Foundations of Restoration Ecology

By Margaret A. Palmer, Joy B. Zedler, Donald A. Falk and Karen Holl

The practice of ecological restoration, firmly grounded in the science of restoration ecology, provides governments, organizations, and landowners a means to halt degradation and restore function and resilience to ecosystems stressed by climate change and other pressures on the natural world. Foundational theory is a critical component of the underlying science, providing valuable insights into restoring ecological systems effectively and understanding why some efforts to restore systems can fail. In turn, on-the-ground restoration projects can help to guide and refine theory, advancing the field and providing new ideas and innovations for practical application.

This new edition of Foundations of Restoration Ecology provides the latest emerging theories and ideas in the science of restoration ecology. Fully one-third longer than the first edition and comprehensive in scope, it has been dramatically updated to reflect new research. Included are new sections devoted to concepts critical to all restoration projects as well as restoration of specific ecosystem processes, including hydrology, nutrient dynamics, and carbon.  Also new to this edition are case studies that describe real-life restoration scenarios in North and South America, Europe, and Australia. They highlight supporting theory for restoration application and other details important for assessing the degree of success of restoration projects in a variety of contexts. Lists at the end of each chapter summarize new theory introduced in that chapter and its practical application.

Written by acclaimed researchers in the field, this book provides practitioners as well as graduate and undergraduate students with a solid grounding in the newest advances in ecological science and theory.

• Language: English
• Publisher: Island Press
• Release date: Nov 1, 2016
• ISBN: 9781610916981

Book preview

JBZ

PART I

Introduction to Restoration and Foundational Concepts

Chapter 1

Ecological Theory and Restoration Ecology

Margaret A. Palmer, Joy B. Zedler, and Donald A. Falk

Rather than explaining ecosystem structure and function under a single unified theory, ecologists deploy a strong and diverse body of theory to address a wide range of ecological problems (Weiner 1995; Pickett et al. 2007; Hastings and Gross 2012). Theories come in many forms—predictive statements, explanatory concepts, and mathematical and computational models (Scheiner and Willig 2011); yet all share a focus on causal explanation. In restoration, theories help to explain historical events, understand current observations, and predict future states. This last application is particularly important because ecosystems, and the task of restoring them, take place in an increasingly altered world (Steffen et al. 2015). Grounded in theory and empiricism from the ecological sciences, restoration ecology provides the science essential to the practice of ecological restoration, which in turn can be used to test those theories in real world contexts (Palmer and Ruhl 2015; Suding et al. 2015).

What Is Restoration Ecology?

Population, community, and ecosystem ecology are well-established branches of ecological science that focus on specific levels of organization, while restoration ecology is much younger and more comprehensive. As the study of the relationships among organisms and their environment in a restoration context, restoration ecology draws on all branches of ecological science and spans genes to entire landscapes (Falk et al. 2006). The homology with the general definition of ecology is not coincidental; restoration ecology can be thought of as a special domain of ecological research, defined by context. Typically, this context includes a natural system of some kind that has been altered in composition, structure, or function. The central aim of restoration ecology is thus to describe and quantify those departures from a characteristic ecosystem state (including the full range of spatial and temporal variation), understand what drives and regulates them, and then project how the system can be moved back toward a less disturbed state (Hobbs and Suding 2009). Restoration ecology also integrates a number of related disciplines, including hydrology, geomorphology, oceanography, and others, particularly various social science disciplines.

Ecological theories can inform the design, implementation, and assessment of restoration projects that range in area from small sites to watersheds. Conceptual theories tend to be the broadest, such as the theory of evolution by natural selection, or models of macroecology (chaps. 3, 15, and 16). Other theories may be specific to a particular type of ecosystem or group of organisms, such as biogeo-chemical cycling, community assembly, or disturbance ecology (Young et al. 2001; chaps. 2, 9, 12, 13). Theories that employ mathematical or statistical models may take the form of simple equations derived from first principles or complex sets of equations drawn from extensive empirical observations. For example, a recent set of theoretical models links a wide range of organism and community traits on the basis of energetics (Schramski et al. 2015). Predictions from models can be very general: for example, that restoring large or well-connected parcels of land will enhance restoration of biodiversity (Tambosi et al. 2014; chap. 4). They can also be specific to a particular time period or ecological system; for example, restoring seagrass in Middle Tampa Bay (FL) to levels observed in the 1950s requires reducing chlorophyll below some threshold (Sherwood et al. 2016), and restoring grasslands benefits from relating plant functional diversity to nutrient cycling and soil carbon (Bach et al. 2012).

