In a recent editorial, Whitten et al. (2001) lament the loss of Sumatran lowland rainforests and encourage conservation biologists to reexamine their efforts to save nature. The Sumatran rain forests are one of a number of critically endangered terrestrial ecoregions of outstanding biological value. Other examples are the Philippines moist forests and the dry forests of New Caledonia—areas characterized by extraordinary levels of endemism—where we need to fight for every scrap left and begin restoration immediately. In these places, recommending new biological surveys or more refined reserve-selection algorithms is akin to fiddling while Rome burns. For many other places on Earth, however, where substantial natural habitat still remains and the best course of action for protecting biodiversity is less obvious, conservation biology has a critical role in identifying what needs to be accomplished and in what order of priority. Well-designed conservation landscapes—with representative systems of conservation areas of sufficient size, condition, and connectivity to maintain even the most sensitive species and ecological processes—are an essential foundation of any conservation strategy, whether they be for the vast forests of New Guinea or the vanishing fragments of lowland Sumatra. The kind of information required to design conservation landscapes is precisely what empowers us to engage effectively in “on-the-ground management and policy decisions,” as exhorted by Whitten et al.
Critical Areas for Research
We believe that the problem lies as much in the nature of the questions being asked by conservation biologists as in our lack of engagement in the nonbiological aspects of conservation. Pimm and Lawton (1998) urged us to focus more intensively on “the most pressing problems involving many species and their fate across decades to centuries, over large geographical areas.” Unfortunately, most research questions remain so narrow in scope or of such limited scale that their results are unlikely to contribute to stemming widespread loss of habitat, species, and phenomena. We urge conservation biologists to allocate greater effort to four critical areas: (1) conserving those species and ecological processes that require the greatest area to persist (minimum-area requirements), (2) conserving widespread species and continental-scale phenomena, (3) quantifying patterns of beta diversity and endemism, and (4) predicting the location and intensity of threats to biodiversity.
Area-sensitive species with large spatial requirements are increasingly being adopted as umbrellas for establishing size limits for reserves protecting other biodiversity features. Yet we know little about the amount and quality of habitat that these area-sensitive species require to persist. For example, we can accurately predict that an isolated, 20-km2 fragment of rainforest will quickly lose its jaguar (Panthera onca) population, but we have only crude estimates of the area needed to sustain jaguars or any other large predator over decades. How do such estimates vary among regions containing a diversity of habitat types or prey densities? How many breeding individuals are necessary for the survival of a population over the next century; is it as low as 50 or higher than 500? Research on range size, movements, and demography will inform the design of effective, interconnected reserve networks, which will likely be our only option for saving area-sensitive species in the wild.
Large carnivores are not the only species whose persistence requires intact natural landscapes. Many larger herbivores, primates, and birds are area-sensitive, as are species that track patchy or low-density resources such as macaws, Purple-throated Fruit Crows ( Querula purpurata), hornbills, mandrills (Mandrillus sphinx), and white-lipped peccaries (Tayassu pecari). Most have received scant attention. Minimum-area requirements for tropical plants are also grossly understudied. For example, how large should tropical forest reserves be to conserve populations of trees that occur at low densities? Guidelines for area-sensitive freshwater and coral-reef species are also woefully lacking. We need a concerted effort to gather empirical data and develop predictive models that can be applied to a wide array of species, guilds, and habitats.
Ignorance about the spatial requirements of ecological processes may be an even greater impediment to conservation than the paucity of information on species. Maintaining ecological processes that operate over larger scales is a primary reason for designing networks of natural habitats and protected areas. For example, we are working with biologists in Asia to design a network of core protected areas, corridors, and buffer zones to maintain metapopulations of tigers (Panthera tigris), Asian elephants (Elephas maximus), and greater one-horned rhinoceros (Rhinoceros unicornis) across northern India and southern Nepal. Where is the sorely needed textbook on corridor biology and dispersal? In Chile we are collaborating with local conservation biologists to propose a network of reserves in the temperate rain forests of Valdivia. How large should reserves be to ensure that extremely rare (500-year) natural fire events still allow fire-dependent reproduction by long-lived trees such as Alerce fitzroya? Process-related questions are even more pressing for freshwater and marine conservation, where hydrologic processes and the movement of materials shape ecosystems.
Large-Scale Processes and Widespread Species
Restricted-range species and rare habitats are clear conservation targets, but species and processes occurring over vast areas require our attention as well. Some widely distributed species may actually be highly sensitive to changes in natural features across landscapes. For example, biologists have warned of the impending ecological extinction of formerly widespread and abundant species such as migratory songbirds. Will certain bird species meet the fate of the Passenger Pigeon (Ectopistes migratorius), plummeting to extinction as thresholds of forest loss, fragmentation, isolation, or population size are exceeded? Are specific habitats for resting and feeding during migration equally important? Research into such questions will promote conservation of key landscape features and habitats.
