Coastal Services Center

---

For the Expert: National Review of Innovative and Successful Coastal Habitat Restoration

---

< Methods and Results | Funding and Partnerships>

Restoration Research

Adequate assessment of coastal restoration requires not only a long-term, systematic approach to monitoring, but also a coordinated experimental research program to explain patterns that emerge from the data (Zedler 2001). By incorporating the principles of adaptive management into the experimental approach, scientists can provide information and suggestions that are considered in management decisions (Holling 1978; Walters 1986; Walters and Holling 1990, Thom 1997, Thom 2000). The process suggests corrective measures and can lead to improved restoration results in future efforts in similar ecosystems. In the absence of highly predictable outcomes, restoration sites offer important opportunities to learn by doing. As such, some of the most effective restoration programs integrate pilot studies or experimental designs into their monitoring programs prior to restoration implementation.

Research Organizations and Programs

A number of unique models exist wherein experimental research has become an integral component of successful restoration projects. In general, the ingredients for a successful program include a large funding base, a stable knowledge base (i.e., institutional knowledge), ongoing monitoring programs, and extensive resources for field studies (i.e., labor and equipment). Most often, these circumstances coalesce in a university setting, with leadership provided by a prominent faculty member with a number of supporting graduate students, postdoctoral associates, and collaborators. Often these programs integrate National Estuarine Research Reserve (NERR) support and study sites. Prominent examples include Rutgers University programs involved in Delaware Bay marsh restoration through the Public Service Enterprise Group (Weinstein et al. 1997), and research conducted by San Diego State University's Pacific Estuarine Research Laboratory (PERL) on southern California wetlands under sponsorship of the Earth Island Institute (Zedler 1996; Zedler 2001). Other research institutions involved in coastal restoration research include the following:

• NOAA Center for Coastal Fisheries and Habitat Research, Beaufort, North Carolina
• U.S. Geological Survey (USGS) National Wetlands Research Center, Lafayette, Louisiana
• National Coral Reef Institute, Fort Lauderdale, Florida
• Oregon Institute of Marine Biology, Charleston, Oregon
• Jackson Estuarine Lab, University of New Hampshire, Durham, New Hampshire
• Virginia Institute of Marine Science (VIMS), Gloucester Point, Virginia
• Wetland Ecosystem Team, University of Washington, Seattle, Washington
• Coastal Research Lab, University of New Orleans, Louisiana.

Several federally funded, nationally and regionally significant programs have been instituted to promote the study and protection of estuarine areas, the development of restoration tools and technologies, and communication and educational outreach. These programs, discussed below, offer opportunities for research, collaboration, and restoration project funding.

The previously mentioned NERR program was established by the Coastal Zone Management Act of 1972 to protect and study estuarine areas through a network of 25 reserves from different biogeographic regions of the United States (NOAA 2003e). One focus topic for the program is habitat restoration (NOAA 2004a). An inventory of restoration activities within the NERR program is currently underway (Crawford 2003).

The Cooperative Institute for Coastal and Estuarine Environmental Technology (CICEET) supports the scientific development of innovative technologies for understanding and reversing the impacts of coastal and estuarine contamination and degradation (CICEET 2003). This research includes finding new approaches for restoring coastal habitats.

The Coastal Restoration and Enhancement through Science and Technology (CREST) program in Louisiana and Mississippi is a research initiative developed through an alliance between NOAA, 11 universities, and the USGS National Wetlands Research Center. The program office opened in 2002, and the first call for proposals is expected in 2003. The goal of the program is to integrate research toward improving coastal habitat restoration through the following:

• better coordination of programs and projects
• assessment and improvement of existing methods
• development of new approaches and modeling
• improvement of tools, communications, and outreach
• increased understanding between scientists, managers, and the public (Chapman 2003).

The NOAA Restoration Research Program, part of the NOAA Restoration Center, was developed to advance the science of restoration ecology in coastal habitats (NOAA 2003f). The program supports research on coastal ecosystem structure and function, and focuses on studying the recovery process of restored coastal habitats, developing and testing innovative restoration methods, and establishing success criteria and monitoring protocols. NOAA staff work in partnership with the scientific community to provide expertise and develop improved restoration techniques.

