any other reason, state regulation of that tribe in accordance with the. standards set forth in this opinion will be sanctioned. I conclude also that defendants have not yet shown a basis for a general allocation of the muskellunge and walleye resource (or of any other resource) between the plaintiffs and the non-Indian users of the resources, and I will reserve a resolution of that issue until further development of the record by the parties. Both sides have filed proposed regulations that would govern plaintiffs’ harvest of walleye and muskellunge. Defendants’ version is in the form of an emergency rule, NR 13, and plaintiffs’ is denominated Off-Reservation Code. Through negotiation the parties have reduced substantially the number of regulatory provisions in dispute. For example, the parties have agreed on the type of documentation required for commercial sales and on the requirement of permits for particular fishing activities. They have agreed that the plaintiff tribes have the capacity to enforce effectively tribal regulations contained in the Off-Reservation Code using tribal enforcement personnel and tribal courts, and that Wisconsin conservation wardens may issue citations to tribal members for violation of regulations conforming to this court’s order. The parties have reached agreement in other areas including food processing, gear marking and identification requirements, and certain biological and other issues. Left for trial is the question of the role to be played by either side in the management of the walleye and muskellunge resources, and a dispute over the following specific regulatory issues: 1) the lakes in which the tribes will be allowed to fish by spearing, fyke and dip netting, seining, and gillnetting; 2) the method of establishing spearing and netting harvest limits on muskellunge; 3) the biological information about any lake required in order to open it to tribal spearing and netting; and 4) the number of walleye and muskellunge that may be harvested by spear or net from any lake. Under defendants’ plan, NR 13, spearing and netting other than gillnetting would be permitted only on lakes over 500 acres; under plaintiffs’ plan, the Off-Reservation Code, spearing and netting would be allowed on lakes of any size, subject to certain conditions. Defendants would prohibit gillnetting on lakes under 1000 acres; plaintiffs would permit gillnetting on lakes of any size. Defendants would require a quota of muskellunge by spear and net on each lake, with permits issued for each tribal harvester; plaintiffs would establish an individual bag limit of two muskellunge a day, with tribal harvest prohibited when the total harvest on any one lake is 50 percent of the safe harvest. Defendants would require a mark-and-recapture population estimate no more than two years old before opening any lake to spearing or netting, and would limit spear or netting harvest totals to a percentage of the estimated number of fish available for harvest; plaintiffs would permit spearing and netting on a lake for which a population estimate has been made by one of a number of methods, including mark-and-recapture, and would limit harvest totals based on the reliability of the estimate used. Defendants contend that their proposed regulation, NR 13, is a necessary and reasonable conservation measure; that it restricts the tribal harvest no more than is necessary to preserve the resource; and that it does not discriminate against the tribal harvest. Plaintiffs disagree. They maintain that defendants have failed to show the necessity for their management plan, and they contend that the plan’s primary thrust is to preserve a non-Indian harvest, rather than the conservation of the species, and that it does discriminate against the tribal harvest. As a basis for resolving these issues, I make the following findings of fact from the evidence adduced at the fourteen-day trial of this part of Phase II. I note, with appreciation for their efforts, that the parties’ comprehensive and focused proposed findings of fact have made this task much easier. FACTS I. NATURE OF THE WALLEYE AND MUSKELLUNGE RESOURCE IN THE CEDED TERRITORY A. Walleye The walleye (stizostedion viteum) is the largest North American member of the per-cidae family. (It is not a pike, although it is often referred to as one.) It is characterized by spiny dorsal and anal fins and large eyes with no iris to shut out light. It lives as long as 20 years and when fully grown, may be as long as 32 inches and weigh up to 15 pounds. (Male walleye tend to be smaller than females.) It is crepuscular in nature, that is, it is particularly active at dawn and dusk. The walleye is a predator that maintains stability in lake communities. It is sometimes referred to as the “keystone” in the ecology of a lake, for the role it plays in improving the lake’s panfish population and water quality. Walleye begin spawning at ages three to six and continue throughout their lives. Ordinarily, the male walleye matures and starts spawning in the third year; the female, in the fourth or fifth years. Because the walleye is a multiple spawning animal (returning to spawn year after year), the production of a particular year class of walleye is not dependent upon a single year class of spawners. (A “year class” is that group of fish hatched in a given year; a “population” is made up of several year classes). Walleye reproductive success is influenced by many factors, one of which is the number of adult spawners. In Wisconsin, male walleye move into the spawning areas in April and May, when the water temperature reaches approximately 47 degrees. They are followed by the females, which deposit their eggs while the males swim alongside and deposit their milt to cover the eggs. Lake spawning walleye congregate at night to spawn over rock or gravel substrata along the shore, in water less than three feet deep. Some lake populations of walleye migrate up tributaries to spawn on rock or gravel bottoms or marsh areas. Walleye tend to return to the same spawning areas each year. Spawning lasts between one to two weeks, although individual females deposit most of their eggs in one night. Various physical factors on individual lakes affect walleye spawning, but as a general rule, spawning starts earlier in the western part of the ceded territory, and moves eastward, so that all the spawning does not occur on all lakes in the ceded territory at the same time. Fertilized walleye eggs incubate for about one month, depending on the temperature. Usually, fewer than 1% of the eggs survive to adulthood. The walleye that hatch each spring (called the “young-of-the-year”) are highly vulnerable to predation and other natural mortality. There is