1. What is restoration ecology?
2. Concepts in restoration ecology
3. Key steps in ecological restoration
4. Repairing damaged ecosystem processes
5. Directing vegetation change: succession and assembly rules
7. Landscape-scale restoration
8. Prevention versus restoration
Restoration ecology is the science underpinning the practice of repairing damaged ecosystems. Restoration ecology has developed rapidly over the latter part of the twentieth century, drawing its concepts and approaches from an array of sources, including ecology, conservation biology, and environmental engineering. We are faced with an increasing legacy of ecosystems that have been damaged by past and present activities, and it is increasingly recognized that, in many situations, successful conservation management will need to include some restoration. This may take many different forms, such as the reintroduction of particular species, removal of problem species such as weeds or feral animals, or the reinstatement of particular disturbance regimes (including fire and flood regimes).
alternative stable state. A relatively stable ecosystem structure or composition that is different from what was present before disturbance
disturbance. Episodic destruction or removal of ecosystem components
resilience. The ability of an ecosystem to recover following disturbance
restoration. The process of assisting the recovery of an ecosystem that has been degraded, damaged, or destroyed
succession. The process of vegetation development following disturbance, often characterized by relatively predictable sequences of species replacement over time
threshold. A situation where there has been a nonlinear (i.e., sudden or stepped) change in the ecosystem in response to a stress or disturbance, which is often difficult to reverse
Restoration ecology is the science behind the term ecological restoration, which covers a range of activities involved with the repair of damaged or degraded ecosystems and is usually carried out for one of the following reasons:
Ecological restoration occurs along a continuum, from the rebuilding of totally devastated sites to the limited management of relatively unmodified sites, and hence merges with conservation biology. Restoration aims to return the degraded system to a less degraded state that is valuable for conservation or other use and that is sustainable in the long term.
An array of terms has been used to describe these activities, including restoration, rehabilitation, reclamation, reconstruction, and reallocation. Generally, restoration has been used to describe the complete reassembly of a degraded system to its undegraded state complete with all the species previously present, whereas rehabilitation describes efforts to develop some sort of functional or productive system on a degraded site. In addition, some authors use the term reallocation to describe the transfer of a site from one land use to a more productive or otherwise beneficial use. However, the term restoration is often used to refer broadly to activities that aim to repair damaged systems.
Ecosystem characteristics to be restored can include the following:
Restoration has often been viewed as returning an ecosystem or community back to a previous state, i.e., the ecosystem that existed at the site before human disturbance or alteration. However, ecosystems are naturally dynamic entities, and hence, the setting of restoration goals in terms of static compositional or structural attributes is problematic. Often, past system composition or structure is unknown or partially known, and past data provide only static snapshots of system parameters. Current undegraded reference systems can therefore act as potential reference systems against which the success of restoration efforts in degraded systems can be measured. An alternative approach is to explicitly recognize the dynamic nature of ecosystems and to accept that there is a range of potential short- and long-term outcomes of restoration projects. Increasingly, the focus is on having a transparent and defensible method of setting restoration goals that clarify the desired characteristics for the system in the future rather than in relation to what these were in the past. Using past characteristics to guide restoration is still useful, but there is increasing recognition that continuing environmental change, including climate change, means that returning ecosystems to past states may not always be possible.
Where it is impossible or extremely expensive to restore composition and structure, alternative goals may be appropriate. These may aim to repair damage to ecological function or ecosystem services or to create a novel system using species not native to the region or suited to changed environmental conditions. Often, partial restoration or the development of alternative ecosystems with some desirable elements of structure, composition, or function can have positive conservation outcomes. For instance, plantations of timber trees may not develop all the characteristics of a native forest but still may be used by some fauna species—especially if plantation management is modified slightly to improve their value as habitat. Clearly, however, a risk analysis is needed to ensure that the changes do not lead to further problems in the future—for instance, using nonnative species in restoration may lead to these species becoming problematic in the future.
Disturbance, or episodic destruction or removal of ecosystem components, is an integral part of the functioning of many ecosystems. Disturbance often initiates massive ecosystem change and triggers a period of regeneration or recovery. Disturbance is a natural feature of many ecosystems, and disturbances range in extent and severity from localized events such as animal diggings or individual tree falls to large events such as catastrophic fires, large storms (hurricanes, cyclones), and floods. Ecosystems have a degree of resistance to disturbance, termed inertia. In other words, ecosystems can absorb a certain amount of disturbance or stress and remain more or less unchanged. For instance, a lowintensity fire in a forest may only burn surface litter and leave the major components of the ecosystem intact. Or a river system may be able to tolerate a certain level of pollution without undergoing large changes in its biota. When the disturbance is large enough, however, this inertia is overcome, and the system changes. The disturbance could be a discrete event such as a wildfire or windstorm, or it could be an accumulated chronic impact such as increasing pollution. The ability of the system to recover from that change is termed resilience. A resilient system will be able to recover quickly after a disturbance or when a degrading factor is removed and will return to more or less the same structure and composition as was present previously. Ecological restoration is required only where the system’s resilience has been diminished in some way or where the normal recovery processes are too slow to achieve management goals within a desirable time frame. As an example, an arid system that has been overgrazed loses its capacity to regulate water flows because the grazing removes the mosaic structure of plants and debris that intercepts surface flows. As a result, erosion occurs, water does not enter the soil, and conditions are not suitable for plant establishment. Simply removing grazing will not initiate recovery—instead, some active intervention to reinstate some surface heterogeneity is required to “kick-start” the process of recovery.
Human activities frequently either modify the original disturbance regimes (e.g., by changing fire or grazing regimes) or add a further set of disturbances that the ecosystem had not previously experienced. In some cases, this human disturbance pushes ecosystems beyond the limits of their resilience. It is in these cases that active restoration is required.
Ecosystems become degraded when human use or alteration modifies ecosystem characteristics such that ecosystem structure and/or function is changed beyond acceptable limits. For instance, vegetation structure may be altered so that it no longer provides adequate habitat for a range of animal species, or the ecosystem may no longer provide ecosystem services, such as provision of clean water or production of food or fiber. Degradation may result in changes to the biological component of ecosystems or more fundamental changes to the system’s physical or chemical characteristics.
Following disturbance, ecosystems undergo a process of recovery known as succession. Succession describes the sequence of species and groups of species that are present at various times since a disturbance. Pioneer species appear early in the recovery sequence and are able to tolerate open, often harsh conditions. Later successional species are either slower growing or appear only after conditions are modified by pioneer species. The successional trajectory describes the direction and rate of change. Relay floristics describes the process whereby species appear in a recognizable sequence, whereas initial floristic composition relates to the situation in which all species that will take part in the successional sequence appear shortly after a disturbance but grow and/or assume importance at different rates. Species may either facilitate, inhibit, or tolerate other species.
In some ecosystems, a relatively predictable postdisturbance recovery sequence can be expected, but in others, there is the possibility that recovery will follow different trajectories and result in different ecosystem compositions. The trajectory followed may depend on the arrival of particular species in the system, the method of management imposed, or the sequence of climatic and disturbance events during the recovery process. The ecosystem may reach a state from which little further change occurs. This state is known as an alternative stable state, which means that the ecosystem has developed a relatively stable structure or composition that is different from what was present before the disturbance. The presence of such a state often indicates the operation of system thresholds that need to be crossed before further system change can occur. A threshold usually indicates a situation where there has been a nonlinear change in the ecosystem in response to a stress or disturbance: often it may be relatively easy to degrade an ecosystem past such a threshold but much more difficult to restore the ecosystem to a less-degraded state (figure 1). For example, in some ecosystems, progressive addition of nitrogen via air pollution can tip the balance between native plant species and invasive grass species. Once invasive grasses are dominant, they prevent the reestablishment of native species, and the ecosystem becomes stuck in an altered state. Reversing this situation involves not just removing the invasive species but also dealing with the elevated nutrient status of the site. Hence, dealing with such thresholds may involve quite intensive management. The identification of system thresholds is an important element of assessing the appropriate restoration measures needed in any given situation.
Figure 1. Processes involved in ecological restoration.
Ecosystem degradation can result in changes to either the biological component of an ecosystem or its abiotic (physical and/or chemical) characteristics. Biotic changes can include loss of particular species or changes in vegetation structure and composition, whereas abiotic changes can include changes in substrate physical or chemical characteristics or alteration of the hydrological regime. Damage to primary ecosystem processes may result in more fundamental system changes than simple biotic changes. Correct assessment of which ecosystem characteristics have been altered during degradation is essential for effective restoration.
