Biodiversity in Heterogeneous and Dynamic Landscapes
Summary and Keywords
Although the concept of biodiversity emerged 30 years ago, patterns and processes influencing ecological diversity have been studied for more than a century. Historically, ecological processes tended to be considered as occurring in local habitats that were spatially homogeneous and temporally at equilibrium. Initially considered as a constraint to be avoided in ecological studies, spatial heterogeneity was progressively recognized as critical for biodiversity. This resulted, in the 1970s, in the emergence of a new discipline, landscape ecology, whose major goal is to understand how spatial and temporal heterogeneity influence biodiversity. To achieve this goal, researchers came to realize that a fundamental issue revolves around how they choose to conceptualize and measure heterogeneity. Indeed, observed landscape patterns and their apparent relationship with biodiversity often depend on the scale of observation and the model used to describe the landscape. Due to the strong influence of island biogeography, landscape ecology has focused primarily on spatial heterogeneity. Several landscape models were conceptualized, allowing for the prediction and testing of distinct but complementary effects of landscape heterogeneity on species diversity. We now have ample empirical evidence that patch structure, patch context, and mosaic heterogeneity all influence biodiversity. More recently, the increasing recognition of the role of temporal scale has led to the development of new conceptual frameworks acknowledging that landscapes are not only heterogeneous but also dynamic. The current challenge remains to truly integrate both spatial and temporal heterogeneity in studies on biodiversity. This integration is even more challenging when considering that biodiversity often responds to environmental changes with considerable time lags, and multiple drivers of global changes are interacting, resulting in non-additive and sometimes antagonistic effects. Recent technological advances in remote sensing, the availability of massive amounts of data, and long-term studies represent, however, very promising avenues to improve our understanding of how spatial and temporal heterogeneity influence biodiversity.
Keywords: biodiversity, landscape ecology, spatio-temporal heterogeneity, landscape mosaic, patch dynamics, ecological disturbance, ecological succession, time-lag, remote-sensing, social-ecological landscapes
Biodiversity, or biological diversity, can be defined as the variety of life forms at all levels of biological systems—organismic, population, species, and ecosystem (Wilson, 1988). The term biological diversity was first used by wildlife scientist and conservationist Raymond F. Dasmann, in 1968, but only came into common usage in science and environmental policy in the 1980s. The term’s contracted form, biodiversity, first appeared in Wilson’s publication (Wilson, 1988). The term rapidly achieved widespread use among scientists, politicians, and citizens with the increasing recognition of global impacts of human activities on the environment.
Although the concept of biodiversity has only emerged in the last 30 years, ecological diversity has been studied for several centuries. Historical observations of natural ecosystems showed that species diversity was not evenly distributed and varied greatly both spatially and temporarily at different scales. These observations led to the development of several scientific disciplines studying processes influencing spatial and temporal patterns of biodiversity, in particular the disciplines of community ecology, biogeography, and landscape ecology. Community ecology has traditionally focused on local mechanisms influencing biodiversity, whereas biogeography has addressed large-scale patterns and processes. Landscape ecology merged both approaches in order to understand mechanisms that influence biodiversity in heterogeneous and dynamic landscapes across multiple scales.
The study of patterns and processes influencing biodiversity at large spatial scales (e.g., planetary scale) and temporal scales (e.g., geological times) began in the mid-18th century with Linnaeus (1707–1778), grew out of the work of people like Alexander von Humboldt (1769–1859), and fully developed during the early part of the 20th century. Studies of the geographic distribution of species showed that large-scale variations in biodiversity are primarily influenced by abiotic conditions such as temperature, precipitation, altitude, geographical barriers or soils, and by historical factors such as glaciations and continental drifts (Millington, Blumler, & Schickhoff, 2011). They also showed that spatial and temporal variations in biodiversity are influenced by biotic factors, local interactions between species, which are the focus of community ecology. Later, the publication of The Theory of Island Biogeography (MacArthur & Wilson, 1967) showed that the number of species found in an undisturbed island is determined by island size and the distance between the island and the mainland. The application of the island biogeography theory to terrestrial habitat fragments later contributed to the conceptual development of landscape ecology (see “Spatial Heterogeneity”). Limitations of the initial theory have been highlighted, and modifications and alternatives have been proposed over the years (e.g., Lomolino, 2000). Despite the theoretical, conceptual and methodological overlaps between biogeography and landscape ecology, the nature of the relationship between these two disciplines is seldom addressed in textbooks (Kupfer, 2011).
Landscape is a central concept of many disciplines, including ecology, geography, architecture, or literature (Burel & Baudry, 2003). As a result, there are many definitions of “landscape,” including the most popular one referring to a scenery of forests, mountains, and streams. Ecologists define landscape as “a spatially heterogeneous geographic area made of diverse interacting ecosystems” (Forman & Godron, 1986). Within the ecological hierarchy, a landscape represents a level more inclusive than an ecosystem, yet nested within a biome (Wiens, 2002). Ecological landscapes include relatively natural terrestrial and aquatic systems such as forests, grasslands, and lakes, and human-dominated environments including agricultural and urban systems. The German biogeographer Troll coined the phrase landscape ecology in 1939 (Troll, 1971). Throughout the following decades, the field of landscape ecology developed in Europe building on concepts from geography, botany, and land-use planning. It focused primarily on cultural landscapes, that is, human-dominated landscapes, and provided a novel understanding of the interactions between humans and their environments, at scales broader than those usually studied by ecologists. To this day, the European school of landscape ecology retains a more applied, transdisciplinary approach (Metzger, 2008). The 1980s were marked by the 1983 Allerton Park workshop, which established the American school of landscape ecology, more focused on natural systems and more invested in ecological theory than the European school. Landscape ecology gained prominence as a discipline due to the growing awareness of broad scale environmental issues and the importance of scale as well as technological advances in remote sensing and computing methods. The 1980s and 1990s were marked by a greater emphasis of landscape ecology on testing scientific theories, the widespread acceptance of the non-equilibrium concept, and the recognition that ecosystems cannot be understood without considering flows of energy and material across ecosystem boundaries (Pickett & Cadenasso, 1995). After a period of largely descriptive studies, landscape ecology is now using a great diversity of approaches to test the generality of concepts across systems and scales and to become a more predictive science.
The aim of this article is to present major milestones and theories in the field of landscape ecology that have led to our current understanding of processes influencing biodiversity in heterogeneous and dynamic landscapes (see Table 1). I will illustrate how landscape ecology theories have influenced current landscape management practices in the fields of conservation biology, agroecology, and restoration ecology. I will conclude by introducing the main challenges that landscape ecology is currently tackling. Although landscapes include both terrestrial and aquatic landscapes, most of the research in the field of landscape ecology has been focusing on terrestrial and freshwater systems; seascape ecology has only recently developed. This article will focus primarily on studies conducted on biodiversity in terrestrial landscapes, but most, if not all, concepts presented here can easily be extrapolated to biodiversity in aquatic landscapes.
Table 1. Synthesis of Main Theories That Have Contributed To Our Current Understanding Of Processes Influencing Biodiversity In Heterogeneous And Dynamic Landscapes, Associated Assumptions, and Predictions And Corresponding References.
Island biogeography theory
A species will persist in the landscape as long as the colonization rate exceeds the extinction rate.
All successions within a given region lead to the same community.
Species dominance varies with the ratio of the availabilities of two resources.
Theory of hierarchy
The spatial and temporal scales of processes are correlated.
Species have perfectly overlapping niches.
Intermediate disturbance theory
Species diversity peaks at intermediate disturbance levels.
The Recognition of Landscape Heterogeneity and Dynamics
Environmental heterogeneity has been a component of ecology from its inception and is therefore hardly a new concept in ecology. Continuous variations in the distribution of species or communities were described and related to spatial variations in topography or climate. Despite the fact that the progenitors of ecology like Hutchinson, Clements, and Whittaker did focus on heterogeneous gradients, local ecological homogeneity and stability became more and more often assumed, probably because it provided the simplification of a complex reality that was necessary for theory to develop (Wiens, 2000). In the 1950s, the view that ecological dynamics were played out in local habitats that are spatially homogeneous and temporally at equilibrium rapidly gained force. Although acknowledged, spatial heterogeneity was then considered as a constraint to be avoided (Pickett & Cadenasso, 1995). Ecological studies mainly focused on homogeneous systems, often natural systems, and tended to be conducted at local scales because they lent an apparent uniformity to processes under study. This early stage of ecology had a powerful and persistent effect on how we view ecological systems. In the early 1970s, theoreticians began to introduce temporal and spatial variation in their models and generate new interesting predictions. This renewed focus on spatial and temporal variations triggered a paradigm shift. Heterogeneity, both spatial and temporal, was confirmed as the center of attention for ecologists (Wiens, 2000).
Despite being intuitively clear, the concept of heterogeneity is hard to define. Heterogeneity may be broadly defined as the uneven, non-random distribution of objects (Forman, 1995) or the existence of discontinuities in space or time (Wiens, 1995). However, there are various components of heterogeneity—for example, spatial vs. temporal, observed vs. functional. Ecologists have often considered spatial heterogeneity separately from temporal heterogeneity, perhaps because these two dimensions are perceived as fundamentally different. For example, spatial heterogeneity is multidirectional and influences species traits such as mobility, whereas temporal heterogeneity is unidirectional and influences species traits such as phenology. Landscape ecology has historically focused on spatial heterogeneity (see “Spatial Heterogeneity”), although the role of temporal heterogeneity was also acknowledged (see “Temporal Heterogeneity”). More recently, new conceptual frameworks on landscape dynamics have been proposed to integrate both spatial and temporal heterogeneity (see “Spatio-Temporal Heterogeneity”).