We are also witnessing major changes in large-scale ecological and physical processes, such as rainfall patterns, as human activities degrade ecoregions. At what point will rainfall diminish drastically in the Amazon as forests are cleared and fragmented? The answer could provide us with one of the most compelling arguments for keeping the richest part of Amazonia intact.
Large-scale processes in marine systems are critical research targets. For example, the design of an effective network of marine protected areas for the Mesoamerican Reef in the western Caribbean Sea requires analyses of both large- and small-scale disturbances. How dispersed does a network of marine reserves need to be to conserve some intact reefs in the face of inevitable 50-km-wide hurricanes over the next half century? Can reserve networks be designed to allow some reefs to escape sediment-laden, 100-year storm events, the effects of which are exacerbated by anthropogenic deforestation of watersheds? How large and far apart should smaller “source-pool” reserves be to buffer fish and invertebrate populations from intense fishing during larval and adult dispersal? Transient and ephemeral features—such as upwellings or the highly productive edges of sea ice that shift regularly in polar seas—need greater attention. How does one design a system of marine reserves that shifts over space and time to provide sufficient refuge for populations of species that track patchy and moving resources?
Beta Diversity and Endemism
The conservation of large, intact habitat blocks, intact biotas, and patterns of beta diversity and local endemism should guide the design of reserve networks. Minimum-area requirements, mentioned above, determine the effective size and connectivity of conservation areas, whereas patterns of local endemism and turnover should determine the placement and number of reserves in a representative network. Thus, we require a better understanding of how species and communities are distributed in many parts of the world to effectively save all the pieces.
For several large, biologically rich regions, such as the Amazon, Orinoco, and Congo basins, lack of data on the distribution and turnover rates of plant and animal assemblages impedes reserve design. Currently, we cannot determine if reserves need to be 50, 100, or 500 km apart for effective representation of plant and invertebrate assemblages in the tropics, or even for temperate rainforests. Only a handful of studies have begun to look at these issues. To address these gaps, biologists need to determine the relative influence of biophysical features on local endemism and betadiversity. With this information, we can use predictive models to identify and map distinct assemblages or at least begin to understand the frequency with which protected areas are required to address species turnover. We must also identify indicator taxa and efficient, repeatable sampling methods for field-testing these models. Morphospecies would suffice in many instances.
The data gap regarding beta diversity, local endemism, and the distribution of distinct communities is not limited to rainforests. We lack similar data for ecosystems as diverse as deserts and large rivers. The latter harbor some of the most diverse and endangered vertebrate faunas anywhere. Yet geographic patterns of freshwater biodiversity remain poorly known, especially for the Amazon, Orinoco, Congo, Mekong, and upper Yangtze River basins. Few data exist for geographic patterns of richness, endemism, and beta diversity for invertebrate and fish faunas of coral reefs.
How can we more accurately predict the location and intensity of threats to biodiversity? Several attempts to model the effects of deforestation, road building, and human settlement on faunal communities provide new insights into threatened forests of the Amazon and Congo basins. Bushmeat hunting is so pervasive in most tropical forests that larger forest vertebrates may rapidly disappear from relatively intact habitats. To propose effective reserve networks, biologists must map where hunting pressure is greatest and construct models to predict where it will be a future threat. Analyses of hunting and access patterns, commercial and local markets, and the sensitivity of different wildlife species are needed to construct useful models.
Exotic species pose another threat that we are ill equipped to address. In the Galapagos archipelago, how long will native biota persist if invasive alien species continue to spread? How are native species affected differentially by patterns of human settlement and movement? The effects of introduced vertebrates are well documented, but what about invasive invertebrates, such as the ant Wasmannia auropunctata, in the Galapagos and in other tropical islands such as New Caledonia and Hawaii? Will this alien ant wipe out only native invertebrates or will it also affect vertebrate populations? Can patterns of introduction and spread be predicted and mitigated? The fate of many island biotas will be determined partly by the speed with which we can provide answers to these questions.
The Ecoregion Approach
The major conservation organizations have embraced ecoregion- and landscape-scale conservation as a powerful framework to address such questions. Answers to these questions will lead to robust biodiversity visions—a science-based blueprint of what success should look like 50 years from now. These far-reaching strategies identify biodiversity goals and targets. They propose representative systems of conservation areas of sufficient size, condition, and connectivity to maintain even the most sensitive species and ecological processes. All of these features strengthen the conservation community's credibility in predicting the consequences of different resource-use scenarios and its bargaining position during negotiations with stakeholders for whom conservation is a lower priority. Without the recommendations of a science-based, ecoregion strategy to serve as a bottom line, conservationists are likely to enter into dangerous compromises because we will not know when, where, and what to fight for.
Conservation biologists typically are the last stakeholders invited to comment on development strategies that will permanently alter the fabric of the remaining natural world—if they are even consulted at all. In the case of the Sumatra lowland forests, it is unclear if intervention by conservation biologists would have turned the tide, given that nearly half of all the logging that goes on is illegal. But in many other ecoregions, rapid answers to the questions outlined above and the design of conservation landscapes that work for both biodiversity and people may help us avoid future Sumatras.