often little correlation between the number of young-of-the-year that are hatched and the number that survive to become “age one walleye.” In some cases the entire year class may fail to survive. Annual natural mortality of a walleye population may range from 5% to 50%, and may be caused by predation, parasitism, or other diseases. Even in unfished lakes, walleye stock may vary dramatically from year to year as a result of climatic influences, episodic events, floods and other natural events. Biologists do not know whether current or future walleye spawning is affected adversely by spearers’ disturbance of the spawning areas or their use of bright lights to help locate the target fish. B.Muskellunge The muskellunge (esox masquinongy) is the largest member of the pike family. It is a long, slender fish, 28 to 34 inches at adulthood, characterized by a stripe like a lightning flash down its back, and very large jaw teeth. It is a solitary, sedentary, predatory fish that aggregates only during the spawning season. It inhabits lakes with broadleaf pond weed, arrow leaf, pick-erelweed and water lilies, and among yellow perch, brown bullheads, white suckers and sunfish. It is a long-lived fish (20 years or more). The male muskellunge matures sexually at four to seven years, the female, at four to six years, depending on location, latitude, and size of the body of water. Muskellunge spawn in the spring in either river systems or marshy areas on vegetation mats in water approximately three feet deep. They usually congregate to spawn about one to two weeks after the walleye spawning season, returning to specific spawning locations. (Generally muskellunge and walleye do not use the same spawning ground.) A large female muskellunge may deposit as many as 300,000 eggs in a season, of which approximately two will survive to adulthood. C.Walleye lakes Within the ceded territory, there are 2,359 lakes greater than 25 acres. (Throughout this opinion I will use “lakes” to refer to any body of water from which walleye or muskellunge can be harvested.) The total area of these lakes is 483,926 acres. Of the known lakes, approximately 861 contain walleye. These tend to be the larger lakes, covering 66 percent of the total lake surface area in the ceded territory, or 350,129 acres. The Department of Natural Resources estimates that there are approximately 408,000 walleye distributed in lakes smaller than 500 acres and approximately 1.2 million in lakes greater than 500 acres. Approximately 40% of the lakes are stocked to supplement natural reproduction; only 6 or 7 percent of all walleye lakes are dependent completely upon stocking for reproduction. There is a loose relationship between the size of a given lake and the size of the walleye population, but populations are highly variable among lakes and even within the same lake. Walleye abundance varies considerably not only from lake to lake, but over time within the same lake. The abundance of adult walleye (walleye age 3 and over) may vary from .2 to 30 per acre in lakes ranging in size from 42 to 15,000 acres. In Escanaba Lake, which is a 300-acre lake used by the Department of Natural Resources for research, the number of adult walleye has varied from 5.2 to 29.4 per acre over a 29 year period. Adult walleye abundance varies as a consequence of reproduction, harvest, natural mortality and environmental factors. In any given lake, there may be more than one stock of walleye present. (A stock is a separate group of walleye that regularly interbreed.) D.Muskellunge Lakes In the ceded territory there are approximately 42,564 muskellunge distributed in lakes smaller than 500 acres, and approximately 46,848 muskellunge in lakes larger than 500 acres. Adult muskellunge populations range from .02 to 2.94 fish per acre. The average adult abundance of muskellunge is about one fish per acre. Shoreline development along the lakes has damaged muskellunge spawning habitat over the last hundred years. II. FISHING METHODS A. Methods at Issue Since the last century, and prior to the judicial recognition of plaintiffs’ usufructu-ary rights, the state prohibited the catching of walleye and muskellunge from off-reservation inland waters by any method other than angling. It has prohibited the use of gill nets on inland lakes since 1878, with the exception of a few, very large lakes, where it has permitted netting for buffalo and carp. These prohibitions no longer apply to plaintiffs when they are exercising their usufructuary rights within the ceded territory. Plaintiffs have been engaged in spearing walleye and muskellunge in the last few years, and they have advised defendants they intend to engage in netting fish, with both “live trap nets” (fyke and dip nets) and “dead trap nets,” such as gill nets. B. Classification of Fishing Methods Fisheries biologists from the Department of Natural Resources and the Great Lakes Indian Fish and Wildlife Commission (GLIFWC) have classified fishing gear in terms of efficiency and selectivity and the level of effort with which they are utilized, and have developed minimum requirements for harvest monitoring, population monitoring, and regulations. In chart form, their classification is as follows: EFFICIENCY EFFORT Low High Low Low probability long-term risk Low probability short-term risk Gear types Unattended lines Set lines/bank poles Ice spearing (walleye) Summer/fall spearing Gillnets Fyke nets Spring seining Weir Harvest Monitoring Periodic Voluntary reporting Mail surveys Short term Total counts Registration Mandatory reporting Regulations Permits Permits Effort restrictions Size restrictions Population monitoring Per generation Relative abundance Length distribution Per generation Relative abundance Age and growth High High probability long-term risk High probability short-term risk Gear types Hook and line Spring spearing Harvest monitoring Annual Statistical surveys Mandatory reporting Registration Daily On-site Total counts Regulations Permits Efficiency restrictions Quota Short-term permits EFFORT Low .High. Size limits Size limits Population monitoring Per generation Relative abundance Age and growth Undecided frequency Absolute abundance Age and growth 1. Angling As the chart shows, angling is an inefficient, high effort method of harvesting fish. Because of its inefficient nature it is largely, but not completely, self-regulating: as fish populations decrease, it becomes harder to catch a fish, and anglers move on to other waters. The catch rate for angling depends on such factors as stock density, availability of other food, and angler skill. Fisheries managers can spread the fish-catching opportunity among more participants by restricting the number of fish caught, the minimum size of the fish caught, and the gear used. 2. Spearing On lakes within the reservations occupied by plaintiffs and recently on certain lakes within the ceded territory, members of the plaintiff tribes have engaged in spring spearing. Spearing is usually done from a boat containing two people, one rowing or guiding the boat and one standing to spear, wielding a fish spear that is 10 to 14 feet long. The spearer uses a light that is either mounted on his headgear or on his boat. Spearing is classified as a high efficiency, high effort method of fishing that tends to select for the larger, mature fish that are on the spring spawning reefs. Harvesting a walleye by spearing takes an average time of five minutes; harvesting a walleye by an angler targeting for walleye takes an average of 9.65 hours. The time of spawning, mercury advisories, water clarity, water surface disturbances and threats of violence by groups opposed to spearing affect plaintiffs’ selection of lakes for spearing and netting. 3.Gillnetting Gillnetting is classified as high efficiency, low effort fishing. Gillnets are set and lifted about once every 24 hours; they require little or no attention between the time they are set and the time they are lifted. The nets are a form of entanglement gear that captures fish as they swim into the net and become entangled. Gillnets are particularly effective for catching walleye and muskellunge because of the fishes’ sharp teeth, and the walleye’s spiny rays, which catch on the nets. In addition, the bottom feeding habits of walleyes and their difficulty in seeing and avoiding the nets during their early morning, late evening feeding make them particularly susceptible to bottom set nets. Also, at certain times of the year, walleye exhibit twilight movements just off shore that enable fishers to set nets to intercept the fish. Muskies often get caught in a gillnet because they attempt to feed on a smaller fish that has been caught in the net and, in the process, bite into the net or otherwise get entangled in it. A gillnet is considered a “dead trap net,” because most of the fish caught in the net will be dead when the net is lifted even if it is lifted daily. The fish die from getting caught with their gills closed, from stress, which builds up lactic acid in their muscles as they struggle to free themselves from the net, or from being eaten by predators such as turtles or larger fish. Even those fish that are alive when the net is lifted may die shortly thereafter from loss of scales and the protective mucous coating on the scales which leaves them vulnerable to bacterial infection, particularly in warm water. The number and kind of fish caught in a gillnet can be regulated to some extent by controlling the length of the net and the size of the openings. However, the use of gillnets results frequently in the incidental catch of species other than the targeted ones. Little is known about intensive gillnet-ting on small lakes. Tribal and state biologists conducted a study of the effects of gillnetting on Escanaba Lake in 1984. From their assessment studies, they derived statistics that they attempted to translate into a curve that would show how vulnerable particular-sized fish are to the mesh being fished. The studies also provided information of the variability surrounding gillnet catches. This information indicated that the bulk of the catches are between zero and ten fish per hundred feet of net, but that there are occasional catches of as many as 45 fish per hundred feet. Although the frequency of such an unusually large catch is less than half a percent, the fact that it occurs indicates both that biologists need more information on the effects of gillnet harvests and that this fishing method poses a significant risk of overfishing. The risk is increased by the fact that the catch cannot be known before the net is lifted, when most of the fish are dead or injured and cannot be returned to the water. Unlike angling, in which population and catch tend to decrease simultaneously, in gillnetting, the catch rate may stay up while the population is decreasing dramatically. This phenomenon has been observed in the perch population in Green Bay. 4. Fyke nets, weir nets, and seining Fyke nets, weir nets, and seining are classified as high efficiency, low effort fishing gear. They are trap nets, and most of the fish caught in them will be alive and uninjured when the nets are lifted. In further discussion of these methods, I will use the term live nets to include fyke and dip nets and seining. 5. Other fishing methods Tribal members participate in other fishing efforts, such as ice fishing with lines, open water hook and line fishing, unattended lines, set lines, bank poles, ice spearing of walleye, and summer and fall spearing. These efforts are classified as low effort, low efficiency fishing methods. By stipulation, the parties have agreed that tribal members using these methods will not be subject to any bag limits, size restrictions, or seasons. However, defendants have proposed subjecting all spearing to minimum sized lakes for which population estimates have been made, whether the spearing is conducted during the spring spawning season or at other times. III. DETERMINING FISH POPULATIONS AND FISH HARVESTS A. Population Estimation It is inaccurate to speak of “measuring” the actual population of any given body of water, because live fish cannot be seen to be counted. Fish biologists can only estimate the biological statistics of a population, that is, the size of the population, the rate of growth, rate of mortality, and exploitation rate, using different mathematical models. These population estimates can then be evaluated on the bases of their precision, their accuracy, and their bias. (Precision is a measure of the variance of the estimate; accuracy is a measure of the likelihood that the estimate approximates the actual population; and bias is a measure of the likelihood that the estimate tends to overestimate or underestimate the population.) 