Mitigation banking is a tool now widely used in wetland conservation programs but is only beginning to be used in endangered species conservation programs. In the wetland programs, mitigation banking has long been promoted as a way to conserve wetlands at a lower cost than by traditional forms of mitigation. It also creates an incentive for at least some landowners to create, restore, enhance, or preserve wetlands, in order to be able to sell "credits" to others who need to mitigate projects that damage wetlands. Traditional wetland mitigation has a dismal track record, and it is not clear whether wetland mitigation banking will fare better. Its detractors are many.

The first wetland mitigation banks were established nearly two decades ago. It was not until 1995, however, that the several federal agencies with wetland conservation responsibilities could agree on uniform guidance concerning the establishment and operation of mitigation banks. Since then, wetland mitigation banking has become more common and consistent.

Because the laws protecting wetlands and endangered species permit otherwise prohibited activities if they are properly mitigated, the interest in endangered species mitigation banking may well grow, just as wetland mitigation banking did. Like the first wetland mitigation banks, the first endangered species mitigation banks are functioning without the benefit of any relevant, overarching federal policy. Indeed, California is the only state that does have a formal policy governing the use of mitigation banking (which it calls conservation banking) for endangered species. The principles of the California policy differ in many respects from those of the federal wetlands guidance.

Since the California policy was instituted in 1995, many endangered species mitigation banks have been established there, and they are now beginning to be established outside California as well. Interest in using this tool for endangered species purposes is clearly growing. Consequently, the U.S. Fish and Wildlife Service needs policy guidance for its own staff and for private interests concerning how endangered species mitigation banks should be established, operated, and overseen.

Because endangered species and wetlands differ in the extent of legal protection provided to each and in the laws and programs pertaining to their conservation, the policies for endangered species mitigation banking and wetland mitigation banking should differ as well. Accordingly, this report describes some of the issues that must be addressed in a federal policy on endangered species mitigation banking, suggests how those issues should be resolved, and proposes a draft policy for adoption by the U.S. Fish and Wildlife Service. This report also includes short case studies of several endangered species mitigation banks that are now in operation or are under development. Finally, the report contains a model endangered species mitigation banking agreement that will effectively conserve endangered species.

A recent headline on the front page of the Wall Street Journal hailed the opening of the nation’s first "butterfly bank." The "deposits" in this unusual bank are conservation credits earned as a result of having preserved an important area of habitat for the Quino checkerspot butterfly (Euphydryas editha quino), an endangered species restricted to California. The bank’s intended customers are other landowners who hope to develop other sites where the butterfly is found. In order to do so, they can buy credits from the private entrepreneur who established the butterfly bank.

Meanwhile, just a week earlier on the nation’s other coast, the state of North Carolina announced that it was purchasing a large tract of land containing a number of endangered red-cockaded woodpeckers (Picoides borealis). The state’s intention is to earn conservation credits that it can use to meet future mitigation requirements when the state’s transportation department builds new roads in woodpecker habitat elsewhere. The California and North Carolina examples illustrate two forms of a new phenomenon, generally known as either conservation banking or mitigation banking for endangered species.

Mitigation banking originated in the nation’s regulatory program aimed at preventing the filling of wetlands. Wetland mitigation banking began nearly two decades ago, in a largely ad hoc fashion. Few banks were developed until several federal agencies promulgated uniform guidance in 1995 concerning how they would evaluate and approve wetland mitigation banks (see Part 2). Today, mitigation banking for endangered species is in a position similar to that of wetland mitigation banking nearly two decades ago. That is, a few endangered species mitigation banks are beginning to be developed, but largely in an ad hoc fashion and with no federal policy or written guidance. The banks are coming first; the policies and rules will come later, if at all.

Developing an intelligent policy for endangered species mitigation banking should not take as long as it did to develop a policy for wetland mitigation banking. The reason is simply that the experience of nearly two decades of wetland mitigation banking should be helpful in designing an appropriate policy for endangered species mitigation banking. However, as this report explains, the policy cannot merely replace the word wetland with the words endangered species each time it appears in existing wetland mitigation banking policy. There are significant differences not only between endangered species and wetlands but also between the regulatory programs that seek to conserve them, and these differences may warrant quite different mitigation banking policies. The aims of this report, then, are to explore those differences, examine the role that mitigation banking could play in achieving the goals of the Endangered Species Act (ESA), and suggest a policy that would best accomplish that result.

In its most general sense, the word mitigation means the "abatement or diminution of something painful, harsh, severe, afflictive, or calamitous," a way, in other words, of making a bad thing less bad. In environmental contexts, mitigation generally refers to efforts to reduce or offset the negative environmental consequences of activities that are permitted despite their negative impact.

The origins of environmental mitigation date back at least as far as the Fish and Wildlife Coordination Act of 1934. As subsequently amended, that law sought to ensure that fish and wildlife conservation was given "equal consideration" with other objectives of major water resource development projects. The construction of dams, channelization of streams, and similar public works projects can have a variety of wildlife impacts, such as blocking fish migration, destroying spawning areas, and inundating wetlands and free-flowing streams. To minimize these inherent impacts, the Fish and Wildlife Coordination Act tried to make sure that every major water resource development project included conservation features. Some of these features were intended to benefit the wildlife resource, such as fish ladders to enable fish to surmount otherwise impassable dams or hatcheries to supplement natural spawning opportunities. Other mitigation measures under the Coordination Act were intended not so much to benefit the wildlife resource as to facilitate human recreational use of it. For example, the construction of boat-launching facilities, the acquisition of existing high-quality habitats, and similar measures were undertaken to offset the loss of public hunting and fishing opportunities. The units of measure for both the adverse impacts and the compensating mitigation were often not acres of habitat or numbers of animals but numbers of human "user days."

At present, somewhat more rigorous mitigation principles are used to implement the wetland conservation provisions of the Clean Water Act. Section 404 of that law prohibits the discharge of fill or dredged material into the waters of the United States, including wetlands, without a permit from the U.S. Army Corps of Engineers (Corps). The Corps routinely requires compliance with specific mitigation measures as a condition of such permits, and it follows a three-step sequence in developing such measures: avoidance, minimization, and compensation. The first step explores the availability of practical alternatives that avoid wetland impacts altogether. If there is a practical alternative to siting a project in a wetland, the Corps will not issue a permit. If there is no practical alternative that completely avoids the wetland, the second step is to minimize the impact on it by reducing the project’s "footprint," restoring temporarily disturbed habitats, or other means. Any remaining, unavoidable impacts on the wetland must then be remedied through compensating measures that try to offset any loss of the wetland’s functions and values because of the project by creating or enhancing the functions and values elsewhere.

In addition to this sequencing requirement for wetland mitigation, other principles have been clearly established. For example, the Corps generally prefers "on-site" mitigation measures (undertaken on or very near the project site) to "off-site" measures (undertaken away from the project site). The reason is that certain wetland functions (e.g., floodwater retention and water purification) are truly local. That is, less retention of floodwater in the Illinois River watershed cannot be fully compensated for by more retention of floodwater in the Iowa River watershed. Similarly, the Corps strongly prefers "in-kind" to "out-of-kind" mitigation. For example, the Corps generally tries to mitigate the loss of a particular kind of wetland by requiring the restoration or enhancement of the same kind of wetland. One rationale for this preference is that the suite of species associated with each particular type of wetland differs from the suite of species associated with other types of wetlands. Requiring in-kind mitigation therefore attempts to maintain ecological values.

A final preference is to mitigate through restoration or enhancement. The preference of restoring and enhancing wetlands over creating them reflects scientific doubts about being able to create fully functioning wetlands where they never occurred naturally. The preference of restoring and enhancing over simply purchasing or preserving existing wetlands recognizes that mitigating the loss of some wetlands by purchasing or preserving others guarantees a net loss of existing wetlands and their associated functions and values. Thus, under the Clean Water Act, the old notion of measuring environmental losses and mitigation gains by the currency of human "user days" has largely been abandoned. Later we will consider whether the principles that are now well established for wetland mitigation are equally relevant to endangered species mitigation.

Project-by-project mitigation, in which on-site, in-kind mitigation measures are drawn up for each new project affecting wetlands, has a number of potential drawbacks, for both regulated and conservation interests. First, designing an appropriate mitigation element for each new small development project is costly, as one cannot take advantage of economies of scale when designing small mitigation sites to compensate for small development projects. Second, the restoration and enhancement of wetlands are uncertain arts; if development projects proceed concurrently with mitigation efforts, the development may be complete long before one can determine the success or failure of the mitigation effort. Third, even if successfully established, small and often isolated mitigation wetlands may be seriously degraded over time by the invasion of exotic species, illegal dumping, off-road vehicles, and other threats. Without some mechanism to "defend" and manage these mitigation sites over the long term, they may cease to provide the full range of functions and values for which they were intended.

Wetland mitigation banking was developed to overcome these deficiencies in traditional mitigation and to create opportunities for entrepreneurial landowners to profit from wetland conservation. Essentially, wetland mitigation banking is the creation, restoration, enhancement, or preservation of wetlands in advance of any specific project requiring mitigation, with the "credits" earned from such efforts made available to meet the mitigation requirements for future projects of the same or a different landowner. State highway departments initiated most of the early wetland mitigation banks. Looking into the future, they knew that they would have a continuing need to mitigate wetland losses as they built more new highways. Rather than develop postage-stamp-sized on-site mitigation projects for each new highway project, they looked for a means of mitigation that was both more efficient for them and more beneficial for the environment. One large-scale wetland restoration effort was often cheaper, on a per-acre basis, than many smaller projects.

Proponents of wetland mitigation banking asserted that the environment would also benefit, for at least three reasons. First, mitigation banking offered the opportunity to locate mitigation sites where they would offer a significant environmental benefit, rather than at the site of the proposed development, which might or might not have such benefits. Second, by consolidating the mitigation for many small projects into one large mitigation site, banking could secure certain environmental benefits (e.g., complexity of habitats, viability of populations, buffering from edge effects) unattainable at smaller sites. Finally, if mitigation banks extended credits only after demonstrating success in creating, restoring, or enhancing wetlands, then banking could offer a means of improving the generally poor record of traditional mitigation.

The track record of traditional, project-by-project wetland mitigation is dismal. One example is a 1991 study done for the South Florida Water Management District that examined more than one hundred projects that required some form of wetland mitigation, but for which the mitigation was actually carried out for only forty. The mitigation for thirty-one of them required the creation of wetlands. A total of 1,058 wetland acres were to be created for the thirty-one projects, but only about half the acres (531) were actually created. The study also found that thirty-two of the forty mitigation sites had been colonized by undesirable plants. Furthermore, only three of the sites had long-term management plans. Finally, although postconstruction monitoring was required at nearly every site, it was never done at fifteen of them.

Whether these results should properly be considered a failure of traditional mitigation or, rather, a failure of the government to monitor and enforce traditional mitigation requirements, the outcome is still the same. Traditional, project-by-project wetland mitigation has often not lived up to its promise. Benjamin Tuggle, a former U.S. Fish and Wildlife Service field supervisor in Chicago, explained why wetland mitigation in urbanizing areas so frequently fails:

The typical project requiring mitigation is a residential or commercial development impacting several acres of wetland. The mitigation typically is sited in a basin which also provides stormwater detention or compensatory floodplain storage, both of which involve large influxes of contaminated urban stormwater. Water quality is often limiting in these situations, and the large water level fluctuations preclude the establishment of normal wetland hydrology. These basins typically become shallow open water ponds, providing few wetland functions. Because they are small and usually surrounded by development, they provide little in the way of wildlife habitat. For similar reasons, it is difficult to find appropriate entities to take over management responsibilities, so there usually is no provision for long-term management and stewardship. All of these factors result in a high failure rate.

Problems with wetland mitigation are not unique to federal programs but plague state programs as well. A recent example is a study of mitigation results in Massachusetts under that state’s generally well regarded Wetlands Protection Act. The goal of the Massachusetts study was to generate more recent and more statistically reliable information than that contained in a 1989 Corps of Engineers study that found a 36 percent failure rate for wetland "replication" (i.e., creation) projects in the state. The later study examined 391 mitigation projects in forty-four randomly selected towns, including site visits to 114 projects. It found that 54.4 percent of the projects did not comply with regulatory requirements and that 38.6 percent had produced no wetland at all. Many of the failed projects were designed as stormwater detention basins, but because they were either too wet or too dry, the basins often failed to create a wetland. In nearly 22 percent of the projects examined, no wetland had even been built; for a nearly identical proportion of projects, a wetland was built but was smaller than required.

Another significant finding of the Massachusetts study was that "the plant communities in replicated wetlands differed significantly from those in wetlands they were designed to replace." Although a majority of the projects for which mitigation was required affected forested wetlands, and creation of a forested wetland was the goal of a quarter of the projects, no forested wetland was successfully created at any mitigation site. Moreover, the study found that plant communities in the created wetlands did not become more like those at impact sites over time, even though the created wetlands were as much as twelve years old. In the words of the study, "This means that, at best, there is a significant temporal loss of wetland function for at least 12-15 years following creation of a replication site, and this loss may possibly last much longer." The study also found evidence suggesting that "replicated wetlands are becoming drier over time."

Both the impacted wetlands and the mitigation sites included in the Massachusetts study were generally quite small; nearly 80 percent of the impacted wetlands were less than five thousand square feet (slightly more than a tenth of an acre). A few somewhat larger projects (up to two acres) were authorized as "variance projects." According to the study, these larger variance projects were "comparable to pilot mitigation banks or other centralized mitigation approaches." Interestingly, all the variance projects were found to be in compliance with regulatory requirements and "were much more carefully designed," although the plant communities associated with them were, again, different from those at the impacted wetlands.

In addition to the biological problems associated with the mitigation sites, the study uncovered deficiencies in program administration. "Some projects that were clearly not in compliance, including two that were never built, had been issued certificates," indicating that they were in compliance. The study also found that "most towns are not systematically tracking the progress of replication projects and determining if they are in compliance with the regulations." Although Massachusetts regulations require that replicated wetlands function similarly to the wetlands they replace, this requirement "appears to be poorly understood by Conservation Commissions and by permit recipients,

and approval is routinely given for wetland replication projects that are designed to be dissimilar from the impacted wetlands in both structure and function."

The Massachusetts study shows that traditional wetland mitigation efforts often fail whether they are overseen by state or federal authorities. There are, as yet, no similar studies assessing the effectiveness of endangered species mitigation efforts. It is clear, however, that some of the facts contributing to the ineffectiveness of wetland mitigation efforts also are present in the case of endangered species mitigation. These include diffuse and poorly coordinated mitigation efforts, technical challenges, and limited resources for monitoring and enforcement.

It was because of difficulties like these and the repeated failure of traditional wetland mitigation efforts that the idea that wetland mitigation banking might be more successful took hold. Whether wetland mitigation banking will in fact achieve better results than traditional, project-by-project wetland mitigation is not yet clear. Nevertheless, some people are beginning to embrace endangered species mitigation banking as a better alternative than traditional approaches to endangered species mitigation. Before deciding whether they are right, we will consider some of the important similarities and differences between wetlands and endangered species.

Many private landowners who want to earn income from their land or to hold onto it as a long-term investment regard wetlands or rare species as a liability rather than an asset. It is not the wetlands or the species themselves that pose a threat; rather, it is the potential land-use restrictions. Under the ESA, no one can "take" an endangered animal without a permit, and this prohibition against "taking" extends to activities that alter the animal’s habitat. Land clearing, timber harvesting, and other habitat-altering activities may therefore expose a landowner to criminal penalties or civil injunction. Landowners whose land houses either endangered animals or plants may also be unable to secure any of the federal permits required for the activities they wish to carry out. From the landowner’s perspective, then, any endangered animal may well be an "anima non grata."

For our purposes, the main similarity between wetlands and endangered species is that both are protected and thus, without a permit, cannot be harmed by a variety of activities. At the same time, there are some important differences between the two. Wetlands are relatively permanent features on the landscape, but many endangered species are only short-term occupants of any given parcel. For example, a wetland with threatened California red-legged frogs (Rana aurora draytonii) living in it remains a wetland even after nonnative bullfrogs have found their way to the site. Once the bullfrogs arrive, however, they are likely to displace the red-legged frogs. When the red-legged frogs are gone, the site will continue to be protected as a wetland under the Clean Water Act, but it will no longer receive protection under the ESA.

It is not just introduced species that contribute to the impermanence of many endangered species. The young stands of jack pines that serve as breeding habitat for Kirtland’s warblers (Dendroica kirtlandii) stop supporting these endangered birds once the stands are more than about twenty years old. A parklike stand of older longleaf pine trees will serve as suitable habitat for the endangered red-cockaded woodpecker only as long as the hardwood understory is kept at bay through prescribed burning or other means. If the hardwood understory and the woodpecker’s ideal open pine forest habitat are not properly managed, they will be transformed into an unsuitably dense mixed pine-hardwood forest. Small, isolated wetlands will likely persist on the land regardless of what happens to other, similar wetlands nearby. In contrast, however, a small patch of habitat supporting the endangered Bay checkerspot butterfly (Euphydryas editha bayensis) may be too small to continue doing so once nearby habitat patches have disappeared.

These examples illustrate a common and important difference between wetlands and endangered species habitats. In order for wetlands to remain wetlands, they often need only to be protected; they generally do not need to be actively managed (though active management may be needed to prevent their degradation by alien species and other factors). In contrast, many endangered species habitats must be actively and continuously managed in order to support endangered species. Simply "protecting" them by putting a fence around them and prohibiting human activities inside is not sufficient. Without active management, many endangered species habitats cease to be endangered species habitats and thus lose the protection they formerly had under the ESA. Without active management, wetlands also may suffer serious degradation, but they are unlikely to cease being wetlands altogether. Despite their degraded condition, they will continue to be protected against the full range of activities regulated by the Clean Water Act. This important difference has significant ramifications for mitigation strategies.

In addition to the inherent differences between wetlands and other endangered species habitats, there also are differences in the purposes of the federal programs that seek to conserve them that, too, can give rise to different mitigation strategies. For example, the goal that has guided wetland policy during the past decade has been "no net loss" of wetland acreage and function. Whereas measuring wetland function is difficult, measuring wetland acreage is easy. That is, if there are x million acres of wetlands in the United States today and will be x million acres ten years from now, the goal of no net loss (at least with respect to acreage) will have been achieved. The goal of the ESA, however, cannot be achieved simply by freezing the status quo. Rather, one must reduce the likelihood of extinction--or, conversely, increase the probability of survival--to a safe level. When a species is no longer in danger of extinction in the foreseeable future and is not likely to become so, it is considered to have recovered. This, in turn, allows it to be removed from the endangered species list.

Recovery requires that the threats to a species’ survival be reduced. It does not necessarily mean an increase in the numbers or distribution of a species (though it usually does). Indeed, it may be possible for a species to decline in total numbers even as its likelihood of survival increases. This could happen, for example, if none of the habitat of a declining species were initially under the sort of ownership that ensured the active management needed to perpetuate it. If, through public acquisition or otherwise, one could be assured that some of its habitat would be appropriately managed, that fact might improve its prospects of survival enough for recovery, even though the rest of the habitat (and the species occupying it) were lost.

For many endangered and threatened species, recovery objectives are often expressed as a number of populations of a given size occupying a secure habitat. To meet these objectives, it is not necessary for every occurrence of a species to be maintained where it now is. Instead, if occurrences of the appropriate size and distribution can be secured, the recovery goals can be achieved regardless of what happens to the smaller, more isolated populations. This fact makes the conservation of endangered species fundamentally different from the conservation of wetlands. It also provides opportunities for mitigation (and mitigation banking) under the ESA that are distinct from those for wetlands under the Clean Water Act. Before considering which policies would be most suitable, we should examine the policies governing wetland mitigation banking.

II. FEDERAL GUIDANCE FOR THE ESTABLISHMENT, USE, AND OPERATION OF WETLAND MITIGATION BANKS

Although wetland mitigation banking began in the early 1980s, it had no uniform federal guidance until 1993, when the U.S. Army Corps of Engineers and the Environmental Protection Agency issued "interim national guidance." Two years later, on November 28, 1995, the Corps, Environmental Protection Agency, U.S. Fish and Wildlife Service, National Marine Fisheries Service, and Natural Resources Conservation Service jointly issued the more detailed, final "Federal Guidance for the Establishment, Use and Operation of Mitigation Banks." This 1995 Guidance remains in place today as a comprehensive account of the policies, procedures, and criteria applicable to the use of mitigation banks to provide compensatory mitigation for authorized adverse impacts to wetlands under Section 404 of the Clean Water Act and Section 10 of the Rivers and Harbors Act.

The 1995 Guidance reflects a generally positive view of wetland mitigation banking and begins by reciting a number of potential advantages of mitigation banking over individual mitigation projects, including

The ecological advantage of consolidating what might otherwise be numerous small, unconnected mitigation parcels into a single, large parcel.

The technical advantage of mustering greater financial and scientific resources with which to establish and manage mitigation bank sites than are typically available in connection with small, individual mitigation sites.

The economic advantage of reducing permit-processing times and increasing the cost effectiveness of mitigation opportunities.

The temporal advantage that mitigation banks offer, in that their mitigation measures are "typically implemented and functioning in advance of project impacts," thus improving the certainty of success.

The administrative advantage to the agencies of reducing the burden of reviewing and monitoring mitigation projects.

The programmatic advantage of providing a mitigation opportunity in circumstances in which mitigation might otherwise not be possible.

The practical advantage of giving greater flexibility to permit applicants.

The 1995 Guidance is silent on the potential disadvantages of mitigation banking. It also does not explicitly discuss the conditions that must exist to realize the advantages. Instead, both the advantages and disadvantages are, to some degree, implicit in the criteria and other requirements of the Guidance.

The 1995 Guidance describes both the procedural and the substantive requirements for wetland mitigation banks. In regard to procedure, the Guidance suggests that a mitigation banking initiative begin with informal discussions between a prospective bank sponsor and the appropriate agencies. Based on those discussions, if the prospective sponsor wishes to proceed, it submits to the Army Corps of Engineers a "prospectus," a very general statement of the sponsor’s plans for the bank. The Corps then provides notice and a brief opportunity for the public to comment on the prospectus.

Submission of a prospectus triggers a formal agency review of the banking proposal. This formal review is conducted by a "mitigation bank review team" chaired by a representative of the Corps and composed of representatives of all the federal agencies with an interest in the proposed bank, as well as state, local, and tribal agencies with regulatory authority pertaining to the proposed bank. Working in consultation with the review team, the prospective bank sponsor next prepares a "banking instrument" that describes in detail the characteristics of the bank and how it will be established and operated. Among other things, banking instruments are typically used to address the geographic area to be serviced by the bank, methods for determining credits and debits, performance standards, monitoring plans, contingency and remedial responsibilities, financial assurances, and provisions for long-term management.

When completed, the banking instrument is signed by the bank sponsor and concurring members of the review team. By signing, each agency signals its agreement with the terms of the instrument. The 1995 Guidance states that review teams should "strive to obtain consensus," but if a consensus cannot be reached, it gives "the responsibility for making final decisions regarding the terms and conditions of the banking instrument" to the chair of the team (i.e., the Corps). Once a bank has been established, the Corps also is responsible for authorizing the use of credits from the bank to mitigate specific projects. When exercising this responsibility, the Corps is to consider the comments of the other resource agencies, just as it is generally required to do for wetland permit applications.

What must a wetland mitigation bank do to generate credits? The 1995 Guidance recognizes four categories of activities: creation, restoration, enhancement, and preservation. Creation refers to the establishment of a wetland where none existed. Technically, this is usually the most challenging, least certain, and most controversial of the mitigation activities.

Preservation refers to the protection of existing wetlands in perpetuity through legal and physical mechanisms. Technically, this is the least challenging and generally most certain of the mitigation options. Nevertheless, for reasons discussed later, the 1995 Guidance does not favor preservation as a source of mitigation banking credits but allows it only "in exceptional circumstances."

Restoration and enhancement refer to a potentially overlapping set of activities. Restoration can mean the reestablishment of a wetland at a site where one once existed but no longer does. Often, restoration is relatively easy and usually successful if it is simply removing drainage tiles or levees, or otherwise restoring hydrological conditions to a site with hydric soils capable of supporting wetland plants. Because the art of wetland restoration is better developed than that of wetland creation or enhancement, the 1995 Guidance specifies that "restoration should be the first option considered when siting a bank."

Restoration, however, is not limited to the reestablishment of wetlands at former wetland sites. It also includes the reestablishment of wetland characteristics and functions at a site where they still exist but are degraded. It is here that it begins to blend with enhancement, which refers to activities that increase wetland functions at an existing (though not always degraded) wetland. When restoration is given this meaning, both it and enhancement activities present a big challenge in quantifying the values gained and ensuring their comparability to those lost by the filling of wetlands elsewhere.

Although the 1995 Guidance explains reasonably clearly the types of activities that can generate credits in a wetland mitigation bank, it is far less clear about how those credits are to be quantified to compensate for the impacts on wetlands elsewhere. Recall that the 1995 Guidance requires the bank sponsor and the agencies on the "mitigation bank review team" to agree on a "banking instrument." Among other things, the Guidance specifies that the banking instrument should address "methods for determining credits and debits." It does not say what those methods should be, other than to say that "an appropriate functional assessment methodology . . . acceptable to all signatories should be used." Such methodologies may include "Habitat Evaluation Procedures, hydrogeomorphic approach[es] to wetlands functional assessment, [or] other regional assessment methodolog[ies]." Whatever methodology is chosen, the Guidance makes clear that it should be used in assessing both credits and debits.

The assessment methodology ultimately used must have a common "currency" to ensure that the environmental gains from the mitigation activities and the environmental losses from the wetland-filling activities are measured consistently. The 1995 Guidance’s preference is a methodology that measures both wetland acreage and "function." However, the Guidance does recognize that in some instances "an appropriate functional assessment methodology is impractical to employ" and that in such instances "acreage may be used as a surrogate for measuring function."

The use of acreage as a surrogate for more sophisticated measures of function is not a problem when the mitigation involves creating or restoring the same general type of wetlands as those being lost. Acreage does become a more problematic surrogate when the mitigation takes the form of preserving an existing wetland or of enhancing the functions of a degraded wetland. The practical and difficult question then is how to determine what amount of wetland enhancement or preservation at the mitigation site is equivalent to the loss of an acre of wetland at the impact site. However credits and debits are measured, the actual exchange of credits for debits requires yet another determination, the specification of a compensation ratio.

Among the many things to be specified in each banking instrument are "compensation ratios." The banking instrument should specify the number of bank credits required to be exchanged in order to compensate for each unit (however measured) of wetland loss. Surprisingly, though, the 1995 Guidance fails to consider this in any detail.

The virtual silence of the 1995 Guidance on the topic of compensation ratios might be thought to reflect an implicit assumption that in nearly every case the ratio will be one-to-one. A one-to-one ratio would ensure that each unit of wetland loss was replaced by a comparable unit of wetland gain. This approach would presumably be consistent with the stated goal of "no net loss" of wetlands governing the nation’s wetland programs since 1990. In practice, however, compensation ratios are frequently set higher than one-to-one.

Although the 1995 Guidance offers few clues to setting compensation ratios at anything other than one-to-one, it does at least suggest that the likelihood of success in creating, restoring, enhancing, or preserving wetlands at the mitigation site should not influence the compensation ratio. In a lengthy discussion of the technical feasibility of mitigation efforts, the Guidance states that "when uncertainties surrounding the technical feasibility of a proposed mitigation technique exist, appropriate arrangements (e.g., financial assurances, contingency plans, additional monitoring requirements) should be in place to increase the likelihood of success." Because using a greater than one-to-one compensation ratio could serve as a hedge against the mitigation’s possible failure, its omission from this list is surprising. The unstated premise, reflected in a number of places throughout the Guidance, is that the conditions attached to approved mitigation banks must virtually ensure their success. One place where the Guidance departs from this premise--and the only place where the adjustment of compensation ratios is addressed--is in its discussion of "advance crediting."

As described earlier, the 1995 Guidance begins by reciting a number of advantages that mitigation banks have over individual mitigation projects. Among these is that with banks, "compensatory mitigation is typically implemented and functioning in advance of project impacts, thereby reducing temporal losses of aquatic functions and uncertainty over whether the mitigation will be successful in offsetting project impacts." This statement contains some important qualifiers. First, instead of "always," bank mitigation is "typically" implemented and functioning in advance of project impacts. Second, the Guidance does not claim that banking arrangements "eliminate" all uncertainty about the mitigation’s success; rather, such arrangements only "reduce" such uncertainty.

All these qualifiers would be unnecessary if the 1995 Guidance required that mitigation bank credits be earned and used only after the bank’s creation, restoration, enhancement, or preservation measures had been proved successful. Instead, it authorizes the "limited debiting of a percentage of the total credits projected for the bank at maturity" before the successful achievement of the full wetland functions that are to be the basis for any credits. This practice, known as advance crediting, was a concession to the economic argument that prospective bankers often need to be able to generate some return relatively quickly in order to finance the costs of mitigation at a bank site, particularly when a large restoration or creation effort is planned at the site. While that argument may often be valid, it undermines one of the strongest arguments frequently offered for mitigation banking--that mitigation is first achieved and then "banked" for later use.

When the 1995 Guidance was originally proposed, it included an illustrative example in which a bank was allowed to sell 15 percent of its projected credit total in advance of the credit’s maturing. The final draft, however, omitted the example and explained that the percentage of advance credits permitted for a particular bank would be decided on a case-by-case basis and "may be higher or lower than the 15 percent example included in the proposed guidance." Although 15 percent (or any other figure) may have been arbitrary, the absence of a ceiling on advance crediting in the Guidance has virtually ensured that mitigation bankers will bargain for as much advance crediting authority as they can get, which often leads to squabbling among the agencies on the mitigation bank review team.

If advance credits are issued, there will be at least a temporary loss of wetland functions before the mitigation actions meant to compensate for that loss take effect. The 1995 Guidance tries to balance this situation in two ways. First, it seeks to limit the duration of such temporary losses by specifying that "initial physical and biological improvements" associated with the advance credits "should be completed no later than the first growing season following initial debiting of a bank." Second, the Guidance suggests that it may be appropriate to impose higher mitigation ratios to compensate for such temporal losses.

This unexplored inconsistency between the practice of advance crediting, clearly allowed by the 1995 Guidance, and the definition of a mitigation bank found in that same Guidance is a source of tension between the regulatory agencies and the banking community. The definition refers to "a site where wetlands and/or other aquatic resources are restored, created, enhanced, or . . . preserved expressly for . . . compensatory mitigation in advance of authorized impacts to similar resources" (emphasis added). When advance crediting is allowed, however, the compensatory mitigation takes place after the authorized impacts, not before them. This obvious inconsistency is a good example of the tension between trying to apply an abstract model of mitigation banking with clear theoretical benefits, and the practical necessity of adapting that model to economic considerations that might sacrifice some of those benefits. Other similar tensions can be seen in the Guidance’s discussion of on-site versus off-site mitigation.

In 1990, five years before the Guidance on mitigation banking, the Corps and the EPA signed a memorandum of agreement concerning mitigation under Section 404 of the Clean Water Act and Section 10 of the Rivers and Harbors Act. That memorandum expressed a preference for on-site mitigation over off-site mitigation. By definition, however, mitigation banks are a form of off-site mitigation. Nevertheless, the 1995 Guidance concludes that the 1990 memorandum "should not preclude the use of a mitigation bank when there is no practicable opportunity for on-site compensation, or when use of a bank is environmentally preferable to on-site mitigation." Although this formulation appears to limit the circumstances in which off-site mitigation would be appropriate, it must be remembered that the 1995 Guidance begins by reciting the many advantages, including several environmental advantages, that mitigation banking offers over traditional project-by-project mitigation. Thus, the preference expressed in 1990 for on-site mitigation may not in fact be as strong as it first appears. Indeed, the 1995 Guidance asserts that "in general, use of a mitigation bank to compensate for minor aquatic resource impacts (e.g., numerous, small impacts associated with linear projects; impacts authorized under nationwide permits) is preferable to on-site mitigation." Thus, for this class of activities, the Guidance actually reverses the preference expressed in the 1990 memorandum.

The rationale for preferring on-site to off-site mitigation is that many of the benefits that wetlands provide, and that the law protects, are inherently local. Floodwater retention, nutrient uptake, and other functions of wetlands may be replicable through wetland restoration elsewhere, but the replication of those functions at a distant mitigation site still leaves the original wetland site without the local benefits that the lost wetland formerly provided. Unless mitigation is carried out very close to the impacted site, some local problems (e.g., flooding and the loss of locally valued wildlife) may actually be exacerbated, notwithstanding the mitigation.

Although the 1995 Guidance declares that the use of a mitigation bank may sometimes be preferable to on-site mitigation, it still tries to limit how far from the impact site a mitigation bank may be located. It does this by requiring that the banking instrument specify a "geographic service area" within which credits from the bank can be used to compensate for wetland impacts. Service areas may be nearby watersheds, adjacent counties, or other areas close by. The determination of the service area has major implications for the bank’s economic viability. In general, bank sponsors seek as large a service area as possible so as to widen the market of potential buyers. The preference of the federal resource agencies for on-site mitigation, however, pushes them toward smaller service areas, to keep the benefits of mitigation as close as possible to the site of the wetland losses. The delineation of a bank’s service area, however, does not necessarily preclude the use of credits from the bank to mitigate for losses outside the service area. The 1995 Guidance does allow such use of credits "on a case-by-case basis, where it is determined to be practicable and environmentally desirable."

The 1995 Guidance allows the preservation of existing wetlands to generate mitigation credits only "in exceptional circumstances." The rationale for this policy is that Section 404 of the Clean Water Act already protects existing wetlands from a broad range of destructive activities. To allow credits for preserving what is already substantially protected from most threats would thus produce a largely illusory benefit.

While the scope of harmful activities regulated under Section 404 is broad, it is not limitless, and some serious threats to the viability of wetlands are beyond its reach. For example, land uses on adjacent or nearby upland areas can substantially degrade wetland values by increasing siltation, adding polluted runoff, lowering water tables, or other means. Section 404 also offers no protection against the considerable threat of invasive, nonnative plants, such as melaleuca in Florida or purple loosestrife in the Midwest. These plants may not wholly eliminate an existing wetland, but they effectively destroy many of its critical functions.

For these reasons, the 1995 Guidance allows the preservation of existing wetlands to serve as a basis for generating mitigation credits under a variety of exceptional circumstances. These circumstances include situations in which existing wetlands "are under demonstrable threat of loss or substantial degradation due to human activities that might not otherwise be expected to be restricted." If, through preservation, a bank prevents the degradation of an existing wetland, then the difficult issue becomes how to quantify the credits. The Guidance specifies that credits in such circumstances "should be based on the functions that would otherwise be lost or degraded if the aquatic resources were not preserved, and the timing of such loss or degradation." In other words, the more imminent the anticipated threat to the existing wetland is, the greater the number of credits earned will be. In practice, this rule is often exceedingly difficult to apply. Because preservation typically protects a wetland from the loss of some, but not all, functions, the Guidance asserts that "a greater number of acres from a preservation bank" is usually required to compensate for a given loss than would be the case if the bank were based on restoration, creation, or enhancement.

The 1995 Guidance prefers that preservation not generate mitigation credits because existing wetlands already have the benefit of significant legal protection. Somewhat similar considerations arise when publicly owned lands are proposed as sites for mitigation banks. Public agencies that manage public lands, particularly those established for conservation purposes, commonly carry out a variety of environmental improvements, including wetland creation, restoration, or enhancement. If private parties finance such improvements, should they be able to earn mitigation credits for doing so?

The 1995 Guidance says yes but specifies that the number of credits in such circumstances "should be based solely on those values in the bank that are supplemental to the public program(s) already planned or in place." Put differently, "baseline values represented by existing or already planned public programs, including preservation value, should not be counted toward bank credits." This presents another, very difficult, variant of the familiar problem of quantifying mitigation credits. It also introduces a potentially troublesome new issue. A private entrepreneur entering into a cooperative arrangement with a public landowner to carry out a wetland restoration project on public land in order to generate credits has no land costs associated with the project. In contrast, a private entrepreneur who carries out a similar project on his own land must factor the cost of the land into the credit price. These public-private cooperative arrangements therefore have a competitive advantage. The Guidance never addresses this, though commentators have pointed to it as an undesirable outcome. It reduces the incentive for entrepreneurs to establish mitigation banks on their own land, and it gives public agencies an incentive to look to the private sector to fund environmental improvements on public land. Ultimately, it may shortchange the environment by providing a convenient rationale for reducing the public funding of environmental improvements on public land.

Although the U.S. Fish and Wildlife Service (Service) was one of the agencies that promulgated the 1995 Guidance, it apparently had misgivings about allowing mitigation banks to be sited on public land, at least if that land were part of the National Wildlife Refuge System. Four years after the 1995 Guidance was drawn up, the Service published its own policy that it "will not allow the use of National Wildlife Refuge System lands for mitigation banks" under the Clean Water Act, although it may accept a successfully completed bank as an addition to a unit of the system. The Service’s policy also generally bars the use of refuge lands for traditional, nonbanking mitigation purposes. The rationale for the policy is that although some refuge lands may have been degraded by past activities, the Service is "authorized to restore degraded habitats within the National Wildlife Refuge System and . . . will be restoring these lands in the future, irrespective of off-Refuge development." The policy, however, does not absolutely prohibit the use of refuge lands for traditional, nonbanking mitigation purposes. An exception may be granted if, among other things, "the mitigation would result in significantly increased natural resource benefits, when compared to other appropriate, off-site mitigation options."

The Service’s policy applies only to wetland mitigation under the Clean Water Act, explicitly stating that it "does not apply to threatened or endangered species." This is a somewhat puzzling exclusion, since the Service’s authority to restore endangered species on refuge lands is no less clear than its authority to restore wetlands and is arguably more so. Section 7(a)(1) of the ESA imposes on the Service an affirmative obligation to use its various authorities also to conserve threatened and endangered species. No comparable statutory command requires the Service to restore wetlands.

A similar question is whether mitigation banks can be sited on land where wetlands have been restored or created with federal funding. The Wetlands Reserve Program and the Partners for Fish and Wildlife Program are two examples. The 1995 Guidance states that federally funded wetland conservation projects undertaken under such programs "cannot be used for the purpose of generating credits within a mitigation bank." Similarly, the Service’s 1999 policy on the use of refuge lands states that "where habitats are protected or restored under" such programs, the Service "will not recommend, support, or advocate the use of such lands as compensatory mitigation, including mitigation banks" during the term of the agreement and may do so after the term of the agreement only "in limited and exceptional circumstances." The rationale for these policies, though not explained, appears to be that because such projects have been at least partially federally funded, private banking interests should not be able to sell mitigation credits for actions that would not have taken place were it not for the federal funding. That rationale is difficult to square with the fact that the 1995 Guidance allows mitigation banks to be sited on public (including federal) land. In those instances, there is also a public contribution to the endeavor (in the form of the public land made available for the bank), yet the Guidance allows credits to be sold from banks sited on public land.

Many of the federally funded wetland restoration programs require a private landowner to maintain a restored wetland for a limited time only. The Partners for Fish and Wildlife Program, for example, typically requires a landowner to maintain a restored wetland for only ten years, after which he may fill it under the authority of a nationwide permit issued by the Corps. If a landowner with the right to fill a restored wetland under such a program elects to waive that right, it is not clear why such a waiver should not qualify as compensatory mitigation. Similarly, a wetland that has been restored under a federal program might subsequently be enhanced by actions undertaken without the aid of federal funds. Here, too, it is not clear why the fact that the wetland was originally restored with the help of federal funding should disqualify any later enhancement of it from earning mitigation credits. Examples like these may in fact be contemplated by the 1995 Guidance, which, despite its general prohibition against using federally funded wetland projects for mitigation banking purposes, does allow mitigation credit "for activities undertaken in conjunction with, but supplemental to, such programs in order to maximize the overall ecological benefit of the conservation project."

Should the loss of one type of wetland (e.g., a forested wetland), be compensated for by the creation or restoration of another type of wetland (e.g., a shrub/scrub wetland)? Such arrangements are known as out-of-kind mitigation. Although the 1995 Guidance is not in favor of them, it does not prohibit them altogether. Instead, it allows them if they are "practicable and environmentally preferable to in-kind compensation." An example of this is a type of wetland at a mitigation site that is "of greater ecological value to a particular region." The Guidance is more emphatic in its opposition to the use of nontidal wetlands to compensate for the loss of tidal wetlands. Although it singles out this particular form of out-of-kind mitigation for special disapproval, even then it stops short of prohibiting the practice altogether.

Out-of-kind mitigation arrangements present a special challenge in quantifying credits and debits, since by definition, one is dealing with apples and oranges. Implicitly, the 1995 Guidance recognizes that it may be environmentally desirable to compensate for the loss of relatively common and ordinary types of wetlands by restoring or protecting relatively more rare or valuable types of wetlands. Alternatively, the relative rarity or value of two different types of wetlands might be addressed through specially tailored mitigation ratios. The Guidance, however, gives no real direction on this difficult issue.

A central tenet of mitigation banking is that beneficial activities carried out at a mitigation site generate credits that compensate for harmful activities at an impact site. Once used, a credit is extinguished and may not be reused. It is hardly surprising, then, that the 1995 Guidance asserts, in one of its few unqualified statements, that "in no case may the same credits be used to compensate for more than one activity."

Immediately following this statement, however, the Guidance goes on to say that "the same credits may be used to compensate for an activity which requires authorization under more than one program." It then cites a number of examples of other potentially relevant programs, including state or local wetland regulatory programs, the National Pollutant Discharge Elimination System program, and the Superfund removal and remediation programs. Although the Guidance does not specifically mention the ESA, it presumably would qualify as a program under which separate authorization, in addition to that required by the Clean Water Act, is needed for at least some activities.

Filling a wetland is, in most cases, a permanent loss. To balance that loss, compensatory measures should be permanent as well. That, at least, is the conclusion of the 1995 Guidance, which generally stipulates that wetlands in a mitigation bank be protected in perpetuity through a conservation easement, title transfer, or similar arrangement.

Once again, however, the general rule in the Guidance has exceptions. Less-than-permanent protection of the resources in a mitigation bank can be approved "in exceptional circumstances." One example given in the Guidance is for "coastal protection projects which prolong the ecological viability of the aquatic system." Although not offered as an example, less-than-permanent protection might also be appropriate when the impact site is unlikely to persist in perpetuity, such as might be the case as a result of nearby groundwater withdrawals, heavy siltation, or the invasion of exotic plants capable of destroying all or most wetland functions.

When less-than-permanent mitigation is approved, the 1995 Guidance makes clear that it should never "extend for a lesser time than the duration of project impacts for which the bank is being used to provide compensation." The Guidance clearly is aiming for a temporal balance between impacts and mitigation: permanent impacts must be counterbalanced by permanent conservation commitments; less-than-permanent impacts should be counterbalanced by conservation commitments that last at least as long as the impacts.

Because of the uncertainties inherent in most forms of wetland mitigation, there must be sufficient financial resources to monitor the success of the mitigation effort and to respond to problems. For the ordinary project-by-project mitigation, the failure to carry out the mitigation requirements or the failure of the mitigation itself may result in revocation of the permit that authorized the impact. Revoking a permit often has little practical value, however, because after the project is completed, the builder no longer owns the affected property.

An important feature of mitigation banking is that the responsibility for the successful implementation of the mitigation requirements is transferred from the developer to the bank. As a practical matter, this provides a better opportunity to ensure the mitigation’s success, since the bank sponsor is more likely to remain on the scene for a longer time. The 1995 Guidance attempts to ensure the success of mitigation at approved banks by requiring that bank sponsors secure sufficient financial assurances to monitor and maintain the bank and to cover future contingencies. Such assurances can take the form of performance bonds, casualty insurance, or other mechanisms. Similar requirements are seldom imposed on those undertaking traditional project-by-project mitigation.

The federal guidance on wetland mitigation banking is similar to the guidance of many states. Although many states have promulgated laws, policies, or regulations pertaining to wetland mitigation banking, only California has adopted a policy governing endangered species mitigation banking. The California policy was promulgated in April 1995 as an initiative of the then governor, Pete Wilson. The roots of that initiative can be traced to federal efforts, under the ESA, to protect a California songbird, the coastal California gnatcatcher.

In mid-1990, the Service began a formal review of the status of the gnatcatcher to determine whether it warranted protection under the ESA. That review set off alarm bells in southern California, as development interests there recognized that federal listing of this bird could have an enormous impact on them. The reason was straightforward. The gnatcatcher depended on coastal sage scrub habitat, and it was precisely that sort of habitat that was being rapidly converted to subdivisions, shopping malls, highways, and other uses throughout Southern California. Once the gnatcatcher was listed as an endangered or threatened species, further development could bump up against the stringent requirements of the ESA.

The gnatcatcher was just the tip of the iceberg, as dozens of other species were largely confined to coastal sage scrub habitat. These, too, were potential future additions to the federal endangered species list. Developers foresaw that even if they could satisfy the requirements applicable to the gnatcatcher, the subsequent listing of other species would put still more obstacles in their path. They desperately needed a solution that would relieve them from the need to run a new series of gauntlets as each new species was added to the list.

On Earth Day in 1991, Governor Wilson unveiled a new initiative that he called "Natural Community Conservation Planning" (NCCP). Its core idea was to initiate conservation planning on a broad geographic scale. The aim of such planning was to protect sizable and strategically situated parcels of land that could perpetuate an entire natural community, rather than just one or a few of its constituent species. The hope was that by doing so, species not yet on the endangered list would never be put on it and those already on the list could be protected in enough key places that they would no longer need the protection of the federal law. The first test of the new idea was to take place in southern California.

In September 1991, the Service formally proposed adding the gnatcatcher to the endangered species list. The following month, Governor Wilson signed into law the Natural Community Conservation Planning Act. Over the next eighteen months, the state and the Service worked to make sure that if the gnatcatcher were to be federally listed, the NCCP initiative could provide the framework within which the ESA’s requirements would be met. In March 1993, the Service designated the gnatcatcher a "threatened" species. This designation enabled it to customize a set of regulations that promised relief to developers and local governments if they took part in the NCCP process.

The NCCP process initially divided the coastal sage scrub habitat into eleven planning subregions in five counties. Each subregion was to develop a habitat conservation plan (HCP) that met the requirements of the ESA and the NCCP law. While the subregional plans were being developed, a limited amount of incidental taking of gnatcatchers was allowed under the special regulations promulgated for that species.

It quickly became apparent that the strategy likely to be applied to all the subregional plans would be to establish a system of preserves. These preserves would be composed of "key" parcels that, because of their size, location, and other characteristics, had special conservation value. An important question then was how to generate the financing that would be needed to acquire these key parcels. Local governments in California had been prohibited from raising property taxes by Proposition 13, a voter initiative passed many years earlier that forced local governments to resort to more creative means of funding services and infrastructure. Although the state had land acquisition resources, these were almost certain to fall well short of what would be needed. Perhaps the owners of the key parcels might simply be prevented by regulatory action from developing them, but that approach would face the near certainty of litigation in support of a constitutional claim for compensation. It was against this backdrop that the idea of "conservation banking" began to receive attention.

Six months after the gnatcatcher was added to the threatened list, the Bank of America foreclosed on a 263-acre parcel in San Diego County known as the "Carlsbad Highlands." An appraisal at the time of foreclosure gave the parcel a low value, primarily because of environmental constraints, including the presence of coastal sage scrub habitat that limited its development potential. Facing major regulatory hurdles if it tried to develop the property, the Bank of America looked instead for opportunities to benefit from the land’s ecological value. At the same time, the state’s highway department, CalTrans, was seeking ways to mitigate the impact on the gnatcatcher of one of its projects. Eventually, the two found each other. CalTrans bought eighty-three acres of the Carlsbad Highlands property and placed a conservation easement on it to mitigate its highway project.

The sale of a part of the parcel to CalTrans left the Bank of America with 180 acres of undeveloped and ecologically significant land, and it was eager to unload the remainder for a similar purpose. Because of the Bank of America’s prominence in the state, it was able to draw attention to a new state policy that was about to be publicly unveiled.

On April 7, 1995, the heads of California’s two principal environmental agencies, the Resources Agency and the California Environmental Protection Agency, jointly issued their "Official Policy on Conservation Banks." At the same time, the state announced that the Bank of America’s Carlsbad Highlands property had been transformed into the first conservation bank created under the new policy.

California’s "Official Policy on Conservation Banks" is a brief document with a broad scope. It provides guidance on the use of banks to compensate for impacts on both wetlands and endangered species, as well as on "Environmentally Sensitive Habitat Areas, mudflats, sub-tidal areas, and less sensitive resources." The broad scope of the policy reflects the several different environmental laws in California that may require mitigation, including the California Environmental Quality Act (mitigation required for projects that "substantially diminish habitat for fish, wildlife or plants"), the California Coastal Act (requiring applicants for projects along the coast to demonstrate that all feasible mitigation measures have been provided to minimize adverse environmental effects), and the California Endangered Species Act.

Interestingly, although all these various laws impose "mitigation" requirements, the policy never uses the term mitigation banks. Instead, it refers throughout to the banks it seeks to encourage as conservation banks. At least in part, this terminology reflects a conscious choice to try to avoid the controversy that long dogged the wetland "mitigation banks." By calling their banks something different, California policymakers apparently hoped that those who had already made up their minds about "mitigation banks" would be willing to take a fresh look at "conservation banks."

Apart from their names, however, there are few real differences. One rather conspicuous difference between the California policy and the federal wetland mitigation banking guidance is the treatment afforded "preservation" as a source of bank credits. Recall that under the federal wetland banking guidance, the preservation of existing resources is permitted as a source of credits only under "exceptional circumstances." Instead, the restoration and creation of wetlands are the preferred activities. Under the California policy, however, no presumption against preservation is made. Indeed, it is listed first among the four types of resource management that can generate credits. The others are resource enhancement (the enhancement of a degraded resource), resource restoration (the restoration of a resource to its historical condition), and resource creation (the creation of a specified resource condition where none existed before). In practice, most of the conservation banks thus far approved in California simply preserve existing parcels and promise little or no resource creation, restoration, or enhancement.

Like the federal guidance, the California policy allows advance crediting, or the recognition and use of credits before the "full realization of the targeted resource value at the bank." The availability of credits is to be determined "in accordance with agreed upon performance criteria for the development of the resource value in question." The California policy has one distinctive feature with respect to credits, however. It provides that "upon sale of the first credit in the bank area or subarea, the land in the bank or subarea must be permanently protected through fee title or conservation easement." Since the policy requires that any bank and each of its subareas be "large enough to be ecologically self-sustaining or part of a larger conservation strategy that has a reasonable expectation of being accomplished," this stipulation serves an important conservation purpose: it ensures that the mitigation associated with any sale of credits will be neither too small nor too transitory to be ecologically significant.

From the point of view of California’s bankers, however, the requirement to convey an easement over an entire bank area or subarea upon the sale of the first credit is problematic. According to interviews with several of them, they believe the resource agencies have no motivation to help complete the sale of credits in a bank once an easement is in place. Rather, the bankers believe that once the first credit has been sold and an easement is in place, the agencies will regard the site as locked into conservation and therefore prefer to seek new sites to lock into conservation rather than to direct project proponents to banks with remaining, unsold, credits. If the bankers’ perception is correct, this requirement will ultimately be self-defeating for the agencies, since it will discourage entrepreneurs from entering conservation banking, exactly the opposite result from what the California policy seeks. In the meantime, bankers are trying to overcome this difficulty by writing into their banking agreements stipulations that will obligate the resource agencies to inform potential credit buyers of the availability of credits at their banks and that will require the agencies to give them a sort of "most-favored nation" guarantee. Such a guarantee would prevent the agencies from recognizing credits as having a mitigation value that is less on an acre-for-acre basis than that of any other mitigation banking or habitat mitigation credit program available for the same species in the same area.

Although the California policy says nothing about the use of existing public lands for conservation banking purposes, this issue soon came up when San Diego County asked for the views of both the California Department of Fish and Game and the Service about the suitability of county-owned lands for such purposes. The two agencies prepared a joint response concluding that several types of county land would not be appropriate for banking purposes, such as (1) land used as mitigation for a previous project; (2) land already designated or dedicated for passive park or open space use, when that use was generally compatible with sustaining biological values; and (3) land acquired or given for park or natural open space purposes. This response does not seem objectionable if preservation is the only activity that the bank is contemplating: the position of the two agencies prevents the use of lands already committed to preservation. A somewhat different response might have been appropriate for banks contemplating resource creation or enhancement. As mentioned earlier, the federal wetland mitigation banking guidance recognizes this distinction and allows mitigation credits to be earned on publicly owned lands, but "based solely on those values in the bank that are supplemental to the public program(s) already planned or in place."

The joint response of the California Department of Fish and Game and the Service to San Diego County’s request illustrates the coordination between these two agencies. Although the California policy on conservation banking does not have the Service’s imprimatur, the two agencies did jointly issue in 1996 its "Supplemental Policy Regarding Conservation Banks Within the NCCP Area of Southern California." This supplemental policy, which begins by noting that the two agencies "support the creation of conservation banks," largely fills in some of the interstices of the state policy, including that the number of banks will be determined by the free market rather than by the agencies and that banking agreements should specify the service areas within which their credits may be used. It also tries to offer some guidance on what has become a contentious issue regarding the interchangeability of species and their habitats. According to the supplemental policy, in general only in-kind mitigation (involving the same habitat and species) is permitted. However, an exception to this in-kind mitigation requirement will be made when "the bank is located within a jurisdiction that has an approved subarea plan, or if the wildlife agencies determine that the bank achieves regional conservation goals." That exception was at issue in the only litigation thus far concerning endangered species mitigation banking.

The premise of the NCCP program was that it would focus on distinctive natural communities (like the coastal sage scrub), rather than on their constituent rare species. The ESA, however, focuses on individual species, generally prohibits their taking, and requires mitigation when their taking is allowed. Natural communities can sometimes usefully serve as a rough surrogate for the individual species generally associated with such communities, but the fit is not always precise. For example, if a landowner develops a one-hundred-acre parcel of coastal sage scrub habitat that is occupied by California gnatcatchers, it is by no means clear that the loss of this particular habitat could be satisfactorily mitigated under the ESA by protecting another one-hundred-acre parcel that is not occupied by gnatcatchers (although even an unoccupied parcel may be of conservation value if, for example, it is situated in a key corridor connecting major occupied sites). Yet the NCCP process tries to avoid this very need to account for each and every species on each and every parcel.

This issue surfaced recently in the only court challenge thus far to an endangered species mitigation bank. In San Bernardino Valley Audubon Society v. Metropolitan Water District, the narrow issue decided was whether the creation of an endangered species mitigation bank required the preparation of an "environmental impact report," or "EIR," under the California Environmental Quality Act (under this law, an EIR is essentially analogous to an "environmental impact statement" under the National Environmental Policy Act). Although the trial court found no need for an EIR, the court of appeals reversed the trial court’s decision.

The mitigation bank in the San Bernardino case was established by the Metropolitan Water District (MWD) primarily for its own future projects in southern California. However, the agreement creating the bank anticipated that there would likely be credits in the bank beyond those needed for the MWD’s projects and provided that these could be sold to third parties. It was this potential for sale to third parties that most concerned the court, plus the fact that the bank purported to provide mitigation for projects affecting some sixty-five different "target species." In order to establish an equivalence of values when mitigating the impacts of future projects by the purchase of credits from the bank, the agreement adopted "a complex habitat value formula to match habitat values in affected outside project areas to the available mitigation bank credits." The parties had different interpretations of how this complex formula was to work in practice, and the court declined to analyze it in detail. Nevertheless, in a key passage, the court concluded that there was "a fair argument that mitigation banking on a habitat basis will allow for a result different from an acre-for-acre or specie-by-specie [sic] exchange." To illustrate its concern, the court offered the following example: "If an outside project has six endangered or threatened species on one acre, it appears . . . that the habitat value equivalent of one acre of the mitigation bank could be used to provide mitigation for all six species. This compression of habitat could have a significant effect on the six species." The court went on to agree with the plaintiff’s concern that significant effects were possible because "occupied habitat may be replaced by unoccupied habitat in the mitigation bank."

The court’s decision would appear to create a significant problem for using habitat-based rather than species-based mitigation banks, at least when the habitat encompassed more than one species of concern. Since a major goal of the NCCP was to design a system of reserves that in their aggregate would meet the conservation needs of a whole suite of ecologically associated species without the need to account for each and every species on each and every parcel, the court’s decision potentially undermines the ability to use mitigation banks as part of the NCCP’s strategy. However, it may be noteworthy that the case addresses only the California Environmental Policy Act and not the California or federal Endangered Species Acts. The court did not rule on the compatibility of the mitigation bank with the requirements of either of those laws. It held only that the bank’s approval required the preparation of an EIR because there was a "fair argument" that the approval would have a significant effect on the environment.

Notwithstanding the recent court decision, banks continue to be established under the California banking policy. Some appear to have had notable success, whereas others have struggled. The case studies in Appendix I illustrate both the challenges and opportunities that have faced conservation bankers in California.

The ESA prohibits the "taking" of endangered animals, a prohibition that has been interpreted quite broadly to include the destruction of habitat under some circumstances. This prohibition, however, is not absolute. The government may permit landowners (and others) to take endangered species incidental to otherwise lawful activities. To secure such a permit, a landowner (or other entity) must prepare an HCP that mitigates the impact of the authorized taking.

Three basic approaches to mitigation are reflected in HCPs. The simplest is a plan for a single landowner to take prescribed measures on his own land to mitigate the impacts of the authorized activity. This approach is functionally equivalent to the traditional project-by-project, on-site mitigation typical of the Clean Water Act’s implementation.

A much more complex mitigation scheme is representative of the many HCPs initiated by local governments on behalf of some or all landowners in their jurisdictions. A common arrangement under these HCPs is for the local government to agree to purchase identified land parcels and to manage them for conservation purposes. In return for that mitigation, landowners whose land has not been identified for acquisition can develop it as they choose. Often, the funds for the acquisition and management of the conserved parcels are generated from special assessments levied on other land as it is developed. These arrangements are thus closely akin to the "in lieu" mitigation arrangements widely used under the Clean Water Act. Under these, in lieu of actually performing wetland mitigation, a developer pays a predetermined sum to a public or private conservation agency, which then uses the money to support wetland conservation activities.

The third form of mitigation is mitigation banking. This is the acquisition (usually by purchase from a third party, at a negotiated price) of conservation "credits" that have been recognized by the appropriate regulatory agency (typically the Service) as a result of some action--such as the preservation, creation, restoration, or enhancement of habitat. This sort of private entrepreneurial mitigation banking is exemplified by the "butterfly bank" mentioned in the opening paragraph of this report. Thus far, mitigation banking has been used relatively infrequently under the ESA, although interest in it appears to be growing.

Although the Service has authorized the purchase of credits from various banks to mitigate adverse impacts to endangered species, it has not yet issued any formal or informal policies with regard to endangered species mitigation banking. As just described, the state of California’s 1995 "conservation banking" policy purports to govern wetlands, endangered species, and certain other forms of mitigation banking. The federal wetland mitigation banking guidance and the California conservation banking policy differ significantly. Whether either of these serves as a good model for the development of federal endangered species mitigation banking policy is open to debate. It is clear, however, that any effective federal endangered species mitigation banking policy must address at least those issues explored next. The recommended resolution of these issues is given in the draft mitigation banking policy in Appendix II.

Federal wetland mitigation banking policy strongly discourages the preservation of existing wetlands as the basis for generating bank credits and favors the restoration of former wetlands or the enhancement of degraded wetlands. When preservation is accompanied by restoration, creation, or enhancement, the federal policy allows the generation of credits for the preservation if "it is demonstrated that the preservation will augment the functions of the restored, created or enhanced aquatic resource." However, the "preservation of existing wetlands . . . may be authorized as the sole basis for generating credits . . . only in exceptional circumstances." The rationale for this policy preference is that most existing wetlands are already "protected" (at least from being filled) by Section 404. Thus, the preservation of already protected wetlands as mitigation for the loss of other existing wetlands would result in a net loss of wetlands.

Do the same policy considerations apply to endangered species habitats? Should an endangered species mitigation banking policy also generally prohibit the preservation of existing habitats as a means of generating credits? The California conservation banking policy has no comparable provisions against mitigation through preservation, and most of the conservation banks established in that state simply preserve existing habitats. At first glance, there appear to be some important differences that may justify an approach more receptive to preservation in the case of endangered species. First, habitat that is currently not occupied by an endangered species is also not protected by the ESA, yet it may nevertheless be valuable for the species in the future. Preserving suitable but currently unoccupied habitat may make it more likely for the species to expand its current distribution and to recover. Second, there is no prohibition against taking endangered plants and therefore no prohibition against destroying their habitat. That is, endangered plants and their habitat are not "protected," at least not off federal land. Finally, although endangered animals benefit from the ESA’s taking prohibition, that prohibition is no guarantee that today’s habitats will still be tomorrow’s. Successional change, catastrophic events, and even the absence of natural disturbances can result in the long-term or permanent loss of endangered species from the habitats they now occupy. Once endangered animals are gone, the law’s taking prohibition no longer matters.

Based on these considerations, any formal policy governing endangered species mitigation banking should embrace preservation as an appropriate means of generating credits, though it should note that in some exceptional circumstances, preservation may not be appropriate. For example, if currently occupied habitat is likely to continue to be occupied more or less indefinitely without active management, then the protection afforded that habitat by Section 9 of the ESA may be sufficient to ensure the site’s long-term protection. In those circumstances, agreement by the landowner to protect a site that is already, as a practical matter, likely to be protected anyway, should not give rise to mitigation credits.

Because endangered species habitats often require some form of active management to continue their value to endangered species, a mitigation banking policy should require that preservation of existing sites be accompanied by a strong management commitment (and the resources to fulfill it). Without a commitment to manage and the resources to do so, mitigation credit could be given for sites that will eventually cease to offer any benefit to the species they were intended to help conserve.

Federal wetland mitigation banking policy generally prefers that banks "service" nearby areas, usually in the same watershed. That is, banks should provide mitigation credits to offset the loss of wetlands in the same watershed (or smaller area). The rationale for this policy is that wetlands provide many site-specific benefits (e.g., stormwater retention, groundwater recharge, pollution filtration, biodiversity, and recreation) that cannot be replaced by compensating in another watershed. Do valid policy considerations dictate a similar approach for endangered species?

Again, the differences between wetlands and endangered species may justify a different result. First, except for aquatic species, "watersheds" do not often serve as useful demarcations for endangered species. Second, if the ESA’s goal is to improve the survival prospects of a species sufficiently that it can be considered recovered, that goal is to some degree independent of where a species is found today. Although endangered species offer significant local benefits, the ESA is not principally concerned with those local benefits. Rather, its overriding concern is that a species be sufficiently secure in enough places that it is not likely to become endangered again in the foreseeable future. To meet this goal, it may be enough to ensure the species’ survival in some, but not all, the localities where it now occurs.

These considerations suggest that recovery plans or similar undertakings should guide the determination of bank service areas. For example, many recovery plans set goals of establishing one viable population on protected habitat in each of several identified geographic areas. At a minimum, those geographic areas should serve presumptively as banking service areas: banks located in a given area can sell credits to offset impacts elsewhere in the same area. In some circumstances, however, credit trading across recovery areas might serve recovery goals (e.g., when recovery goals for the red-cockaded woodpecker have been met in the Sandhills regions of North Carolina, future mitigation for impacts in the Sandhills could properly go toward meeting recovery goals in others areas where recovery goals have not yet been met). Such an approach would both hasten the recovery of endangered species and enhance the economic viability of mitigation banks.

Where no recovery plan exists, or where it fails to specify recovery goals for specific areas, the location of mitigation banks is probably more important than the delineation of their service areas. Typically, the goal should be to locate mitigation banks where they can support (or contribute to) viable populations by linking or buffering already protected areas. By securing these linkages or bufferings, recovery prospects for an imperiled species may be improved, notwithstanding the loss of habitats elsewhere that are isolated and too small to support viable populations. Thus, in these circumstances, the service area is essentially unlimited, with strategically sited banks able to sell credits to offset impacts anywhere else that the species occurs. The important point, however, is that such banks must genuinely be "strategically sited," with a particular conservation objective clearly in mind.

An important and difficult issue in wetland mitigation banking is establishing a common currency for valuing credits created by restoration, enhancement, or other activities at the bank site and similarly valuing what is lost at the impact site. Wetland mitigation policy seeks to ensure no net loss of wetland "function," but in the absence of a clear measure of function, acreage often serves as a rough surrogate for it. How should the currency of credits and debits be measured in the case of endangered species mitigation banks? Will the answer to this question be different for each species? Would it ever be appropriate to use some measure of habitat as the currency, without knowing the precise extent of use of that habitat by a particular species? Any useful policy governing endangered species mitigation banking must resolve these issues.

In theory, any action that either hurts or helps an endangered species can be reduced to a common currency, the likelihood that the species will survive for some period of time. If we assume that a particular endangered species has a 50 percent probability of surviving for another hundred years and that a proposed development action would, if unmitigated, reduce its survival probability to 40 percent, then mitigation measures would fully compensate for the development impacts if they restored the species to a 50 percent survival probability.

Unfortunately, as a practical matter, our ability to quantify precisely current survival probabilities and the impacts of helpful or harmful actions is rudimentary to nonexistent. But even though they cannot quantify such probabilities precisely, regulators nevertheless base many of their mitigation requirements on intuitive or crudely quantified assessments of such probabilities or use a convenient substitute. For wetland mitigation, wetland acres often serve as a convenient surrogate for a more sophisticated assessment of functions and values lost and gained. For endangered species, acreage is more of a problem, because the value of any given acre to a particular endangered species always depends on a host of variables. These variables include configuration, the size of the habitat patch of which it may be a part, its proximity to other habitat patches, its position upwind or downwind of pollution sources, the presence or absence of exotic species, its overall condition, and others. We have no neat formula by which to weight each of these many variables and produce a meaningful index value to assign to the acre.

Because of these difficulties, the practice of existing mitigation banks varies widely. A recent agreement establishing a North Carolina Department of Transportation bank for the red-cockaded woodpecker refers to mitigation credits without ever defining what they are. The International Paper agreement profiled in Appendix I concludes that a credit is earned when a new "active cluster" (i.e., a spatially associated group of trees with one or more cavities currently being utilized by one or more red-cockaded woodpeckers) is established at the mitigation site. Under the proposed Saipan Upland Mitigation Bank, also described in Appendix I, a mitigation credit consists of protecting habitat associated with an existing pair of nightingale reed-warblers (Acrocephalus luscinia), plus enhancing sufficient unoccupied habitat to provide the conditions needed by a pair of reed-warblers. Some of the conservation banks in California equate credits with acres of a particular type of habitat preserved or destroyed. These varied approaches reflect not merely the absence of a clear policy but also the practical reality that the varied circumstances and needs of particular species will inevitably produce different "currencies" to define bank credits and debits.

Federal wetland mitigation banking policy limits the use of mitigation banks on publicly owned lands. Part II.B.2 of that policy provides that banks may be sited on public or private lands. However, the policy goes on to say that credits generated by mitigation banks on public lands "should be based solely on those values in the bank that are supplemental to the public program(s) already planned or in place, that is, baseline values represented by existing or already planned public programs, including preservation value, should not be counted toward bank credits." On September, 10, 1999, the U.S. Fish and Wildlife Service issued a policy that generally prohibits compensatory wetland mitigation on National Wildlife Refuges because the Service is already "authorized to restore degraded habitats within the National Wildlife Refuge System and . . . will be restoring these lands in the future, irrespective of off-Refuge development." Ironically, in the case of endangered species, it is not uncommon for the Service to approve HCPs in which the mitigation for the loss of habitat on private land is carried out on public land, including federal land, even though federal land managers are also authorized (indeed, arguably required) to restore and improve habitats for endangered species on their land.

The argument for allowing mitigation on public land is that public agencies may lack the resources to carry out the restoration, enhancement, or management activities on their land that would help endangered species. Supplementing agency budgets with revenues derived from mitigation assessments enables conservation activities to be completed sooner on public lands than would otherwise be the case. That argument is sometimes true, yet it is troubling. There is a danger that the private resources contributed by the banker will not simply supplement agency budgets but will also displace them. As revenues for conservation management from private banking sources rise, appropriated public funds for those same activities may fall. To the extent that happens, the apparent mitigation on public lands will fail to compensate for losses elsewhere. This is especially worrisome in the case of federal lands, because of the affirmative duty imposed on federal agencies by the ESA to promote the conservation of endangered species. Another concern is that allowing banks to be located on public lands (federal or nonfederal) may create incentives for entrepreneurs to enter the banking business, but it does not create an incentive to use private land for conservation purposes.

For these reasons, a formal policy governing endangered species mitigation should generally disallow the use of federal lands for mitigation purposes, including mitigation banking. For other public lands (e.g., state or local), a formula similar to that in the federal wetland banking interagency guidance seems appropriate. The number of credits in such circumstances "should be based solely on those values in the bank that are supplemental to the public program(s) already planned or in place."

Federal wetland mitigation banking policy allows limited "advance sales" of credits (i.e., credits can be sold before the mitigation has been successfully carried out) in order to capitalize the banks. Bankers cite the need for a more liberal allowance of advance credit sales as a financial necessity, whereas many environmentalists urge that no credits be given until the mitigation is a proven success. Are the considerations different for endangered species than they are for wetlands? This is not an issue when the preservation of existing habitats generates mitigation credits, because the credits are earned as soon as the site is dedicated to conservation purposes. Note also that landowners who have created, restored, or enhanced endangered species habitat under "safe harbor" agreements may be able to become mitigation bankers by relinquishing their rights to remove the habitat improvements they have made. When the habitat has been improved under a safe harbor agreement, the concern about advance credit sales also disappears, because the credits are realized as soon as the landowner gives up his rights to return to baseline conditions under the safe harbor agreement.

In other circumstances, when a bank wants to earn credits by restoring or enhancing currently unoccupied habitat, the question of whether any credits should be available for sale before the restoration or enhancement has begun (or before its success has been demonstrated) should perhaps hinge on whether the restoration or enhancement techniques are well established and known to be generally successful or whether they are largely experimental. If they are well known and generally successful, advance credit sales could be permitted with greater confidence. Conversely (and probably more typically), if they are experimental and the likelihood of their success is highly uncertain, advance credit sales should not be permitted. A formal endangered species mitigation banking policy should try to reflect these principles, though any rule short of an absolute prohibition against advance credit sales will unavoidably put a considerable amount of discretion in the hands of those who implement the policy at the field level.

Federal wetland mitigation banking policy requires that once all the credits in a bank have been sold, the underlying property must be given permanent protection through deed transfer or conservation easement. Should that same policy always apply to endangered species mitigation banking, or in some circumstances can a less-than-permanent-protection arrangement legitimately generate credits?

Some endangered species occupy early successional habitats that in time will mature and no longer support those species; other endangered species may survive on sites that are so small and so isolated that the species has a low probability of remaining on the site much longer. Once such habitats are no longer occupied by an endangered species, no ESA prohibitions encumber those sites, at least according to how the law has been interpreted thus far. Therefore, when the owners of such sites want to develop them and are required to mitigate that development, it is at least an open question whether the required mitigation should necessarily be in the form of a site protected in perpetuity. One might address this problem in either of two ways: (1) compensating for the loss of transitory habitats with less-than-permanent mitigation or (2) insisting that mitigation always be permanent but adjusting the mitigation ratios downward when the habitat being lost is transitory. A formal endangered species mitigation banking policy should probably favor the permanent protection of the mitigation site while allowing one or both of the preceding options in appropriate circumstances.

Whatever the duration of the mitigation commitment, it is important to have in place both the resources and an effective mechanism for maintaining the biological values of the mitigation site throughout that period. Often, mitigation sites are subject to edge effects from incompatible land-uses on neighboring lands. Others suffer from a policy of "benign" management resulting in erosion, invasion by exotic plants, and a lack of physical security. When such sites are near population centers, active management is often necessary to prevent the loss of the very values that the sites were established to preserve. It is critical, therefore, that the expected costs of future management needs be assessed and that sufficient financial resources be available to meet those costs.

The basic yardstick for deciding how much is needed is the average annual cost of management over the long term. The objective is to establish a funding source that provides enough income to cover annual stewardship costs and that also keeps pace with inflation. Unfortunately, there is no easy way to determine this, and managers around the country are struggling to devise formulas for calculating these costs. The costs vary widely with the nature of the land, the type of ownership (fee or easement), the scope of permitted activities (public use, education, etc.), and year-by-year circumstances. California has a system of establishing expected management costs and the size of endowments needed to meet them that is widely used by both public agencies and nonprofit land trusts. This system was developed by the Center for Natural Lands Management and is known as the "Property Analysis Record" (PAR).

The PAR is a computerized database methodology that helps land managers calculate the costs of land management for a specific site. It does this by analyzing the characteristics and needs of the property from which the management requirements are derived. It helps specify the management tasks and estimates both their costs and the necessary administrative costs. The PAR generates a concise report that serves as a well-substantiated basis for long-term funding, including endowments. The development of the PAR was funded by the U.S. Environmental Protection Agency and the David and Lucille Packard Foundation. Seminars on its use have been funded by the National Fish and Wildlife Foundation.

As a general matter, wetland impacts must be mitigated in a prescribed sequence: first avoiding impacts, then minimizing unavoidable impacts, and finally compensating for the remaining impacts. The federal wetland mitigation banking policy purports to be faithful to this policy, although some conservation interests worry that the practical effect of the mitigation banks is to tempt regulators to skip rather lightly past avoidance and minimization and proceed instead directly to compensation in the form of purchasing credits from a bank. Bankers deny that this is the case but urge that the policy be less demanding with respect to sequencing. The argument for sequencing is that every wetland is important and that our ability to compensate for wetland losses through restoration and enhancement (or wetland creation) is imperfect at best.

Are those arguments as persuasive for endangered species? The idea that it is always better to protect any habitat that is currently occupied by an endangered species than to compensate for its loss elsewhere is hard to defend, given that many currently occupied habitats are practically indefensible in the long run because they are too small, too degraded, or too isolated to contribute meaningfully to the recovery of the endangered species now occupying them. Even though our ability to create, restore, or enhance habitats for endangered species is clearly imperfect, in some cases it may be less imperfect than our ability to maintain currently occupied habitats so that they remain occupied.

Perhaps because of these differences between wetlands and endangered species, there has never been, for endangered species mitigation purposes, a formal sequencing requirement comparable to that used for wetlands. It is worth noting that whereas avoiding wetland impacts serves the stated goal of no net loss of wetlands, merely avoiding endangered species impacts does not necessarily advance the goal of recovering imperiled species. By itself, it simply perpetuates the existing, unsatisfactory status quo. In most instances, to enable recovery, those areas currently lacking formal legal protection or beneficial management must receive it, and those areas not currently occupied by endangered species must become occupied, or both. Unless public funds to accomplish this are made available, the only alternative may be to do so through well-designed mitigation programs that trade other areas less important to the recovery of the species. For these reasons, endangered species mitigation banking policy need not track the sequencing requirements found in the wetland banking guidance.

Some wetland mitigation banks can sell credits to compensate for wetland losses in connection with only some kinds of projects. For example, some can sell credits only to mitigate the impacts of small projects authorized by nationwide general permits; others can sell credits only for projects affecting wetlands beyond the Corps’s jurisdiction (but still regulated by state or local laws). The rationale for these limitations is that larger projects should mitigate on-site wherever possible, whereas mitigation for very small projects on-site is often expensive, of little ecological value, and likely to fail. Thus, banks provide a convenient mitigation alternative for the less sophisticated and less wealthy "little guys" who are attempting to satisfy Section 404’s requirements. Should endangered species mitigation banks also be designed primarily to serve little guys (e.g., those that qualify for "low-effect" HCPs)?

Although some banks may be purposely designed to cater to some subset of projects requiring endangered species mitigation, there does not appear to be a compelling reason that endangered species mitigation banking policy should limit the availability of bank credits to any particular category of potential purchasers. As a practical matter, bigger projects affecting larger areas may have more on-site mitigation opportunities available to them, but even for these projects, the alternative of buying credits from a strategically located and well-designed bank may yield greater environmental benefits.

Wetland mitigation banks may be either publicly or privately owned. Private bankers complain that publicly owned banks compete unfairly because (1) they use underpriced public goods; (2) they are sometimes owned by the same agencies that act as regulators, who can steer permittees toward their own banks rather than the private banks; and (3) public banks are generally not required to meet the same financial assurance requirements that private banks must meet (e.g., posting of performance bonds). Some independent observers echo these criticisms and urge that public agencies stay out of the banking business (other than to generate credits for their own internal use). Should endangered species mitigation banking policy allow or encourage both public and private mitigation banks?

Some public endangered species mitigation banks (e.g., the North Carolina Department of Transportation bank) may be established primarily or exclusively to meet the future mitigation obligations of the public agency establishing the bank. Unless they sell credits to private, third parties, they do not compete with private mitigation banks in any meaningful sense. Conversely, the proposed Saipan Upland Mitigation Bank is a public bank established for the purpose of facilitating private development by selling mitigation credits to private developers. Its circumstances, however, are unique because of the legal restrictions on who may own land in the Commonwealth of the Northern Mariana Islands. Elsewhere, publicly owned mitigation banks offering credits to third parties could stifle the development of genuine entrepreneurial banks. Although that would be an undesirable result, preventing it by means of an outright prohibition against public banks could have undesirable conservation consequences, too, particularly if the market for endangered species mitigation banking remained too uncertain to attract much private investment. To the extent that mitigation banking policy addresses this issue at all, it ought to seek to minimize the potential conflicts of interest in which public agencies serve as both regulators and mitigation bankers.

Federal wetland mitigation banking policy imposes uniform requirements on federally approved mitigation banks. The process of creating banks is prescribed; their basic elements are delineated; and other features are mandated in a policy that stresses uniformity rather than creativity, flexibility, and spontaneity. Should that also be the case for endangered species, or are the circumstances concerning endangered species too varied, the experience with mitigation too limited, and the potential for banking too fluid to warrant similar uniformity? Should endangered species mitigation banks be allowed to develop more or less spontaneously out of safe harbor agreements or other arrangements?

At least initially, the argument for spontaneity appears persuasive. An endangered species mitigation banking policy should aim at preventing potential abuses and securing real conservation gains. At the same time, however, it should avoid overly prescriptive and rigid rules that are unlikely to fit well with the many different circumstances that a host of different endangered species with highly varied needs will present.

SELECTIVE BIBLIOGRAPHY

In addition to the sources cited in the footnotes, the following bibliography identifies information sources that should be useful to anyone seeking a greater understanding of mitigation banking. Although the literature on mitigation banking is now extensive, almost all of it pertains to wetland mitigation banking, with as yet, nearly no published literature on endangered species mitigation banking.

Although dated, one of the most comprehensive sources of information about wetland mitigation banking is

Environmental Law Institute. Wetland Mitigation Banking. Washington, DC: Environmental Law Institute, 1993. 159 pp. plus appendices.

In addition to its own insightful analysis, this report includes an extensive annotated bibliography. A revised version of this report is included with the several studies done as part of the National Wetland Mitigation Banking Study by the Institute for Water Resources of the U.S. Army Corps of Engineers. The several reports that comprise that study are available from the Corps or can be downloaded from the web at

Several of the reports in the National Wetland Mitigation Banking Study were written before the 1995 federal guidance on wetland mitigation banking. This Guidance is thoughtfully and critically examined in

Gardner, Royal C. Federal Wetland Mitigation Banking Guidance: Missed Opportunities, 26 Envt’l L. Rep. 10075-10079 (February 1996).

A longer and more detailed account of the guidance and the history that led up to it is by the same author,

Gardner, Royal C., Banking on Entrepreneurs: Wetlands, Mitigation Banking, and Takings. 81 Iowa L. Rev. 527-587 (1996).

Another post-Guidance source of useful information about wetland mitigation banking and the various perspectives on it can be found in the following two recent congressional hearings:

Hearings on Wetlands Mitigation Banking Before the Senate Committee on Environment and Public Works. March 16, 1996, S. Hrg. 104-644, 104th Cong., 2d sess., 242 pp.

Hearings on Wetlands Protection and Mitigation Banking Before the Subcommittee on Water Resources and Environment of the House Committee on Transportation and Infrastructure. December 9, 1997, H. Hrg. 105-49, 105th Cong., 1st sess., 135 pp.

The state of California not only has the only statewide policy on mitigation (or "conservation") banking that includes endangered species, but it also maintains a website where much information on that subject can be found. The website includes the official state policy, related press releases and explanatory material, and a catalog of conservation banks approved in the state. Unfortunately, the information in the catalog of approved banks has not been regularly updated and therefore must be used with caution. The website address is

http://ceres.ca.gov/topic/banking/

A good reference that spans wetland mitigation banking and the California conservation banking efforts is

Marsh, Lindell L., Douglas R. Porter, and David A. Salvesen. Mitigation Banking: Theory and Practice. Washington, DC: Urban Land Institute, 1996. 300 pp.

Although there is still relatively little information about mitigation banking specifically for endangered species, two references worth consulting are

Toyon Environmental Consultants, Inc. "Conservation Banking--A Technical Report." Prepared for the California Department of Fish and Game and the California Wildlife Foundation with funding from the Ford Foundation. Undated. 24 pp.

Lawhead, David. "Bucks in the Bank . . . Land Bank That Is." Outdoor California (July-August 1997), pp. 4-8.

This report was prepared by Michael J. Bean, Robert Bonnie, and Dr. David S. Wilcove of the Environmental Defense Fund, with the assistance of Lynn Dwyer, senior project manager; and Krista Thomas, project manager, of Sustainable Conservation. Many people from state and federal agencies, the private sector, and academia willingly shared their experiences, opinions, and recommendations regarding mitigation banking. The conclusions of this report, however, are solely those of the Environmental Defense Fund and Sustainable Conservation.

Among those who offered helpful insights are Dr. Charles Bruton, North Carolina Department of Transportation, Raleigh, NC; Donald Carr, Winthrop, Stimson, Putnam and Roberts, Washington, DC; Jack Chowning, U.S. Army Corps of Engineers, Washington, DC; Mary Lynn Coffee, TMC Companies, Newport Beach, CA; Craig Denisoff, Wildlands, Inc., Citrus Heights, CA; Nancy Gilbert, U.S. Fish and Wildlife Service, Carlsbad, CA; Kim Gould, Southern California Edison Company, Rosemead, CA; Mike Horton, U.S. Fish and Wildlife Service, Washington, DC; Terry Huffman, Huffman Associates, Larkspur, CA; Dr. Donald Hunsacker, The Environmental Trust, El Cajon, CA; Jim Jackson, Bank of America, San Diego, CA; Paul Jones, Environmental Protection Agency, San Francisco, CA; Chris Kelly, The Nature Conservancy, San Francisco, CA; David Lawhead, California Department of Fish and Game, San Diego, CA; Mark Littlefield, U.S. Fish and Wildlife Service, Sacramenton, CA; Michael McCollum, McCollum Associates, Sacramento, CA; Katharine Miller, Commonwealth of the Northern Mariana Islands Division of Fish and Wildlife, Saipan; Scott Miller, Northern Mariana Island College, Saipan; Jim Monroe, U.S. Army Corps of Engineers, Sacramento, CA; Jeff Olberding, Olberding Environmental, San Jose, CA; Todd Olson, TMC Companies, Newport Beach, CA; Carlton Owen, Champion International, Greenville, SC; George Platt, Florida Wetlands Bank, Inc., Fort Lauderdale, FL; Ron Rempel, California Department of Fish and Game, Sacramento, CA; Edwin G. Sauls, The Sauls Company, Laguna Beach, CA; Dr. Leonard Shabman, Virginia Polytechnic Institute, Blacksburg, VA; Candace Skarlotis, Bank of America, San Francisco, CA; Margaret Strand, Oppenheimer, Wolff & Donnelly, Washington, DC; Greg Sutter, Wildlands, Inc., Citrus Heights, CA; Ted Tchang, Tech-Bilt, Inc., San Diego, CA; Sherry Teresa, Center for Natural Lands Management, Fallbrook, CA.; and Benjamin Tuggle, U.S. Fish and Wildlife Service, Washington, DC. Special thanks go to Michael Hertel and Kim Gould, Southern California Edison, and the Bank of America Foundation. These organizations supported the work of Sustainable Conservation on this project. We apologize to any whose names we may have omitted.

Finally, we wish to acknowledge and thank the National Fish and Wildlife Foundation and the Sand County Foundation for their support of this report. We also thank the Doris Duke Charitable Foundation for its support or our work regarding mitigation banking and other potential incentives for conserving rare species on private land.

This appendix presents brief profiles of six endangered species mitigation banks. The first three are based in California, where mitigation banking for endangered species conservation purposes has been used far more widely than elsewhere. Under a 1995 state policy initiated by then Governor Pete Wilson, California’s "conservation banks" are specifically encouraged in areas where they support continuing regional habitat conservation strategies. Comparable to wetland banks, a conservation bank is privately or publicly owned land managed for its natural resource values. The banks are normally based on habitat and designed to conserve a number of rare species.

It is not surprising that endangered species banking should have gotten its start in California. It is the state with the second largest number of such species, and it has a pressing need to find creative new approaches to the growing conflict between growth and protection of the natural environment. Early architects of California’s conservation banking policy describe endangered species banking as a combination of mitigation banks created to create, restore, or enhance wetlands and a totally new form of natural resource banking focused on species and habitats.

California’s new enterprise, Natural Community Conservation Planning (NCCP), is designed to protect threatened and endangered species at the scale of the natural community. The proving ground for the program is the coastal sage scrub natural community and the range of plant and animal species that occupy this habitat mosaic across five southern California counties covering six thousand square miles. The NCCP is intended to be an improvement on the prior model of project-based, single-species conservation efforts. Under this new approach, wildlife agencies identify areas of important habitat for conservation before they are degraded or jeopardized by development. Planning is on a regional basis in order to maintain areas large enough to retain species diversity. Each subregion and locality is expected to contribute a share to the reserve system, and each has a predominant role in land-use planning.

One of the NCCP’s goals is to identify the funding for, and secure the management of, reserve lands from public and private sources. Conservation banks were originally envisioned as a tool to provide long-term protection of habitat and to offer landowners economic incentives to protect natural resources. Because the planning is on a large regional level and then broken down into a subregions and subareas, the region and the community know where the preserve system will be. The conservation banks are part of the effort to consolidate environmental mitigation requirements and apply them to high-priority sites for maximum regional environmental benefit.

Thus far, the NCCP has been the impetus for more than twenty conservation banks in California. Because the center of this tool has been San Diego County, we explore its advantages and disadvantages in two conservation banks from that area, the San Vicente and Manchester Avenue Conservation Banks. The Kimball Island Mitigation Bank, in the delta of the San Joaquin and Sacramento Rivers, is an example of a quite different bank that seeks to restore a mix of aquatic and riparian habitats. Of the remaining banks briefly summarized, two are for the red-cockaded woodpecker (the Southlands and Champion International Banks) and one is located on the island of Saipan in the South Pacific.

The San Vicente Conservation Bank started as a "feel-good" story, but its ending is still uncertain. The bank became the dream of the Boys and Girls Clubs of the East County Foundation after a wealthy donor deeded the land to the group. Without the associated costs of purchasing the land, the foundation hoped to finance its recreational, cultural, and physical education programs for disadvantaged youth by creating the bank from what had been a fifteen-hundred-acre cattle ranch.

The East County Foundation is a nonprofit organization with a mission "to inspire and enable all young people, to realize their full potential as productive, responsible and caring citizens." Other nonprofits have used the conservation banking tool in California, including The Nature Conservancy, the Trust for Public Land, and the American Land Conservancy. The experience of the Boys and Girls Clubs has been mixed and provides many useful lessons concerning the pitfalls as well as the opportunities of banking.

Biologically, the area of the San Vicente Conservation Bank is exceptional. The 320-acre bank is dominated by high-quality coastal sage scrub and southern mixed chaparral communities and houses numerous pairs of the threatened California gnatcatcher. It is located near a great deal of open space and undeveloped land, some private and some public, in the vicinity of the Sycamore Canyon County Open Space Preserve and the San Vicente Reservoir and along the boundary with the Iron Mountain Preserve. The bank is relatively undisturbed and requires minimal management such as maintenance of fencing and signage, control of exotic plants, removal of trash, and patrolling for trespassers which is done voluntarily by friendly, but protective, neighbors. Since the