Constraints to Crassostrea ariakensis aquaculture: season and method of culture strongly influence success of grow-out.

ABSTRACT The proposed introduction of the nonnative Suminoe oyster Crassostrea ariakensis to the east coast of the United States to restore the wild oyster fishery and/or for commercial aquaculture is contingent on perceived benefits outweighing costs. Trials conducted at six sites distributed across North Carolina evaluated the likely biologic success of aquaculture of C. ariakensis by assessing the oyster's growth, mortality and fouling under alternative methods and seasons of deployment. Of the six sites at which oysters were deployed, growth of C. ariakensis was greatest at the high-salinity sites, Hoop Pole Creek (29-32 ppt) and Newport River (25-30 ppt) and lowest at the low-salinity site, Nags Head (4 ppt). Across sites, growth was consistently lower (by 50%) in suspended floats than in fixed racks held 15 cm off the bottom. Only on the muddiest sediments (Newport River), where growth in tissue mass was 50% greater and growth in shell mass 20% greater on raised than bottom racks, did growth differ detectibly with rack elevation. In spite of low mortality and fouling over winter, predation in summer by blue crabs resulted in mortality approaching 50% at one of the sites (Swan Quarter) and heavy settlement by Polydora spp. produced a greater than 25% cover of mud-blisters on 84% to 97% of oysters. Consequently suitability of C. ariakensis for sale on the high-valued half-shell market may be contingent on completion of grow-out before summer Polydora spp. infestation. Results indicate that whereas at high salinity sites grow-out from a 27-mm deployment size is typically achieved within 5 mo, at medium salinity sites >7 mo is required. Failure to remove oysters at medium salinity sites from the water prior to summer may result in heavy Polydora spp. infestations that necessitate sale on the lower-valued shucked market. Irrespective of target market, it is recommended that C. ariakensis be grown in fixed racks as opposed to floating structures to maximize rates of growth without compromising survivorship.

KEY WORDS: aquaculture, fouling, growth, mortality, Crassostrea ariakensis, oyster

INTRODUCTION

Interactions among over-harvest, habitat destruction, disease and declining water quality along the mid-Atlantic coast of the United States have reduced standing stocks of the native oyster, Crassostrea virginica (Gmelin, 1791), to an all time low (MacKenzie 1997, Lenihan & Peterson 1998, Coen & Luckenbach 2000). As a result, oyster landings in Maryland, Virginia and North Carolina have dropped to 1% of levels in the mid 1900s (Rothschild et al. 1994, Frankenberg 1995, Brumbugh et al. 2000, Kirby 2004) and the region's oyster-processing industry and restaurants survive only by importation of oysters to satisfy public demand (Kirkley 1997). Various efforts have been made over the past few decades to restore oyster populations and their ecosystem services (Breitburg et al. 2000, Mann 2000). A lack of any dramatic progress in rebuilding the native oyster fishery has, however, forced an increasing number of watermen to consider aquaculture.

In contrast to the wild oyster fishery that depends upon natural recruitment onto suitable substrate, aquaculture frequently utilizes hatchery-reared spat. Hatchery-production of spat enables selection for traits that maximize growth and survivorship in the local area. Selective breeding efforts have yielded strains of C. virginica with improved survival and resistance against one of the major diseases threatening populations of oysters along the east coast, Haplosporidium nelsoni (Haskin & Ford 1988, Matthiessen et al. 1990, Paynter & DiMichele 1990). Until recently (see Calvo et al. 2003), limited success has, however, been achieved in producing C. virginica that are also resistant to Perkinsus marinus, another major oyster disease along the Atlantic coast of the United States. Consequently, there has been growing interest over the last several decades in the introduction to the east coast of the United States of a nonnative species of oyster that is resistant to P. marinus and H. nelsoni (Mann et al. 1991, Shatkin et al. 1997, National Research Council [NRC] 2003).

The Pacific oyster, C. gigas (Thurnberg 1793) and the Suminoe oyster, C. ariakensis (Fujita 1913), are two species that are used in other parts of the world for aquaculture and suffer little mortality from P. marinus or H. nelsoni. Earlier trials of these oysters in mid-Atlantic waters suggest that C. ariakensis is the more suitable species for introduction (Calvo et al. 1999, Calvo et al. 2001, Grabowski et al. 2004). Although C. gigas consistently outperformed the native oyster and C. ariakensis at high salinities (>30 ppt), C. ariakensis grew the fastest and suffered the least mortality at lower salinities (16-20 ppt), characteristic of the large estuarine systems of Pamlico Sound and the Chesapeake Bay (Grabowski et al. 2004). As a result of the favorable performance of C. ariakensis in initial trials, several mid-Atlantic states are presently considering introduction of this normative oyster (NRC 2003).

According to protocols of the International Council for the Exploration of the Sea (ICES), introductions of nonnative species should proceed only if: (1) their economical, sociological and ecological benefits are perceived to be substantial and (2) benefits are anticipated to grossly out-weigh potential costs. In the case of aquaculture, economical benefits will depend on culture season and method, taste, shelf life and market acceptance. Season profoundly impacts profitability of an aquaculture operation through influences on growth (Bayne et al. 1977, Doering & Oviatt 1986), the prevalence of oyster diseases (Burreson & Ragone Calvo 1996, Burreson et al. 2004) and the activity of predators (Micheli 1997). Elevation of oysters off the bottom can affect growth and mortality by determining the temperature at which the animals are held, their exposure to tides, waves and currents and their overgrowth by fouling organisms such as ascidians and barnacles that compete for resources and may inhibit feeding (e.g., Lesser et al. 1992, Moroney & Walker 1999, Lodeiros et al. 2002, Soletchnik et al. 2005). Additionally, suspension of oysters can reduce settlement of Polydora spp. mud worms (Loosanoff & Engle 1943) that lower consumer value of oysters by boring through the shell and producing mud blisters (Korringa 1951, Korringa 1952, Handley 1997).

Necessary information on the effect of culture season and method on benefits of oyster aquaculture may be obtained without appreciable risk of unplanned introduction of self-sustaining populations through experimental culture of nonreproductive triploids under cautious constraints (NRC 2003). Here we compare the relative rates of growth, mortality, fouling and infestation by Polydora spp. of triploid C. ariakensis when cultured on fixed racks (on and off the bottom) and suspended in floats just below the surface of the water. Any difference in performance of oysters between methods may be related to the distance off the bottom and/or the mobility of the structure holding the oysters. To separate these effects, we compare the performance of oysters across methods when oysters are held at the same tidal height and when the distance between the bottom and the oysters differs between the two methods. By repeating all methods for culturing oysters in winter and in summer we also assess the oysters' performance at different times of the year.

MATERIALS AND METHODS

Deployment of Oysters

We cultured oysters in research sanctuaries at: Buzzards Bay, Nags Head (Nags Head; 35[degrees]59' N75[degrees]40'W); east of Herring Island Shoal, Walterslough (Walterslough; 35[degrees]49.4'N, 75[degrees]35.2'W); the Narrows, Swan Quarter (Swan Quarter 1; 35[degrees]21'N, 76[degrees]24'W); Deep Cove, Swan Quarter, (Swan Quarter 2; 35[degrees]24'N, 76[degrees]23'W); Hoop Pole Creek, Atlantic Beach (Hoop Pole Creek; 34[degrees]42'N, 76[degrees]45'W) and in the Newport River at Mill Creek (Newport River; 34[degrees]46.2' N, 76[degrees]44.3'W; Fig. 1). These sites were chosen to cover as many different combinations of salinity and sediment-type as possible and varied in aspect and exposure (Table 1). Sites at Nags Head, Walterslough, Swan Quarter (1 and 2) and Hoop Pole Creek had a sloping bottom, ranging in depth from 0.2-1.4 m at mean low water (MLW), and minimal tidal influence (<0.1 m range). The site in the Newport River had a flat bottom, a tidal range of about 1 m and a depth of 1 m at mean low water.

[FIGURE 1 OMITTED]

We compared three culture methods--raised racks, bottom racks and suspended bags. The racks used for both raised and bottom culture were constructed of steel bars, were 0.8 m wide and 2.6 m long and held five 0.85 x 0.45 x 0.05 m oyster bags at the desired elevation off the bottom (0.15 m for the raised treatment; 0 m for the bottom treatment). Suspension of bags was achieved by inserting a 0.8 x 0.4 m piece of Styrofoam into each and stringing rows of five bags between four wooden posts, deployed in a 1.5 x 8 m array (Fig. 2). Two parallel ropes ran between the posts and were connected by crosspieces of PVC pipe, 1 cm in diameter. Each oyster bag rested on two PVC crosspieces, which helped dissipate strain on the ropes resulting from rough seas. The parallel ropes had sufficient slack to allow the bags to sit just below the air-water interface at high and low tide.

[FIGURE 2 OMITTED]

To compare growth, mortality, fouling and Polydora spp. infestation of oysters between fixed racks and suspended bags, at sites ranging in salinity and sedimentology, we deployed oysters at Nags Head, Walterslough, Swan Quarter 1, Swan Quarter 2 and Hoop Pole Creek in fall 2003 (Table 1). At all sites except Hoop Pole Creek, three racks and three suspended arrays (floats), each consisting of five 19-mm mesh polypropylene bags, were interspersed at each of 3 depths--shallow (MLW--0.2 m), medium (MLW--0.8 m) and deep (MLW--1.4 m; Fig. 3). The shallow depth was chosen so that racks and floats held oysters the same distance above the bottom; at the medium and deep depths, oysters in floating bags were a greater distance off the bottom than oysters on racks. This design enabled us to separate effects of (1) suspended float versus fixed structures and (2) distance off the bottom on the growth, mortality and fouling of oysters. If the distance of oysters off the bottom influences growth and mortality, but floating versus fixed culture is unimportant, the magnitude of the effect of method of culture should vary simply with depth. At Hoop Pole Creek, oysters were deployed only on racks at the deep depth. Oysters were maintained in the field for 24 wk, which was sufficient to achieve marketable size (75 mm shell height) in >50% of oysters in racks at Hoop Pole Creek. We visited sites every 8 wk at which time we scrubbed bags with a wire brush to remove fouling organisms, killed any crabs that had settled within the bags and checked oyster bags for damage and repaired these as necessary.

[FIGURE 3 OMITTED]

Our comparisons of raised and bottom culture of oysters commenced in spring 2004 at Walterslough, Swan Quarter 1 and the Newport River and extended for a 16-wk period into summer (Table 1). This was sufficient for >50% of oysters to achieve marketable size in raised and bottom racks at Newport River. At Walterslough and Swan Quarter 1, three raised (0.15 m) and three bottom (0 m) racks were deployed at each of the shallow, medium and deep depths defined earlier. At the Newport River site, a different grow-out configuration was used. We interspersed 18 raised and 18 bottom racks in a 6 x 6 rack matrix, 25 m x 25 m in area. Because of limited numbers of oysters, only 17 racks of each treatment were filled with C. ariakensis. As in winter, we cleaned and maintained oyster bags every 8 wk over summer.

The oysters, spawned at the Virginia Institute of Marine Sciences (VIMS) Aquaculture Genetics and Breeding Technology Center on July 2, 2003, were triploid C. ariakensis resulting from crosses between first-generation diploid females from the west coast of the United States (Taylor Shellfish, Shelton Washington) and first-generation tetraploid males (also from the west coast). Batch testing of the oysters indicated the genetic production of triploids to be 99.87% effective (S. K. Allen, pers. comm.). Between receipt of oysters on August 27, 2003 and deployment in bags in the fall of 2003 and spring of 2004, oysters were maintained at high density (1,000-2,000 per bag) in finer (2 mm) mesh bags at Swan Quarter 1. On transfer to the 19-mm mesh grow-out bags, oysters were thinned to a density of 150 per bag. At the time of deployment, oysters were at least 27 mm in shell height Table 1).

Sampling

Water-temperature at each of our sites was recorded hourly during the period of oyster grow-out using HoboTemp H8 Data Loggers attached to racks deployed at the medium depth at Nags Head, Walterslough, Swan Quarter 1 and Swan Quarter 2 and racks deployed at the deep depth at Hoop Pole Creek and Newport River. At Swan Quarter 1, additional HoboTemps were attached to floats at shallow, medium and deep depths and racks at shallow and deep depths, allowing comparison of temperature among treatments.

We ascertained mortality of C. ariakensis and measured sizes, percent cover on both valves of fouling organisms and infestation by Polydora spp. of oysters every 8 wk over the grow-out periods. Mortality of oysters was determined in the field by enumerating the number of dead oysters in three bags per rack. We then haphazardly selected 10 oysters from each of the same three bags per rack, transported them back to the laboratory and froze them at -20[degrees]C until measurements could be made. Use of this destructive method of sampling prevented us from tracking the growth of individual oysters through time. Growth was consequently expressed as an average daily increase in mean size (shell-height, dry mass of shell and dry mass of tissue).

On defrosting of samples, the exterior of each oyster was examined for the presence of fouling organisms, the cover of which was not influenced by freezing and thawing. We used visual estimates to document percentage of the shell surface (right and left valves combined) covered by fouling plants and animals because (1) confirmation of surface area with foil cut-outs revealed that rapid visual assessment provided estimates within 5% of true values 100% of the time (n = 100) and (2) visual estimation provides data that are more accurate and repeatable than other quantitative methods (Dethier et al. 1993). After estimation of cover, fouling organisms were removed by careful scraping with a knife from the exterior of oyster shells so that shell-heights and masses could be measured without bias.

Shell-heights (mm), dry shell masses (g) and dry tissue masses (g) were used to calculate growth of oysters. These measures were chosen because they are the most commonly reported by studies on growth and because tissue and shell growth in bivalves is not necessarily synchronized (e.g., Harvey & Vincent 1990). At each time of sampling, the maximal shell-height (distance between the hinge and opposite margin) of each collected oyster was measured to the nearest mm using digital calipers. Oysters were opened and the tissues carefully excised from the shell. The dry mass of tissue and shell were determined by drying at 80[degrees]C to constant (<0.1% variation) mass (2-5 d).

Prior to drying oyster shells, we examined their interior for the presence of mud blisters. We recorded the percentage of the interior surface of the shell (right and left valves) covered by blisters. Shells with greater than 25% internal cover of mud blisters are considered unsuitable for the half-shell market (Handley & Bergquist 1997).

To assess the susceptibility of Crassostrea ariakensis to Perkinsus marinus, the oyster parasite that causes Dermo, we sampled 15 oysters from raised and 15 from bottom racks at Walterslough, Swan Quarter 1 and also the Newport River in early October 2004. At Nags Head, 15 oysters were collected in October from raised racks only. A 3-5-mm-long section of the rectum was removed from each oyster and analyzed for the presence of P. marinus using fluid Thioglycollate medium (RTFM) assays (Ray 1952, Paynter & Burreson 1991).

Statistical Analyses

We tested for significant effects (at [alpha] = 0.05) of method of deployment on growth using ANOVAs on the final (1) shell-height, (2) shell mass and (3) dry tissue mass attained by oysters at each site, for winter and summer deployments. Separate analyses were done for each site because of differing initial sizes of oysters and differing times of deployment (Table 1). With the exception of analyses for the Newport River site, where the depth factor was not involved, the mixed model ANOVAs had 3 factors: method of deployment (raised [15-cm] racks versus floats or raised [15-cm] versus bottom [0-cm]; fixed); depth (shallow, medium, deep; fixed and orthogonal to method of deployment) and block (5-bag rack or float, random; nested within method of deployment). The experimental unit was the culture bag, with three sampled per rack. Analogous ANOVAs were done to compare mortality, fouling and infestation by mudblisters of oysters between methods of deployment and depths. Before each analysis, the assumption of homogeneity of variances was examined using Cochran C-test. In the case of heterogeneous variances, data were transformed to homogeneity using In (x +1). Analyses of variance were followed by a posteriori Student-Newman-Kuels (SNK) tests to identify those treatment means that differed significantly.

RESULTS

Measurements of water temperature revealed spatially concordant patterns of seasonality among our sites (Fig. 4). After deployment of oysters in fall of 2003, water temperature steadily declined from 21.6[degrees]C-23.0[degrees]C to a minimum of 0.7[degrees]C-4.8[degrees]C in early February. From late February to the conclusion of our first trial in May 2003, water temperatures warmed. Water temperatures continued to rise into the second trial from 27.3[degrees]C-27.8[degrees]C in early June to a peak of 29.3[degrees]C-30.4[degrees]C in late July. Overall, water temperatures were warmer during the second (May to October) than the first (October to May) deployment. On any date the greatest deviation in temperature among sites at Nags Head, Walterslough, Swan Quarter 1, Swan Quarter 2 and Newport River was 2[degrees]C and on most days was less than 1[degrees]C. Between December 12, 2003 and February 26, 2004, the water temperature at Hoop Pole Creek was on average 3[degrees]C warmer than at the other sites. At Swan Quarter 1, where temperature loggers were placed on racks and floats at each depth, temperatures were indistinguishable among methods of deployment.

[FIGURE 4 OMITTED]

Across our study sites there was no consistent difference in mortality among methods of deployment. Although at two of our sites, Walterslough and Swan Quarter 2, mortality was greater (by 76[degrees]C-133%) on raised racks than floats over the winter months (Fig. 5), at Swan Quarter 1 winter mortality was greatest in the floating racks deployed at a deep depth and at Nags Head there was no detectable difference between methods. Over summer, mortality was 74% greater among bottom than raised racks at Newport River but did not differ between methods at Walterslough or Swan Quarter 1.

[FIGURE 5 OMITTED]

Mortality of oysters varied among sites (Fig. 5). Over the cooler months, we recorded the greatest mortality of 16 [+ or -] 1% at the low salinity site, Nags Head. The higher salinity sites (Walterslough, Swan Quarter 1, Swan Quarter 2, Hoop Pole Creek) each had average mortalities of less than 10%. During the warmer period, mortality at Swan Quarter 1 was 37%. At this site, 100% mortality was recorded in four of the suspended bags and seven of the bags attached to fixed racks. Adult blue crabs were found in each of these decimated bags as were crushed oysters, characteristic of crab predation. Less mortality occurred at Walterslough (15 [+ or -] 1%) and Newport River (16 [+ or -] 7%) where settlement of crabs into bags was negligible.

We observed strong effects of method of deployment on growth of oysters during our field trials (Fig. 6, 7). During winter, growth in shell-height, shell mass and tissue mass of oysters at three of the sites, Walterslough, Swan Quarter 1 and Swan Quarter 2, was greater on fixed racks than floating structures (Fig. 6; Table 2, 3, 4). At the lowest salinity site, Nags Head, greater growth in shell mass in racks than floats was evident at medium and shallow depths, but dry-tissue mass did differ between methods of deployment. At the Newport River site, summer growth in shell-height, dry shell mass and dry tissue mass was greater in raised than bottom racks (Fig. 7; Table 5, 6, 7). At Swan Quarter 1, however, summer growth was similar between raised and bottom racks and at Walterslough, only growth in shell-height differed among methods, and only at the deep depth where it was greater in raised than bottom racks (Table 5).

[FIGURES 6-7 OMITTED]

Large differences in rate of growth were evident among our sites (Fig. 6, 7). During winter, oysters at the lowest salinity site, Nags Head, displayed negative growth in shell height and much slower rates of growth in shell and tissue mass than oysters at the others sites. Of the other four sites, Walterslough, Swan Quarter 1, Swan Quarter 2 and Hoop Pole Creek, growth in shell height and shell mass was greatest at Hoop Pole Creek and tissue mass was greatest at Swan Quarter 1. During the second summer grow-out, oysters at the Newport River site generally displayed higher growth than oysters at Walterslough and Swan Quarter 1. At all three sites, growth rates were greater than in winter.

At several of the sites, the degree to which oysters were fouled was dependent on method of deployment (Fig. 8). At Nags Head and Swan Quarter 2, oysters deployed on fixed racks had a greater percent cover of fouling organisms than suspended oysters. At Swan Quarter 1, oysters deployed on racks in deep water were more fouled than other oysters. During the summer deployment, fouling of oysters at Swan Quarter 1 was greater at shallow than deeper depths, independent of method of deployment. At Newport River, oysters on raised racks were more fouled than those on bottom racks. In both winter and summer, fouling was greatest at Walterslough (Fig. 8). At Walterslough, green algae covered 30% to 40% of the surface of oyster shells, more than double the total cover of fouling organisms on oysters at any of the other sites.

[FIGURE 8 OMITTED]

Infestation of oysters by Polydora spp. was negligible over the winter months. At all sites, mean cover of the interior surface of oyster shells by mud blisters was less than 4% (Fig. 9). At Walterslough and Swan Quarter 1, not one of the oysters had a cover that was greater than 25%, the acceptable limit for the half-shell market. At Nags Head and Swan Quarter 2, an insignificant 0.4% had greater than 25% cover of mud blisters. At Hoop Pole Creek this figure was 4%. Thus, although some patterns of difference in infestation among methods of deployment were evident (Fig. 9) they are of little economical consequence.

[FIGURE 9 OMITTED]

Over the summer months, in contrast, mean cover of the interiors of oyster shells by mud blisters was well over 50% (Fig. 9). At Walterslough, 96% of oysters had greater than 25% cover of mud blisters on the interior surface of their shell. This proportion was similarly high at Swan Quarter 1 (97%) and Newport River (84%). At none of the sites did cover of mud blisters significantly differ between raised and bottom racks.

We did not detect Perkinsus marinus in any of the 105 oysters sampled from raised or bottom racks at Nags Head, Walterslough, Swan Quarter 1 and Newport River in early October 2004.

DISCUSSION

Aquaculture of triploid oysters entails ecological and economical risk (NRC 2003). Because the efficiency of triploid production is seldom 100% and triploids themselves may revert to diploidy, controlled aquaculture of triploids may lead to establishment of diploid, self-sustaining populations. Nonnative pathogens may not be effectively excluded by quarantine protocols and the nonnative oyster may act as a vector for the spread of endemic pathogens or parasites (NRC 2003). Thus, decisions regarding the introduction of triploid shellfish to novel environments should be based on the same careful consideration of perceived economical, sociologic and ecological costs and benefits as introductions of diploids. Our study sought to determine: (1) methods and seasons of culture that are most appropriate for culture of C. ariakensis and (2) rates of C. ariakensis growth and mortality under these optimal conditions, necessary information for modeling projected economical benefits of a C. ariakensis introduction for aquaculture. We found that, unlike other species of oyster, thin-shelled C. ariakensis grow fastest when cultured in fixed racks as opposed to suspended bags, and like other species suffer the least mortality and fouling when cultured from fall through spring.

In our trials done at six sites across North Carolina, shell and tissue growth of C. ariakensis were consistently less in suspended bags than in fixed racks held 15 cm off the bottom. In the late fall and early winter (November to December), chipping of shell margins resulted in negative growth in shell-height of oysters in suspended culture at all sites. Oysters on stable racks were generally less chipped except at Nags Head, the most exposed site, where winter storms produced large holes in >80% of C. ariakensis. Although, in most instances, these holes that exposed oyster tissue did not kill the C. ariakensis, their repair presumably reduced the amount of energy available for growth in shell and tissue. Unlike oysters cultured on floats at Walterslough and Swan Quarter, which increased in tissue mass throughout the year, at Nag's Head, oysters lost tissue mass over the winter. In the spring, shell growth was 36% to 132% and tissue growth 7% to 61% greater on the fixed racks than in the suspended bags. In contrast to the sizeable difference in the performance of C. ariakensis in float racks and fixed racks, the performance of C. ariakensis was generally similar on raised (15 cm high) and bottom racks. Only at our muddiest site, Newport River, was a difference in performance evident between elevations. This difference was most evident in growth of tissue, which was 50% greater in racks raised above the surface of the sediment. Growth in shell-height was only 6% greater on raised than bottom racks and growth in shell mass differed by 20% between elevations. The lower growth of C. ariakensis when deployed on muddy sediment may be explained by the resuspension of fine sediments decreasing the efficiency of feeding (Rhoads & Young 1970, Muschenheim 1987).

Of the six sites at which oysters were deployed, growth was greatest at the high-salinity sites, Hoop Pole Creek (29-32 ppt) and Newport River (25-30 ppt) and lowest at the low-salinity site, Nags Head (4 ppt). This pattern is consistent with the relationship between growth of C. ariakensis and salinity suggested by previous studies done in North America. Calvo et al. (1999) found that growth of C. ariakensis in the Chesapeake Bay was reduced at an intermediate salinity range of 15-25 ppt in contrast to sites with salinities consistently >25 ppt. In North Carolina, C. ariakensis grew fastest at salinities >25 ppt (Grabowski et al. 2004). Although in our study, warmer winter water temperatures at Hoop Pole Creek than at the other sites may have contributed to the great rate of growth of oysters at this high salinity site, in summer water temperatures at Newport River were similar as at other sites such that temperature alone cannot explain the faster growth of oysters at high salinity sites. The extremely low growth of C. ariakensis at our low (4 ppt) salinity site, Nags Head, confirms the conclusion that salinities of <10 ppt are unsuitable for culture of C. ariakensis (Grabowski et al. 2004), although the large fetch of our low salinity site probably also contributed to decreased rates of growth by increasing the frequency of rough seas, which may inhibit feeding and cause abrasion of shells. Growth at the medium-salinity site Walterslough, where fetch is also great, was poorer than at sheltered medium-salinity sites.

Our study recorded up to 50% less mortality of C. ariakensis during the winter than the summer deployment. Whereas winter mortality at all sites was less than 16%, summer mortality at Swan Quarter approached 40%. The great summer mortality of oysters at Swan Quarter is unlikely to be caused by the parasite Bonamia sp. that was detected for the first time in North Carolina waters in 2003 (Burreson et al. 2004) and appears to increase in prevalence over the warmer months (R. B. Carnegie unpublished data). Bonamia sp. is purged from oysters at salinities <20 ppt and seems to only infect spat (<25 mm; R. B. Carnegie, C. Audemard unpublished data). Similarly the summer mortality cannot be attributed to Perkinsus marinus, one of the major contributors of mortality in the native oyster (e.g., Lenihan et al. 1999). Perkinsus marinus was notably absent from all 105 C. ariakensis samples tested.

Instead, it seems that the site-specific summer mortality was caused by predation by blue crabs. Crabs have been identified as one of the major predators in shellfish culture on shores bordering the Atlantic and the Gulf of Mexico (Menzel & Hopkins 1955, Parsons 1974, Walne & Davies 1977). Whereas crabs primarily attack newly-planted oysters on the sea-bed, megalops larvae have previously been observed to settle into oyster trays (Behrens Yamada et al. 1993). Tray-raised oysters are thinner shelled (Paynter & DiMichele 1990) and thus more easily opened by crabs than intertidally raised oysters. In the 11 bags of oysters at Swan Quarter where 100% mortality of oysters occurred, we discovered blue crabs with carapace widths as large as 15 cm. Moreover, these bags contained large numbers of crushed oyster shells and molted crab carcasses, lending further support for the hypothesis that mortality was caused by crabs that had settled into the bags. At sites where settlement of blue crabs was not observed, mortality averaged a low 16%. Results of feeding trials show that the exotic Suminoe oyster, C. ariakensis, is treated as prey by even naive North American blue crabs lacking previous experience with the oyster and that this nonnative oyster is preyed on by blue crabs at far higher rates than the native oyster of the Atlantic and Gulf coasts of North America (Bishop & Peterson submitted [a]). Even C. ariakensis well outside the size-range of native oysters usually consumed by blue crabs are eaten in sizeable numbers (Bishop & Peterson submitted[a]). The high summer mortality of C. ariakensis observed in this study, in combination with feeding trials examining predation by blue crabs on C. ariakensis, suggests that in the absence of regular culling of crabs within oyster bags, predators may contribute to significant losses of C. ariakensis from culture operations.

Our results indicate that by growing oysters over winter it is possible to avoid heavy settlement of fouling organisms that may slow the growth of oysters (Bishop & Peterson submitted[b]) and lower their value on the half-shell market (Handley & Bergquist 1997). With the exception of the site at Walterslough, where anomalous fouling by algae covered 30% to 40% of the exterior of oyster shells, the cover of fouling organisms on oysters was much less than 10% at the end of the winter grow-out period. Of particular importance was the minimization of Polydora spp. infestation by winter grow-out. At all sites, mean cover of the interior surface of oyster shells by mud blisters was less than 4% for the winter deployment. After the summer grow-out season, in contrast, almost all oysters (84% to 97%) had covers of mud blisters on their interior shell that exceed 25% and hence rendered them unsuitable for the half-shell industry (Handley & Bergquist 1997). Although the greater infestation by Polydora spp. of oysters deployed in the summer may be explained by the longer period of time these oysters were in the water, in Virginia and Maryland where C. ariakensis have been deployed at different times of the year high levels of Polydora spp. infestation have also been reported among C. ariakensis over summer months (K. T. Paynter pers. com., R. B. Carnegie unpublished data) supporting the hypothesis that Polydora spp. infestation is a widespread seasonal problem. Periodic treatment of oysters with freshwater, heated seawater (Nel et al. 1996) or a hypersaline dip (J. Swartzenberg, J&B Aquafoods, pers. com.) may reduce Polydora spp. infestation (Nel et al. 1996). It remains, unclear, however, whether treatment of thin-shelled C. ariakensis would sufficiently reduce infestation of oysters to an acceptable level for half-shell consumption.

Because of the complex relationship between season and patterns of C. ariakensis growth, mortality, fouling and infestation by Polydora spp., the success of C. ariakensis culture will depend on season of harvest. Because of the heavy infestation of C. ariakensis by Polydora spp. over summer, only oysters deployed in fall and harvested by mid spring are likely to be suitable for sale on the half-shell market, which places great importance on the appearance of oyster shells. Based on the rates of C. ariakensis growth we observed at Hoop Pole Creek and Newport River (Fig. 6, 7), we estimate that 27-mm oysters (i.e., those just large enough to be retained by 3.4" shellfish bags) deployed at high (>20 ppt) salinity sites in early fall could reach a shell height of 75-mm, the legally harvestable size of C. virginica, by early spring, well before summer infestation by Polydora spp. and possible seasonal decline in tissue quality corresponding to the reproductive season. At medium (10-20 ppt) salinity sites such as Walterslough, Swan Quarter 1 and 2, more than 7 mo would be required for oysters to reach 75-mm (Fig. 6, 7). Thus, oysters planted at medium salinity sites in early fall would need to be harvested at smaller size if they are to be sold on the half-shell market before summer Polydora spp. infestation. Failure to remove oysters at medium salinity sites from the water before summer may result in heavy Polydora spp. infestations that necessitate sale on the lower-valued shucked market. At high salinity sites, it may be possible to grow out two crops of C. ariakensis targeting the shucked market per year if one is over the summer months when grow-out may be completed in 3 months (Fig. 7). Irrespective of target-market, it is recommended that C. ariakensis be grown in fixed racks as opposed to floating structures to maximize rates of growth without compromising survivorship.

ACKNOWLEDGMENTS

The authors thank J. Grabowski and three anonymous reviewers for improving the quality of this manuscript. This research was funded by the North Carolina Division of Marine Fisheries and the Golden Leaf Foundation. J. Braddy, D. Schmitt and M. Ulery reared the oysters. They were assisted in the laboratory by N. Chaplinski, J. Farmer, T. Hutcherson, M. Moretz, C. Peterson and C. Waters. Dr S. Allen Jr. of the Virginia Institute of Marine Sciences provided triploid C. ariakensis.

LITERATURE CITED

Bayne, B. L., J. Widdows & C. W. Worrall. 1977. Some temperature relationships in the physiology of two ecologically distinct bivalve populations. In: F. J. Venberg, A. Calabrese, F. P. Thurberg and W. Vernberg (eds.). Physiological Responses of Marine Biota to Pollutants. New York: Academic Press. pp. 379-400.

Behrens Yamada, S., H. Metcalf & B. C. Baldwin. 1993. Predation by the crab, Cancer oregonensis Dana, inside oyster trays. J. Shellfish Res. 12:89-92.

Bishop, M. J. & C. H. Peterson. a. Predation as a mechanism of invasion resistance: high allocation of resources to rapid development does not predict invasion success. Ecol. App. (Submitted).

Bishop, M. J. & C. H. Peterson. b. Direct effects of physical stress can be counteracted by the indirect: invertebrate growth on a tidal elevation gradient. Oecologia (Submitted).

Bishop, M. J. & Peterson C. H. 2005. Consumer ratings of the non-native oyster, Crassostrea ariakensis, during home cooking. J. Shellfish Res.

Breitburg, D. L., L. D. Coen, M. W. Luckenbach, R. Mann, M. Posey & J. A. Wesson. 2000. Oyster reef restoration: convergence of harvest and conservation strategies. J. Shellfish Res. 19:371-377.

Brumburgh, R. D., L. A. Sorabella, C. Johnson & W. J. Goldsborough. 2000. Small scale aquaculture as a tool for oyster restoration in Chesapeake Bay. Marine Tech. Soc. J. 34:79-86.

Burreson, E. M. & L. M. Ragone Calvo. 1996. Epizootiology of Perkinsus marinus disease of oysters in Chesapeake Bay, with emphasis on data since 1985. J. Shellfish Res. 15:17-34.

Burreson, E. M., N. A. Stokes, R. B. Carnegie & M. J. Bishop. 2004. Bonamia sp. (Haplosporidia) found in nonnative oysters Crassostrea ariakensis in Bogue Sound, North Carolina. J. Aquat. Anim. Health 16:1-9.

Calvo, L. M. R., G. W. Calvo & E. M. Burreson. 2003. Dual disease resistance in a selectively bred eastern oyster, Crassostrea virginica, strain tested in Chesapeake Bay. Aquaculture 220:69-87.

Calvo, G. W., M. W. Luckenbach, S. K. Allen & E. M. Burreson. 1999. Comparative field study of Crassostrea gigas (Thunberg, 1893) and Crassostrea virginica (Gmelin, 1791) in relation to salinity in Virginia. J. Shellfish Res. 18:465-473.

Calvo, G. W., M. W. Luckenbach, S. K. Allen & E. M. Burreson. 2001. A comparative field study of Crassostrea ariakensis (Fujita 1913) and Crassostrea virginica (Gmelin 1791) in relation to salinity in Virginia. J. Shellfish Res. 20:221-229.

Coen, L. D. & M. W. Luckenbach. 2000. Developing success criteria and goals for evaluating oyster reef restoration: ecological function or resource exploitation? Ecol. Eng. 15:323-343.

Dethier, M. N., E. S. Graham, S. Cohen & L. M. Tear. 1993. Visual versus random-point percent cover estimations: 'objective' is not always better. Mar. Ecol. Prog. Ser. 96:93-100.

Doering, P. E. & C. A. Oviatt. 1986. Application of filtration rate models to field populations of bivalves: an assessment using experimental mesocosms. Mar. Ecol. Prog. Ser. 31:265-275.

Frankenberg, D. 1995. North Carolina blue ribbon advisory council on oysters. Final report and recommendations. Raleigh, NC: NC DENR.

Grabowski, J. H., C. H. Peterson, S. P. Powers, D. Gaskill & H. C. Summerson. 2004. Growth and survivorship of non-native (Crassostrea gigas and Crassostrea ariakensis) vs. native eastern oysters (Crassostrea virginica). J. Shellfish Res. 23:781-793.

Handley, S. J. 1997. Optimizing subtidal oyster production, Marlborough Sounds, New Zealand: spionid polychaete infestations, water depth, and spat stunting. J. Shellfish Res. 1:143-150.

Handley, S. J. & P. R. Bergquist. 1997. Spionid polychaete infestations of intertidal Pacific oysters Crassostrea gigas (Thunberg), Mahurangi Harbour, northern New Zealand. Aquaculture 153:191-205.

Harvey, M. & B. Vincent. 1990. Density, size distribution, energy allocation and seasonal variations in shell and soft tissue growth at two tidal levels of a Macoma balthica (L.) population. J. Exp. Mar. Biol. Ecol. 142:151-168.

Haskin, H. H. & S. E. Ford. 1988. Characteristics of inbred oyster strains selected for resistance to Haplosporidium nelsoni (MSX). J. Shellfish Res. 7:162.

Kirby, M. X. 2004. Fishing down the coast: historical expansion and collapse of oyster fisheries along continental margins. Proc. Natl. Acad. Sci. USA 101:13096-13099.

Kirkley, J. 1997. Virginia's commercial fishing industry: its economical performance and contributions. Virginia Sea Grant Publication VSG97-02, VIMS, College of William and Mary, Gloucester Point, VA.

Korringa, P. 1951. The shell of Ostrea edulis as a habitat: observations on the epifauna living in the Oosterchelde, Holland, with some notes on polychaete worms occurring there in other habitats. Arch. Neerl. Zool. 10:32-152.

Korringa, P. 1952. Recent advances in oyster biology. Q. Rev. Biol. 27: 266-365.

Lenihan, H. S. & C. H. Peterson. 1998. How habitat degradation through fishery disturbance enhances impacts of hypoxia on oyster reefs. Ecol. Appl. 8:128-140.

Lenihan, H. S., F. Micheli, S. W. Shelton & C. H. Peterson. 1999. The influence of multiple environmental stressors on susceptibility to parasites: An experimental determination with oysters. Limnol. Oceanogr. 44:910-924.

Lesser, M. P., S. E. Shumway, T. Cucci & J. Smith. 1992. Impact of fouling organisms on mussel rope culture: interspecific competition for food among suspension-feeding invertebrates. J. Exp. Mar. Biol. Ecol. 165:91-102.

Lodeiros, C., D. Pico, A. Prieto, N. Narvaez & A. Guerra. 2002. Growth and survival of the pearl oyster Pinctada imbricata (Roding 1758) in supended and bottom culture in the Golfo de Cariaco, Venezuela. Aqua. Int. 10:327-338.

Loosanoff, V. L. & J. B. Engle. 1943. Polydora in oysters suspended in the water. Biol. Bull. 85:69-78.

MacKenzie, C. 1997. The molluscan fisheries of the Chesapeake Bay. The history, present condition, and future of molluscan fisheries of North and Central America, and Europe, V.1, Atlantic and Gulf Coast. US Department of Commerce. NOAA Tech. Rep. 127:141-169.

Mann, R. 2000. Restoring the oyster reef communities in the Chesapeake Bay: a commentary. J. Shellfish Res. 19:335-339.

Mann, R., E. M. Burreson & O. K. Baker. 1991. The decline of the Virginia oyster Fishery in the Chesapeake Bay: considerations for introduction of a non-endemic species, Crassostrea gigas (Thunberg, 1973). J. Shellfish Res. 10:379-388.

Matthiessen, G. C., S. Y. Feng & L. Leibovitz. 1990. Patterns of MSX (Haplosporidium nelsoni) infection and subsequent mortality in resistant and susceptible strains of the eastern oysters, Crassostrea virginica (Gmelin, 1971), to Perkinsus marinus. J. Shellfish Res. 9:359-365.

Menzel, R. W. & S. H. Hopkins. 1955. Crabs as predators of oysters in Louisianna. Proc. Nat. Shellfish. Assoc. 46:177-184.

Micheli, F. 1997. Effects of predator foraging behavior on patterns of prey mortality in marine soft bottoms. Ecol. Monogr. 67:203-224.

Moroney, D. A. & R. L. Walker. 1999. The effects of tidal and bottom placement on the growth, survival and fouling of the eastern oyster Crassostrea virginica. J. World Aqua. Soc. 30:433-442.

Muschenheim, D. K. 1987. The dynamics of near-bed seston flux and suspension-feeding benthos. J. Mar. Res. 45:473-496.

National Research Council. 2003. Study on non-native oysters in the Chesapeake Bay. Committee on non-native oysters in the Chesapeake Bay, Ocean Studies Board, division on earth and life sciences. Washington, DC: National Academies Press.

Nel, R., P. S. Coetzee & G. Van Niekerk. 1996. The evaluation of two treatments to reduce mud worm (Polydora hoplura Clapare'de) infestations in commercially reared oysters (Crassostrea gigas Thunberg). Aquaculture 141:31-39.

Parsons, J. 1974. Advantages of tray culture of Pacific oysters (Crassostrea gigas) in Strangford Lough, N. Ireland. Aquaculture 3:221-229.

Paynter, K. T. & L. DiMichele. 1990. Growth of tray-cultured oysters (Crassostrea virginica Gmelin) in Chesapeake Bay. Aquaculture 87: 289-297.

Paynter, K. T. & E. M. Burreson. 1991. Effects of Perkinsus marinus infection on the eastern oyster, Crassostrea virginica: 2. Disease development and impact on growth rate at different salinities. J. Shellfish Res. 10:425-431.

Ray, S. M. 1952. Perkinsus. Science 116:360-361.

Rhoads, D. C. & D. K. Young. 1970. The influence of deposit-feeding organisms on sediment stability and trophic community structure. J. Mar. Res. 28:150-178.

Rothschild, B. J., J. S. Ault, P. Goulletquer & M. Heral. 1994. Decline of the Chesapeake Bay oyster population: a century of habitat destruction and overfishing. Mar. Ecol. Prog. Ser. 111:29-39.

Shatkin, G., S. E. Shumway & R. Hawes. 1997. Considerations regarding the possible introduction of the Pacific oyster (Crassostrea gigas) to the Gulf of Maine: A review of global experience. J. Shellfish Res. 16:463-477.

Soletchnik, P., C. Lambert & K. Costil. 2005. Summer mortality of Crassostrea gigas (Thunberg) in relation to environmental rearing conditions. J. Shellfish Res. 24:197-207.

Walne, P. R. & G. Davies. 1977. The effect of mesh covers on the survival and growth of Crassostrea gigas Thunberg grown on the sea bed. Aquaculture 11:313-321.

MELANIE J. BISHOP, (1,2) * CHARLES H. PETERSON (1)

(1) University of North Carolina at Chapel Hill, Institute of Marine Science, Morehead City, North Carolina 28557; (2) Department of Environmental Sciences, University of Technology Sydney, Gore Hill NSW 2065 Australia

* Corresponding author: E-mail: [e-mail]

TABLE 1.
Summary of environmental conditions and deployments of oysters
at the six study sites.

                  Salinity                          Maximal
Site               (ppt)     Substrate    Aspect   Fetch (km)

Nags Head            4       Sand         West        4.8
Walterslough        5-15     Sand         South       1.2
Swan Quarter 1     10-14     Muddy-sand   East        0.4
Swan Quarter 2     10-15     Sandy-mud    East        2.4
Hoop Pole Creek    29-32     Muddy-sand   West        0.1
Newport River      20-30     Sandy-mud    South       1.5

                   Period of Deployment
Site                    of Oysters             Methods of Deployment

Nags Head         Nov 04 2003-May 10 2004   Raised racks, floats
Walterslough      Nov 05 2003-May 11 2004   Raised racks, floats
                  May 20 2004-Oct 12 2004   Raised racks, bottom racks
Swan Quarter 1    Oct 16 2003-Apr 19 2004   Raised racks, floats
                  May 28 2004-Sep 21 2004   Raised racks, bottom racks
Swan Quarter 2    Oct 20 2003-Apr 28 2004   Raised racks, floats
Hoop Pole Creek   Dec 10 2003-May 21 2004   Raised racks
Newport River     Jun 11 2004-Sep 27 2004   Raised racks, bottom racks

                      Shell-height
                  (+1 SE) of Oysters at
Site                 Deployment (mm)

Nags Head           50.6 [+ or -] 1.1
Walterslough        50.6 [+ or -] 1.1
                    37.1 [+ or -] 0.9
Swan Quarter 1      48.0 [+ or -] 0.8
                    37.7 [+ or -] 0.9
Swan Quarter 2      48.8 [+ or -] 1.3
Hoop Pole Creek     29.9 [+ or -] 5.9
Newport River       49.0 [+ or -] 0.8

TABLE 2.
Results of ANOVAs testing for the effect of method of deployment
and depth on growth in shell-height of C. ariakensis over the
winter grow-out period.

                          Nags Head              Walterslough

               df     MS      F       P       MS     F       P

Me              1   378.0   43.31   0.000   896.8   38.1   0.000
De              2     0.5    0.06   0.942    48.5    2.1   0.170
B1 (Me x De)   12     8.7    0.68   0.758    23.5    1.6   0.147
Me x De         2    35.1    4.02   0.046    24.1    1.0   0.389
Res            36    12.8                           15.1
Cochran test        C = 0.21, P > 0.05       C = 0.16, P > 0.05
SNK test            Me x De S: R = F;        Me R > F
                    M: R > F; D: R > F

                   Swan Quarter 1           Swan Quarter 2

                  MS       F       P       MS       F       P

Me             3438.1   213.2   0.000   5967.6   149.5   0.000
De               99.6     6.2   0.014     57.7     1.4   0.274
B1 (Me x De)     16.1     2.0   0.053     39.9     2.4   0.021
Me x De          51.8     3.2   0.076     40.9     1.0   0.388
Res               8.0                     16.7
Cochran test   C = 0.18,P > 0.05        C = 0.22, P > 0.05
SNK test       Me R > F                 Me R > F
               DeS = M < D

Me = method (2 levels; rack [R], float [F]), De = depth (3 levels;
shallow [S], medium [M], deep [D]), B1 = block [3 levels]) Res =
residual, n = 3.