1. Direct methods of estimating population a. Mark-and-recapture The most accurate fish population estimation method now known is based on a “mark-and-recapture” survey. In this kind of survey, two separate “catches” are made of the fish in the lake. After the first catch, the fish are marked, and in the second catch, the number of marked fish is compared to the number of unmarked fish. The catches are made usually with a fyke net during spawning season when the fish are congregated in shallow water; the marking is done by snipping a fin or inserting a tag into the fish’s mouth. Each catch proceeds under a set of assumptions that exist in the statistical framework determining the study. The ratio of marked and unmarked fish in the second catch is analyzed according to a standardized formula, and the outcome is used to estimate the abundance of each species and age group within the lake. (The mark-and-recapture survey is not as accurate for muskellunge as it is for walleye, because the lower relative density of muskellunge produces much smaller survey samples.) The mark-and-recapture survey is subject to a certain level of sampling error. The survey result is an estimate that should be within a certain range of the true population if the survey has been performed correctly. The accuracy of this population estimate declines quickly as time passes. With respect to walleye, for example, the population measured in one year will be almost completely replaced by a different population in two to three years. The cost of doing one mark-and-recapture survey on one lake is about $15,000. The Department of Natural Resources spends approximately $320,000 a year for assessment programs of this kind. GLIFWC has a considerably smaller budget for its assessment programs. It has conducted mark-and-recapture surveys on two lakes in each of the years 1987 and 1988, and intends to conduct two to four such surveys in 1989. b. Abundance sampling Department of Natural Resources biologists have tried sampling young-of-the-year classes in the fall, or yearling classes in either the spring or fall, as a means of determining adult abundance, but the results have not been useful for making population estimates. Such sampling does indicate how many fish are in a lake for different sizes and different species. 2. Indirect methods of estimating population a. Regression model Regression analysis provides a means of determining a relationship between a dependent variable and independent variable for a given population. In fish biology, it may be used to determine the relationship between a known fact, such as the size or water quality of a lake, and an unknown fact, such as the size of the walleye population in the lake. Department of Natural Resources and tribal biologists have tested regression models using average numbers of fish per acre to estimate walleye population on lakes lacking any previous direct (mark- and-recapture) estimate of population. Pri- or to 1988, the state and tribal biologists used a regression model that incorporated any population estimate, regardless of its age, or, if multiple population estimates existed, an average of those estimates. In 1988, the model incorporated all existing population estimation data to produce a more accurate estimate of the average density of walleye populations in relation to the size of the lake. When department and tribal biologists have used the mean confidence level derived from the analysis, their estimates have been inaccurate: they have indicated a larger population than actually existed. Department biologists believe that using the low confidence level (or 95th percentile) would give them a reliable basis for a population estimate. They anticipate also that the accuracy of the estimate can be improved as additional information is gathered about the known facts with which they are working. b. Morphoedaphic index A morphoedaphic index is another form of indirect population estimator that utilizes the size of the lake and the amount of nutrients contained in the lake to derive the total fish yield of the lake. It may be possible to make a separate computation to estimate the biomass of a particular species, but such a computation is still being developed. The morphoedaphic index gives a “factor of two” precision, that is, if the prediction is that a lake would produce one pound of fish per acre per year, in actuality the lake would produce between .5 and two pounds per acre per year. The morphoedaphic index is not suitable for use in lakes that have very rapidly fluctuating water levels in the course of a year, or in which the ions are in inappropriate ratios. It has been used successfully in northern Ontario in opening up for commercial fishing several hundred of the largest 100,000 lakes in a one hundred and fifty thousand square mile area. c. Catch per unit of effort Catch per unit of effort measures relative abundance of a fish population, in terms of number of fish per acre. It can be used to derive an estimate of absolute population by multiplying the surface acreage of the lake by the number of fish per acre. CPE estimates are used to make relative comparisons of population numbers with other lakes or with the same lake in prior years. The index may be used to monitor catch. B. Harvest Measurement 1. Creel survey A creel survey is a statistical sample of the fish harvest in a particular body of water. It can be conducted in a variety of ways: by interviewing all or many of the individual fishers on a lake on any given day, or by requiring all persons fishing on a particular lake to register their catch every day; or by monitoring a particular harvest. Such a survey provides no information that can be used to make population estimates; it is a measure of exploitation only. The survey will allow the detection of significant changes in sport angling effort and exploitation rates from year to year, although it is not effective in detecting small changes of this kind. 2. Creel census A creel census differs from the a creel survey in that it is an actual count of each harvested fish by on-the-site monitors. Both the Department of Natural Resources and GLIFWC have engaged in such censuses of the Indian spearing harvest. IV. FISH MANAGEMENT A. In General 1. Primary and secondary objectives Fisheries managers have as their primary objectives the maintenance or achievement of naturally-reproducing, self-sustaining stocks of fish, and the prevention of the elimination of fish populations through overfishing or other causes. Maintaining a naturally-reproducing population requires preservation of some minimum number of adult spawners in each stock, as well as the minimum number of individuals necessary for genetic diversity. The latter is important, as it appears that small communities of animals tend toward genetic drift — a condition in which the diversity of genetic material is lost with the result that the animals become more vulnerable to extinction from a common disease, climatic change, or other natural occurrence. Biologists have not determined what the minimum genetic number is for walleye and muskellunge populations; the parties to this case have agreed that the basic mínimums for healthy, reproducing walleye populations are three adult spawn-ers per acre, and five year classes of females in a sample or three year classes in a sample of 100 females that each contribute at least fifteen percent of the sample. Secondary management objectives are determined by the kind of fisheries that are being managed. In commercial fisheries, the objective will be the greatest safe utilization of the resource; in a sport fishery, the objective may be a broad spread of number of year classes and a minimal rate of exploitation. Commercial fishers have a strong incentive to take whatever portion of the harvest they can capture, which makes regulating them more difficult than regulating fishers for whom the effort and the trip are often as important as the harvest. 