Ecosystem Concepts for Sustainable Bivalve Mariculture (2010)

The role of suspension-feeding bivalves in estuarine and marine ecosystems has been extensively documented through research in ecology, physiology, biogeochemistry, mariculture, interdisciplinary marine science, and fisheries science. Suspension-feeding bivalve molluscs consume at the lowest trophic level, feeding largely as herbivores (Duarte et al., 2009). This chapter is divided into three sections to characterize: (1) the biological activities of molluscs (whether wild or cultured) and the effects of their biogeochemical modifications and habitat provision; (2) the incidental impacts of bivalve mariculture operations on multiple components of the ecosystem caused by mariculture structures and activities and by the biological activities of the molluscs; and (3) consequences of actions taken by culturists to alter ecological interactions purposely to manage the effects of pests, competitors, and predators on mariculture systems. The purpose of these sections is to illustrate issues that have been or could be addressed in best management practices—a complete description of the ecosystem services provided by molluscs in both natural systems and in mariculture is provided in Chapter 7 (also see National Research Council, 2009). The last section of the chapter (Uncertainties, Unknowns, and Recommended Research) summarizes issues where additional research will be necessary to determine ecosystem impacts and develop effective mitigation approaches.

Benthic suspension feeders, such as many species of bivalve molluscs, influence the nutrient and organic coupling of benthic and pelagic systems (Dame, 1996) through their ability to filter a wide size range of particles and deposit organic wastes that sink to the bottom (biodeposition). Suspension-feeding bivalves perform this function in a range of habitats and physiographic conditions (e.g., estuaries, lagoons, coastal oceanic systems) where they filter out and deposit significant amounts of suspended material, as well as excrete dissolved nutrients. In estuarine systems, the influence of benthic suspension-feeding bivalves on benthic-pelagic coupling, turbidity, nutrient remineralization, primary production, deposition, and habitat complexity has been well documented (reviewed in Dame and Olenin, 2005). Kaiser (2001) reviews the effects of molluscan cultivation on the ecology of systems, identifying a similar set of mechanisms of influence, and concludes that such processes have a generally positive influence on the overall water quality of a system. Suspension-feeding bivalves also drive many other biogeochemical processes and cycles, which are well described for intertidal oysters by Dame (2005).

Molluscs influence nutrient dynamics through direct excretion and indirectly through microbially mediated remineralization of their organic deposits in the sediments (McKindsey et al., 2006a). Therefore, nutrient regeneration is related to the abundance and location (shallow versus deep water) of bivalves in a system. The extent to which this affects overall nutrient budgets and thus primary production is related to the system flushing rate and residence time (Dame, 1996; Newell et al., 2005). The subsequent proportions of elements in the system will influence the levels of recycling and possibly result in one or more being limited (Dame, 1996).

The majority of studies of bivalve effects on nutrient recycling have focused on nitrogen because this is the most common nutrient-limiting biological production in marine and estuarine systems (Parsons et al., 1983; Howarth, 1988; National Research Council, 2000). Benthic bivalves are important contributors of nitrogen (usually in the form of ammonium, NH₄⁺) to both subtidal and intertidal systems. Nixon et al. (1976) conclude that nitrogen flux across oyster reefs is highly variable and is heavily influenced by tidal flow. Dame (1986) reviews a body of work relating to nutrient fluxes involving Crassostrea gigas in northern France and concludes that 15–40% of nitrogen in the system was derived from the oysters. In addition, measured values were always higher than the estimated values, likely due to remineralization occurring in adjacent

sediments. The suggestion that macroalgal cover of mussel beds will intercept nitrogen (Asmus and Asmus, 1991), thereby making it unavailable for phytoplankton production, has also been proposed for other systems (Mazouni et al., 1998; Lin et al., 2005; see Wadden Sea box in Chapter 4). In contrast, nitrogen is retained within some systems through direct recycling of nitrogen from bivalves (e.g., Crassostrea virginica) to phytoplankton (Dame and Libes, 1993; Newell et al., 2005). Numerous studies have demonstrated that nutrients derived from biodeposits and/or excreted nitrogen serve to enhance growth of eelgrass and other submerged aquatic vegetation (see below).

In the Marennes-Oléron culture region in France, Leguerrier et al. (2004) show that higher oyster production increased benthic-pelagic coupling, which in turn increased secondary production (in the form of meiofauna), providing food for juveniles of predatory nektonic species. Also, Mazouni (2004) and Newell et al. (2005) demonstrate that other planktonic organisms (bacteria, ciliates, and flagellates) can act as sources of nitrogen for bivalve molluscs in the absence of suitable autotrophic phytoplankton.

Phosphorus is important to biological systems, and phosphorous budgets constructed in and around mollusc assemblages show considerable removal of this nutrient from the system through biodeposition. Asmus et al. (1990) demonstrate that mussel beds with large macroalgal populations released less phosphate than beds without a large macroalgal component. Silicon is an important element for diatoms and can be limiting in systems dominated by diatoms. Bivalve molluscs contribute to recycling of silicate through transfer of this nutrient from the water column to the sediment with little being sourced from the bivalves (Prins and Smaal, 1994). Molluscs, such as mussels, may also selectively feed on components of particulate matter and thereby concentrate certain metals like copper in their pseudofeces (Allison et al., 1998).

The production of pseudofeces in large quantities is an important mechanism by which bivalves couple the water column to the bottom (see review in Dame, 1996). Epifaunal bivalves (oysters and mussels) have a plastic response to increasing levels of plankton and detritus in the water column with ever-increasing filtration capacity and production of pseudofeces. However, this response is not observed in infaunal bivalves (clams and cockles), which regulate ingestion rates at high-seston concentrations by adjusting clearance rates rather than by increasing production of pseudofeces (e.g., Foster-Smith, 1975; Bricelj and Malouf, 1984; Bricelj et al., 1984; Prins et al., 1991; Iglesias et al., 1996). Oysters and mussels are also known to tolerate relatively high levels of suspended inorganic particles and continue to filter and produce higher levels of biodeposits.

The positive and negative feedback mechanisms observed in aquatic systems as a consequence of nutrient dynamics mediated by molluscs

have been the subject of numerous studies (Dame, 1996; Prins et al., 1998; Newell et al., 2005). Their high filtration capacity, rapid response to high levels of food (e.g., plankton), and relative permanence in aquatic systems give bivalves the ability to stabilize systems and enhance resilience to perturbations (Jackson et al., 2001a; Newell, 2004). Large bivalve assemblages can regulate the abundance of phytoplankton in shallow seas (see Newell et al. [2005] and McKindsey et al. [2006a] for list of relevant studies), and intense filtering can reduce phytoplankton bloom intensity while extending the duration of less intense blooms (Herman and Scholten, 1990). Filtration and biodeposition of phytoplankton and other suspended materials by extensive beds of bivalves also reduce downstream transport, thereby moderating effects of excess nutrients or sedimentation in outlying waters. Thus, bivalves provide the system with a capacity to buffer against sudden perturbations (DeAngelis et al., 1986; Jackson et al., 2001a; Lotze et al., 2006). The large-scale removal of bivalves from a system has resulted in some well-documented shifts in system processes and has contributed to general degradation of water quality or, more appropriately, a reduction in the resilience of the system to perturbations like nutrient loading and sedimentation (e.g., oysters in Chesapeake Bay; see Newell et al. [2005, 2007] and Pomeroy et al. [2006]).

Many estuaries, such as Chesapeake Bay, and coastal oceans suffer from eutrophication, in which excess nutrients enter waterways from land-based sources and atmospheric deposition (e.g., sewage treatment plants, farm animal wastes, agricultural use of fertilizers, industrial releases of nitrogen oxides or ammonia) and trigger massive blooms of phytoplankton and other algae. Phytoplankton blooms reduce water clarity and deplete the water of oxygen as they die and decompose. Bivalves can reduce excessive growth of phytoplankton and, at high density, can counteract symptoms of eutrophication, thereby improving local, and in some cases downstream, water quality. Yet many bivalve molluscs have been depleted by overfishing, especially oysters (Jackson et al., 2001a; Kirby, 2004; Lotze et al., 2006; Beck et al. 2009), but also clams (Peterson, 2002; Kraeuter et al., 2008) and scallops (Peterson et al., 2008). Consequently, augmenting suspension-feeding bivalves, preferably native, through restoration and mariculture has the potential to enhance suspension-feeding activity and controls in systems where natural populations have been depleted (Jackson et al., 2001a).

In addition to nutrient cycling, molluscs contribute to biogeochemical processes through shell formation, which captures carbon in the form of calcium carbonate and can lead to sequestration of carbon in marine

sediments after natural mortality of wild molluscs or terrestrial burial of shells after consumption of wild-caught or cultured molluscs. The shells of molluscs (living and dead) accumulate in various types of structures in estuarine, coastal, and oceanic systems (Shumway and Kraeuter, 2000). Surface shell accumulations provide a range of ecosystem services, primary among which are structural habitat (e.g., refuge, complexity) and erosion reduction (Coen and Grizzle, 2007).

Shell is also an important source of sedimentary carbonate content. The carbonate budget of estuarine and coastal waters is now of concern because of extensive shell extraction (through mollusc harvesting and mining for construction), the prohibitive cost of long-term continuous substrate provisioning to support fisheries, and the loss of shell via reduced bivalve populations resulting from fishing and disease processes (Mann and Powell, 2007). Moreover, growing ocean acidification caused by increasing concentrations of atmospheric CO₂ has serious implications for seawater carbonate chemistry (Brewer, 1997; Caldeira and Wickett, 2003; Feely et al., 2004; Doney et al., 2009). Recent studies have shown that bivalve growth, development, and survival are negatively affected by decreased pH (e.g., Berge et al., 2006; Fabry et al., 2008; Kurihara, 2008). The change in carbonate water chemistry and concomitant decrease in viability of bivalve molluscs potentially will reduce both the provisioning and persistence of shell in coastal and estuarine systems, particularly those in high-latitude areas with low alkalinity seawater (Feely et al., 2004; Lee et al., 2006). Availability of abundant mollusc shells in the surface sediments can provide local buffering against increasing acidity.

The importance of the interactions between ecological communities and sedimentary carbonate content was articulated in a conceptual model that described a positive feedback process between benthic molluscs and carbonate addition to the sediments. The taphonomic (process of fossilization) feedback hypothesis underlying this conceptual model (Kidwell and Jablonski, 1983) states that increasing shell content encourages settlement of calcifying organisms, and their deaths increase the rate of carbonate addition, forming a positive feedback process. Recent studies have shown that the interaction between carbonate content and community dynamics is critical to ecosystem dynamics in estuarine systems (Gutierrez et al., 2003; Powell et al., 2006; Powell and Klinck, 2007). The species benefiting most are the carbonate producers, particularly bivalves that, through their own deaths, provide a critical sedimentary constituent promoting the long-term survival of their species.

Shell is an essential component of present-day estuarine and coastal ecosystems; however, it is not a stable resource (Powell et al., 2006). Shell must be continually renewed and will disappear rapidly if the processes that support this renewal are slowed or stopped. Carbonate loss possibly

exceeds gain in shallow-water marine ecosystems today (see discussion in Powell et al. [2006]). It is likely that current environmental conditions and commercial mariculture practices, when coupled with predicted changes, such as ocean acidification, will facilitate and accelerate carbonate loss in estuarine and coastal systems. Thus, management of shell-producing commercial species must also include management of the habitat that will maximize production of carbonate. Long-term sustainability of mollusc stocks depends upon the maintenance of a positive shell budget for carbonate, as well as provision of habitat that supports recruitment, growth, and survival of bivalves. Mariculture of bivalve molluscs can contribute favorably to shell production and preservation in coastal ecosystems if the operators return the shell resource to the environment after harvest. However, regulations requiring the return of shells to the estuarine, lagoonal, or coastal bottom after shucking may be required to achieve this goal.

Shell adds hard substrate and habitat complexity to soft substrates, thereby increasing species diversity (Wells, 1961; Larsen, 1985; Coen et al., 1999; Harding and Mann, 2000; 2001; Mann, 2000; Gutierrez et al., 2003) and enhancing recruitment and survival of bivalves (Haven and Whitcomb, 1983; Abbe, 1988; Kraeuter et al., 2003; Bushek et al., 2004; Green et al., 2004; Soniat and Burton, 2005). When present in significant amounts, shell adds bottom-habitat complexity to the ecosystem (Haven and Whitcomb, 1983; DeAlteris, 1988; Grizzle, 1990; Powell et al., 1995; Allen et al., 2005). Fish have been shown to associate with both biogenic and artificial structures on the bottom, such as eelgrass, bivalve reefs, and the legs of oil platforms, as a consequence of attraction to structured habitat for protection or feeding (Franks, 2000; Heck et al., 2003; Peterson et al., 2003; Coen and Grizzle, 2007; Horinouchi, 2007; Jablonski, 2008).

Seagrasses are often considered to be an extremely important plant in estuaries and lagoons where they form emergent structural habitat for fish and invertebrates in these soft-sediment systems (Jackson et al., 2001b; Williams and Heck, 2001; Heck et al., 2003; Bostrom et al., 2006). Local improvements in water clarity induced by filter-feeding bivalves can promote the spread of eelgrass, especially to depths where light would otherwise be limiting (Dennison et al., 1993). Augmentation of nutrient concentrations in sediments can also stimulate eelgrass growth, as has been shown to occur for eelgrass growing alongside mussels in Europe, Florida, and southern California (Reusch et al., 1994, Reusch and Williams, 1998; Peterson and Heck, 1999; 2001a, b). Many estuaries on the west coast of the United States are flushed with relatively nutrient-rich ocean waters, and under these circumstances, eelgrass may not benefit as

much from the additional nutrients released by bivalves (Dumbauld et al., 2009). Both the reduction of turbidity and fertilizing effects of bivalve molluscs have been demonstrated experimentally for modest densities (16 per m²) of hard clams (Mercenaria mercenaria) in a relatively oligotrophic Long Island estuary (Carroll et al., 2008). Positive effects of a modest number of suspension-feeding bivalves are more likely to benefit eelgrass in relatively oligotrophic water bodies, where functional enhancement of water clarity may be achieved without a huge increase in filtering capacity (Carroll et al., 2008).

Several factors contribute to the rate of production of biodeposits, including the distribution, density, and the species of bivalves coupled with environmental conditions, such as food concentrations, water temperature, turbidity, and feeding rates of the bivalves (Jaramillo et al., 1992; Dame, 1996). Rates of accumulation or dispersion of the biodeposits also depend on water movements close to the seafloor (Widdows et al., 1998; Callier et al., 2008). Generally, mariculture activities in well-flushed intertidal areas are likely to result in dispersal of the organic biodeposits, whereas subtidal mariculture in quiescent areas has the potential of producing a greater accumulation of biodeposits and consequently a greater localized impact on the benthos. The vast majority of the literature pertaining to organic enrichment has focused on mussel farming. Most studies have concluded that the effect of bivalve mollusc farming is relatively small and much less than that caused by finfish farming where organic matter is added to the system as food (e.g., Baudinet et al., 1990; Grant et al., 1995; Buschmann et al., 1996; Cranford et al., 2007; Zhang et al., 2009). Only a few studies have characterized organic loading from mollusc farms as high (e.g., Dahlbäck and Gunnarsson, 1981; Mattsson and Linden, 1983; Metzger et al., 2007), and these are cases in which cultured mussel densities are high and/or tidal circulation is low.

Culture operations for bivalves interact with aquatic plants through displacement of seagrass by the cultured bivalves and associated culture structures, through disturbance caused by shellfish planting and harvesting, through provision of unnatural hard substrates involved in culturing, through physical modification of flows regimes and sediments, and through water clarification and nutrient delivery to the bottom. Facili-

tation of benthic plants can occur when bivalve molluscs or associated culture structures provide attachment sites for macroalgae, the growth of which provides ecosystem services (habitat and nutrient sequestration; DeAlteris et al., 2004; Luckenbach and Birch, 2009). Eelgrass and other submerged aquatic vegetation (SAV) species can benefit from increased light penetration that expands the range of suitable bottom for occupation by SAV and from fertilization of the plants with biodeposits, as discussed earlier. In addition, bivalve mariculture activities can have negative effects on SAV. In Willapa Bay, total production of eelgrass was lower in areas with oyster mariculture (Tallis et al., 2009). The relative growth rate of eelgrass was unaffected by the presence of oysters or geoducks in Willapa Bay and Totten Inlet, respectively. However, in these examples, shoot size varied and may have been responding to increased porewater ammonium or reduced intraspecific competition when molluscs were present (Dumbauld et al., 2009; Tallis et al., 2009). Augmentation of sediment nutrient concentrations is known to stimulate eelgrass growth in some locations (see earlier section, Habitat Creation and Maintenance). Theoretically, high levels of biodeposits could lead to toxic sulfide concentrations, but this has only been shown to occur when conditions were already eutrophic (Vinther et al., 2008). Finally, bivalve culture can stimulate growth of several species of macroalgae (DeCasabianca et al., 1997; Vinther et al., 2008), which can in turn negatively affect seagrasses (Hauxwell et al., 2001).

Seagrasses are subject to multiple anthropogenic disturbances, which have been shown to be at least partly responsible for a general worldwide decline in their abundance (Orth et al., 2006; Waycott et al., 2009). Seagrasses are highly susceptible to rapid changes in their environment because of their requirement for high-incident light levels and their restriction to relatively shallow nearshore coastal waters (Dennison and Alberte, 1985; Orth et al., 2006). Eelgrass, Zostera marina, is one of the more common species studied in relation to bivalve mariculture because of its worldwide distribution in temperate seas. The upper distributional limit of Z. marina is determined primarily by desiccation (Boese et al., 2005) and the lower limit determined by light penetration, which is affected by turbidity in the estuary. Z. marina distribution overlaps directly with the area where most bivalve culture occurs, extending to almost –10 m where water clarity is high on both coasts of the United States (Phillips, 1984; Moore et al., 1996; Thom et al., 2003; Kemp et al., 2004). The enhancement of water clarity by suspension-feeding bivalves thus relieves an intrinsic limitation to the spread of eelgrass.

In some areas, mollusc culture operations and aquatic vegetation compete for space. However, this relationship is not one-to-one. In Willapa Bay, Washington, an apparent threshold has been detected above which

eelgrass declined by more than the area covered by ground-cultured oysters, while at lower levels of oyster cover, eelgrass was more abundant than predicted from simply the amount of space available (Dumbauld et al., 2009). Part of the threshold effect has been attributed to the severing of eelgrass blades by the sharp tips of the oyster shells (Schreffler and Griffen, 2000), reducing its percent cover and possibly reproductive capacity. Shading from overwater structures is another form of negative interaction. Work conducted by Everett et al. (1995) in Coos Bay, Oregon, found 100% loss of eelgrass directly under oyster racks, presumably resulting from shading and sediment erosion (10–15 cm at the base of the structure). Smaller reductions in eelgrass cover and density have been documented with other forms of off-bottom culture, such as longlines and stakes, but losses tended to scale with density or spacing and were restricted primarily to the area beneath lines and stakes where shading or sedimentation occurred (Everett et al., 1995; Rumrill and Poulton, 2004; Tallis et al., 2009). In one of the few landscape-scale studies that monitored changes for a long period of time, eelgrasses in Bahia de San Quentin, Mexico, did not decline as might be expected from shading by oyster culture racks (Ward et al., 2003).

The degree to which benthic invertebrate populations and communities are impacted by bivalve mariculture is typically related to the scale of operation, the species and culture techniques being used, and the physical and hydrodynamic characteristics of the culture site. As a result, scientific studies demonstrate a broad range of responses of benthic infauna to mariculture, ranging from no or moderate negative effects to positive effects. In addition to the relatively complex nature of the impacts of bivalve culture on benthic invertebrate populations and communities, many of the studies have focused only on the grow-out phase of cultivation rather than assessing all aspects of the cultivation process (Kaiser et al., 1998). For instance, although collection of wild mussel seed for most commercial cultivation is done by the use of spat collectors, in a few locations (e.g., Maine in the United States, the Wadden Sea in Germany and the Netherlands, the Irish Sea) seed is harvested by bottom dredging in subtidal areas (see Box 4.2 on the Wadden Sea), resulting in greater impacts on benthic habitat. Lastly, disturbances to benthic habitats associated with routine maintenance, harvesting, and handling of the molluscs are also not normally evaluated in published studies. Much of the research regarding the effects of bivalve mariculture on benthic invertebrate populations has focused on the following two areas: (1) effects of increased organic loading to the sediments from bivalve biodeposits and (2) habitat

modification associated with the off- and on-bottom mariculture gear (e.g., racks, cages, bags) and the replacement, reduction, or enhancement of the local fauna by the cultivated bivalve species. The relative influence of each of these on benthic habitats varies depending upon the factors previously mentioned.

Bivalve culture can modify benthic habitats in a number of positive and negative ways. For example, growing a species on the seafloor (e.g., oysters) increases habitat structure and enhances local biodiversity relative to soft-sediment landscapes (e.g., Ferraro and Cole, 2007). Folke and Kautsky (1989) suggest that large-scale mussel culture can result in structural changes in marine ecosystems by indirectly affecting the recruitment of other commercially important species. In addition, adult bivalves can remove larvae of some invertebrate species through their filtering activities. Pechenik et al. (2004) demonstrate that adult Pacific and European flat oysters were capable of filtering the larvae of the slipper shell snail, Crepidula fornicata, although ample numbers of C. fornicata larvae survived through settlement and metamorphosis. Similarly, Troost et al. (2008) show that an escape response was elicited when Pacific oyster and blue mussel larvae were subjected to suction currents similar to those of adult Pacific oyster feeding currents. However, both studies acknowledge that experimental conditions were not necessarily reflective of natural conditions where many other factors come into play. Thus, the potential for high-density bivalve culture to impact recruitment of benthic species with planktonic larvae requires further study.

Structures used in some types of mariculture operations, such as racks, bags, and ropes, can increase biodiversity by providing more habitat for fouling species (e.g., Powers et al., 2007) but also can alter the hydrodynamics of an area to some degree (see review by Kaiser et al. [1998]). These structures can redirect water flow and produce either scouring or accretion of sediment around the structures, depending on the local hydrodynamic regime (Hecht and Britz, 1992; Everett et al., 1995). At an intertidal Pacific oyster farm in Dungarvan Bay, Ireland, tides and strong currents around the farm site prevented organic enrichment beneath oyster trestles by dissipating biodeposits, but in access lanes that were subject to compaction and dispersal of the sediment by boat traffic, the species composition and abundance of certain epibenthos and infauna differed significantly when compared with those parameters at a distant control site (de Grave et al., 1998). Castel et al. (1989) note that Pacific oyster culture on suspended racks in Arcachon Bay, France, increased sedimentation and enhanced the accumulation of debris (e.g., shells, macroalgae). An investi-

gation of the effects of two types of oyster mariculture on sediment surface topography by Everett et al. (1995) found that stake culture resulted in a significant increase in sediment deposition, whereas rack culture resulted in more erosion compared with reference sites.

Re-seeding large areas of the seabed with cultured or wild-collected bivalve seed stock and then harvesting market-sized individuals by dredging is common culture practice in many parts of the world (e.g., United States, France, the Netherlands, Ireland, Japan). Dredging has been widely reported to cause significant habitat and community changes (Dayton et al., 1995; Jennings and Kaiser, 1998; National Research Council, 2002). Dankers and Zuidema (1995) found that the most obvious impact of mussel culture on the Dutch Wadden Sea environment was dredging of seed mussels, which reduced the food supply for several bird species (see Chapter 4 for a more detailed discussion of harvest effects and Box 4.2 on the Wadden Sea as a case study). In some regions, the culture area is also mechanically worked to remove predators and prepare the substrate for re-seeding. For example, in Japan, re-seeded scallop beds are scraped with a “mop” to remove predators. Relatively large areas (e.g., square kilometers) can be affected, and the mopping activity can substantially alter the benthic epifaunal community structure.

Studies of bivalve mariculture operations, mostly off-bottom, have shown higher abundances of some fish and crustaceans in areas with mariculture structures in comparison to nearby areas with unstructured open mudflats, eelgrasses, or even nearby oyster reefs and rocky substrates, although eelgrass generally harbors more unique species (DeAlteris et al., 2004; Clynick et al., 2008; Erbland and Ozbay, 2008). A study of flatfish behavior showed that juvenile sole utilized oyster trestles for protection during the day and foraged over adjacent sand flats at night (Laffargue et al., 2006). A number of studies have documented the positive influence of suspended mussel mariculture on food resources and therefore abundance of large macroinvertebrates and fish (Freire and GonzalezGurriaran, 1995; D’Amours et al., 2008). A study in Narragansett Bay, Rhode Island, found that scup (Stenotomus chrysops) grew slightly faster on adjacent rocky habitats than in oyster mariculture bottom cages; tagging suggested that they had greater fidelity to the oyster cages (Tallman and Forrester, 2007).

Powers et al. (2007) demonstrate that densities of fish and free-swimming invertebrates in North Carolina are as high over cultured clams in plastic bottom net bags (and associated fouling epibiota) as in eelgrass beds, with much lower fish and invertebrate densities over

unvegetated bottoms. However, abundance estimates are not necessarily an indication of how structured habitats benefit fish because structures can attract fish without enhancing their productivity (e.g., reproduction, growth, survival). This is the classic “production versus aggregation” debate. Nevertheless, experimental research has shown that artificial reef structures provide nektonic organisms with protection against predation, thereby offering a survival advantage, especially to more vulnerable juvenile life stages (Dempster and Taquet, 2004). Also, gut contents reveal that demersal fish associated with structures are consuming organisms found on and enhanced by the availability of the hard substrates (Posey et al., 1999; Peterson et al., 2003).

Studies of fish around bivalve mariculture operations in U.S. west coast estuaries provide useful insights into the interactive processes that may occur between mariculture structures associated with bivalve mariculture and mobile species. In Humboldt Bay, California, oyster long lines were found to harbor more fish than either eelgrass or open mudflats (Pinnix et al., 2004). In Willapa Bay, Washington, few statistically significant density differences were found among the more than 20 species of fish and crabs collected at intertidal locations when oyster bottom culture, eelgrass, and open mudflats were compared (Hosack et al., 2006). In both studies, some individual species like tube-snouts (Aulorhynchus flavidus) were more abundant in structured habitats. In a preliminary study submitted as a project report to the National Park Service (Elliott-Fisk et al., 2005), Wechsler (2004) examined the potential effects of oyster mariculture on fish communities in Drakes Estero, California. No significant differences in fish abundances or species richness were detected among three sampling sites; however, there was an indication that fish assemblages were modified near oyster racks by enhanced numbers of the guild characterized as “structure-associated fishes.” This pattern was driven primarily by increases in one species (kelp surfperch, Brachyistius frenatus), typically associated with hard substrate (Wechsler, 2004; Elliott-Fisk et al., 2005).

Larger mobile invertebrates have also been shown to display modified species-specific and even life-stage-specific behaviors around structure. In one study, juvenile Dungeness crabs (Cancer magister) utilized artificial structures, but older individuals utilized open mudflats, whereas red rock crabs (Cancer productus) preferred on-bottom oyster culture structures (Holsman et al., 2006).

The following are three areas in which bivalve genetics are pertinent to the development of best practices for mariculture: (1) domestication

and genetic improvement of molluscs for mariculture; (2) genetic impacts of translocations or introductions of molluscs; and (3) genetic impacts of interbreeding between hatchery stocks and wild populations, such as might arise in either bivalve restoration programs or commercial mariculture.

Bivalve mariculture is faced with the dual challenge of becoming more efficient (producing more from less area) and of adapting to a changing ocean. Genetic improvement and domestication are proven routes to increase the efficiency of agricultural production across a range of environments, but research toward these ends in bivalve mariculture is in a primitive state. Unique challenges are, moreover, presented by the high fecundity of these animals.

Though cultivated since Roman times (Günther, 1897), bivalve molluscs are in a proto-domestication phase (Harris and Hilman, 1989): diverse species are no more than exploited captives (Clutton-Brock, 1981; Duarte et al., 2007). Obvious candidates for concerted domestication efforts are the seven bivalve molluscs among the top-40 species of global aquaculture (C. gigas, Ruditapes philippinarum, Patinopecten yessoensis, Sinonovacula constricta, Mytillidae, Anadara granosa, and Perna viridis), yet the knowledge base for domesticating and improving these top-producing bivalves is shockingly narrow. A principal limitation to assessing genetic improvement and domestication is a lack of basic, detailed, mariculture statistics. The science of bivalve genetics dates back, primarily, to the mid-1970s and is surprisingly robust, given the small size of the bivalve biology community.

Work on bivalve population genetics to date has focused primarily on geographic subdivision and the causes of marker-associated heterosis (superiority of marker heterozygotes to homozygotes with respect to growth, survival, and other fitness traits) in natural populations (e.g., Zouros et al., 1980; Fujio, 1982; Buroker, 1983; Gaffney, 1994; Zouros and Pogson, 1994; Bierne et al., 1998; David, 1998; Launey and Hedgecock, 2001); heterosis in yield—the product of growth and survival—in experimental populations (Hedgecock et al., 1995; Hedgecock and Davis, 2007); the heritability of production characteristics, mostly in oysters (Lannan, 1980; Newkirk, 1980; Sheridan, 1997; Langdon et al., 2003; Dégremont et al., 2007); and the development of genomic approaches to understanding complex traits and physiological ecology, mostly in oysters and mussels (Hedgecock et al., 2005; Saavedra and Bachere, 2006; Hedgecock et al., 2007a; Gaffney, 2008; Gracey et al., 2008; Tanguy et al., 2008).

Most bivalve genetic diversity resides in natural populations, from which mariculture stocks are derived continuously. As demonstrated by early allozyme studies and reinforced now by numerous DNA studies, bivalves are among the most genetically variable animals. This diversity extends to additive genetic variance in quantitative traits, such as

response to selection, as has been recorded in studies of disease resistance (Hershberger et al., 1984; Haskin and Ford, 1987; Dégremont et al., 2007) and yield (Langdon et al., 2003). Non-additive genetic variance is also important in highly fecund bivalves, as evidenced by yield heterosis (hybrid vigor) in the Pacific oyster that is as dramatic as that in maize (Shull, 1908; Crow, 1998), even in crosses among inbred lines derived from the same wild population (Hedgecock and Davis, 2007). Yield heterosis is associated with equally dramatic inbreeding depression (Evans et al., 2004), attributed to a remarkably large load of deleterious recessive mutations (Bierne et al., 1998; Launey and Hedgecock, 2001). Inbreeding depression can easily eliminate or reverse gains from selection. A large mutational load in bivalves was predicted by Williams (1975), in the Elm-Oyster model for the advantages of sexual reproduction in species with high fecundity and high early mortality. Since high fecundity and high early mortality are the dominant life history features among marine fish (Winemiller and Rose, 1992) and invertebrates (Thorson, 1950), considerable scope for genetic improvement likely lies in crossbreeding of inbred lines.

The best practice for bivalve breeding is to take advantage of both additive genetic variance, through selection, and non-additive genetic variance, by identification of selected inbred lines for crossbreeding. Development of genomic resources promises to accelerate discovery of phenotypic-genotypic associations, the genes underlying economically important traits, and methods for determining the breeding or crossbreeding values of broodstock at early life stages (Pace et al., 2006; Hedgecock et al., 2007a).

Since the oyster and other bivalve industries have shifted heavily toward use of triploids because non-reproductive oysters enhance production (Nell, 2002)—also a welcome trend for minimizing impacts on natural populations, as discussed below—breeding programs seek to improve triploid, as well as diploid, seed. Triploid seed is currently produced by fertilizing diploid eggs with sperm from tetraploid males (Guo et al., 1996; National Research Council, 2004). Existing tetraploid stocks of the Pacific oyster were derived haphazardly from a rather narrow genetic base of wild diploid oysters. To take full advantage of additive and non-additive genetic variance for yield, breeders will need to build new tetraploid lines that incorporate good genes and genetic combinations from diploid lines. Biosecurity of reproductively competent tetraploid stocks in the environment is an issue that is just beginning to be addressed (Piferrer et al., 2009); early experience with tetraploid Pacific oysters suggests that they are not robust enough, at present, to have a negative impact on reproductive success of diploid stocks.

Even when native molluscs are used in mariculture, the natural genetic structure can be disrupted via interbreeding between wild and cultured

genotypes, potentially jeopardizing wild populations by decreasing their adaptive potential (Lynch, 1991; Allendorf et al., 2001). The risks depend on the amount of genetic divergence between the wild and cultured populations. Marine molluscs, with widely dispersing planktonic larvae, typically show minimal genetic divergence over broad scales (Hedgecock et al., 2007b). The Eastern oyster (C. virginica) is a notable exception, with a major genetic discontinuity between Gulf of Mexico and Atlantic populations (Buroker, 1983; Reeb and Avise, 1990; Karl and Avise, 1992; Cunningham and Collins, 1994; McDonald et al., 1996). A regional sub-population divide along the mid-Atlantic coast has, further, been identified with molecular markers (Hoover and Gaffney, 2005; Gaffney, 2006) and may correspond with the races identified earlier on physiological grounds (Loosanoff and Nomejko, 1951; Barber et al., 1991). Although genetic impacts from historical translocations of Eastern oysters have yet to be reported, the precautionary approach dictates that proposed translocations ought to be preceded, at least, by a determination of the population genetic structure of the target species (Bell et al., 2005; Ward, 2006) and, ideally, also by quantitative analysis of local adaptation.

The majority of marine bivalve molluscs share a suite of life-history traits—relatively late maturation, high fecundity, small eggs, long-lived plankton-feeding larvae with relatively high-dispersal potential, and broad geographic ranges (Winemiller and Rose, 1992)—that renders them more vulnerable to loss of variation and extinction than might be expected from their sheer abundance (Palumbi and Hedgecock, 2005). Reproductive success, because it involves a complex chain of events for molluscs, may vary dramatically among individuals, perhaps even among individuals adjacent to one another in space but spawning at slightly different times. Consequently, reproductive success in marine organisms is hypothesized to resemble, at times, a sweepstakes lottery, in which there are a few big winners and many losers (Hedgecock, 1994). Support for this hypothesis has come from both empirical (e.g., Li and Hedgecock, 1998; Hauser et al., 2002; Turner et al., 2002; Hedgecock et al., 2007c; Lee and Boulding, 2007; 2009) and theoretical (Waples, 2002; Hedrick, 2005; Eldon and Wakeley, 2006; Sargsyan and Wakeley, 2008) studies. The conservation implication is that even abundant bivalve stocks may have effective population sizes (as reflected in genetic diversity) that are orders of magnitude smaller than census sizes and, thus, rates of genetic drift and inbreeding that can erode biodiversity on ecological time scales.

Adverse interactions of wild and hatchery-propagated stocks are growing with the global expansion of mariculture for finfish, such as salmon (McGinnity et al., 2003; Hindar et al., 2006), and stock enhancement programs, including shellfish restoration efforts (Born et al., 2004; Gaffney, 2006). High fecundity and large variance in reproductive suc-

cess in hatchery stocks (Gaffney et al., 1993; Boudry et al., 2002) create a risk of diluting the genetic diversity of wild populations with hatchery-propagated bivalves (Allen and Hilbish, 2000; Gaffney, 2006).

One way to eliminate the risk of interaction between wild and hatchery stocks is to render farmed stocks sterile. Triploidy is commonly induced in bivalves to reduce reproductive effort, divert energy to growth, and improve meat quality during the normal spawning season (Allen and Downing, 1986; Nell, 2002). Because triploids are effectively sterile, their use in bivalve mariculture dramatically reduces but does not eliminate the risk of spawning and mixing with local native or naturalized stocks. If an introduced or farmed species is a nonnative, however, triploidy may offer only a short-term reduction in the risk of an introduction (National Research Council, 2004). Gene knockout offers another means of sterilization (Grewe et al., 2007; Wong and van Eenennaam, 2008), but public resistance to genetically modified organisms makes this a less attractive strategy.

To augment or replace depleted natural stocks or to diversify the number of species used in mariculture operations, managers of molluscs in the past have employed translocations of native species and introductions of nonnative species. No new nonnative bivalves have been introduced for mariculture purposes for several decades (Naylor et al., 2001), although introduction of the nonnative Asian oyster (Crassostrea ariakensis) was proposed by Virginia and Maryland as a strategy for replenishing the oyster population in Chesapeake Bay (National Research Council, 2004). Virginia and Maryland have since decided not to move forward with the introduction of the Asian oyster following the U.S. Army Corps of Engineers’ preferred alternative for the use of native Eastern oysters over the nonnative Asian oyster in restoration activities (U.S. Army Corps of Engineers, 2009). The introduction of nonnative species in mariculture has also been responsible for the unintentional importation of other nonnative species (i.e., “hitchhikers”). In most cases, current bivalve mariculture best management practices prevent the unintentional introduction of hitch-hiking species.

There are several reviews on the importation of nonnative molluscs for mariculture (e.g., Andrews, 1980; Chew, 1990), particularly Pacific oysters (C. gigas) (Coleman, 1996; Shatkin et al., 1997; Ruesink et al., 2005). In some instances, these importations have resulted in the establishment of naturalized (breeding) populations of the nonnative molluscs that has affected resident oyster species. For example, there is evidence that naturalized populations of C. gigas have become a significant competitor of

native oyster species in France (Goulletquer and Heral, 1991), Australia (Ayres, 1991), New Zealand (Dinamani, 1991), and the western United States (Trimble et al., 2009). While it should be noted that not all nonnative shellfish introductions have led to negative consequences on the native species, nonnative species often exhibit faster growth rates than equivalent native species (e.g., C. gigas; Ruesink et al., 2005) and thus are apt to be superior competitors for resources. However, some faster growers, such as the triploid Asian oyster (C. ariakensis), allocate fewer resources to shell thickness and are thus more susceptible to predation by crabs and perhaps other predators (Bishop and Peterson, 2006).

In addition to affecting native, economically important species, culturing of nonnative bivalve species may influence native biodiversity, have direct and indirect influences on local community composition, and influence the performance of ecosystems with resultant economic impacts. Although there is a burgeoning literature cataloging and assessing the impacts of introduced species in coastal waters (e.g., reviews of Carlton, 1985; 1987; 1989; Ruiz et al., 1997; 1999; 2000; Grosholz, 2002), Ruesink et al. (2005) note there is a surprising lack of information on the effects of nonnative oyster introductions on community- and ecosystemlevel structure and function and on how similar the ecosystem services provided by nonnative species are to native species. There also appears to be a similar general lack of knowledge regarding the impacts of other nonnative bivalve species (e.g., clams and scallops) that are commonly used in mariculture operations (Whiteley and Bendell-Young, 2007). The National Research Council (2009) details the impacts and risks of non-native species introductions, focusing on Pacific oysters.

Several practices are used to reduce the risk of the establishment of naturalized populations from nonnative cultured bivalves, including the use of triploid seed or the culture of bivalves in areas with low potential for the establishment of a wild population. To date, the use of triploid nonnative bivalves on a commercial basis has largely been restricted to C. gigas, and in 2002, about one-third of the “eyed larvae” (i.e., larvae that have an eye spot and a foot, which indicate readiness to set on a growing surface) produced by U.S. west-coast hatcheries were triploids (Nell, 2002). Interest in mariculture of the nonnative oyster C. ariakensis in Chesapeake Bay led to considerable research for improving techniques to reduce the percentage of reversion of triploid oysters toward diploidy, as well as screening procedures to reduce the risk of inadvertent introduction of reproductive C. ariakensis (Allen and Burreson, 2002). As discussed in an earlier report on nonnative oysters (National Research Council, 2004), there is no federal statute establishing criteria for deliberate, nonnative marine species introductions. States have the authority to set criteria for introductions, but “the existing regulatory and institutional frame-

work is not adequate for monitoring or overseeing the interjurisdictional aspects” of nonnative marine species introductions (National Research Council, 2004). As mentioned above, Virginia and Maryland have decided to forego the use of C. ariakensis in mariculture to focus instead on the restoration of the native C. virginica oyster.

Intentional introductions of nonnative molluscs have also resulted in the unintentional transfer of a wide variety of other nonnative plants and animals that “hitchhiked” with the introduced shellfish (Carlton, 1992a, b). For instance, it has been estimated that about 20% of the nonnative species found in San Francisco Bay are the result of shipments of Eastern (C. virginica) and Pacific (C. gigas) oysters, particularly during the early 19th century. Some of these species have become important predators and competitors of the resident fauna and flora, as well as pests in mariculture operations.

In recent years, tighter controls have been invoked for the importation and transfer of nonnative shellfish species around the world. A Code of Practice for the introduction of nonnative species, developed by the International Council for the Exploration of the Seas (ICES), has been adopted in many countries (Sinderman et al., 1992). The Code requires that the species being considered for introduction be studied in its native habitat for known pests, predators, and diseases, as well as for its biological characteristics, such as genetic makeup. Only broodstock of the nonnative species may be brought into the recipient country and only into quarantine facilities for breeding so that only first-generation offspring can be released into open waters after testing to ensure that no diseases or pests are present.

In response to the past decade’s rich scientific literature on the negative impacts of nonnative “hitchhikers” on shellfish production and on the altering of structure and function of the native populations and communities (see review of Ruesink et al. [2005] for the Pacific oyster), bivalve mariculture industry practices have been adopted to reduce the potential spread of nonnative species. For instance, the use of hatchery-reared seed on the U.S. west coast, coupled with the application of the ICES protocols, can greatly reduce the risk of co-introductions.

When nonnative oysters were brought to the U.S. west coast in the early 20th century, regulatory agencies and the shellfish industry were not fully aware of the threat posed by diseases carried by the imported bivalve mollusc stock (Sinderman, 1984). The resulting introductions of exotic diseases created a number of persistent problems that still have not been solved (e.g., Andrews and Frierman, 1974; Naylor et al., 2001; Bower, 2006).

Current industry practices are designed to prevent the spread of molluscan diseases associated with the use of nonnative species (e.g., Elston, 2004; Office Internationale des Epizooties, 2006). The culture of native species is frequently recommended as an alternative to reduce or avoid harmful interactions among cultured nonnatives and wild species (e.g., Naylor et al., 2001). This strategy does not, however, preclude epidemiological and genetic impacts on the resident populations of conspecifics. Diseases naturally present at low densities in wild populations can achieve epidemic status in culture (e.g., Multinucleated Spore X [MSX] disease in oysters [C. virginica; Ewart and Ford, 1993], Quahog Parasite X [QPX] disease in northern quahogs [M. mercenaria; Lyons et al., 2007]).

The importance of disease management and prevention is well recognized in the mariculture community (see Office Internationale des Epizooties, 2006). Typically regional and or national guidelines and policies exist to reduce the potential introduction or transfer of a disease agent or parasite to a new location. In addition to these policies, the World Organization for Animal Health via representatives from member countries develops health management plans, policies, and diagnostic methods for known (and also novel) disease agents.

Numerous examples of disease transfer via movement of infected stocks have been documented (Harvell et al., 1999; Burreson et al., 2000; Naylor et al., 2001). In the majority of these cases, the fact that the translocated animals harbored a disease agent was unknown, as a consequence of either a lack of basic knowledge of the diseases themselves or inadequate testing and monitoring before translocation. Approaches to successful health management of any species, wild or cultured, is predicated on prior knowledge of typical symbiotic, commensal, and pathogenic organisms associated with that species. The consensus among human and animal health experts is that such baseline health data are lacking for most species impacted by a disease (Haaker et al., 1992; Harvell et al., 1999). Without this information, it is difficult to predict potential health problems, such as disease outbreaks, or to determine the source of emerging epidemic infections. For instance, no baseline data were available for abalone (a gastropod not a bivalve, yet relevant to this problem) in California prior to the outbreak of withering syndrome in 1985 (Haaker et al., 1992). Because of the complexity of the host–parasite relationships and the variability among abalone species, it was difficult to establish which among several newly observed parasites was the causative agent of the withering syndrome outbreak (Haaker et al., 1992; VanBlaricom et al., 1993; Friedman et al., 1993; 1997; 2000; 2007; Gardner et al., 1995; Moore et al., 2000; 2001). In addition, the identification and understanding of new or emerging diseases is dependent on baseline data; emergent disease has been frequently associated with both climatic change and anthropogenic

activities, including animal movements associated with molluscan mariculture (Friedman and Perkins, 1994; Harvell et al., 1999; Burreson et al., 2000; Daszak et al., 2001; Naylor et al., 2001).

Clearly, adequate baseline information on the presence or absence of particular pathogens is crucial to the management of both wild and cultured stocks because it allows us to identify both potentially problematic pathogens and the locales in which they occur. When Pacific oysters were brought to the U.S. east coast after successful culture on the west coast, the oysters failed to thrive, but a parasite that infects Pacific oysters, Haplosporidium nelsoni (MSX), came with them and became established in the native Eastern oysters, C. virginica (Burreson et al., 2000). Because MSX appears to cause little disease and mortality in adult Pacific oysters, it had not been detected in the Pacific oyster. However, the same parasite causes a fatal disease in the Eastern oyster that has contributed to the population decline in many areas of the east coast, such as Chesapeake Bay and Delaware Bay (Andrews, 1976; Ford and Haskin, 1982; Friedman et al., 1991; Ford, 1992; Friedman, 1996; Burreson et al., 2000). In Australia, a recently observed (December 2005 to present) herpes-like virus has caused severe losses of wild abalones, and a lack of baseline health information has made it impossible to determine whether the pathogen emerged from native stocks or was introduced (Hooper et al., 2007; Carolyn Friedman, personal observation). Similar deficiencies in background information have been observed in many marine species (Harvell et al., 1999).

In the aquatic environment, invertebrate hosts and pathogens are subject to many abiotic stressors, such as thermal shifts related to climate (Harvell et al., 1999; 2002; Daszak et al., 2001). A thermal shift as small as 1^oC can alter the dynamics of a disease from causing minor infections and little disease to population-wide epidemics (Harvell et al., 2002; Burge et al., 2006; 2007; Travers et al., 2008a). For example, significant alterations in host–parasite dynamics have been observed in recent years associated with climatic changes and small thermal increases in several species of marine gastropods (abalones: Haliotis spp.) with bacterial pathogens, such as Vibrio spp. (Travers et al., 2008b), rickettsia-like organisms (e.g., Moore et al., 2000), and viruses (Burge et al., 2006; 2007). Alternatively, thermal increases may reduce the pathogenicity and associated disease load if the ambient temperature is beyond the tolerable range of the parasite (Lafferty, 1997).

Mariculture can have both positive and negative effects upon populations of large marine vertebrates, such as birds, marine mammals, and marine turtles. Almost all research on these interactions has focused on

finfish farming and the economic and ecological impacts that result from depredation (e.g., Nash et al., 2000). Of the limited research on the impact of bivalve mariculture on wildlife populations, most relate to bird populations. Three published studies explore impacts on marine mammals (Markowitz et al., 2004; Watson-Capps and Mann, 2005; Becker et al., 2009), but they were not designed specifically to detect ecological impacts on these species, and only Becker et al. (2009) relate to bivalve mariculture methods currently used within the United States. No published studies on potential interactions with marine or estuarine turtles were identified.

Drawing upon a broader understanding of the ecology of these species, potential impacts of bivalve mariculture upon these wildlife populations have been identified in one published review (Kemper et al., 2003). In addition, a National Oceanic and Atmospheric Administration workshop (Moore and Wieting, 1999) explored broader interactions between mariculture and marine mammals and marine turtles, and a discussion paper on the potential effects of mussel farming on marine mammals and seabirds was produced by the New Zealand Department of Conservation (Lloyd, 2003). The potential impacts identified from these sources are summarized in Table 3.1. It should be noted that direct demonstrations of these impacts are rare, and in most cases, potential effects are therefore predicted from the best existing information.

Entanglement in fishing gear and marine debris is a major cause of mortality for seabirds, marine mammals, and marine and estuarine turtles (Lewison et al., 2004; Read et al., 2006). Entanglement in mariculture gear appears to be rare, but two Bryde’s whales have reportedly died in separate incidents after entanglement in mussel spat collection ropes in New Zealand (Lloyd, 2003), and marine turtles have also been entangled in ropes (Godley et al., 1998; Kemper et al., 2003). The introduction of any lines or netting, for example to exclude predators, may therefore pose a risk of entanglement to birds, marine mammals, and marine and estuarine turtles. Where bivalve mariculture operations expand into offshore areas, this may increase the likelihood of interactions with large whales and sea turtles, which are protected in the United States under the Endangered Species Act. Based on experience with other types of mariculture and fishing operations, the risks of entanglement can be reduced by using heavier lines and ensuring that lines and anti-predator nets are kept taut.

Ingestion of marine litter is also known to cause mortality in birds, marine mammals, and marine turtles (Derraik, 2002). Mariculture operations are recognized as a major source of marine litter (Johnson, 2008). For example, young Australian gannets in New Zealand’s Marlborough Sound have been found entangled in rope ties from mussel farms that have been incorporated into their nests (Lloyd, 2003).

Mariculture activity may also influence prey availability for birds,

marine mammals, and marine and estuarine turtles in several ways. As discussed earlier, the presence of culture bags will alter the structure of benthic communities and the extent of eelgrass beds, indirectly affecting prey availability. Also, Manila clam cultivation in bags negatively affects the use of favored foraging areas by oystercatchers (Godet et al., 2009). Alternatively, farm structures may increase food availability by providing a substrate for biofouling organisms suitable as prey, such as mussels. In British Columbia, for example, densities of Surf Scoter (Melanitta perspicillata) and Barrow’s Goldeneye (Bucephala islandica) were positively associated with the presence of oyster farms (Žydelis et al., 2009). Because these species do not feed upon oysters, this association appeared to be driven by the high densities of mussels recorded on mariculture structures (Kirk et al., 2007). In this case, seaduck predation on wild mussels was not perceived as negative by shellfish farmers (Žydelis et al., 2009). However, experiments on natural mussel beds have shown that predation by eiders can reduce mussel biomass by 50% (Hamilton, 2000), demonstrating the impact that seaducks can have upon commercial mussel farms in some areas.

In intertidal areas, the presence of culture bags may directly exclude shorebirds that probe in the sediment from foraging habitat (Kelly et al., 1996). Würsig and Gailey (2002) also highlight the need to consider potential loss of feeding and breeding habitat for cetaceans due to the physical presence of bivalve mollusc farms, particularly given predicted increases in these facilities in inshore environments. Subsequent studies in Western Australia and New Zealand (Markowitz et al., 2004; Watson-Capps and Mann, 2005) indicate that bottlenose dolphins (Tursiops spp.) and dusky

dolphins (Lagenorhynchus obscurus) avoid farmed areas where oyster and mussel cultivation hanging lines are present.

Displacement from key areas may also result from disturbances attributable to the activities of mariculture workers (Becker et al., 2009). This disturbance may be caused directly by the presence of workers on intertidal areas or by boats associated with mariculture activity. In addition, marine mammals may respond to noise from mariculture-related boat traffic.

Mariculture structures can provide shelter, roost, or haul-out sites for birds and seals. This is unlikely to have negative effects on bird or seal populations, but it may increase the likelihood that these species cause fecal contamination of mollusc beds. It has also been noted that the presence of mariculture structures could attract juvenile marine turtles, which usually aggregate under patches of floating weed, thereby disrupting natural dispersal behavior (National Oceanic and Atmospheric Administration, 1999).

Information on the potential effects of mariculture outlined above is largely based upon a general understanding of wildlife ecology and the relationships of these species to the physical and biological environment rather than based upon directed studies built around mariculture operations. Even where studies have been carried out around shellfish farms, uncertainty over spatial and temporal variation in both the location of structures (Watson-Capps and Mann, 2005) and levels of disturbance (Becker et al., 2009) constrain the conclusions that can be drawn about the impacts of mariculture. However, there is less uncertainty about the general effects of “disturbance.” The tending of any mariculture operation requires a human presence, and many studies have used avoidance distances to establish buffer zones to minimize disturbance from other human activities (e.g., Rodgers and Smith, 1997; Blumstein et al., 2003). However, it is important to recognize that some species may not show marked avoidance if they lack suitable alternative habitat, even where the fitness costs are high, and disturbance costs may therefore be underestimated or unrecognized (Gill et al., 2001). Consequently, assessing whether disturbance has a population consequence, estimated as increased mortality or decreased fecundity, is a much more difficult proposition (Stillman et al., 2007). In addition, limited understanding of the foraging distribution of birds, marine mammals, and marine turtles from spatially localized breeding colonies also makes it extremely challenging to assess population-level impacts of disturbance, entanglement, or habitat loss resulting from bivalve mariculture. Thus, if there is increased mortality around a culture site, there could be consequences for breeding colonies hundreds of kilometers away.

Bivalve molluscs are cultured in the marine or estuarine environment, which exposes them to competitors associated with biofouling (i.e., undesirable sedentary organisms that settle on shells and mariculture structures, such as racks, stakes, lines, and bags) and to mobile competitors and predators, including other invertebrates, finfish, birds, and marine mammals. Growers have responded by developing control and management practices, including placing the bivalve molluscs on or under protective structures (i.e., racks, cages, bags, or under netting), physical removal of pests and predators, chemical control, and in some cases biological control.

Biofouling is a common and potentially increasing problem for growers. Epifaunal mussels and oysters are especially vulnerable because their shells and culture structures provide hard substrate for settlement of fouling organisms, and such hard surfaces are often rare in soft-bottom estuarine and coastal systems. The fouling organisms, mostly filter feeders, reduce water flow and can compete with the cultured animals for food (Michael and Chew, 1976; Claereboudt et al., 1994; Taylor et al., 1997), although the magnitude of the effects, if any, will depend on location and species (Arakawa, 1990; Lesser et al., 1992; Ross et al., 2002; LeBlanc et al., 2003; Mallet et al., 2009). Several fouling organisms are nonnative species that came as hitchhikers with the introduction of the cultured bivalves (reviewed by McKindsey et al., 2007). Although current international protocols, typically enforced at the state level in the United States, have reduced unintentional species introductions associated with culture of nonnative bivalve molluscs, fouled hulls and ballast water releases associated with global trade and marine transport have resulted in more introductions of nonnative fouling organisms, including various species of algae and tunicates (e.g., the algae Sargassum muticum, Undaria pinnatifida, and Codium fragile and the tunicates Didemnum spp. and Ciona intestinalis).

Shellfish culture on the seafloor (e.g., oysters, mussels) or suspended off the bottom (e.g., oysters, mussels, scallops) adds substrate area for the colonization of a variety of native and nonnative fouling species or epibionts (e.g., barnacles, tunicates, sponges, bryozoans, macroalgae). In many benthic habitats, the hard substrate surface area provided by bivalve shells on the seafloor may be equal to or greater than the amount of natural inert hard substrate (Railkin, 2004), and it is well recognized that adding more structure to benthic habitats results in an increase in the overall biodiversity to those habitats (e.g., Dumbauld et al., 2001;

Peterson et al., 2003). In addition, off-bottom mariculture activities typically employ a variety of gear types that have the potential for greatly enhancing the abundance and diversity of species through greater provision of additional substrata for the colonization of fouling species. These include ropes and netting used in mussel culture; racks, trays, and bags used in oyster culture; and nets used in scallop culture.

While the addition of structure may increase overall local biodiversity in a system, as compared to unstructured habitats, there is evidence that the biofouling community structure can differ greatly from that on natural hard substrates (e.g., Karlson, 1978; Anderson and Underwood, 1994; Glasby et al., 2007). In addition, there is some evidence that artificial substrates may disproportionally favor the colonization of nonnative fouling species by increasing local sources of propagules of these species (Tyrrell and Byers, 2007). In some cases, the proliferation of nonnative biofoulers has resulted in reductions in local biodiversity (e.g., Blum et al., 2007), which have the potential to facilitate further invasions (Stachowicz et al., 2002) and to lead to potential alterations in population and community structure in coastal food webs (Byrnes et al., 2007).

While some studies have shown that cultured mollusc growth is unaffected (e.g., Lesser et al., 1992; Lopez et al., 2000) or even enhanced (Ross et al., 2002) by fouling, most studies have found that fouling results in reduced mollusc growth and survival and in increased costs to the industry (Watson et al., 2009). In one especially dire circumstance, the invasive tunicate Ciona intestinalis threatens 77% of Canadian mussel farms; at Prince Edward Island, some mariculturists may lose their livelihoods (Edwards and Leung, 2009). Because biofouling by both native and nonnative species increases production costs for the industry, several practices have been developed and implemented to reduce or control it. The general trend is to use techniques that reduce labor costs, ensure product quality, and minimize potential environmental impacts. Techniques include mechanical, chemical, and biological control methods with mechanical and chemical techniques being the most common methods used to remove fouling species from cultured bivalve molluscs and mariculture gear (Watson et al., 2009) (Box 3.1). However, specific applications and their effectiveness typically depend upon the species being cultured, the nature and degree of the biofouling community, and the local environmental conditions. For instance, one-minute exposures to vinegar are 100% effective in mitigating C. intestinalis biofouling (Carver et al., 2003).

Growers use various methods to control biofouling, most often based on physical removal or inhibition by turning over nets and bags (Mallet et

al., 2009) but sometimes by using antifouling agents and other chemical treatments (e.g., acetic acid brine) that are typically applied as dips and followed by brief aerial exposure of the affected organisms or structures (Shearer and MacKenzie, 1961; Huguenin and Huguenin, 1982; Carver et al., 2003; Forrest et al., 2007; LeBlanc et al., 2007; Locke et al., 2009). Some growers have experimented with biological control agents, such as crabs, littorinid snails, and even fish, but this method does not appear to have been widely adopted (Hidu et al., 1981; Enright et al., 1983; 1993; Cigarria et al., 1998). Physical removal of fouling organisms has the potential effect of spreading marine invasive species and increasing the bottom deposition of organic material when conducted over water. With proper disposal techniques, both physical and chemical treatments conducted offsite or in separate holding areas would have little additional environmental effects, but this is economically feasible only on the small scale. Impacts of direct application of chemical control agents in the field at larger scales have not been examined (see Shumway et al. [1988] for details on the use of calcium oxide).

Predation on commercially raised bivalve molluscs, particularly small juveniles planted directly in marine or estuarine growing areas, continues to constrain mariculture in many areas. Predators range in size from diminutive flatworms to birds and mammals (Woelke, 1956; Jory et al., 1984). Some predators, such as the Japanese oyster drill (Ocenebrellis inornatus), were introduced along with the nonnative bivalves and have remained problematic for both cultured and non-cultured species (Chapman and Banner, 1949; Buhle and Ruesink, 2009). Birds are recognized predators and are often more abundant in areas with mussel culture than nearby controls (Caldow et al., 2004; Roycroft et al., 2004), yet the direct effect of bivalve mariculture operations on their behavior varies by species.

Where depredation of the cultured species is a problem, farmers use a wide range of both passive (Dionne et al., 2006) and active deterrents (Ross et al., 2001; Thompson and Gillis, 2001) to reduce losses (Table 3.2). These practices can in turn influence the distribution patterns and behavior of the species preying upon molluscs in their farms or upon other species coexisting in the area. If the use of anti-predator netting leads to entanglement or if shooting is used to reduce predation, these interactions may also result in a reduction in the abundance of affected predator populations. These interactions can raise both ethical and legal issues, particularly where migrating wildfowl or shorebirds are protected under international treaties. In the United States, turtles and some marine mammals are protected

by the Endangered Species Act, and the Marine Mammal Protection Act prohibits the intentional killing or harassment of all marine mammals.

Shellfish growers have responded to predation threats primarily by providing physical protection measures like raising the bivalves off the bottom to protect them from crawling benthic predators or growing the bivalves in protective bags, under netting, surrounded by fences, in tubes, or by adding gravel and shell fragments to the substrate (Castagna and Kraeuter, 1977; Kraeuter and Castagna, 1985; Beattie, 1992; Thompson, 1995). Protective structures modify water flow; affect sediment deposition; provide attachment sites for fouling organisms; and some structures, such as racks, create shaded spots that inhibit the growth of seagrasses (Everett et al., 1995; Rumrill and Poulton, 2004; Tallis et al., 2009). Clam mariculture conducted in bags has been shown to affect sediment but not water column characteristics. Macroalgae and bryozoans attached to bags were shown to attract mobile invertebrates and fish (Powers et al., 2007). Predator netting can result in slightly enhanced sediment organic content but has little consistent effect on sediment grain size or presence of indigenous bivalves (Munroe and McKinley, 2007; Whiteley and Bendell-Young, 2007). Adding gravel and shell to the substrate in Puget Sound, Washington, appears to have site-specific effects on the benthic community, with a general trend of enhanced gammarid amphipod and nemertean abundance and reduced abundance of glycerid, sabellid, and nereid polychaetes (Simenstad and Fresh, 1995; Thompson, 1995).

Though mussels are sometimes grown under protective covering, they are still highly vulnerable, and thus both visual and acoustic deterrents to disturb birds that prey on mussels have been investigated (Ross et al., 2001) (see Table 3.2). These practices would seem to have little direct environmental impact but could change local predator–prey relationships.

In some cases, predators may be trapped, removed by hand, or mechanically removed. For example, starfish have been removed by towing mops, cotton bundles tied to a metal frame, across the bottom (MacKenzie, 1970). Chemical means of controlling predators on bivalve molluscs were extensively investigated in the 1960s (Loosanoff et al., 1960) and applied on small scales for oyster drills and sea stars (Glude, 1957; Huguenin, 1977; Shumway et al., 1988), but chemicals have been rarely used on large estuary-wide scales.

One exception has been the use of the pesticide carbaryl to control burrowing shrimp on oyster beds in Washington State (Feldman et al., 2000). The shrimp are not direct predators but strong bioturbators, which indirectly cause mortality by burying and smothering the oysters under sediment. Because this practice of poisoning has raised persistent concerns about effects on the resident ecological community, it has been studied reasonably well. Long-term changes in the structure of the com-

munity are driven by the removal of one ecosystem engineer (the shrimp) and replacement with another (oysters and even eelgrass; Dumbauld et al., 2001; Dumbauld and Wyllie-Echeverria, 2003). Bioturbation by shrimp oxygenates sediments, thereby accelerating degradation of organic matter and nutrient cycling (Dewitt et al., 2004; D’Andrea and DeWitt, 2009). Abundances of the commensal bivalve, Cryptomya californica, crashed after experimental ghost shrimp removal in a southern California lagoon, whereas recruitment of another bivalve (Sanguinolaria nuttalli) was dramatically enhanced (Peterson, 1977; 1984). The scale of ghost shrimp removal programs is small relative to the size of most estuaries where carbaryl is used. For example, <1% of the intertidal in Willapa Bay is treated annually, and the shrimp are abundant in untreated areas (at least 20% of the intertidal area in Willapa Bay; Dumbauld et al., 2008).

Ecological uncertainties associated with managing the environmental consequences of mariculture will depend on the species cultured, the characteristics of the resident ecosystem, and the scale of the culture operation. This section summarizes some of the areas in which additional research would help to address key questions about the ecological effects of molluscan mariculture to improve best management practices.

The impact of a small mariculture operation (possibly defined by stocking density) on the ecological community in a large, well-flushed system will probably be undetectable relative to the natural “noise” of the system. With an increase in stocking density relative to the supply of food or other resources, the ecological effects could become measurable in at least three aspects. First, there could be direct competition for resources, especially food and space, between the farmed species and the other residents of the system. Second, the biodeposition of organic materials could induce local oxygen depletion and mortality of natural bottom invertebrates where shellfish loading is high and physical flushing low. Third, the cultured suspension-feeding bivalves could conceivably function as predators on the eggs and dispersing larvae of resident species. Knowledge of these effects is critical for evaluating system carrying capacity and addressing concerns about potential impacts on biodiversity.

Not much is known about the factors that cause seagrasses to alter their reproductive strategy (seed or spore production versus asexual expansion via rhizomes and blade growth); how plants respond to disturbance from bivalve mariculture operations relative to natural disturbances; and how response to disturbance varies by season (plant density varies naturally across seasons), location, environment, and species.

The use of nonnative species in bivalve mariculture is likely to persist in areas, such as the Pacific Northwest, where there is a long history of culturing nonnatives, such as the Pacific oyster and the Manila clam. In some cases, these nonnatives have become naturalized—reproductive populations have become established in ecosystems well removed from the immediate vicinity of the shellfish farms. Even in areas where the cultured species has not established a self-replicating population, there is still the possibility that the cultured nonnative bivalve may become naturalized. The presence of nonnative molluscs may suppress the recovery of native species. For example, Trimble et al. (2009) show conclusively that competent larvae of the native oyster O. lurida are lured into settling in unfavorable environments by the presence of shells of the nonnative C. gigas. This contributes to the lack of recovery of O. lurida populations even though remnant populations in some estuaries and lagoons reproduce annually. There are also risks associated with nonnative molluscs as vectors of invasion for hitchhiking species and disease agents that may affect economically important resident species, as well as having potential impacts on population-, community-, and ecosystem-level structure and function. The implementation of current nonnative bivalve transfer practices, such as the ICES Code of Practice, has greatly reduced the potential introduction of nonnative hitchhiking species. However, there are still concerns about the importation of pathogens and other organisms that may not be detected by normal screening procedures.

Infectious diseases can be key drivers shaping local community structure and biodiversity. Despite this, parasites and pathogens are commonly overlooked or underappreciated elements of the ecology and biodiversity of many systems. Although the general roles of infectious diseases in population regulation are recognized, the roles of specific disease agents are often disregarded or have not been well studied (see review by Thomas et al. [2008]). Characteristics of the host, pathogen, and environment shape the ecology of infectious diseases and may cause dramatic fluctuations in populations. Although parasites and disease agents are natural components of ecosystems, their expression may be magnified or altered in an environment where animals are in high density. Such potential for changing impacts of parasites and diseases can be easily monitored in a bivalve mariculture setting. High densities favor parasite transmission via higher levels of parasite release and/or greater contact between infected and uninfected organisms (e.g., Stiven, 1964; Anderson and May, 1981). Many examples exist in which the introduction or transfer of marine molluscs has resulted in the inadvertent introduction of a pathogen (e.g., Elston et al., 1986; Burreson et al., 2000; Naylor et al., 2001; Friedman and Finley, 2003; Wetchateng, 2008). Should a parasite be introduced into a new environment with new potential hosts, one cannot predict the outcome of such encounters (Lafferty et al., 2004). In addition, with global climate changes, current host–parasite relationships that appear to be in equilibrium may shift in or out of favor for the parasite and result in epidemics or improved health in the host population(s).

Information on the potential effects of mariculture outlined above is largely based upon a general understanding of wildlife ecology and the relationships of these species to the physical and biological environment rather than directed studies built around mariculture operations. In addition, limited understanding of the foraging distribution of birds, marine mammals, and marine turtles from spatially localized breeding colonies makes it extremely challenging to assess population-level impacts of disturbance, entanglement, or habitat loss resulting from bivalve mariculture.