Theories provide us with templates and logic paths for predictions. Theories are used to guide the framing of research questions, the design of experiments, collection of data, and ways to organize information to understand the natural world. Theories can be used to explore the impacts of assumptions we might make about ecosystems, and deviation from theories can help inform future research. For all of these reasons theory is fundamental, not only to restoration ecology but to the practice and advancement of ecological restoration.

What is Ecological Restoration?

The Society for Ecological Restoration (SER 2004) defines ecological restoration as, the process of assisting the recovery of an ecosystem that has been degraded, damaged, or destroyed. Restorationists attempt to move the composition, structure, and dynamics of a damaged system to an ecological state that is within some acceptable limit relative to a less altered and (probably) more sustainable system (Falk 1990; Allen et al. 2002). Informed by the work of Clewell and Aronson (2013), we simplify the SER (2004) framework and attributes of restored systems to define science-based ecological restoration as in Palmer and Ruhl (2015) (table 1–1).

In this simplified framework, features refer to the structural components of an ecosystem. For example, floodplains are a key component of river ecosystems; their connection to the water and land is an aspect of pattern. Similarly, the structure of a forest includes tree size classes and canopy properties; one aspect of its pattern is the spatial distribution of trees on the landscape. Processes (also called functions) include a wide range of dynamic attributes, such as primary production, stream discharge, fire regimes, dispersal and migration, population dynamics, and biogeochemical cycling. Processes vary over time and space, and ideally lead to recovery of a self-sustaining dynamic system that requires less human intervention than during the process of restoration (Beechie et al. 2010).

TABLE 1-1.

The framework for defining ecological restoration specifies that it aims to recover the properties of an intact system such as species assemblages, food webs, and functional attributes similar to reference systems (chaps. 7, 8, 9, and 11). However, restoration can take decades, and even when a design is science based, unexpected alternative states or incomplete recovery may result (chap. 2). An unexpected outcome is different from knowingly targeting an end state other than full ecosystem recovery and fidelity to an appropriate reference system (Clewell 2000; Egan and Howell 2005). Attempts to reverse environmental degradation that are not ecological restoration include the use of hardscapes or nonnative species to reduce excessive soil erosion and run-off, and other types of engineered systems that cannot be self-sustaining given their design or placement in a highly modified landscape context, such as strip mines, chemically polluted brownfields, or severely eroded sites (Palmer and Ruhl 2015). Other examples of projects with limited objectives (Suding et al. 2015) include maximizing a single ecosystem service, such as stabilizing a steep slope using a monoculture of nonnative, deep-rooted trees (Mao et al. 2012), or postmining reclamation of a formerly forested region to a nonnative grassland (Yeiser et al. 2016). So long as objectives are not too narrowly focused (e.g., on a single social goal), recovering a broad range of ecosystem services is possible and the chances of this may be increased if efforts are invested in maximizing functional biodiversity and associated ecological processes (chap. 3).

A fully restored ecosystem is inferred to be self-sustaining and resilient; that is, it has the capacity for recovery from expected change and stress (SER 2004). In cases where landscape-scale processes no longer occur naturally, restoration can compensate for some constraints on self-sustainability by reintroducing the missing process (chaps. 4 and 16). Examples include controlled burning to restore grasslands or forests, and flood pulsing to restore riparian habitat and stream channels. Moreover, invasion by an aggressive nonnative species or an uncharacteristic disturbance may trigger the need for ongoing maintenance (Shaish et al. 2010; Dickson et al. 2014). In such cases, restoration sensu stricto may never be finished. It is uncertain whether fully restored ecosystems will be resilient to all future stressors, especially with changing climate and other stressors that occur at a greater rate or magnitude than the system has experienced over recent evolutionary time (chaps. 5, 6, 15, and 17).