Several models have been proposed over the last decades to describe spatial heterogeneity (Figure 1).
A first family of models conceptualizes spatial heterogeneity as discrete: the island model, the patch-matrix-corridor model, and the landscape mosaic model. Other discrete models include the point pattern model (e.g., map of wetland locations), the linear network model (e.g., river streams network), or graph matrix models, corresponding to a network composed of nodes and links between nodes (e.g., wetland connectivity). A second group of models considers that spatial heterogeneity is rather continuous: the variegation model, the landscape contour model, and the gradient model. All these models can be positioned along a continuum from pattern-based to species-orientated approaches. They have led to the development of parallel but complementary research paradigms in landscape ecology. For each of these models, spatial heterogeneity can be measured at different spatial scales of the landscape, i.e. at different spatial extents (size of the landscape) or different spatial resolutions (level of details included in the map; Figure 2).
Spatial heterogeneity is highly dependent on the spatial scale at which the landscape is studied. A landscape can have a low heterogeneity at the local scale but high heterogeneity at large scale (Wiens, 1989b).
In the 1960s, the island biogeography and metapopulation theories rapidly and profoundly influenced the way we view terrestrial systems and led to the conceptualization of the landscape as a patchy environment. The island model considers that the landscape can be represented as a combination of habitat patches and non-habitat patches for all species. In this model, environmental conditions are considered relatively homogeneous within a given patch, and clear patch boundaries can be defined. Habitat patches surrounded by unsuitable habitats can then be considered as analogous to oceanic islands in an “inhospitable sea” (Haila, 2002). In this model, spatial heterogeneity is measured as patch size, patch number, or distance between patches. The strong influence of island biogeography on landscape ecology led to the dominance of the island model and the development of many related theories and concepts. Because of its strong focus on the relationship between patch area and species diversity, it also had a huge impact on the design of protected areas (see “Landscape Management for Biodiversity”). However, the island model failed to consider that terrestrial habitats between patches are not equivalent to a sea and may actually have a wide range of influences on individuals, for example in terms of movements between patches.
The patch-matrix-corridor model was therefore proposed by Forman (1995), as an extension of the island model. This new model considered that patches are separated by the matrix, the dominant patch type in the landscape, rather than by an “inhospitable sea.” It also recognized that patches are linked to each other through corridors, linear habitat elements that connect two or more patches. The patch-matrix-corridor model was rapidly and widely adopted in conservation biology and led, for example, to the design and implementation of green infrastructures (see “Landscape Management for Biodiversity”). In this model, spatial heterogeneity is measured as patch size, patch number, and distance between patches, but also as edge length, edge/area ratio, or patch connectivity. Assumptions underpinning the patch-matrix-corridor model are reasonable in many situations, particularly in landscapes where there are patch-dependent species, in forested landscapes undergoing habitat fragmentation, for example. However, this model also relies on oversimplification. In particular, it fails to take into account that few species perceive the landscape as binary (suitable vs. unsuitable), and that the matrix is often heterogeneous, has a strong influence on species occurrence and spatial dynamics, and may even provide suitable habitat for some species.
More recently, the landscape mosaic model was proposed as a multi-patch extension of the patch-matrix approaches (Wiens, 1995). Indeed, landscapes are spatially complex, heterogeneous combinations of patch types, which cannot simply be categorized into discrete elements such as patches, matrix, and corridors. Patches may be more or less suitable to a species rather than suitable vs. unsuitable. In this model, connectivity corresponds to the extent to which movement is facilitated or impeded through different patch types across the landscape rather than by the presence or absence of a corridor. The landscape mosaic model has only recently emerged as a viable alternative to the island model. It gained prominence through the development of software products allowing the computation of a wide variety of landscape metrics for categorical maps (e.g., Fragstats software). This model was the first one to highlight that spatial heterogeneity can be separated into two components: compositional heterogeneity (the number and proportions of different habitat types) and configurational heterogeneity (the spatial arrangement of habitat types; Fahrig & Nuttle, 2005). These two components of heterogeneity can be defined based on a priori landscape maps, or based on the way organisms perceive landscape heterogeneity, i.e. functional heterogeneity (Fahrig et al., 2011). Despite its aim to provide a more realistic representation of how organisms perceive the landscape, this model still fails to consider that patch boundaries may not always be clear and discrete habitat types may be hard to define a priori.
Repeated observations of continuous variations in vegetation cover have led several authors to propose alternatives to the patch mosaic model for situations where spatial heterogeneity is continuous rather than discrete. The variegation model was, for example, proposed by McIntyre and Hobbs (1999) to take into account that boundaries between patch types are diffuse and spatial variations in vegetation cover are often gradual rather than abrupt. Originally developed for grazed Australian landscapes characterized by small patches of woodland and relatively isolated trees scattered through the landscape, this model is particularly relevant to study landscapes where vegetation cover is heterogeneous and patch boundaries are not easily defined. In this model, the landscape is seen as a fuzzy-edged mosaic resulting from species’ perception. However, this alternative conceptual model is habitat-centric, in that it proposes a gradient model of species’ habitat but does not provide a general model of the landscape (McGarigal, Tagil, & Cushman, 2009).
The landscape contour model was then developed by Fischer, Lindenmayer, and Fazey (2004). It recognizes that habitat is a species-specific concept, that no two species respond to landscape change in precisely the same way, and that habitat suitability is influenced by multiple ecological processes. The landscape is actually characterized by gradual changes (e.g., temperature, vegetation height, prey availability). As a result, each species has its own habitat contour map with peaks and troughs, spacing of contours representing changes in habitat suitability through space. The landscape can then be represented by overlaid species-specific habitat contour maps. Although this model rightfully recognizes that different species respond in unique ways to habitat changes, we often don’t know enough about species habitat requirements to build such models.
Finally, the gradient model was introduced by Cushman, Gutzweiler, Evans, and McGarigal (2010). It is based on the fact that organisms experience landscape structure as pattern gradients that vary through space according to the distance at which a particular organism perceives or is influenced by landscape patterns. Therefore, instead of analyzing global landscape patterns (e.g., average tree density within the landscape), it is usually more appropriate to quantify the local landscape pattern across the space delimited by an organism’s perceptual abilities (e.g., local tree density within 100 m). This model does not presuppose discrete structures but will identify them if they do exist. As a result, this model subsumes the landscape mosaic model as a special case. The increasing availability of continuous large scale remote-sensed data and recent developments in computing abilities have contributed to the emergence of this new landscape model. In this model, spatial heterogeneity can be measured, for example, by the degree of spatial autocorrelation. The gradient model facilitates the simultaneous analysis of multiple landscape variables and provides a flexible framework for conducting organism- or process-centered analyses. The gradient model currently constitutes the main alternative to the conceptualization of the landscape as a patchy environment.
Environmental heterogeneity also results from temporal variations of the environment. Ecological processes of growth, death, and replacement ensures that ecosystems are dynamic. These processes operate at different time scales. For example, trees live for decades and go through annual phenological changes, predators create pluri-annual cyclic variations of prey populations, and insects live only few days and adapt their activities to daily variations of the temperature. In most ecosystems, however, other factors contribute to temporal heterogeneity. Although landscape ecology has primarily focused on the causes and consequences of spatial heterogeneity, other sub-disciplines of ecology, in particular ecosystem ecology and community ecology, have studied the temporal dynamics of ecosystems for more than a century. These disciplines have largely contributed to our current understanding of the causes and consequences of temporal heterogeneity. Most of this literature has primarily focused on two main ecological processes that create temporal heterogeneity: ecological succession and ecological disturbance.
Ecological succession was historically defined as a directional, predictable change over time in species or groups of species called communities (Grime, 1979). Succession is one of the longest-studied ecological concepts. The idea of ecological succession goes back to the beginning of the 19th century but was only formalized in 1899 by Henry Chandler Cowles. His observation of vegetation along sand dunes of different ages, moving from bare sand beach, to grasslands, to mature forests, led him to define the ecological succession as a repeatable sequence of community changes specific to particular environmental circumstances (Cowles, 1899). Temporal heterogeneity created by ecological succession can be characterized for a given site by the rate of change in parameters such as humidity, light, or vegetation cover. In 1916, Clements suggested that ecological succession is highly predictable and deterministic regardless of starting conditions. Clements’ theory became dominant, and ecological succession was, until the 1960s, considered as a process leading to a stable end-stage called “climax,” sometimes referred to as the “potential vegetation” of a site, and shaped primarily by the local climate (Whittaker, 1967). Rigorous data-driven tests of this theory, however, started showing the role of historical contingency and the existence of alternate pathways in the actual development of communities. Moreover, the importance of ecological disturbances was increasingly acknowledged. Indeed, most natural ecosystems experience disturbance at a rate that makes the “potential vegetation” unattainable. As a result, the notion of climax was abandoned in favor of the non-equilibrium theory of ecosystems dynamics (Whittaker, 1953). It was acknowledged that the trajectory of successional changes can be influenced by site conditions (pH, soil structure) or the species initially present. Moreover, there is now ample evidence that ecological succession is strongly influenced by the character of the event initiating succession, the type of disturbance, as well as by spatial heterogeneity.
Pickett and White (1985) provided the most widely quoted definition of ecological disturbance: “any relatively discrete event in time that disrupts ecosystems, community, or population structure and changes resources, substrate availability, or the physical environment” (p. 7). Disturbances can either be natural or anthropogenic. Natural disturbances include earthquake, volcanic eruption, tsunami, fire, flood, windstorm, insect or disease outbreak, grazing, trampling, drought. Anthropogenic disturbances include agriculture, road construction, urbanization, climate change, anthropogenic fire, clear-cutting, forest clearing, or the introduction of invasive species. Different types of disturbance trigger different types of succession. Primary succession corresponds to the colonization of habitats devoid of life, such as newly exposed rocks, sand surfaces, lava flows, or glacial tills. The rate of primary succession usually tends to be slow because of the long process of soil development, which can take up to a thousand years. Secondary succession corresponds to the successional dynamics, following a less severe disturbance that leaves seeds, spores, or the subterranean part of plants. It usually occurs at a faster rate than primary succession because the habitat is readily accessible to species either already present or colonizing the disturbed area. Secondary succession is much more commonly observed than primary succession. Indeed, disturbance is a natural component of most ecosystems, from the tundra and forests to savannas or deserts. Temporal heterogeneity created by disturbance can be characterized by a combination of parameters such as disturbance size, frequency, intensity or season. This combination of parameters is referred to as the disturbance regime.
Different ecological processes occur at different time scales. Ecological succession creates a relatively slow and continuous temporal heterogeneity, whereas ecological disturbance creates a more rapid and abrupt temporal heterogeneity. Just like spatial heterogeneity depends on the spatial scale considered, temporal heterogeneity depends on the temporal scale considered and should be measured at the appropriate scale based on the species, the system, and the question studied. Temporal heterogeneity can be assessed based on measurements at different times during a year, which will highlight seasonal variations due, for example, to phenology or agricultural perturbation. It can also be assessed based on measurements in different years at the same season; this will highlight the effect of slower processes such as succession or less frequent and more catastrophic disturbances such as fire. There is still a lack of conceptual and methodological frameworks to study temporal heterogeneity. Some measures of temporal heterogeneity do exist, including the rate of change, the frequency and intensity of change, the habitat lifespan, or temporal autocorrelation. Several authors have recently highlighted the lack of adequate consideration for the temporal dimension of environmental heterogeneity (e.g., Metzger, 2008). Although temporal heterogeneity is now increasingly acknowledged in landscape ecology research, further efforts are still needed to better account for temporal heterogeneity (Bertrand, Burel, & Baudry, 2015).
Landscapes have long been perceived by ecologists as static spatial patterns, probably as a result of the relatively short time scale of human observations and the historical dominance of equilibrium theories (see introduction of “The Recognition of Landscape Heterogeneity and Dynamics”). From the second half of the 20th century, numerous studies started showing that spatial and temporal heterogeneity interact in multiple and complex ways. Ecological disturbance and succession both create and respond to spatial heterogeneity. For example, the influence of local conditions on succession can result in temporal asynchrony of succession at larger scales that creates spatial heterogeneity. Moreover, ecological succession in a given site is influenced by the landscape context: succession is faster in smaller disturbed patches surrounded by undisturbed patches, due to the proximity of colonizing organisms in neighboring patches. Under certain conditions, disturbance may increase spatial heterogeneity, which will then limit the impact of a later disturbance by spreading the risk. Moreover, several studies have shown that there are complex spatial interactions between disturbances, for example between fire and grazing (Turner & Bratton, 1987). As a result, spatial heterogeneity can act either as a stabilizing factor, decreasing temporal heterogeneity (e.g., spatial heterogeneity may decrease fire frequency through the role of more open habitats as fire breaks), or as an accelerating factor, enhancing temporal heterogeneity (e.g., spatial heterogeneity may speed up succession dynamics through increased recolonization processes).
The increasing recognition of both dimensions of landscape heterogeneity led in the late 1970s to the development of new conceptual frameworks on landscape dynamics (Figure 3).
The concept of “patch dynamics” was developed to take into account that landscapes are dynamic mosaics of interconnected patches of environmental conditions or communities (Levins & Culver, 1971; Thompson, 1978). This concept does not suggest that landscapes are patchy in an absolute sense, nor that patches are necessarily homogeneous. It suggests that environmental conditions or communities change over time in a relatively patch-wise manner. A patch can go through three different states: potential, active, and degraded. Patches in the potential state are transformed into active patches through colonization of the patch by dispersing species arriving from other active or degrading patches. Patches are transformed from the active to the degraded state when ecological disturbance occurs and the patch is abandoned. Finally, patches change from the degraded to the potential state through a process of recovery, that is, ecological succession.
Bormann and Likens developed a similar concept called the “shifting mosaic steady-state” (Bormann & Likens, 1979). They considered that vegetation at a given site in the landscape changes over time, but, if averaged over a sufficiently long time or large area, the proportion of the landscape in each stage is relatively constant and can be considered as being in equilibrium. This concept emphasizes that even systems with a high disturbance frequency can be considered as being in a “steady-state” or “equilibrium” if the creation of new patches is balanced by the maturation of old ones, that is, if there is a balance between disturbance and succession at large spatial scales. Such balances are to be expected when the likelihood of a disturbance increases with the time since the last disturbance, when disturbances regimes are stable, and when patch size is small relative to the landscape scale considered (Pickett & White, 1985). The main challenge of the “shifting mosaic steady-state” concept is, however, to determine the type of ecosystems for which and the scale at which the concept is applicable. Despite its practical limits, the concept of landscape equilibrium has contributed to important theoretical advances regarding the central importance of scale hierarchies in landscape ecology.
Ecologists have realized over the last decades that scale (grain and extent; Figure 2) and levels of organization (e.g., patch, stand, forest, landscape; Figure 4) are central issues in ecology (Levin, 1992).
The theory of hierarchy postulates that ecological systems are structured in discrete levels of organization and that ecological processes are associated with characteristic spatial and temporal scales (Allen & Starr, 1982; O’Neill, 1986). Moreover, these spatial and temporal scales are often related: short-term changes tend to affect small areas while long-term changes affect larger areas (Figure 4; Peterson, Allen, & Holling, 1998). As a result, different processes require different spatial and temporal scales of study, and most processes actually require multiple scales of study. Although this theory provides a general guideline for considering pattern-process relationships, further developments are still required to adequately conceptualize and quantify spatio-temporal heterogeneity (Dutilleul, 2011). The relationship between spatial and temporal heterogeneity is currently considered one of the hottest topics in landscape ecology (see “Current Challenges”).
Biodiversity in Heterogeneous Landscapes
One of the major goals of modern landscape ecology is to understand how spatial and temporal heterogeneity influence biodiversity. To achieve this goal, researchers came to realize that a fundamental issue revolves around how they choose to depict and measure heterogeneity (Turner, 1989). Indeed, observed landscape patterns and their apparent relationship with biodiversity often depend on the spatial scale of the observations and on the models used to describe the landscape (Wiens, 1989a). Like landscape heterogeneity, biodiversity can be measured and monitored at several spatial scales. This led Whittaker (1972) to develop the three concepts of local patch diversity, among-patch diversity and total diversity (Figure 5).
Local patch diversity, called alpha diversity, refers to the diversity within a particular habitat patch or ecosystem. It corresponds for example, to the number of species within a patch of woodland. Among-patch diversity, called beta diversity, refers to the diversity between habitat patches or ecosystems. It corresponds, for example, to the total number of species that are unique to each of the ecosystems being compared. Total diversity, called gamma diversity, corresponds to the total number of species across all habitats within a landscape. Ecologists have historically focused on alpha diversity, but research on beta diversity has recently increased. Different landscape models have tended to focus on different components of biodiversity. For example, the island and patch-matrix-corridor model tend to focus on a single patch or group of patches of a single habitat and the associated alpha and beta diversity, whereas the landscape mosaic model tends to focus on multiple habitats and the associated gamma diversity. As a result, different landscape models have highlighted distinct but complementary effects of spatial heterogeneity on species diversity.
The Effect of Patch Structure
One of the most fundamental ecological relationships is that as the area of a region increases, so does the number of different species encountered within that region. This relationship is primarily due to the fact that the number of habitats, and therefore ecological niches, increases as the area of a region increases. While this may seem obvious, this observation first occurred only late in the 18th century and slowly took hold in the 19th century. The species-area relationship was initially elucidated in islands. Later, species diversity was shown to be influenced by isolation between the island and the mainland (MacArthur & Wilson, 1967).
Following the assumption that patches are analogues of islands (see the island model in “Spatial Heterogeneity”), the effects of patch size and isolation were tested in terrestrial landscapes (Figure 6).