Even with correct assessment and treatment of the problems leading to degradation, it may be impossible to return a system completely to its predisturbance state. System recovery may follow a different path from that taken during system decline (hysteresis), and the resulting system may thus differ from the original. Natural ecosystems are also naturally dynamic and hence constantly changing. This also makes it unlikely that the recovering system will return to exactly the same composition and structure as the predisturbance system. Thus, it is important to set realistic restoration goals that take into account the dynamic nature of the ecosystems we are trying to restore.
There are a number of key steps in any restoration program that need to be undertaken to ensure that useful outcomes are achieved (figure 2). These include setting clear goals with associated success criteria, correctly identifying the factors limiting system recovery or leading to further degradation, and instigating restoration activities that reverse or ameliorate these factors. These activities have to be placed in the context of broader management objectives and monitored to ensure that progress is being made toward the agreed goals.
Setting clear and achievable goals is an important element that is often overlooked but that greatly facilitates the process of deciding on restoration options and monitoring progress. Almost anything is possible if enough money and resources are available, but generally, goals have to be selected on the basis of costeffective measures to overcome limiting factors to allow reasonable goals to be achieved. Goals broadly relate to the restoration of ecosystem function and/or ecosystem structure or composition. The choice of restoration options needs to be guided by both the ecological constraints operating in any given situation and the range of individual and societal goals.
Figure 2. Summary of ecosystem degradation and restoration in terms of alternative states and transitions between them. States are indicated in boxes, and possible transitions between states are shown by arrows. Hypothesized thresholds that prevent transition from a more degraded state to a less degraded state are indicated by the vertical gray bars; such transitions can be either biotic (for instance, competition or grazing) or abiotic (for instance, changed physical or chemical conditions).
Often goals for restoration are set in relation to a particular ecosystem or species composition—for instance, a goal for postmining restoration in a forest ecosystem may be the return of the complete forest ecosystem with all the species that were there previously, or it may be simply the stabilization of the mined surface and the return of the major tree species. An alternative goal for a postmining ecosystem may be the creation of a grass pasture that can be used for livestock production.
Although the cause of degradation may be obvious in some cases (e.g., in mine sites), the factors that are important in influencing system recovery may not be so obvious. A clear identification of the environmental factors that are either causing ongoing degradation or preventing system recovery is essential. Failure to do this properly can result in costly mistakes that do not fix the problem. These factors can affect either primary abiotic processes such as nutrient and water retention or biotic processes such as plant recolonization and survival. As an example, in a mine-site restoration project, there is no point in replanting the area if there are problems with the stability or chemical composition of the substrate. Similarly, in an area where the hydrological regime has changed, restoration of an area of riparian habitat may be ineffectual if the broader hydrological condition of the watershed is not also considered.
Once goals have been defined and methods identified for overcoming the degrading or limiting factors, the restoration project has to be actually implemented. This involves the practical considerations surrounding logistics, budgets, timing of operations, etc. Implementation includes not only the initial activities early in restoration but also ongoing management of the restored site to ensure that the restored system continues to develop along an appropriate trajectory that will result in the restoration goals being achieved.
Monitoring progress is essential to the success of a project but is often not carried out effectively. The choice of variables to be monitored must relate to the goals set for the restoration. Variables must also be relatively inexpensive and simple to monitor. Restoration projects often have success criteria set that have to be met to satisfy contractual or legal obligations. Hence, tracking progress toward success is important. Monitoring can also allow adaptive management, which can identify situations where the management treatment is not having the desired result and hence can be changed.
Ecosystem processes include the cycling and retention of nutrients, carbon, and water. These processes are essential to ecosystem function. Degraded systems frequently have less control over ecosystem flows, and the reinstatement of structures and processes that regulate these flows is often a first step in restoration. This is also true in areas such as mine sites, where ecosystems have to be established de novo.
Removal of vegetation cover and degradation of soil structure can influence the ability of an ecosystem to mediate system flows. This can be manifested as increased erosion and loss of nutrients, altered hydrology leading to increased runoff, rising water tables, flooding, and salinization. Physical manipulation of the substrate (e.g., modifying soil structure or surface microtopography), introducing physical barriers (e.g., brushing), or reestablishing vegetative cover may be necessary to reinstate local control of water and nutrient flows. Remediation of chemical composition of soil or water may also be necessary to facilitate the reestablishment of vegetation.
