Annual Reviews

www.annualreviews.org/aronline

Annu.Rev. Ecol. Syst. 1989.20:171-97

Annu. Rev. Ecol. Syst. 1989.20:171-197. Downloaded from arjournals.annualreviews.org
by "UNIV. OF WISCONSIN, MADISON" on 04/13/05. For personal use only.

LANDSCAPEECOLOGY:The Effect
Pattern on Process1
Monica Goigel

of

Turner

EnvironmentalSciences Division, OakRidgeNational Laboratory, OakRidge, TN
37831
INTRODUCTION
A Historical
Perspective
Ecologyand natural history have a long tradition of interest in the spatial
patterning and geographic distribution of organisms. The latitudinal and
altitudinal distribution of vegetative zones was described by VonHumboldt
(154), whose work provided a major impetus to studies of the geographic
distribution of plants and animals (74). Throughoutthe nineteenth century,
botanists and zoologists described the spatial distributions of various taxa,
particularly as they related to macroclimaticfactors such as temperature and
precipitation (e.g. 21, 82, 83, 156). The emerging view was that strong
interdependenciesamongclimate, biota, and soil lead to long-term stability of
the landscape in the absence of climatic changes (95). The early biogeographical studies also influenced Clements’theory of successional dynamics,
in which a stable endpoint, the climax vegetation, was determined by macroclimate over a broad region (14, 15).
Clementsstressed temporal dynamicsbut did not emphasizespatial patterning. Gleason (36-38) argued that spatially heterogeneous patterns were important and should be interpreted as individualistic responses to spatial gradients in the environment. The developmentof gradient analysis (e.g. 17,
164) allowed description of the continuous distribution of species along
environmentalgradients. Abrupt discontinuities in vegetation patterns were
believed to be associated with abrupt discontinuities in the physical environmerit (165), and the spatial patterns of climax vegetation were thought
reflect localized intersections of species respondingto complexenvironmental
gradients.
~TheUSgovernmenthas the right to retain a nonexclusive, royalty-flee license in and to any
copyright covering this paper.
171

Annual Reviews
www.annualreviews.org/aronline

Annu. Rev. Ecol. Syst. 1989.20:171-197. Downloaded from arjournals.annualreviews.org
by "UNIV. OF WISCONSIN, MADISON" on 04/13/05. For personal use only.

172

TURNER

A revised concept of,vegetation patterns in space and time was presented by
Watt (157). The distribution of the entire temporal progression of successional stages was described as a pattern of patches across a landscape. The
orderly sequence of phases at each point in space accounted for the persistence of the overall pattern. The complexspatial pattern across the landscape
was constant, but this constancyin the pattern was maintainedby the temporal
changesat each point. Thus, space and time were linked by Watt (157) for the
first time at the broader scale that is nowtermedthe landscape. The conceptof
the shifting steady-state mosaic(3), which incorporates natural disturbance
processes, is related to Watt’s conceptualization.
Consideration of spatial dynamics in manyareas of ecology has received
increased attention during the past decade (e.g. 1, 89, 99, 103,135,161). For
example,the role of disturbance in creating and maintaininga spatial mosaic
in the rocky intertidal zone was studied by Paine &Levin (99). Patch size
could be predicted very well by using a model based on past patterns of
disturbance and on measured patterns of mussel movementand recruitment.
The dynamicsof manynatural disturbances and their effects on the spatial
mosaichave received considerable study in a variety of terrestrial and aquatic
systems (e.g. 103).
This brief overviewdemonstratesthat a long history of ecological studies
provides a basis for the study of spatial patterns and landscape-level processes. However,the emphasis previously was on describing the processes
that created the patterns observedin the biota. The explicit effects of spatial
patterns on ecological processes have not been well studied; the emphasison
pattern and process is what differentiates landscape ecology from other
ecological disciplines. Therefore, this review focuses on the characterization
of landscape patterns and their effects on ecological processes.
Landscape

Ecology

Landscapeecology emphasizesbroad spatial scales and the ecological effects
of the spatial patterning of ecosystems. Specifically, it considers (a) the
development and dynamics of spatial heterogeneity, (b) interactions and
exchanges across heterogenous landscapes, (c) the influences of spatial
heterogeneity on biotic and abiotic processes, and (d) the management
spatial heterogeneity (107).
The term "landscape ecology" was first used by Troll (138); it arose from
Europeantraditions of regional geographyand vegetation science (the historical developmentis reviewed in 90, 91). Manydisciplines have contributed to
the recent development of landscape ecology. For example, economists and
geographers have developed many of the techniques to link pattern and
process at broad scales (e.g. 53, 172), such as the developmentof spatial
models to address questions of humangeography (reviewed in 42). Landscape

For example. 1989.org by "UNIV. the structure. and change(29).g. the distribution of energy. a landscape can be considered a spatially heterogeneousarea. and a variety of practical applications are being developed concurrently (e. 26. and change of landscapes are themselves scale-dependent. kinds and configurations of components. 61. For personal use only. 153. materials. 86. 150). Most simply. OF WISCONSIN. 112. 151. "Structure" refers to the spatial relationships betweendistinctive ecosystems. 29.107. and ecological processes in landscapescan be studied at different spatial and temporalscales (106). Syst. 81. 111. 48. that is. changes in landscape structure or function are scale-dependent. Downloaded from arjournals. such as the flow of organisms.20:171-197. Landscapes can be observed from manypoints of view. for example.e. For example."Change"refers to alteration in the structure and function of the ecological mosaic through time. we are unlikely to discern the important elements of patch structure or dynamics unless we adopt an organism-centered view of the environment (165). Gardner et al (3~1) demonstrated that the number. and organisms amongthe component ecosystems. 150). Finally. 121.org/aronline Annu. 2. also dependon scale.annualreviews. numbers.g. and species in relation to the sizes.Annual Reviews www. that is. In Czechoslovakia. LANDSCAPEECOLOGY 173 ecology is well integrated into land-use planning and decision-making in Europe (e. function. 56. 140. Rev. the flow of energy. function. shapes. a dynamic landscape may exhibit a stable mosaic at one spatial scale but not at another.annualreviews. "Landscape" commonlyrefers to the landforms of a region in the aggregate (Webster’s NewCollegiate Dictionary 1980) or to the land surface and its associated habitats at scales of hectares to manysquare kilometers. 169).Observationsof landscape function. abiotic processes such as gas fluxes maybe controlled spatial heterogeneity that is not intuitively obviousnor visually apparent to a humanobserver. and shapes of patches in a landscape were dependentupon the linear dimensionof the map. sizes.if we are interested in a particular organism. The measurementof spatial pattern and heterogeneity is dependent uPonthe scale at which the measurements are made. For example. Consideration of Scale The effects of spatial and temporal scale must be considered in landscape ecology (e. 7. Landscapeecology is also developing along moretheoretical avenuesof research with an emphasison ecological processes (e. 93). 145.g. environmentalmosaics). Ecol. The scale at which studies are conducted may profoundly influence the . The scale at which humansperceive boundaries and patches in the landscape mayhave little relevance for numberous flows or fluxes.g. materials. landscape-level studies serve as a basis for determiningthe optimal uses of land across wholeregions (113). Similarly. MADISON" on 04/13/05. "Function" refers to the interactions betweenthe spatial elements. Three landscapecharacteristics useful to consider are structure. Becauselandscapes are spatially heterogeneous areas (i.

8. Landscape indexes derived from information theory (Table 1) have been applied in several landscape studies. 5.most of the variance in litter decomposition rates at local scales is explainedby properties of the litter and the decomposercommunity. Rev. This spatial patterning is a unique phenomenon that emergesat the landscape level (59). The spatial patterns observed in landscapes result from complexinteractions between physical. Most landscapes have been influenced by humanland use. Considerable progress in analyzing and interpreting changes in landscape structure has already been made(for detailed methodsand applications. and patchiness were calculated for a subalpine portion of YellowstoneNational Park and related to the fire history of the site since 1600(109. Ecol. 61. 174 TURNER conclusions: Processes and parameters important at one scale maynot be as important or predictive at another scale. The indexes developed by Romme(109) were adapted by Hoover (51) . Changesin landscape diversity were also hypothesizedto have effects on species diversity. 1989. and the nutrient content and productivity of aquatic systems (11o). habitat use by wildlife. Thus. The trends observed in the landscape pattern and the disturbance regimesuggested that YellowstonePark is a non-steady-state systemcharacterized by long-term cyclic changes in landscape composition and diversity. conclusionsor inferences regarding landscape patterns and processes must be drawnwith an acute awareness of scale. see 146. current approachesto the analysis of landscape structure are reviewed. and the resulting landscape mosaic is a mixture of natural and human-managed patches that vary in size.annualreviews. Seedlingmortality at local scales decreases with increasing precipitation.annualreviews. The distribution of oak seedlings also explaineddifferently at different scales (92). 28.whereas climatic variables explain most of the variance at regional scales (79. 29. 148). For example. In this section.20:171-197. and social forces. identify significant changesthrough time. Syst. OF WISCONSIN. For personal use only. Downloaded from arjournals. Table 1 reviews several methodsthat have been applied successfully in recent studies.Annual Reviews www.org/aronline Annu. and relate landscape patterns to ecological function. statistical approachesare reviewed in 149).g. shape. 80). whereasmortality at regional scales is lowest in the drier latitudes. biological. Indexes of landscape richness. CHARACTERIZING LANDSCAPE STRUCTURE Landscapestructure must be identified and quantified in meaningful ways before the interactions betweenlandscape patterns and ecological processes can be understood. and arrangement (e. MADISON" on 04/13/05.org by "UNIV. 110). Quantifying Landscape Patterns Quantitative methodsare required to comparedifferent landscapes. evenness.

