Key issues and research priorities in landscape ecology: An idiosyncraticsynthesis

Jianguo Wu1,* and Richard Hobbs2

1Department of Plant Biology, Arizona State University, Tempe, AZ 85287-1601, USA; 2School ofEnvironmental Science, Murdoch University, Murdoch, 6150, WA, Australia; *Author for correspondence(e-mail: jingle@asu.edu)

Received 15 November 2001; accepted in revised form 8 February 2002

Key words: Key issues, Landscape ecology, Research priorities and challenges

Abstract

Landscape ecology has made tremendous progress in recent decades, but as a rapidly developing discipline it isfaced with new problems and challenges. To identify the key issues and research priorities in landscape ecology,a special session entitled “Top 10 List for Landscape Ecology in the 21st Century” was organized at the 16thAnnual Symposium of the US Regional Association of International Association of Landscape Ecology, held atArizona State University (Tempe, Arizona, USA) during April 25–29, 2001. A group of leading landscape ecolo-gists were invited to present their views. This paper is intended to be a synthesis, but not necessarily a consen-sus, of the special session. We have organized the diverse and wide-ranging perspectives into six general keyissues and 10 priority research topics. The key issues are: (1) interdisciplinarity or transdisciplinarity, (2) inte-gration between basic research and applications, (3) Conceptual and theoretical development, (4) education andtraining, (5) international scholarly communication and collaborations, and (6) outreach and communication withthe public and decision makers. The top 10 research topics are: (1) ecological flows in landscape mosaics, (2)causes, processes, and consequences of land use and land cover change, (3) nonlinear dynamics and landscapecomplexity, (4) scaling, (5) methodological development, (6) relating landscape metrics to ecological processes,(7) integrating humans and their activities into landscape ecology, (8) optimization of landscape pattern, (9) land-scape sustainability, and (10) data acquisition and accuracy assessment. We emphasize that, although this syn-thesis was based on the presentations at the “Top 10 List” session, it is not a document that has been agreed uponby each and every participant. Rather, we believe that it is reflective of the broad-scale vision of the collective asto where landscape ecology is now and where it may be going in future.

Introduction

In the past two decades, landscape ecology has expe-rienced rapid and exciting developments in both the-ory and applications, and has transformed from a “re-gional” discipline practiced mainly in central andeastern Europe to a “global” science with its presencefound in university curricula and a variety of ecologi-cal applications. With the recent and unprecedentedadvances, landscape ecology has been enriched anddiversified greatly in theory, methodology, and appli-cations. This is evident as one browses through thefew dozens of books in landscape ecology published

in less than a decade (e.g., Haines-Young et al.(1993), Naveh and Lieberman (1994), Forman(1995), Hansson et al. (1995), Zonneveld (1995), Bis-sonette (1997), Ludwig et al. (1997), Nassauer(1997), Farina (1998, 2000), Barrett and Peles (1999),Klopatek and Gardner (1999), Mladenoff and Baker(1999), Sanderson and Harris (2000), Wu (2000),Dale and Haeuber (2001) and Turner et al. (2001)).While a number of different definitions of “land-scape” and “landscape ecology” can be found in ex-isting literature, a sample of views from some prom-inent scientists in this field (Wiens and Moss 1999)confirmed the proliferation and divergence of per-

355Landscape Ecology 17: 355–365, 2002.© 2002 Kluwer Academic Publishers. Printed in the Netherlands.

spectives and approaches. The recent diversificationin landscape ecology has apparently caused someconcerns with the “identity” of landscape ecology. AsWiens (1999) put it, “landscape ecology continues tosuffer from something of an identity crisis”. Moss(1999) warned that landscape ecology’s “healthy,youthful development will be cut off before it maturesif it does not recognize and develop its own distinc-tive core and focus”.

Do we really need to be concerned with the iden-tity of landscape ecology? What are, or should be, the“distinctive core and focus” of landscape ecology?Given the multidisciplinary origins of the field,should we embrace and solidify the interdisciplinar-ity of landscape ecology or move away from it? Howdo we integrate humans and their activities into land-scape ecology, or should we at all? To move land-scape ecology forward steadily and successfully inthe 21st century, what are the priority issues? To ad-dress these questions, a special session, entitled “Top10 List for Landscape Ecology in the 21st Century”,was organized at the 16th Annual Symposium of theUS Association of the International Association ofLandscape Ecology (US-IALE) held at Arizona StateUniversity, Tempe, April 25–29, 2001. A group ofleading landscape ecologists (Table 1) were invited topresent their views on what the most important issuesin landscape ecology are in the 21st century.

It was not surprising that the “Top 10 Lists” pre-sented by the participants varied considerably in sev-

eral ways in terms of their scope and specifics. Be-hind this seemingly overwhelming diversity andvariability, however, some common themes didemerge from an arduous analysis and synthesis. Tounderstand what the major issues in landscape ecol-ogy are according to the group of participants, we firstsorted the presented materials into different catego-ries, then tried to identify commonalities and differ-ences, and finally reorganized them into two broadgroups: (1) general issues characterizing landscapeecology as a scientific discipline and guiding its di-rections in future development, and (2) priority re-search topics defining the fundamental core and de-velopmental fronts. While classification and synthesisare two common approaches to achieving a higherlevel of organization of information and, thus, under-standing of complex phenomena, they inevitably havea certain degree of subjectivity introduced by the per-son who does the classification or synthesis. Here itis a case in point. Also, we note that the number ofboth key issues and research topics discussed belowis reflective of the authors’ logic of organizing infor-mation, not of the order of importance.

Table 1. Participants in the special session, “Top 10 List for Landscape Ecology in the 21st Century”, at the 2001 Annual Symposium ofUS-IALE held at Arizona State University, Tempe, Arizona, USA, April 25–29, 2001.

Jack Ahern Department of Landscape Architecture and Regional Planning, University of Massachusetts, Amherst, USA

Marc Antrop† Department of Geography, University of Ghent, Belgium

Bill Baker Department of Geography and Recreation, University of Wyoming, Laramie, USA

Gary and Terry Barrett Institute of Ecology, University of Georgia, Athens, USA

Virginia Dale Dale, Virginia H. Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge, USA

Almo Farina Faculty of Environmental Sciences, The Urbino University, Urbino, Italy

Richard Forman Harvard University, Graduate School of Design, Cambridge, USA

Richard Hobbs† School of Environmental Science, Murdoch University, Murdoch, Australia

Tony King† Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge, USA

Simon Levin† Department of Ecology and Evolutionary Biology, Princeton University, Princeton, USA

Arthur Lieberman† Professor Emeritus of Physical Environmental Quality at Cornell University, Ithaca, USA

David Mladenoff Department of Forest Ecology and Management, University of Wisconsin, Madison, USA.

Zev Naveh Israel Institute of Technology, Haifa, Israel

Bob O’Neill† Oak Ridge National Laboratory, Oak Ridge, USA

Monica G. Turner† Department of Zoology, University of Wisconsin, Madison, USA

John Wiens Department of Biology, Colorado State University, Fort Collins, USA

Jianguo (Jingle) Wu Symposium Organizer, Department of Plant Biology, Arizona State University, Tempe, USA

† Participants who sent in written lists, but did not give oral presentations at the meeting.

356

Key issues of landscape ecology

Interdisciplinarity or transdisciplinarity

All participants, one way or another, suggested thatlandscape ecology is and should be an interdiscipli-nary or transdisciplinary science. Currently, landscapeecology is more of a multidisciplinary endeavour, andcross-discipline fertilization and collaborations withinand beyond the realm of ecology are needed to makeit a truly interdisciplinary or transdisciplinary science.For many of the problems landscape ecologists dealwith, they should work directly with landscape de-signers, planners and managers as well as social sci-entists and decision makers.

Integration between basic research and applications

Most participants indicated that landscape ecologyshould be an integrative science in which basic re-search and applications are fully integrated. Such in-tegration should be reciprocal: research guides appli-cations and applications feedback to research. Atpresent, the degree of integration is far from beingsatisfactory. To enhance it, several activities havebeen suggested: developing research projects thatdeal with real-world problems; ensuring communica-tion between landscape ecologists and practitioners(designers, planners, and resource managers), andpresenting landscape ecology as a science of integrat-ing theory and practice in university education. Thereis an urgent need for developing landscape ecologicalprinciples and pragmatic guidelines for applicationsin resource management, land use planning, andbiodiversity conservation. On the other hand, the ap-plications are necessary and essential to the develop-ment of a science core of landscape ecology. In gen-eral, as the behavior of complex systems likelandscapes may not be predictable, some have sug-gested that landscape ecology should be perceived asmore of an anticipative and prescriptive environmen-tal science.

Conceptual and theoretical development

Landscape ecology is still lacking a generally ac-cepted conceptual and theoretical basis on whichprinciples, methods and applications can be devel-oped. Naveh and Lieberman (1994) proposed thatgeneral systems theory, biocybernetics, and ecosyste-mology form the conceptual and theoretical frame-

work for landscape ecology. However, this frame-work does not seem adequate to account for somerecent theories, principles, and methods that deal ex-plicitly with spatial heterogeneity. There seems to bean emerging view that the rapidly developing scienceof complexity (e.g., nonlinear dynamics, catastrophetheory, chaos theory, fractals, cellular automata, self-organization, hierarchy, complex adaptive systems)may provide a broader and, hopefully, sounder theo-retical basis for landscape ecology. Also, for concep-tual and theoretical developments, landscape ecologyneeds to go beyond “land” to include aquatic envi-ronments, and go beyond pecularities of specific land-scapes to also seek generalities. Several importantconceptual and theoretical topics are to be discussedin the next section.

Education and training

Education and training were considered one of themost important and pressing issues in landscape ecol-ogy by essentially all the participants. Comprehensiveand integrative university curricula and professionaltraining programs (within and outside academic insti-tutions) in landscape ecology need to be establishedand strengthened. These curricula and training pro-grams must emphasize the interdisciplinarity and ho-listic nature of landscape ecology, as well as the in-tegration between science and applications. They alsoneed to accommodate the diverse needs of studentsand professionals who have different interests andbackgrounds.

International scholarly communication andcollaborations

Several of the participants suggested that interna-tional communication and collaborations among land-scape ecologists are important to the development ofthis field. Landscapes are shaped by physical and eco-logical processes as well as socioeconomic and cul-tural factors, and landscape ecologists are inevitablyshaped by their science and cultural backgrounds andtraditions. The exchange of ideas, methods, and inter-pretations of landscapes among ecologists and prac-titioners who are accustomed to different physical,socioeconomic, and cultural environments throughcommunication and collaborations (e.g., forums,scholarly exchange programs, joint research projects)seems necessary for developing a comprehensive andcoherent core of landscape ecology.

357

Outreach and communication with the public anddecision makers

Almost all participants recognized that communica-tion and outreach are a key to the success of land-scape ecology in the decades to come. Effective com-munication between landscape ecologists and thepublic and decision-makers presently is lacking, butwill be essential for the future development of thescience and applications of landscape ecology. Com-munication is not only necessary to the integrationbetween landscape ecological research and applica-tions, but also can enhance the interdisciplinarity thisfield exemplifies. Effective communication with thepublic and people outside the “landscape ecology cir-cle” requires willingness, desire, and commitment onthe landscape ecologists’ part. It may also entail a de-parture from the traditional way of doing science.Landscape ecologists need to be engaged and proac-tive in helping shape the landscape, while being sci-entifically honest and responsible. Advanced informa-tion technologies can certainly enhancecommunication, and outreach programs of differentkinds should be promoted.

The six broad issues discussed above are all re-lated to each other, and may be generally applicableto all interdisciplinary sciences (Figure 1). Develop-ing a comprehensive and coherent scientific core maybe most essential for the future of landscape ecology.This core is not likely to be adequate if it is merely aspatial expansion (larger-scale), or a spatialized ver-sion (considering space explicitly), of existing ecolo-gies. Because the structure, functioning, and dynam-ics of landscapes are influenced by a myriad ofphysical, biological, socioeconomic, cultural, and po-litical forces, it is evident that landscape ecologyought to be interdisciplinary in theory, methodology,and practice. Landscape ecology is expected to pro-vide a scientific basis for resource management, landuse planning, biodiversity conservation, and otherbroad-scale environmental issues, which makes theintegration between basic research and applicationseven more essential than in other ecological disci-plines. The interdisciplinarity and integration of land-scape ecology reinforce each other. For example,landscape ecology provides a theoretical basis, aswell as methods, tools, data and experiences, for land-scape and urban planning and design, whereas theplanned and designed landscapes may serve as fieldexperiments to test hypotheses and theories in land-scape ecology (Golley and Bellot 1991). While inter-

disciplinarity and integration characterize the field oflandscape ecology, they must be accomplishedthrough properly designed education and training pro-grams, effective communication means and channels,and fertile collaborations of global reach.

Top 10 research topics

To develop a comprehensive and coherent scientificcore of landscape ecology, it is useful to identifysome of the major research topics. Based on the di-verse views from the group, we have derived ten pri-ority research topics that cover a wide range of theo-retical, methodological, and applied issues.Recognizing that the reciprocal integration betweentheory and application is a salient characteristic oflandscape ecology, we do not try to separate thesetopics into these three categories. It would also bedifficult to do so because each topic needs to dealwith the three aspects. However, some topics aremore of theoretical or methodological developments,and some others highlight important areas of applica-tions (Figure 2).

Ecological flows in landscape mosaics

A primary goal of landscape ecology is to understandthe reciprocal relationship between spatial pattern andecological flows or processes. This goal is far frombeing reached. While much of the attention has beengiven to spatial pattern analysis, research emphasisnow should be directed towards processes themselvesand how they affect, and are affected by, landscapepattern. Understanding the fundamental mechanismsand spatial dynamics and variability of ecologicalflows of materials (including organisms), energy, andinformation across landscape mosaics is central tolandscape ecology.

In particular, the study of the interactions betweenpopulation processes and spatial pattern has mademuch progress, but there is a need to integrate socio-economic theory of landscape change into metapopu-lation models to make them more relevant to the is-sues of biodiversity conservation and landscapesustainability. The spread of invading species has be-come an increasingly important ecological and eco-nomic problem which deserves more research efforts.In addition, little is known about the interrelationshipbetween spatial heterogeneity and ecosystem pro-cesses. For instance, how do ecosystem process rates

358

vary in space and across scales? What control suchvariations in diverse landscapes that are influenced byhuman activities with different characters and inten-sities? Apparently, integrating population, commu-nity, and ecosystem ecology into landscape ecologyshould be a high priority as well as an exciting chal-lenge.

Causes, processes, and consequences of land useand land cover change

Land use and land cover predominantly determine thestructure, functioning, and dynamics of most land-scapes throughout the world. Land use and land coverchange is driven primarily by socioeconomic forces,and is one of the most important and challenging re-search areas in landscape ecology, and indeed in glo-bal ecology. More research efforts are needed to un-derstand the causes, processes, and ecologicalconsequences of land use and land cover change.Landscape ecology needs to incorporate the insights

of economic geography which studies how economicactivity is distributed in space and resource econom-ics which determines how land will be used (O’Neill1999). Long-term landscape changes imposed byeconomies and climate change, as well as “land uselegacies” (i.e., the types, extents, and durations ofpersistent effects of prior land use on ecological pat-terns and processes), need to be considered in thestudy of land use and land cover change. In addition,highly dynamic or chaotic landscapes (e.g., urbaniz-ing landscapes or land areas under political, economicor military conflicts) may provide unique opportuni-ties for studying land use and land cover change.

Nonlinear dynamics and landscape complexity

Landscapes are spatially extended complex systemsin which heterogeneity, nonlinearity, and contingencyare the norm. While emergent properties, phase tran-sitions, and threshold behavior often characterizelandscapes of all kinds, they are the manifestations of

Figure 1. Six key issues in landscape ecology based on a classification of the presentations by the session participants. Interdisciplinarity andintegration between research and application are two crucial issues identified by essentially all the session participants. The other four issuesare all indispensable for an interdisciplinary and integrative science. The six issues and their interactions are important to sciences other thanlandscape ecology, but the emphasis on beyond-bioscience interdisciplinarity and real-world problem-solving seems one of the several char-acteristics distinguishing landscape ecology from the traditional bio-ecological disciplines such as population or community ecology.

359

nonlinear dynamics of spatially heterogeneous sys-tems. Ecological theories that can account for thesecharacteristics need to be developed and tested. Toeffectively deal with the complexity of landscapes,insights from the science of complexity and nonlin-ear dynamics may play an important role. WhileNaveh and Lieberman (1994) emphasized the rele-vance of general systems theory, cybernetics, andnonlinear thermodynamics, only in recent years do wesee a wide range of landscape ecological applicationsof concepts and methods from the science of com-plexity, particularly fractals and cellular automata.The science of complexity and nonlinear dynamics,with established and newly developing methods, mayhelp the establishment of a theoretical and method-ological basis for landscape ecology.

In supporting the above assertion, several partici-pants advocated a number of concepts and theories incomplexity science, including self-organization, com-

plex adaptive systems (CAS), nonlinear dynamics,phase transition, and metastability. Levin (1999) hasargued that ecosystems and biosphere are complexadaptive systems. In his view, heterogeneity, nonlin-earity, hierarchical organization, and flows are fourkey elements of CAS that together allow for self-or-ganization to occur. That is, CAS typically becomeorganized hierarchically into structural arrangementsthrough nonlinear interactions among heterogeneouscomponents, and these structural arrangements deter-mine and are reinforced by the flows of energy, ma-terials and information among the components (Levin1999). However, the theoretical potential and practi-cal implications of studying landscapes as CAS areyet to be fully explored.

Scaling

Scaling refers to the extrapolation or translation ofinformation from one scale to another in space andtime. Most participants thought that scaling is mostessential in both the theory and practice of landscapeecology. While scale effects are widely recognized inlandscape ecology, questions are yet to be addressedof how to determine appropriate scales for under-standing particular patterns and processes and how toscale up or down across heterogeneous landscapes.Specifically, how can information gained at finescales be extrapolated to broad scales in space andtime, or vice versa? How can knowledge at one orga-nizational level be translated to another? How can re-sults of experimental systems be extrapolated to real-world systems? What are the theoretical bases andpragmatic guidelines for aggregating and disaggregat-ing data and variables in landscape ecological re-search? In the recent decades, there has bee a greatdeal of interest in scaling issues across all earth sci-ences, and the literature in this area is expanding rap-idly. Yet, both general “rules of thumb” and specifictechniques for scaling landscape patterns and pro-cesses need to be developed and tested more widelyand rigorously. While the science of complexity maylikely facilitate the search for scaling rules and strat-egies in landscape ecology, an integrated approachthat combines field measurements, experimental ma-nipulations, remote sensing, GIS, and modelingseems imperative for developing a science of scale.

Figure 2. Top 10 research topics in landscape ecology. All the top-ics need to address the issues of theory, methodology, and applica-tion. However, some of them focus more on theoretical or meth-odological developments, whereas others emphasize more onapplications. The exact positions of the 10 topics in the 3-dimen-sional space are somewhat arbitrary, but a proper balance amongthe three components may be desirable in the study of each topic.

360

Methodological advances

Many landscape ecological problems need to be stud-ied over large and multiple scales in a spatially ex-plicit manner. The spatial heterogeneity and complex-ity of landscapes pose new methodologicalchallenges. For example, the lack of replicability atthe landscape scale often results in the problem of“pseudoreplication” (Hargrove and Pickering 1992).Apparently, this creates a serious hurdle for using tra-ditional scientific methods that hinge primarily on ex-perimentation, although Oksanen (2001) has recentlyargued that this need not be as big a problem as isoften stated. An integrative approach that combinesobservation, experimentation, and modeling seemsnecessary to deal with multi-scaled complex land-scapes. The use of meta-analysis (the statistical syn-thesis of the results of separate studies) may alsoprove valuable. How these approaches should bemixed properly needs to be addressed in landscapeecological research.

Also, the ubiquitous existence of spatial autocor-relation in landscapes violates the fundamental as-sumptions of traditional methods in statistical analy-sis and data sampling and, thus, landscape ecologistsneed to be cautious and innovative when using statis-tical methods in experimental design and data analy-sis. At the same time, more attention should be givento the proper use, evaluation of effectiveness, andecological interpretation of various spatial and geo-statistical methods in landscape ecological research.Whatever techniques (GIS included) or methods areused, they must be preceded by, and aimed at, mean-ingful landscape ecological questions. We need toavoid having powerful methodologies in search ofmeaningful questions to answer; rather we need toseek the right techniques to answer pressing ques-tions.

In addition, landscapes are often composed ofphysical, ecological, socioeconomic, and cultural pat-terns and processes. Most, if not all, of them arehighly nonlinear complex systems whose behaviormay be inherently unpredictable. This is especiallytrue when human activities and processes must beconsidered essential to the system under study. Sev-eral participants suggested that more emphasis isneeded for holistic and systems approaches as well ascomplexity theory and associated methods (e.g., self-organization, CAS, fractals, cellular automata, geneticalgorithm, neural networks). To effectively deal withthe methodological problems caused by spatial heter-

ogeneity, lack of replicability, scale-multiplicity, au-tocorrelation, and interdisciplinary complexity, land-scape ecologists may need to go beyond whattraditional sciences can offer to invent new ap-proaches. In particular, the traditional hypothetico-de-ductive doctrine is not adequate, and other scientificapproaches need to be explored (e.g., Pickett et al.(1994)).

Relating landscape metrics to ecological processes

Many landscape metrics have been developed andwidely used in the past two decades, but a soundtechnical and ecological understanding of these met-rics is still lacking. For example, the basic question,how landscape metrics relate to ecological processes,remains largely unanswered. The claim that processescan be inferred by pattern needs to be critically ex-amined in landscape ecological research. Clearly, theempirical relationships between pattern and processneed to be better documented and the underlyingmechanisms understood. Numerous studies haveshown that landscape metrics are sensitive to chang-ing scale (grain and extent), but are there general scal-ing functions across different types of landscapes?How much does a landscape need to change before ametric can detect the change? How does one deter-mine whether or not changes in metrics are signifi-cant both statistically and ecologically? Is it possibleto develop a set of standards for metrics selection andchange detection? Can a suite of “vital landscape at-tributes” (Aronson and Le Floc’h 1996) be developedfor monitoring and predicting landscape changes?How can we develop synthetic or holistic metrics thatreflect social, cultural, and ecological diversity andheterogeneity? The above questions need to be ad-dressed by combining both empirical and theoreticalapproaches. To make landscape metrics truly the met-rics of landscapes, we must “get inside” the numeri-cal appearance of metrics to find ecological essence,“move out” of the confines of the presumption thatpattern must somehow be related to processes, and“go beyond” the patch-based metrics to incorporateother forms of heterogeneity.

Integrating humans and their activities intolandscape ecology

Landscape ecology focuses on relatively large-scaleecological systems that are increasingly influencedand determined by human activities. As most of the

361

participants unequivocally indicated, socioeconomicprocesses are the primary drivers for land use andland cover change which in turn determines the struc-ture, function, and dynamics of most landscapes.Therefore, it is evident that humans themselves andtheir activities (be they rational or radical) must con-stitute an integral part of the ecology of landscapes,and they should be treated as such in research. It isbecoming increasingly apparent that this is the caseeven in areas such as North America, where the em-phasis on “natural” landscapes is slowly but steadilygiving way to a perception of the importance of hu-mans in shaping the landscape. In addition, the ideasrelating to landscape planning and design need morecareful interweaving with the biophysical aspects oflandscape ecology, particularly if we aim to allowlandscape ecology to be forward looking and to assistin preventing the recurrence of current land use di-lemmas and designing landscapes for the future.

Thus, a more humanistic perspective is needed. Infact, landscape ecology, especially in Europe, has atradition of considering humans and their activities aspart of the whole landscape. In recent years, a “holis-tic landscape ecology” perspective–a systems viewthat links natural and human systems - has been ad-vocated (Naveh and Lieberman 1994; Naveh 2000).The need for incorporating humans, including theirperceptions, value systems, cultural traditions, and so-cioeconomic activities, into landscape ecology re-quires interdisciplinarity. As such, the reciprocal inte-gration between basic research and applications hasto be the norm, not just an ideal. Although some the-ories and methodologies exist, effectively integratinghuman-related processes into ecology may remainone of the ultimate challenges for ecologists and thelike in the new century.

Optimization of landscape pattern

A fundamental assumption in landscape ecology isthat spatial patterns have significant influences on theflows of materials, energy, and information while pro-cesses create, modify, and maintain spatial patterns.Thus, it is of paramount importance in both theoryand practice to address the questions of landscapepattern optimization (e.g., optimization of land usepattern, optimal landscape management, optimallandscape design and planning). For example, canlandscape patterns be optimized in terms of both thecomposition and configuration of patches and matrixcharacteristics for purposes of biodiversity conserva-

tion, ecosystem management, and landscape sustain-ability? Are there optimal ways of “spatially meshingnature and culture”? Are there ecologically optimumnetwork forms? Research into the spatial optimizationof landscape pattern for ecological processes maypresents a new and exciting direction for landscapeecology. Such studies are likely to require theoriesand methods more than those in traditional operationsresearch (e.g., different types of mathematical pro-gramming), as well as the participation of scientistsand practitioners in different arenas.

Landscape conservation and sustainability

In view of the continuing human population growthand associated land use change and global environ-mental changes, the dynamic nature of landscapes isapparent and profound. Biological organisms andhigher-level organizations composed of them live inincreasingly fragmented landscapes. Thus, a paradoxarises: on the one hand, the conservation and sustain-ability of landscape systems ought to be an ultimategoal of landscape ecology in action; on the otherhand, such goal may not be attainable, especially, onlarge scales considering the persisting and pervasivechanges. Most of the participants recognized the im-portance of applying landscape ecological principlesin biodiversity conservation and maintaining the sus-tainability of landscapes. However, specific landscapeecological guidelines for biodiversity conservationare needed, and a comprehensive and operationaldefinition of landscape sustainability is yet to be de-veloped. Such definition may have to incorporate thephysical, ecological, socioeconomic, cultural, and po-litical components of the landscape, with explicit ex-pression of scale in time and space. A related issuethat is equally important and similarly challenging isto develop a scientifically justifiable basis and a setof pragmatic guidelines for valuing ecosystem ser-vices of landscapes. Such valuation must properlytake into account the non-marketable and intangibleaesthetic, cultural, spiritual, and non-instrumental in-trinsic nature values. Although ecologists have beenaddressing the issue of sustainability primarily interms of species and ecosystems, the reality is thathow humans perceive and value landscapes will sig-nificantly influence both the science and practice oflandscape sustainability.

362

Data acquisition and accuracy assessment

The availability and quality of data over large areasand extended time periods are critical to the develop-ment of landscape ecology because its research focusis usually on, but not restricted to, broad-scale pat-terns and processes. Indeed, it was no coincidencethat giant strides in landscape ecology often occurredin the wake of technological advances in surveyingand information-processing. Today, a suite of ad-vanced technologies are readily available to land-scape ecologists. For example, various remote sens-ing techniques provide continuous streams of digitalinformation over large areas with multiple spectral,spatial, and temporal resolutions; evolving geo-graphic information systems continue to revolution-ize the way of storing, manipulating, and analyzingspatial data; and global positioning systems allowecologists to get “spatial” quickly and accurately.Landscape ecologists as a whole are among arguablythe best equipped “high-tech” ecologists today. But,“high tech” not only often comes with “high cost”,but sometimes “high risk” as well. Several problemswere identified concerning the acquisition, quality,and analysis of landscape data.

First of all, detailed biological understanding oforganisms and species is essential to understandingmany aspects of landscape structure and function, andthis requires the collection of basic biological data onorganisms and species. Second, the problems inher-ent in sampling across large regions in a way thatpermits inference of the effects of spatial heterogene-ity remain challenging. Innovative sampling methodsare needed, using appropriate (and possibly new) sta-tistical methods and creative combinations of avail-able methods, including field sampling, experimenta-tion, remote sensing, and modeling. Third, the qualityof landscape data is often unsure due to the lack ofmetadata, error/uncertainty analysis, and accuracy as-sessment. However, the quality of data and metadatawill directly determine landscape ecologists’ abilityand effectiveness of detecting patterns and relatingthem to processes, and consequently affect researchresults and practical recommendations. Developingand testing methods of error/uncertainty analysis oflandscape data and assessing the effects of data qual-ity on the results of landscape pattern analysis andmodeling represent an extremely important and chal-lenging research priority. Fourth, to understand thestructure, function and dynamics of landscapes re-quires time series of spatial data, which in turn call

for long-term landscape monitoring programs. Land-scape monitoring is not only essential for testing land-scape ecological theories and principles, but also formaintaining landscape sustainability through adaptivestrategies. A sound landscape monitoring programneeds to be interdisciplinary in science, integrative inmethodology, and multiple in scale.

Discussion

Since 1939 when the term, “landscape ecology”, wascoined by Carl Troll, the field has certainly mademany gigantic strides in theory, methodology, and ap-plications. In the same time, new problems and chal-lenges are also abundant, begging for solutions. It isan important first step to identify what the priority is-sues and challenges are, and the special session on theTop 10 List for Landscape Ecology in the 21st Cen-tury was an attempt towards this end. There was a di-versity of views made at the special session, whichwas due, in part, to the fact that there had not been aprescribed set of rules for the participants to producetheir “lists”. We realize that this synthesis paper is aresult of “nonlinear” interactions among “fractal”components with one particular set of initial condi-tions. We are not sure that the “whole” in this case isnecessarily larger than the “sum of parts”, but it iscertain that the whole is not exactly equal to the sum.Although several common themes seem clear to us,the details may not be agreed upon by all the partici-pants. In addition, we acknowledge that neither theselection of participants nor the process of synthesiswas based on a rigorous statistical design, thus theresults reported here may not be reproducible or veri-fied in that sense.

The title of John Wiens’ presentation at the spatialsession, “Looking ahead by looking back”, seems tosuggest a historical and dialectical way of conductingthe science and practice of landscape ecology. Indeed,many of the important issues brought up at the spe-cial session have much to do with the history of land-scape ecology. The early developments of landscapeecology took place mainly in the central and easternEurope focusing on issues directly related to land-scape planning, management, conservation, and res-toration. This research emphasis on the interactionsbetween human activities and land resources necessi-tated the development of holistic, interdisciplinary,and somewhat pragmatic views and approaches. Incontrast, landscape ecology began to develop in North

363

America in the 1980s with an apparent emphasis onspatial heterogeneity and its effects on ecological pro-cesses where quantitative methods, particularly spa-tial pattern analysis and modeling, are central. Theconceptual framework for the North America per-spective is often traced back to Watt’s (1947) seminalwork on patch dynamics and MacArthur and Wilson’s(1967) theory of island biogeography. In short, therehave been two contrasting and complementary per-spectives in landscape ecology: one is more human-istic and holistic and the other more bio-ecologicaland analytical.

However, this largely geographically-based dichot-omy of the schools of thought is an oversimplifica-tion of the reality, and maybe has been exaggeratedin view of the state-of-the-science of landscape ecol-ogy. Both perspectives have been practiced by ecolo-gists worldwide, and the current trend is more of aconvergence rather than divergence (Naveh and Lie-berman 1994; Wu 1999; Wiens and Moss 1999;Turner et al. 2001). Developments in landscape ecol-ogy in recent years clearly indicate the necessity andfeasibility of integrating these two perspectives into amore comprehensive one that is holistic and with sci-entific vigor. Naveh (1988) pointed that “One of themajor challenges for landscape ecology is... to form aunified theoretical and methodological framework fora transdisciplinary science that is oriented to bothproblem-inquiry and problem-solving”. Wiens (1999)also believed that “While many interdisciplinary ap-proaches are simply traditional disciplines dressed innew clothes, “landscape ecology truly is interdiscipli-nary.” Yet, how to do holistic and transdisciplinaryresearch with acceptable scientific rigor remains agrand challenge to landscape ecologists and the like.Given the interdisciplinary and application-orientednature of landscape ecology, as perceived by most ofthe participants, how well we meet this challenge hasa major bearing on the future of the field.

The rapid diversification and inconsistency ofideas and approaches characterize most young andimmature sciences (Kuhn 1970). While this is appar-ently the case with landscape ecology, diversity,rather than divergence, of perspectives may be a last-ing hallmark of all interdisciplinary sciences. The cur-rent developmental stage of landscape ecology maybe called a stage of discovery — somewhere betweenthe infant stage and mature stage. It certainly has anumber of characteristics of immature science, suchas the lack of consistency and coherence in conceptsand theories. As discussed above, landscape ecology

is expected to be a genuinely interdisciplinary fieldthat emphasizes reciprocal integration between theo-retical developments and empirical testing and appli-cations. Because of its scientific immaturity and be-cause human activities and socioeconomic processesneed to be considered as an integral part of the land-scape under study, a dilemma often occurs which mayhinder the integration between theory and applicationas well as interdisciplinary fertilization. The commenton scientific thought by Kuhn (1983) seems quite ap-propriate for describing this predicament: “There arepolicy decisions to which scientific findings are rel-evant, but for which these findings are not preciseenough nor the theories developed enough to permitanalysis of outcomes in any but the vaguest terms. Ifscientists then respond to pressure for definite, factualanswers, they mislead policymakers. But if policy-makers insist that only precise, factual answers willdo, they reject the only help scientists can sometimesgive” (cited in Putnam (1986)). On the one hand, suchsituation is apparently not unique to landscape ecolo-gists. On the other hand, it is in the middle of thisdilemma do we see great potentials of ecologists in-fluencing landscapes! While the accuracy of the “top10 lists” can not be certain — much like any projec-tion of the highly nonlinear dynamics of complexlandscapes, the challenges landscape ecologists haveto face up are certainly grand and multifaceted.

Acknowledgements

We would like to thank all the participants of the spe-cial session, “Top 10 List for Landscape Ecology inthe 21st Century” (Table 1). While their participationis greatly appreciated, the authors are solely respon-sible for any misinterpretation or misunderstanding oftheir views. W. L. Baker, G. W. Barrett, V. Dale, A.Farina, R. T. T. Forman, S. A. Levin, D. J. Mladenoff,R. V. O’Neill, and M. G. Turner provided valuablecomments on the draft of the paper. We are particu-larly grateful to R. H. Gardner, F. Golley, and P. Op-dam for their critical reviews. All abstracts for thespecial session are available at http://www.public.as-u.edu/˜jingle/LEML/iale2001. JW acknowledges sup-port from US EPA (R827676-01-0) and NSF (DEB97-14833; CAP-LTER).

364

References

Aronson J. and Le Floc’h E. 1996. Vital landscpae attributes: Miss-ing tools for restoration ecology. Restor. Ecol. 4: 377–387.

Barrett G.W. and Peles J.D. 1999. Landscape Ecology of SmallMammals. Springer-Verlag, New York, New York, USA.

Bissonette J.A. 1997. Wildlife and Landscape Ecology. Springer-Verlag, New York, New York, USA.

Dale V.H. and Haeuber R.A. 2001. Applying Ecological Principlesto Land Management. Springer-Verlag, New York, New York,USA.

Farina A. 1998. Principles and Methods in Landscape Ecology.Chapman & Hall, London, UK.

Farina A. 2000. Landscape Ecology in Action. Kluwer AcademicPublishers, Boston, Massachusetts, USA.

Forman R.T.T. 1995. Land Mosaics: The Ecology of Landscapesand Regions. Cambridge Univ. Press, Cambridge, UK.

Golley F.B. and Bellot J. 1991. Interactions of landscape ecology,planning and design. Landsc. Urban. Plan. 21: 3–11.

Haines-Young R., Green D.R. and Cousins S.H. 1993. LandscapeEcology and GIS. Taylor & Francis, London, UK.

Hansson L., Fahrig L. and Merriam G. 1995. Mosaic Landscapesand Ecological Processes. Chapman and Hall, London, UK.

Hargrove W.W. and Pickering J. 1992. Pseudoreplication: A sinequa non for regional ecology. Landsc. Ecol. 6: 251–258.

Klopatek J.M. and Gardner R.H. 1999. Landscape Ecological Anal-ysis: Issues and Applications. Springer-Verlag, New York, NewYork, USA.

Kuhn T.S. 1970. The Structure of Scientific Revolutions. Univ. ofChicago Press, Chicago, Illinois, USA.

Kuhn T.S. 1983. Roundtable: Kuhn and Lederberg on ScientificThought. New York Times, New York, New York, USA, March13, 1983.

Levin S.A. 1999. Fragile Dominion: Complexity and the Com-mons. Perseus Books, Reading, Pennsylvania, USA.

Ludwig J., Tongway D., Freudenberger D., Noble J. and Hodgkin-son K. 1997. Landscape Ecology, Function and Management.CSIRO, Collingwood, Australia.

MacArthur R.H. and Wilson E.O. 1967. The Theory of Island Bio-geography. Princeton Univ. Press, Princeton, New Jersey, USA.

Mladenoff D.J. and Baker W.L. 1999. Spatial Modeling of ForestLandscape Change: Approaches and Applications. CambridgeUniversity Press, Cambridge, UK.

Moss M.R. 1999. Fostering academic and institutional activities inlandscape ecology. In: Wiens J.A. and Moss M.R. (eds), Issuesin Landscape Ecology. International Association for LandscapeEcology, Snowmass Village, Colorado, USA, pp. 138–144.

Nassauer J.I. 1997. Placing Nature: Culture and Landscape Ecol-ogy. Island Press, Washington DC, USA.

Naveh Z. 1988. Biocybernetic Perspectives of Landscape Ecologyand Management. In: Moss M.R. (ed.), Landscape Ecology andManagement. Polyscience, Montréal, Québec, Canada, pp. 23–34.

Naveh Z. and Lieberman A.S. 1994. Landscape Ecology: Theoryand Application. 2nd edn. Springer-Verlag, New York, NewYork, USA.

Naveh Z. 2000. What is holistic landscape ecology? A conceptualintroduction. Landsc. Urban. Plan. 50: 7–26.

Oksanen L. 2001. Logic of experiments in ecology: is pseudorep-lication a pseudo issue? Oikos 94: 27–38.

O’Neill R.V. 1999. Theory in landscape ecology. In: Wiens J.A. andMoss M.R. (eds), In Issues in Landscape Ecology. InternationalAssociation for Landscape Ecology, Snowmass Village, Colo-rado, USA, pp. 1–5.

Pickett S.T.A., Kolasa J. and Jones C.G. 1994. Ecological Under-standing: The Nature of Theory and the Theory of Nature. Ac-ademic Press, San Diego, California, USA.

Putnam S.H. 1986. Future directions for urban systems models:some pointers from empirical investigations. In: Hutchinson B.and Batty M. (eds), Advances in Urban Systems Modeling.,New York, North Holland,, pp. 91–108.

Risser P.G., Karr J.R. and Forman R.T.T. 1984. Landscape Ecol-ogy: Directions and Approaches. Illinois Natural History Sur-vey Special Publ. 2, Champaign, Illinois, USA.

Sanderson J. and Harris L.D. 2000. Landscape Ecology: A Top-Down Approach. CRC, Boca Raton, Florida, USA.

Troll C. 1939. Luftbildplan and okologische bodenforschung.Zeitschraft der Gesellschaft fur Erdkunde Zu, Berlin, 241–298.

Turner M.G., Gardner R.H. and O’Neill R.V. 2001. LandscapeEcology in Theory and Practice: Pattern and Process. Springer-Verlag, New York, New York, USA.

Watt A.S. 1947. Pattern and process in the plant community. J.Ecol. 35: 1–22.

Wiens J.A. 1999. Toward a unified landscape ecology. In: WiensJ.A. and Moss M.R. (eds), Issues in Landscape Ecology. Inter-national Association for Landscape Ecology, Snowmass Vil-lage, Colorado, USA, pp. 148–151.

Wiens J.A. and Moss M.R. 1999. Issues in Landscape Ecology in.International Association for Landscape Ecology, SnowmassVillage, Colorado, USA.

Wu J. 1999. Hierarchy and scaling: Extrapolating informationalong a scaling ladder. Can. J. Rem. Sens. 25: 367–380.

Wu J. 2000. Landscape Ecology: Pattern, Process, Scale and Hier-archy. Higher Education Press, Beijing, China.

Zonneveld I.S. 1995. Land Ecology. SPB Academic Publishing,Amsterdam, The Netherlands.

365