Apart from standard management methods described earlier which involve enhancement of stocks by reduction of fishing intensity, two main strategies of stock enhancement for sessile invertebrates are: (a) transplanting juveniles or adults from areas of recurrently successful settlement; and (b) seeding with spat previously reared in the laboratory or caught in collectors placed in the natural habitat.
Hatcheries for shellfish such as clams, scallops, oysters, and lobsters have operated in North America, Japan and elsewhere, starting as early as the late 1800s (see e.g. Rice, Valliere and Caporelli, 2000) with the intention of providing spat for public or private reseeding of grounds. Most of these establishments, funded by local or national governments, were closed around the mid-twentieth century or earlier, largely due to a general failure to demonstrate any effects of their operation on commercial fishery landings.
Earlier hatchery operations tended to release larvae into the wild fairly soon after hatching, and the general impression, confirmed by more recent studies, was that survival was low. The problem of demonstrating their contribution to the wild stock remains a major issue, but a lack of data has prevented any realistic cost-benefit evaluation. This is not surprising since under the open access conditions that generally prevailed, collecting adequate data to evaluate even the state of the stock was not a simple matter. More recent research on European lobsters, conch and other species described later, indicates that rearing to a later juvenile stage improves survival, but also drives up costs. The use of microwire tags or other means of identification of released juveniles also allows identification of recaptured hatchery progeny that is not easily accomplished with larval release.
Although revival of large-scale hatchery facilities is not a reasonable option, it is worth noting that at the time these operations were carried out, natural populations were generally fairly healthy, so that density-dependent mortality of releases may have been accentuated. The idea of using natural colour variants, especially where stocks are very depleted, might have been worth further consideration to test possible contributions from shore-based rearing facilities. For example, naturally blue lobsters occasionally occur in the wild, and the release of unusually coloured offspring might be easily detected in the wild as a natural tagging experiment - similar colour variants might be used for other invertebrate species. However, with respect to hatchery and release operations, the emphasis nowadays has switched to the use of relatively low cost equipment for local rearing that can be installed at the end of a wharf for example, and mainly involves bivalve resources. The use of variants of the "upweller" technology (see below and Figure 4.2), has shortened considerably the holding period, which is now effectively confined to the larval life in vitro, and has reduced the costs of hatchery operations. This has allowed local entities and organizations to rear shellfish seed with minimal plant or investment. This appropriate-scale technology appears to have made small-scale molluscan shellfish relaying a commercially feasible proposition, and in theory, the methodology might be extended to other invertebrate resources.
Due to degraded habitats and overharvesting, replanting schemes for clams is an operational methodology (Rice, Valliere and Caporelli, 2000). A common procedure has been relaying shellfish from natural populations in contaminated bays, to allow their depuration on clean private leases. Rice, Valliere and Caporelli (2000) describe the history of shellfish management and restoration efforts in Rhode Island, which began in the late 19th century, and these provide a general perspective on the evolution of shellfish enhancement activities.
From the late 1890s up to the Second World War, the Rhode Island Fisheries Commission operated a lobster hatchery in response to a decline in local lobster catches. Eggs were collected from wild broodstock, hatched, and larvae reared to fifth stage juveniles before being released. The project was terminated mainly for cost considerations, but also due to the failure to demonstrate any improvement in lobster catches. Currently there is an effort underway to restore lobsters to artificial reefs using settlement funds provided in compensation for an oil spill in Narragansett Bay in 1989.
From the 1930s to the 1980s, hatcheries were used to produce bivalve spat for public and private culture, but these efforts were not economically viable. The programme of longest duration was for relaying Mercenaria mercenaria, from dense beds in waters closed to shellfishing due to pollution problems. Large-scale operations began in the 1950s, but were terminated in the 1960s when power dredging for shellfish was banned in Narragansett Bay. Since the late 1970s a small-scale programme pays a small fee to hand-diggers who transplant quahogs under supervision, from closed waters into clean, managed areas for harvest after depuration. Since 1997, dredge boats have again been hired to relay shellfish into management areas. A calculation based on maximum sustainable yield (MSY) considerations, restricts annual relays to not more than 10.3 percent of the standing crop. Finally, the Rhode Island Public Benefit Aquaculture Project, a joint educational effort with commercial fisheries involvement, is involving secondary level students in the nursery culture of shellfish (though marina-based upwellers) for seeding of public shellfish beds.
In conclusion, shellfish hatcheries have had an uncertain history as government-run institutions, but still operate locally in response to a growing demand for seed, especially for clam and oyster fisheries.
Experiments with New Zealand littleneck clams by Stewart and Creese (2002) showed that growth was highest but that mortality was also high, when transplanted low in the intertidal zone, and vice versa, high in the intertidal. A best compromise was to transplant to mid-tide level: this gave a high recovery rate of 60-90 percent, and clams tended to remain in the transplanted area. Transplanting would seem therefore feasible, although season of transplant and densities of transplanting need to be tested carefully.
Declines in wild production of soft-shell clams in Maine have led to a revival of interest in enhancing stocks from hatchery-grown seed, but predation is a serious problem (Beal and Krouse, 2002). In one experiment, seed were transplanted into boxes with mesh covers, which showed a 13 percent greater survival than uncovered boxes. Interestingly, survival was independent of density in the unprotected boxes, but inversely density-dependent in protected boxes. The strategy proposed is to transplant juveniles from the hatchery to near or below mid-tide levels, and to cover them with a flexible netting (6.4 mm aperture) raised several cm above the sediment surface, to protect them from predation. This can be removed before seasonal storm conditions ensue. By this time, (which precedes winter in Maine), lengths of 25-30 mm have been reached, and clams can burrow to escape predators. For soft-shell clams, growth to harvest will then take another 2-4 year depending on temperature, but will be considerably faster for other species in less extreme climatic conditions.
An example of the use of hatchery-produced spat in upwellers was the evaluation by Heasman et al. (2002) of two alternative nursery-rearing protocols for hatchery-produced Pecten fumatus larvae. This provides an example of possible pilot scale and experimental approaches using hatchery-reared spat and their grow-on in upwellers. Larvae were initially settled and on-grown on mesh downweller screens in a conventional hatchery. Two experimental protocols were then followed:
Spat were retained on downweller screens until large enough to transfer to a field nursery consisting of stacks of mesh screens located in an upwelling system. Stocking density per unit surface area of screen was critical in determining growth rate of P. fumatus spat in field upwellers. Irrespective of growth-limiting factors such as food, the stocking rate at which maximum growth rate was maintained was approximately 70 percent screen coverage.
An alternative nursery-rearing protocol settled spat on mesh screens using cheap nylon curtain material and retained these in the hatchery for 1-5 weeks post-settlement, before the mesh was removed from each screen, cut into sections, and placed in spat collector bags filled with coarse plastic netting before being deployed in the field for grow-on.
Subsequent survival after 30 days depended on spat size at deployment and on handling methods. Some 25-30 percent of 500-750 mum spat at 2-3 weeks post-settlement, were recovered at a size of >5 mm, suitable for transfer to grow-out facilities and comparable with that from tiered upweller nurseries. Screen to collector bag transfer required less capital and was less labour-intensive than tiered upweller systems.
The operations just described may best be described as aquaculture operations, but also form useful procedures before field seeding during enhancement.
The use of hatchery recruits to enhance a stock will need to take into account the existence of natural bottlenecks in the habitat which could prevent the enhanced recruits contributing effectively to the population (Figure 4.1). This requires knowledge of constraints that operate in the wild, and the failure of many enhancement procedures appears to stem from inadequate knowledge of these natural factors. The genetic makeup of the seed used for enhancement should reflect the range of genotypes in the local population, since although it may be advantageous over the short term to add faster-growing strains, this risks creating an "enhanced" stock that is not well adapted to the environmental changes that will certainly occur, which the local stock through behaviour or hardiness may have adapted to overcome.
Figure 4.1 A listing of some of the key problems to be faced in a shellfish enhancement programme (after Tanaka, Seikai and Furuta, 1998).
The ideal size for stocking, as well as the likely size-dependence of predation mortality, may be tested by tethering animals of different sizes in the grow-out environment. This approach revealed a very high mortality for stocked conch (Strombus gigas) compared with juveniles from the native stock. This suggests that predation will be a difficult and costly obstacle to overcome (Ray, Stoner and OConnell, 1994), since it requires cultivating conch to a minimum release size of 75-90 mm, which is a costly operation. Predation may also occur through the effects of a wide range of micro-predators on small settling shellfish (see e.g. Ray-Culp, Davis and Stoner, 1997). These include crabs, shrimps, lobsters and a variety of polychaete worms, and emphasize the importance of releasing juveniles at about a centimeter or more in size.
