Proc. Nati. Acad. Sci. USAVol. 85, pp. 5986-5991, August 1988Ecology

Global change: Geographical approaches (A Review)*(environmental systems/anthropogenic factors/landscape dynamics/natural hazards/protected areas)

V. M. KOTLYAKOVt, J. R. MATHERO, G. V. SDASYUKt, AND G. F. WHITE§tInstitute of Geography, Soviet Academy of Sciences, Moscow, U.S.S.R. 109017; tUniversity of Delaware, Newark, DE 19716; and §University of Colorado,Boulder, CO 80309-0482

Contributed by G. F. White, May 17, 1988

ABSTRACT The International Geosphere Biosphere Pro-gram sponsored by the International Council of ScientificUnions is directing attention to geophysical and biologicalchange as influenced by human modifications in global energyand mass exchanges. Geographers in the Soviet Union and theUnited States have joined in critical appraisal of their experi-ence in studying environmental change. This initial report is onsome promising approaches, such as the reconstruction ofearlier landscape processes, modeling of the dynamics ofpresent-day landscapes, analysis of causes and consequences ofanthropogenic changes in specified regions, appraisal of socialresponse to change, and enhanced geographic informationsystems supported by detailed site studies.

Recognition, among people of all countries, of human abilityto modify global energy and mass exchanges between theatmosphere, oceans, and biota of the planet occurred in thepast century. Deepening understanding, nourished by re-search, showed the complex ways that flows of matter andenergy link together the Earth's physical elements and alsothe proliferation of social and economic bonds across na-tional borders, bonds compounding the potential for humandisruption of natural processes. Appreciation of these phe-nomena was expressed in a widely shared environmentalmovement and in an international scientific program-theInternational Geosphere Biosphere Program (IGBP)-thatexamines present and prospective geophysical and biologicalchanges (1, 2).The IGBP grows out of scientific evidence that terrestrial

environmental instability threatens human well-being andsurvival. United action is important and urgent because thelimits of changes in biosphere variables within which civili-zation exists form a surprisingly narrow ecological corridor.

Therefore, geographers in the Union of Soviet SocialistRepublics and the United States of America joined tocritically appraise their experience in studying global envi-ronmental change. This first report from that collaborationindicates the selection ofproblems being examined, the rangeof methods considered, and some of research opportunitiesahead. A full report will be published during the coming year;we invite critical response from scientists in geography andrelated disciplines to this preliminary communication.

Background

Before the current wave of concern for global environmentalmatters, relatively few voices called for broad approaches tothe cluster of questions about environment, development,and population. Six individuals stood out among Americanand Russian scholars.

In the 1860s, G. P. Marsh was among the first to raisequestions about the long-term deterioration of soils and

forests, chiefly in the ancient Mediterranean world (3). By theend of the 19th century, V. I. Vernadsky had pointed out thesignificance of the biogeochemical flows of nutrients inmaintaining life, and his concept of "biosphere" eventuallylaid the foundation for the United Nations Educational,Scientific, and Cultural Organization program for Man andthe Biosphere (4). His concept of "noospherogenesis"-therole of conscious human activity leading to changes in thegeobiosphere-has proved useful in approaching global is-sues (5). V. V. Dokuchaev later showed the close relation-ships among climatic, vegetation, water and soil systems, andhe began scientifically studying how these relationships werealtered by humans (6). In 1955, a biologist (M. Bates), ageographer (C. 0. Sauer), and an urban planner (L. Mum-ford) organized the first monumental interdisciplinary reviewof Man's Role in Changing the Face of the Earth (7). Sucha comprehensive appraisal ofhuman impact on the Earth wasnot repeated until 1987, when an interdisciplinary symposiumentitled "The Earth Transformed" was held by the ClarkUniversity School of Geography, the International Instituteof Applied Systems Analysis, and the World ResourcesInstitute.Many Soviet geographers conceptualize the Earth's com-

ponents as atmosphere, hydrosphere, and lithosphere-arranged in layers through which move matter and energy.On the continents, along the upper limit of the lithosphere, athin landscape or "geosystems" mantle of soil or weatheredmaterial and biota is maintained in a quasistable state bymatter and energy transfer. Flows and interactions throughthe landscape or geosystem mantle are influenced by bothexternal and internal processes, including human activities.Interaction ofthese processes produce still more complicatedcycles in the landscape. A kind of homeostatic regulationmay even exist on a global level, in which biota controlatmospheric and hydrologic conditions affecting the bio-sphere in any given combination of external processes.Generations of explorers and geographers have marveled

at the vast extent, complexity, and variability of the Earth'sbiosphere. Systematic exploration into the spatial interactivebehavior of the Earth's landscapes and the human environ-ment, indeed, has become the raison d'etre of the discipline

Abbreviations: IGBP, the International Geosphere Biosphere Pro-gram; EIA, environmental impact assessment; SIA, social impactassessment; ISLSCP, International Land Satellite Surface Climatol-ogy Project; SCOPE, Scientific Committee on Problems of theEnvironment of the International Council of Scientific Unions.*This report grows out of discussions held in Moscow in March andMay, 1987, and in Portland, OR, in April, 1987, and in Washington,DC, in July, 1987, and involved the following as contributors to thelarger book now being prepared for publication by Progress Pub-lishers in Moscow: from the U.S.S.R.-T. V. Bochkareva, A. V.Drozdov, N. F. Glazovsky, S. P. Gorshkov, A. N. Krenke, M. Y.Lemeshev, S. M. Myagkov, V. N. Solntsev, A. A. Velitchko, andV. V. Vladimirov; from the U.S.A.-R. G. Barry, J. Dozier, S. N.Goward, D. Greenland, J. K. Mitchell, W. E. Riebsame, and C. J.Willmott. This report will also be published in Russian in theProceedings of the Academy of Sciences, U.S.S.R.

5986

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

7, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988) 5987

of geography. Global maps of climate, vegetation, soils,relief, and other terrestrial phenomena contained within ouratlases represent the best efforts of our predecessors tosynthesize the then current knowledge of the Earth. Butknowledge of global-scale processes is often meager andcouched in empirical terms, for which no biophysical expla-nation has been derived (e.g., the global distribution ofvegetation is thought to be largely a function of climate; yet,the biophysical linkages that determine this relation arerelatively obscure).For the current IGBP effort to be genuinely productive, it

needs to undertake vigorously at least five lines ofgeographicresearch: (i) examination of the evolution of geosystems ofdifferent scales; (ii) analysis of landscape-change dynamics;(iii) analysis of factors, mechanisms, and effects of anthro-pogenic changes in geosystems and their socioeconomic andenvironmental causes and consequences; (iv) appraisal of thehuman dimensions of forces generating these changes and ofthe conditions in which they may be guided or controlled; (v)improvement of methods for stimulating geosystems dynam-ics, including creation of a new generation of geographicinformation systems.

Investigations should involve intensive collaborationamong disciplines, but our primary concern must be with themode of analysis rather than with the discipline.

Evolutionary Geography

A realistic assessment of the state of terrestrial landscapesand their future development needs to consider changes overevolutionary time scales. Spatial reconstruction of previousclimates and landscapes helps illuminate regularities in thedevelopment ofthe geosystem climatic fluctuations over timeand individual components of landscape systems duringthose fluctuations. Reconstructions also provide a basis forecological and climatic scenarios for the future. Such re-search may be called evolutionary geography (8).An evolutionary analysis indicates that the present-day

landscapes form a heterochronous system with strong inher-ited features, a polarized structure, and large instability. Theresponse of these landscapes to new temperature regimesmay involve quite different alterations in the zonal distribu-tion of vegetation and moisture than have occurred histori-cally. Scenarios from paleogeographical reconstructionsmust be corrected for rate differences in component modifi-cations due to natural factors (thousands of years) andanthropogenic factors (decades).

Modeling the Dynamics of Landscapes

The concept of developing a fully integrated numerical modelof the Earth stems from the success achieved by numericallymodeling atmospheric and oceanic circulation of the planet(9). Some progress has been made in developing regionalmodels of vegetation dynamics (10) and hydrologic systems(11), but no global models of such phenomena have beendeveloped. If truly interactive models of land phenomena areto be incorporated into integrated Earth models, then muchmore attention must be given to developing analytical de-scriptions of land processes.Developing dynamic models of land phenomena is difficult

because simple concepts of turbulent energy/mass balancedo not adequately describe the spatiotemporal evolution ofland conditions (12). Discrete and highly contrasting landconditions are maintained in close proximity over long timeperiods.Landscape heterogeneity varies in space and time not only

because of the diurnal, seasonal, and long-term variability ofclimate but also because of geologic events recorded in thetopography and stratigraphy of continents. Additional effects

of human activity on landscape heterogeneity date backseveral centuries, the history of which is insufficientlydocumented to incorporate into a dynamic land model (7). Aneffective macroscale description of terrestrial biophysicalprocesses is needed before progress can be achieved inglobal-change studies. Derivation of such a description willscientifically challenge the IGBP.Models of climate, biogeochemical cycles, ground water

and surface runoff all apply to global-change studies. Bylinking them with resource and economic models, effects canbe studied. For example, existing statistical and deterministiccrop-yield models have been applied to climate scenariosgenerated by global-circulation models in estimating possibleimplications of greenhouse gases (13). By varying environ-mental factors with potential future climate conditions, themodels simulate effects on future yields.

Unfortunately, certain weaknesses apply to most uses ofsimulation models in assessing environmental effects: (i) theenvironmental projections themselves may be unrealistic; (ii)the impact models often assume constant human manage-ment and input, which could vary over time; and (iii) themodels may be particularly unreliable when used to projecteffects of environmental factors substantially altered fromthose under which the models were developed. Althoughsimulation models cannot replace empirical impact assess-ment, they can help guide empirical impact studies and,perhaps, make them more efficacious.

Analysis of Anthropogenic Causes and Social Consequencesof Change

Causes of anthropogenic change in landscape variablesdeserve careful examination. Prediction of where and howlong human alterations may prevail involves cumulativeempirical evidence. Identifying possible lines of correctiveaction or positive adaptation also requires knowledge ofhuman activities as a process function. Social consequencesof environmental changes likewise deserve analysis to learnhow societies perceive these changes as hazards and thencope with them. These matters were approached tangentiallyby environmental and social impact assessments (EIAs andSIAs).

Impact Assessments in the IGBP

SIAs of global environmental changes are needed to evaluatehuman ability to cope as a species and to formulate preven-tive or adaptive responses. Since the emergence of contem-porary concerns for environmental quality in the late 1960s inthe United States, the phrase "impact assessment" connotesefforts to measure the effects of specific events on the naturalenvironment. The impact assessment process, of course,focuses on human effects on the environment, not vice versa,and the process was generally restricted to local or, at themost, regional effects extrapolated to about a decade. Theprocess lacks clear-cut guidelines for measuring effectsbecause the focus was on forecasting results to make pre-project decisions, rather than on measurement of actualsignals. Both approaches are needed in global-change stud-ies. As now pursued, the EIA process only partly applies tostudies of global-environmental changes, their regional re-sult, or their ramifications for social systems. Yet, some EIAmethods could be adapted for global-change impact studies,such as (i) the identification and monitoring of key speciesand other indicators or effects, (ii) the use of impact matrices,and (iii) methods for anticipating the interaction of multipleenvironmental effects (14, 15).The field of SIA has not focused on social implications of

environmental changes, but rather on direct social implica-tions of specific projects and development plans. Some

Ecology: Kotlyakov et A

Dow

nloa

ded

by g

uest

on

Aug

ust 1

7, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988)

topics, such as weather modification and large water-development schemes, the subject of SIAs (16), may in manyways emulate the effects of global-environmental changes atthe regional level.The fields of EIA and SIA offer guidance for assessing

complex interactions between environmental and social ele-ments, methods for extrapolating effects into the near future,designs of monitoring schemes, and approaches to handlinglarge amounts of data. Properly appropriated, these methodswill be useful as efforts begin to assess global-change effects.On request ofthe Council on the Problems ofConservation

and Improvement of the Environment of the Council ofMutual Economic Assistance Committee on Scientific andTechnical Cooperation, geographers of socialist countries in1981 prepared a report entitled "Methodological Recommen-dations on Economic and Extraeconomic Evaluation of theImpact of Man on the Environment." Those recommenda-tions were based on assessment ofthe social consequences ofeffects upon different economic sectors of specific projects inmodel regions, chosen by the countries especially for thispurpose. The following conditions were examined in thestep-by-step research: (i) the effect of human activity as a"trigger mechanism" of interaction; (ii) change in geosys-tems due to this impact; and (iii) consequences for humanhealth and economic activity resulting from the alteredenvironment. Methods were suggested for integrated study ofthe spatial unity of the interaction of people, economy, andnature to preserve and improve the environment (17).

Regions and Resource Systems

Experimental evidence suggests that sustainable manage-ment of anthropogenic-natural processes is possible if suchmanagement relies on a spatial framework consisting ofcomplex, heterogeneous regional systems (nature-resource-ecological, demographic-social-ethnic, productive-eco-nomic), which are closely interconnected. This shifts thefocus from tracing effects to regional analysis with full use ofconcepts of resource cycles, defined as the "sum total oftransformations and spatial transfer of a certain matter, or agroup of substances at all stages of its use by man" (18).Principal cycles receiving attention from geographers arethose of energy, water (19, 20), and nutrients (21). Closerlinkage of analysis of global cycles with regional analysis isneeded, as well as a recognition that each anthropogenicchange, such as industrial waste generation, originates incertain regions and affects particular regions, not necessarilythe same areas.

Socioeconomic development of any region obviously de-pends, in part, on the environment. These constraints andopportunities may be identified only when the naturalresource/ecological potential of natural biospheric subdivi-sions is known. Often, river basins serve as such subdivisions(22, 23). The basin approach to resource assessment does notexclude, but supplements the traditional physicogeograph-ical landscape approach. The approach may suggest a matrixof landscapes and drainage areas within which to investigatechange under an anthropogenic load and, in particular, in theprocess of materials migration.

Appraisal of Social Coping with Environmental Change

In considering how individuals and societies may and do copewith change once it is seen or predicted with confidence,geographers work along several lines, two of which arenoted: (') appraisal of economic development and environ-mental conservation strategies within regional settings and(ii) analysis of the ways people perceive and adapt to actualor potential environmental hazards.

Regional Development and Environmental Quality

In many parts of the world there is earnest searching foreffective means of coordinating economic development withenvironmental conservation. Much research was devoted tosimulation of ecological systems at a local scale and eco-nomic systems at regional and world scales. Primarily globalmodeling focuses on economic and sectoral systems. Muchless developed is thoughtful simulation of regional naturalresource use. Building a single management model of ascomplex a system as a region, a city, or an agglomeration isextremely difficult. Yet, in such areas much anthropogenicchange is being generated.

In these situations the challenge seems to be to buildinterrelated models of social, economic, and ecologicaldevelopment. One obstacle is that no agreed-upon criterionof socioecological/economic efficiency presently exists. Thislack is due to insufficient analysis and to incompatibility ofindices characterizing economic and mostly social and eco-logical effects of nature-conservation measures. One simpli-fied approach by geographers confronted the possibility ofclimate change. The rapidly evolving field of climate impactassessment emerged in the mid-1970s from concerns over thegrave consequences of climate fluctuations such as theSahelian drought, recent El Nifio episodes, and, of course,the potential for global climate change associated with thebuildup of greenhouse gases. Climate impact assessment is,by definition, an attempt to assess social implications ofglobal change and its regional and local manifestations, andthe field is the most developed and broadly applicable area ofglobal-change impact studies (16).

Natural Hazards and Risk Assessment

One disturbing tendency of our day is intensification ofhazards from extreme natural events and of the resultingdamage. Such events present concrete instances of howsocieties and individuals react when nature threatens theirwell-being. Geographers sought answers to this question byevaluating ways to reduce, prevent, and mitigate any conse-quences and by systematically classifying hazardous areas,and they worked to refine methods of environmental-riskassessment.Much experience has accumulated in assessing impacts

and response to environmental extremes under the rubric ofnatural-hazards research (24-28). Demographic, technolog-ical, and political circumstances, errors in the assessment ofnatural situations, and the management of production areamong causes of social losses from natural hazards.

In associated studies geographers contributed concepts,models, and methods to the emerging interdisciplinary fieldof environmental-risk assessment (29-33). Research on thehuman ecology of natural hazards established that the degreeof threat experienced by a society is a function of fourinteractive variables: risk, exposure, vulnerability, and re-sponse. Subsequent research identified previously neglectedalternative adjustments to hazard and focused attention onbehavioral factors affecting risk assessment. Later work ontechnological hazards also established the interactive char-acter of human/machine/environment systems under condi-tions of failure, underlined the importance of accurate riskcommunication, and addressed various social constraints onrisk assessment.Methods of assessment for long-established local risks are

already well developed; much is known about the natural riskcomponents ofdramatic events like earthquakes, storms, andfloods and, to a lesser extent, about drought, erosion, andother long-term, quasinatural risks. Information about hu-man exposure to risk and physical vulnerability is also widelyavailable. Fewer studies of socioeconomic vulnerability have

5988 Ecology: Kotlyakov et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

7, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988) 5989

been done, and less information has been published about theefficacy of available public and private responses to envi-ronmental hazards. Although only recently begun, the paceof research on assessment of new universal risks acceleratesrapidly. Pioneering work on the environmental consequencesof nuclear war and climate-impact assessment studies guidefuture work on other universal risks (16, 34).

Geographic Data and Information Systems

Unraveling the mechanism of global change in the coupledbiosphere-atmosphere system and determining direction andrate of such change requires modeling studies and compati-ble, homogeneous global data sets for a variety of keyterrestrial variables. Many, but not all, of these data may begathered using remote-sensing technology. A continual needexists to conduct regional field studies to coordinate remotelysensed data as well as to collect data not acquired withremote-sensing technology. Over the long term, there will beneed for consistent, quantitative global measurements ofsurface albedo, surface temperature, vegetation cover, mois-ture, snow cover (extent and depth), evapotranspiration, andprecipitation (frequency and intensity) for terrestrial ecosys-tems, and ocean color, tropospheric gases, ocean topogra-phy, sea-surface temperature, sea ice, and wind stress for theoceans (35). Atmospheric and space observations shouldinclude solar flux (and spectral variability); stratospherictemperatures, stratospheric aerosols and ozone; and tropo-spheric aerosols. Cloud cover is not specifically listed,although it is a key element of the World Climate ResearchProgram under the International Satellite Cloud ClimatologyProgram (36). Recent remote-sensing research by geogra-phers includes the estimation of precipitation rates fromvisible and IR satellite images (37), the modeling of radiationin mountainous terrain from multispectral Landsat data (38,39) analysis of the climatic significance of cloud cover (40),the estimation of large-scale net primary productivity (41),and the remote sensing of snow and ice parameters (42),especially from passive microwave data (43, 44).Model building and data collecting are corequisites for any

successful evaluation of the interactivity between terrestrialbiosphere and atmosphere. Sophisticated biosphere models(45-47) are of little value without adequate observations andprecise characterization of the near-surface environment.But, concomitantly, data that purport to characterize thebiosphere are of limited value if (i) variables required byclimate models are not sampled, or (ii) the time or spacescales of observations or statistical summary are inconsistentwith model needs. For example, daily changes in planttranspiration, in general, cannot be adequately inferred frommonthly climate data. Data collection and modeling effortsmust not be conducted in ignorance of one another. Thismeans that collected data should be representative at spacescales of ==100 kmi2, although this resolution will improvewith faster computers and more efficient algorithms. Modeltime scales, by contrast, are much more variable and rangefrom hours to millennia, again depending on the variable ofinterest and the model.The conceptual bases for conducting regional and global

studies with remotely sensed observations and computer-based geographic information systems are only now devel-oping (48, 49). A current major conflict exists between theconcepts of raster and vector storage of geographic data indigital computers, which has effectively halted the marriageof remote-sensing and geographic information systems (50).Vector data descriptions tend to preserve the precision ofboundary definitions, whereas raster systems tend to pre-serve measures of spatial heterogeneity. Because remotelysensed observations are now viewed as contiguous, butdiscrete, observations of heterogeneous phenomena, such

vectorization is not possible. Inasmuch as the IGBP funda-mentally depends on the marriage of remote-sensing togeographic information systems and global models, a solutionto these conflicts should be a focus of early IGBP research.

Further, research is needed on the interfacing of differentscale sizes. In some physical fields, such as numericalmodeling of the atmosphere, this interfacing was approachedby nesting of grids of various sizes. There might be a veryhigh density array ofdata points existing within a more sparsenetwork. This approach often works for the atmospherebecause of its commonly smooth gradients in values of thetreated variables. In some cases this treatment might work inEarth-surface problems, but in others it would not. The lattercases are those where great heterogeneity of land coverexists, and distinct changes occur over a short distance; theurban/rural boundary would be illustrative. Geographersneed new methods of varying grid size-methods that do notnecessarily treat the problem as linear in time or space. Inmental or perceptual maps, for example, we have not beenconstrained by linearity. Nor should there be need forconstraint in the present context, save only for the constraintof consistency and the adherence to physical and geometriclaws.

Interpolation between locations implies that the way inwhich the particular observed phenomena change in spaceand time between the observations is understood. Any effortto map point information without that knowledge introducesunknown errors, which may increase rapidly away from theobservation point. The combination of remotely sensedobservations and ground measurements has the potential tostudy these patterns, but carefully designed experiments,done over several years, are necessary. Preliminary experi-ments of this type are now being done under the InternationalSatellite Land Surface Climatology Project (ISLSCP) (51,52). Similar studies are needed to evaluate other types of landphenomena. In each case a major focus of such researchshould be the scaling question; that is, how do local phe-nomena aggregate to produce continental-to-global descrip-tions of land conditions and dynamics?

Full evaluation of terrestrial dynamics must include humanactivities. Greenland (53) suggested, in the context of re-source management, that many indicators can be developed,limited only by the imagination. In the IGBP studies weconsider here, however, somewhat more restraint operateswhen studies rely solely on those variables monitored fromspace or air. There is no a priori reason why only remotelysensed data should be used to monitor socioeconomic indi-cators except that consistency might be greater.

Need for Site Studies

From the foregoing the effectiveness of the data base inmonitoring and helping predict environmental changes willapparently be greatly enhanced by detailed site studies.American geographers tend to emphasize representativetarget areas, whereas Soviet geographers tend to emphasizea system of protected natural areas.

Representative Target Areas

Initially, existing network and individual study sites shouldbe used as representative target areas. Long-term ecologicalresearch sites and sites belonging to the ISLSCP immediatelysuggest themselves for use in these geographical studies. Asignificant contrast exists between long-term ecological re-search sites, which generally represent a particular biomewith minimum human impact, and ISLSCP sites chosen forthe degree of impact already experienced (54).

Ecology: Kotlyakov et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

7, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988)

System of Protected Natural Areas

One important link in building understanding of regionalenvironments would be to create a system of protectednatural areas, within which the stability of natural environ-ment would be sustained by a variety of administrativeregimes of human use. Forming a peculiar "natural skele-ton," each region in a system of protected natural areaswould have at its center a preserved natural core. A numberof subsidiary units would be managed to maintain a dynamicecological balance necessary for harmonious development ofthe economy and society. Physical and social planning wouldbe aimed at a variety of land uses and at predicting itsdevelopment. This planning would include classificationaccording to the natural and social properties, as well asdeveloped, undeveloped, and completely reserved territoryin each natural zone.

In connection with the responsibility of IGBP, it would bedesirable to elaborate criteria for choosing appropriate areasfor the organization of biosphere reserves as well as fordeveloping integrated monitoring for the entire protectedarea. Using N. I. Vavilov's ideas of maximum diversity andof migration of plants, Soviet scientists (55) developed ascheme for locating stations for integrated monitoring withincontinents and parts of the oceans' water surface. Thisscheme rests on an analysis of climatic zones, of lithological-orographic heterogeneity of continents, and of ocean circu-lation. Areas with a maximum diversity of ecosystems as wellas those with greater sensitivity to changes in thermal andprecipitation regimes are identified. The number of suchareas all over the planet would probably range between 150and 200.

In areas bordering subdivisions of the biosphere (broadlydefined as ecotones) the main features of the structure andfunction of ecosystems change significantly-primarily un-der the influence of external factors. As for the slow self-development of ecosystems and their evolution in the purestform, they can best be studied not in the interfaces and notin zones of maximum diversity of environmental conditions,but, rather, in regions with the greatest homogeneity-i.e., inthe foci of corresponding areas.

Concluding Discussion

These observations illustrate investigative lines needing ex-tention and refinement for IGBP to achieve its goals. Ade-quate understanding of global change cannot be gainedwithout reconstruction of earlier landscape processes, moresophisticated modeling of the dynamics of contemporarylandscapes, analysis of the causes and consequences ofman-made changes in specific regions, appraisal of theconditions under which societies will respond to new knowl-edge about those changes, and a greatly enhanced geographicinformation system supported by detailed site studies.

We thank the following for comments on a preliminary draft: JohnR. Borchert, S. A. Evteev, L. N. Karpov, Thomas F. Malone, andB. L. Turner. George Demko played a major role in arranging for theInternational Research and Exchange Board (IREX) encouragementand support under its Subcommission on Geography. We thank theNational Academy of Sciences-Soviet Academy of Sciences SeniorScholar Exchange Program for making possible the initial discus-sions of this project during 1987. Thereafter, meetings in Washing-ton, DC, during July, 1987, and in Newark, DE, during May, 1988,were facilitated by support from the Association of AmericanGeographers, the Soviet Academy of Sciences, and the IREX.

1. Malone, T. F. (1986) Environment 28, 6-11, 39-42.2. National Research Council (1986) Global Change in the Geo-

sphere-Biosphere: Initial Priorities for an IGBP (Natl. Acad.Press, Washington, DC).

3. Marsh, G. P. (1864) Man and Nature or Physical Geography as

Modified by Human Action (Scribner, New York); reprinted(1965) by Belknap-Harvard Univ. Press, Cambridge, MA.

4. Vernadsky, V. 1. (1981) Isbrannii trydi po istorii nauki (inRussian) (Selected Works on the History of Science) (Nauka,Moscow).

5. Glazovsky, N. F. (1988) Struktura noosphery i zadachi geo-graphii (in Russian) (Noosphere Structure and Tasks of Geog-raphy) 1, 38-48.

6. Dokuchaev, V. V. (1892) Nashi ctepiprejdi i teper (in Russian)(Our Steppes Past and Present) (Moscow).

7. Thomas, W. L., Jr., ed. (1956) Man's Role in Changing theFace of the Earth (Univ. Chicago Press, Chicago).

8. Velitchko, A. A. (1985) Proc. Acad. Sci. USSR Sect. Geogr. 6,25-35.

9. Washington, W. M. & Parkinson, C. L. (1986) An Introductionto Three-Dimensional Modeling (University Science Books,Mill Valley, CA).

10. Shugart, H. H. (1984) A Theory ofForest Dynamics (Springer,New York).

11. Camillo, P. J., Gurney, R. J. & Schmugge, T. J. (1987) WaterResour. Res. 19, 371-380.

12. Miller, D. H. (1978) in Sourcebook on the Environment, eds.Hammond, K. A., Macinko, G. & Fairchild, W. B. (Univ.Chicago Press, Chicago), pp. 63-88.

13. Warrick, R. A., Gifford, R. M. & Parry, M. L. (1986) in TheGreenhouse Effect, Climatic Change, and Ecosystems, eds.Bolin, B., Doos, B. R., Jager, J. & Warrick, R. A. (Wiley,Chichester, U.K.), Scientific Committee on Problems of theEnvironment of the International Council of Scientific Unions(SCOPE) Rep. 29, pp. 393-473.

14. Munn, R. E., ed. (1980) Environmental Impact Assessment(Wiley, Chichester, U.K.), SCOPE Rep. 5.

15. Parry, M. L., Carter, T. R. & Konijn, N. T., eds. (1987)Assessment of Climate Impacts on Agriculture (Reidel, Dor-drecht, The Netherlands), Vols. 1 and 2.

16. Kates, R. W., Ausubel, J. H. & Berberian, M., eds. (1985)Climate Impact Assessment: Studies in the Interaction ofClimate and Society (Wiley, Chichester, U.K.), SCOPE Rep.27.

17. Committee on Scientific and Technical Cooperation (1986)Vliyanie khozyastva na prirodu: Otshenki, modeli, karty (inRussian) (Economy Impact upon Nature: Assessment, Models,Maps) (Council of Mutual Economic Assistance, Budapest).

18. Komar, I. V. (1975) Ratsionalnoye ispolzovaniye prirodnykhresursov u resursnye tsikly (in Russian) (Rational Use ofNaturalResources and Resource Cycles) (Nauka, Moscow).

19. L'vovitch, M. I. (1986) Voda i zhizn: vodnye resursy, ikhpreobrazovanie i okhrana (in Russian) (Water and Life: WaterResources, Their Transformation and Protection) (Mysl', Mos-cow).

20. Mather, J. R. (1984) Water Resources Distribution, Use, andManagement (Wiley, New York).

21. Riabchikov, A. M., ed. (1980) Crugovorot veshchestva v pri-rode i ego izmenenie khozyaistvennoi deyatel'nostyu (in Rus-sian) (Matter Cycle in Nature and Its Change by EconomicActivity) (Moscow State Univ. Press, Moscow).

22. Chorley, R. J., ed. (1969) Water, Earth and Man (Matheun,London).

23. Gorshkov, S. P. (1982) Ekz dinamicheskie protsessy os-voennykh territoriy (in Russian) (Exadynamic Processes ofDeveloped Territories) (Nedra, Moscow).

24. Burton, I., Kates, R. W. & White, G. F. (1978) The Environ-ment as Hazard (Oxford Univ. Press, New York).

25. Heathcote, R. L. (1985) in Climate Impact Assessment, eds.Kates, R. W., Ausubel, J. H. & Berberian, M. (Wiley, Chi-chester, U.K.), SCOPE Rep. 27, pp. 369-401.

26. Mitchell, J. K. (1984) in Environmental Perception and Behav-ior: An Inventory and Prospect, eds. Saarinen, T. F., Seamon,D. & Sell, J. L. (Univ. Chicago Press, Chicago), Dept. Geogr.Res. Pap. 209, pp. 33-69.

27. Myagkov, S. M. (1986) Problemy geografli razrushitelnykhprirodnykh yavleniy v svete zadach uskoreniya nauchnotekh-nicheskogo progressa (in Russian) (Problems of Geography ofDestructive Natural Phenomena in the Light of Scientific-Technological Progress Acceleration) (Vestnik Moscow Univ.Press, Moscow), Ser. 5 (Geography), Vol. 1, pp. 9-15.

28. O'Riordan, T. (1986) in Geography, Resources and Environ-

5990 Ecology: Kotlyakov et A

Dow

nloa

ded

by g

uest

on

Aug

ust 1

7, 2

020

Proc. Natl. Acad. Sci. USA 85 (1988) 5991

ment, eds. Kates, R. W. & Burton, I. (Univ. Chicago Press,Chicago), Vol. 2, pp. 272-309.

29. Kasperson, R. E. & Kasperson, J. X., eds. (1988) NuclearRiskAnalysis in Comparative Perspective (Allen & Unwin, Win-chester, MA).

30. Kasperson, R. E. & Pijawka, K. D. (1985) Public Admin. Rev.45, 7-18.

31. Kates, R. W. (1978) Risk Assessment ofEnvironmental Hazard(Wiley, Chichester, U.K.), SCOPE Rep. 8.

32. Kates, R. W. & Kasperson, J. X. (1983) Proc. Natl. Acad. Sci.USA 80, 7027-7038.

33. Whyte, A. V. & Burton, I., eds. (1980) Environmental RiskAssessment (Wiley, Chichester, U.K.), SCOPE Rep. 15.

34. Harwell, M. A. & Hutchinson, T. C., eds. (1985) Environmen-tal Concequences ofNuclear War (Wiley, Chichester, U.K.),SCOPE Rep. 28.

35. Rasool, S. I. (1987) Adv. Space Res. 7, 1.36. Schiffer, R. A. & Rossow, W. B. (1983) Bull. Am. Meteorol.

Soc. 64, 779-784.37. Barrett, E. C. & Martin, D. W. (1981) The Use ofSatellite Data

in Rainfall Monitoring (Academic, London).38. Dozier, J. (1987) in Large-Scale Effects of Seasonal Snow

Cover, eds. Goodison, B. E., Barry, R. G. & Dozier, J. (Int.Assoc. Hydrol. Sci. Press, Wallingford, U.K.), Publ. 166, pp.305-314.

39. Dozier, J. (1980) Water Resour. Res. 16, 709-718.40. Barry, R. G., Henderson-Sellers, A. & Shine, K. P. (1984) in

Climate Processes and Climate Sensitivity, eds. Hensen, J. &Takehashi, T. (AGU, Washington, DC), Monogr. 29, pp. 221-237.

41. Goward, S. N. & Dye, D. G. (1987) Adv. Space Res. 7, 165-174.42. Hall, D. K. & Martinec, J. (1985) Remote Sensing ofSnow and

Ice (Chapman & Hall, New York).43. Anderson, M. C. (1987) in Large-Scale Effects of Seasonal

Snow Cover, eds. Goodison, B. E., Barry, R. G. & Dozier, J.

(Int. Assoc. Hydrol. Sci. Press, Wallingford, U.K.), Publ. 166,pp. 329-342.

44. Hall, D. K., Chang, A. T. C. & Foster, J. L. (1987) in Large-Scale Effects of Seasonal Snow Cover, eds. Goodison, B. E.,Barry, R. G. & Dozier, J. (Int. Assoc. Hydrol. Sci. Press,Wallingford, U.K.), Publ. 166, pp. 403-414.

45. Dickinson, R. E. (1985) Adv. Geophys. 28, 99-129.46. Mintz, Y., Sellers, P. J. & Willmott, C. J. (1983) On the Design

ofan Interactive Biosphere for the GLAS General CirculationModel (Natl. Aeronaut. Space Admin., Greenbelt, MD), Tech.Memo 84973.

47. Sellers, P. J., Mintz, Y., Sud, Y. C. & Dalcher, A. (1986) J.Atmos. Sci. 43, 505-531.

48. Kushkarev, A. B. & Karakin, V. A. (1987) Regionalnue geoin-formatsionnye sistemu (in Russian) (Regional GeoinformationSystems) (Nauka, Moscow).

49. (1987) Banki geographicheskik dannuh dly tematicheskoyo kar-tographirovanya (in Russian) (Geographical Data Banks for The-matical Cartography) (Moscow State Univ. Press, Moscow).

50. Marble, D. F. & Peuquet, D. J. (1985) in Manual of RemoteSensing, eds. Marble, D. F. & Peuquet, D. J. (Am. Soc. Photo-gramm., Falls Church, VA), 2nd Ed., Vol. 1, pp. 923-958.

51. Sellers, P. J., Hall, F. G., Asrar, G., Strebel, D. E. & Murphy,R. (1988) Bull. Am. Meteorol. Soc. 69, 22-27.

52. Ruttenberg, S. (1983) Draft Workshop Report on the Interna-tional Satellite Land-Surface Climatology Project (Int. Assoc.Meteorol. Atmos. Phys., Comm. Space Res., Univ. Corp.Atmos. Res./Natl. Center Atmos. Res., Boulder, CO).

53. Greenland, D. (1983) Guidelines for Modern Resource Man-agement (Charles Merrill, Columbus, OH).

54. Callahan, J. T. (1984) Bioscience 34, 363-367.55. Sokolov, V. E., Gunin, P. D., Drozdov, A. V. & Puzachenko,

Yu. G. (1983) in Proceedings of the First International Con-gress on Biosphere Reserves (GCNT, Moscow), pp. 70-87.

Ecology: Kotlyakov et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

7, 2

020