Various studies organized by habitat type, are discussed in the sections below. The studies were selected because they improve the understanding and methodology of restoration and were presented in the published literature and at scientific conferences.

Salt-marsh Research

Hydrological restoration and tidal channel development

Natural hydrology is necessary for restoring functional coastal marshes, and this is often accomplished by returning tidal inundation via breach or removal of barriers such as dikes and levees, or excavation of fill. A recent journal issue (i.e., Restoration Ecology Vol. 10, 2002) was dedicated to exploring the potential and pitfalls of dike/levee breach restoration projects, including examination of experimental manipulations that have been useful for designing and evaluating these projects.

The Model Marsh project in the Tijuana Estuary, California, is providing researchers with an opportunity to develop a better understanding of how marsh channels develop following reintroduction of tidal hydrology to a system (Zedler 2001). Tidal channels are one of the most difficult aspects of a system to predict or design. In the Model Marsh, part of the system is divided into two treatments (each replicated three times). In one treatment, channels were dug to mimic what was understood to be natural channel morphology for this type of system. The second treatment allows self-design, in which channels develop naturally under the existing tidal hydrology. The rate, pattern, and morphology of the channels will be assessed through time, along with the role of marsh-plant colonization and fish use.

A number of studies shed light on how hydraulic-geometry relationships and monitoring of physical evolution can assist in the planning of regionally specific tidal wetland restoration projects. Zeff (1999) examines tidal salt-marsh channel morphometry in New Jersey and identifies principles that can be applied to the design and construction of channel networks in tidal-marsh creation and restoration projects. Williams and Orr (2002) analyzes the rate and pattern of evolution of vegetated marsh plains and tidal channels in 15 formerly leveed salt-marsh sites in San Francisco Bay. In this study, vegetation establishment was retarded in sites that had limited suspended-sediment supply, high wind-wave erosion, and restricted tidal exchange. Williams and Orr (2002) also found that formation of marsh tidal channels was greatly dependent on whether and how high the site was filled before breaching, reporting little channel development at high intertidal sites. In a second San Francisco Bay study, Williams et al. (2002) provides hydraulic, geometric relationships for predicting the depth, width, and cross-sectional area of mature tidal channels as functions of contributing marsh area or tidal prism. These relationships can be used to predict the direction and rate of evolution in an immature or perturbed system.

Hydraulic geometry and other indices provide useful guidelines for physical restoration and creation of estuarine tidal channels but do not clarify the ecological consequences of channel form. To provide this linkage, Hood (2002) investigated whether slough geometry is scaled in parallel with ecological processes, including current velocities, detritus export, organic matter deposition, and benthic community composition. Hood found that organic material in bottom sediment scaled negatively with channel size, as did the abundance of benthic-surface deposit feeders, suggesting that fish-feeding functions could be concentrated in smaller channels or the distal portions of large channels. Similarly, Williams and Zedler (1999) linked fish assemblage composition to channel morphology in a series of restored and natural marsh channels in San Diego Bay. Findings from this study highlight the importance of choosing proper assessment criteria and reference sites to avoid misleading interpretations of restoration success.

Restoration of tidal hydrology should proceed with caution in some situations because biogeochemical processes that mobilize nutrients and other elements can have temporary detrimental effects to water quality, soil nutrient levels, and plant vigor. For example, salt-marsh cores from diked and ditched salt marshes in Cape Cod, Massachusetts, were exposed to experimental treatments in greenhouse mesocosms (Portnoy 1999). This study found that salination of drained peat increased porewater pH, ammonium, and iron, and caused increased sulfate reduction that led to significant subsidence and decreased vigor of transplanted Spartina alterniflora.

Elevation manipulation

In a study by Cornu and Sadro (2002), the marsh surface of a diked and subsided estuarine wetland in Coos Bay, Oregon, was manipulated to examine structural and functional recovery at three intertidal elevations. Results demonstrated that marsh surface elevations subsided at all three treatments, primarily as a consequence of fill consolidation, with vertical sediment accretion driven largely by marsh vegetation density. Tidal-channel development was influenced by marsh surface gradient as much as by marsh surface elevation.

The use of dredged material has been used in numerous ways to offset marsh deterioration or to create proper elevations for marsh creation. For example, to offset the decrease in marsh elevation in North Carolina due to sediment deficits and sea-level rise, a thin layer of dredged material was applied using high-pressure spray dredging (Leonard et al. 2002). The dredged material was added such that the thickness of added sediment ranged from 0 cm to 10 cm in four 40 m 2 plots of both deteriorating and non-deteriorating marsh. The results of the study indicate that both plant stem densities and microalgal biomass increased with the addition of sediment. Their study also suggests that further research is needed to determine an upper limit for the effective thickness of added sediment. Another project using high-pressure spray dredging was effectively employed in Louisiana on a larger-scale plot (0.5 ha) to offset wetlands losses due to subsidence (Ford et al. 1999). The dredged material was applied to both vegetated marshes and shallow water areas adjacent to marshes (i.e., previously vegetated marshes). The results showed that the material enhanced vegetative cover in the marsh and emergent vegetation growth in the shallow water areas. However, shallow subsidence and erosion were concerns in the shallow water areas.

In the Sonoma Baylands restoration project, California, dredged material was used as a supplement to subsidence in previously diked marshes (Marcus 2000). The study estimated that the use of dredged material reduced the time needed for habitat development by several decades. Dredged material is also currently being used successfully in technologies such as geo-textile stabilization tubes, marsh terracing, barrier-island creation, shallow-water enhancement (see Section 3.4 for more information).

The effectiveness of using dredged material for creating or restoring salt marshes has been evaluated in numerous studies. For example, in southwest Louisiana wetland structural characteristics were compared in three natural and four created marshes ranging in age from 3 to 19 years (Edwards and Proffitt 2003). This study found that as created marshes age several structural characteristics approach the levels found in natural marshes. Specifically, vegetation species composition was similar to those found in natural marshes within a few years and also developed along a predictable successional path. However, the percent of organic matter and the bulk densities of the soils were quite different than those of the natural marsh and the authors speculate that it may take several decades for the values to be similar to those in a natural marsh. In Galveston Bay, researchers are experimenting with various marsh configurations of dredge-material marshes to evaluate the effect of landscape on functional value (Minello and Rozas 2002; Rozas and Minello 2001).

The use of dredged material in salt marsh restoration poses many advantages, however numerous factors should be considered in determining whether dredged material should be used for restoration. In a study by Edwards and Proffitt (2003) the dredged material was composed of silty clay (30-65% clay) compared to some other dredge-filled created marshes that used sandier substrates. The authors speculate that the higher clay content may contain greater initial nutrient levels and thus may encourage more rapid vegetation growth. Turner and Streever (2002) recommend that the suitability of the dredged sediment be considered during the planning stages of restoration. Specifically, the authors recommend that contaminated sediment should not be used and caution that the use of clays and silts usually require construction of confining dikes.

Plant propagation and reintroduction

A common area of experimental research has focused on increasing the establishment, growth, and functional benefits of native salt-marsh vegetation. Soil amendments, including organic matter (e.g., composted kelp and municipal sewage sludge) or inorganic fertilizers (e.g., urea or ammonium nitrate), are experimental treatments that have been used in restoration sites with coarse soils, such as dredge spoils or sandy upland areas, where nitrogen is limiting (Callaway 2001). Boyer and Zedler (1999) found nitrogen additions increased Spartina foliosa growth in the short-term, but few long-term impacts on aboveground growth. Added nitrogen also promoted growth of the annual, Salicornia bigelovii, over S. foliosa (Boyer and Zedler 1999), although it did not appear to promote accumulation of soil organic matter or nutrients in the long-term (Boyer et al. 2000). Preliminary findings in other ongoing studies have suggested that organic matter amendments, such as alfalfa, peat, kelp, or sewage, may "jumpstart" the marsh food chain by encouraging greater microbial growth (Levin and Currin 2002). Soil enrichment and propagation techniques compared in Seal Beach, California, to test establishment of pickleweed (Salicornia virginica) suggest that pickleweed mulch rototilled into the soil increases establishment over transplanting and cuttings (Disney and Miles 2000).