The re-stocking or enhancement of Homarid populations has been the focus of several research projects in the last decades. However, despite being able to successfully rear lobsters in captivity, there are few reports of newly settled European lobsters (Homarus gammarus) in the wild (Linnane et al., 2001). One of the few well-documented enhancement experiments for crustaceans was conducted by Bannister, Addison and Lovewell (1994) on the European lobster, Homarus gammarus, using over 50 000 hatchery reared juveniles released at an age of three months; released animals being identified by microwire tags and conventional T-bar tags. Lobsters showed a considerable "site fidelity", remaining within six km of release sites. They survived in the wild up to six years, to be caught at legal size of 85 mm 4-5 year after release in the commercial fishery, and some animals already were egg bearing on recapture. An independent experiment with T-bar tagged animals allowed the fishery exploitation rate to be determined independently. Using this information, estimates of survival from release to recapture averaged between 50 and 84 percent, depending on assumptions made about tagging mortality and tag shedding during the experiments. This experiment seems to suggest that hatchery operations may be useful, at least for stock restoration of lobsters in depleted areas, if the problems of conserving the wild genotype can be avoided through use of brood stock from local populations. Bannister and Addison (1998) describe the associated lobster research programme on enhancement from the technical perspective of ensuring survival to commercial size of up to 50 percent of microwire wire tagged lobsters released. They concluded however that stock enhancement of this slow-growing species is unlikely to be economically worthwhile when the cost of running a hatchery is taken into consideration. They note however that a programme aimed at enhancing the natural breeding stock, or for tourism purposes associated with public visits to the lobster hatchery, may still be viable. Bannister and Addison (1998) emphasise that experimental testing is needed of whether hatchery-reared juveniles supplement or replace naturally settled shellfish before assessing the biological and economic benefits of enhancement programmes. Again, experiments must be designed to answer ecological questions and to make predictions about how stocking density of natural stock will affect survival and recapture rate. This is of serious concern for managers when trying to ascertain if restocking or enhancement programmes are going to be successful at the fishery level (Linnane et al., 2001).
Site-fidelity following colonization is an important issue where private or local resource users are considering such an operation. Site fidelity was observed by Jensen et al. (1994) after colonization by European lobsters of an artificial reef erected on a flat sandy bottom some three km from the nearest natural reef. Colonization occurred after several weeks, and 48 percent of the 114 tagged individuals were recaptured at least once on the reef. "Vagrants" tagged on the reef were captured elsewhere, but generally less than 16 km away from the reef.
An experiment described by Shiota and Kitada (1992) involved the release of tagged individuals of a more mobile swimming crab, Portunus trituberculatus, in different seasons in shallow coastal waters of Japan, where the species is confined within the 30 m isobath. This gave recoveries of between 11-46 percent; 90 percent of which were recaptured less than 20 km from the release site, despite a spawning migration to shallow water from a hibernation area at 20-30 m depth. Again, this shows that restocking, even of mobile crustaceans, may be a practical proposition, although the economic effectiveness of this operation was not evaluated. A better documented experiment for a tagged inshore flatfish, hatchery-reared flounder can be mentioned in this connection, in the same waters, which was supported by a 2-stage random sampling survey described by Kitada, Kishino and Taga (1993), who found the operation to be profitable, even with a recovery rate as low as 15 percent.
Beal, Mercer and OConghaile (2002) note that success is uncertain when releasing large number of lobster larvae from a hatchery; however, costs are high for grow-on of post-settlement stages to a larger size before release. A strategy they found to be cost-effective, was a field-based nursery system for rearing cultured lobster juveniles in cages deployed on the bottom. Feeding was on the plankton, or by foraging on the fouling community in the cages. Rates of recovery in on-bottom growth cages were site-specific, but of the order of 25-40 percent to 5-7 mm carapace length. Presumably these juveniles might then be used for transplanting to favourable juvenile habitats as described by Bannister and colleagues, but doubts as to the economic viability of the whole operation stem from the slow growth rate of European lobsters.
In the tropics, the stocking of lagoons for harvest by artisanal fisheries has been practiced. Davenport et al. (1999) describe the stocking of a lagoon in Sri Lanka with Penaeus indicus: a species which does not breed in the lagoon, where the outlet of the lagoon is blocked seasonally by a sand bar. Over the winter to spring months the shrimp stock in the lagoon is almost entirely harvested. Previous twice-annual stocking with larvae and post-larvae could therefore be linked to subsequent catches, which corresponded to roughly 3.5 percent of the post-larvae released: catches were enhanced by some 1 400 percent over previous levels. Bioeconomic calculations demonstrated this to be ecologically and economically sustainable.
The possibility of using hatchery-reared queen conch to rehabilitate overfished Florida populations was evaluated by Stoner and Davis (1994) through a 15-month field experiment in the Bahamas, comparing hatchery-reared and wild juveniles of 85-120 mm shell length at two experimental sites; one with a wild population, the other without. Survival of hatchery reared conch after seven months was low (nine percent) compared with wild stock (28 percent), and thin shells, short spines and low burrowing frequency may have increased the vulnerability of hatchery stock to predation, and their growth rate was half that of the wild stock. Survival was somewhat higher where resident populations provided some density-dependent protection for introduced individuals. In an experiment with tethered animals, hatchery conch showed twice the mortality to predation of wild individuals, confirming their higher vulnerability. Despite this, the authors considered this a possible method of rehabilitation, noting that this approach to enhancement requires release of large numbers of high-quality large juveniles into appropriate habitats.
Successful enhancement programmes for conch, as for other species of invertebrate, will have to consider habitat requirements and how they shift during ontogeny. Sandt and Stoner (1993) found juvenile conch (35-54 mm shell length) to remain buried in clean sand during the day. At 1-2 years they moved into seagrass and coral rubble habitats where growth rates were much higher. This movement may be in response to food requirements or predator avoidance, but needs to be taken into account in stock enhancement programmes. A further complication noted by Stoner and Ray (1993), is that juveniles remain aggregated within only a small fraction of the vegetated habitat, in an area where tethered individuals showed lower mortalities. Evidently, the requirements for survival of transplanted individuals vary on a local scale for reasons that are not clear, suggesting that experiments with tethered animals (or for other species, pilot-scale trials with juveniles grow-out in cages), in different habitats, could usefully precede any large scale transplantation.
Biological pollution of habitats containing significant shellfish biomasses has given rise to replanting schemes for clams in a number of localities. Relaying shellfish from populations in contaminated bays allows their natural depuration on clean private leases prior to marketing (Rice, Valliere and Caporelli, 2000). Such operations may allow the local management authority to charge a modest fee to private shellfish diggers transplanting stock - digging often being facilitated by the use of mechanical harvesters. A best compromise was to transplant to mid-tide level: this gave a high recovery rate of 60-90 percent, and clams tended to remain in the transplanted area. Transplanting would seem therefore feasible, although season of transplant and densities of transplanting need to be tested carefully.
Transplantation is intended to maintain or improve depleted or overexploited populations, or even to extend distribution areas to new grounds in order to establish new fisheries. Different approaches to transplanting are (a) relocation of seed or juveniles from dense beds to depleted areas; (b) collection and culture of local and imported seeds from e.g. long-lines or seed collectors, after which they are transplanted in a habitat suitable for species development (see e.g. Kristensen and Hoffmann, 1991); and (c) transplanting subadults or adults to supplement reproduction of natural populations, or in the case of "empty" habitats, in the hope of developing new self-sustaining populations (Peterson, Summerson and Luettich Jr., 1996).
There are many examples of the gathering of juveniles (oysters, mussels, clams) from one area and their transfer for on-growing to another (Quayle and Newkirk, 1989; Brand et al., 1991). For example, juveniles of the New Zealand scallop Pecten novaezelandiae which settled on the outside of collector bags were redistributed to areas where natural settlement had been unsuccessful (Bull, 1994). Alternatively, juveniles can be transplanted on grounds that may be unsuitable for releasing very small spat because of predators, or adverse hydrographic conditions (see Tegner, 1989 and references therein for examples on sea-urchins). Kristensen and Hoffman (1991) transplanted seed of Mytilus edulis dredged from natural beds to 3000 m2 culture plots, in order to evaluate individual growth rates and production within the period of transplantation. The effects of wave action and starfish and shore crab predation were mentioned as factors limiting success of their enhancement operations.
Addition/colonization experiments by transplanting adults from high-density sites to new areas are poorly documented for molluscs. Peterson, Summerson and Luettich Jr. (1996) reported a successful transplantation experiment with adult scallops Argopecten irradians concentricus from an abundant site to four receiver sites where the species had been virtually eliminated by a red tide outbreak. Transplanted sites enhanced local adult densities from 1-3 to 15 scallops m-2 in two years, and local recruitment was up to five times greater than in years when no transplantation had occurred. However, settlement indices, as estimated from spat collectors, did not confirm that the transplants succeeded through the enhancement of larval abundance. Despite this partial success story, local transplantations have usually been conducted on a trial and error basis, without evaluating the implications of such introductions in the colonized area. Active restocking should also be conducted carefully in order to avoid further unexpected ecological damage (Peterson, Summerson and Luettich Jr., 1996).
On larger spatial scales, experimental restocking trials from other sea areas (Brand et al., 1991) or even from one ocean to another, have on occasions had at least temporary success (Larkin, 1991). For instance, a long distance extension of the distributional range of high-valued species has been achieved (e.g. for Japanese oyster Crassostrea gigas), and this species in many cases has replaced native oysters which have been decimated by disease or environmental change (see also Chapter 5 for additional information on invasions and species introductions).
As noted by Hannesson (1986), thinning of a population subject to density-dependent growth influences increases the potential growth rate of the survivors, and in the case of shellfish, small, young individuals may command a separate market price, making this two-stage harvesting a close to optimal strategy. One of the characteristics of harvesting by dredges or towed gear is the inevitable presence of small individuals in the catch, which ideally should be returned to the sea, but may not be since they command a market price. Fishers have ways of reducing the proportion of small individuals in the catch if there is a legislative incentive to do so. A scheme was introduced for the Georges Bank scallop fishery in the 1970s which did not rely on a minimum size limit (which would be difficult to apply where the animal was "shucked" from the shell prior to landing).
Enhancement of wild shellfish stocks with hatchery-raised seed has been seen as a useful and, in some cases, economic way of restoring depleted stocks (Saito, 1984; Schiel, 1993), although on occasions its economic validity has been questioned, especially where large hatcheries are used. Natural production can be augmented through the use of collectors to catch spat in their natural habitat, or by inducing spawning and rearing larvae in the laboratory. In fact, due to the decreasing trend in landings of most important species, the number of spat released from enhancement programmes in some countries is continually increasing. For example, Kitada, Taga and Kishino (1992) documented a considerable increase in the number of spats of scallops and abalones released in Japan between 1983 and 1989 where scallops are one of the most important species used in stock enhancement programmes (ca. 3 231 million fingerlings were released in 1989). Although hatcheries may achieve economies of scale by providing spat to growers over a wide area, there are significant advantages in small local spat production, which conserves local genotypes. A new technological development, the "upweller" is now being used to produce seed or spat for local growers.
Despite a growing demand for entry into small-scale hard clam aquaculture in the USA, this has been limited by the cost of hatchery seed ready to replant at a suitable planting size (usually eight mm SL or larger) and formerly it was necessary to grow-on small seed to this size in a nursery. Ponds or impoundment facilities are other alternatives for grow-on, but often lack access to ready supplies of clean seawater and suffer from seasonal algal blooms. The use of floating trays for cultivation of spat led to an "upwelling system" located in a sheltered impoundment (Bayes, 1981: in Hadley et al., 1999). Nurseries on land give high survival and rapid growth, but require expensive waterfront property and are energy- and labour-intensive to operate. Field-based nurseries are inexpensive, but seed survival is often very low and success is site-specific. Hadley et al. (1999) described a floating upwelling system (FLUPSY) which has the advantages of land-based hatcheries, such as good survival and growth, with the low-cost operation of field-based systems. This approach can avoid high real estate costs by for example, being incorporated into a dock or floating pontoons on a wharf. The system described by Hadley et al. (1999) cost US$ 4 500 to construct, with operating costs below US$ 5 000 annually, suitable for small-scale growers.
Currently, small scale upwelling systems may be powered by a pump, airlift or water wheel, to ensure an upward flow of seawater through the facility, using the natural plankton therein to feed spat in stacked or side-by-side trays in the facility. Hadley et al. (1999) describe a tidally powered unit (Figure 4.2). Using this system, after larval rearing to settlement in a relatively limited facility, animals can be grown to a size where they are more resistant to predation at relatively low cost. Upwellers have proven to be extremely effective as bivalve nursery units and their use is steadily increasing in North America (Appleyard and Dealteris, 2002). Another variant of small-scale spat production is the wave-operated nursery system was developed by Hickman et al. (1999) for growing hatchery-produced spat of the New Zealand dredge oyster Tiostrea lutaria through to 20 mm, suitable for on-growing using conventional oyster farming techniques.
Figure 4.2 Representation of a tidal-powered upwelling system for spat rearing (redrawn from Hadley et al., 1999).
Linnane and Mercer (1998) compared five tagging methods to follow survival of juvenile European lobsters after release through multiple moults, and found that implanted coded microwire tags and elastomer implants were more efficient than rostrum ablation; in fact the rostrum was rapidly replaced after three moults. Survival rate, as usual, was higher for somewhat larger juveniles. Abdominal streamer tags showed good retention and survival, though they occasionally interfered with moulting. Branding resulted in somewhat lower survival. This type of comparative study seems advisable prior to any large scale enhancement programme involving restocking.
Studies with another cryptic species, the gastropod Trochus niloticus, showed that metal tags of folded aluminium foil can be easily picked up by a metal detector even in complex coral habitats for at least three months, and may have a wider application in following survival of other species of hatchery-released shellfish. From parallel experiments described elsewhere, the suggestion is that survival rate is relatively low in most localities, suggesting the desirability of an experimental approach prior to large scale stocking.
Seed collectors are often used as a preliminary stage in transplantation programmes. However, they have also been used to collect early life history stages, which are then either transferred to a controlled environment for faster grow out to a marketable size, or released at the larger size needed to avoid high predation rates (Brand et al., 1991). Kristensen and Hoffmann (1991) described successful transplantation experiments of the blue mussel Mytilus edulis in Denmark. The approach was based on transplantation of seeds collected from long lines and from natural mussel beds to bottom culture plots. Individual growth rates, production and biomass were evaluated and compared with natural mussel beds.
Timing and placement of shellfish collectors has to be varied according to life history characteristics of the cultured species for optimal results. Spatial and temporal patterns in spat distribution (both vertically and horizontally), duration of the spawning season, and sequences of larval stages, are all information required for spatfall forecasting and timing of collector placement, especially for those species with short duration peaks in settlement within the year. For example, settlement of the puerulus stage of the palinurid Jasus edwardsii was determined by crevice collectors at sites along the east coast of the North Island of New Zealand (Booth et al., 1991). Depth of greatest settlement found on collectors varied with locality and time, but was within the upper 12 m. The ability of J. edwardsii to settle over a wide depth range may improve its chances of recruitment. The authors also described a device (closing crevice collector) for measuring puerulus settlement of J. edwardsii at depth. Phillips et al. (2001) tested different collector designs for the rock lobster Panulirus cygnus at different depths and distances offshore; they also examined the effect of collector size, and tested the effect of frequency of servicing the collectors. Five collector designs were set in shallow waters < five m, and were checked over four lunar months during peak settlement. Sandwich collectors had significantly better catch rates than others, and settlement rates were highly correlated with collector dimensions. Daily servicing for seven days around the time of new moon yielded catches 170 percent higher than those from a single monthly servicing.
In some cases collectors failed to reveal enhancement following adult transplantation. In fact, Peterson, Summerson and Luettich Jr. (1996) showed that spat collectors are not a reliable indicator of recruitment enhancement, and this was confirmed by a poor correlation between larval settlement of scallops and subsequent recruitment data on the grounds. This suggests that collectors are not always effective for following natural enhancement of shellfish populations, and results using these techniques are species and site/time-specific. Detailed description of seeding techniques of this type is beyond the scope of this review but interested readers are referred to Quayle and Newkirk (1989) or Shumway (1991).
Shellfish culture in Japan using hatchery juveniles to supplement wild stocks of fish and invertebrates has been implemented on a commercial basis for decades, and a review of programme successes and failures was provided by Masuda and Tsukamoto (1998). Scallop stock enhancement appears particularly successful, with stress on the necessary high quality and viability of seedlings, and habitat improvement. Crop rotational practices may be followed, supporting a steady increase in yield, but for all species, questions of habitat improvement and the preservation of genetic diversity remain priorities.
The effects of scallop culture on total wild plus cultured production from a marine area, has been well documented by the Japanese scallop industry of Hokkaido (Kitada and Fujishima, 1997). The most striking feature is not only the considerable increase in landings following the use of culture techniques, but the reduction in the coefficient of variation of annual landings that has accompanied it. Seed release appears to be economically successful, but the most interesting implication is that a greatly increased level of spawning resulting from the artificially enhanced population has improved overall recruitment in the culture areas, which contrasts interestingly with the earlier mention in this text of the very low minimal %SPR levels for recovery of Placopecten magellanicus.
Three approaches to enhancement of estuarine bay scallop populations were tested by Goldberg, Pereira and Clark (2000) in an area where high densities of natural spat are uncommon: (a) collecting and distributing natural spatfall (which for shellfish may occur locally in very high densities reducing growth rates and promoting increased mortality); (b) introduction and overwintering of hatchery-reared stock to serve as spawning stock the following season; and (c) overwintering of the same in suspension culture, to create mobile spawner sanctuaries. The first option, involving planting of hatchery-reared spat at different times in different densities, showed that predation probably is the major factor influencing survival. Time of planting rather than planting density was a key factor, emphasizing the importance of deciding on an optimal season of release. Scallops released in an eelgrass bed had high overwinter survival and spawned the following season, thus contributing to the enhancement programme in two ways. Overwintering in suspension culture gave a 60-80 percent survival to spawning the following year: the advantage of this approach which mimics that just described for Japanese scallops, is that spawners could be transferred to "source" areas where spawning is believed to ensure optimal survival of seed. This seems to illustrate the contention that for shellfish, there is the potential to use a range of methods in areas where recruitment is poor but environmental conditions are not limiting, but that timing of the enhancement intervention is critical.
Other stock enhancement programmes are based on the production of larvae or post-larvae under controlled conditions, and their subsequent release onto the fishing grounds (Tong, Moss and Illingworth, 1987; Schiel, 1993). As an alternative to adult rearing, which in almost all cases is economically infeasible, one option is the direct placement of early hatchery-raised benthic stages into natural habitats favourable to survival (Schiel, 1993). This by-passes high predation and density-dependent effects (e.g. competition, cannibalism) and avoids the environmentally induced mortality that occurs in the larval phase.
Over recent decades much research has focussed on the rearing of shellfish spat or juveniles for enhance overfished stocks. Large-scale juvenile production units, and reseeding programmes have been carried out, particularly in Japan (Saito, 1984), but with variable success. Tong, Moss and Illingworth (1987) documented an enhancement strategy of a natural population for the abalone Haliotis iris using larvae reared in the laboratory and released after 13 days post-fertilization. Schiel (1993) detailed a comprehensive enhancement programme for the abalone Haliotis iris in which 80 000 hatchery-raised juveniles were placed on rocky boulders at eight sites around Chatham Island, New Zealand. Economic analyses showed that only in three of the eight sites analysed was the internal rate of return positive; high natural mortality rates determined negative returns on the other five sites. Nevertheless, an overall positive financial return was obtained when all sites were combined.
High production costs and high predation on spat released directly into the natural habitat thus appear to be two major bottlenecks to consider when planning such enhancement programmes (Ogawa, 1988; Castilla, 1990; Kristensen and Hoffmann, 1991; Schiel, 1993; Parrish and Polovina, 1994).
Genetic factors are now recognized as playing an important role in any enhancement plan. The application of genetic techniques to invertebrates is in many ways essentially similar to that in finfish (Thorpe, Sole-Cava and Watts, 2000). However, relative differences in the life history of shellfishes lead to particular problems in the use of genetic data to study invertebrate species and the potential for enhancement from a genetic perspective. The main role for genetics is the identification of groups of interbreeding individuals as the basis for a fishery or stock enhancement programme (Tringali and Bert, 1998). In the genetic assessment of invertebrate stocks, the large evolutionary range of invertebrates exploited and their widely different life history attributes, notably the mobility and the relative extent of the dispersive larval phase (see Chapters 2 and 3), poses additional problems when compared with finfish, which deserve special consideration.
At high densities of stocking, enhancement programmes can conflict with conservation considerations, and lead to loss of genetic diversity (e.g. Baltic salmon). Genetic differences between possible broodstocks also need to be taken into account (e.g. Kristensen and Hoffmann, 1991). Strains resistant to diseases or capable of faster growth or meat yield may be selected for, but the irony is that continued resistance to environmental changes is likely to be conferred by maintaining a wide range of genetic components in the broodstock population (Gaffney and Bushek, 1996). Slow-growing shellfish for example may be inefficient competitors for food, but live longer and be more resistant to environmental change. Genetic factors also played a significant role during large-scale transplantations (see Brand et al., 1991 and references therein). The way that transplanted individuals affects wild stocks is not clear, although serious concern has been shown in the case of Baltic salmon that genetic "contamination" from escaped cage-reared fish is reducing the ability of natural populations hybridized with them to find their natural spawning grounds. All of these issues deserve more investigation, but in general, selection to improve commercial characteristics of shellfish in culture will inevitably continue.
Stock enhancement programmes should use information on population structure to optimize enhancement strategies (Shaklee and Bentzen, 1998) in such a way that the genetic diversity and character of existing wild stocks is protected. Genetic methods of stock identification allow tests of reproductive isolation and gene flow between populations using naturally occurring marks, thus avoiding the need for physical tagging. Protein electrophoresis has been widely used, but nuclear and mitochondrial DNA-based methods have less stringent tissue sample requirements, and their higher cost appears destined to decline, and relevant data bases are likely to grow as focus is directed at these more recent methods.
Boulding, Boom and Beckenbach (1993) used empirical parameter estimates from coding regions of mitochondrial DNA to assess genetic-variation in one bottlenecked and two wild populations of the Japanese scallop (Patinopecten yessoensis). The genetic diversity of a population bred in a small experimental hatchery in British Columbia for three generations was compared with its wild source population at Mutsu Bay (Aomori, Japan) and with a second wild population at Uchiura Bay (Hokkaido, Japan). The three populations were similar in the frequency distributions of the 11 mitochondrial clonal lines. This suggests that the experimental hatchery stock was not severely inbred and that gene flow between the two wild Japanese populations has been sufficient to prevent divergence. Genetic analysis of proteins and/or DNA variation have revealed the existence of multiple isolated stocks in what was formerly treated as a single population, but now must be referred to as a metapopulation, with a requirement to conserve the individual population components (Boulding, Boom and Beckenbach, 1993).
Design of enhancement programmes requires an understanding of larval transport in and out of selected areas reserves, and the understanding of whether selected sites will be self-seeding, and whetherrecruits from surrounding areas will intermix with the local stock, and hence what is the rate of exchange in recruits. Direct measurements of mean larval dispersal are needed to understand connectivity between members of a metapopulation, but such measurements are extremely difficult. Genetic patterns of isolation by distance have the potential to add to direct measurement of larval dispersal distance and can help set the appropriate geographic scales on which marine reserve systems will function well (Taylor and Hellberg, 2002; Palumbi, 2003). While low connectivity does not imply limited gene flow, the opposite is true: geographic genetic differentiation is strong evidence for low connectivity (Orensanz et al., 2003). Some studies in invertebrates show that the genetic structure of populations, using polymorphic allozymes, revealed an expected congruence between the larval strategies and spatial differences in allele frequencies. For example, allele frequencies for planktotrophic species lack significant spatial heterogeneity over distances of > 1000 km, while there was significant spatial heterogeneity in allele frequencies over distances of as little as 100 km between populations of lecithotrophic species (Lambert, Todd and Thorpe, 2003). This long-term study (1985-1995) corroborated the spatial studies, showing that the population structure for these species is closely related to their realized larval dispersal. The authors show that the scale of larval dispersal and recruitment could be successfully examined in the field by means of transplant experiments between genetically and/or phenotypically different pairs of populations. The results obtained by the authors for two intertidal nudibranchs, however relevant for shellfish and other invertebrates, suggest that small, local populations are not totally open demographically and receive at least a proportion of their recruits from larvae generated within that population, thus increasing the probability of local extinction (Lambert, Todd and Thorpe, 2003).
Thus, the degree of genetic differentiation between locations can provide important indirect evidence, reflecting the pattern and scale of effective larval dispersal. Genetic studies conducted by Heipel et al. (1998) and Heipel, Bishop and Brand (1999) also showed that a stock of Pecten maximus from a semi-enclosed area (Mulroy Bay, Ireland) differed significantly from open water populations, suggesting that scallop populations in semi-enclosed coastal systems tend to be self-sustaining. The lowest genetic variability was recorded from this enclosed habitat, probably reflecting the relative isolation of Mulroy Bay, whereas dynamic hydrographic conditions in the Irish Sea and the Channel may generally ensure extensive mixing of the planktonic larvae. Lewis and Thorpe (1994) evaluated scallop stock enhancement through transplantation, which could reduce the fitness of receptor local populations. They found highly significant inter-site genetic heterogeneity in a study of twelve populations of Aequipecten opercularis around the British Islands. Preliminary evidences of differentiation though transplant experiments could also be assessed through reproductive ecology studies. Ansell, Dao and Mason (1991) showed that differences in the reproductive ecology of different populations of Pecten maximus indicated relative genetic isolation among stocks (see also Mackie and Ansell, 1993). However, results should be contrasted through genetic studies and could not be supported by e.g. allozyme polymorphisms (Wilding, Latchford and Beaumont, 1998: see Orensanz et al., 2003).
As the potential of enhancement of invertebrate stocks through aquaculture becomes increasingly realized, transplanting and introductions are becoming more common. To predict the genetic consequences of transfers, information on genetic differences between source and recipient populations is critical (Beaumont, 2000). This author highlighted that potential risks and consequences of hybridization should be experimentally assessed before introductions of scallops are carried out, because hybridization is unpredictable and can lead to loss of genetic diversity or breakdown of co-adapted gene complexes.
The possible effects of restocking in diluting, through mixing, a relatively small but locally well-adapted genotype, has been referred to as "genetic contamination". It is now widely recognized that hatchery introductions of genotypes differing from the local population through cross-breeding with the locally adapted stock can negatively affect adaptation of the local race to its particular environment, and we should recognize that hatchery strains selected for fast growth in culture may not be adapted genetically or behaviourally for life in the wild. Tringali and Bert (1998) point out that conservation of a sufficiently large genetically discrete population could be affected by stocking programmes. One hypothetical example could be a project to reintroduce the queen conch, an important commercial mollusc, to island shelves of the Caribbean where populations have declined to low levels. An economically efficient approach would be to use a central hatchery to cultivate conch from one of the remaining abundant populations, and simply to distribute the juvenile conch from a small plane flying over each island shelf. The problem is that isolated conch populations are likely to have adapted to the hydrographic and ecological situation of each individual island shelf, such that the introduced animals would have a limited chance of completing their life history, and thus this enhancement methodology would compromise the possibility of recovery of any remnants of the local native stock.