An ecosystem consists of both biotic and abiotic components, which are interlinked via transfers and flows of nutrients, energy, and water. Restoration of primary processes requires attention to both the abiotic and biotic components. Attempts to reestablish vegetation on areas where primary processes have not been repaired are likely to fail. On the other hand, attention to the biotic component can also speed up repair of primary processes. For instance, nutrient capture may be enhanced by the reestablishment of mycorrhizae or the inclusion of plant species with an array of different root architectures.
As discussed above, successional processes result in the redevelopment of vegetation on an area following a disturbance. In order for this to occur, plant propagules have to be available at a site (via dispersal or from seed stored in the soil or held on the canopy of adult trees), the site has to be suitable for establishment, and the species have to be able to grow and reproduce. In small disturbed areas surrounded by native vegetation, it may be possible to let species disperse in unaided, but often assisted reestablishment is needed. Hence, for instance, in mine-site restoration, seeds may be returned to the area in topsoil taken from adjacent areas to be mined and in seed mixes containing an appropriate mix of species. In addition, seedlings or parts of plants may be planted into the area.
Restoration frequently aims to speed up vegetation development or to direct its course to a predetermined goal. In order to do this, site characteristics, plant colonization, and subsequent survival can all be manipulated. Attention to the repair of primary processes such as water retention and nutrient cycling is essential, and factors such as soil structure and chemistry and nutrient availability need to be considered. The colonization of a site by species can be effected by ensuring that seed is available on site, either in the soil or canopy seed store or by seed dispersal. Where seeds are mobile or dispersed by birds and other animals, they may be effectively dispersed without further intervention. However, this process may be too slow, and seed may need to be introduced. In extreme cases, seed germination may be too unreliable, and planting of seedlings or other plant material may be necessary. Once species are established at a site, continued vegetation development depends on their survival and how they interact with other species. Survival can be increased by ensuring that site characteristics are favorable (e.g., through water or nutrient retention) and that damage via herbivory and disease is minimized. The development of a functioning biotic community depends on achieving a good mix of species with different life forms (trees, shrubs, grasses, etc., depending on which sort of community is being restored) and the development of species interactions such as mycorrhizal associations, pollination, and seed dispersal.
In addition, large numbers of introduced plant species cause significant alteration to ecosystems around the world. Frequently, restoration has to involve the removal of these undesirable plant species, which either prevent system development or cause the successional trajectory to divert from the desired goal.
How communities are built is a central question in restoration ecology, and recently there have been attempts to consider what factors affect the characteristics of the developing community. Such factors include the timing and extent of arrival of different species on the site and a series of abiotic and biotic filters that influence the success of establishment and survival of each species based on its physiological tolerances, competitive abilities, and interactions with other species. These factors may be considered to be a set of “rules” for what species persist at any given site. Restoration efforts can be guided by these “rules” and aim to modify the abiotic and biotic filters to allow colonization and persistence of species that will allow restoration goals to be achieved.
As indicted earlier, in some instances, restoration can take advantage of autogenic processes by which species recolonize degraded areas unaided, and a plant and animal community reassembles. This is likely to occur where substrate conditions are favorable and where species are effective colonizers; it is the cheapest form of restoration. In some situations, the types of species that colonize may not be desirable (e.g., invasive weed species), or they may prevent the further development of the vegetation to a desired state. In others, there may be little or no colonization because of the continued adversity of the site or because the native species have low dispersal capabilities. In these cases, assisted recovery is necessary and involves ensuring that the desired mix of species is available on site and can persist.
Animals are important elements of ecosystems, and yet restoration projects generally focus on the reestablishment of vegetation, on the assumption that they will provide habitat for fauna to return. However, it is not always clear that this is the case, and often the precise habitat requirements for individual animal species or suites of species are poorly known. Hence, what to put back and which processes are essential to reinstate are not always clear.
Particular faunal elements, known as “ecosystem engineers,” can play important roles in structuring ecosystems and modifying ecosystem processes. The removal or introduction of such species can have dramatic ecosystem effects. Particular examples include beavers in North America and Europe, large herbivores in Africa, and digging marsupials in Australia. Such animals actively modify the environment; for instance, beavers build dams on streams, which create ponds and alter the structure and flow of the water course; large herbivores such as elephants in Africa structure the vegetation by knocking over and feeding on particular types of tree; and digging marsupials in Australia locally change soil characteristics, which affects water infiltration and creates suitable sites for plant establishment. These are particularly important considerations for restoration because the reintroduction of these species can also be construed as restoration programs that are restoring key ecosystem processes.
The reintroduction of particular species of fauna is often conducted in isolation from other restoration activities. However, the reintroduced fauna may have an impact in modifying ecosystem dynamics, and this may be either to the benefit or to the detriment of the system as a whole, depending on the overall goals of management. Similarly, restoration may involve the removal of particular problematic species, including introduced predators. Many restoration projects in such places as Australia and New Zealand and many island ecosystems involve the removal of introduced herbivores and predators that have had severe impacts either on the overall habitat or on native fauna species.
Landscapes are heterogeneous areas of land, usually square kilometers in extent, composed of interacting ecosystems or patches. The spatial distribution and interrelationships among landscape patches determine functions such as biotic movement and fluxes of water, energy, and nutrients.
Ecosystem modification and use often lead to dramatic changes in landscape structure and function. The most obvious manifestation of this is landscape fragmentation, which occurs when the original vegetation of an area is cleared and the land is transformed for agriculture or other use. Fragmentation results in the native vegetation being left as remnants of varying sizes and degrees of isolation. Reduction in size and connectivity of habitat reduces the probability of persistence of many species and results in declining species abundances (see chapter V.1). In addition, landscape processes such as water flows can be dramatically altered, leading, for instance, to changes in wetland habitats or salinization.
As with site-based restoration, management interventions need to be based on a sound assessment of the causes of degradation. In the case of landscape fragmentation, the causes of species loss and decline are often the loss of habitat and connectivity. Thus, replacing habitat and increasing connectivity are often goals for landscape restoration. This is best done by using the existing native vegetation as a skeleton on which to build restoration efforts. Hence, additional habitat or buffer strips can be established next to existing remnants, and corridors or other connecting vegetation can be developed. Restoration actions are best planned in relation to the needs of particular species so that specific recommendations on dimensions of habitat and corridors required can be determined. In addition to habitat re-creation, broader-scale revegetation and other activities may be needed to reverse or slow down hydrological changes and influence other landscape-level processes.
Restoration activities are often relatively costly because they involve management intervention, often on a large scale. In general, it is much more cost effective to prevent damage in the first place rather than repair it once it has happened. Thus, restoration forms part of a spectrum of management options that include prevention and conservation. For biodiversity conservation, the top priority must be to retain areas that remain in good condition. The next priority will be to repair damaged areas of native vegetation. Finally, and as a last resort, restoration of areas that have been transformed (e.g., by agriculture) may be needed to increase areas of habitat or landscape connectivity. It is much more costly to re-create a natural habitat than it is to protect or repair an existing one.
Falk, Don, Margaret Palmer, and Joy B. Zedler, eds. 2006. Foundations of Restoration Ecology. Washington, DC: Island Press.
Hobbs, Richard J., and David A. Norton. 1996. Towards a conceptual framework for restoration ecology. Restoration Ecology 4: 93–110.
Society for Ecological Restoration International Science & Policy Working Group. 2004. The SER International Primer on Ecological Restoration. Downloadable at http://www.ser.org/content/ecological_restoration_primer.asp.
Temperton, Vicky M., Richard J. Hobbs, Tim J. Nuttle, and Stefan Halle, eds. 2004. Assembly Rules and Restoration Ecology—Bridging the Gap between Theory and Practice. Washington, DC: Island Press.
van Andel, Jelte, and James Aronson, eds. 2006. Restoration Ecology: The New Frontier. Oxford: Blackwell.
Walker, Lawrence R., Joe Walker, and Richard J. Hobbs, eds. 2007. Linking Restoration and Ecological Succession. New York: Springer.
Whisenant, Steve G. 1999. Repairing Damaged Wildlands: A Process-orientated, Landscape-scale Approach. Cambridge, UK: Cambridge University Press.
Young, Truman P. 2000. Restoration ecology and conservation biology. Biological Conservation 92: 73–83.