2. Risk avoidance Fish biologists try to prevent overfishing events that lead to the elimination of fish populations, such as those that have occurred in Lake Erie and in Lake Michigan in recent years. Overfishing occurs when total extraction is greater than the harvest-able surplus. Although overfishing in one year does not mean necessarily that the stock will be destroyed or “collapse” the next year, repeated overfishing events pose serious risk to the stock. (In collapse, the number of adult spawners declines, there is recruitment failure (failure of juveniles that reach sexual maturity), and the stock reaches such a low level it cannot be harvested.) Exploitation greater than 50 per cent of an adult fish population on two occasions within five years puts the stock at high risk. The risk is accentuated for walleye because of the variation among year classes, and for muskellunge because of their poor reproductive capacity. Overfishing is difficult to detect unless the stock is monitored scientifically. Catch information may not indicate the strength of the stock: in some situations, record catches have been observed immediately preceding the collapse of a fishery. Overfishing causes population instability in the target stock. Age structure becomes truncated, which can lead to a loss of recruitment. In extreme cases it can lead to collapse of the stock or to total extinction of the species. Overfishing can have ramifications for an entire lake community as well, particularly if the target stock is a predator. If predators are fished to very low levels, another species may assume the predator role and prevent the target species from resuming its function. Small lakes are more vulnerable to overfishing than are larger ones. In small lakes, inaccurate estimates of population and safe harvest pose higher risks simply because of the smaller number of adult spawners and the smaller number of individual fish generally found in the such lakes. Stocking is not a cure for overfishing. Not only is it difficult and expensive to produce the fry and fingerlings for stocking, the stocked fish often have difficulty reproducing in the wild. Biologists think that this is a result of either the hatchery-rearing process, which may deprive the fish of the learned behavior it needs for reproduction, or of a lack of genetic information. Rebuilding depleted populations has been more successful when remnant wild populations exist from which the stocked fish can learn reproductive behavior. Overfishing is not the only risk to fish stocks. Other anthropogenic (human-caused) risks include degradation or elimination of the spawning environment and pollution. B. Fisheries Management in Wisconsin The Department of Natural Resources manages the roughly 14,000 inland lakes within Wisconsin, 11,000 of which are within the ceded territory; the portion of the Great Lakes under Wisconsin’s jurisdiction; and the major river systems within or abutting the state. The fisheries within the state are a combination of recreational and commercial fisheries. Commercial fisheries exist on Lake Michigan and Lake Superior, on some inland lakes (limited to rough fish), and a set line fishery for catfish on the Mississippi River. The state’s non-commercial fishery is limited to angling for recreational purposes. Estimates of its value range from about $500 million to $700 million in direct expenditures by anglers for activities relating to their fishing. Since the middle of the 19th century, the trend in the state’s fisheries management has been to reduce the efficiency of gear that could be used for fishing on the inland lakes. Gillnets and spearing were banned in the 19th century, and later, sales of trout and other species were prohibited. The present management strategy is to disperse the fishery widely and keep it relatively inefficient so that stocks can be sustained. The state’s management objectives for the walleye fishery are the following: 1. To maintain or achieve self-sustaining walleye populations, at abundance levels commensurate with available habitat and food. Indicators of healthy reproducing populations are three adult spawners per acre, and five year classes of females in a sample or three year classes in a sample of 100 females that each contribute at least fifteen percent of the sample. Populations that fall below these levels typically require management actions such as harvest restrictions or stocking. These levels might be referred to as “action levels.” 2. To maintain or achieve walleye populations that optimize fishing opportunities, by increasing abundance or growth rates. Increase abundance in lakes with poor reproduction, over-harvest, or high natural mortality. Increase size of fish in lakes with over-harvest or over-population. 3. To maintain or achieve walleye populations that, through predation, optimize abundance and/or growth rates of other species. 4. To promote stability of desired populations. Given the number of walleye and muskellunge in the ceded territory, and the wide distribution of the two species, it is unlikely that allowing unregulated tribal spearing or gillnetting would result in the extinction of the species. However, it is likely that such fishing could result in the collapse of fish stocks in many lakes. 1. Management of angling fishery Although angling is a high effort, low efficiency method of fishing, the number of individuals engaged in it makes it an intensive fishery, in terms of its effect upon fish stocks. In the years 1980-1987, sport angling effort in the ceded territory targeted at walleye averaged 16,150,710 hours annually, resulting in an average annual harvest of 672,303 fish. On the larger lakes, the angling exploitation is estimated to be 20-25% percent of the fish population. The Department of Natural Resources calculates exploitation rates for sport angling exploitation by conducting mark-and-recapture surveys followed later by creel surveys to determine what component of the population has been harvested. Despite the intensity of the angler fishery, it has not posed a significant threat of overfishing, and what threat it does pose can be dealt with by imposing additional restrictions on bag size, length of seasons, minimum sizes, etc. For example, the Department of Natural Resources will be imposing additional restrictions on angling fishing in the next few years, pursuant to a comprehensive long-term fisheries plan it developed in the late 1970’s. (These restrictions would have been imposed even if the tribes’ treaty rights had not been judicially recognized. It is purely fortuitous that the time for their implementation came shortly after the start up of Indian spring spearing.) Recreational sport angling is allowed on lakes under 500 acres in size. On Escana-ba Lake, the department’s research lake, angling walleye exploitation has exceeded 37% of the population in three of the last four years. The average exploitation has been 28%. Escanaba is a 294 acre lake located within the ceded territory. The department imposes almost no restrictions on fishing in the lake, such as seasons or bag or size limits, except that each fish caught must be registered. Angling tends to harvest younger walleye and muskellunge. Minimum size limits for muskellunge are intended to preserve individual fish so that members may spawn and contribute to reproduction. The management of an angling fishery does not require intensive monitoring when stocks are relatively abundant. The Department of Natural Resources has been successful in preserving and enhancing walleye and muskellunge stock in inland lakes for an angling fishery throughout the state under the current management approach. 2. Tribal treaty fishery a. Operation since 1983 under interim agreements Since plaintiffs’ usufructuary rights were judicially recognized in 1983, the tribes and the Department of Natural Resources have worked together to implement and manage a tribal treaty fishery for walleye and muskellunge. To this end, the tribes and the department have negotiated four interim agreements covering the spring spearing fishery and a gillnet assessment plan. Beginning in 1985, the tribes have engaged in off-reservation spring spearing during walleye spawning season. From these efforts, the department has learned that it is possible to harvest as many as 78 walleye in one hour on one trip; that tribal members have been able to harvest as many fish as permitted, restricted only by the weather and by the number of fish present at spawning; and that the tribal harvest tends to take older fish than the angling harvest. The department has not learned whether the disturbance of the spawning grounds, the use of lights, or any other aspect of spearing has any effect on walleye or muskellunge reproduction. Under the interim agreements, both the Department of Natural Resources and GLIFWC have been involved in monitoring the spearing harvest. The agencies provided monitoring teams that met the spearers at the landings when they returned with their harvest and counted the total catch and also measured and determined the sex of a sampling of the fish. In 1985, plaintiff tribes harvested 2,791 walleye; in 1986, they harvested 6,940 from 31 lakes; in 1987, they harvested 21,321 from 67 lakes; and in 1988, they harvested 25,969 from 92 lakes. In 1985, 1987, and 1988, plaintiffs’ spring walleye harvest was limited to lakes 500 acres or larger. In 1986, their spring walleye harvest was limited to lakes 1000 acres or larger. During 1986, plaintiffs harvested over 80% of the quota provided them under the interim agreement from 20% of the lakes in which they speared. In 1987, plaintiffs harvested over 80% of their quota from 34% of the lakes in which they speared. In 1988, plaintiffs harvested over 80% of their quota from 39% of the lakes in which they speared. Plaintiffs have the capability of spearing 30,000 to 35,000 walleye during the 1989 spring spearing season. Plaintiffs harvested 85 muskellunge in 1985, 55 in 1986, 196 in 1987, and 158 in 1988. Their muskellunge harvesting was limited to lakes of 500 acres or more in 1985, 1987, and 1988, and to lakes of 1000 acres of more in 1986. Under the interim agreements, muskellunge were not targeted for spearing. The muskellunge harvest was incidental only. The resumption of a spearing harvest on inland lakes has created a need for additional, more reliable information about fish stocks because of the tribes’ predicted ability eventually to harvest up to the limits of the quotas negotiated under the interim agreements. It is particularly important to have accurate population estimates because of the impact of intensive fishing methods upon fish stock. Unlike the largely self-regulating angling fishery, the spearing fishery has an immediate effect upon the fish stock and it has no self-regulating features. Safe management of a spearing fishery requires a precise estimation of the number of fish available for harvest, or a “Total Allowable Catch.” Under the interim agreements, state and treaty biologists have used a variety of methods to estimate walleye stock and derive a total allowable catch from which the tribal quota has been set on a lake by lake basis. The parties have developed independent, conflicting proposals for determining total allowable catch and tribal quota on a long term basis. V. MANAGEMENT PROPOSALS FOR TREATY HARVESTING OF WALLEYE AND MUSKELLUNGE A. NR 13 The department has formulated a special administrative regulation covering the usu-fructuary activities of the plaintiff tribes. The measure was approved by the department’s board in February 1988, and became effective with publication on October 5, 1988. (The two subdivisions of NR 13 of particular relevance to this opinion, NR 13.13 and NR 13.14, are reproduced as an appendix to this opinion.) The measure governs all aspects of walleye and muskellunge harvest through two kinds of provisions: those that limit the plaintiffs’ conduct directly and those that set criteria by which the department will manage the resource. It was drafted for a shared, or mixed, fishery. NR 13 leaves almost unrestricted the plaintiffs’ use of less efficient harvest methods, such as set poles and lines, bank poles, hook and line and ice fishing. Under the proposal, tribal members may catch as many fish by less efficient methods as they want, from whatever public access lake or stream they may choose, unrestricted by seasons, bag limits or size limitations. Spring spearing, netting, and snagging are governed by a quota system similar to the process used under the interim agreements. Under NR 13.14, spearing is limited to lakes greater than 500 acres; gillnetting is limited to lakes greater than 1000 acres. No spearing or gillnetting may occur on any lake of whatever size unless the lake has had a population estimate made within the previous two years. The estimate must be either a mark-and-recapture survey or another method agreed upon by the parties. One of the purposes of requiring that a lake have a population estimate no more than two years old is to ensure that lakes not be subjected to efficient fishing methods more than two years in a row, without the opportunity for the department to assess the impact of the fishery on the particular lake. NR 13 sets a tribal quota (TQ) of 20% of the total allowable catch (TAC). The quota works as follows: beginning with a theoretical population in the lake (the population estimate), the biologists apply an exploitation rate of 35% for walleyes and 27% for muskies that has been stipulated to by state and tribal biologists. (These rates are subject to review and revision as scientific information about the fish stocks permits more precise management.) The product of the population estimate times the exploitation rate is the Total Allowable Catch. The concept of a Total Allowable Catch is used worldwide in the management of efficient fisheries. It represents the number of fish in excess of the number needed for spawning or, in other words, that component of the fishery not needed for reproduction. If the population of a particular lake were known, the Total Allowable Catch could be harvested without risk to the fish stock. For lakes that have a one-year old mark-recapture survey, defendants use a safety factor of 35% to derive the safe harvest for walleye. For lakes that have a two-year old mark-recapture survey, defendants use a safety factor of 30%. For lakes that have had a mark-recapture survey done within the preceding two years, defendants use a safety factor of 35% to derive the safe harvest for walleye. (Another way of achieving the same result is to do as defendants have done in NR 13: allocate plaintiffs 20% of the Total Allowable Catch (TAC being population estimate (PE) times exploitation rate of 35%). The resulting safe harvest estimate is then divided between the high efficiency methods used by plaintiffs and the low efficiency angling harvest. Thus, in a lake estimated to contain 100 walleyes, 12 walleye (100 X .35 x .35) could be caught without unduly risking the viability of the stock. Of these, NR 13 would make 7 available for spearing and 5 for anglers. In the real world, only the statistically average lake contains 100 walleyes; using defendants’ method of calculating the population estimate on the basis of a mark-and-recapture survey, the probability is that the lake contains no fewer than 40 and no more than 170. In the lake with the smallest number of fish, a harvest of 35 would leave only five fish in the lake. On the other hand, in the lake with the largest number, a harvest of 35 would leave 135 fish unharvested that could have caught without risk to the stock. Defendants made the decision to limit plaintiffs’ harvest by efficient fishing methods to 20% of TAC for several reasons: they believed it would provide plaintiffs with approximately the same number of fish that plaintiffs had been taking under the interim agreements; it provided a margin of safety for a fishing method whose effects were still largely unknown; and it was a means of allocating the fish in a mixed fishery. If plaintiffs were permitted to take the total safe harvest, that is, the product of the safety factor times the exploitation rate times the population estimate, or 12 walleye from a lake estimated to contain 100) of any species from any lake, it would be necessary to close that lake to further fishing for that species by any method by Indians or non-Indians. NR 13 contains no requirement that the department make a minimum number of mark-and-recapture surveys in any given year, or that it make any at all. The rule contains no provision for population estimates made by any other method, although it does provide that other standardized sampling methods may be developed and used “by agreement with the tribes.” NR 13.13(d). It leaves to the discretion of the department the choice of which lakes to survey. (However, during the course of the trial defendant’s counsel stated that it had no objection to a court order requiring it to conduct a certain number of surveys each year.) NR 13 provides no method of negotiating or reviewing the calculations used by the department to determine Total Allowable Catch. It does not provide explicitly for tying the amount of tribal quota to the amount of harvest sought by plaintiffs. B. Tribal Code and Management Plan 1. Tribal organization Each plaintiff tribe is a member of the Great Lakes Indian Fish and Wildlife Commission, which is an inter-tribal, interstate organization operating under a constitution ratified by each member tribe. It was established in 1984, and has three committees organized by the geographic area of the member tribes’ treaties. One of these is the Voigt Inter-Tribal Task Force Committee, which directs the commission’s activities related to the land and resources in the 1837 and 1842 cession areas and is composed of representatives of the eight tribes with reservations located within the area ceded by these treaties in Wisconsin, Michigan and Minnesota. The member tribes have delegated to GLIFWC the authority to employ conservation wardens to monitor and enforce the off-reservation conservation regulations enacted by the member tribes. The Voigt Task Force has primary responsibility for intertribal co-management and regulation of off-reservation resources. It reviews and approves management plans, develops and recommends seasonal agreements with the Wisconsin Department of Natural Resources and the United States Fish and Wildlife Service and recommends implementing regulations to the plaintiffs. GLIFWC provides biological and resource management support services to the Voigt Task Force to assist the task force in carrying out its responsibilities. • Each plaintiff tribe has authority to enact harvest regulations, which may be more, but not less, restrictive than the recommendations of the Voigt Task Force. The combined harvest of any species by the members of the plaintiff tribes is limited to the overall harvest goals and quotas for each species approved by the task force and developed by the commission staff. GLIFWC’s Biological Services Director has authority to terminate the combined harvest of any species whenever it is the director’s judgment that continued harvest is likely to exceed the harvest goals or quotas adopted by the task force or would otherwise result in biological harm. The task force establishes harvest goals and quotas annually for walleye and muskellunge, relying upon the recommendations of GLIFWC’s Biological Services Division. The task force sets a harvest level for walleye for each body of water on which harvest by any method is contemplated. It bases the harvest levels upon the determination of the Biological Services Division of those off-reservation waters within the ceded territory that are capable of sustaining a harvest of walleye during the fishing year. In June 1988, GLIFWC drafted a Management Plan and Model Code of Ordinances Governing Off-Reservation Harvest by Tribal Members. The plaintiff tribes have approved the tribal management plan through their Tribal Governing Bodies, and two of the tribes have enacted it. Under the Model Code, GLIFWC’s staff will transmit harvest and other biological data to the Department of Natural Resources. Commission personnel will consult representatives of the department regarding review of harvest goals, management plans, seasonal agreements, and emergency closures, and will participate in joint technical working groups and in the review and comment on department management plans and regulations. 2. Structure of Management Plan Plaintiffs’ management plan is set out in four separate documents. The first, the Management Plan, describes biological characteristics of the walleye population, on which are based the Plan’s regulatory principles, guidelines, and plans for further research and regulatory development. The second document is the Intertribal Agreement, which describes the delegation of authority for the making of critical regulatory decisions. The third document is the Voigt Task Force Protocol, which sets out in detail the procedure to be followed in determining annual walleye harvest and quota. Finally, the Model Conservation Code sets forth a number of substantive fishery regulations and the mechanism for enforcing quota limits. 3. Tribal code Prior to trial the tribes developed a management plan setting forth methods for computing the tribal quotas to be applied to each body of water that would be subjected to intensive fishing methods. During the course of trial, the Biological Services Director of GLIFWC, Thomas Bu-siahn, revised the computations downward, producing smaller tribal quotas. Although the revised computations had not been submitted to the tribes or to GLIFWC for approval, plaintiffs’ counsel have used the new figures as the basis for their proposed findings of fact. In summarizing and evaluating the tribes’ management plan, therefore, I will assume that the revised figures are incorporated into that plan. Under the tribal management plan, for walleye populations that do not exhibit indicators of a healthy stock, and for walleye populations without information regarding such indicators, the Biological Services Director will set the Total Allowable Catch at 20 per cent of the estimated population of each lake on which plaintiffs intend to use high efficiency fishing methods. For walleye populations that do exhibit indicators of healthy stock, the Biological Services Director will set a Total Allowable Catch at 35 per cent of the estimated population. To arrive at the tribal quota from the Total Allowable Catch, the Biological Services Director will apply a discount to the Total Allowable Catch, based on the nature and age of the population estimate. Thus, for walleye populations that have had a mark-and-recapture estimate within the preceding two years, the safe harvest level will be set at 50 per cent of the total allowable catch. The safe harvest level will be the tribal quota. For walleye populations with mark-and-recapture estimates more than five years old, and for walleye populations estimated by the regression formula rather than by mark-and-recapture, the Biological Services Director will calculate the safe harvest level as 20 per cent of total allowable catch. For walleye populations dependent wholly or primarily on stocking for recruitment, the Biological Services Director will assume the population size to be 1.4 per acre, and will calculate total allowable catch as 35 per cent of the population size, and the safe harvest level as 50 per cent of the total allowable catch. The following chart sets forth the tribal managment plan, together with the accompanying level of risk that the particular tribal quota will result in a harvest in excess of the target exploitation. Population meets objectives Population doesn’t meet objectives Risk that TQ exceeds target exploitation Lakes with mark-recapture P.E.’s 1-2 years old .35 X .50 .20 X .50 10.0% Population meets objectives Population doesn’t meet objectives Risk that TQ exceeds target exploitation .20 X .30 3-5 years old CO Cn X Or o .20 X .20 5+ years old CO cn X to o to Lakes without mark-recapture P.E.’s Regression P.E. P.E. X .35 X .20 Regression P.E. P.E. X .20 X .20 5.0% All stocked lakes (Use recent P.E. or 1.4 per acre) P.E. x .35 X .50 Case-by-case per management objectives 16.7% Under the tribal plan, as revised by Bu-siahn, the total allowable catch for muskellunge will be set as follows: the Biological Services Division will estimate the muskellunge population on the basis of the lowest observed population density for the species, that is, .1 fish per acre in lakes with stocked muskellunge populations, and .2 fish per acre in lakes with naturally reproducing populations. The division will determine the maximum safe harvest of muskellunge by multiplying the estimated population in a particular body of water by the exploitation rate of 27 per cent, and multiplying the result by 50 per cent to calculate the safe harvest. Harvest will be terminated when the limit of the safe harvest is approached. The safe harvest levels for lakes with stocked populations is one muskellunge per 75 acres; for lakes with naturally reproducing populations, the safe harvest level is one muskellunge per 37 acres. On or before January 15th of each year, the Biological Services Division is to provide the Voigt Task Force with the tribal quotas it has determined for each body of water. The tribes then register their harvest goals in terms of their potential harvest capability and designate the lakes and the portions of the allowable tribal quota they intend to take. By February 15, the task force must set the number of walleye that the tribes will seek during the season, which runs from April 1 to the following March 31. The task force harvest level determination includes all fishing methods: hook and line, ice fishing, netting, and spearing. It establishes a quota for walleye for each lake named by a member tribe. The quota cannot be exceeded. If the task force does not establish a quota for any lake, tribal members may not harvest fish from that lake by any method. By March 15th of each year, the task force will issue and provide to the Department of Natural Resources the walleye quotas for ea