1Department of Biology, University of New Mexico= ; 2Santa Fe Institute; 3School of the Environment, Levine Science Research Cent= er, Duke University
ABSTRACT
We develop methods for quantifying habitat connectivity at multiple scale= s and assigning conservation priority to habitat patches based on their c= ontribution to connectivity. By representing the habitat mosaic as a math= ematical "graph," we show that percolation theory can be used to quantify connectivity at multiple scales from empirical landscape= data. Our results indicate that connectivity of landscapes is highly sca= le dependent, exhibiting a marked transition at a characteristic distance= and varying significantly for organisms with different dispersal behavio= r. More importantly, we show that the sensitivity and importance of lands= cape pattern is also scale dependent, peaking at scales associated with t= he percolation transition. In addition, the sensitivity analysis allows u= s to identify critical "stepping stone" patches that, when removed from t= he landscape, cause large changes in connectivity.
KEY WORDS:
connectivity,
conservation in fragmented landscapes,
dispersal,
fragmentation,
habitat connectivity vs. dispersal distance,
landscape,
landscape graphs,
metapopulation,
percolation,
quantifying habitat connectivity at multiple scales,
"stepping stone" patch,
Strix occidentalis lucida.
INTRODUCTION
= Connectivity of landscapes depends not only on the spatial distribution of habitats across a landscape, but also on the scale at which organisms interact with landscape pattern (Merriam 1984, Gardner et al. 1989, Noss 1991). In fragmented landscapes, where islands of high-quality habitat are punctuated by stretches of poor habitats, species that disperse over long distances will perceive a given habitat distribution as more connected than will a species with short-range dispersal. Thus, landscape pattern can act as a scale-dependent "filter" acting differentially on the movement of species with different degrees of vagility, in much the same way as high- and low-pass filters remove high and low frequency components in a communication channel (Shannon and Weaver 1949). In relatively continuous landscape mosaics, where habitat quality varies smoothly, it is likely that the filtering effect of landscape pattern will also be smooth, reducing the rate of movement of some species while enhancing that of others (Johnson et al. 1992a, b). In highly fragmented landscapes, the filter response might be much more abrupt: species that perceive the habitat distribution below a critical scale could be effectively isolated on individual habitat patches. Therefore, particularly in fragmented landscapes, identification of critical scales associated with abrupt changes in landscape connectivity is an important consideration in quantifying habitat pattern and the influence of habitat pattern on the movement of organisms.Landscape connectivity does not depend on scale alone; the configuration = or spatial arrangement of habitats in a landscape is also an important determinant of connectivity = (Forman and Baudry 1984, Gardner et al. 1992, Henein and Merriam 1990, T= aylor et al. 1993). Heterogeneity in habitat quality, differences in patc= h shape and size, and variation in isolation among patches lead to spatia= l variation in landscape connectivity (Gustafson and Gardner 1996). Thus,= although indices of landscape connectivity averaged over entire landscap= es are useful for comparisons among ecosystems, these indices provide lit= tle insight into the local connectivity structure of landscapes. To fully= characterize the spatial structure of landscapes, it is necessary to con= sider connectivity measures associated with local regions or individual h= abitat patches. Just as overall measures of landscape connectivity are sc= ale dependent, local measures of connectivity are likely to change with s= cale.
Here, we present a multiscale analysis of landscape connectivity, based on an extension of uniform percolation theory (Stauffer and Aharony 1985, Gardner et al. 1987, Gould and Tobochnik 1988, Creswick et al. 1992) to non-uniform landscape graphs (Cantwell and Forman 1993, Harary 1969). We develop both aggregate measures of landscape connectivity and patch-based measures of individual patch contributions to overall connectivity. We examine the sensitivity of landscape connectivity to changes in landscape configuration, and the relationship between sensitivity and scale. Finally, we consider the implications of our analysis for habitat preservation efforts with regard to threatened species.
METHODS
Landscape data
Our study site consisted of a 1.5 x 106 km2 region of the southwestern United States, including the states of Arizona, Colorado, New Mexico, and Utah. The Southwest physiographic region is largely comprised of desert, rocky canyons, semiarid grassland, and piñon-juniper woodland, interspersed with numerous mountain ranges supporting both conifer and deciduous hardwood forests (McLaughlin 1986). From a digital cover map of the Southwest (Evans et al. 1993, Evans and Zhu 1993), we selected two forest cover types, mixed-conifer and ponderosa pine, to represent the habitat pattern of interest; all other cover types were considered "nonhabitat." The mixed-conifer habitat type was principally composed of Douglas-fir (Pseudotsuga menzesii ), white fir (Abies concolor ), and several species of pine (Pinus ponderosa, P. contorta ). The = ponderosa pine habitat type was dominated by ponderosa pine (P. ponderosa ), but probably included some pine-oak woodlands as well. The habitat map had a grid-cell resolution of 1 km x 1 km.Our original motivation for choosing ponderosa pine and mixed-conifer forest types was to approximate the distribution of suitable nesting habitat for Mexican Spotted Owls (Strix occidentalis lucida) within the southwestern United States. The data were used in developing a recovery plan for the Mexican Spotted Owl (Keitt et al.1995, USDI Fish and Wildlife Service 1995).