Ecologists have since then produced hundreds of examples of increasing species diversity with increasing habitat patch size (Debinski & Holt, 2000). This relationship is primarily due to the fact that species have minimum patch size requirements. Numerous studies have also shown that patch isolation, the lack of habitat in the landscape surrounding the patch, has a negative effect on species diversity (McCoy & Mushinsky, 1999). This relationship is primarily due to the fact that the chance of species immigration into a given patch decreases with the distance between this patch and the species pool. Immigration occurs predominantly from habitat within the neighborhood of the patch rather than from a common “mainland.” Each patch is influenced by its own species pool, which is either assumed to be the nearest patch, the nearest patch weighted by area, or the summed or weighted summed areas of all patches within an appropriate distance. The effects of patch size and patch isolation have historically been considered as distinct. Lenore Fahrig recently challenged this assumption and offered an alternative hypothesis called the “habitat amount hypothesis” (Fahrig, 2013). This new hypothesis predicts that “species richness in equal-sized sample sites should increase with the total amount of habitat in the ‘local landscape’ of the sample site, where the local landscape is the area within an appropriate distance of the sample site” (p. 1649). The validity of this alternative hypothesis remains to be tested.
Upon the realization that explanations based only on processes operating at the patch scale do not fully account for diversity patterns at larger scales, new conceptual frameworks were developed. The term metapopulation arrived in the ecological literature in 1970 to describe a group of populations of the same species that are separated by space but interact, as individuals move from one population to another (Levins, 1970). The metapopulation theory states that (a) the smaller the population (and by extension, the habitat patch), the more likely it is to go extinct through population fluctuations and (b) the group of populations remains stable at the landscape scale because individuals immigrate from one population to colonize vacant patches or rescue other populations from extinction. The development of the metapopulation theory, in conjunction with the development of source-sink dynamics, emphasizes the role of both patch size and patch isolation/connectivity when considering biodiversity in heterogeneous landscapes. Since its early development, the metapopulation ecology has played an increasing role in landscape ecology and conservation science (Gilpin & Hanski, 1991). The metapopulation theory has led to the development of several new concepts including the threshold condition for persistence, the contributions that individual habitat patches make to metapopulation dynamics and persistence, the time to metapopulation extinction, and the effective size of a metapopulation living in a heterogeneous patch network (see Hanski & Ovaskainen, 2003 for further details on these concepts). It has also lead to the extension of the power-law species-area relationship to the species–fragmented area relationship, which takes into account the effect of fragmentation in addition to the effect of habitat loss on the number of species (Hanski, Zurita, Bellocq, & Rybicki, 2013).
More recently, the concept of metacommunity was developed as an extension of metapopulations. It defines a metacommunity as a group of communities that are linked through the dispersal of individuals from multiple potentially interacting species (Wilson, 1992). The metacommunity theory states that the rate and frequency of dispersal mediates the spatial distribution of species diversity, the abundance and the flux of energy across the metacommunity. Leibold et al. (2004) identified four types of metacommunities based on the relative importance of structuring processes: the patch-dynamic, the species-sorting, the mass effects, and the neutral views. Cottenie (2005) then showed that, among 158 published data sets, the majority of metacommunities were explained by both environmental and spatial variables and were structured by species-sorting dynamics (SS), followed by a combination of SS and mass-effect dynamics, while neutral processes were the only structuring process in 8% of studies.
The Effect of Patch Context
The profound influence of island biogeography and the development of the metapopulation and metacommunity theories have contributed to the historical dominance of the island model in landscape ecology. However, studies comparing diversity in island and terrestrial habitats have started showing that fragments are actually not islands (Norton, Hannon, & Schmiegelow, 2000). Indeed, fragments are not embedded in a true inhospitable matrix. This realization was the foundation for the second major axiom of contemporary landscape ecology: patch context matters. Studies showed that the matrix surrounding patches can have a strong influence on species diversity observed within patches and can be even more important than patch size and spatial arrangement of patches (Bender & Fahrig, 2005).
The matrix influences conditions found near the edge of patches. Abiotic effects correspond to changes in the environmental conditions that result from the proximity to a structurally dissimilar matrix (e.g., more light entering a forest patch). Direct biological effects correspond to changes in species abundance and distribution caused directly by physical conditions near the edge (e.g., more shade-intolerant plant species). Indirect biological effects involve changes in species interactions such as predation, brood parasitism, competition, herbivory, and biotic pollination and seed dispersal (e.g., more predators entering a forest patch near the edge). These effects have been extensively studied in the context of forest fragmentation (Laurance et al., 2002). Some species are only weakly affected by these variations in environmental conditions (generalist species). However, other species completely avoid patch edges and occur only within the core part of habitat patches (interior species). The distance that separates the patch border from the point where environmental conditions do not differ from those found in the interior of the patch can sometimes be substantial. As a result, some species actually occur only within patch edges (edge species). The variation in abundance or occurrence of a species near the edge is called the edge effect; it can be either positive or negative.
The matrix also influences movements between patches. The degree to which the matrix facilitates or impedes movement among patches is called connectivity. Connectivity can be assessed based on structural connectivity (the physical properties of patches and the matrix) or based on functional connectivity (the actual movement of individuals among patches). Connectivity tends to increase when the matrix is structurally more similar to the patches and decreases when the matrix alters the behavior of organisms or increases mortality (Driscoll, Banks, Barton, Lindenmayer, & Smith, 2013). For example, hedgerows enhance movements of forest species between forest fragments. Connectivity is species dependent. It is influenced by factors such as species’ perception of structural connectivity or individuals’ mobility. Connectivity can be assessed based either on habitat permeability or on movements of individuals. Several methods have been developed to model connectivity, such as Circuitscape, which is based on circuit theory (McRae, Dickson, Keitt, & Shah, 2008). The emergence of landscape genetics over the last decade or so has greatly improved our understanding of gene flow in heterogeneous and fragmented landscapes, therefore providing estimates of functional connectivity (Manel & Holderegger, 2013).
The Effect of Landscape Spatial Heterogeneity
The landscape mosaic model fully took into account that the matrix is not homogeneous. Moreover, it triggered a shift from studies focusing on single habitats, often natural habitats, to studies focusing on multiple habitats, that is, all major habitats found within a landscape.
In 1992, Dunning et al. defined four landscape processes that affect population in mosaic landscapes: landscape complementation, landscape supplementation, source-sink relationships, and neighborhood effects (Figure 7; Dunning, Danielson, & Pulliam, 1992).
Landscape complementation stems from the need of animals to utilize resources available in distinct habitats. For example, leopard frogs require ponds for breeding, grassy meadows for foraging, and a stream or lake for overwintering. Similarly, some bird or bat species require one type of habitat for foraging and another type of habitat for roosting. These resources are all required for the organism to survive and are therefore non-substitutable. Landscape complementation occurs when all required habitats occur in close proximity within a landscape and allow a population to be maintained (Andren, Delin, & Seiler, 1997). Landscape supplementation corresponds to a similar situation, where species can use different resources but these resources are substitutable. For example, bird species that are normally restricted to large forest patches may be able to occur in small patches of forest if they are able to forage in other nearby habitat patches. Deriving from the metapopulation theory, source-sink relationships within the landscape mosaic occur when the landscape contains both high quality and low quality habitat. The population in the high quality habitat has a positive growth rate, whereas the population in the low quality habitat has a negative growth rate. However, both populations are maintained thanks to a constant flow of emigrating individuals from the high quality to the low quality habitat. Source-sink relationships are influenced both by the proportion of high-low quality habitats, i.e. landscape composition, and by the distance between the source and the sink habitats, i.e. landscape configuration. Neighborhood effects occur when species abundance within a patch is more strongly affected by the characteristics of the contiguous patches than by those of more distant patches. Landscape complementation has been documented for vertebrates as well as for invertebrates (Tscharntke et al., 2012). These results caution against the use of binary description of landscapes (habitat vs. non-habitat) in the context of the metapopulation and metacommunity theories.
In 1961, MacArthur and MacArthur proposed the habitat heterogeneity hypothesis, which stipulates that an increase in the number of habitats leads to an increase in species diversity in a landscape (MacArthur & MacArthur, 1961). This increase can be explained by the fact that species have adapted to particular ecological niches (Grinnell, 1917), and that increasing the number of habitats increases the number of available ecological niches. Moreover, increasing the number of habitats increases the chance of landscape complementation for species that require resources found in distinct habitats. More recently, Fahrig et al., (2011) highlighted that the positive effect of landscape spatial heterogeneity on biodiversity through landscape complementation processes could result either from a change in landscape composition (the number or proportion of habitats) or from a change in landscape configuration (the spatial pattern of habitat patches).
Modifications of landscape heterogeneity are directly related to the processes of habitat loss and fragmentation: an increase in compositional heterogeneity implies some degree of habitat loss and an increase in configurational heterogeneity implies some degree of habitat fragmentation. Habitat loss and fragmentation have been shown to have dramatic negative impacts on biodiversity. As a result, there is an apparent contradiction in the literature between the expected positive effects of landscape heterogeneity versus the expected negative effects of habitat loss and fragmentation on biodiversity. The literature on landscape heterogeneity is focusing on the effect of niche diversity, landscape connectivity, and landscape complementation on gamma diversity, whereas the literature on habitat fragmentation focuses on the effect of patch size requirements and negative edge effects on alpha diversity. This apparent contradiction may therefore be explained by the fact that they focus on different biodiversity scales and distinct mechanisms. More recently, it was highlighted that any increase in environmental heterogeneity must lead to a reduction in the average amount of effective area available for individual species (Allouche et al., 2012). This “area-heterogeneity trade-off” provides the foundation for a prediction combining the heterogeneity and fragmentation predictions: biodiversity is expected to increase with increasing heterogeneity at relatively low levels of compositional and configurational heterogeneity, but it is expected to decrease above some critical threshold when total area or patch area become significant limiting factors. There is now widespread evidence supporting positive, negative, and hump-shaped relationships between habitat heterogeneity and biodiversity. The generality of heterogeneity–richness relationships is therefore still debated (Stein, Gerstner, & Kreft, 2014).
Biodiversity in Dynamic Landscapes
Landscape ecology has primarily contributed to our understanding of the role of spatial heterogeneity for biodiversity while making only modest advances regarding the role of temporal heterogeneity. Most of our understanding of the role of temporal heterogeneity for biodiversity actually comes from other sub-disciplines of ecology, in particular community ecology. Indeed, landscapes are dynamic due to ecological succession and ecological disturbance, which are among the longest-studied ecological processes.
The Complexity of Succession Mechanisms
Henry Cowles was the first ecologist to thoroughly characterize successional patterns, which he did in his classic 1899 study of sand dunes along the shores of Lake Michigan (Cowles 1899). He showed that, after a disturbance, plant species occupying or colonizing the site modified the environmental conditions at the site, and that different species were adapted to different parts of the environmental gradient. It was shown that there is usually an inverse correlation between species characteristics, called traits, which confer success during early and late stages of succession (McCook, 1994). As a result, early-succession plant species tend to be short-lived annuals that grow rapidly, whereas late-succession plant species tend to be long-lived perennials that grow slowly. This inverse correlation results in the sequential replacement of species from early to late-succession species (Figure 8).
Classic examples of plant succession studies include volcanic island succession, post-glacial succession, old-field succession, or post-fire succession. Although earlier succession studies have focused predominantly on plants, ecologists have also studied succession processes in animal communities. This abundant work showed that the turnover of animal species depends on vegetation structure, food, or nesting site availability provided by plants. But they also suggested that some animals actually have a strong influence on plant succession, for example through zoochory.
Several contradictory hypotheses about succession mechanisms were developed since Cowles’ study, including Clements’ theory (Clements 1916), triggering vivid debates. It later became evident that a universal succession model was unlikely, since numerous aspects of historical and environmental circumstances influence the succession process. After decades of controversy, Connell and Slatyer (1977) formulated a more global conceptual framework based on three mechanisms influencing succession. This framework assumed that plant species initiating the succession were early pioneer species that produce numerous seeds, germinate early, and grow quickly. Succession could then occur through facilitation, the ability to alter the environment, whereby early successional species modify the environment for later species that replace them. The second mechanism was tolerance, the ability to withstand a wide range of conditions, which allows later successional species to grow more slowly and eventually replace early successional species. Last, inhibition corresponded to the ability of late successional species to inhibit the growth of early colonists. Later, David Tilman developed another mechanistic view of succession based on the resource-ratio theory (Tilman, 1985). His model predicted that changes in the proportion of resources over time could shift community composition, because species specialize on specific proportions of those resources. Reviewing 26 studies testing the predictions of the resource-ratio theory, Miller et al. (2005) showed that the predictions were supported 75% of the time.
The current view of succession acknowledges the complex array of factors and contingencies that influence this ecological process. Experiments have shown that animals, in particular keystone animal species, play crucial roles in the succession process (MacMahon, 1981). Concepts developed in the field of community ecology are increasingly being integrated within the framework of “succession.” Priority effects, the impacts of particular species on community development due to prior arrival at a site, have been increasingly documented. However, the respective role of deterministic processes and historical contingencies has become increasingly controversial. Young, Chase, & Huddleston (2001) argued that both convergence (i.e., a return to the predisturbance state regardless of history) and divergence (i.e., contingent successions) actually occur. Models and methods are becoming more sophisticated and are now including complex spatial processes such as dispersal, species traits, or species interactions. Fukami et al. (2005) showed that community can be both convergent and divergent at different levels of community organization, with individual species remaining unique across different community replicates, whereas species traits generally become more similar. Finally, our current vision of succession has also been greatly influenced by recent advances within the ongoing debate between the niche theory (i.e., the theory upon which communities are assembled through local environmental filters and interspecific competitive exclusion; Hutchinson, 1957) and the neutral theory (i.e., the theory upon which species are considered equal in terms of competitive abilities and fitness, and communities are mainly structured by dispersal mechanisms; Hubbell, 2001). The neutral theory has generated considerable controversy, and its usefulness has been heavily criticized (McGill, 2003). Jonathan Chase recently suggested that we simply cannot unambiguously disentangle the relative importance of niche and neutral processes without an explicit consideration of the spatial scale on which patterns are examined, highlighting once again the central role of scale in ecology (Chase, 2014).
The Intermediate Disturbance Hypothesis and its Alternatives
Disturbance has long been recognized as influencing species coexistence and the maintenance of biodiversity (Connell & Slatyer, 1977). Disturbance is actually a natural component of most ecosystems such as savanna, forest, or chaparral. As a result, there is a considerable body of research examining the effects of disturbance regime on species diversity. The idea that the disturbance regime can actually influence species diversity goes back to the 1940s, but began to be studied in the 1960s. Much of this work has focused on testing the intermediate disturbance hypothesis, which predicts that intermediate frequency or intensity of disturbance will maximize species diversity (Grime, 1973). This relationship is expected because a) at high disturbance frequency, only species that quickly colonize and reach maturity are able to survive; b) at low disturbance frequency, competitively dominant species exclude competitively inferior species; and c) at intermediate levels of disturbance, both colonizers and competitors co-exist (Figure 9).
The intermediate disturbance hypothesis has been supported by several studies but has also been subjected to heavy criticism since its inception. Recently, Jeremy Fox called for a critical reassessment of the intermediate disturbance hypothesis on both empirical and theoretical grounds (Fox, 2013). He highlighted that a review of over 100 published diversity–disturbance relationships found that the predicted peak of diversity at intermediate disturbance levels rarely occurred (< 20% of studies). He also suggested that the intermediate disturbance hypothesis suffers from major theoretical flaws. First, disturbance reduces species’ densities, thereby weakening competition, but it also reduces the strength of competition needed for species exclusion; that is, it does not necessarily decrease species diversity. Second, intermediate disturbances slow competitive exclusion by increasing the long-term average mortality rate, thereby reducing differences in the average growth rates of competing species. As a result, it affects species’ abundance but not coexistence, it does not increase species diversity. Third, intermediate disturbances temporarily affect relative species fitness, but the species with favored fitness will out-compete the rest of the species.
Alternative hypotheses to the intermediate disturbance hypothesis have been developed. One of them proposes that species diversity is maximized by a disturbance regime resembling the historical disturbance regime because species generally adapt to the level of disturbance in their ecosystem through evolution, whether disturbance is high, intermediate, or low level. Indeed, in ecosystems where disturbance is naturally frequent, many species even depend on disturbance. This is the case, for example, in the Fynbos, which must burn every 6 to 45 years in order to sustain most plant species. Many Fynbos plant species store their fruit in fire-safe cones for release after a fire, and ants are enticed to bury fruits where they are safe from rodents and fire. After a fire, many plant species resprout, but the majority rely on the predictability of fires and only regenerate from seeds after a fire.
Time Lags in Biodiversity’s Response
Landscape ecologists usually assume a causal relationship between current landscape patterns and current species distribution patterns. This assumption may not always be correct, however, because species distribution patterns can reflect past as well as present landscape conditions. Indeed, when the time necessary for a species to respond to landscape changes is longer than the rate of landscape change itself, this species will respond to landscape changes with a delay (Figure 10).
This time lag results in what was called a legacy effect of past landscapes on current species distribution. These legacy effects are likely to be more important in habitats that have historically been relatively stable (e.g., old growth forests). In central Massachusetts, for example, the legacy of agricultural land-uses from the 18th century is still reflected in the vegetation of today’s forests. However, legacy effects have also been observed in highly disturbed landscapes such as European agricultural landscapes.
It was suggested that species most likely to exhibit lagged responses to landscape change are species with poor colonization abilities, species with large or stable local populations, species with long individual lifespans or seed banks, species with low turnover or population growth rates, and species with low sensitivity to environmental fluctuations (With, 2007). However, it was shown that even mobile birds and short-lived carabid beetles display time-lagged responses of up to several decades. In many cases, species distribution patterns are therefore likely to be explained by past landscape patterns more than by current landscape patterns.
In 1994, Tilman et al. predicted that habitat loss could cause a time-delayed extinction, an “extinction debt” (Tilman et al., 1994). Accumulating evidence suggests that such extinction debts are widespread and highly underestimated, occurring across a wide range of taxa and ecosystems (Kuussaari et al., 2009). Species with long generation times and populations near their extinction threshold are most likely to suffer from an extinction debt. The other side of the coin is a “colonization credit,” which is the slow reappearance of species after habitat gain. In heterogeneous and dynamic landscapes, both extinction debt and colonization credit are likely to occur and have complex effects on the reorganization of communities and ecosystem functioning. More recently, it was highlighted that the combination of global change drivers (climate, land use, invasion) and response mechanisms (metapopulation dynamics, dispersal limitation, successional dynamics, adaptation) is likely to result in cumulative time-lags (Essl et al., 2015).
Landscape Management for Biodiversity
Landscape ecology has been defined as a highly interdisciplinary and transdisciplinary science of environmental heterogeneity that aims to understand and improve the relationship between spatial patterns and ecological processes on a range of scales with the goal of achieving landscape sustainability (Wu, 2013). The landscape scale is usually the scale of action. As a result, the landscape represents a boundary concept shared by many disciplines, in particular conservation biology, land use planning, and agroecology or restoration ecology. This section provides four examples of the way landscape ecology theories have major implications on landscape management.
First, landscape ecology and metapopulation theories on the role of patch structure and context on species persistence with unstable local populations have influenced the field of conservation for decades (Hanski, 1999). The influence of island biogeography and the historical dominance of the island model led to conceptualizing protected areas as archipelagos of islands surrounded by an “inhospitable sea.” The recognition of the role of patch size and isolation for species diversity triggered the question about whether a single large reserve will conserve more species than several small reserves (Diamond, 1975). This initiated a vivid debate named the SLOSS debate (single large or several small), which started in the 1970s and 1980s and never really came to a resolution (Tjørve, 2010). It later became widely acknowledged that protected areas are not sufficient to protect biodiversity anyway. This triggered a shift in conservation policies, which now include the design of ecological networks to combat fragmentation and to protect both endangered and ordinary species. Long-term corridor experiments have shown that corridors increase species dispersal of plants and animals and, as a result, increase species diversity, although they also increase edge effects (Haddad et al., 2011). The creation and maintenance of wildlife corridors and green infrastructures are direct applications of the patch-matrix-corridor models. Finally, recent advances in metapopulation models have been critical to the management of many species in highly fragmented landscapes (McCullough, 1996). The listing of several species on the Endangered Species List, as well as their management and recovery plans, are based in part on the analysis of their metapopulation dynamics (Macdonald & Service, 2009).
Second, landscape ecology theories on the role of landscape spatial heterogeneity have influenced agricultural policies. In 2003, Benton and Vickery conducted a review of the impact of agriculture intensification on biodiversity suggesting that the loss of ecological heterogeneity at multiple spatial and temporal scales is one of the main factors explaining the decline of farmland biodiversity (Benton, Vickery, & Wilson, 2003). Current agricultural policies are therefore aiming at increasing landscape complexity by increasing semi-natural habitats such as woodlots or hedgerows and by increasing crop diversity to favor landscape complementation. Moreover, biodiversity provides many ecosystem services to humans, including food provision, pest regulation, pollination, or pollution reduction. Current land management policies are therefore aiming at improving landscape structures to regulate plant and animal populations as well as nutrient and water fluxes, with the goal to maximize ecosystem services.
Third, landscape ecology theories on succession have influenced the field of ecological restoration (Young, Chase, & Huddleston, 2001). Following an anthropogenic disturbance, a damaged ecosystem can return to a stable, healthy, and sustainable state through the ecological succession process. This ability to recover is called resilience. After a moderate disturbance, restoration can consist of merely assisting or accelerating this existing process. However, in a system that has experienced a more severe disturbance, for example extreme physical or chemical alterations of the environment, return to a sustainable state may require intensive restorative efforts to recreate environmental conditions that favor natural successional processes. These efforts will prevent succession from moving in unpredictable or undesirable directions—for example, favoring the invasion of exotic species. The aim of restoration is then to constrict environmental conditions within a narrow range that will increase the likelihood of returning to the desired stable, healthy, and sustainable state.
Finally, landscape ecology theories on disturbance have influenced fire management strategies both inside and outside protected areas. Fire is a natural component in many biomes and, as such, extremely important in shaping many of our landscapes. However, there have been and still are important gaps in our understanding of the effects of fire on ecosystems and biodiversity. As a result, fire management strategies have been modified and debated for decades. In the Kruger National Park in South Africa, fire suppression policies were implemented at the beginning of the 20th century, which resulted in a decrease of small fires and an increase of large fires, decreasing landscape heterogeneity and therefore biodiversity. Upon this realization, the park implemented different strategies one after the other: prescribed burning, rotational burning, natural burning. The lack of success of these strategies in this park, as well as in the rest of the world, led researchers to propose that a diversity of fire regimes, named pyrodiversity, may be necessary to maintain biodiversity (Parr & Andersen, 2006). Although evidence for this relationship remains scarce, many fire management strategies in conservation areas now seek to increase fire variability, hoping to create landscape heterogeneity and maintain biodiversity.
Recent technological advances are rapidly transforming the way we describe spatial and temporal heterogeneity. Indeed, remote sensors are now providing continuous data collected on large spatial extents (planetary), at fine spatial and temporal resolutions (few centimeters, every few days), from across the electromagnetic spectrum (hyperspectral), or in three dimensions (LIDAR). Such data represent an incredible opportunity to a) describe landscape heterogeneity in a much more objective and ecologically relevant way than what we have been doing so far, and b) better understand the impact of fine scale ecological processes operating over large spatial extents (Davies & Asner, 2014). The rise of the information age is generating massive volumes of data called “big data.” Such data is not easily handled by usual conceptual and methodological frameworks, which require the new development of “big science” (Hampton et al., 2013). Our current challenge is therefore to achieve a greater integration of diverse remote sensing and ecological big data to improve our understanding of relationships between landscape heterogeneity and biodiversity but also ecosystem function, stability, and resilience.
A better description of spatial and temporal heterogeneity should also lead to a better understanding of their relative and combined effects on biodiversity. In 1992, Fahrig’s simulation study suggested that temporal scale far outweighs the effect of spatial scale on population abundance (Fahrig, 1992). Yet, landscape ecology has mainly focused on understanding the role of spatial heterogeneity. Some studies have suggested a positive role of temporal heterogeneity for biodiversity, such as carabids in agricultural landscapes, showing that carabid species respond differently to spatial and temporal landscape heterogeneity based on their dispersal abilities (Bertrand, Burel, & Baudry, 2015). One of the current challenges for landscape ecology therefore remains to develop conceptual and methodological frameworks that truly combine spatial and temporal scales.
A better understanding of the role of temporal scale will involve taking into account the role of time lags. Our understanding of delayed biodiversity responses to landscape changes has increased rapidly over the last two decades. However, currently available frameworks and studies concern isolated mechanisms of delayed responses (e.g., extinction debt) of one biodiversity component (e.g., species). Cumulative time lags critically limit our ability both to understand the relationship between landscape patterns, species distribution, and ecosystem functioning and, therefore, to manage landscapes in an appropriate and efficient way. Our current challenge now consists in developing and testing new frameworks, for example on cumulative time lags (Essl et al., 2015). This will require interdisciplinary approaches combining disciplines such as historical ecology, landscape ecology, and metacommunity ecology. Making use of existing long-term monitoring programs such as the Long-Term Ecological Research network represent a promising avenue to better understand mechanisms causing cumulative biodiversity lags.
Due to the complexity of ecosystems and global change drivers, one of the main challenges for landscape ecology will be to develop new conceptual and methodological frameworks integrating multiple layers of information. Most disciplines are becoming increasingly aware of the need to take into account the spatial dimension. Predation ecology now refers to “the landscape of fear”. Road ecology is developing the concept of “landscape of noise”. The concept of “hidden mosaic” was recently proposed in the field of agroecology to take into account spatial variations in agricultural practices within fields. The role of landscape structure for microclimate is also being increasingly acknowledged. All these dimensions of the landscape are likely to influence biodiversity in non-additive ways; that is, their global effects on biodiversity will not only be the sum of each effect in isolation. These interactive effects might be synergetic or antagonistic, making it even more challenging to identify appropriate landscape management strategies. For example, increasing connectivity to favor species range shift in response to climate change may also decrease the chance of local adaptation or may favor species invasions (Haddad et al., 2014).
Finally, the most important challenge for landscape ecology will be to understand the complex interconnections between social and ecological processes. It was suggested that most problems encountered by societies in managing natural resources, including biodiversity, may arise because of a mismatch between the scales of management and the scales of the ecological processes (Cumming, Cumming, & Redman, 2006). For example, agricultural policies tend to concern the farm scale, whereas most ecological processes occur at the landscape scale. Similarly, ecosystem services provided by the environment tend to be produced and managed at the local or landscape scale, whereas these same services tend to affect people at much larger spatial and temporal scales. The diversity of societal demands for incompatible ecosystem services, including incompatible aspects of biodiversity, will increasingly trigger tensions. Our ability to adequately manage landscapes and biodiversity will therefore rely on identifying spatial trade-offs between ecosystem services at multiple spatial and temporal scales. The last decades have witnessed a greater development of the American school of landscape ecology, more focused on testing ecological theories than the European school, which has focused on the human dimensions of landscape. The challenge will be to move towards a unified landscape ecology discipline, considering both ecological and social processes to understand and manage social-ecological landscapes.
Allen, T. F. H., & Starr, B. (1982). Hierarchy: Perspectives for ecological complexity. Chicago: University of Chicago Press.Find this resource:
Burel, F., & Baudry, J. (2003). Landscape ecology: Concepts, methods, and applications. Enfield, NH: Science Publishers.Find this resource:
Clements, F. E. (1916). Plant Succession: An Analysis of the Development of Vegetation. Carnegie Institution of Washington.Find this resource:
Cowles, H. C. (1899). The ecological relations of the vegetation of the sand dunes of Lake Michigan. Botanical Gazette, 95–117; 167–202; 281–308; 361–369.Find this resource:
Forman, R. T. T. (1995). Land mosaics: The ecology of landscapes and regions. Cambridge, U.K.: Cambridge University Press.Find this resource:
Forman, R. T. T., & Godron, M. (1986). Landscape ecology. New York: Wiley.Find this resource:
Hansson, L., Fahrig, L., & Merriam, G. (Eds). (1995). Mosaic landscapes and ecological processes. Dordrecht, Netherlands: Springer.Find this resource:
Hutchings, M. J., John, L. A., & Stewart, A. J. A. (2000). The ecological consequences of environmental heterogeneity. New York: Cambridge University Press.Find this resource:
Pickett, S. T. A., & White, P. S. (1985). The ecology of natural disturbance and patch dynamics. Oxford: Elsevier.Find this resource:
Rosenzweig, M. L. (1995). Species diversity in space and time. Cambridge, U.K.: Cambridge University Press.Find this resource:
Turner, M. G., Gardner R. H., & O’Neill, R. V. (2001). Landscape ecology in theory and practice: Pattern and process. New York: Springer Verlag.Find this resource:
Wiens, J. A., & Moss, M. R. (2005). Issues and perspectives in landscape ecology. Cambridge, U.K.: Cambridge University Press.Find this resource:
Allen, T. F. H., & Starr, B. (1982). Hierarchy: Perspectives for ecological complexity. Chicago: University of Chicago Press.Find this resource:
Allouche, O., Kalyuzhny, M., Moreno-Rueda, G., Pizarro, M., & Kadmon, R. (2012). Area–heterogeneity tradeoff and the diversity of ecological communities. Proceedings of the National Academy of Sciences, 109(43), 17495–17500.Find this resource:
Andren, H., Delin, A., & Seiler, A. (1997). Population response to landscape changes depends on specialization to different landscape elements. Oikos, 80, 193–196.Find this resource:
Bender, D. J., & Fahrig, L. (2005). Matrix structure obscures the relationship between interpatch movement and patch size and isolation. Ecology, 86, 1023–1033.Find this resource:
Benton, T. G., Vickery, J. A., & Wilson, J. D. (2003). Farmland biodiversity: Is habitat heterogeneity the key? Trends in Ecology & Evolution, 18, 182–188.Find this resource:
Bertrand, C., Burel, F., & Baudry, J. (2015). Spatial and temporal heterogeneity of the crop mosaic influences carabid beetles in agricultural landscapes. Landscape Ecology, 31(2), 451–466.Find this resource:
Bormann, F. H., & Likens, G. E. (1979). Pattern and process in a forested ecosystem. New York: Springer Verlag.Find this resource:
Burel, F., & Baudry, J. (2003). Landscape ecology: Concepts, methods, and applications. Enfield, NH: Science Publishers.Find this resource:
Chase, J. M. (2014). Spatial scale resolves the niche versus neutral theory debate. Journal of Vegetation Science, 25, 319–322.Find this resource:
Connell, J. H., & Slatyer, R. O. (1977). Mechanisms of succession in natural communities and their role in community stability and organization. The American Naturalist, 111, 1119–1144.Find this resource:
Cottenie, K. (2005). Integrating environmental and spatial processes in ecological community dynamics. Ecology Letters, 8, 1175–1182.Find this resource:
Cumming, G., Cumming, D., & Redman, C. (2006). Scale mismatches in social-ecological systems: Causes, consequences, and solutions. Ecology and Society, 11(1), 14.Find this resource:
Cushman, S. A., Gutzweiler, K., Evans, J. S., & McGarigal, K. (2010). The gradient paradigm: A conceptual and analytical framework for landscape ecology. In S. A. Cushman & F. Huettmann (Eds.), Spatial complexity, informatics, and wildlife conservation (pp. 83–108). Tokyo: Springer.Find this resource:
Davies, A. B., & Asner, G. P. (2014). Advances in animal ecology from 3D-LiDAR ecosystem mapping. Trends in Ecology & Evolution, 29, 681–691.Find this resource:
Debinski, D. M., & Holt, R. D. (2000). A survey and overview of habitat fragmentation experiments. Conservation Biology, 14, 342–355.Find this resource:
Diamond, J. M. (1975). The island dilemma: Lessons of modern biogeographic studies for the design of natural reserves. Biological Conservation, 7, 129–146.Find this resource:
Driscoll, D. A., Banks, S. C., Barton, P. S., Lindenmayer, D. B., & Smith, A. L. (2013). Conceptual domain of the matrix in fragmented landscapes. Trends in Ecology & Evolution, 28, 605–613.Find this resource:
Dunning, J. B., Danielson, B. J., & Pulliam, H. R. (1992). Ecological processes that affect populations in complex landscapes. Oikos, 65, 169–175.Find this resource:
Dutilleul, P. (2011). Spatio-temporal heterogeneity: Concepts and analyses. New York: Cambridge University Press.Find this resource:
Essl, F., Dullinger, S., Rabitsch, W., Hulme, P. E., Pyšek, P., Wilson, J. R. U., et al. (2015). Historical legacies accumulate to shape future biodiversity in an era of rapid global change. Diversity and Distributions, 21, 534–547.Find this resource:
Fahrig, L. (1992). Relative importance of spatial and temporal scales in a patchy environment. Theoretical Population Biology, 41, 300–314.Find this resource:
Fahrig, L. (2013). Rethinking patch size and isolation effects: The habitat amount hypothesis. Journal of Biogeography, 40, 1649–1663.Find this resource:
Fahrig, L., Baudry, J., Brotons, L., Burel, F. G., Crist, T. O., Fuller, R. J., et al. (2011). Functional landscape heterogeneity and animal biodiversity in agricultural landscapes. Ecology Letters, 14, 101–112.Find this resource:
Fahrig, L., & Nuttle, W. K. (2005). Population ecology in spatially heterogeneous environments. In G. M. Lovett, M. G. Turner, C. G. Jones, & K. C. Weathers (Eds.), Ecosystem function in heterogeneous landscapes (pp. 95–118). New York: Springer.Find this resource:
Fischer, J., Lindenmayer, D. B., & Fazey, I. (2004). Appreciating ecological complexity: Habitat contours as a conceptual landscape model. Conservation Biology, 18, 1245–1253.Find this resource:
Forman, R. T. T. (1995). Land mosaics: The ecology of landscapes and regions. New York: Cambridge University Press.Find this resource:
Forman, R. T. T., & Godron, M. (1986). Landscape ecology. New York: Wiley.Find this resource:
Fox, J. W. (2013). The intermediate disturbance hypothesis should be abandoned. Trends in Ecology & Evolution, 28, 86–92.Find this resource:
Fukami, T., Martijn Bezemer, T., Mortimer, S. R., & van der Putten, W. H. (2005). Species divergence and trait convergence in experimental plant community assembly. Ecology Letters, 8, 1283–1290.Find this resource:
Gilpin, M. E., & Hanski, I. A. (1991). Metapopulation dynamics: Empirical and theoretical investigations. San Diego, CA: Academic Press.Find this resource:
Grime, J. P. (1973). Competitive exclusion in herbaceous vegetation. Nature, 242, 344–347.Find this resource:
Grime, J. P. (1979). Plant strategies, vegetation processes, and ecosystem properties. Chichester, U.K.: Wiley.Find this resource:
Grinnell, J. (1917). The niche-relationships of the California thrasher. The Auk, 34, 427–433.Find this resource:
Haddad, N. M., Brudvig, L. A., Damschen, E. I., Evans, D. M., Johnson, B. L., Levey, D. J., et al. (2014). Potential negative ecological effects of corridors. Conservation Biology, 28, 1178–1187.Find this resource:
Haddad, N. M., Hudgens, B., Damschen, E. I., Levey, D. J., Orrock, J. L., Tewksbury, J. J., et al. (2011). Assessing positive and negative ecological effects of corridors. In J. Liu (Ed.), Sources, sinks, and sustainability (pp. 475–503). New York: Cambridge University Press.Find this resource:
Haila, Y. (2002). A conceptual genealogy of fragmentation research: From island biogeography to landscape ecology. Ecological Applications, 12, 321–334.Find this resource:
Hampton, S. E., Strasser, C. A., Tewksbury, J. J., Gram, W. K., Budden, A. E., Batcheller, A. L., et al. (2013). Big data and the future of ecology. Frontiers in Ecology and the Environment, 11, 156–162.Find this resource:
Hanski, I. (1999). Metapopulation ecology. Oxford: Oxford University Press.Find this resource:
Hanski, I., & Ovaskainen, O. (2003). Metapopulation theory for fragmented landscapes. Theoretical Population Biology, 64, 119–127.Find this resource:
Hanski, I., Zurita, G. A., Bellocq, M. I., & Rybicki, J. (2013). Species–fragmented area relationship. Proceedings of the National Academy of Sciences, 110, 12715–12720.Find this resource:
Hubbell, S. P. (2001). The unified neutral theory of biodiversity and biogeography. Princeton, NJ: Princeton University Press.Find this resource:
Hutchinson, G. E. (1957). Concluding remarks. Cold Spring Harbor Symposia on Quantitative Biology, 22, 415–427.Find this resource:
Kupfer, J. A. (2011). Theory in landscape ecology and its relevance to biogeography. In A. Millington, M. Blumer, & U. Schickhoff (Eds.), The SAGE Handbook of Biogeography (pp. 57–74). London: Sage.Find this resource:
Kuussaari, M., Bommarco, R., Heikkinen, R. K., Helm, A., Krauss, J., Lindborg, R., et al. (2009). Extinction debt: A challenge for biodiversity conservation. Trends in Ecology & Evolution, 24, 564–571.Find this resource:
Laurance, W. F., Lovejoy, T. E., Vasconcelos, H. L., Bruna, E. M., Didham, R. K., Stouffer, P. C., et al. (2002). Ecosystem decay of amazonian forest fragments: A 22-year investigation. Conservation Biology, 16, 605–618.Find this resource:
Leibold, M. A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J. M., Hoopes, M. F., et al. (2004). The metacommunity concept: A framework for multi-scale community ecology. Ecology Letters, 7, 601–613.Find this resource:
Levin, S. A. (1992). The problem of pattern and scale in ecology. Ecology, 73, 1943–1967.Find this resource:
Levins, R. (1970). Extinction. Lectures on Mathematics in the Life Sciences, 2, 75–107.Find this resource:
Levins, R., & Culver, D. (1971). Regional coexistence of species and competition between rare species. Proceedings of the National Academy of Sciences of the United States of America, 68, 1246–1248.Find this resource:
Lomolino, M. V. (2000). A call for a new paradigm of island biogeography. Global Ecology and Biogeography, 9, 1–6.Find this resource:
MacArthur, R. H., & MacArthur, J. W. (1961). On bird species diversity. Ecology, 42, 594–598.Find this resource:
MacArthur, R. H., & Wilson, E. O. (1967). The theory of island biogeography. Princeton, NJ: Princeton University Press.Find this resource:
Macdonald, D., & Service, K. (2009). Key topics in conservation biology. New York: Wiley.Find this resource:
MacMahon, J. A. (1981). Successional processes: Comparisons among biomes with special reference to probable roles of and influences on animals. In D. C. West, H. H. Shugart, & D. B. Botkin (Eds.), Forest Succession (pp. 277–304). Springer Advanced Texts in Life Sciences. New York: Springer.Find this resource:
Manel, S., & Holderegger, R. (2013). Ten years of landscape genetics. Trends in Ecology & Evolution, 28, 614–621.Find this resource:
McCook, L. J. (1994). Understanding ecological community succession: Causal models and theories, a review. Vegetatio, 110, 115–147.Find this resource:
McCoy, E. D., & Mushinsky, H. R. (1999). Habitat fragmentation and the abundances of vertebrates in the florida scrub. Ecology, 80, 2526–2538.Find this resource:
McCullough, D. R. (1996). Metapopulations and Wildlife Conservation. Washington, DC: Island Press.Find this resource:
McGarigal, K., Tagil, S., & Cushman, S. A. (2009). Surface metrics: An alternative to patch metrics for the quantification of landscape structure. Landscape Ecology, 24, 433–450.Find this resource:
McGill, B. J. (2003). A test of the unified neutral theory of biodiversity. Nature, 422, 881–885.Find this resource:
McIntyre, S., & Hobbs, R. (1999). A framework for conceptualizing human effects on landscapes and its relevance to management and research models. Conservation Biology, 13, 1282–1292.Find this resource:
McRae, B. H., Dickson, B. G., Keitt, T. H., & Shah, V. B. (2008). Using circuit theory to model connectivity in ecology, evolution, and conservation. Ecology, 89, 2712–2724.Find this resource:
Metzger, J. P. (2008). Landscape ecology: Perspectives based on the 2007 IALE World Congress. Landscape Ecology, 23, 501–504.Find this resource:
Miller, T. E., Burns, J. H., Munguia, P., Walters, E. L., Kneitel, J. M., Richards, P. M., et al. (2005). A critical review of twenty years’ use of the resource-ratio theory. The American Naturalist, 165, 439–448.Find this resource:
Millington, A., Blumler, M., & Schickhoff, U. (2011). The SAGE Handbook of Biogeography. London: SAGE.Find this resource:
Norton, M. R., Hannon, A. S., & Schmiegelow, F. K. (2000). Fragments are not islands: Patch vs. landscape perspectives on songbird presence and abundance in harvested boreal forest. Ecography, 23, 209–223.Find this resource:
O’Neill, R. V. (1986). A hierarchical concept of ecosystems. Princeton, NJ: Princeton University Press.Find this resource:
Parr, C. L., & Andersen, A. N. (2006). Patch mosaic burning for biodiversity conservation: A critique of the pyrodiversity paradigm. Conservation Biology, 20, 1610–1619.Find this resource:
Peterson, G., Allen, C. R., & Holling, C. S. (1998). Ecological resilience, biodiversity, and scale. Ecosystems, 1, 6–18.Find this resource:
Pickett, S. T., & Cadenasso, M. L. (1995). Landscape ecology: Spatial heterogeneity in ecological systems. Science, 269, 331–334.Find this resource:
Pickett, S. T. A., & White, P. S. (1985). The ecology of natural disturbance and patch dynamics. Oxford: Elsevier.Find this resource:
Stein, A., Gerstner, K., & Kreft, H. (2014). Environmental heterogeneity as a universal driver of species richness across taxa, biomes, and spatial scales. Ecology Letters, 17, 866–880.Find this resource:
Thompson, J. N. (1978). Within-patch structure and dynamics in pastinaca sativa and resource availability to a specialized herbivore. Ecology, 59, 443–448.Find this resource:
Tilman, D. (1985). The resource-ratio hypothesis of plant succession. The American Naturalist, 125, 827–852.Find this resource:
Tilman, D., May, R. M., Lehman, C. L., & Nowak, M. A. (1994). Habitat destruction and the extinction debt. Nature, 371, 65–66.Find this resource:
Tjørve, E. (2010). How to resolve the SLOSS debate: Lessons from species-diversity models. Journal of Theoretical Biology, 264, 604–612.Find this resource:
Troll, C. (1971). Landscape ecology (geoecology) and biogeocenology: A terminological study. Geoforum, 2, 43–46.Find this resource:
Tscharntke, T., Tylianakis, J. M., Rand, T. A., Didham, R. K., Fahrig, L., Batáry, P., et al. (2012). Landscape moderation of biodiversity patterns and processes: Eight hypotheses. Biological Reviews, 87, 661–685.Find this resource:
Turner, M. G. (1989). Landscape ecology: The effect of pattern on process. Annual Review of Ecology and Systematics, 20, 171–197.Find this resource:
Turner, M. G., & Bratton, S. P. (1987). Fire, grazing, and the landscape heterogeneity of a Georgia barrier island. In M. G. Turner (Ed.), Landscape heterogeneity and disturbance (pp. 85–101). New York: Springer.Find this resource:
Whittaker, R. H. (1953). A consideration of climax theory: The climax as a population and pattern. Ecological Monographs, 23, 41–78.Find this resource:
Whittaker, R. H. (1967). Gradient analysis of vegetation. Biological Review, 42(2), 207–264.Find this resource:
Whittaker, R. H. (1972). Evolution and measurement of species diversity. Taxon, 21, 213–251.Find this resource:
Wiens, J. A. (1989a). The ecology of bird communities: Processes and variations. New York: Cambridge University Press.Find this resource:
Wiens, J. A. (1989b). Spatial scaling in ecology. Functional Ecology, 3, 385–397.Find this resource:
Wiens, J. A. (1995). Landscape mosaics and ecological theory. In L. Hansson, L. Fahrig, & G. Merriam (Eds.), Mosaic landscapes and ecological processes (pp. 1–26). London: Chapman & Hall.Find this resource:
Wiens, J. A. (2000). Ecological heterogeneity: An ontogeny of concepts and approaches. In M. J. Hutchings, E. A. John, & A. J. A. Stewart (Eds.), The Ecological Consequences of Environmental Heterogeneity (pp. 9–31). Oxford: Blackwell Science.Find this resource:
Wiens, J. A. (2002). Central concepts and issues of landscape ecology. In K. J. Gutzwiller (Ed.), Applying landscape ecology in biological conservation (pp. 3–21). New York: Springer.Find this resource:
Wilson, E. O., & Peter, F. M. (1988). Biodiversity. Washington, DC: National Academy Press.Find this resource:
Wilson, D. S. (1992). Complex interactions in metacommunities, with implications for biodiversity and higher levels of selection. Ecology, 73, 1984–2000.Find this resource:
With, K. A. (2007). Invoking the ghosts of landscapes past to understand the landscape ecology of the present . . . and the future. In J. A. Bissonette, & I. Storch (Eds.), Temporal dimensions of landscape ecology (pp. 43–58). New York: Springer.Find this resource:
Wu, J. (2013). Key concepts and research topics in landscape ecology revisited: 30 years after the Allerton Park workshop. Landscape Ecology, 28, 1–11.Find this resource:
Young, T., Chase, J., & Huddleston, R. (2001). Community succession and assembly: Comparing, contrasting and combining paradigms in the context of ecological restoration. Ecological Restoration, 19, 5–18.Find this resource: