Habitat Modification and Coastal Protection by Ecosystem-Engineering Reef-Building Bivalves

Reef-building bivalves like oysters and mussels are conspicuous ecosystem engineers in intertidal and subtidal coastal environments. By forming complex, three-dimensional structures on top of the sediment surface, epibenthic bivalve reefs exert strong bio-physical interactions, thereby influencing local hydro- and morphodynamics as well as surrounding habitats and associated species. The spatial impact of the ecosystem engineering effects of reef-building bivalves is much larger than the size of the reef. By influencing hydrodynamics oysters and mussels modify the sedimentary environment far beyond the boundaries of the reef, affecting morphological and ecological processes up to several hundreds of meters. that serve two the service coastal shoreline and the supporting habitat for species service (enhancement of biodiversity and diversification of the landscape). Due to their wave dampening effects, reef-building reefs shoreline protection and erosion control eroding coastlines, artificial shoreline Abstract in Chinese

facilitative interactions at long-distances of bivalve reefs provide biodiversity benefits and more specifically facilitate or protect other valuable habitats such as intertidal flats, sea grasses, saltmarshes and mangroves.
Two case studies are used to demonstrate how bivalve reefs can be restored or constructed for shoreline protection and erosion control, thereby focusing on oyster reefs: (1) Oyster reefs for shoreline protection in coastal Alabama, USA, and (2) Oyster reefs as protection against tidal flat erosion, Oosterschelde, The Netherlands.
It is argued that bivalve reefs should be promoted as nature-based solutions that provide biodiversity benefits and coastal protection and help in climate change mitigation and adaptation. In order to successfully restore these habitats practitioners should consider a general framework in which habitat requirements, environmental setting and long-distance interdependence between habitats are mutually considered.

Bivalves as Ecosystem Engineers
Ecosystem engineers are organisms that have morphological and/or behavioral traits that enable them to modify, maintain and/or create habitats (Jones et al. 1994). They induce physical state changes in abiotic and biotic materials, thereby regulating the availability of resources to other organisms. This alters the availability of ecological niches to other species, facilitating certain species and inhibiting others (Bruno et al. 2003;Bouma et al. 2009). Jones et al. (1994Jones et al. ( , 1997 distinguished two types of ecosystem engineers. Autogenic ecosystem engineers change environments through their physical structures (e.g. vegetation, coral reefs), in other words they are part of the engineered habitat, whereas allogenic ecosystem engineers transform living or non-living materials from one physical state to another (e.g. beavers, bioturbating worms).
Bivalves are conspicuous ecosystem engineers that often have both autogenic and allogenic characteristics. In general, in soft-sediment environments, bivalves can be divided into epibenthic and endobenthic organisms depending on whether they spend most of their lifetime above or below the sediment, respectively. Many endobenthic bivalves such as cockles and clams modify the sedimentary habitat through their behavior and can be considered allogenic ecosystem engineers (Bouma et al. 2009). They affect a number of resources mainly through bioturbation and bioirrigation (Reise 2002;Ciutat et al. 2007; Montserrat et al. 2009), but in high densities they also can increase sediment stability (Donadi et al. 2013). The most prominent epibenthic ecosystem engineers inhabiting bare coastal sediments are reef building bivalves. The best-known examples of reef-building bivalves are intertidal and shallow subtidal mussels (Mytilidae) and oysters (Ostreidae). These epibenthic ecosystem engineers modify the sedimentary habitat mainly through their physical structures, and thus are true autogenic ecosystem engineers (Gutiérrez et al. 2003, Bouma et al. 2009). At the same time they are active filter feeders, removing large quantities of suspended material from the water column, influencing water clarity and quality, and producing faecal and pseudofaecal biodeposits that accumulate in the reef and its surroundings (Newell 2004;Kellogg et al. 2013). So, these epibenthic ecosystem engineers have both autogenic and allogenic properties.
Coastal engineers characterise structures such as coral and oyster reefs as low crested or submerged breakwaters and have modelled and validated the effect of these structures on wave height and current velocities (Van der Meer et al. 2005). However, the literature describing the impact of natural reefs on disrupting wave energy is much better developed for coral reefs than bivalve reefs (Sheppard et al. 2005;Rogers et al. 2013;Ferrario et al. 2014). This has included how reef parameters and morphology influence the physics of wave attenuation and flooding (Monismith 2007;Lowe et al. 2010;Quataert et al. 2015).
In this paper, we focus on the ecosystem engineering effects of intertidal and shallow subtidal reef-building bivalves in soft-sediment environments, thereby focusing on the bio-physical interactions these organisms have with their environment and two ecosystem services these interactions provide, (1) coastal protection (i.e. erosion control and shoreline stabilization) and (2) habitat for species by facilitation of other marine habitats and species (i.e. increasing biodiversity at large landscape scale). Using two case studies, one in the US and one in the Netherlands, we subsequently show how constructed oyster reefs contribute to shoreline and salt marsh protection or erosion control.
The shift in species composition at the spatial scale of the engineered habitat, this is within the footprint of the reef structure, is discussed in two other papers within this book, namely the effect on associated benthic macroinvertebrates (Craeymeersch and Jansen 2019) and the effect on finfish and crustacean production (Hancock and zu Ermgassen 2019).

Characteristics of Epibenthic, Reef-Building Bivalves
Epibenthic, reef-forming bivalves create spatially and topographically complex habitats and can be found in a wide range of spatial scales, from small clumps to large patches to extensive beds and reefs that cover thousands of square meters and extend kilometres in length (Gutiérrez et al. 2003.1). Bivalves attach themselves to the substrate or to each other via byssal threads (e.g. mussels like Mytilus edulis, Musculista senhousia) or via calcification, this is cementing themselves to the substrate or to each other (oysters like Crassostrea gigas, Crassostrea virginica, Ostrea edulis). This makes individual mussels more mobile compared to oysters, on the other hand reefs of mussels will be less persistent compared to oysters, whose structure will persist for a long time, even after dying off of the reef.
The often distinct spatial patterns observed in reef-building bivalves are caused by feedback loops and self-organization processes at different spatial scales and lead to complex habitat forms (Bertness and Grosholz 1985;van de Koppel et al. 2005van de Koppel et al. , 2008. Mussels like the Blue mussel (Mytilus edulis) typically attach themselves to a hard substrate or to each other by forming byssal threads. To find protection or food, mussels can move by releasing the byssal threads and using its foot to move to a new location. Liu et al. (2014) demonstrated that in mussel beds, selforganization generates spatial patterns at two characteristic spatial scales: smallscale, net-shaped patterns due to individual movements (i.e. behavioural aggregation) of individuals, and large-scale, regular banded patterns due to the interplay betweenmussel facilitation and long-distance competition for algae. The interplay between self-organizing processes at individual and ecosystem level is therefore a key determinant of the functioning and resilience of mussel beds (Liu et al. 2014). A decrease of water depth over blue mussel Mytilus edulis mounds results in an increase in water flow, enhancing food transport to the mussels on top of mounds. This locally lowers algal depletion resulting in higher net growth on top of the mounds (positive feedback loop) (Liu et al. 2012).
Oysters typically build reefs through the gregarious settlement of multiple generations onto existing oyster substrate. The shells provide a solid substrate on which new oyster larvae can attach, increasing larval survival by providing shelter from predators and preventing burial in sediment (Mann and Powell 2007). Each generation of larvae settle on top of the preceding generation so that the reef grows vertically and becomes highly complex, with many structural irregularities and infoldings (Dame 2005;Rodriguez et al. 2014). Once settled, oysters are less mobile compared to mussels. But also oyster reefs appear in distinct reef morphologies and spatial patterns, thought to arise from feedback mechanisms between oysters and local hydrodynamics (Lenihan 1999;Dame 2005;Colden et al. 2016). Colden et al. (2016) describe three different reef types or morphologies for the Eastern oyster (Crassostrea virginica) based on historical work by Grave (1905): string reefs, fringe reefs and patch reefs.

Interaction with the Local Environment
Reef-building bivalves add hard substrate to an otherwise soft, more unstable, and often relatively flat bottom in sedimentary environments. By forming complex, three-dimensional structures on top of the sediment surface, epibenthic bivalve reefs influence local hydrodynamics and sediment dynamics (Lenihan 1999;Gutiérrez et al. 2003). Epibenthic bivalves enhance near-bed turbulence and vertical mixing in the lower water column and slow down current velocities through increased roughness of the bed (Widdows et al. 2002). Epibenthic bivalve reefs, especially oyster reefs, are often mentioned to act as breakwaters that can attenuate waves (Grabowski and Peterson 2007;Scyphers et al. 2011), but empirical evidence is scarce. Flume studies by Borsje et al. (2011) and Manis et al. (2015) showed that oysters effectively reduce wave energy compared to bare sediment. Lunt et al. (2017) measured in the field wave height and current speed at the windward and leeward side of oyster (Crassostrea virginica) reefs in St Charles and Aransas Bays (Texas, USA), with average wave height (average: leeward 0.05 m, windward 0.22 m) and current speed (average: leeward 3.75 cm s −1 , windward 6.68 cm s −1 ) highest on the windward side of oyster reefs.
Epibenthic bivalve reefs exert a strong effect on the benthic-pelagic coupling through their suspension feeding behavior, feeding on seston in the water column and transferring undigested organic and inorganic material in their faeces and pseudofaeces to the sediment surface (Dame 1996;Newell 2004). When abundant, bivalves can exert top-down control on the phytoplankton, thereby reducing turbidity and increasing water transparency (see Cranford 2019).

Ecosystem Engineers Offer Essential Ecosystem Services Including Coastal Protection and Habitat for Species
Coastal ecosystems are increasingly recognized as essential elements in ecosystem restoration and coastal adaptation (Cheong et al. 2013;Temmerman et al. 2013). The capacity of marshes and shellfish reefs to maintain their own habitat via biophysical feedbacks, and their ability to grow with sea-level rise is seen as an important advantage over traditional man-made hard engineering structures (Rodriguez et al. 2014;Kirwan et al. 2016;Walles et al. 2015b). Ecosystem engineers, in particular, are increasingly recognized for their role in delivering important ecosystem services that serve human wellbeing. In case of epibenthic bivalve reefs, ecosystem services include a myriad of provisioning, regulating, habitat and cultural services (Coen and Luckenbach 2000, Coen et al. 2007, Beck et al. 2009, Grabowski et al. 2012, and papers within this book). Here we focus on two services, namely the regulating service coastal protection (coastal erosion prevention, shoreline stabilization) and the supporting/habitat service habitat for species (enhancement of biodiversity and diversification of the landscape). The latter specifically deals with the extending effects outside the boundary of the reef structure itself. Figure 13.2 visualizes these ecosystem services in a cross-section along a shoreline. In their natural setting, bivalve reefs attenuate wave energy and stabilize sediments, protecting adjacent habitats such as intertidal flats, sea grasses and salt marshes against erosion. As coastal erosion is an increasing problem worldwide, due to human interventions, worsened by climate induced sea level rise and more frequent storms, shellfish reefs are therefore an attractive living shoreline approach that can be used for shoreline protection and erosion control along eroding coastlines, offering an alternative to artificial shoreline hardening (Meyer et al. 1997;Piazza et al. 2005;Scyphers et al. 2011;La Peyre et al. 2015;Walles et al. 2016a) ( Fig. 13.3). For example, 30% of the oyster restoration projects in the US involve coastal protection as one of the targets (http://projects.tnc.org/coastal/). Besides relatively small restoration projects, increasingly large projects are in development. In Alabama there is a long range goal to construct 100 miles of oyster reef for shoreline protection and conservation of coastal marshes and seagrasses. This work has developed into large shoreline protection projects using oyster reef in all the US Gulf of Mexico states, valued at over $178 Million US (See case study 1). In the Netherlands, 1.3 km of oyster reefs were constructed in the Eastern Scheldt for erosion control of intertidal flats (Walles et al. 2016a, see case study 2).
Another important ecosystem service delivered by bivalve reefs is the provision of habitat(s) for species, leading to enhancement of biodiversity and diversification of the landscape. Indeed, the spatial impact of the ecosystem engineering effects of reefs is much larger than the size of the reef footprint. By influencing hydrodynamics oysters and mussels modify the sedimentary environment and physical landscape far beyond the boundaries of the reef, affecting morphological and ecological processes up to several hundreds of meters. Biodeposition of organic material in the form of faeces and pseudofaeces changes the sediment composition and enhance primary production of microphytobenthos in the surroundings (Newell 2004;Donadi et al. 2013;Engel et al. 2017). Walles et al. (2015a) demonstrated alteration of tidal flat morphology by Pacific oyster Crassostrea gigas reefs up to tens or  to be scale-dependent: at short distances cockle (Cerastoderma edule) abundance was suppressed, whereas at larger distances increased spatfall and better survival of adult cockles was observed due to reduced sediment erosion (Donadi et al. 2013). This in turn influenced higher trophic levels, such as shorebirds that feed upon cockles (Donadi et al. 2013, van der Zee et al. 2012. Other studies showed cross-habitat interactions between epibenthic, suspension feeding bivalves and seagrass habitats (Peterson and Heck 2001;Newell and Koch 2004) and marshes (Meyer et al. 1997;Piazza et al. 2005). Therefore, at the landscape level, ecosystem engineering typically enhances environmental heterogeneity, thereby increasing niche opportunities and eventually the diversity of communities and habitats such as sea grasses, marshes and mangroves (van de Koppel et al. 2015). The facilitation of these vegetated habitats in turn also deliver essential ecosystem services (Barbier et al. 2011) and coastal plant communities are increasingly recognized for their capacity for climate change adaptation and mitigation (Duarte et al. 2013). Marshes and mangroves overall have higher potential as natural defences for coastal protection, but restoration of bivalve reefs in front of these vegetated habitats can strengthen their growth and survival. Therefore, epibenthic bivalve reefs are as such not suitable for use as primary flood defences, due to their relatively low position in the intertidal zone, but indirectly, through their long-distance effects and facilitative interactions on other habitats, will add to the protection and flood defence of our coasts and improve the resilience of ecosystem-based coastal defence practices (Temmerman et al. 2013).

Study Case 1: Oyster Reefs for Shoreline and Salt Marsh Protection in Coastal Alabama, USA
In the northern Gulf of Mexico (USA), several coastal habitats have suffered declines of more than 50% over the last century, including seagrass, wetlands, and oyster reefs. Locally, Alabama has lost some 70% of its seagrass habitat, 85% of its oyster habitat, and thousands of acres of wetlands. Loss of nearshore habitats has been caused by decreased water quality, dredge-and-fill activities, construction of seawalls, jetties and groins among other causes, but in the case of oyster habitat mainly from overfishing (Beck et al. 2011). Increased erosion is due to increased ship wakes from the channel to the Alabama Port in Mobile as well as seasonal storms, sea level rise and climate change, and has been exacerbated by the loss of habitat such as oyster reef that can reduce wave energy before it reaches the coast. The declines, along with increasing appreciation of the value of the services these habitats provide, have prompted increased efforts to restore these habitats and ecosystem services. A goal of many habitat restoration projects has been shoreline protection (e.g. living shorelines, constructing oyster reef to act as breakwaters and beneficial use of dredge material) (Kroeger and Guannel 2014). Coastal Alabama hosts some advanced and well documented ecological studies that demonstrate the potential for restoring degraded habitats and enhancing many ecosystem services for community resilience, wave attenuation, shoreline protection, and fish production (Coen and Luckenbach 2000;Piazza et al. 2005;Grabowski and Peterson 2007;Gregalis et al. 2009;Powers et al. 2009;Scyphers et al. 2011;Grabowski et al. 2012;La Peyre et al. 2014;Scyphers et al. 2015;zu Ermgassen et al. 2016).
In 2009 the Obama administration provided funds for economic stimulus to counter the recession triggered by the real estate collapse in 2008 (https://www. treasury.gov/initiatives/recovery/Pages/recovery-act.aspx). Some of these funds were directed toward habitat restoration and used to move what had generally been 'proof of concept' scale work to an ecosystem scale. In Mobile Bay, AL, a coalition of partners led by The Nature Conservancy used this opportunity to implement large scale oyster habitat restoration directed primarily toward shoreline protection (Kroeger and Guannel 2014). This approach has been continued through the successful engagement of municipalities and communities across the Alabama coast, the development and implementation of watershed management plans (WMP) and additional shoreline protection projects. Socioeconomic studies of coastal Alabama residents have demonstrated an awareness of the decline of coastal habitats, a broad appreciation that they provide important ecosystem services, and substantial support for enhancing coastal resilience (Scyphers et al. 2014a(Scyphers et al. , 2014b. A substantial investment (~$50 M US dollar) has been made in more than 13 completed and 5 progressing restoration projects in coastal Alabama to the end of 2017 (Table 13.1). Data from 13 projects completed by multiple partners is currently being synthesised to determine which techniques have worked best for oyster recruitment, fisheries production and shoreline protection, including an economic valuation of the outcomes. At the time of writing, project monitoring has highlighted a need for some specific enhancements on the initial habitat restoration projects to improve functionality, as well as to add additional benefits beyond the original objectives. For these existing restoration projects, adaptive management strategies informed by long-term monitoring will be implemented to enhance these current investments and further improve future projects.
For example, in early 2011 following the Deepwater Horizon oil spill, at a site known as Helen Wood Park, 12 intertidal reefs were constructed to protect the Louisiana has substantial investment but data are not readily available 392 m shoreline and provide habitat. The site has been impacted by adjacent development (seawall) and has experienced significant erosion. The reefs measured ~25 m long and 2-3 m wide with 7.5 m gaps. A combination of natural and engineered materials was used, with 4 reefs constructed using concrete Reefballs, 7 reefs were constructed using bagged oyster shell, and 1 reef being partly made of Reefballs and partly bagged oyster shells. Shoreline position, reef footprints and recruitment of sessile organisms was monitored annually. Approximately one year post-construction, the marsh at the south end had expanded by some 7% and accumulation of sediment leeward of the structures had occurred (Fig. 13.4). As monitoring continued, by year 3, the bagged shell reefs had noticeably broken down. Their footprint expanded as the shell spread, and the vertical structure had decreased ( Fig. 13.5). While the use of bagged oyster shell was preferred for conspecific recruitment, as well as aesthetics, this substrate broke down after 3-4 years, providing reduced wave attenuation and shoreline protection (Unpublished data).
The long-term monitoring conducted at the Helen Wood Park site was used to inform subsequent projects to increase project success and longevity. For example, the 940 m Swift Tract project was completed in the summer of 2012. The construction technique utilized at the Swift Tract consisted of Hesco barriers, galvanized steel modular baskets that were installed and then filled with gabion stone. A 0.15 m layer of oyster shell was placed on top of the gabion stone within the cages. Pockets on the front and rear sides of the Hesco barriers ~0.15 m deep were filled with oyster shell. Five individual segments were constructed, each ~125 m long and 5.5 m wide, with 12 m gaps between. This hybrid technique used cages and rock to retain vertical structure and oyster shell for oyster and mussel settlement.
This shoreline at Swift Tract had been experiencing erosion, so one of the project objectives was to reduce wave energy impacting the shore. Historically (1957- With 4 years of data following reef construction, the annual shoreline loss has dropped to ~0.02 m/yr. (Fig. 13.6). The materials used have not lost their vertical structure during the first 5 years of monitoring. It is possible that given a longer monitoring period (>4 years) we may find that the shoreline has stabilized. While progress has been made toward reducing shoreline erosion in Alabama, a coordinated and comprehensive effort to restore shorelines is necessary to help maximize these benefits for both the communities and the natural habitats. Like traditional bulkheads, nature-based or hybrid shoreline restoration options are not one-size-fits-all. By developing a ready-to-implement plan for a range of individual sites, communities and residents will have a resource available to guide restoration activities that help make their communities more resilient, while also improving habitat and water quality. Broader implementation of sound restoration techniques at the place-based level can: • improve coastal and community resilience by incorporating multiple individual efforts to maintain and improve municipal Community Rating System (CRS) rankings for coastal insurance incentives, • enhance coastal habitat, such as marsh, to uptake nonpoint source nutrients and pathogens and improve water quality before runoff enters coastal tributaries, • improve estuarine habitat for fish and shellfish, coastal birds and other wildlife.
These projects, with the benefit of the lessons learned and adaptive management, can provide long term benefits for the broader community by enhancing fisheries, providing recreational opportunities, improving water quality, retaining property and land values, and increasing community resilience. Collectively, these benefits contribute to a stronger economy and better quality of life.

Study Case 2: Oyster Reefs as Protection Against Tidal Flat Erosion, Oosterschelde, The Netherlands
In the Netherlands three intertidal oyster reefs were constructed in 2010 in the Oosterschelde to investigate their contribution to coastal protection, acting as a natural buffer for erosion control. The Oosterschelde is a coastal bay (tidal range at the study site 3 m, salinity 30 ppt) that suffers from eroding intertidal flats, due to infrastructural works (i.e. construction of a storm-surge barrier) in the 1980s (Nienhuis and Smaal 1994;Walles et al. 2016a). Tidal flats are eroding on average 1 cm. yr. −1 in height, and it is predicted that more than half of the 11,000 hectares will be lost by the end of this century. This has consequences for both nature values as well as coastal protection, as the tidal flats are foraging grounds for an internationally important number of waders like Oystercatchers (Haematopus ostralegus), and at the same time protect the dikes against wave erosion. The constructed reefs, made of 25 cm high gabions filled with Pacific oyster shells, provided substrate for settlement of new oyster recruits. To attenuate waves, the 200 m long reefs were positioned perpendicular to the dominant wave direction at three different elevations: 23%, 35% and 50% emersion time (Walles et al. 2016a) (Fig. 13.3). Over a 5-year period, the development of these reefs and their effect on tidal flat morphology (erosion/sedimentation) were monitored.
Reef development was related to vertical position in the intertidal zone. Recruitment rate, shell growth, and the condition of the oysters (Crassostrea gigas) were correlated with tidal emersion (Walles et al. 2016b). The reef positioned at 50% tidal emersion lacked sufficient settlement of new oyster recruits to maintain the reef structure. As such, the constructed reef shows a continuous deterioration. The reefs positioned at 35% and 23% tidal emersion had sufficient recruitment to maintain their structures. Oyster grew at both reefs, however a loss in weight and a low condition of the oysters on the reef at 35% tidal emersion indicated stress at this elevation ( Fig. 13.7). Rodriguez et al. 2014 showed that reefs have an upper limit with respect to tidal range, a so-called growth ceiling above which reefs cannot grow as stress from limited inundation is too high. For natural reefs in the Oosterschelde the growth ceiling occurs above 60% tidal emersion (Walles et al. 2016b). Based on growth rates observed on the lowest reef, a vertical accretion of the reef base in the order of 7.0-16.9 mm year −1 is expected (Walles et al. 2016a). This is an underestimation of the real reef growth as vertical accretion of the taphonomically active zone is much higher. Over the course of 5 years this reef increased on average 10 cm (unpublished data).
Measurements of waves windward and leeward of the reefs showed that wave attenuation is depending on the size of the incoming waves and the water depth over the reef structure. Close to submergence, reefs potentially attenuate waves up to 95% of the incoming waves. For larger water depths, the attenuation depends on the wave height and wave length. We found that wave height leeward of the reef relates to water depth on top of the reef with a factor of 0.3. This is comparable to rubble mount breakwaters (Van der Meer et al. 2005). For example, for a water depth of 1 m on top of the reefs, waves smaller than 30 cm will not be influenced by the reef anymore.
Elevation measurements showed that the reef positioned at 23% tidal emersion, effectively reduced erosion leeward of the reef, as predicted (Walles et al. 2015a). Up to 90 m leeward of the reef there was a reduction of 51 ± 29% in the erosion measured. The reef however influences the morphology over a much larger area, up to 360 m leeward of the reef.

Management Applications and Considerations
There is growing interest in the restoration and conservation of epibenthic bivalve reefs for coastal protection and shoreline stabilization, also because of their potential to adapt to climate change impacts and sea level rise. This is evidenced by the growing number of projects and studies done worldwide, as well as the increasing funding for nature-based defences or living shoreline initiatives. Despite increasing interest in restoration, the success of bivalve reef restoration is variable (Coen et al. 2007;Mann and Powell 2007). Yet, there are few syntheses of information on what kind of projects meet the goals or are cost effective (Grabowski et al. 2012;Bayraktarov et al. 2016). Data from monitoring the success or failure of restoration projects are not always consistent or readily available, or projects lack sound monitoring efforts at all (Baggett et al. 2015). In order to properly develop guidelines and methodologies for project design, monitoring whether a restoration project designed for coastal protection (or another ecosystem service) has achieved its stated objectives is necessary and should be incorporated in each project setup. A general framework that considers both project success and cost-benefits of restoring epibenthic bivalve reefs for coastal protection or facilitation of other habitats is needed for each project. Therefore, it is necessary to consider carefully how to design and implement restoration projects given the stated goals and targeted ecosystem services. Understanding the interaction between reef-forming ecosystem engineers and the surrounding environment is of utmost importance and several aspects need to be considered when restoring or creating epibenthic bivalve reefs for coastal protection or facilitation of other habitats. First, using epibenthic bivalve reefs for coastal protection or facilitation of other habitats requires knowledge about the habitat requirements of the targeted bivalve species, or 'site suitability' for the selected species (La Peyre et al. 2015). Recruitment, growth and survival of epibenthic bivalves depend on many factors, including salinity, temperature, dissolved oxygen, hydrodynamics, tidal emersion and wave exposure (Baggett et al. 2015;Walles et al. 2016c;Theuerkauf et al. 2017). In addition, biological factors, such as availability of larvae, predation pressure and diseases will influence the long-term survival and sustainability of the constructed reef. This all determines whether a certain site is suitable for the construction of a bivalve reef.
Second, to get optimal effect on shoreline stabilization or facilitation of other habitats, reefs should be constructed at the right position in the intertidal or shallowsubtidal zone. Wave dampening by shellfish reefs depends on water depth (tidal range), wave height/length, and reef height/width. The latter implies that design parameters are also crucial aspects for successful restoration for shoreline protection. Based on the slope of the area the long-term distance of influence by the reef can be estimated. One should realise that there are limitations on the extent that oyster reefs offer shoreline protection or erosion control. When positioned properly, and with the right conditions, reefs can dampen waves considerably, interacting over long distances with the seabed, reducing erosion and facilitating neighbouring habitats and ecosystems such as tidal flats, seagrasses and salt marshes. In other cases, for instance when the bed slope is very steep or the tidal range large, the effect will be strongly reduced. Additionally, the role of reefs in stabilizing the substrate underneath their footprint is a currently under-appreciated service that helps in armouring the coast against erosion.
Thirdly, when one knows where a bivalve reef can potentially grow and act as coastal defence or habitat facilitator, the restoration itself also requires attention. Restoration of epibenthic bivalve reefs can be done in several ways. For oysters, typically some kind of substrate is offered on which oyster larvae can settle. Oyster larvae preferentially settle on conspecifics, so adding loose oyster shells is often used in low-dynamic, sheltered environments, whereas shells packed in bushels or gabions are needed in more dynamic situations. Often there is a shortage of oyster shells, and other substrates are being used such as reef balls, oyster castles, etc. (zu Ermgassen et al. 2016). For restoration of mussel beds, it has been proven that taking into account the behavioural self-organisation of mussels improved restoration success considerably (De Paoli et al. 2017).

Conclusions
In conclusion, epibenthic bivalve reefs are conspicuous ecosystem engineers that modify the soft-sediment environment in which they live to a great extent. Both their physical structure and suspension feeding activity strongly influence the neighbouring sedimentary environment up to 100 s of meters outside their own footprint. Wave attenuation by the reefs reduces erosion, facilitating other species and protecting habitats such as seagrasses and salt marshes. These ecosystem engineering effects of epibenthic bivalve reefs are increasingly recognized and bivalve reefs are promoted as nature-based solutions that provide biodiversity benefits, support many ecosystem services including coastal protection, and contribute to climate change mitigation and adaptation. Much has been learned about how bivalve reefs function and deliver these services, which has accelerated growth and interest in bivalve reef restoration projects ((zu Ermgassen et al. 2016). Objectives for restoration of these reefs should be framed on the basis of the desired ecosystem services provision. This requires a general framework in which habitat requirements of the considered species, environmental setting and long-distance interdependence between habitats are mutually considered (zu Ermgassen et al. 2016).
The coast cannot be restored to its historic condition, but innovation and coordination of techniques and projects can enhance ecosystem function and quality of life and make coastal communities more resilient in the decades to come.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.