The Restoration-Theory Linkage

The acid test of our understanding is not whether we can take ecosystems to bits on paper, however scientifically, but whether we can put them together in practice and make them work.—A. D. Bradshaw (1983)

Restoration ecology and ecological restoration are reciprocal concepts. Ecological theory informs the practice of restoration but the converse is also true: restoration science and practice can contribute to basic ecological theory (Bradshaw 1983; Jordan et al. 1987; Perring et al. 2015). Ecological restoration is especially useful for testing theories associated with understanding the processes that govern ecosystem trajectories (assembly rules, postdisturbance succession, alternative stable states; chaps. 2, 8, and 9). For example, work in Poland showed that restoration of drained fens may result in communities with different levels of plant functional diversity due to the differential effects of competition and habitat filters (Hedberg et al. 2014). This work demonstrated that knowing which filters act in a particular setting is essential to predicting which species or functional groups are likely to dominate (chaps. 3, 6, and 9). Efforts to restore natural fire regimes in forested communities have informed general understanding of fire ecology where disturbance dynamics have been disrupted (Falk et al. 2011; Young et al. 2015). Restoration studies have informed our understanding of the link between biodiversity and habitat heterogeneity or complexity (Bell et al. 1997; Zedler 2000; Palmer et al. 2010; chap. 10). Because restoration scientists work, by definition, in systems that have been disrupted, their observations and experiments have especially informed—and been informed by—theories of the ecology of disturbance (Temperton et al. 2004; Lake 2013; chap. 9).

As Bradshaw’s (1983, 1987) famous remark anticipates, the use of restoration research to test theory and challenge dogma has grown dramatically (Young et al. 2005; Zedler et al. 2007). For example, recent work by Ford et al. (2015) showed that the expected trophic cascade effect of restoring a top predator (Kenyan wild dogs that significantly reduced a dominant ungulate herbivore) did not increase tree abundances, even though the herbivores were known to suppress tree abundance and despite a positive correlation between trees and dog abundance. The authors suggest alternative hypotheses including significant time delays in indirect effects and the possibility of a reticulate food web such that, once the dominant herbivore declines, herbivory by other species increases. In a very different type of ecosystem, work by Hamilton et al. (2014) supported ecological theory linking dietary niche breadth to the size structure of a predator population. Fishing pressures selectively removed large sheepshead fish in a California kelp bed, but when the size structure of the fished population was restored, the predator’s dietary niche expanded, with implications for urchins, algae, and kelp.

Conducting large-scale experiments that test basic ecological theory while restoring a site simultaneously can advance both the practice and the science of restoration (Zedler and Callaway 2003). For example, a nitrogen-addition experiment coupled with restoration of an endangered tidal marsh plant demonstrated that nutrient levels affected the annual plant’s abundance, suggesting that in some instances strategic modification of biogeochemical properties can reinforce species-level responses (Parsons and Zedler 1997). Efforts to reintroduce wildland fire as a keystone ecosystem process have enabled forest scientists to study fire effects on vegetation dynamics, biogeochemical cycling, and carbon sequestration in more detail than would be possible in uncontrolled wildfires (Schoennagel and Nelson 2010). Working on strip-mined areas in Brazil, Silva et al. (2015) showed that nutrient limitation, plant community composition, and microbially mediated biogeochemical reactions interacted to determine soil development, carbon sequestration, and restoration trajectories. Their tests included trait-based, functional ecology theory, which Laughlin (2014) recommended for experimentation in restoration sites to advance both theory and practice.

Ecological theory and contemporary modeling approaches can also be coupled to explore ways to enhance restoration projects. As an example, a recent study by Crandall and Knight (2015) used spatially explicit modeling to explore factors that may weaken the positive feedbacks that often allow exotic species to replace natives. Theory suggests that dominance by exotics is due either to major fitness advantages, or because they create positive feedbacks that benefit conspecifics more than individuals from other species. As the population size of the exotic species increases, self-reinforcing feedbacks may become stronger, making it difficult to eradicate nonnatives (Stevens and Falk 2009; Larkin et al. 2012; Yelenick and D’Antonio 2013). Crandall and Knight’s (2015) work suggests that once exotics have become established and dominate a system, it is less likely that the system can be moved back to a native state unless disturbance strongly reduces the fitness of the exotic relative to the native. This theoretical work implies that intervention before an exotic becomes dominant is essential, but in later stages of invasion, experimenting with different disturbance regimes, perhaps implementing them even more frequently than was the case historically, can be productive (e.g., manipulating disturbance; chap. 8).

These examples illustrate our fundamental premise: ecological restoration benefits from a strong grounding in theory, while at the same time ecological theory benefits from the unique opportunities to test theory in restoration contexts. Specific examples of this reciprocity are provided throughout this book, covering major areas of ecological theory spanning multiple levels of ecological organization from genetics to whole ecosystems (table 1–2).

Foundations of Restoration Practice

Some view ecological restoration as an art or a skill honed by practice, experience, and tutelage. As we have suggested, ecological theory is a foundation for restoration practice; however, restorationists can also learn from empirical vernacular experimentation and traditional ecological knowledge (TEK) (Martinez 2014). Both modes of learning develop over time, based on varying trials (intentionally or otherwise) and selecting approaches that best achieve desired outcomes, informing others of advances and adapting practices to new knowledge. Both approaches also identify what works over multiple trials, which can sometimes extend over many years. Together, science-based, experiential, and TEK approaches can guide restoration goals, treatments options, and experimental designs (Rieger et al. 2014). Implementation of the project should be accomplished in an adaptive management framework in which scientific monitoring informs each step of the process, including the need for additional actions to move the restoration project forward (fig. 1–1).

As a general principle, the first step in restoring an ecosystem is to remove or at least reduce causes of degradation so the system can begin to recover on its own via natural processes (Batchelor et al. 2015). Following this, the preferred or lowest-cost approach to the restoration of degraded ecosystems is often to allow them to recover on their own. This approach, sometimes referred to as passive restoration, is based on the premise that natural systems have their own recovery pathways, mechanisms, and timetables, which may not be mirrored in human-driven designs or implementation. In a sense, this approach is a null test of the potential for spontaneous recovery without human intervention. For instance, many forests recover essential attributes following low-severity wildfire, because biota are pre-adapted to such events (Keeley et al. 2011). Healthy stream systems adjust their channel morphology in response to flooding and seasonal variation in streamflow as it interacts with sediment inputs and redistribution; river forms remain dynamically stable if this process is not interfered with by human actions such as building of dams or levees to constrain channels (Wohl et al. 2015). Perhaps the classic example of passive restoration is fisheries management, when populations can recover spontaneously (Hilborn and Ovando 2014) once overfishing or harmful harvest practices are eliminated. Similarly, eradication of nonnative mammalian predators on islands (>800 projects to date) allowed passive recovery of seabird colonies with stable metapopulations on New Zealand islands near source populations (Buxton et al. 2014). Likewise, removal of livestock grazing led to passive recovery of native riparian vegetation in rangelands of the US Central Basin (Hough-Snee et al. 2013), and fencing areas to limit human access resulted in the recovery of Mediterranean coastal dunes (Acosta et al. 2013).

Figure 1–1. Steps in ecological restoration are informed by theory and methods of restoration ecology science. Photo on lower right courtesy of Joshua Viers.

More often, however, restoration requires multiple efforts, because species, sites, or their landscapes, have been pushed beyond the ability to recover passively. In these cases, the preferred approach is process-based restoration, which is well grounded in ecological theory. Here, actions aim to restore underlying processes that create and maintain ecological systems, such as dispersal, biogeochemical cycling, hydrologic dynamics, watershed infiltration capacity, and fire, rather than micromanaging every aspect of community composition and structure (Zedler 1996; Beechie et al. 2010). The water-flow regime is a master variable that influences almost all aspects of stream and river ecosystems; thus, restoring flows affects almost every other aspect of stream restoration from a biophysical and organismal point of view (chap. 14). Similarly, fire regimes are integral to the restoration of many forests, woodlands, and grasslands; restoring this keystone process can allow the systems to reequilibrate in other respects such as forest structure and composition (Bond and Keeley 2005; Falk 2006). Process-based restoration is far more likely to result in a more sustainable system than efforts focusing only on compositional or structural elements, such as the shape and size of a stream channel or planting the desired species in soils that cannot supply sufficient nitrogen. Without restoring the critical processes that maintain these ecosystems, ongoing maintenance will likely be required, for example, to clean the water, maintain species metapopulations, or maintain forest structure (Palmer et al. 2014; Roccaforte et al. 2015; chaps. 4 and 7).

A comprehensive restoration approach that focuses on restoration of both processes and structures is, of course, ideal and supported strongly by ecological theory and empirical observation. One of the first large-scale grassland restorations began at the University of Wisconsin–Madison Arboretum, where land dedicated by Aldo Leopold in 1934 achieved diverse prairie via combinations of plantings and prescribed burning (Meine 1988) (fig. 1–2). Unfortunately over time, urbanization increased runoff and eutrophication, and regulations reduced opportunities for burning. Woody and weedy plants expanded and displaced native plants. Recent research has quantified links between structure and processes in experimental wetlands in the Arboretum, showing that prolonged hydroperiod alone fostered cattail invasions that decreased plant diversity and increased discharges of dissolved phosphorus, contrary to project design. Adjacent wetlands with higher infiltration rates had less cattail cover, more diverse vegetation, and lower rates of nutrient discharge (Doherty et al. 2014).

Figure 1–2. Curtis Prairie (University of Wisconsin–Madison) was cultivated and then used to pasture horses before being planted to establish tallgrass prairie, from 1935 to the 1950s (left photo). In a 1966 census, P. and J. Zedler recorded 212 native plant species and 33 exotics (e.g., center photo); in 2002, T. Snyder found 230 natives and 35 exotics. The largest shift during those thirty-six years was the expansion of shrubs, notably the native gray dogwood, Cornus racemosa, which increased from 15% to 53% frequency of occurrence in square-meter plots. The 2011 control burn (right photo) left many live woody stems and rhizomes. In 2015, mowing was added to burning to control woody plant invasion. Management to achieve tallgrass prairie continues, despite eighty years of restoration effort. Photo collage courtesy of Sarah Friedrich.

Ecological Theory and the Reference Concept

The concept of a reference system is a bedrock principle in restoration, and such systems often serve as a template to guide restoration and to identify when recovery is complete (Egan and Howell 2005). However, in practice the most important role for reference systems may be to provide ecological information about the system of interest (Higgs et al. 2014). A scientific analysis of historical or contemporary environmental conditions within which such a system exists (existed) involves the use of ecological theories (e.g., on species interactions, biogeochemical fluxes, disturbance regimes, or dispersal) and applicable experiments to determine whether the composition, biophysical processes, or structural aspects of a potential restoration site are outside the reference range (fig. 1–3). The historical or contemporary range of variability in species assemblages, biophysical processes, and ecosystem features derives from the factors that control system states, the rate of change in response to environmental change, and the ongoing direction of change (Gildar et al. 2004; Wiens et al. 2012).

In fact, ecological restoration has never been limited to the literal use of historical conditions as a target for restoration (i.e., returning to the past) (Clewell 2000). Some projects may attempt to recover species native to a region, but the modern science of restoration emphasizes that the ability to do this is highly context dependent (Zedler et al. 2012). If the range of conditions necessary to support historical species is not readily recoverable, then more aggressive interventions or alternative targets may need to be explored (e.g., functional targets; chap. 3). Even in these cases, however, the role of reference systems remains: to provide a guide to the dynamics of the degraded system, in essence suggesting where that system would be currently had it not been disrupted (Falk 1990). In the end, maintaining or restoring resilience or adaptive capacity may be as, or more, important than literal historical authenticity (chaps. 2, 15, and 17).

Identifying the appropriate spatial and temporal scale in evaluating ranges of variability is not a trivial task (White and Walker 1997). How long a history and how large an area provide a relevant baseline for a particular restoration site? Some have suggested that the most appropriate spatial scale for identifying the historical range of variability should be where the intensity of ecological interactions, system components, or states of environmental variables changes substantially, for example, major discontinuities or steep gradients (Post et al. 2007). Rapid temporal shifts in the environment might help explain a major state change, for example, when an earlier state was sustained by processes and ecological components that no longer exist, or that may not persist into the future (Jackson and Hobbs 2009). This approach can be combined with projections of climate-induced changes in dominant system processes or major environmental variables, such as temperature, to identify when historical bounds are likely to be exceeded (Mora et al. 2013). Such analyses may determine the extent to which historical ranges should influence the restoration approach and its goals (Maschinski and Haskins 2012; chap. 17).

Figure 1–3. Ecological systems and their structural and functional attributes are dynamic in time and space. Understanding such variability is critical to determining if a system needs restoring and to what extent. Once the system is restored it is unlikely initially to occupy the same ecological space as predisturbance. The figures show the distribution of some ecological attribute, such as native species richness, for a reference ecosystem (black) and for a second ecosystem of interest (gray). In the prerestoration state (left panel), variability in species richness for the ecosystem (gray) is outside the range of variability for reference systems, indicating that the ecosystem needs restoring. In the postrestoration state (right panel), the two distributions overlap, indicating that the system has been restored (but note that both systems have changed). These images highlight how critical it is to recognize that reference systems are themselves not static, and also that restored systems need not be identical to a particular reference system in every respect.

Other ecologists argue that the field has advanced sufficiently that selection of a single spatial or temporal scale of reference makes little sense; rather, we are equipped with theory to integrate across scales, including ecological to evolutionary timeframes and organism to ecosystem targets (Chave 2013). Given the growing availability of environmental data along with new analytical tools and advanced computer capabilities, macrosystems ecologists (Heffernan et al. 2014) now use dynamic linear modeling methods to explore how broad scale and local phenomena interact, and they can predict how patterns and processes at local scales are likely to respond to environmental changes at multiple scales (Levy et al. 2014). Methodological advances such as the use of stable isotopes for food-web analyses have revealed linkages across spatially distinct ecosystems that can inform restoration designs (chap. 11), and advances from combining species distribution modeling with methods from landscape genetics have resulted in a better understanding of what restoration outcomes are possible under future conditions (chap. 5). These examples suggest that we view restoration not simply in the context of what is or is not possible given the pace of global change; rather, we ask how consideration of ecological and evolutionary processes such as metapopulation dynamics, trophic interactions, dispersal, range shifts, and microevolution can be integrated to address conservation and restoration issues even in the face of uncertainty (chaps. 4, 7, 15, 16, and 17).

The difference between variability (an ecosystem property) and uncertainty (a sampling and prediction property) is a key issue, not only in the analysis of reference conditions but more broadly in all restoration research. In some ecosystems (for example, conifer forests in western North America), certain historical reference conditions can be determined with a high degree of accuracy, such as species composition and the mean fire return interval (Friederici 2003). But many other historical properties are not known or even knowable, especially fine-scale ecological composition, structure, and processes (table 1 in Falk 2006), such as stream geomorphology or water chemistry, local-scale species distributions, dispersal and migration pathways, predation patterns, reproductive output, or disease outbreaks. The problem becomes a kind of ecological Heisenberg effect in the sense that the earliest documented observations were made while systems were being significantly altered, such as in North America following European settlement.

Using contemporary ecosystems (spatial references: same time, another place) would seem to avoid some of these problems of historical references (White and Walker 1997). Least impacted contemporary ecosystems may represent ongoing interactions of biotic and abiotic elements, all functioning in complex ways not directed by humans. Moreover, their properties can be observed in detail and can be measured whenever researchers need data. However, contemporary references are confounded by their own significant problems. First, the supposed reference may already be altered in unknown ways, due to pervasive drivers that affect the region and the target system, such as climate, altered species distributions, or land use. Second, the site may not be a good analogue for the target system, in which case the information derived may not be appropriate for the restoration setting. Third, both systems may be nonequilibrial, meaning that, even if they are comparable, they may reflect alternative successional or metastable states. This issue could be important in ecosystems characterized by high temporal variability and a wide range of potential postdisturbance successional sequences and alternative states (Fletcher et al. 2014).

The Imperative to Test and Advance Theory

The challenges associated with using contemporary ecosystems to define reference conditions become even greater when considered in the context of the pace of anthropogenically induced environmental change. It is thus critical to consider the role ecological restoration can play within the broader context of strategies to cope with such change (Aronson and Alexander 2013). Ecological restoration is an essential component of a sustainability agenda but it is not sufficient to ensure the future health and well-being of natural systems and the people they support. Global human impact continues to expand, vital resources, such as freshwater and arable soils, undammed rivers, and wild forests are increasingly threatened and depleted (Freedman 2014). Conservation is always preferable to resource degradation followed by restoration, but, even with this, it may not be possible to hold the line given rates of resource consumption on our crowded planet (Corlett 2015). In some cases, engineered solutions may be needed even though they do not bring the diverse benefits that conservation provides and that restoration has the potential to provide. To realize this potential, scientists are challenged to ask under what circumstances can we grow a science of restoration ecology that is soundly rooted in ecological theory? One way to address this challenge is to ask which research questions or settings require an extension of our theories and models or even the development of theories de novo. Extending existing theory and developing new theory to guide restoration is difficult, because restoration typically takes place across a multidimensional spectrum of specific sites within various kinds of landscapes, and where goals range from highly specific (e.g., enhance the population of one rare animal species at a specific site) to general (e.g., encourage development of a diverse and complex ecological community over a large landscape).

The need to align ecological restoration closely with a sound theoretical base is imperative for at least three reasons. First, growing demand has made restoration a booming business that requires the support of a knowledge base and research innovations. Billions of dollars are spent annually to restore polluted and sediment-clogged streams (BenDor et al. 2015,) and to reforest lands in areas that have been degraded or fragmented (Rodrigues et al. 2011), yet many restoration efforts are still trial and error improvisations. Second, ecosystem restoration can regain essential ecosystem services such as soil fertility, carbon sequestration, and water purification (chaps. 10, 12, 13, and 14). The stakes are too high not to restore ecosystem integrity wherever possible, especially because services are often what motivate the public and policy makers to invest in restoring natural systems (Aronson et al. 2007; Schaefer et al. 2015). In certain regions of the world, such as the global tropics, human well-being cannot be separated from the sustainability of ecosystems (Lamb et al. 2005). A third reason to enhance the linkage between ecological theory and restoration is to grow the field of ecology. Regardless of what motivates ecologists, their research will likely benefit by being tested in a restoration context, as described earlier and throughout this book. Clearly, practice can be used to both test and grow ecological theory and basic understanding (Young et al. 2005). Whether realized or not, virtually all restoration projects rely on ecological theory, from the initial stages of envisioning a project through completion.

Closing Remarks

Some theories and concepts are so fundamental to understanding the dynamics of ecological systems and how to manage them that they will always be relevant to restoration. Many of these relate to the critical role that diversity, in all its forms, plays in the structure and continuity of populations, communities, and ecosystem processes. Others relate to historical legacies and contemporary context combined with biotic and abiotic filters that can result in persistent alternative states or unanticipated trajectories following the implementation of restoration actions. As we describe in detail in the last chapter of this book, these fundamental concepts and theories are the basis for persistent themes that can be found in both editions of this book; seven such themes are discussed (table 18.1). However, advances in ecological science have been substantial since the last edition and thus new, emergent themes, are apparent in this edition, which are also outlined in the last book chapter (table 18.2). A number of these have been fostered by quantitative advances in modeling, development of new analytical or sampling methods, or creative application of existing statistical methods. Other themes have emerged as a result of growing societal interest in the provision of ecosystem services and the recognition that the current pace of environmental change challenges some conventional assumptions about what it means to restore a system. Readers may wish to read chapter 18 both before and after the other book chapters in order to develop their own list of fundamental, timeless concepts and ask if the growth in ecological science reflected in emerging themes can help meet the growing global imperatives.

References

Acosta, A. T. R., T. Jucker, I. Prisco, and R. Santoro. 2013. Passive Recovery of Mediterranean Coastal Dunes Following Limitations to Human Trampling. In Restoration of Coastal Dunes, edited by M. L. Martinez, J. B. Gallego-Fernández, and P. A. Hesp, 187–98. Verlag, Germany: