Science and Uncertainty in Habitat Conservation Planning

Laura H Watchman, Martha Groom, John D Perrine. American Scientist. Volume 89, Issue 4. Jul/Aug 2001.

A town meeting in eastern Texas virtually erupts at the mention of a small songbird. The marbled halls of the U.S. Congress echo with angry debate over federal regulation of private property. Two thousand miles away in a rapidly growing area of Tucson, Arizona, a biologist’s heart races when she hears an endangered cactus ferruginous pygmy owl reply to calls from her tape recorder. Underlying these seemingly disparate scenes is a common theme that has recently risen to the forefront of both conservation science and national politics: how to effectively manage privately owned habitat for threatened and endangered species.

In the past decade, habitat conservation plans (HCPs) have emerged as tools of compromise, allowing nonfederal landowners to develop their land and thereby damage habitat for imperiled species, provided that the impacts are offset through minimization and mitigation measures. These plans have caused considerable controversy, largely because of uncertainty about their ultimate impacts on the survival and recovery of imperiled species. A recent exhaustive review of the use of science in HCPs has illustrated the threads of uncertainty in the planning process, and the results suggest some general measures that can reduce uncertainty for both landowners and wildlife.

Conservation and Controversy

The primary U.S. law protecting species from extinction is the Endangered Species Act of 1973, which promotes survival and recovery of imperiled species by prohibiting the “take” of individuals of species listed as endangered or threatened. Legally, take includes not only killing, harming or harassing listed species but also destroying the habitat on which these species rely. Habitat loss and degradation are the primary threats to imperiled species in the U.S., and most of these species have a significant proportion of their habitat on privately owned lands. This situation has created conflicts between landowners and federal regulators and created a dilemma for policymakers: Heavy-handed enforcement of the Endangered Species Act on private lands could erode public support for the law, but ignoring the need to protect wildlife on these lands could accelerate the decline of the very species the law was designed to protect.

In an effort to resolve this conflict, Congress added a provision to the Endangered Species Act in 1982 allowing landowners to receive permits to take imperiled species on their lands, provided that such takings were incidental to otherwise lawful activities such as construction or timber harvest. To receive such a permit, a landowner must first prepare a habitat conservation plan outlining how much take would occur as a result of the proposed activity and what steps would be taken to minimize and mitigate the impacts. The U.S. Fish and Wildlife Service (for terrestrial and freshwater species) and the National Marine Fisheries Service (for marine species) usually provide guidance and advice to the landowner developing the HCR The landowner then submits the plan to one or both of the these agencies for approval. If the agency determines that the plan minimizes and mitigates take to the maximum extent practicable, and that the taking is unlikely to jeopardize the continued survival and recovery of the imperiled species affected, the HCP is approved, the “incidental take permit” is issued, and the project may commence.

The first HCP, approved in 1983, addressed housing development on 3,600 acres of habitat on San Bruno Mountain near San Francisco, California. This area supported about 14 percent of the remaining population of the endangered Mission blue butterfly (Icaricia icarioides missionensis). The plan involved phasing construction to minimize habitat loss, conveying a portion of the land to county ownership for conservation and charging residents an annual fee to fund management activities such as habitat restoration and removal of exotic weeds.

After a slow start, HCPs have quickly become the primary means by which the federal government addresses imperiled species on private lands. Fewer than a dozen plans were implemented between 1982 and 1992, but the number of approved plans skyrocketed during the Clinton administration. By the end of 2000, more than 330 HCPs had been approved, and another 200 were under development or pending approval. While proliferating across the landscape, habitat conservation plans have diversified in size, duration, types of land uses, number of landowners involved and number of species addressed. A modern HCP might cover one acre of beach inhabited by a single imperiled species or thousands of acres of forest providing habitat for dozens of species of conservation concern.

Far from preventing conflicts, the HCP process has instead generated its own storm of controversy. Noting that habitat loss is the primary cause of population declines for most imperiled U.S. species, critics argue that the plans make an already bad situation even worse since they inherently result in additional habitat destruction. Proponents of the process, on the other hand, point out that enforcement of the Endangered Species Act’s taking prohibitions on private lands was weak to nonexistent and that HCPs can greatly benefit imperiled species by enticing landowners to protect habitat and mitigate take. Central to the debate is the following question: Do HCPs ensure some certainty of survival and recovery for imperiled species, or are they risky measures that further imperil species already menaced by extinction? The answer to this question depends in part on the quality of the scientific data and judgment used while preparing the plan.


In the fall of 1997, the National Center for Ecological Analysis and Synthesis (NCEAS) and the American Institute of Biological Sciences (AIBS) sponsored an evaluation of the use of scientific data and analyses in habitat conservation plans. Peter Kareiva, then with the University of Washington, organized a team of 13 faculty members and 106 graduate and undergraduate students at eight major universities to conduct the evaluation. The team dissected 43 HCPs representing the full range of size, duration, geographic location and other characteristics typical of those approved to date, addressing 97 imperiled species from longhorn beetles to loggerhead sea turtles.

To determine what scientific information was available to each plan’s developers, the team performed an exhaustive search of the literature, both peer-reviewed and “gray” (non-peer-reviewed reports prepared by government agencies, consulting firms and other sources), and conducted interviews with key participants (such as landowners, consulting firms and agency biologists) who prepared and implemented the HCPs in question. Then the review team scrutinized the plans and their numerous supporting documents to track how well they incorporated the available information into each of the five primary components of the planning process: determining the status of the species affected, calculating the amount of take that would be caused by the proposed activities, estimating the biological impact of the proposed activities, selecting the mitigation measures to offset the anticipated take and developing a process to monitor and evaluate the success of the plan’s implementation. Each component was graded using a six-point scale, ranging from “excellent” to “extremely poor,” reflecting the overall adequacy of its use of scientific data and analysis. Throughout the evaluation, the team maintained a narrow focus on science and did not attempt to analyze the role of economic and political considerations in HCP formulation and approval.

The review team answered hundreds of specific questions on each plan, making this the most comprehensive and exhaustive review of HCPs ever performed. (The complete list of questions, participants, plans reviewed and other details can be found through a link in the Internet resources box at the end of this article.) The level of detail in the data set allowed the team to ask sophisticated questions about the patterns of scientific strengths and weaknesses in HCPs, such as the kinds of data the plans typically included, the common data gaps, and how the mix of available and missing data shaped each plan’s management strategies. Although not an explicit goal of the study, the results painted a clear picture of how data gaps generate and propagate uncertainty within the HCP-development process.

Status, Take and Impact

One of the first steps in creating an HCP is to estimate the size of the local population of any imperiled species that will be affected by the plan’s activities. Once the population is determined, plan preparers and agency officials can estimate the total amount of take expected and what impact that would have on the survival probability of the species as a whole. Unfortunately, generating an accurate estimate of the local population size of a threatened or endangered species is often not an easy task, since comparatively little is known about the basic biology of many of these species. The team found that even general trends (whether the species was improving or declining) were unknown for 49 percent of the species addressed by the reviewed HCPs. The San Bruno Mountain HCP started with a two-year study to determine the size of the Mission blue butterfly population on the site and to survey for other sensitive biological resources. But this type of rigorous, population-based survey became the exception rather than the rule as HCPs proliferated in the 1990s. Instead, plan preparers often used general habitat quantity and quality measures or took density estimates from the existing scientific literature for the species in question and multiplied them by the size of the affected area. Although this approach has some validity, density estimates based on research conducted outside the plan area may not accurately represent conditions in the local population. Additionally, the estimated population size may range across several orders of magnitude depending on which density estimates were used. For example, estimates in the scientific literature of the density of the endangered Tipton kangaroo rat (Dipodomys nitratoides nitratoides) range from 0.4 to 10 per acre. This range yielded an unwieldy prediction of 64 to 1,600 individuals in the area of the Metropolitan Bakersfield HCP. Such an estimate may be an important first step in determining population size, but plan preparers have little ability to verify the accuracy of such estimates without conducting additional site-specific surveys.

Uncertainty about the size of the local population of affected species often propagated throughout the remainder of the HCP reviewed, starting with estimates of the amount of take that would occur. Forty-seven of the 97 cases examined presented no estimate of how many individuals would be taken as a result of the proposed activities, and an additional 21 cases gave estimates that were so vague as to provide little aid to planners. This is more serious overall than it is for the local population, since the Fish and Wildlife Service often requires that HCPs assume total loss of the local population, a conservative approach that helps compensate for uncertainty. But without sound information on local population size, plan reviewers cannot evaluate how crucial the population in the plan area may be to conservation of the species. Such assessments of impact were often found to be quite weak, largely because important information about the size of the local population was unavailable. And recall that for an HCP to be approved, it must not jeopardize the continued survival and recovery of any imperiled species.

Assessment of the potential impacts was also clouded by uncertainty about the cumulative effect of multiple plans addressing the same species in different portions of its range. Such cumulative impacts are clearly relevant to the planning process but were unknown for 35 percent of the cases examined by the review team. For example, the threatened Northern spotted owl (Strix occidentalis caurina) is affected by numerous HCPs covering more than a million acres of privately owned habitat in the Pacific Northwest, as well as by the Northwest Forest Plan for federally owned forests. The review team found that several of the HCPs addressing Northern spotted owls contained little analysis of the cumulative impact that the plan would have in combination with other existing or proposed plans. Or consider the case of the endangered Alabama beach mouse (Peromyscus polionotus ammobates). This species is confined to a 13-mile strip of dune habitat on Alabama’s Fort Morgan Peninsula, and its habitat is threatened by extensive resort construction, commercial development and periodic hurricanes. Although the impact of a single plan may be small, numerous HCPs have been enacted or proposed for this tiny area, and each plan whittles away at the remaining habitat. In aggregate, these plans may pose a considerable threat to the survival and recovery of the beach mouse, yet this cumulative threat was discounted in the plans reviewed.

Minimization, Mitigation and Monitoring

To be approved, an HCP must include a strategy to minimize and mitigate the expected take. Minimization involves modifying the proposed activity to reduce the disturbance to imperiled species in the area, whereas mitigation compensates for the actual take. For example, when a development company in Brevard County, Florida, proposed to develop 3.78 acres of habitat for the threatened Florida scrub jay (Aphelocoma coerulescens coerulescens), it planned to minimize harm by avoiding construction activities and land clearing during the jay’s nesting season, and to mitigate harm by purchasing and protecting 7.5 acres of scrub jay habitat elsewhere. In general, HCPs relied on a variety of mitigation techniques, including land acquisition and habitat restoration, and most plans used more than one technique.

The effectiveness of the mitigation strategy is central to the success of the entire plan: If the impacts are not effectively offset, the plan will result in more harm than expected to imperiled species. Planners must determine the type and extent of mitigation action appropriate to offset the expected take. Uncertainty about the amount of expected take complicates the development of the mitigation strategy, and additional uncertainty arises from the difficulty in predicting how a species will respond to a given mitigation measure.

The mitigation measures in the plans reviewed appropriately identified and addressed the primary threats to 85 percent of the species examined. For example, when species were declining because of habitat destruction, the mitigation measures generally involved protecting other habitat instead of some other approach such as captive breeding. But the team also found that the specific mitigation strategies chosen often lacked sufficient data to demonstrate their efficacy. For example, the mitigation strategies for several plans addressing the endangered golden-cheeked warbler (Dendroica chrysoparia) relied on purchasing and preserving other lands, but they provided no indication that the proposed acreage would sufficiently offset the expected loss of habitat.

Without scientific data demonstrating the efficacy of a proposed mitigation measure, its use in an HCP represented little more than an educated guess about what might defray the impacts to the species. In the most troubling cases, plan preparers omitted or ignored existing information about the effectiveness of mitigation measures for certain species. In perhaps the worst example, several HCPs relied heavily on relocation of threatened Utah prairie dogs (Cynomys parvidens), despite a study available to plan preparers concluding that nearly 80 percent of 480 Utah prairie dogs disappeared or died within three months following relocation. In this case, the proposed mitigation strategy was little better than killing the animals outright. Overall, the review team concluded that 45 percent of cases failed to demonstrate that the proposed mitigation measure would offset the expected take. To their credit, most plans avoided completely unproven mitigation strategies, but the long-term effectiveness of many mitigation measures is simply unknown.

A good monitoring program can alleviate much of the uncertainty in HCPs by tracking the actual take and how well the mitigation measures are offsetting this take. Should problems arise during plan implementation, such as more take than expected or an inability to effectively implement the mitigation measures, a thorough monitoring program can provide the opportunity to recognize and correct them before excessive harm is done to imperiled species. Unfortunately, the team found that most of the HCPs reviewed did not provide a specific program for monitoring the plan’s effects on either the population of the imperiled species or overall habitat quality. Only 22 of 43 plans analyzed contained clear biological monitoring programs, and only seven were considered sufficient for actually evaluating the success of the HCP The reviewed plans generally contained at least some minimal monitoring program, but most were not described in enough detail to demonstrate that they would effectively inform future management efforts. Since two-thirds of the clearly described monitoring plans were inadequate, it is likely that many of the undescribed monitoring programs were also inadequate. The lack of an explicit monitoring program generating publicly accessible data and results hinders the ability of biologists and planners to fully understand the impacts of the HCP and to modify its terms to make it more effective. On the one hand, an effective monitoring program could compensate somewhat for uncertainties associated with the size of the local population of imperiled species, the amount of expected take and responses of species to minimization and mitigation. On the other, the lack of such monitoring can exacerbate risks by increasing the difficulty of detecting potential problems for species affected by HCPs.

Addressing Uncertainty

The review team found that uncertainty for imperiled species arises in several different ways during HCP development and implementation, chiefly as a result of incomplete data. Little is known about the basic population ecology of many imperiled species, and often the data necessary to guide management efforts were simply unavailable. Plan preparers often had to interpret incomplete data to make decisions for the specific circumstances of their HCP In fewer cases, uncertainty arose because the relevant data existed but were omitted from the plan. Often the lack of important basic information, such as the ecological requirements of the affected species, influenced not only the estimates of the local population size and take but also the selection of effective mitigation and monitoring strategies. Uncertainty may arise at any step in plan preparation, but often it then propagates throughout the entire process.

The resulting uncertainty produces risks for landowners as well as for affected species. HCPs are expensive to prepare and implement, and landowners may be hesitant to invest in developing one if they fear that the terms of their agreement will be changed if unanticipated problems eventually arise. To promote HCPs, the Clinton administration developed a variety of measures to reduce risks to landowners. One of the most controversial is a regulation known as “no surprises,” which assures landowners that they will not be held financially responsible for implementing increased mitigation efforts resulting from unintended consequences of the plan. The large geographic scope and long duration of some HCPs, combined with the regulatory certainty just mentioned, provide further assurances to landowners.

Minimizing the risks to imperiled species has apparently been a lower priority, since many of the measures that reduce uncertainty and risks for landowners actually magnify the risks to species. The no-surprises regulation, for example, begs the question of whether perpetually under-funded government agencies will actually be able to provide funding for increased mitigation efforts should they become necessary because of limitations or oversights in an HCP. Current policies also do not require landowners to fully offset the take of imperiled species or demonstrate that their efforts increase a species’ prospects for survival and recovery. In light of the NCEAS review team’s results, increased attention should be paid to how uncertainty and risk can be reduced for species addressed by HCPs. Standardized protocols should be developed to recognize where uncertainty exists and take it into account while the plan is still under development. Although important information about imperiled species is almost always limited or absent, planners can address the resulting uncertainty by collecting new data, utilizing precautionary measures and incorporating provisions to improve the plan’s effectiveness over time.

A habitat conservation plan offers an ideal opportunity to collect new data about the basic biology of imperiled species, but far too often this opportunity is missed. Calculating population sizes, densities and trends for wild species is difficult under the best of circumstances-and much more so when time, funds and staff are in short supply, as is often the case during HCP preparation. Conducting a two year population inventory similar to the one done at San Bruno Mountain may not be practical or even possible in many cases. Nevertheless, increased emphasis should be placed on collecting data on local population size, characteristics and response to mitigation efforts. The collection of such data would be a major step toward reducing the uncertainty currently associated with many HCPs. Over time, the data collected across multiple HCPs would help alleviate the paucity of information about imperiled species on private lands, thereby aiding future management efforts.

A variety of precautionary measures could minimize uncertainty, but such measures have not been utilized in many cases. For example, plans lacking good biological information about the affected species can reduce the risks to these species by beginning mitigation prior to the take whenever possible. This approach allows managers to ensure that effective and adequately funded mitigation measures are in place before irreversible harm is done to a species or its habitat. Commendably, many HCPs used such an approach, with take occurring before mitigation in just 23 percent of cases reviewed. Unfortunately, plans with irreversible impacts were actually more likely to have take occur before mitigations measures were in place-the opposite of the precautionary approach. Another way to reduce risk would be for federal agencies to reduce the duration of the permit (the amount of time developers or logging companies are guaranteed take permits for their activities) in cases where uncertainty exists because of a lack of information. For the plans reviewed, permit duration was unrelated to the level of uncertainty about impacts. A third precautionary approach would be to build more independent scientific review into the HCP development process. Teams of qualified scientists could be assembled to review topics such as data gaps for species commonly addressed by HCPs or the evidence of the efficacy of specific management and mitigation techniques. Such scientific reviews could reduce uncertainty for both landowners and imperiled species by vastly improving the information available for future plans.

Lastly, uncertainty could be further minimized by providing greater opportunities to adjust plans during implementation in response to new information or changing environmental conditions. Planners can develop ways to monitor and modify HCPs while continuing to provide financial assurances to landowners. Fifteen of the 43 plans analyzed included some mechanism for mid-course correction as additional information was obtained. Unfortunately, plans with a high level of uncertainty about the effectiveness of mitigation were actually less likely to contain a discussion of future changes in management strategies. To be effective, mid-course corrections require biological monitoring programs that are sophisticated enough to detect the effects of management and the population trends of target species. Such monitoring programs would also help alleviate data gaps about the biological status and trends of the nation’s most imperiled plants and animals, thereby promoting additional recovery efforts for those species. But to date, biological monitoring programs have not become sufficiently sophisticated or prevalent to adequately reduce uncertainty in HCPs.

The U.S. government has already taken major steps to reduce uncertainty for landowners. In June 2000, roughly 18 months after the NCEAS review team published its findings, the Fish and Wildlife Service and the National Marine Fisheries Service finalized new guidelines designed to reduce the risk and uncertainty for species affected by HCPs. If fully implemented, these guidelines will make biological monitoring and mid-course corrections more prevalent and will require that biological goals and objectives be specified in each plan. Following these guidelines is not required, but at least some HCPs will be based on better information and precautionary principles in the future.

One of the “new and improved” HCPs addresses approximately 250,000 acres of habitat for the endangered Kamer blue butterfly (Lycaedes melissa samuelis) in Wisconsin. The plan has a stated biological goal of no net loss of butterfly habitat and includes an adaptive management component designed to arrest any declines in Kamer blue butterfly populations. The plan’s biological monitoring program combines monitoring activities by local partners on their lands with a broader, statewide effort to track Kamer blue butterfly populations and their response to management under the HCP. In addition, the planner incorporated contingency strategies into the plan to deal with changing conditions such as gypsy moth infestations and forest fires. With the new federal guidelines and growing experience in conservation planning, improved HCPs such as this one will likely become more common.

In an ideal world, a scientifically based habitat conservation plan process would include conservation strategies based on knowledge of the basic population biology of the affected species, their ecological requirements and a quantitative estimate of the impact the taking would have on the viability of the species. Mitigation measures would have good track records in offsetting impacts to species, and rigorous monitoring would follow the plan’s actual effect on species and their habitat. Ongoing research and monitoring programs would guide adjustments to minimization and mitigation activities over time. This scenario is impossible to achieve in the real world. Several important steps outlined here, however, could be taken when limited information generates uncertainty in HCPs. Without better assurances of the success of such plans, the long-term result could be the tragic destruction of the very species these plans were designed to protect.