Guest Opinion: Science and the Art of the Solvable In Hawai`i's Extinction Crisis

“Insanity is doing the same thing over and over and expecting a different result.”

On November 29, 2004, a male po`ouli died of what was effectively old age while in captivity in Hawai`i. The only two other known birds of this species, both at least five years old, have not been seen in more than a year and the species is now presumed extinct. With the passing of the po`ouli, it is appropriate to ask what Hawai`i’s latest avian loss tells us about the state of terrestrial conservation and conservation science in the archipelago. Does the po`ouli’s demise reflect the ineffectiveness of conservation efforts in Hawai`i or is it merely an unlucky fluke?

Of the 107 endemic bird species present before the arrival of humans around 400 A.D., only 10 percent are either not extinct or not threatened with extinction. Almost 300 plant species, or about 30 percent of the endemic native flora, are threatened or endangered, and a similar number are in decline. For invertebrates, we have no idea how many species we have already lost, although the number is surely large. Finally, 20 percent of the total species in Hawai`i are now non-native.

The forces that have driven this ecological holocaust include massive, largely historical, habitat conversion by humans in the lowlands; ongoing habitat destruction by insects and alien wild mammals; wholesale replacement of native vegetation by alien plants; changes in fire frequency; and extensive mortality inflicted on native species by alien predators and diseases.

The Hawaiian Conservation Alliance stated that 737 staff and $63 million are devoted to conservation in Hawai`i each year. Based on this impressive figure, one might think that biodiversity conservation in Hawai`i is a mature and effective operation. To the contrary, we argue here that the demise of the po`ouli accurately reflects the status of conservation practice and science in Hawai`i, which suffer from a suite of problems that routinely compromise conservation effectiveness.

In what follows, we won’t discuss the failures of conservation practice (a topic of considerable scope) but instead will focus on the role that less-than-optimal conservation science plays in limiting conservation effectiveness. We describe the gaps in knowledge that exist for conservation science, discuss some of the institutional problems and poor practices that contribute to those gaps, and propose changes in practice in the hope that far more of Hawai`i’s biological wealth can be preserved in the future than would be the case if the present trajectory continues. Failing that, Hawai`i may become an ecological Potemkin Village of lush, alien vegetation, with a few remnant native birds and plant communities clinging to life in the hills, as much of the island of O`ahu arguably is already.

Providing effective conservation in Hawai`i requires that conservation science attend to three broad topics: invasive species, endangered species, and habitat restoration. The first is the overwhelming cause of biotic impoverishment in these islands. The second is its most imperative set of symptoms. The last is the primary means by which the problem may be rectified.

Invasive species
The process of invasion by alien species involves three stages: transport, establishment, and spread. Each stage poses a different set of research questions appropriate to informing meaningful management decisions.

For suspected or known invasive species that haven’t arrived yet but are likely to be introduced, we need to know the most probable means of introduction and how best to intercept or exclude them. There is a growing body of work on how to identify potential invasives, but little of it has been Hawai`i-focused. Information about modes of invasion in Hawai`i are available for vertebrates, plants, some disease vectors, and a few parasites, but not for most taxa. We have few scientific studies for Hawai`i on the best ways of preventing introductions, but, for certain taxa, we can borrow profitably from New Zealand and Australia.

More than 5,000 alien species are established in Hawai`i, but limited information exists as to how many are invasive. An early study provided a preliminary attempt for vascular plants, although that estimate is now quite dated. An estimate of invasive potential among the 8,000 already cultivated plants is available, but no similar compendium has been attempted for animals. Potentially, any could become invasive, even after a long lag time that can span centuries. Documented ecological impacts have largely been restricted to a few vertebrates, plants, and some insects. Taxa of major ecological importance in structuring continental ecosystems (for example, earthworms, mycorrhizae) have largely gone uninvestigated, and even studies on hymenopterans, a group of major ecological significance, have been restricted to a handful of taxa.

At present, managers rely on sporadic observations and personal impressions to identify new species that appear invasive, but such impressions are difficult to convey to others in a compelling manner or to use in ranking response activities. Decision-support systems to establish priorities for action among the large number incipient invasives are lacking. We are aware of only a single study that attempts a systematic survey (of plants on Maui) to identify potentially invasive species before they are too widespread to stop. This study, laudably, has resulted in immediate eradication of four high-risk species, but it needs to be systematically duplicated and periodically updated for other taxa and other islands.

Simple, tactical population models for invasive species have rarely been used in Hawai`i but could assist in setting eradication priorities. We need, but rarely have, at least crude estimates of age at first breeding/fruiting, intrinsic rates of increase, seed persistence in the soil (for plants), present range, and propagule dispersal distance. Such models need not be exact to help estimate the feasibility and cost of containing or eradicating an invasive species. Data-based planning and budgeting allow for effective attack on recently established species before they become firmly entrenched and allow realistic allocation of effort to contain entrenched species; otherwise, control actions are apt to merely chase after such species as they expand their ranges. For example, current legislative attempts to divert $2 million to $3 million from rapid-response programs against incipient invasive species to coqui control on Hawai`i island seem senseless in light of conservative estimates that more than $25 million a year would be needed to treat the 5,000-plus infested acres in the Hilo-Puna districts alone.

For long-term control of established alien species that are damaging native ecosystems, only two options exist: biocontrol across landscapes or mechanical/chemical control in limited sensitive areas, such as implemented by the National Park Service in its Special Ecological Areas. In Hawai`i, agricultural interests have been able to set priorities for biocontrol, but these have eluded natural resource managers and researchers. Biocontrol, exclosures, or direct control are being pursued against several widespread invasive species, such as strawberry guava, wild pigs, and apple snails, but it is unclear what criteria managers are using to target such species or to judge success. Formal triage models to prioritize taxa for control within reserves have been developed and are widely used in New Zealand, but are not yet applied in Hawai`i. Much could probably be gained by their general application in land management within Hawai`i but this would, again, be predicated on having basic natural-history information for the proposed target species.

In Hawai`i, efforts to identify novel mechanical and chemical control methods are virtually nonexistent, although fencing against ungulates and management after fire have met with success. As a result, rapid response to eradicate incipient pest species is routinely hindered by lack of knowledge on how to efficiently kill them. The need for simple herbicide trials on scores of incipient invasive plant species has been recognized for years but has gone unaddressed, in part due to lack of funding. Absence of reliable control mechanisms has also been partly responsible for allowing feral parrot and frog populations to explode on Maui and Hawai`i islands. Useful detection methods are often lacking (making it difficult, for example, to evaluate or respond effectively to reports of mongoose on Kaua`i) and need to be developed for many invasive taxa, animals and plants alike.

Studies to determine the costs that terrestrial invasive species impose on Hawai`i’s economy and quality of life would help establish political support for making the difficult decisions necessary for effectively addressing this problem, but are only now being pursued in earnest. Elsewhere, such studies have documented the tremendous economic impacts of these species.

Endangered Species
For those Hawaiian species officially listed as endangered – a few invertebrates, many plant species, and most land and wetland birds – a series of science and management problems hinders or precludes their recovery. We will focus here on birds, as they have received most of the attention and funding and, thus, presumably represent the ‘state of the art’ for endangered-species conservation in Hawai`i. For plants and invertebrates, the situation is much worse. For most of the hundreds of endangered plants, pollination biology, seed dispersal, and genetic structure of populations are unknown, but knowledge of all three is essential to directing successful recovery efforts. For large numbers of invertebrate taxa, such as land snails and many insect orders, and micro-organisms, we do not even know which species still exist, their numbers, or their geographic ranges. Hence, it is currently impossible to decide even which taxa require intervention efforts, much less how to intervene, despite the fact that this group has likely already experienced the greatest number of extinctions and currently has the largest number of biologically endangered species.

According to one estimate, $37.8 million was spent between 1987 and 1997 on research on Hawaiian birds. The author of that estimate, William Steiner, determined that only one species, nene, emerged as a success story and he provided a long list of basic natural-history information that remains ungathered. He also estimated that at least an additional $57.7 million has been spent on management and habitat acquisition. A recent draft plan calls for another $3.6 billion to be spent on forest bird research and conservation in the next 50 years. Despite large research expenditures, some of the most basic science necessary for understanding the target species has never been done. For example, Population Viability Analysis (determining a minimum population of the species that has some acceptably low chance of going extinct over some time period) has been done for just seven birds: Laysan teal, `alala, akohekohe, palila, `elepaio, Hawaiian stilt, and Laysan finch. This despite the fact that in 1992 the Hawaiian Forest Birds Conservation Assessment and Management Plan called for an additional 21 such analyses for species or for island populations, the draft recovery plan for Hawaiian forest birds 11 years later ranked population viability analysis as “Priority 2,” and population viability is one of the criteria for downlisting of Hawaiian forest birds.

Many factors threaten Hawaiian birds, including habitat loss, disease, mammalian predators, food loss, and competition from introduced birds, but we do not know if all are equally powerful, if their importance varies among bird taxa, or even if all are currently at work. Managers would benefit from knowing against which threats they should focus their efforts. We also don’t know the carrying capacity of the habitats in which most endangered species occur (the palila being an exception), or even whether currently occupied habitats are optimal. Conservation efforts directed at suboptimal habitats may ultimately be futile, but we cannot currently assess the degree to which this is a problem. If we did know the most serious threats and the optimal habitats, we could then test methods for managers to use for converting suboptimal habitats to optimal ones.

Habitat Protection and Restoration
The first step in restoration is to ask to what period the habitat or ecosystem should be restored. We have some general knowledge of pre-contact and pre-Polynesian conditions, but, as with the recovery of native species, we are hindered by the lack of details. It would be useful to have clear data on the location, amount, and quality of native Hawaiian habitats for use in establishing priorities for habitat stabilization and recovery efforts. We know a little of the present and past distribution of ecosystems, but this knowledge is not as well-developed as it is in other states, despite our small geographic size. Hawai`i is now involved in the national GAP program, which may eventually make such data available.

We know little of past ecosystem processes. Hawaiian ecosystems originally lacked earthworms and ants, two dominant forces in continental terrestrial ecosystems and now ubiquitous across Hawai`i. Their introduction may have forced massive shifts in ecosystem structure or functioning, but the ramifications have not been investigated. Many pollinating and seed-dispersing species are believed to have gone extinct and have likely affected plant populations. Following the introduction of parasitic wasps, moths ceased to be mass-defoliating agents in ecosystems, probably altering the cycle of forest disturbance and nutrient cycling. Snails used to occur at great densities, potentially providing food for birds and significant grazing pressure on fungi, but are now ecologically extinct. Aboriginal hunting, land clearing, and introduction of mammalian predators may have led to massive decreases in seabirds nesting in vegetation, decreasing importation of marine nutrients to forests. Finally, many plant species disappeared or became very rare following the introduction of ungulate herbivores. Indeed, many forests in Hawai`i exhibit wholesale recruitment failure of native vegetation, and it is not clear which of the many possible alien agents is most at fault. Each of these changes is of potentially profound importance for understanding current Hawaiian ecosystems and how to more closely approximate original ones, but each has eluded rigorous study.

Abiotic perturbations such as rain, floods, storms, El Niño events, hurricanes, fire, erosion, and volcanism may have caused ecosystems to oscillate between different stages of secondary succession, creating a natural mosaic of habitats in pre-human Hawai`i. We know little to nothing about such ‘rhythms of the land,’ nor have we much knowledge of current patterns of perturbation, but such forces, combined with habitat loss following the arrival of humans, probably play a critical role in the fate of many rare species. If conserved areas, networks of areas, or secondary succession and recovery of habitat are smaller or slower than the scale of major perturbations, suitable habitats will not persist over time, resulting in the loss of their constituent species. The smaller, fewer, or more widely dispersed the conservation areas, the more active must management be to maintain islands of suitable habitat within their borders, but we have no idea of the minimum areas needed for any of our natural habitats, ecosystems, or even species. Of our existing natural areas, perhaps only upland East Maui and upland Mauna Loa are big enough to survive intact over the next 500 to 1,000 years without intensive human intervention. Even the Alaka`i Swamp area on Kaua`i, the last refuge of several endemic bird species, proved too small or ecologically simple to shelter them from two hurricanes. No analysis exists of how long Hawai`i’s Natural Area Reserve System is likely to preserve its stunning ark of Hawai`i’s biodiversity, nor what would be necessary to slow or stop loss in individual units. Finally, climate change needs to be incorporated into such considerations, as it is likely to significantly alter native and alien species distributions and to seal the fate of a number of endangered species.

Most restoration, as well as management of established natural areas, has been tactical rather than strategic, in the sense of lacking an overarching recovery goal. With a few nascent exceptions, such as watershed partnerships, we have no consensus ‘best practices’ beyond fencing, removing ungulates, and outplanting native plants, with some efforts to change fire regime. Because of the data gaps just listed, few if any plans have been based on consideration of long-term goals or how the immediate efforts will achieve them, despite an emerging body of knowledge elsewhere on how to do just this.

Hawaiian conservation biology has to confront the formidable research agenda outlined above if it is to give managers of land and natural resources the information they need to guide their efforts, but several major institutional challenges loom.

Academic ecological science has two components: basic natural history and hypothesis testing of models or theories. Unfortunately, the two are often used in inappropriate contexts in Hawai`i. Collection of natural history data is commonplace in Hawai`i but largely restricted to a handful of attractive and larger taxa that, for the most part, don’t need that work. Its application to invertebrates, where it is most needed, occurs far more infrequently. This is in part because describing the distribution, abundance, and attributes of species, communities and ecosystems is widely viewed as ‘unsexy’ (and sometimes as ‘unscientific’) in the broader scientific community, making it difficult to find support and attract interest among talented researchers. Nonetheless, such work is absolutely imperative for baseline ecological knowledge for conservation. Its absence is a critical bottleneck for conservation of the mass of Hawai`i’s endemic diversity. Alternatively, for well-studied taxa in Hawai`i, too often science never advances beyond the merely descriptive, failing to test alternative hypotheses or to explore models that might explain species rarity, invasion, or how ecosystems work or collapse. For example, despite years of research, we don’t know the relative current contributions of different factors to the endangerment of Hawai`i’s native forest birds or to wholesale replacement of native forests by alien vegetation, and thus we can’t suggest priorities to managers. More generally, the effectiveness of much conservation science is compromised because broader research agendas are lacking and projects are rarely coordinated to build logically toward on-the-ground solutions.

Conservation science in Hawai`i generally doesn’t use ongoing management operations as opportunities for research. Hence, it frequently fails to provide answers in a timeframe useful to managers. Managers are too often left to assess their own effectiveness without a scientific or even quantitative basis. There is truth to the aphorism, “You can’t manage what you can’t measure.” Science should be able to measure success of management and provide feedback that can be used to fine-tune management. Such adaptive management of natural resources has been used successfully for species restoration in New Zealand and is routinely used in forestry and fighting wildfires on the mainland. Two rare and welcome exceptions to this rule are fire-restoration and weed-control management approaches developed at Hawai`i Volcanoes National Park.

Hawai`i has been singularly unsuccessful at organizing any focused multi-institutional research sites since the days of the International Biological Program (1970-1975). It has no terrestrial field stations generally open to field researchers, and it has few field sites with long runs of environmental or biological data, although the Forest Service is in the process of establishing an experimental forest on the island of Hawai`i. It has no site in the National Science Foundation’s Long-Term Ecological Research (LTER) system and, unless the research community can coalesce around a shared research vision, appears destined to play only a satellite role in the new NSF National Ecological Observatory Network (NEON). We believe the establishment of such a dedicated resource in Hawai`i could facilitate many of the ecological studies needed for better understanding how Hawai`i’s ecosystems are structured and the actions that likely are needed to rehabilitate them. Barro Colorado Island in Panama and the Charles Darwin Station in the Galapagos are both examples of ecological research facilities that have had major positive effects on research and management in their respective ecosystems by achieving a critical mass that allowed detailed understanding of ecosystems impossible to achieve with solitary or short-term projects. Both have generated substantial bodies of sustained basic and applied research that have informed and supported management. Finally, both have helped support their local economies, generating support for conservation, science, and the environments they deal with.

Failure to publish is a persistent limitation of Hawaiian conservation biology. The Hawai`i Ecosystems at Risk website (www.hear.org) and the Technical Reports of the Pacific Cooperative Studies Unit (www.botany.hawaii.edu/faculty/duffy/techrep.htm) have helped make agency ‘gray’ literature available, but too many studies never make it to the refereed literature, casting doubt on the competence of Hawaiian science and management and leading managers and scientists needlessly to repeat suboptimal methodologies. For example, while there have been numerous attempts to control rats in Hawai`i, few studies have been published that provide the details that are needed to allow rigorous evaluation of optimal methodology. Similarly, we lack published details of stock, planting techniques, subsequent care, and eventual survival for almost every outplanting effort in the state. Word of mouth is not the most effective means of transmitting best management practices. Moreover, it is singularly ineffective at convincing funding sources that further efforts deserve support.

Even when relevant scientific knowledge is produced, it may well be ignored by agency officials in reaching management decisions. Reasons for this include political and funding limitations, the cultural divide that often separates managers and academics, and a tradition of consensus-based decisions. As a result, decisions as to which species or issues should receive funding are frequently based on inchoate emotional or political criteria instead of a scientifically or economically driven triage system. Consequently, much recovery funding is directed at species with the least likelihood of recovery, such as forest birds with low population numbers, and which are subject to threats not readily ameliorated. This approach to conservation stands in stark contrast to the model taken for granted in public health, in which one rightly tries to save the most people, not the sickest. Management in Hawai`i merits its own review, which is beyond the scope of this paper.

“If politics is the art of the possible, research is surely the art of the soluble. Both are immensely practical-minded affairs.”

It is not our intention to propose a research agenda (though a partial one is implicit in what we have already said), but instead we want to suggest six broad areas where change could help create a research and management culture that could go far to rectify many shortcomings and would more effectively protect the native biodiversity remaining in Hawai`i.

Focus on the research agenda
We need to stop open-ended applied research financed by conservation and natural-resource agencies. Instead, we need tightly focused research agendas having benefits (answers) clearly identified with costs. That doesn’t mean that all research will be guaranteed to yield answers, but it should be aimed at specific, solvable problems. Instead of broadcasting general requests for proposals and supporting the most relevant of them, identifying research priorities in advance and soliciting investigators to do the identified projects might better guarantee relevance of research to managers. Identification of research priorities could perhaps use an agenda-setting process to be conducted by both scientists and managers at regular (say, three-year) intervals. This process should include personnel from outside Hawai`i in an effort to broaden perspective.

Teach ‘The Art of the Soluble’
Conservation science also needs to be applied in an explicit ‘art of the soluble’ paradigm. Graduate students having an interest in this area need formal courses that train them in the skills required once they enter conservation management professions. These include courses in planning, dealing with human conflict, and dealing with uncertainty. Getting faculty and graduate students involved in hands-on recovery or restoration programs would help keep research grounded in practical concerns. In Hawai`i, U.S. Forest Service researchers have provided good examples of the benefits of this approach for both basic research and practical conservation. State and federal agencies need to invest in institution-building, supporting such students and their graduate programs as these will be their future employees. Also needed is a commitment by the administration of the state university system to provide the recognition and credit to faculty who conduct research directly applicable to resolving conservation problems. A fundamental expansion of institutional values is needed to ensure that studies of the Hawaiian environment are valued as much as those of Mars.

Push Publications
More needs to be done to ensure that when science is completed, it gets used by managers. The first step toward achieving this is ensuring that results are published and disseminated to managers. To encourage the former, funding agencies should evaluate the publication records of those applying for funds and factor that into their granting decisions. To achieve the latter, the Hawaiian Conservation Alliance could profitably provide single-paragraph summaries of each of the previous year’s relevant research papers to people who attend HCA’s annual conservation conference, which is the focal point for interaction of the two communities in Hawai`i.

Incorporate Science in Management
State and federal land-management programs should institute regular peer-reviewed processes by which joint teams of scientists and managers ensure that the best available scientific knowledge is being applied in the field. Peer-review is widely accepted as denoting that minimum standards of scientific credibility have been met; similar standards should be in place for ensuring that land managers are using best available information and practices.

Establish Field Stations
We need one or two larger stations covering wet, mesic, and dry forest ecosystems, open to all qualified researchers whose projects meet clear criteria of not interfering with others or damaging the environment. The stations need to be large enough to have a critical mass of researchers present, but only require simple lab facilities, communal eating facilities, dorms for visiting classes, and private quarters for longer stays. In addition, managers of natural areas might consider establishing simpler field stations to attract researchers, remembering that one such humble facility, now known as Aldo Leopold’s “The Shack,” was the birthplace of restoration ecology and much of current ecological thought.

Target Funds
We need strategically targeted funding to close the knowledge gaps. Increased state and federal funds for natural-resource and conservation biology are going to be difficult to obtain, however, unless the links between costs and benefits can be more clearly demonstrated than they are in the present open-ended campaigns against invasive species and on behalf of endangered species. It may sound cynical to say that so many dollars will “buy” the removal of one species from the Endangered Species list, but such an approach may be more likely to receive funding than current efforts. Funds for the environment can also help advertise Hawai`i, much as the publicity resulting from scientific research in the Galapagos Islands helped make them the premier ecotourism destination in the world. Such marketing has yet to be tried in Hawai`i.

We suggest that dedicated sources of new funding are required to provide proof-of-concept for problem-based research and to support investigation into crucial but locally unfashionable topics such as invertebrate inventories and ecosystem “services.” Such funds are not likely to come from traditional state and federal sources because existing programs merely cater to pre-existing research values. Well-focused support from private institutions could help jump-start eventual increased governmental funding in this area, as the conservation pay-off becomes clear. A longer-term view would be the development of a new National Science Foundation program with a mandate to foster rigorous hypothesis-driven research to basic and applied conservation problems, thereby providing a rigor and independence not always apparent in agency-funded science.

Conservation in Hawai`i is sometimes seen as a lost battle. This pessimistic view stems from recognition of the hundreds of native species already lost, the thousands of invasive species introduced, and the fact that the tide of aliens is not being stemmed. Such pessimism ignores both the thousands of native species still at risk and the fact that some of the major ecological perturbations driving these losses (for example, ungulates, fire, habitat alteration) are amenable to human control. We suggest that many of Hawai`i’s conservation problems are solvable and that their future social costs may be avoided by sustained, informed action taken now. But conservation management and its supporting science must change how they do business, becoming more accountable to each other and for the resources they request. They must focus on problem-solving and explicit planning that adequately reflects biological and fiscal realities, rather than impractical, unfocused, and unachievable wish lists.

Conservation biology must provide both relevant basic biology and hypothesis-driven problem-solving on which to base management decisions, as well as focusing and measuring the outcome of management efforts. Managers in turn must be open to having their work assessed objectively and must be willing to respond accordingly. An explicit problem-solving approach provides a method to assess and justify the total costs of species recovery or ecosystem restoration. This in turn will make it easier for the public and their representatives to support conservation of species, as they will have an idea of their return on investment.

Too many recovery, restoration, and invasive-species-control efforts in Hawai`i either are failing at present or are unsustainable, but failure is not inevitable. If the definition of insanity is repeating the same behavior while expecting a different result, it is time to end the insanity and cultivate the ‘art of the soluble.’

By David Cameron Duffy and Fred Kraus*

*David Duffy (dduffy@hawaii.edu) directs the Pacific Cooperative Studies Unit and is professor of botany at the University of Hawai`i-Manoa. Fred Kraus is a herpetologist with the Bishop Museum in Honolulu.

Volume 16, Number 11 May 2006