The use of seedlings continues to be an effective approach for establishing diverse salt-marsh vegetation in smaller restoration projects. Empirical studies have compared the effects of species diversity on ecosystem function (Sullivan 2001; 2001). Findings from these studies support the general hypothesis that diversity is important in southern California marsh restoration sites, with high-diversity plantings resulting in more complex canopies, increased biomass, and increased nitrogen accumulation. In a study by Lindig-Cisneros and Zedler (2002), natural seedling recruitment was evaluated in a restored salt marsh as a function of established plant-species diversity. Abiotic effects (i.e., elevation, salinity, and canopy cover) preceded biotic interactions in determining recruitment early in marsh development, although species richness did have a scale-dependent effect on overall recruitment. Results from this study indicate that natural dispersal of seeds at a restoration site can be maximized by introducing tidal inundation in early winter before all seeds have dropped, when high spring tides could disperse seeds, and when rainfall might lower salinities and enhance germination.

Researchers in southwestern Louisiana have evaluated the genetic diversity of native plant species that have been allowed to recolonize large expanses of mud flats constructed from dredged sediments (Travis et al. 2002). The study was conducted on three restored marshes with ages of 1, 4, and 14 years at the time of sampling. Overall, they found that restored populations of Spartina alterniflora maintained levels of genetic diversity comparable with natural populations. The researchers also theorize that natural recolonization may actually stimulate an increase in genetic diversity relative to natural marshes occupying neighboring sites.

Biotechnology and classical plant improvement methods are currently being evaluated for use in salt-marsh restoration activities through cooperative efforts of scientists from Louisiana State University and the U.S. Natural Resources Conservation Service (NRCS). The objectives of the program are to develop a seed-based system of propagating Spartina alterniflora over large areas and to genetically improve the performance of this species (Harrison et al. 2001). Specific ongoing research efforts are focusing on seed encapsulation techniques and evaluation of clones with high seedling vigor and exceptional resistance to disease (Croughan et al. 2002).

Functional Assessment

Research on monitoring methods often is focused on determining the functions provided by a restored habitat and comparing the functions to those provided by natural systems. The discussion below is primarily focused on the various functions that salt marshes provide for fish. Other functional assessments can include the use of habitat by birds and mammals, the production of invertebrates, the ability to improve water quality, or the reduction in shoreline erosion.

One novel method to assess ecological function is through the evaluation of bacterial activity in transforming organic matter into available forms for secondary consumers. Researchers have found that bacterial growth efficiency (BGE) is an indicator of ecosystem function, where pristine marshes have the highest BGE (del Giorgio and Newell 2002). The researchers conducted long term monitoring of BGE in both natural and restored marshes and determined that restored marshes have a lower BGE then natural marshes. While there may be a trend in restored marshes toward the values observed in natural marshes, the restored systems continued to have a lower BGE over the six-year study period. These results suggest that there are considerable differences in elemental and organic carbon cycling between restored and natural marshes.

Studies often monitor both the opportunity for fishes to access restored habitats and the capacity of these habitats to promote fish resilience and production (Simenstad and Cordell 2000). Most research has shown that restoration of tidal flows in previously restricted salt-marsh habitats result in rapid changes in the composition, density, size, and distribution of fish and crustacean species (Able et al. 2000;Raposa 2002) although the degree of tidal restriction may also influence parameters of community composition (Raposa and Roman 2003). In some cases, assemblage development may peak soon after restoration, then later decline as sediment and hydrologic processes of a site change and stabilize (Williams and Zedler 1999; Williams and Desmond 2001). In other cases, assemblage structure may be relatively simple in terms of species richness and trophic composition immediately after restoration, and may take a decade or more to approach more natural systems (Simenstad and Thom 1996; Warren et al. 2002).

Landscape characteristics within the marsh can also effect marsh function for fish access and use. Research in Texas and Louisiana continues to assess habitat values for fishery species of dredged material marshes in various landscape configurations using terracing (Minello and Rozas 2002; Rozas and Minello 2001). This method of salt marsh restoration involves the creation of ridges in some pattern that maximizes intertidal edge and minimizes fetch between ridges, with the intertidal areas planted with marsh vegetation. The researchers found that while the terrace marsh was not functionally equivalent to the natural marsh for fishery species, the terrace field supported higher standing crops of most fishery species compared with shallow marsh ponds of similar size (Rozas and Minello 2001). The study recommends that future terracing projects should increase the proportion of marsh in a terrace field to enhance the habitat value for fishery species. In Chesapeake Bay, ongoing studies have started to evaluate whether different planting configurations of Spartina alterniflora in restored areas of Eastern Neck National Wildlife Refuge influence access to the marsh surface by fish (NOAA 2003g).

Empirical studies are increasingly being used to confirm the value of restored salt-marsh habitats in terms of realized survival and growth to fishes. Miller and Simenstad (Miller and Simenstad 1997) detected no significant differences in the relative growth of juvenile Pacific salmon in a created estuarine slough as compared with a natural reference site using fish otolith microstructure. In Delaware Bay, an intensive mark-and-recapture program with coded wire tags determined that growth and production of mummichog (Fundulus heteroclitus) appeared higher in a restored marsh than in previous studies in natural marshes (Teo and Able 2003). Bioenergetics models that estimate growth of fishes under various conditions (Gray et al. 2002) (e.g., temperature and other physicochemical measurements during fish occupation) have also been proposed as a tool for designing coastal wetland restoration projects (Madon et al. 2001). Researchers at the University of Washington are currently using these mass-balance models to determine how estuarine marsh habitats in different stages of recovery contribute to the growth of juvenile salmon (Gray 2003).

Food-web relationships and fish feeding have also been explored in empirical studies that examine the functional benefits of restored salt-marsh habitats. As with the previous studies, results vary by site. Tupper and Able (2000) found that creek utilization and diets of striped bass were similar between natural and restored Delaware marsh habitats. However, juvenile salmonids occupying restoration sites in southwest Washington (Miller and Simenstad 1997) showed diet composition different from what might be expected in natural reference systems. Restoration sites of different age may reflect differences in fish prey and diets as these systems develop and mature (Gray, Simenstad, Bottom, and Cornwell 2002;Simenstad and Cordell 2000). Stable isotope methods developed as a tool for the analysis of food webs in tidal salt-marsh systems (Kwak and Zedler 1997;Page 1997;Weinstein et al. 2000) are also beginning to be used to assess recovery of trophic function in created or restored systems (Levin and Currin 2002).

In summary, these studies indicate that overall restored salt marshes are providing fish access to usable habitat and the systems are functioning to increase growth, production, and resilience of fish populations. However, in some cases restored systems may be structurally and/or functionally different from natural marshes. Continued research will help determine whether improved restoration methods could improve functional equivalency.

Seagrass Research

Seagrass restoration research has evaluated the effects of restoration on seagrass genetics, methods to improve the cost effectiveness of seagrass restoration, and ways to improve the conditions for seagrass growth.

Genetics

Several studies have looked at the effects of transplanting seagrasses on genetic diversity in Chesapeake Bay and southern California (Williams 1997; Williams and Orth 1998; Williams 2001). Findings indicate that transplanted beds were not reduced in genetic diversity compared to natural beds and the use of seeds as donor material could improve the genetic diversity of restored areas. In addition, there is a positive association between genetic diversity and propagation. Eelgrass (Zostera marina L.) populations with higher genetic diversity developed more flowering shoots, achieved greater seed germination, and had a higher leaf-shoot density.

Transplant Methods

Transplanting adult eelgrass plants has been the focus of many restoration efforts (Davis and Short 1997; Fonseca et al. 1998; Calumpong and Fonseca 2001). Recent research has primarily centered on finding more successful and cost-effective techniques. Short (2003) has developed the Transplanting Eelgrass Remotely with Frame System (TERFS). Two hundred eelgrass shoots are tied with biodegradable ties to a 50- by 50-cm, plastic-coated wire frame, which is then dropped into position from a boat. Several weeks later, after the ties have degraded, the frames are removed and the shoots remain in place. Three out of four one-acre sites continue to show an increase in eelgrass shoot abundance after preliminary monitoring, with failure in the fourth attributed to an algal bloom smothering the plants (Short 2003).

Propeller scarring and vessel groundings have caused considerable damage to seagrass beds in Florida. Two methods have been tested to determine an effective means of restoring the damaged areas are 1) the use of sediment-filled tubes and 2) the placement of bird stakes to promote natural fertilization in transplant areas. The use of sediment-filled, biodegradable, fabric tubes to fill propeller scars in seagrass beds was tested to see if they enhance or prevent seagrass recovery (Hammerstrom 2003). Early results indicate that the sediment tubes are an effective means of deploying fine sediment and preventing further erosion in the propeller scars.

The second method tested in propeller-scarred areas is transplanting fast-growing seagrasses in damaged areas among bird roosting stakes (Kenworthy 2003). The theory is that while roosting, the birds defecate into the water where the seagrasses are planted, and Halodule wrightii , a colonizing species, uses the nutrients and rapidly covers the damaged bottom. After 1.5 to 2 years, the stakes are removed and the slower growing, climax species reestablish. The idea can be thought of as a compressed succession in which recovery is accelerated by fertilizing the faster growing species that naturally occur but are normally in lower abundance than the climax species, Thalassia testudinum. This method was tested against application of water-soluble fertilizers and plant-growth hormones. The results indicated that the fertilizer application technique failed to increase the recovery rate of T. testudinum or H. wrightii , whereas the bird stakes produced extremely high recovery rates for H. wrightii (Kenworthy et al. 2000).

Seeds

Recent research has begun to focus on the use of seeds for restoration to reduce costs and promote larger-scale restoration efforts. Researchers at the University of Rhode Island (Nixon et al. 2002) have developed a mechanized underwater seed planter to efficiently plant large areas of eelgrass. The system is comprised of a pump, similar to that used to fill jelly donuts, mounted on a sled that is pulled along the sediment. The seeds are mixed with a planting media then pumped just below the sediment surface. Current research is comparing organic and inorganic planting media to reduce anoxia and increase resulting seed germination (Nixon et al., 2002).

Harwell and Orth (1999) tested the effectiveness of burlap bags (1.0-mm mesh size) to protect seeds from predation, burial, or lateral transport. The results showed that seeds survived better to the seedling stage with protection (41% to 56%) than without protection (5% to 15%). Seeds planted in the laboratory with and without bags had similar seedling survival (50% in both treatments), indicating that predation and transport may be causes for the differences observed in the field trials.

A method for eelgrass restoration that shows considerable promise is being tested by Orth (2003) in Chesapeake Bay. In this large-scale experiment, a total of 9.1 million seeds were broadcast from a boat over 74 one-acre (0.4-ha) plots. Preliminary results showed that the eelgrass plots were visible in aerial photos after eight months. The use of seeds for seagrass restoration will likely be the focus of further research in efforts to decrease the costs of seagrass restoration.

Improving light conditions under docks and piers

Improving light under docks is one method of restoring conditions conducive to seagrasses growth. Blanton et al. (2002) compared the relative amount of light produced by deck prisms, solar tunnels, and metal halide lights. Recommendations considered many variables, including the height of the deck above the water surface, maintenance, structural integrity, and costs; the solar tunnels proved to be the best overall alternative. Another study examined the effectiveness of reflective material under a dock to reflect light from the water surface to depths where seagrasses grow (Gayaldo et al. 2001). These results indicated a positive correlation between the placement of the material and improved survival of the seagrass. A study in Florida looked at the use of deck prisms for increasing light under docks in a freshwater system and found a consistently higher percentage of vegetation cover under docks with deck prisms than under those without (McKinney 2001).

In summary, the research discussed above is focused on improving the conditions for seagrass growth and the cost effectiveness of planting techniques. Through these efforts, seagrass habitat restoration is likely to become more successful and able to occur on a larger scale than in the past.

Reef Research

Coral reef

One theme that emerged at the International Conference on the Scientific Aspects of Coral Reef Assessment, Monitoring, and Restoration in 1999 (Thomas 2001) was the need to conduct hypothesis-based research into the efficacy of methods for coral-reef restoration and rehabilitation. One method being evaluated is transplanting donor stock, either collected or grown, to natural and artificial substrates. Studies by Rinkevich and others describe the mariculture of clonal coral fragments in protected areas for eventual transplanting to denuded areas of the reef, as well as the transplanting and growth of gravid colonies in the field or the laboratory to increase the diversity of sites restored by transplanting (Rinkevich 1995; Epstein et al. 2001).

The National Coral Reef Institute has several studies under way in South Florida, some located at ship-grounding sites (Gilliam et al. 2003; Glynn et al. 2003; Quinn et al. 2003). These experiments evaluate the use of potential coral-larvae attractants, such as iron additive, CaCO3, and coral transplants to enhance larval recruitment. They also evaluate the effects of structural complexity affixed to Reef Balls™ (i.e., concrete structures) on fish assemblages, and the interaction between fish assemblages and coral recruitment. Monitoring is being conducted at both the transplanted areas and the coral communities from which donor cores were extracted.

The potential benefits of coral transplants are critically evaluated in a global review of projects by Edwards and Clark (1998), which argues that transplanting should be carried out only when natural recovery processes fail. The authors do, however, identify some exceptions to the rule. One key finding of their study is that because branching corals (e.g., Acropora, Pocillopora) naturally recruit rapidly but transplant relatively poorly, there is little justification for transplanting. In contrast, massive species (e.g., Porites, Pavona) recruit slowly, grow slowly, and transplant well, suggesting that these species make better subjects for transplantation.

Oyster reef

Scientists at the Virginia Institute of Marine Science use oyster reef restoration projects to conduct research on reef biology, community development, and trophic dynamics. These studies provide insight into the success of the restoration in mirroring ecological function provided by historically prevalent oyster reefs in Chesapeake Bay. Harding (2001) examined the horizontal spatial and temporal variation in zooplankton community abundance and composition at a restored oyster reef in a Chesapeake Bay estuary as a potential metric of restoration progress over time. Zooplankton communities are an important component of intermediate trophic levels in estuarine food webs. The seasonal abundance patterns and community composition documented at this restored reef site were similar to that observed in other mid- and south Atlantic estuaries.

Mangrove Research

Hydrology

An Army Corps of Engineers Wetland Research Program technical note (Lewis and Streever 2000), shows that hydrology is the most critical factor in mangrove restoration, as demonstrated by the reestablishment of all three Florida mangrove species at West Lake near Fort Lauderdale, Florida, in just 78 months after hydrologic restoration without any planting. Research on this and other systems shows that although restoration managers have frequently tried planting first, a better approach is to "determine the causes for mangrove loss, remove these causes, and work with natural recovery processes." Typically, the cause of mangrove loss is altered hydrology, so it is concluded that planting should only be used as a restoration tool when tidal hydrology has been reestablished and if waterborne seeds or seedlings will not reach the restoration site. Only 15 to 30 years are required for self-repair or secondary succession if normal hydrology and propagules from adjacent stands are present, and dense mangrove shrubs can develop within 5 years of plant establishment. Lewis and Streever (2000) provides five steps critical to successful mangrove restoration, which are abbreviated here: 1) understand the ecology of local mangrove species (especially patterns of reproduction, propagule distribution, and successful seedling establishment), 2) understand the normal hydrologic patterns, 3) assess habitat modifications preventing natural secondary succession, 4) design a restoration program to restore hydrology and, if possible, utilize natural recruitment, and 5) plant only if natural recruitment will not meet the objectives of the restoration project. This approach has been successfully applied in the Cross Bayou Mangrove Restoration Site in Pinellas County, Florida, as mitigation for the Tampa Bay oil spill of August 10, 1993, which after 2 years of monitoring was rapidly achieving the performance criteria (Lewis III 2003b).

Plantings

Mangrove plantings have typically resulted in monocultures because of the ease with which the propagules of the red mangrove (Rhizophora mangle) are collected and transplanted (Gayaldo 2003). Natural regeneration tends to result in a community of mangrove species that more closely represents nearby communities, which may be similar to the original mangrove vegetation in the area (Field 1998). According to Field (1998), only about 30% of mangrove species have been used in planting projects. However, while monoculture is less desirable than a more natural assemblage of species for various ecological reasons, establishing mangroves may be desirable for stabilizing shorelines (e.g. along boat channels experiencing erosion or on the shore of a new dredge-spoil island). Fortunately, species can be allowed to self-sort along environmental gradients during reestablishment of mangroves, without competition from invasive species, because of the limited species pool that is adapted to the extremes of the mangrove environment such as salinity and flooding (Lugo 1998).

In summary, restoration research is an important aspect of restoration. Experimentation should be incorporated in a restoration project or program whenever possible. The results of research efforts are helping to reduce uncertainties associated with restoration, improve the cost-effectiveness of projects, and develop better implementation methods and assessment techniques.