Landscape graphs
Because our study concerned the spatial relationships among landscape pat= ches, we chose to represent the habitat distribution as a mathematical "g= raph" (Harary 1969). Graphs are composed of sets of points ("vertices") a= nd lines ("edges") connecting vertices. Previous work has shown graphs to= be a useful representation of landscape structures (Cantwell and Forman = 1993), but have relied on purely subjective assignment of vertices and ed= ges to landscape features.We developed an objective means of generating landscape graphs (Fig. 1 ). The input data consisted of a raster habitat map with each grid cell assigned the value one if that cell corresponded to the chosen habitat type, and zero if the cell was considered nonhabitat. A patch was defined as a spatially contiguous set of habitat cells. Two habitat cells were considered to be in the same habitat patch if they were adjacent in the four cardinal directions or diagonally (northeast, northwest, and so on). Once the habitat patches were labeled, we assigned a graph vertex to each patch. Edges were placed between vertices if their corresponding patches were connected according to the criteria to be described. Thus, the landscape graph consisted of a set of vertices, one for each habitat patch, and a set of edges indicating connections among patches. For display purposes, vertices were placed at the geometric center of each patch.
FIG. 1. Construction of a landscape graph. This example landscape
contains three separate habitat patches and a single habitat cluster, or
``subgraph.'' The three patches belong to a single cluster because there
exists a path along the graph edges (solid lines) that connects all three=
patches. If either one of the edges shown were removed, there would be
two habitat clusters. If all edges were removed, then there would be thre=
e
habitat clusters, each consisting of a single patch.
![](3D"figure1.gif")
Our initial labeling of the forest data produced > 6000 individual
habitat patches, some as small as a single grid cell. We were
concerned that the smallest patches might not represent true habitat
patches, but rather noise introduced by classification errors. Thus,
we developed a novel method of filtering potential classification
errors from the data. Our approach was based on applying neutral
models (Gardner et al. 1987) to divide patches into two groups: those
that were likely to result from random error and those that were more
likely to represent true habitat patches.
In classified landscape data, small habitat patches are much more likely to be the product of random error than are large patches. For example, if errors are introduced into an image with probability p << 1, then the probability of creating a patch of errors of size n is approximately p n. Random errors are unlikely to produce large patches, because as n increases, pn becomes very small. The forest data covered ~ 38% of the map area. Of the patches identified in the forest data, roughly 50% were single-cell patches. We calculated that a random map with 18% cover would also have 50% single-cell patches. Thus, an 18% error rate was sufficient to account for all single-cell patches (i.e., those most likely to be the result of classification error) in the forest data.
We then asked, over what patch sizes was the frequency of forest patches similar to the expected frequency of patches created by random noise? We plotted the frequencies of forest patch sizes and of patch sizes observed in 100 random maps with 18% density (= Fig. 2 ), and found that the two distributions were similar up to a patch size of 10 km2 . Forest patches > 10 km2 in size occurred at much higher frequencies than patches of equivalent size in the random maps. Thus, we retained only patches >10 km2 in subsequent analyses.
![](3D"figure2.gif")
Scaling analysis
In order to explore the effect of scale on landscape connectivity, we def= ined a patch connection criterion based on minimum distances between patc= hes. We reasoned that the major barrier to dispersal would be gaps in hab= itat distribution across which an animal would have to travel to reach a = neighboring patch. The extent to which these barriers limit animal moveme= nts and, hence, connectivity will, of course, depend on the scale at whic= h an organism perceives landscape pattern. Species capable of long-distan= ce dispersal will be less impacted by gaps in habitat distribution than w= ill species with short-range dispersal. To emulate the effect of dispersa= l behavior on landscape connectivity, we constructed landscape graphs by = placing edges between patches only if the minimum distance between them w= as less than a specified threshold. Sets of patches joined by edges forme= d "subgraphs" or isolated components of the graph. By varying the distanc= e threshold, we were able to resolve the connectivity structure of the la= ndscape across a wide range of scales.
Connectivity measures
Percolation theory is the study of connectivity in stochastically generated structures (Stauffer and Aharony 1985). Thus, connectivity measures developed in percolation theory were a natural choice as measures of landscape connectivity. Most applications of percolation theory in ecology (Gardner et al.1987) are based on lattice percolation, in which cells on a lattice are occupied with probability p and are left unoccupied with probability 1 - p. A standard mathematical result in lattice percolation is the emergence of a large "spanning cluster" at a critical probability pcrit. If connections between adjacent, occupied lattice cells are restricted to occur only in north, south, east, and west directions, then the critical probability is equal to 0.5928 (Stauffer and Aharony 1985). Other critical probabilities can be obtained by changing the connectivity rule, for example, by allowing connections to run diagonally (northeast, southwest, etc.) between occupied cells.Percolation need not occur on a regular lattice. A more general model involves percolation on graph structures, referred to as "bond percolation" (Stauffer and Aharony 1985). Because regular lattices are a subset of possible graph configurations (Harary 1969), bond percolation naturally incorporates lattice percolation as a special case. In bond percolation, sites are either connected with probability p or are unconnected with probability 1 - p. As in lattice percolation, a large cluster of connected sites occurs at a critical value of p. Our application of percolation theory represents an extension of standard bond percolation. In our model, the probability of a connection between two habitat patches depends on the distance between the patches. Thus, the emergence of a spanning habitat cluster occurs at a critical mean dispersal distance instead of at a particular habitat density, as in traditional lattice percolation. Percolation problems where connection probabilities vary across the landscape are referred to as"gradient" or "non-uniform" percolation (Milne et al. 1996).
In percolation theory, connectivity is associated with the average size of connected clusters. A natural measure of the size of a circular cluster is its radius. However, in general, clusters are not round; they can be irregular, sinuous structures. Therefore, a measure of cluster size must control for irregular shapes. A measure of cluster size used in percolation theory is the "radius of gyration," defined as
![](3D"eqn1.gif")
where < x > and < y > are the mean x and y coordinates of lattice cells in the cluster, xi and yi are the coordinates of the ith grid cell in the cluster, and n is the total number of cells in the cluster (Creswick et al. 1992). Habitat clusters were defined as sets of patches connected by a subgraph or component of the thresholded landscape graphs. For a cluster comprised of several habitat patches, the sum in Eq. 1 was taken over all habitat cells among all patches in the cluster.
Unlike unitless indices of landscape connectivity, the cluster radius has units of distance and a direct physical interpretation. Imagine a randomly moving particle placed randomly on a habitat cluster. The radius of gyration is the average distance that the particle will move before encountering the cluster edge. Similarly, if a dispersing animal is restricted to moving on a particular habitat cluster (i.e., it has a low probability of traversing any gap separating it from another cluster), its average dispersal range will correspond to the radius of the cluster.
The size-weighted average connectivity of a set of clusters defines the c= orrelation length of a landscape. The correlation length of a set of clus= ters is given by
![](3D"eqn2.gif")
where m is the number of clusters and ns is the number of grid cells in cluster s (Creswick et al. 1992). We used the correlation length as an overall measure of habitat connectivity in a landscape. As with the radius of gyration, correlation length has units of distance: it is the average distance an individual is capable of dispersing before reaching a barrier, if placed randomly on the landscape. Thus, as the correlation length increases, landscape connectivity increases.
Patch removal experiments
We wanted not only a measure of overall landscape connectivity, but also local measures of patch connectivity. In particular, we wanted to know the effect on landscape connectivity of removing a patch from the landscape. We designed a simple analysis in which each habitat patch was individually removed from the landscape. We recorded the change in the correlation length induced by removing each patch. Because the effect of removing a patch depended on scale, we repeated the patch removal analysis for a large number of distance thresholds. After a patch was removed, a new landscape graph was constructed. The connected subgraphs were then identified and the new correlation length was recorded. Letting C ( d ) be the correlation length of the landscape graph, thresholded at distance d , and C ( d, i ) be the correlation length after patch i was removed, then
![](3D"eqn3.gif")
is the normalized importance index of patch i. The normalized importance is a relative index of the contribution of each patch to overall landscape connectivity. We chose a normalized, unitless index because it was designed precisely for among-patch comparisons.
We expected that large patches would have larger importance indices, simply owing to their size. Thus, we wanted a measure of per area importance as well. The per area importance index quantified a patch's contribution to overall landscape connectivity per unit area. We defined the per area index as
![](3D"eqn4.gif")
where n ( i ) is the number of habitat cells in patch i and ac is the area of a grid cell. The per area index had units of 1/area, e.g., km-2.
Stochastic landscape graphs
To make percolation on landscape graphs equivalent to a bond percolation problem, it was necessary to define a stochastic function determining connections among patches. Because successful dispersal across gaps in the habitat distribution implies moving a distance at least, if not greater than, the minimum distance between patches, we used a cumulative probability distribution to indicate whether two patches were effectively connected for a given dispersal distance. For thresholded landscape graphs, the dispersal function was simply the step function
![](3D"eqn5.gif")
where d was the minimum distance between patches. An important property of the threshold function was that a given threshold distance resulted in only a single landscape graph structure. However, the percolation problem can be generalized to include any probability function. Beyond the threshold function, the simplest probability function is the cumulative negative exponential
![](3D"eqn6.gif")
where p ( d ) is the probability of dispersing at least distance d, and k is the dispersal coefficient. The dispersal coefficient has units of 1/distance; its inverse, 1/k, is the average dispersal distance under the distribution. Distributions requiring more parameters, such as the gamma distribution, could be used as well.
Unlike the threshold function, the negative exponential function resulted in a large number of landscape graphs for any given dispersal coefficient. Following standard techniques in percolation theory, we used a Monte Carlo procedure to generate multiple landscape configurations. We varied the dispersal coefficient, k , from 1.0 (average dispersal distance =3D 1 km) to 0.01 (average dispersal distance =3D 100 km). For each value of the dispersal coefficient, 100 randomly connected landscape configurations were generated and their correlation length was recorded. Connections between patches in each configuration were determined by comparing a uniform random number between 0.0. and 1.0 to the probability of dispersal occurring between patches, according to Eq. 6. If the random number was less than the the dispersal probability, the patches were joined into the same subgraph or cluster. Over many random configurations, the expected frequency of two patches being joined was equal to the probability of successful dispersal between them. Patches far apart were rarely connected, whereas patches whose borders were in close proximity were often connected, the frequency being determined by the dispersal coefficient.
RESULTS
Thresholded landscape graphs
As the maximum interpatch dispersal distance was increased, the forest cover map became increasingly connected and eventually coalesced into a single, large graph spanning the entire habitat distribution (Fig. 3 ). At a 20-km threshold distance, the landscape was largely composed of independent patches and small habitat clusters. For organisms capable of dispersing 20 km, the landscape was highly fragmented. At 40 km, larger subgraphs formed, but the landscape was still divided into several habitat clusters. Above 40 km, most of the habitat distribution was connected. Although most of the habitat was joined at 50 km, only a single edge existed between the large subgraphs in the northeast and southwest of the habitat distribution. The vertices at either end of the single connecting edge are known as "articulation points," because removing either one would bisect the graph (Harary 1969). At a threshold distance of 80 km, the graph was highly interconnected and, in general, there were many alternate pa --------------4BEC5F7F34EB-- From cestvr@ces.iisc.ac.in Wed Jan 14 08:17:19 1998 Received: from energy.ces.iisc.ac.in by ces.iisc.ac.in (ERNET-IISc/SMI-4.1) id IAA20842; Wed, 14 Jan 1998 08:17:19 GMT Message-ID: <34BD3162.34D@ces.iisc.ac.in> Date: Wed, 14 Jan 1998 13:42:58 -0800 From: "Dr. T.V.Ramachandra"Copyright © 1997 by the Ecological Society of America
Peterson, G., G.A. De Leo, J.J. Hellmann, M.A. Janssen, A. Kinzig, J.R. Malcolm, K.L. O'Brien, S.E. Pope, D.S. Rothman, E. Shevliakova, and R.R.T. Tinch. 1997. Uncertainty, Climate Change, and Adaptive Management. Conservation Ecology [online] 1(2): 4. Available from the Internet. URL: http://www.consecol.org/vol1/iss2/art4Young Scholar Dialogue
Uncertainty, Climate Change, and Adaptive Management
Garry Peterson1, Giulio Alessandro De Leo2, Jessica J. Hellmann3, Marco A. Janssen4, Ann Kinzig5, Jay R. Malcolm6, Karen L. O'Brien7, Shealagh E. Pope8, Dale S. Rothman 9, Elena Shevliakova 10, and Robert R.T. Tinch 111Department of Zoology, University of Florida, USA; 2Department of Electronic Engineering and Information Sciences, Politecnico di Milano, Italy; 3Center for Conservation Biology and Department of Biological Sciences, Stanford University, USA; 4National Institute for Public Health and the Environment, The Netherlands; 5Department of Ecology and Evolutionary Biology, Princeton University, USA; 6Faculty of Forestry, University of Toronto, Canada; 7CICERO, Norway; 8Conservation Ecology, Canada; 9Sustainable Development Research Institute, University of British Columbia, Canada; 10Integrated Assessment Group, Carngie Mellon University, USA; 11York University, UK.
- Introduction
- Climate Change and Its Impacts
- Climate and People
- Navigating an Uncertain Future
- Responses to This Article
- Literature Cited
KEYWORDS: adaptive management; climate change; global change; uncertainty; models.
INTRODUCTION
Humanity's transformation of the Earth has increased the concentration of greenhouse gases, thereby altering Earth's climate (Walker and Steffen 1997). The drivers and the potential consequences of climate change are interwoven with a huge variety of biogeophysical and human-caused processes that complicate the analysis of policies designed to mitigate and adapt to climate change. In this paper, we explore how adaptive management (Walters 1997) can be used to grapple with the regional and global scientific, economic, and political uncertainties of climate change.
CLIMATE CHANGE AND ITS IMPACTS
Atmospheric change
Climate change policy has focused upon the reduction of greenhouse gas emissions. Emission reduction must be a central component of any climate policy; however, reorganizing society to reduce greenhouse gas emissions quickly will be difficult. Atmospheric concentrations of greenhouse gases will probably continue to increase in the near future. Existing concentrations of greenhouse gases are likely to alter climate, and future emissions will only add to that alteration. Although the extent of this alteration is uncertain, it could prove significant (Azar and Rodhe 1997). Extreme impacts are likely in some regions (Mahlman 1997).
Resilience
Walker and Steffen (1997) point out that global changes, such as biodiversity loss, land use/land cover change, hydrological modification, and the alteration of global biogeochemical cycles, will interact with climate change to alter ecosystems in complex ways across a broad range of scales. Ecological resilience, the ability of an ecosystem to persist despite disruption and change (Holling 1973), depends upon the continuity of ecological processes at smaller and larger scales (Peterson et al. 1998). The pervasive and synergistic impacts of global change threaten to reduce ecological resilience at local to global scales, producing ecosystems that are increasingly brittle and sensitive to disruption.
Ecological reorganization
Climate change affects species individually (Root 1993, Pitelka et al. 1997). Different species and populations migrate, establish, and become extinct at different rates. Climate change, therefore, will cause the dissolution of existing ecosystems and the formation of new ecosystems. Ecological collapses will probably eliminate some species entirely, and these species losses may cause the elimination of entire ecosystems. The Earth may lose cold-adapted systems such as arctic and alpine communities to warming, and low-elevation islands to sea level rise. New ecosystems will form as a consequence of climate change, but the membership of these new systems will be drawn from the subset of existing species that can survive under the new conditions. Although climate change has occurred many times throughout Earth's history, the ecological impacts of the current climate alterations are amplified by other anthropogenically-driven global changes.
Walker and Steffen (1997) also note that the rate at which existing ecosystems dissolve will exceed that at which new assemblages form. Mortality of existing vegetation through processes such as insect outbreak, flooding, or fire - all processes projected to increases under climate change - will occur rapidly compared to the accumulation of vegetative structure in new ecosystems. Under global change, this rate difference, combined with the human removal of old-growth ecosystems, will increase the area covered by early successional ecosystems (Walker and Steffen 1997). Such a global increase in early successional ecosystems suggests that species dependent upon old-growth ecosystems will become increasingly vulnerable, whereas populations of early successional species, often "weedy" species, will potentially increase.
CLIMATE AND PEOPLE
Ecological services
The extent to which species populations can adapt and ecosystems can shift, disintegrate, or reorganize has implications for humans at cultural, economical, and ecological levels. Particular species and ecosystems hold cultural value for different societies (Colding and Folke 1997). Many species and ecological services are economically valuable. The spread of many diseases is mediated by specific species and ecological processes. Climate change will disrupt these and other relationships in uncertain ways that will benefit some, but will probably harm many.
Basic ecological services, such as carbon fixation, can be produced by simple ecosystems (Ewel et al. 1991); however, the elimination of more complex ecosystems may reduce the flexibility and range of ecological services generated globally. Simplification of ecological systems may also reduce their capacity to respond to novel conditions in the future. Although humans depend upon ecological products and services, there is little understanding of how these are produced, maintained, enhanced, or degraded (Daily 1997).
Vulnerability
Humanity is threatened by both direct and indirect consequences of climate change. Unlike other species, humans have the ability to plan for the future and to invest in technology and learning to mitigate and adapt to future changes. However, just as the climatic and ecological impacts of climate change are unequally distributed, so, too, is the adaptive capacity to cope with these impacts. Wealth, infrastructure, and political stability all contribute to a nation's capacity to anticipate and respond to change. A poorly educated populace, limited physical infrastructure, degraded natural capital, or ineffective governance can all contribute to the vulnerability of a region to climate change. Unfortunately, some areas that are already vulnerable due to social and economic circumstances, such as small island states and the arid tropics, are also predicted to experience larger than average impacts from climate change.
At the recent Climate Change conference in Kyoto, Japan, President Clodumar of the Republic of Nauru, one of the small island nations threatened by rising sea level, argued that it was unethical to ignore these ecological and social inequities, because "...the willful destruction of entire countries and cultures, with foreknowledge, would represent an unspeakable crime against humanity. No nation has the right to place its own, misconstrued, national interest before the physical and cultural survival of whole countries." (Mcilroy 1997)
Regional inequity
People have tried to organize a global response to climate change through the United Nations Framework Convention for Climate Change (FCCC). The Convention has focused attention on reducing emissions of greenhouse gases. In many ways, it is attempting to establish an international property rights regime to regulate the human use and modification of the world's carbon cycle. At the December 1997 meeting in Kyoto, the main issues were the distribution of rights to emit greenhouse gases to nation-states and the conditions under which such rights applied. Deferred for further discussion were trade in these emission rights and the allocation of credits for natural sinks that remove greenhouse gases from the atmosphere.
Allocating property rights for emissions is complicated by the inequities in past and current emissions, population growth, technical capacity, and vulnerability to impacts. These differences underlie the willingness, or lack thereof, of different groups to pursue such a global agreement. A durable agreement must address the relative situation of nations. For example, linking the reduction of CO2 emissions to technology transfers or participation in emissions trading may provide a means to address regional inequities. These mechanisms may also provide a means whereby noncompliance can be punished and compliance can be rewarded.
Intergenerational inequity
To achieve intergenerational equity, we must leave the Earth no worse for our children than we received it from our parents. However, accomplishing such a goal is difficult. The dynamics of both atmospheric carbon and human society are slow. Choices made today will have consequences that extend over decades, but it is difficult to distinguish good choices from bad, given the uncertainties surrounding climate change.
Connecting the present to possible futures is necessary before good choices can be made. Work that demonstrates the future impact of specific individual, corporate, and societal choices provides a starting point. For example, the concept of emission corridors provides a means of linking long-term climate change to current policy decisions. An emission corridor defines the path from a range of short-term global greenhouse gas emission levels through intermediate climate goals to long-term targets such as the change in global average surface temperature or in sea level (Alcamo and Kreileman 1996). Work by the Potsdam Institute for Climate Impact Research (PIK) in Germany on what they call "Tolerable Windows" offers another approach to connecting present conditions with possible futures (Toth et al. 1997).
Reduction, mitigation, and adaptation
International negotiations have focused on limiting increases in greenhouse gas emissions. Decreasing emissions is a necessary, but not sufficient, step to address the consequences of human alteration of the atmosphere. Humanity has already increased atmospheric concentrations of CO2 and other greenhouse gases. If nations follow the Kyoto protocol, levels will still continue to increase over the next two decades. Existing physical, institutional, and "behavioral" infrastructure limit our capacity to mitigate emissions in the short term. The lifetime of an electrical generating plant is several decades; transportation infrastructures are similarly "hardwired" and slow to change. It will also take time to change institutions, because current tax structures, resource and management policies, and lifestyles do not reflect climate change realities. Given these realities, not only must we strive to reduce greenhouse gas production, but also to mitigate and adapt to the consequences of atmospheric alteration to which the Earth may already be committed.
Climate change and global change
Climate change is ecologically and socially intertwined with other forms of global change (Vitousek 1994). Although the scientific investigation of these processes needs to be integrated, it is unclear whether similar integration would facilitate action on climate change at the global policy level. The negotiation of international treaties is likely to be aided by a narrow focus that considers interconnected issues in isolation. Without such isolation, efforts to reach agreement are easily stalled by calls to wait for clarification of all interconnections. However, a more integrated approach might better address synergisms between issues, decrease negative externalities, and reduce administrative costs. Although the first approach may be more practicable in the international political arena, narrowly focused environmental laws in the United States have led to calls for integrative ecosystem management (Christensen et al. 1996).
At the scale of cities or regions, the impacts, adaptation to, and mitigation of global change should be approached in an integrated manner, because the interwoven consequences of individual policy decisions will be more tractable at this smaller scale. For example, building a dam may emit methane, destroy fisheries, and generate "cheap" electricity. Is this better or worse than building a coal-fired power plant or investing in energy conservation? An integrated approach may actually strengthen the imperative to act on climate change. Overlap in benefits may be used to argue for action; for example, reforestation may improve habitat for wildlife, provide timber supplies for the future, and control soil erosion, as well as sequester carbon.
NAVIGATING IN AN UNCERTAIN FUTURE
Novelty and uncertainty
The ecological, social, and economic dynamics of the changing Earth all encompass uncertainties that can be categorized as statistical uncertainty, model uncertainty, or fundamental uncertainty (Hilborn 1987). Statistical uncertainty is the uncertainty that surrounds a variable when its state at any one point is unknown, but the probability distribution that characterizes that variable is known. For example, the probability of a tree being struck by lightning is a form of statistical uncertainty. Model uncertainty occurs when the connections between variables are uncertain. Such uncertainty allows the prediction of outcomes, but makes it difficult to assess their likelihood. For example, the Atlantic conveyor has periodically been turned off, but the processes causing this are not understood well enough to predict the likelihood that the event will occur under possible future climatic conditions (Broecker 1996). Finally, fundamental uncertainty describes novel situations for which existing models do not apply. The discovery of the ozone hole falls into this category of uncertainty. Careful science can reduce, but not eliminate, these uncertainties. However, such science is often expensive, especially for large, weakly replicable systems such as the global climate system.
The uncertainties and complexity of the forces driving the social, biological, and physical dimensions of global change ensure that it will have surprising consequences (Clark 1986, Schneider and Root 1996). The political challenges posed by the novelty of climate change are compounded by its scale. The human domination of the earth defines a new geological epoch (Vitousek et al. 1997). Although historical studies can help scientists to understand ecological processes, they do not provide analogues for a future Earth transformed by global change. As we continue to change processes at a global scale, past experience will serve less often as an accurate model of future conditions, shifting the balance of the uncertainties we face from the more easily managed categories of statistical and model uncertainty to that of fundamental uncertainty.
Adaptive management
Coping with novel situations requires the capacity to learn. Walter's (1997) paper discusses Adaptive Environmental Assessment and Management to manage ecological systems through a structured process of learning by doing (Holling 1978, Walters 1986, Lee 1993). The policy-based experimentation advocated by adaptive management is essential to reduce the ecological, social, and economic costs of learning. Adaptive management focuses upon developing alternative hypotheses, identifying gaps in knowledge, and assessing what knowledge would most effectively distinguish alternative hypotheses and, therefore, could be most useful in setting and updating research and action priorities.
We argue that climate change policy could benefit by taking an adaptive approach. Considering the uncertainty surrounding climate change, knowledge and policy need to be continually updated and challenged. The International Panel on Climate Change (IPCC) has managed such a process with the science of climate change. It would be ideal if a parallel adaptive approach could be incorporated into policy development. Such an approach would produce policies and treaties that are robust to key uncertainties, that test alternative policies, that provide opportunities for learning, and that monitor policy outcomes. Although it is difficult to conduct experiments at a global scale, there are many opportunities, regionally and nationally, to test competing models of effective and fair policies.
Models
Models are important tools for evaluating alternatives in an adaptive management framework. Models can be used to identify the important gaps that exist in understanding. If their dynamics and behavior are clearly communicated, models can serve to communicate possible futures and to bound the range of our uncertainty. They can help to make the link between local actions and the aggregation to global consequences.
It is crucial that models consider not only stable states, or the state at specific future dates, but the dynamic trajectories that are required to reach those endpoints. Models may produce similar endpoints, but exhibit substantial differences in their trajectories. Although the predicted endpoint may be acceptable, a specific path may not. Such trajectory analyses can elucidate regional disparities in impacts of not only climate change but also efforts to adapt to and mitigate climate change.
To be useful, models must help us to understand, rather than hide, important sources of uncertainty. The importance of uncertainties can be determined by assessing the degree to which model behavior is altered by changes in parameter values (sensitivity analysis), model organization, and external disturbances. Because one cannot address all sources of uncertainty, it is often useful to focus these analyses around a few relevant scenarios.
Evaluating alternatives
To evaluate alternatives, scientists and policy makers must develop a more sophisticated approach toward uncertainty than they have traditionally used. Rather than simply testing and rejecting individual hypotheses, scientists and decision makers must consider diverse sets of alternative hypotheses. Alternatives need to be continually revised, modified, and discarded, based upon how they fare in tests against empirical data (Hilborn and Mangel 1996). Maintaining the status quo must be explicitly examined as one alternative among many, with its attendant consequences, benefits, and costs.
Cost-benefit analysis is a technique that is commonly used to evaluate alternative projects or decisions (Cline 1992). The relative costs and benefits of each case over a time period are estimated, are discounted back to present values, and are summed to yield the net present value of the project. Cost-benefit analysis is useful in assessing the relative merits of alternative projects, but political and ethical issues are involved with its application. The choice of a discount rate is fundamentally an ethical choice about intergenerational equity (Howarth and Norgaard 1995). Similarly, when costs and benefits are shared over a group, the aggregation of individual preferences requires some method of assigning value to different preferences. Comparing different individuals' preferences is implicitly ethical and political, and is therefore an area that is contested. Similarly, because the members of a group and their preferences can change over time, cost-benefit analyses should assess the sensitivity of the conclusions reached to preference changes; such changes will affect the value placed on outcomes and discount rates, which will alter the end result of the cost-benefit analysis (Pearce and William 1994).
More often than not, policy decisions have multiple dimensions that are difficult, if not impossible, to convert into a single metric. In these cases, techniques such as multi-attribute utility analysis, wherein tradeoffs between alternatives are evaluated using multiple metrics, may be necessary. In either case, such methods of analysis are best viewed not as authoritative objective procedures, but as modeling processes that provide a means of making underlying valuations open to scrutiny, discussion, and sensitivity analysis.
Metrics
Comparison of alternatives requires the use of common metrics. Determining appropriate metrics to analyze global impacts is difficult because of heterogeneity between regions. Commonly, dollar values, adjusted for purchasing power parity, are used for cross-country comparisons. However, such comparisons do not accurately capture nonmarket services, including ecological services. Alternative indicators that include natural services and capital provide a means to capture a more complete view of the human impacts of climate change. Such methods include ecological footprint (Wackernagel and Rees 1995), the sustainable process index (Krotscheck and Narodoslawsky 1996), and the United Nation's Human Development Index (United Nations Development Programme 1997).
Communicating uncertainty
Uncertainty does not imply "no risk." Rather, it constrains our ability to precisely qualify and quantify the risks associated with different management actions. The precautionary principle suggests that the greater our uncertainty (i.e., the less our capacity to precisely define risk), the more cautious and "reversible" our management actions should be. Although future research may narrow uncertainties, the scale of our actions is creating new uncertainties, further reducing our capacity to predict risk.
We propose that policy-oriented science must actively address uncertainty, rather than simply focus on trying to eliminate it. Often, scientists have approached political questions by emphasizing the uncertainties surrounding particular policy issues and calling for further research. Not all uncertainties can be reduced by further research, and even where reduction is possible, it may come at great cost; hence, scientists should articulate where and how science can continue to reduce uncertainty and where it cannot. This approach acknowledges limits to scientific knowledge and constrains the growth of a technocracy. However, because it secedes scientific control over a policy question to a broader community, such an approach can be controversial.
Walters (1997) discusses how management agencies often suppress scientific dissent in order to present a unified, "certain" front to the outside world, thereby consolidating the political power of the agency. However, political power can rapidly dissolve when an agency's policies fail (Hutchings et al. 1997). We believe that an active approach toward uncertainty is necessary to produce policies that fairly and openly address the uncertain future. Such an approach presents the opportunity to develop more sophisticated public debates on management issues.
Many individuals are concerned about the environment, but neither take nor advocate environmental action. One reason may be that the costs of such actions are near, whereas the benefits are far in both space and time. Another fundamental reason for inaction is uncertainty in determining the relative merits of different actions. In the absence of better information, people may assume that the status quo is preferable, because uncertainties about the positive and negative aspects of action balance one another. A clear discussion of uncertainty and its variety and location can alter people's perception of possible futures. We argue that effective public policy demands that scientists work to clearly communicate the uncertainties surrounding alternative futures, how those uncertainties can be reduced, and what actions provide the best insurance in the face of those uncertainties.
Politics and experimentation
Walters (1997) concludes that some of the most significant barriers and difficulties to applying adaptive management are social. Different people and different ecosystems benefit or lose from specific ecological changes, and, therefore, "conflicts over ecological values are likely to be one of the main impediments to policy design in adaptive management and ecological restoration". Attempting to overcome these social rigidities requires the integration of the political into adaptive management. This presents a challenge.
High levels of uncertainty provide a warning that surprises and unexpected events are likely to occur. A surprise event, such as a change in ocean circulation, could have both negative and positive consequences. What types of policies build social and ecological resilience to allow people and nature to react and adapt to surprise? What types of institutions can be developed to integrate experimental approaches to local emission reduction and adaptation with global agreements and coordination?
The need for scientific action
From a scientific perspective, multiple efforts to integrate the many aspects of global change are necessary. Increased interaction between scientists and policy makers offers the possibility to improve both decision-making and global change science. Often the range of policy alternatives considered is overly restricted. Scientists need to work to expand the range of polices that are proposed, debated, and implemented. Scientists need to make inaction uncomfortable. We need to inject novelty, new ideas about how society and nature can be organized, into the political debate, and we need to honestly test and explore all ideas to assess their relative merits. Science should visualize alternative futures, develop alternative policies, and develop opportunities for learning.
RESPONSES TO THIS ARTICLE
Responses to this article are invited. If accepted for publication, your response will be hyperlinked to the article. To submit a comment, follow this link. To read comments already accepted, follow this link.
LITERATURE CITED
Alcamo, J., and E. Kreileman. 1996. Emission scenarios and global climate protection. Global environmental change - human and policy dimensions 6:305-334.
Azar, C., and H. Rodhe. 1997. Targets for stabilization of atmospheric CO2. Science 276:1818-1819.
Broecker, W. S. 1996. Thermohaline circulation, the Achilles heel of our climate system: will man-made CO2 upset the current balance? Science 278:1582-1588.
Christensen, N., A. M. Bartuska, J. H. Brown, S. Carpenter, C. D'Antonio, R. Francis, J. F. Franklin, J. A. MacMahon, R. F. Noss, D. J. Parsons, C. H. Peterson, M. G. Turner, and R. G. Woodmansee. 1996. The report of the Ecological Society of America committee on the scientific basis for ecosystem management. Ecological Applications 6:665-691.
Clark, W. C. 1986. Sustainable development of the biosphere: themes for a research program. Pages 4-48 in W. C. Clark and R. E. Munn, editors. Sustainable development of the biosphere. Cambridge University Press, Cambridge, UK.
Cline, W. R. 1992. The economics of global warming. Institute for International Economics, Washington, D.C., USA.
Colding, J., and C. Folke. 1997. The relations among threatened species, their protection, and taboos. Conservation Ecology [online]1(1): 6. Available from the Internet. URL: http://www.consecol.org/vol1/iss1/art6
Daily, G. C. 1997. Nature's services: societal dependence on natural ecosystems. Island Press, Washington, D.C., USA.
Ewel, J. J., M. J. Mazzarino, and C. W. Berrish. 1991. Tropical soil fertility changes under monocultures and successional communities of different structure. Ecological Applications 1:289-302.
Hilborn, R. 1987. Living with uncertainty in resource management. North American Journal of Fisheries Management 7:1-5.
Hilborn, R., and M. Mangel. 1996. The ecological detective: confronting models with data. Princeton University Press, Princeton, New Jersey, USA.
Holling, C. S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4:1-23.
_______ . 1978. Adaptive environmental assessment and management. John Wiley, London, UK.
Howarth, R. B., and R. B. Norgaard. 1995. Intergenerational choices under global environmental change. Pages 111-138 in D. W. Bromley , editor. Handbook of environmental economics. Basil Blackwell, Cambridge, Massachusetts, USA.
Hutchings, J. A., C. Walters, and R. L. Haedrich. 1997. Is scientific inquiry incompatible with government information control? Canadian Journal of Fisheries and Aquatic Sciences 54:1198-1210.
Krotscheck, C., and M. Narodoslawsky. 1996. The sustainable process index: a new dimension in ecological evaluation. Ecological Engineering 6:241-258.
Lee, K. N. 1993. Compass and gyroscope. Island Press, Washington, D.C., USA.
Mahlman, J. D. 1997. Uncertainties in projections of human-caused climate warming. Science 278:1416-1417.
Mcilroy, A. 1997. Island leader makes plea for his people: vague U.S. offer doesn't hide fact global-warming summit is short on solutions. The Globe and Mail., December 9, Toronto, Canada.
Pearce, D., and D. William. 1994. Project and policy appraisal: integrating economics and environment. Organisation for Economic Co-operation and Development, Paris, France.
Peterson, G. D., C. R. Allen, and C. S. Holling. 1998. Diversity, ecological function, and scale: resilience within and across scales. Ecosystems 1, in press.
Pitelka, L. F., R.H.Gardner, J. Ash, S. Berry, H. Gitay, I.R. Noble, A. Saunders, R.H.W. Bradshaw, L. Brubaker, J.S. Clark, M.B. Davis, S. Sugita, J.S. Dyer, R. Hengeveld, G. Hope, B. Huntley, G.A. King, S. Lavorel, R.N. Mack, G.P. Malanson, M. McGlone, I.C. Prentice, and M. Rejmanek. 1997. Plant migration and climate change. American Scientist 85:464-473.
Root, T. 1993. Effect of global climate change on North American birds and their communities. Pages 280-292 in P. M. Kareiva, J. G. Kingsolver, and R. B. Huey, editors. Biotic interactions and global change. Sinauer Associates, Cambridge, Massachuetts, USA.
Schneider, S. H., and T. L. Root. 1996. Ecological implications of climate change will include surprises. Biodiversity and Conservation 5:1109-1119.
Toth, F.L., T. Bruckber, H.-M. Fussel, M. Leimbach, and G. Petschel-Held. 1997. The tolerable windows approach to integrated assessments. Paper presented at the IPCC Asia-Pacific Workshop on Integrated Assessment Models, Tokyo, Japan, March 10-12.
United Nations Development Programme. 1997. Human Development Report 1997. Oxford University Press, New York, New York, USA.
Vitousek, P. M. 1994. Beyond global warming: ecology and global change. Ecology 75:1861-1877.
Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human domination of Earth's ecosystems. Science 277:494-499.
Wackernagel, M., and W. E. Rees. 1995. Our ecological footprint: reducing human impact on the Earth. New Society Publishers, Gabriola Island, British Columbia, Canada.
Walker, B., and W. Steffen. 1997. An overview of the implications of global change for natural and managed terrestrial ecosystems. Conservation Ecology [online]1(2): 2. Available from the Internet. URL: http://www.consecol.org/vol1/iss2/art2
Walters, C. J. 1986. Adaptive management of renewable resources. McGraw Hill, New York, New York, USA.
_______ . 1997. Challenges in adaptive management of riparian and coastal ecosystems. Conservation Ecology [online]1(2):1. Available from the Internet. URL: http://www.consecol.org/vol1/iss2/art1
Address of correspondent:
Garry Peterson
Department of Zoology
111 Bartram Hall
University of Florida
Gainesville, Florida, USA 32611
Tel: (352) 392-6913
Fax: (352) 392-3704
garry@zoo.ufl.edu
Return to Table of Contents for Volume 1, Issue
2