with natural disturbances increasing landscape complexity. contagion.org by "UNIV. to compare the geometry different landscapes(61. whereas the diversity of plant species decreased. Fractal geometry(71. Results showed that landscape diversity increased southward from the mountains to the coastal plain. Landscapecomplexityhas not been shownto constant across a widerange of spatial scales (i. with the fractal dimensionand dominanceindexes reflecting broad-scalepattern and the contagionindex reflecting the fine-scale attributes that incorporatethe adjacencyof different habitats. The three indexes also appeared to provide information different scales.Annual Reviews www. as measured by the fractal dimension (61.it remains a focus of current research. 96.org/aronline Annu. Fractals have been used to compare simulated and actual landscapes (34. LANDSCAPEECOLOGY 175 applied to six study areas in Georgia. 1989. The fractal dimensionhas also beenhypothesizedto reflect the scale of the factors causing the pattern (61.g. Syst.(148). Landscape patterns in sites with relatively little humaninfluence were comparedalong a gradient from the mountains to the coastal plain. Applyingpredictions madeat one scale to other scales maybe difficult if landscapestructure varies with scale (84). 148). OF WISCONSIN. however. Thesecharacteristics maybe of particular importancefor species that require habitat patches of a minimum size or specific arrangement [e. and agricultural areas (96). The use of three complementarylandscape indexes (dominance. This type of scale sensitivity could prove useful in selecting measures of pattern that can be easily monitored through time (e.e. and to judge the relative benefits to gained by changingscales in a modelor data set (10). 148). The potential effects that the changesin patch structure created by forest clear- .annualreviews. This lack of constancyprobablyreflects the effects of processesthat operate at different scales. MADISON" on 04/13/05.a study that included human land-use patterns revealed a general trend of decreasing landscape diversity from the mountains to the coastal plain of Georgia. Landscapes influenced by natural rather than anthropogenic disturbances mayrespond differently. Rev. The size and distribution of patches in the landscape is another measureof landscape structure. 141). It has been suggested that human-influencedlandscapes exhibit simpler patterns than natural landscapes. 96. Ecol.g. self-similarity). This apparent contradiction illustrates the sensitivity of these indexes to the schemethat is used to classify the different componentsof the landscape. by means of remote sensing) and that can related to different processes. For personal use only. mountainforest. Downloaded from arjournals. the spotted owl (Strix occidentalis) in the Pacific northwest(41)]. However. and fractal dimension) in the eastern United States discriminated between major landscapetypes. 85. 85). such as urban coastal. which provide a measureof the complexityof the spatial patterns.annualreviews.20:171-197. 72) was introduced as a methodto study shapes that are partially correlated over manyscales. Shapes and boundaries in the landscape have been quantified by using fractals.

20:171-197. For personal use only.org/aronline 176 o x ?~ II II .org by "UNIV. Annual Reviews www. Downloaded from arjournals.Annu. OF WISCONSIN.annualreviews. Syst. Rev. Ecol. 1989. MADISON" on 04/13/05.annualreviews.

For personal use only. OF WISCONSIN.annualreviews.20:171-197.org by "UNIV.org/aronline 177 . MADISON" on 04/13/05. Ecol. Syst. Downloaded from arjournals.Annu.annualreviews. 1989. Annual Reviews www. Rev.

5. For personal use only. drainage.that is.annualreviews. Nearest neighbor probabilities reflect the degree of fragmentation in the landscape and. Ecol. Directionality in the landscape pattern. 62). The amountof edge between different landscape elements maybe important for the movement of organismsor materials across boundaries (e. can be measuredby calculating nearest neighbor probabilities both vertically and horizontally (or even diagonally). slope. which may reflect topographic or other physical constraints.an analysis might be conductedfor a 10.Annual Reviews www. 73. the subsequent deforestation that accompaniedhumansettlement was selective (126). can any serve as "early warning" signals? Answersto these and other questions are necessary for the developmentof broad-scale experiments and for the design of strategies to monitor landscape responses to global change. Rev. it maybe important to monitor changes in edges whenone quantifies spatial patterns and integrates pattern with function. woodlots)have also been studied biogeographic islands for both flora and fauna (e.e. especially fire. the complexity of patch boundaries. 27. Patch size and arrangement may also reflect environmentalfactors. grain and extent). MADISON" on 04/13/05. A variety of other techniques are available for quantifying landscape structure. Small patches of forest (i.g. 8.it important to note that the value of any measurementis a function of howthe landscape units were classified (e. Fine-scale measuresof adjacencypatterns and the directionality of individual cover types can be quantified by using nearest neighbor probabilities. what constitutes a significant change in landscape structure? Which measures best relate to ecological processes? Howdo the measurementsof pattern relate to the scale of the underlying processes? Whichmeasures of structure give the best indications of landscape change. land use categories vs successional stages) and the spatial scale of the analysis (e. Syst. Important questions remainabout landscape patterns and their changes. 47. 168). The quantitative measuresreviewed here could be easily applied to remotely sensed data. Thus. However.annualreviews. soil type. Therefore. 44. which would permit broad-scale monitoring of landscape changes.000-ha study site (extent) by using data with a resolution of 1 ha (grain). indirectly.g. Measurements of landscape pattern do not respond in the same way to changes in grain and extent. For example.g. both classification and scale must be carefully considered in analyses of landscape structure.g. and disturbance regime (126). 178 TURNER cutting patterns have on the persistence of interior and edge species were analyzed by Franklin & Forman(32). "Grain" refers to the level of spatial or temporal resolution within a data set.20:171-197. 163).org by "UNIV. The pattern of presettlement forests was closely related to topographyand the pattern of natural disturbances. 144.g. The size and isolation of forest patches in southern Wisconsinwere correlated with groups of environmental variables for example. and "extent" refers to the size or area of the study. 1989. and to data in a geographicinformation system(GIS). OF WISCONSIN. Downloaded from arjournals. .org/aronline Annu. For example. such as topographyor soil type. and the importanceof edge habitat for various species is well known(e.

128). Rev. The variability of soil organic carbon across the UScentral grasslands was studied through the use of a GIS model of macroclimate. Because most ecological modeling has focused on temporal changes. Yet. Downloaded from arjournals. and methodsare available to assess the error associated with the broad-scale applications (20). 55. Individual-based models incorporate the properties of individual organismsand the mechanismsby which they interact with their environment (52). 43. The JABOWA-FORET models used to predict forest succession are examples (4. In addition. Pastor & Post (101) combinedan individual-based model with a nutrient cycling model and demonstrated that the patterns of soil heterogeneity in the landscape had a strong influence on forest responses to global climatic change. decreased with temperature. 143). and functional modelsare being constructed. From a regional soils data base.20:171-197. the linking of models with geographic information systems and remote sensing technologies has begun (e. Soil organic carbon increased with precipitation.g. (b) transition probability models. For personal use only. Multiple simulations can be done with these modelsto represent a variety of environmental conditions in the landscape (9. socioeconomics) are difficult to represent mechanistically. spatial stimulation modeling is not yet well developed (16). and (c) process models.org/aronline LANDSCAPEECOLOGY Annu. MADISON" on 04/13/05. For example. 129-131.annualreviews. Transition models maybe particularly useful when factors causing landscape change (e. Individual-based models can be linked togetherspatially in a transect or grid-cell formatto represent a heterogeneouslandscape (e. and biotic responses into a grid-cell based spatial modelhas been used successfully to predict changes in a coastal landscape (132). regression . 9. Burkeet al (9) used a GIS develop a regional application of an ecosystemmodel.g. "Process-basedsimulation modelsare also being developed. 43) and human-dominated landscapes (e. 57).org by "UNIV.g. Syst. a modelthat combineshydrology.g.159). Simulation modelingwill continue to play an important role in predicting landscape changes and in developing our understanding of basic landscape dynamics. A review simulation modelingas applied to landscape ecology is beyond the scope of this article (see 133). The developmentof new computer architectures should facilitate the simulation of landscape dynamics(e. 12). nutrient dynamics. and was lowest in sandy soils. Three general classes of ecological modelare presently being applied in the prediction of changesin landscape structure: (a) individual-based models.g. manyopportunities nowexist for linking ecosystemmodels to geographic information systems to study landscape processes. Transition probability models have been used in a spatial frameworkto predict changing landscape pattern~ in natural (e. In a somewhatdifferent application. Predicting 179 Changes in Landscape Structure Modelsare necessary for landscape studies because experiments frequently cannot be performed at the ideal spatial or temporal scale. and managementstatus. 127). 1989. For example.annualreviews.g. Ecol. 141.Annual Reviews www. OF WISCONSIN. 50. soil texture. but somerecent developmentsare highlighted.

Percolation theory (98. 134) was used by Gardner et al (34) to develop neutral models of landscape patterns.annualreviews. Net primary production was driven primarily by precipitation and exhibited a linear relationship.annualreviews.g. and fractal dimension of clusters on these randommapsvary as a function of the size of the mapand the fraction of the landscape occupied by the habitat. at random.5928 for very large arrays). Methodsdeveloped from percolation theory provide a means of generating and analyzing patterns of twodimensional arrays. This section first reviews the use of neutral modelsto predict the effects of pattern or process and then examinescurrent research addressing ecological processes for which landscape pattern is important.Annual Reviews www. Neutral Models of Pattern and Process Anexpected pattern in the absence of a specific process has been termed a "neutral model" (13).e. achieving this goal mayrequire the extrapolation of results obtained from small-scale experiments to broad scales (e. Syst. This is analogous to generating a spatial pattern of sites occupied by a particular habitat.20:171-197.5928. Cluster characteristics change most rapidly near the critical probability. 1989. OF WISCONSIN. Rev.org by "UNIV. 140).A "cluster" (i. Pc. Ecol. such as forest or grassland. Thus. whichis the probability at whichthe largest cluster will "percolate" or connect the mapcontinuously from one side to the other (pc = 0. The need to understand the spatial relationships between driving variables and output variables was demonstrated. and responses were nonlinear. Thus.size distribution.The number. For personal use only. hypothetical animal restricted to a single habitat type might be expected to disperse successfully across a randomlandscape if the probability of occurrence of habitat exceeded 0. A two-dimensional percolating network within an rn by rn array is formed by randomlychoosing the occupation of the mz sites with probability p. This goal is difficult to accomplish.org/aronline 180 TURNER Annu. analysis was used to examinepredictive variables at different spatial scales. however.were driven by soil texture. for example. which are similar to maps of landscape patterns. patch) defined as a group of sites of similar type that have at least one edge in common.because the broad spatial-temporal scales involved make experimentation and hypothesis testing more challenging. Predictions of soil organic matter. RELATING LANDSCAPE PATTERNS ECOLOGICAL PROCESSES AND Elucidating the relationship between landscape pattern and ecological processes is a primary goal of ecological research on landscapes. MADISON" on 04/13/05. Downloaded from arjournals. . The use of neutral models in landscape ecology is promising approach for testing the relationship between landscape patterns and ecological processes (34). however.

109. severity. windstorms. 110. a low-level disturbance of individual pine trees (by lightning).g. or growth patterns expected in a steady-state forest (30).org by "UNIV. pathogens. 1989. community. or the physical environment"(103). including spatial distribution. or population structure and changes resources. Disturbances operate in a heterogeneous manner in the landscape-gradients of frequency. Under conditions favorable for the beetle (stressful conditions for the trees). in concert with previous history and edaphic conditions. 144). This result demonstratesthe need for a better understandingof the geographicrole of disturbance. 60.g. and synergism(e. For example. climatic.g. Neutral models can be used as a baseline from which to measure the improvementin predicting landscape patterns that can be achieved when topographic. It should be possible to determinesusceptibility to disturbance across the landscape.Annual Reviews www. frequency. Landscapesrespond to multiple disturbances. and type are often controlled by physical and vegetational features. species dominance.org/aronline LANDSCAPEECOLOGY 181 Annu.g. OF WISCONSIN. and fire) was controlled by slope position and aspect (30). disturbance effects changefrom physiological damageof an individual tree to the creation of forest patches (bark beetle spots) in which gap phase succession is initiated. such as the spatial distribution of wildlife (e. not only in NewEnglandbut elsewhere. Similarly. using a neutral modelapproach. intensity. predictability.g. area. Ecol. Ecological disturbance regimes can be described by a variety of characteristics. For example. rotation period. Downloaded from arjournals. leads to the vegetation mosaic observed in the landscape. or disturbance effects are included.20:171-197. substrate availability.lightning. 123).return interval.a study of the disturbance history of old-growth forests in NewEngland between 1905 and 1985 found that site susceptibility to frequent natural disturbances (e. maybe propagatedto the landscape level by bark beetles (115). 107. Rev. 162). For personal use only. MADISON" on 04/13/05. and the interactive effects of disturbances are importantbut difficult to predict (e. hierarchical structure. Syst. the beetle . 140). whereasyounger forests are least susceptible (109. mapswith known connectivity. For example. evidence was foundthat the last 350 years have provided the stability. or patterns of environmentalcharacteristics mightbe used. It is also possible to generatethe expectedpatterns of other ecological phenomena. Neutral models need not be restricted to purely randommaps.annualreviews. 114. matureconiferous forest stands in YellowstoneNational Park are generally most susceptible to fire. Foster has also shownthat wind damagein forest stands produces predictable patterns basedon the age of the trees (31). severity. In forested landscapes of the southeastern United States. 88). The differential exposure to disturbance. Landscape Heterogeneity and Disturbance The spread of disturbance across a landscape is an important ecological process that is influenced by spatial heterogeneity (e.annualreviews. With this propagation. Disturbance can be defined as "any relatively discrete event in time that disrupts ecosystem.

35). Other examplesof landscape heterogeneity impeding the spread of disturbance include pest outbreaks and erosional problems in agricultural landscapes.g. MADISON" on 04/13/05. If the disturbance is likely to movebetween communities. pine forests susceptible to bark beetle .high landscape heterogeneity should retard the spread of the disturbance. furthermore.g. the rate of disturbance propogation should be directly proportional to landscape heterogeneity for disturbances that spread between communities.g. For personal use only. the spread of a species-specific parasite through a forest). increased landscape heterogeneity due to "checkerboard"clear-cutting patterns enhances the susceptibility of old growthforest to catastrophic windthrow(32). and (b) those that cross boundaries and spread between different habitat types (e. in which disturbance is generally enhanced by homogeneity. heterogeneityin the spatial patterns of forest by age class tends to retard the spread of fires (e.but inversely proportional for disturbances that spread within the same community.org/aronline Annu. A comparison of the arctic landscape in 1949 and 1983demonstrated that indirect impacts of anthropogenic disturbances mayhave substantial time lags.20:171-197. This suggests a strong need for comprehensivelandscape planning through the use of current technologies (e. In forests of the Pacific Northwest. Estimation of the cumulative impacts of disturbances in a landscape is important for protecting sensitive habitats or environmentalquality. increased landscape heterogeneity should enhance the spread of disturbance. geographic information systems) to address such cumulative synergistic disturbance effects. Landscapeheterogeneity may also retard the spread of disturbance.annualreviews. fire spreading from a field to a forest). Rykiel et al (115) suggests that the bark beetles are amplifying the original disturbance of lightning strikes.org by "UNIV. Can the relationship between landscape heterogeneity and disturbance be generalized? Disturbances can be further characterized by their modeof propagation: (a) those that spread within the samehabitat type (e. the total area influenced by both direct and indirect effects can greatly exceed the area of planned development(155).g. Rev. In some coniferous forests. Furthermore. 182 TURNER populations can expandto becomean epidemicwith quite different effects on the landscape. Ecol. The spatial spread of disturbance maybe enhancedor retarded by landscape heterogeneity. 1989. If the disturbance is likely to propagate within a community. Anotherapproachto generalizing the spread of disturbance across a heterogeneouslandscape is to characterize the landscape in terms of habitat that is susceptible to the disturbance (e. the unusually close proximity of different habitats in the landscape appeared to enhance the disturbance effects that resulted from introduced ungulate grazers in mature maritime forest (144).g. Whetherlandscape heterogeneity enhancesor retards the spread of disturbance may depend on which of these two modes of propagation is dominant. Ona barrier island. Downloaded from arjournals. Syst.annualreviews.Annual Reviews www. OF WISCONSIN.

Syst. 166). and low connectivity (34). 67. and the sizes and numbersof clusters were not substantially affected by the intensity of disturbance. habitat fragmentation) have on the distribution. and (c) disturbanceintensity. The propagation of disturbance and the associated effects on landscape pattern were qualitatively different whenthe proportion of the landscape occupied by disturbance-susceptible habitat was above or beyond the percolation threshold (Pc). It has been sug- .and persistance of species.g. For personal use only. with numerous. Ecol.g. Movement and Persistence of Organisms The spatial patterns of biological diversity have long been of concern in ecology (e.partly because we often tend to study single disturbances in small areas rather than multiple disturbances in whole landscapes. and biogeographical studies have examinedthe regional abundanceand distribution patterns of manyspecies (e. Landscapeecological studies focus on the effects that spatial patterning and changesin landscape structure (e. The interactive effects of disturbances are not well known. Disturbances operate at manyscales simultaneously. suggesting that well-vegetated fencerows may provide interconnections between patches of suitable habitat (158). Habitats occupyingmorethe Pc tended to be highly connected. hardwoodforest. neutral modelapproach can then be used to provide predictions of the spread of disturbance that can be tested against observations. birds and small mammalsin an agricultural landscape use fencerows betweenwoodlots more than they travel across open fields. Landscapeconnectivity maybe quite important for species persistence. and disturbances spread through the landscape even when frequency was relatively low. Downloaded from arjournals. Habitats occupyingless than Pc tended to be fragmented. and their interactions contribute to the observed landscape mosaic. grassland. as by Turner et al (147). the probability that a disturbance. (b) disturbance frequency. The landscape can be considered as a mosaic of habitat patches and interconnections.org by "UNIV. 1989.Annual Reviews www. once initiated.g. wouldspread to an adjacent site. 92). Rev. 68. A better understanding of howdisturbance regimes vary through time and space is needed.org/aronline Annu. MADISON" on 04/13/05. Natural disturbances are likely to vary with a changing global environment.g. movement. and altered disturbance frequency or intensity maybe the proximal cause of substantial changes in the landscape. etc). The spread of a disturbance was constrained by this fragmentedspatial pattern. OF WISCONSIN. The relationship between landscape pattern and disturbance regimes must be studied further.annualreviews. small patches. pine forest that is too youngto be infested. Disturbance was simulated as a function of (a) the proportion of landscape occupiedby habitat susceptible to the disturbance. forming continuous clusters (34).20:171-197. For example. LANDSCAPEECOLOGY 183 infestations) and habitat that is not susceptible to the disturbance (e.the probability of disturbanceinitiation. particularly in light of potential global climatic change.annualreviews.

including soil type.annualreviews. species in isolated patches should have lower probability of persistence. whereassimilar but isolated forests were not (69). and woodyspecies composition. Furthermore.20:171-197.org by "UNIV. 1989.annualreviews. ants . Local extinctions of small mammalsfrom individual forest patches were readily recolonized by animals from other patches whenfencerows were present (49). because the survival of populations in a landscape dependson both the rate of local extinctions (in patches) and the rate of organism movementamongpatches (22). MADISON" on 04/13/05. Within a neutral model framework. regardless of networkconfiguration (24). Theoretical approaches are being developed to identify scale-dependent patterns of resource utilization by organisms on a landscape. passing through riparian areas frequently used by grizzlies. Downloaded from arjournals. whoexaminedthe effects of landscape fragmentation on the wintering areas of white-tailed deer (Odocoileusvirginianus). Simulation of numerous possible network configurations further showedthat one linkage with another patch accounted for most of the variance in survival and that more than two linkages had no significant effects.750 contiguous 0. Bears used habitat within 100 of roads significantly less than expected. 184 TURNER gested that. The effects of road 2 area of the Rocky development on grizzly bear movementswithin a 274-km Mountainswere studied for seven years (75).7% loss of habitat. canopyclosure. OF WISCONSIN. model was developed by using Bayesian probabilities conditional on 12 landscape variables. This approach mayallow the connectivity of a landscape to be described for a variety of species. Fahrig &Merriam(24) demonstratedthat the survival of populations in individual woodypatches was enhanced when patches had more corridors connecting to other patches. but the resource must occur with high probability along the path. the effects of patch isolation were studied by Milne et al (88).Annual Reviews www. Everypoint need not contain a critical resource.4-ha locations. Syst.g. The path length will vary for different organisms (e. Another study reported that small forest patches connected by corridor to a nearbyforest systemwerecharacterized by typical forest interior avifauna. The minimal requirement is that organismsbe able to moveacross a landscape in a path of length n with a high probability of locating a resource. Modifications of habitat connectivity or patch sizes can have strong influences on species abundanceand movementpatterns. For personal use only. In simulations and field studies. Rev. Several studies support this idea. Because roads often followed valley bottoms. the road development represented approximately an 8. Ecol.org/aronline Annu. Deer habitat was predicted independently at each of 22. Minimalscales for resource utilization were predicted by O’Neill et al (97) by considering the spatial distribution of resources. avoidance of roads was independent of traffic volume. Comparisonof the predictions of the neutral model with observed habitat-use data demonstrated that sites containing suitable habitat but isolated from other suitable patches were not used by the deer (88).

use was indiscriminate with respect to size. A similar effect of landscape heterogeneity on cyclic populations of Microtuswas also suggested by Hansson (45). Van Dorp & Opdam(152) concluded that the distribution forest birds in a landscape results from a combination of dispersal flow. From their studies in the Netherlands. Rev. suggesting that regional conservation strategies should maximizeboth patch size and forest heterogeneity (33). shape. The effect of . Minesnear concaveforest boundaries had 2. If resources are clumped. resources spanning the landscape) at lower values of p.org by "UNIV. The interaction between dispersal processes and landscape pattern influences the temporal dynamicsof populations. or straight. patches of suitable habitat may provide refuges from predators during population crashes.organismsmust adjust their scale of resource utilization and operate at larger scales in order to movefrom one resource patch to another. OF WISCONSIN. Downloaded from arjournals. resistance of the landscape(i. in the cyclic northerly populations.Annual Reviews www. such as birth rate and death rate. 1989.annualreviews.In a study of forest fragmentsin an agricultural landscape. For example. Wolff(170) suggested that southerly populations of the snowshoehare (Lepus americanus) maynot be cyclical because of habitat discontinuities resulting from the wide spacing of suitable habitat patches.e. Local populations of organismswith large dispersal distances maynot be as strongly affected by the spatial arrangementof habitat patches. Another study demonstrated that revegetation patterns on reclaimed strip mines in Maryland and West Virginia differed. For personal use only. larger and more heterogeneous forests had more species and bird pairs. protecting the local populations from extinction. Syst. overall patch use was shifted t6~vard the larger patches (167). Nonrandomuse of patches by shrubsteppe birds was reported by Wiens(167). Linear corridors stretching across the landscape wouldpermit percolation (i. whichprevents interpatch dispersal. Studies of patch characteristics and use by two sparrows (Amphispizabelli and Spizella breweri) suggested that the birds may select relatively large patches for foraging.20:171-197. In contrast. MADISON" on 04/13/05. The size. LANDSCAPEECOLOGY 185 and antelope wouldhave different scales of resource utilization). Woodlotsize was also found to be the best single predictor of bird species richness in the Netherlands (152).annualreviews. and populationcharacteristics. moreof the variance in the richness of woodyplant species on peninsulas in Mainewas explained by sample position in relation to the base of the peninsula than by distance from mainland (87). depending on whether the adjacent forest boundarywas convex.5 times more colonizing stems and greater evidence of browsing than mines adjacent to convex forest boundaries (46). governedby local and regional patch density. Ecol.org/aronline Annu. In areas containing large patches.e. concave. The shape of patch mayalso influence patterns of species diversity within the patch. and diversity of patches also influence patterns of species abundance. but where smaller patches predominated. barrier effects).

20:171-197. 105). Helens. few studies have ad- . then the spatial arrangement of habitat will have less effect on population dynamics. Experimentalapproaches (e. or urban development (e.g. The effect that the spatial structure of habitats has on populations is also a focus of conservation biology. Results suggested that if an organism disperses along corridors. Furthermore. MADISON" on 04/13/05. Downloaded from arjournals. although input-output studies of whole ecosystems and watersheds have been extensive. whereas extinction. the organism disperses large distances in randomdirections from patches and does not detect patches from a distance. species richness increased with habitat subdivision.annualreviews. 3.org by "UNIV. Rev. 186 TURNER spatial arrangementof host-plant patches on the local abundanceof cabbage butterfly (Pieris rapae) was studied by Fahrig &Paloheimo(25) through the use of models and field experiments. Regional-scalestudies of the dominancepatterns of six native grass species in the central UnitedStates suggestedthat the spatial patterns of these grasses were limited primarily by dispersal processes or resistance barriers caused by competition from other grasses (6).Annual Reviews www.org/aronline Annu. in an experimentally fragmented California winter grassland. In an urban habitat. For example. forestry. Graphic and geographic migration models were used to examine the relationship between present dominancepatterns and presumedsource areas for the six species. For example. However. 108) wouldbe extremely valuable in studies of landscape heterogeneity and species persistence. and turnover rates were relatively independent of habitat subdivision (108). a blending of concepts developed in conservation biology and landscape ecology could yield muchinsight into these issues (e. The spatial patterns supported a migrating-wave hypothesis of grass species dominanceand did not support the idea that grass species distributions were controlled primarily by climatic factors. then the spatial relationships betweenhabitat patches are important. For personal use only. These results contrast with predictions that habitat subdivision necessarily results in greater rates of extinction. Syst. dispersal distance) did not correlate with either seed abundanceor plant density in revegetated study sites. Dickman(23) found that two small patches retained more species than one large patch of equal area. Redistribution of Matter and Nutrients The redistribution of matter and nutrients across heterogeneouslandscape is not well known. it is well knownthat increased nutrient loadings in water bodies can result fromagricultural~practices. If. immigration. however. OF WISCONSIN. In a study of revegetation of debris avalanches on Mount St.g.annualreviews. Ecol. Dale (18) reported that absolute distance to a seed source (i. 1989.g. Results also suggestedthat the Plains grasses are probably not yet in equilibrium with their environment. 160).e. It remains a challenge to predict quantitatively the dynamicdistribution of a species from the spatial arrangement of habitat patches and the landscape structure of the surrounding region.

and plants potentially alter gas flux across the landscape (40). . Landscapeposition also influences redistribution processes. the coupling of natural and managedsystems within a watershed mayreduce non-point-source pollution (102). Large animals are important because they typically graze (and removenutrients) from patches containing high-quality forage and mayreturn nutrients (by meansof defecation) to areas in whichthey rest or sleep. For personal use only. although there is increasing recognition that such influence is important (e. 118. Characteristics of water quality can vary with a lake’s position in the landscape. 125. Research has shownthat riparian forests reduce sediment and nutrient loads in surface runoff (64. nitrogen and phosphorus) in surface runoff and shallow groundwaterin an agricultural watershed that contained both cropland and riparian forest. Kesner (57) used grid-cell modelto study the spatial variability in the loss. Rev. 122. The horizontal flow of nutrients or sediment in surface waters of humanmodified landscapes maybe affected by spatial patterning.Annual Reviews www. MADISON" on 04/13/05. Syst. New technologies such as Long-path Fourier-Transform Infrared Spectroscopy (F’I~IR) offer r~ew. 1989. 119). whereasthe riparian habitats were removingnitrogen from the surface flow. However. Total nitrogen output (kg/ha) wassubtracted fromtotal nitrogen input for each cell in a geographicinformation system (GIS). 76-78.annualreviews. Landforms such as sedimentdeposits or landslide areas influence the temporaland spatial patterns of material fluxes carried across landscapes by surface water (137). LANDSCAPEECOLOGY 187 dressed the influence that spatial pattern mayhave upon the flow of matter and nutrients.org by "UNIV.20:171-197. The flux of gases between the atmosphereand the biota maybe influenced by spatial heterogeneity. gain. Ecol. microbes.g.g. 124. 171). Lakeslower in the landscape had a higher specific conductance because their surface or groundwater supplies passed through more of the vegetation and soils. Nutrient removalis significant to receiving waters.annualreviews.research has not explicitly addressed the effects that different spatial arrangementsof habitat have on nutrient transport by grazers. For example. accumulatinga greater concentration of dissolved matedal. 39). and scale-dependent processes (40). Downloaded from arjournals. and storage of total nitrogen across an agricultural landscape.org/aronline Annu. potential patterning of biological processes.. Peterjohn & Correll (102) studied concentrations of nutrients (carbon. OF WISCONSIN. powerful methodsto study fluxes between ecosystems. Their study demonstrated that nutrient removalhad occurred in the riparian forest. as demonstrated in the Colorado alpine zone (11) and in Wisconsin forests (70). Nutrients can be transported by grazing animals across landscapes and between patches (e. Results indicated that upland agricultural areas were exporting nitrogen to the surface flow. The source-sink relationship between soils.

65. MADISON" on 04/13/05. for a constant frameof reference. and an interaction term. primary production. Schimelet al (117) demonstratedthat the spatial pattern of soil and forage properties influences cattle behavior and hence urine deposition in grasslands. 66).g. however. makinglarge-scale estimates of nitrogen loss challenging. samplingcannot be done at the appropriate spatial scale. Regionaltrends in soil organic matters across 24 grassland locations in the Great Plains have also been predicted by using a few site-specific variables: temperature. For example. soil texture. 139). but NPP varied amongland uses and across physiographic regions. Ecol.annualreviews. and studies mayneed to rely on data collected for other purposes. Becausethe spatial heterogeneity of manyecosystemprocesses is not well known.Annual Reviews www.g. The first method.4 t/ha during the period from 1935 to 1982 (in comparisonwith a potential natural productivity of ~ 16-18 t/ha).the extrapolation of site-specific measurementsto regional scales is difficult. Estimates of forest canopyETthat are based on data from the ThermalInfrared Multispectral Scanner (TIMS) compared well with estimates made through energy balance techniques (65). soil water-holdingcapacity.5 to 6. ANPPwas explainedby annual precipitation. Sala et al concludedthat. and biogeochemistry.org/aronline 188 TURNER Annu. Ecosystem Processes at the Landscape Level Landscape-level estimates of ecosystem processes (e. This change in the ability of particular variables to explain variability as the spatial scale changes has also been demonstrated for other processes. a model will need to include a large numberof variables to account for the pattern of the same process as the scale of analysis becomesfiner. Sala et al (116) demonstrated that the regional spatial pattern of abovegroundnet primary production (ANPP)in the grasslands region of the United States reflected the east-west gradient in annual precipitation. OF WISCONSIN.g. evapotranspiration. 80) and evapotranspiration (54).org by "UNIV. NPPhas also been extensively studied at regional-scales through the use of remotesensing technology(e. it is important to note that remotesensing technologyoffers considerable promisefor the estimation of other ecological processes at broad scales. Turner (142) used agricultural and forestry statistics to estimate net primary production (NPP)of the Georgia landscape over a 50-year interval. For personal use only.annualreviews. For example. a simple . Accordingto her study. Several recent studies have attempted to examinescale-dependent patterns of productivity. and nitrogen inputs (100). water balance. 1989.evapotranspiration (ET) from forested landscapes can be estimated from remotely sensed data (e. and decomposition) that are influenced by spatial heterogeneity are difficult to obtain. NPPof the Georgia landscape increased from 2. Kinget al (58) tested two methodsof extrapolating site-specific modelsof seasonal terrestrial carbon dynamicsto the biome level. such as decomposition(79.20:171-197. Frequently. Althougha review of this literature beyondthe scope of this article. Syst. Rev. moisture. Downloaded from arjournals. plant lignin content. At the local scale.

g. Syst.org by "UNIV. and in resource managementdecisions. Results from landscape ecological studies strongly suggest that a broad-scale perspective incorporating spatial relationships is a necessary part of land-use planning. Predictions were based on the mathematical expectation of simulated sitespecific exchanges for each region times the area of the region. Ecologists. (c) the frequencydistribution of modelparameters variables that vary across the region and define the heterogeneity of the region. producing more reasonable results. Manyland management activities (e.annualreviews. CONCLUSION Spatial pattern has been shownto influence manyprocesses that are ecologically important. the rarity criterion (for species management) maybe most appropriately applied at regional/global scales (see also 120). forestry practices. The long-term maintenance of biological diversity mayrequire a managementstrategy that places regional biogeographyand landscape patterns above local concerns (93).annualreviews. MADISON" on 04/13/05. Noss&Harris (94) present conceptual schemethat evaluates not only habitat context within protected areas but also the landscape context in which each preserve exists. Four main ingredients were required to extrapolate the site-specific modelsacross heterogeneousregions: (a) the local site-specific model. in decisions about the creation or protection of sustainable landscapes. the effects of pattern on process must be considered in future ecological studies. the landscape (like manyecological systems) represents an interface betweensocial and environmental processes. Therefore. OF WISCONSIN.org/aronline Annu. edaphic. regional planning. Downloaded from arjournals. (b) designation of larger region of interest. Althoughthis review has selectively emphasizedthe effects of spatial patterns on ecological processes. so theory development and empirical testing should continue. Methodssuch as those developedby King et al (58) showpromise for dealing with this difficult problem. was not adequate for biome-level predictions. Rev. and climatic patterns within a biome. 1989. The problemof extrapolation of site-specific measurements to obtain regional estimates of ecological processes remains a challenge.and planners have traditionally ignored interactions betweenthe different elements in a landscape--the elements are usually treated as different systems. For personal use only. and natural resource development)involve decisions that alter landscape patterns.Annual Reviews www. The second methodexplicitly i~ncorporated spatial heterogeneity in the abiotic variables that drive carbon dynamics. Ecol. A workingmethod for landscape planning was presented by Steiner & Osterman (136) and applied to a case study of soil erosion.20:171-197. There . particularly at broadscales. and (d) a procedurefor calculating the expected value of the model. land managers. LANDSCAPEECOLOGY 189 extrapolation that assumed homogeneityin biotic. for example. Withregional diversity and ecological integrity as the goal.

a slight change near a critical threshold can have dramatic consequencesfor the persistence of the population.20:171-197. Habitat fragmentation mayprogress with little effect on a population until the critical pathwaysof connectivity are disrupted. For example. 1989. Rev. processes. 190 TURNER remains a tremendouspotential (and a necessity) for truly interdisciplinary cooperation amongecologists. large tracts of forest can be easily fragmentedand qualitatively changedby disturbances of low to moderate intensity and low to high frequency. evenat high intensities of disturbancepropagation. then. this change mayhave important implications for species persistance. but effects maybe counterintuitive for intermediate levels of frequency and intensity.if there is insufficient landscape connectivity. OF WISCONSIN. managementshould focus on the frequency of disturbance initiation. Newinsights into ecological dynamicshave emergedfrom landscape studies and have led to hypothesesthat can be tested in a diversity of systems and at manyscales. The broad-scale indexes of landscape structure mayprovide an appropriate metric for monitoring regional ecological changes. Suchan application is of particu- .annualreviews.If a habitat type is rare (e. Franklin & Forman(32) presented a convincing argumentfor considering the ecological effects of spatial patterns of forest cutting patterns.annualreviews. In contrast. A threshold level of habitat connectivity maydemarcatedifferent sorts of processes or phenomena. Hypotheses regarding the existence and effects of critical thresholds in spatial patterns should be tested through the use of a diversity of landscapes. Landscape theory may have direct applications to the managementof disturbance-prone landscapes.Annual Reviews www.The numberor length of edges in a landscape changes rapidly near the critical threshold (34).org by "UNIV. the spread of disturbance across a landscape maybe controlled by disturbance frequency whenthe habitat is below the critical threshold. landscape planners. If a habitat type is common.org/aronline Annu.g. Several studies have suggested that the landscape has critical thresholds at which ecological processes will change qualitatively. Similarly. and resource managers to develop an integrated approach to landscape management. geographers. granite outcrops and remnant forests). The relationships betweenindexes. Ecol. MADISON" on 04/13/05. The theoretical studies conductedby Turner et al (147) also have implications for landscape management. Downloaded from arjournals. disturbances with low frequencies may havelittle impact. For personal use only. high frequencies of disturbance initiation can substantially changelandscape structure. Syst.managementmust consider both frequency and intensity. but it maybe controlled by disturbance intensity whenthe habitat is abovethe critical threshold. processes. and scale needs morestudy to understand(a) the factors that create pattern and (b) the ecological effects of changingpatterns on processes. Current research suggests that different landscape indexes mayreflect processes operating at different scales. and scales. The effects of disturbance can be predicted at the extremeends of the ranges of frequency and intensity.

and R. R.g. The identification of instances in whichspatial heterogeneity can be ignored is as importantas the identification of the effects of spatial pattern. Rev. Natural experiments. or ecosystemprocesses such as NPPor the distribution of soil organic matter.and scales at which spatial heterogeneity has a significant influence. A better understanding of the parameters necessary to predict patterns at different scales is necessary. Theseobservations could simplify the prediction of landscape dynamicsif a significant amountof fine-scale variation can be incorporated into a few parameters. under Contract no. Microcosmsor mesocosmsin which spatial pattern can be controlled by the experimentermayalso prove useful. such as disturbances that occur over large areas or regional development . Several studies suggest that.org/aronline Annu. The relative importance of parameters controlling ecological processes appears to vary with spatial scale. improved this manuscript. an entire river drainage basin).g.g. V. in reponse to global change) can be measured with remote-sensing technology. MADISON" on 04/13/05. Forrnan.the effect of landscape heterogeneity on the redistribution of materials is not well known.20:171-197. and an understanding of the pattern-process relationship will allow functional changes to be inferred. DE-AC05-84OR21400 with Martin Marietta Energy Systems. W.annualreviews. ACKNOWLEDGMENTS The commentsand suggestions of V. Milne. 1989. The spatial patterning of habitats maybe important to predict nutrient distribution in landscapes of small extent (e. OF WISCONSIN. Neutral modelsof various types will continue to be helpful in the identification of significant effects of spatial patterns. Gardner. and I sincerely thank them for their thoughtful reviews.. LANDSCAPEECOLOGY 191 lar importancebecause changes in bi~oad-scale patterns (e. US Department of Energy. A few variables maybe adequate to predict landscape patterns.org by "UNIV. For personal use only. Dale. Theoretical and empirical workshould progress jointly. phenomena. H. For example. T. It is important to identify the processes. the spread of disturbances. Funding was provided by the Ecological Research Division. Downloaded from arjournals. Inc. only a few variables maybe required to predict landscape patterns. T. B. King. at the landscapelevel. A. the watershed of a lower-order stream) but less important as extent increases (e. Future research should be oriented toward testing hypotheses in actual landscapes. O’Neill. but the broad-scale nature of manylandscape questions requires creative solutions to experimentaldesign. and by an . Syst. T. R. H. Office of Health and Environmental Research. perhaps through an iterative sequence of model and field experiments. Ecol. Of paramount importanceis the developmentand testing of a general bodyof theory relating pattern and process at a variety of spatial and temporal scales. Methodsfor characterizing landscape structure and predicting changes are nowavailable.Annual Reviews www.annualreviews. also provide opportunities for hypothesis testing.

Hierarchy. Munich 8. Ecology 69:468-75 26. 1979. Wisconsin Press 18. ORNL/TM-7759.. Likens. 1959. E. Onthe survival of populations in a heterogeneousand variable environment. Oak Ridge. F. B. Schimel. Principles of Geographic Information . CarnegieInst. Elevational contrasts in comtemporary geomorphic activity in the Colorado Front Range. T. J. Merriam. 1. Ecol. Sklar. A quantitative analysis of forest island pattern in selected Ohio landscapes. Am. D. Constitution dons le regne vegetal de groupes physiologiques applicables a la geographie ancienne et moderne. 1986. Bd. 1976. Ann. Publication No. E. H. In press 10. 1988.. J. Curtis. A. T. eds. Carpatho-Balcanico 18:531 12. 27:3-22 13. Landscape Ecol. 1981. Syst. A. For personal use only. H. L. Turner. 24:337-51 24. Pattern and Process in a Forested Ecosystem. Gersmehl. H.. 1985. Helens debris avalanche. Janak. Migration models for grasses in the American mid-continent.annualreviews. Downloaded from arjournals. ORNL. Burgess. No. D. Gardner. V.. Conf. Caswell. Baker. 1981. Landscape ecology and nature reserve design in the Boundary Waters Canoe Area. Assessing regional impacts of growth declines using a forest succession model. Habitat patch connectivity and population survival. Tenn.20:171-197. F. 1874. Fahrig. Engelhart. Regional modeling of grassland biogeochemistry using GIS. L. Bormann. Bowen. Bot.. C. Nature and structure of the climax. G: 1985.. Sharpe.. Caine. Communitystructure: a neutral model analysis. 1976. Oecologica 50:39-53 23. H. Oecologia26:1-8 . Assoc. S. 1968. R.C. NewYork: Springer-Verlag 4. J. 1987. Ecol. A. Emergingdirections in landscape ecology and applications in natural resource managment. 1988. 1985. Forman. Start.. R. Fahrig. Rev. Botkin. L. Parallel and vector processing in landscape dynamics. Studia Geomorphol. 60:849-72 5. Grundlagen. W. Gardner.. Using sensitivity and uncertainty analyses to improve predictions of broad-scale tbrest development. E. Forest size and avian diversity in NewJersey woodlots with some land use implications. Clements. 1936. Wallis. Ecology 70:23-35 3. and effectiveness of mathematical models: a review of freshwater wetland applications. on Sci. 1989. Someecological consequences of a computermodelof forest growth. J. Dale. H. Bostedt-Craig. Forest Island Dynamics in ManDominated Landscapes. Ecol. administered by OakRidge Associated Universities. D. F. Hen’mann. Habitat fragmentation and vertebrate species richness in an urban environment. Modelling 42:165-78 21. Math. T. Joyce. A.. Annu. R. H. Clements. W. Dale. N. R. Forman. MADISON" on 04/13/05. 1989. H. Publ. H. L. 24:252-84 16. T. J. M. Ecol. Plant succession: an analysis of the developmentof vegetation. Monogr. G. Ecology 66:1762-68 25. The Vegetation of Wisconsin: An Ordination of Plant Communities. Buchwald. Environ. J. A. to M. 1986.Systems for Land Resources Assessment. Burgess. Allen. D. Ecol. 1981. F.. R. L. T. Chicago: Univ. J. A. Effect of spatial arrangementof habitat patches on local population size. P.. Appl. Wash. Dickman. T. Can. L. 1984. W. A. Rep. F. A. 3317 of the EnvironmentalSciences Divison. Jager. R. V. G. Oak Ridge Natl Lab. J. H. pp. Geneva: Archives des Science Physiques et Naturelles 22. 1982. accuracy.. H. 1987. Costanza.. E. Parton. G. P. DeCanclolle. J. Washington. 59-88. 1916.. BLVVerlagsgesellschaft. Modelling 27:45-68 17. 6. Chicago Press 2. D. Literature Cited 1. F. R. DenBoer. T. R.Annual Reviews www. Geogr. V. 46:327-54 14. Madison: Univ. B. C. W. M. Casey.. Natl. eds. In press 19. 75:383-94 7. E. ed.. M.. Winddispersed seeds and plant recovery on the Mount St... J. Ecol. Rosen.org/aronline 192 TURNER AlexanderHollaenderDistinguished Postdoctoral Fellowship. 1988. OF WISCONSIN. J. F. J. P. 242 15. Ecol. Jameson. New York: Springer-Verlag 9. 1989. P. Yonker. Minnesota. Oxford: Clarendon 11. C.T. Handbuch]~r landschaftpflege und naturschutz. Articulation. R. DC: The George Wright Society 27. R. 1972.org by "UNIV. 1989. Comput.In Proc. 24:83-93 20. Burrough. I.. Brown. I. Burke. K.annualreviews.. Manage.. Parks. Dale. Galli. Appl. Paloheimo. Lock.. R.

1986. R.. DeAngelis.. Post. Merdam. F.. NJ: Prentice-Hall 54. Gleason. pp. Locational Analysis in Geography. Wegner. 1987.20:171-197.. N. Patch-within-patch restoration of man-modifiedlandscapes within Texas state parks. For. Importance of area and habitat heterogeneity to bird assemblagesin temperate forest fragments..H. R. Creating landscape patterns by forest cutting: ecological consequences and principles. W. Sharpe. D. 1987. Godron. Pace. 1971. Freemark.. MSThesis. Conserv. OF WISCONSIN. 1988. Huston. geography. R. No. J. Serotiny. Neutral models for the analysis of broad-scale landscape pattern. Biogeochemistry 2:101-12 40. Hansen. T. P. Stomatalcontrol of transpiration: scaling up from leaf to region. Landscape Ecol. M. Comparativestructure of landscapes across physiographic regions of Georgia.. A. M...C. M. Sta. Int. Foster. Hardt. G.. Chicago: Univ. L. Naiman. Gardner. A.682-91 53. 1987. W. T. Chicago Press 48. 1986.. T. EnglewoodCliffs. USA. Species and stand response to catastrophic windin central New England. Post. Rep. D. Isard. D. D. T.. Gleason. MSThesis.. 1988. W. T. Res. LANDSCAPE 28. Introduction to Regional Science. Ecol.. Portland. 1:19-28 35. Landscape Ecol. A. E. Pacific NWFor. Sci. 1988. Biol. W. Risser. 1985.. communityorganization and vegetation dynamics of the old-growth Pisgah Forest. Midl. ECOLOGY 193 GeoScience Remote Sensing Syrup. Athens 52. 1:5-18 33. Henderson. H. Carey. In pres 47. J. See Ref. F. Botkin. Spec. Manage. Hoover. Onthe importance of landscape heterogeneity in northern regions for the breeding population densities of homeotherms:a general hypothesis. Boundary form effects on woodycolonization of reclaimed surface mines. G. A. 473-82. G. T. G. Adv. Newcomputer models unify ecological theory. Ecotones: what and why?In A New Look at Ecotones. New York: Wiley 43. Gleason. D. M. 1986. Landscape Ecology. Biogeochemistry research needs: observations from the ecosystem studies program of The National Science Foundation. D.. Forman. J. OR 42. Ecol. The structure and development of the plant association. M. Bull. Hoekstra. Long-path I~TIR measurement of atmospheric trace gas concentrations. Forman. 173-98 49.. V. Holland. Joyce. L. Goetz. King.. Torrey Bot. Jarvis. Gosz. Ananalysis of forest dynamics in the north Georgia piedmont. Ecology and Managementof the Spotted Owl in the Pacific Northwest. 53:7-26 38. Conserv. 1987... Issue 17:9-46 45. R. T. G. M. III 1987. 1975. R. Harris. G. D. Athens 58. 1981. Hall. pp. 1984. M. T. Downloaded from arjournals. 22:307-22 56. Ecol. D. 76:105-34 31.annualreviews. Landscape pattern and successional dynamicsin the boreal forest.. Regional multiresource models in a national framework. and fire in the Pine Barrens of New Jersey. Oikos 33:182-89 46. 140. Am. M. NewYork: Wiley 30. R. L. 1989. DiCastri. D. Univ Georgia.. Forman. F. Syst. ed. P. M. Term. Mich. Kesner. J. Club 43:463-81 37. PNW-185. The individualistic conceptof the plant association. Givnish. J. A. 1981. The seasonal exchange of carbon dioxide between the atmosphere ... R. W. Ecology 69:1326-30 41. southwestern New Hampshire. T. BioScience 38. Club. 1984. R.. Evolution 35:101-23 36. J. Riskind. H. H. W. R. D. Range Exp. Gosz. Disturbance history. J. 15:1-49 55. A. The individualist concept of the plant association. Gen. Ecol. R. The Fragmented Forest. McNaughton. 1988. USA. 1917. B. K. Rep.. Proc.annualreviews. F. R. USDAForest Service. Univ. Milne.. The geography of nitrogen in an agricultural watershed:a technique for the spatial accounting of nutrient dynamics. Torrey Bot. 10:761-71 57. ORNL-IBP-71-8. Patchy environments and species survival: chipmunksin an agricultural mosaic. B. Woods. L. Georgia. Merriam. Franklin. DeAngelis.. T. A. 1986. 51. P.. 1979. Bull. K. H. C. Alig. Turner. Ecology. 44. J. Environ. S. T. Dahm. Foster. H. 31:95105 50. Hayes. R. W.. Oak Ridge Natl Lab. T. Patches and structural componentsfor a landscape ecology. Cliff. 76:13551 32. Int. A. MADISON" on 04/13/05. Forman.. 1939. O’Neill. Hett.. Haggett. 1986.R.K. D. 1976. G. Rev. Landuse changesin east Tennessee and a simulation modelwhich describes these changes for 3 counties. S.. Biol. A. A.Annual Reviews www. J. Natl. Di Castri. Johnson. BioScience 31:73340 29. J. Hansen. B. 1985. J. Tech. A. R. For personal use only. 1988. J. B.. Gutierrez. 21:92-110 39. T. Go&on. Biol. J.org by "UNIV. J. Hansson. R. 1987. Ann Arbor. Frey. eds. 1977. Oak Ridge. T. D. 1926. 31:95105 34. L. E. Strebel.org/aronline Annu.

M. H. DC 84. McNaughton. 194 TURNER and the terrestrial biosphere! extrapolation from site-specific modelsto regional models. Ecol. 1987.org by "UNIV. See Ref. Ecology of a grazing ecosystem: the Serengeti. Rev. R. Box.. 1989. G. Milne. Meentemeyer. 1983. P. San Francisco: Freeman 72.. Forman. J. S. In press 87. D. B. M. Downloaded from arjournals. R. R. C. Holbo. T. Kratz. Forman. lightning. and environmentalpattern. Meentemeyer. 1985. Krummel. In Serengeti: Dynamics of an Ecosystem. Appl. Norton-Griffiths. R. Wilson. H. Johnson. Chicago Press 77. J. Lab. 1988. Appl. T. McCoy. Limnol. Coleman. Evidencefor the value of corridors and minimization of isolation in preservation of biotic diversity. Scale-dependent proximity of wildlife habitat in a spatially- . Pickett. K. Whitcomb. In press 67. S. habitat use and demography. R. Fracmls. 1984. O’Neill. B. H.J. McIntosh. J. Graphic Display of Two. The Fractal Geometry of Nature. distance. D. J. S. 10. Gardner. Heterogeneity as a multi-scale characteristic of landscapes. R. V. pp. Todd. MacArthur. Verh. B. Milne.annualreviews. O’Neill. Long-term ecological research (LTER)on north temperate lakes of the United States. 1984. 15-34 82. Meentemeyer. O. O. Press 69. Milne. B. Bowser. J. Scribners 63. Math. Landscape patterns in a disturbed environment. MacArthur. R. H. Lin.. San Francisco: Freeman 73. NewYork: Springer-Vedag. R. See Ref. Magnuson. V. 59. Grasslandherbivore dynamics. L. A. MacClintock. Form.. E. Cambridge: Cambridge Univ. V. A theoretical approachto regional environmental conflicts. B. C. Ecology 67:967-74 88. B. 1983. The geography of organic decomposition rated. J.. Game Management. 146. 1985. Macroclimate and lignin control of litter decomposition rates. 53:291-320 78. Asmussen. J. 13:22-27 65. 1989. K. Hierarchical landscape structure and the forest planning model: discussants comments. Ecol. Nutrient cycling in an agricultural watershed: I. H. K. Oak Ridge.S. Leopold. T. Philadelphia: Harper & Row 68. Environ. 1977. ed. Ann. New York: Van Nostrand Reinhold 64. 1979. H.. H. Arizona. Peninsulas in Maine: woodyplant diversity.Annual Reviews www. 140. L. Waiters. Modeling forest canopythermal response on a landscape scale using remotely sensed data. S. R. Island biogeography and "habitat islands" of eastern forest. Results of a biological survey of the San Francisco Mountainregion and desert of the Little Colorado. The Theory of Island Biogeography. USDA Bull. 1984. L. 1983. J. ORNL/TM-10570. C. S. Int. MADISON" on 04/13/05. Holbo. 27:67-79 86. Identifying biotic boundaries along environmental gradients. RemoteSens.. 1967. D. Harbaugh. pp. Sugihara. McNaughton. Measuring the fractal dimension of landscapes. 1990. Birds 31:6-12 70. B. Comput. 1977. 27:1-10 66.. T. E. pp. Life zones and crop zones of the United States. Press 75.J.. V. A. R. C. Geographical Ecology: Patterns in the Distribution of Species. G. Oak Ridge Natl. Oikos 48:32124 62. 1933. In FORPLAN:An evaluation of a forest planning tool. USDAForest Service 85. R. B. W. Luvall. T. Ecology 59:465-72 80. 1989. T. Merriam. P. 1890. R.20:171-197. and the vegetative mosaic in wilderness landscapes. M. 55:25%94 79. S.22:533-35 71. Shackleton.. Whitcomb. Monogr.. Syst. 140. Geogr. J. Mankin...E.. Parasites. 1898. NewYork. 1972. Krummel. F. J. No. Measurementsof short-term thermal responses of coniferous forest canopiesusing thermal scanner data. Monogr..B. 1987. R. Lowrance.. Sinclair. T. H. Ecology 67:749-59 74.J. Environ. Mandelbrot. Grizzly bears and resource extraction industries: effects of roads on behavior. Phreatlc movement. See Ref. Chance and Dimension. 4681. V. Kolasa. T.. Rep..annualreviews. Qual. R. T. 1986.org/aronline Annu. Milne.. Chicago: Univ. J. 74:551-60 81. Am.. II. The Background of Ecology. Serengeti grassland ecology: the role of composite environmental factors and contingency in community organization. Fauna 3:1-136 83. R. R. Am. In Ecological Heterogeneity. E. R. Verein. 1984. E. 25:451-60 76. C. B. Bell.. Mandelbrot. Manage. N. Assoc. North Am.. Knight. For personal use only. 1989. Scale effects in landscape studies. R. 1986. Merriam.J. Klopatek. Princeton: Princeton Univ. 128-32. C. Ecol. 1988. 1987. T. Washington. McLellan. Environ. Ecol. J. M. B. 16:1-15 60. 1978. Tenn. L. OF WISCONSIN. T. Milne. C. 1987. ed. McNaughton.. Luvall. 59-83 61.and Three-Dimensional Markov Computer Models in Geology. A. pp.

Ecology Am.. Ruzicka. Z. O’Neill. Ecol. 10:299-309 L. V. Biol. J. Disturbance and Ecosystems. Champaign. J.. networks. DeAngelis. Resource 114.. Patton. Extinc120. T. 95. 1984. eds. P. O’Neill... InFlamm. Res. 5:233-43 105. Manage. Orbach. mate change. Environ.R.. pp. Jr. 1983. 1985.org by "UNIV. Sehimel.. 1988. R. A. T. 1987.. 1981. Forman. G... Yellowstone National Park. Paine. 99. 1987. S. J. pp. Oecologia 91.. Sala.. R. W. Naveh. Patton. 1981. Surv.. Water Res... optimization of the east Siovakian lowLandscape Ecol. O. L. Correll. Topical problems of 94. L. tions and Approaches..I11. MADISON" on 04/13/05. T. 3-14 Ecol. F. S.. 51:1173-79 69:40-45 101. Adv. Jr. Czechoslovakia Princeton: Princeton Univ. B. Robinson. Ecol. I. The Ecology of Natural Disflow. D. White. turnover and species diversystem science. V. 34997. Soil Sci.. C. Biogeogr. Harris. "Ecological ner. Mooney. Romme. J. 90. Nodes. Quinn. Vlllth Int... D. Probs. J. Ecol.. Hist. E. gion of the United States. Hierarchical Concept of Ecosystems. Knight. Nature 344:55-58 R. 1:198-208 distribution of rarity. Rykiel. as an emerging branch of human ecoExtinction. 1983. 1982. I. Waide. Ruzicka. M. F.. H. Lieberman. T. Landscape 51:145-78 Ecol. I11. 1:129-39 100. B. Schoener. OF WISCONSIN. No. RipaYork: Academic Press rian vegetation and channel morphology 104. Ecol. Jackson. T.. E. Proc. and MUMs:preserving dining. R. F. 10:275-97 Landscapediversity: the concept applied 93.. The role of cattle in 102. Woodmansee. 1987.. 1984. S. Ecological Diversiimpacton spatial patterns of water quality in agricultural watersheds. R.. Bull. landscape ecological research and plannetworks. 1986. F. F. Joyce. New 119.. M. Lauenroth. Response 117. BioSci32:664-70 ence 33:700-06 111.. Indices of landscape pattern. Ecol. 1987.. For personal use only.. Soc. W. R. Ekologica-CSSR 5:233-38 versity at all scales. R.. Monogr. P. Mildos. Theory and Appli109. Res. ty. H.’" See Ref. J.. J. 1988. Risser.. P. O. Allen. Hastings. l: 153-62 land utilization. 1982. F. Fire and landcation. S. W. dynamics. pact of riparian vegetation during base 1985. 1989. Sugihara. NewYork: Springer-Verlag scape diversity in subalpine forests of 92. Milne. Noss. 1. 1986. E. Pielou. Godron. 1988. NewYork: Wiley-Interscience Manage. nition of ecological disturbance. R. Z. Aust. E. H. H. Ecology 65:1466-75 quality in agricultural watersheds: im103. Risser. Krummel. 1975. The geographic tion in subdivided habitats. J. R. H. T. Ojima. D. J. A regional landscape to Yellowstone Park. BioScience approach to maintain diversity. Vol. 58 M. D. R. Ecol. Syst. W. J. 1988.. Press 113. 2:63-69 Ecol. J. H. Disturbance propatertidal landscapes: disturbance and the gation by bark beetles as an episodic dynamics of pattern.. W. 1983. F. et al. O’Neill. J. M. Rykiel. 1988.R. Landscape ecology 108. Noss. 116. S. J.. W. B. Pickett. In press 107.. 1986. T. W. 140. H. Cole. Monogr. Naveh. 1987. J.. Turner.annualreviews. R. G. NewYork: Springer-Verlag 2. Biol. 1988. Downloaded from arjournals. L. Rev.. See Ref.. Towardsa defiutilization scale and landscape pattern. Karr. A. Syrup. Allen. D. R. Wullstein. J. G. Schlosser. Primary tors controlling soil organicmatter levels production of the central grassland rein Great Plains grasslands.. Kozova. 52:199-221 Biogeography of two southwest American oaks in relation to atmospheric 110. 112. 17:233-40 turbance and Patch Dynamics. W. Karr. H. P. S. Patton. eds. T. Levin. landscape phenomenon. V. CBESSAS. Coulson. J. Senft. Bratislava.20:171-197. Adamof northern forests to CO2-inducedclisen. Karr. M. Peterjohn. J. Landscapeecology: 73 .. Post. A. 12:189sity in an experimentally fragmented 237 California annual grassland. L. S. Romme.. Gardachieved within the target-oriented project of the basic research. Science 231:813-19 Sharpe... T.J. A Exp. 1981.. Water riparian forest. J. R. G. H. 1988. A.Annual Reviews www.. the volatile loss of nitrogen froma shortNutrient dynamics in an agricultural grass steppe. G. V.org/aronline Annu. M. C.Special Publ. Quinn. Analysis of facA. N. Results 96. 10:361~55 98. P. Schimel. 1984. Dynamics of fractal 115.. Nat. G. Landsc. D. L. Landscape Ecol. Conserv. Pastor. D. LANDSCAPE ECOLOGY 195 neutral Bayesian model. T. Landscape Ecology: Direc89. Gardner.annualreviews. Schlosser. Environ. 1982.R. R. Inst... 112. R. R. Ruzicka. M. 76:71-82 Landscape Ecology. Hrnciarova. Biogeochemistry 2:39-52 watershed: observations on the role of a 118. H. J. eds. R. et al. W. 1988. K. A. Oecologia74:161106. L. Neilson. 1986. Landscape state-of-the-art.

139-58 127. Predicting across scales: theory development and testing..Annual Reviews www. S. Can. Dyer. Tall Timbers Res. L. 27:39-51 144.. H. P. R. BioScience 38:92-98 138. M. Appl. Tall Timbers Fire Ecology Conf. In Proc. G. A Theory of Forest Dynamics. C. Introduction to Percolation Theory. 140. Tallahassee.annualreviews. pp. O’Neill... 1988.. Conley. 85-101 145. Steiner. 11:23747 143.. Kratz.. et al. 1:24151 149. Appl.. R. D. Woodmansee. Modelingforest landscapes and the role of disturbance in ecosystems and communities. In Scales and Global Change. 467-72 Simulation Soc. 1987. Turner. Sklar. 125-51.. I-I. Ges. Ellis. USA. Changesin the spatial patterns of land use in Georgia. 1986. E. Predicting the spread of disturbance in heterogeneous landscapes. 1990. 241-98..I. Fire. J. Troll. Swanson. A. Tucker. See Ref. H. Sharpe. O. 9:128-39 122. Shugart. Factors influencing patterns of grazing behavior on shortgrass steppe. Rittenhouse. 1986. P. J. R. See Ref.landscape: 1935-1982. Steams. D. Shugart. R. 1978. No. 1986. D. bl. M. T. 1988. Downloaded from arjournals. Erdkunde. R. Luftbildplan and okologische Bodenforschung... J. J. Fire management in Yellowstone National Park. H. Computer 133. pp. B. Landscapeplanning and protection of the environment. Turner. Luker. H. G. R.. T. Woodmansee. Landscape Ecol.. Rastetter. J..20:171-197. R. Landscape Ecol. L. R. 1987. Coughenour. grazing and the landscape heterogeneity of a Georgiabarrier island. G. D. Oikos.org by "UNIV. pp. New York: Springer-Verlag 141. 1989.. G. Turner. pp.. F.. Landform effects on ecosystem patterns and processes. 196 TURNER 121. Z. A. In press 134... S. M. See Rcf. E. H. R. Osterman. Summer Computer Simulation Conf. A. R. 1989. Large herbivorc foraging and ecological hierarchies.. 7:13951416 140. London: Taylor & Francis 135.annualreviews. U. MADISON" on 04/13/05. G. Turner. 1988. Sci. A spatial simulation of ecosystem succession in a Louisiana coastal landscape. Crossley. L. 1987. M. A spatial simulation modelof land use changesin a piedmont county in Georgia. F. In press 146. Sellers. Berlin 139. H. Sellers. Rittenhouse. F. A. Rev. R. M. R.org/aronline Annu. M.. Ruscher. M. The Analysis and Interpretation of Landscape Heterogeneity. Feedingecology and niche separation in someungulates on the shortgrass prairie. M.. W. 353-68 131. 1985. Math. Sala. Senft. J. Rastetter. ed. 1988. Schwartz. I4. W.. Simulation models of forest succession. Rosswall.1990. Gardner. Washington. C.C.. E. G. Vegetation dynamicsin a southern Wisconsin agricultural landscape. H. M. V. T. Turner. Turner. 1987. 1985. Jr. J. Landscape Heterogeneity and Disturbance. G. Pattern and scale: statistics for . F. Dale. Remote Sens.. D. D. For personal use only. H. 1987. Quantitative Methods in Landscape Ecology. G. Dev. 1987. O’Neill. R. G. Ecol. 1985. M. Spatial Pattern in Plankton Communities. F. Risser. G. NewYork: Springer-Verlag 128. R.. W. 1:29-36 142. P. Dep. R. Gardner. 1987. L. V. Stauffer. pp. Niche theory and community organization. G. J. ed. Bot. H. ed.. 140. BioScience 37:789-99 125. 1977. 2504. C. 1981. Stat. P.. Landscape Ecol. Int. See Ref. 1984. G. Smith. V. 103. pp. T. Costanza. S.. M. V. J. E. 146. In Proc.. Weinstein. Land use changes and net primary production in the Georgia.. C. Computermodels of terrestrial ecosystems. R.. Comput. Crosbie. ed. 18:343-53 123. Turner. Bonan. K. Shugart.. P. M. In press 147. C. Fla. Dunn. New York: Wiley 130. H.. E. 1939. 1990. B. Turner. H.. 14.. Woodmansee. P. H. F.. DC 132. Syst. Range Manage. eds.. Landscape Ecol.... L. pp.. 55:121-29 148. 124. N. Gardner. R. H. State Publ. Shugart. The contribution of landscape ecology. Ecol. Appl. H. M. pp. Shugart. Turner. J. Manage. Sklar. G. Landscape planning: a working method applied to a case study of soil conservation. J... Costanza. NewYork: Plenum 136. New York: Springer-Verlag. ed. 66:2634-39 129. G. 1:213-26 137. L. The development of spatial simulation mod- cling for landscape ecology. S. B.. 1976. In Couplingof Ecological Studies with RemoteSensing: Potentials at Four Biosphere Reserves in the United States. Senft. Bratton. R. G. Smith. Caine. Steele. M. Seagle. No. K. 71-81.38:8187 126. Spatial simulation of landscape changes in Georgia: a comparison of 3 transition models. OF WISCONSIN. 99-113. H. Dale. Michaels. B. Turner. Guntenspergen. Despain. Satellite remote sensing of primary production. 1989. 1988. D. Schreiber. Bailey. Leitner. G. Environ.

Watt. Webber. 1985. 29-45 160. pp. Hairston. A. A.. For personal use only. See Ref 146. Graz: R. Walker. See Ref. VonHumboldt. V. White. Downloaded from arjournals. D. A. 3. A. 1979. Factors influencing input and output of nitrogen in grasslands. 1807. R. Bennett. S. P.. J.. Appl. 103. J.M.. G. 1987. Binnian. T. Monogr. Soil Water Conserv. 1 (Ecology). Pattern.. P. White. Science 238:757-61 156. Ecol. J.. Natural disturbance and patch dynamics: an introduction. Quantitative Geography: A British View. R. Woodman. Van der Maarel. 45:229-99 162. Weinstein. Opdam. 26:1-80 165.. F. H. 1985.. S.H. N. E. 1975. Woebse.. J.R. 1895. H. 35:122 158. 1985. Landscape Ecology and Land Use.org/aronline Annu. A.. 1983. T. Tubingen 155. ed. 1:59-73 153. 1976. Rev.20:171-197. Shugart. Population responses to patch environments. F. F. N. 1983. Landschaftsokologie und Landschaftsplanung. 31:265-87 164. NewYork: Macmillan 166. Ecol. 1979. E. J. French. OF WISCONSIN. 1981. MADISON" on 04/13/05. Wiens. A. D. J.. Wiens. W. F. Communities and Ecosystems. Woodmansee. Landscapeecology. In press 150. 1987. 1989. G. See Ref. Musser. R. 1985. W. in Perspectives in Grassland Ecology. BioScience 37:119-27 151. R. Monogr. 117134. Boston: Routledge & Kegan Paul . NewYork: Springer-Verlag 172. Ecol.Annual Reviews www.org by "UNIV. H. Vol. M. Rev. Verlag 170. Whittaker. J. J. J. Warming. R. B. 1981. J. eds. ed. Ecological principles for physical planning. 413-50.. Cumulative impacts of oil fields on northern Alaskan landscapes. 1978. A. 1981. 89. Lot~dot~: Longman 154.. Vegetation of the Great Smoky Mountains. Ecological modeling of landscape dynamics.annualreviews. Everett. Rev. Ecol. O’Neill.. Perkins. Wegner. 36:172-77 161. Relationship between increased crop acreage and nonpoint-source poilu- ECOLOGY 197 tion: a Georgiacase study. Movements by birds and small mammalsbetween a woodand adjoining farm habitats. Urban. 7:81-120 167.. Whitney. Wrigley.. Syst. ldeenzu einer geographie der pflangen nebat einem naturgemalde der tropenlander. L. 1987. 3-13 163. Biol. Effects of patch size. C. and natural disturbance in vegetation. Lederer. Plantesamfund grundtrak of den okologiska plantegeographi. Boundary dynamics: a conceptual framework for studying landscape ecosystems. D. NewYork: Plenum 152. pp. Annu. R. H. N. P. NATOConf. O. Wolff. Copenhagen:Philipsen 157. process. Pattern and process in the plant community.. Conserv. Vertebrate responses to environmentalpatchiness in arid and semiarid ecosystems.P. et al. pp. Ecol. A. 1975. Vink. Somerlot. Gosz. 1979. White. Syst. A. Wiens. isolation and regional abundance on forest bird communities. F. Ecol. A. G. Ser. Landscape Ecol. Pickett.J. S.annualreviews. Van Dorp. Bot. Oikos 45:421-27 169. J. H. J. D. R. 1947. 103. D. Reed. H. Merriam.. P. M. R. Whittaker. A case study of woodlandcontinuity and change in the American midwest. N. LANDSCAPE landscape ecology.. The role of habitat patchiness in the population dynamicsof snowshoe hares. H. K. 1956. S. E.50:11130 171..W. 169-93 168. H. S. Holgate. J. J. Crawford. C. Shugart. See Ref. In The Breakdown and Restoration of Ecosystems. 16:349-57 159. pp. pp.

OF WISCONSIN. . Rev. Downloaded from arjournals. Ecol.org by "UNIV.annualreviews.Annu. MADISON" on 04/13/05. For personal use only. Syst.20:171-197. 1989.

MADISON" on 04/13/05.20:171-197. For personal use only. Downloaded from arjournals.annualreviews. . Syst.org by "UNIV. 1989.Annu. Ecol. OF WISCONSIN. Rev.