TABLE 3.
Results of ANOVAs testing for the effect of method of deployment
and depth on growth in dry shell mass of C. ariakensis over the
winter grow-out period.

                     Nags Head                 Walterslough

                df    MS     F       P      MS      F        P

Me               1   24.0   19.6   0.001   182.0   26.5   0.000
De               2    0.1    0.1   0.922     8.1    1.2   0.341
B1 (Me x De)    12    1.2    1.1   0.425     6.7    0.8   0.660
Me x De          2    5.3    4.4   0.038     9.8    1.4   0.276
Res             36    1.2                    8.7
Cochran test         C = 0.18, P > 0.05    C = 0.22, P > 0.05
SNK test             Me x De S: R = F;     Me R > F
                     M: R > F; D: R > F

                    Swan Quarter 1          Swan Quarter 2

                MS        F       P      MS       F       P

Me             269.8     48.8   0.000   869.0   294.4   0.000
De              40.7      7.4   0.008     0.9     0.3   0.747
B1 (Me x De)     5.5      2.0   0.056     3.0     0.8   0.614
Me x De          1.8      0.3   0.731     8.1     2.7   0.105
Res              2.8                      3.5
Cochran test   C = 0.26, P > 0.05       C = 0.25, P > 0.05
SNK test       Me R > F                 Me R > F
               De S = M < D

Me = method (2 levels; rack [R], float [F]), De = depth (3 levels;
shallow [S], medium [M], deep [D]), B1 block [3 levels]) Res =
residual, n = 3.

TABLE 4.
Results of ANOVAs testing for the effects of method of deployment
and depth on growth in dry tissue mass of C. ariakensis over the
winter grow-out period.

                        Nags Head           Walterslough

               df    MS     F      P      MS     F       P

Me              1   0.03   1.1   0.313   1.63   18.6   0.001
De              2   0.01   0.2   0.815   0.11   1.3    0.318
B1 (Me x De)   12   0.03   1.7   0.108   0.09   0.6    0.831
Me x De         2   0.04   1.6   0.235   0.01   0.2    0.853
Res            36   0.02                 0.15
Cochran test        C = 0.14, P > 0.05   C = 0.18, P > 0.05
SNK test                                 Me R > F

                 Swan Quarter 1         Swan Quarter

                MS     F       P      MS     F      2P

Me             4.70   13.6   0.003   4.05   50.4   0.000
De             2.06    6.0   0.016   0.11    1.4   0.288
B1 (Me x De)   0.35    1.5   0.162   0.08    0.8   0.698
Me x De        0.16    0.5   0.640   0.28    3.4   0.066
Res            0.23                  0.11
Cochran test   C = 0.19, P > 0.05    C = 0.25, P > 0.05
SNK test       Me R > F              Me R > F
               De S = M < D

Me = method (2 levels; rack [R], float [F]), De = depth (3 levels;
shallow [S], medium [M], deep [D]), B1 = block [3 levels]) Res =
residual, n = 3.

TABLE 5.
Results of ANOVAs testing for the effects of method of deployment
and depth on growth in shell-height of C. ariakensis over the
summer grow-out period.

                       Walterslough         Swan Quarter 1

               df    MS     F      P      MS      F      P

Me              1   79.1   8.2   0.014    97.3   0.7   0.430
De              2   48.1   5.0   0.027    59.1   0.4   0.676
B1 (Me x De)   12    9.7   1.1   0.391   145.7   9.1   0.000
Me x De         2   88.2   9.1   0.004   153.7   1.1   0.378
Res            36    8.8                  16.1
Cochran test        C = 0.21, P > 0.05   C = 0.23P > 0.05
SNK test            Me x De S: R = B;
                    M: R = B; D: R > B

                                   Newport

                         df    MS      F       P

Me             Me         1   829.7   15.4   0.000
De             BI (Me)   32    54.0    6.0   0.065
B1 (Me x De)   Res       68    34.7
Me x De
Res
Cochran test                  C = 0.12, P > 0.05
SNK test                      Me R > B

Me = method (2 levels; raised rack [R], bottom rack [B]), De = depth
(3 levels; shallow [S], medium [M], deep [D]), BI = block [3 levels])
Res = residual, n = 3.

TABLE 6.
Results of ANOVAs testing for the effects of method of deployment
and depth on growth in dry shell mass of C. ariakensis over the
summer grow-out period.

                        Walterslough

               df      MS       F      P

Me              1      0.2     0.0   0.911
De              2     48.4     2.9   0.091
Bl (Me x De)   12     16.5     2.9   0.007
Me x De         2     33.8     2.1   0.171
Res            36      5.6
Cochran test          C = 0.23, P > 0.05
SNK test

                       Swan Quarter 1

                      MS        F      P               df

Me                  0.31 (a)   1.7   0.220   Me         1
De                  0.02       0.1   0.903   Bl (Me)   32
Bl (Me x De)        0.19       7.2   0.000   Res       68
Me x De             0.45       2.4   0.129
Res                 0.03
Cochran test        C = 0.24, P > 0.05
SNK test

                            Newport

                      MS       F       P

Me                  1015.8     6.3   0.018
De                   162.5     4.9   0.000
Bl (Me x De)          33.3
Me x De
Res
Cochran test        C = 0.12, P > 0.05
SNK test            Me R > B

(a) Data were ln (x + 1) transformed prior to analysis

Me = method (2 levels; raised rack [R], bottom rack [B]), De = depth
(3 levels; shallow [S], medium [M], deep [D]), B1 = block [3 levels])
Res = residual, n = 3.

TABLE 7.
Results of ANOVAs testing for the effects of method of deployment
and depth on growth in dry tissue mass of C. ariakensis over the
summer grow-out period.

                        Walterslough

               df    MS       F      P

Me              1   0.05     1.8   0.202
De              2   0.04     1.4   0.274
Bl (Me x De)   12   0.03     4.7   0.000
Me x De         2   0.02     0.6   0.588
Res            36   0.01
Cochran test        C = 0.28, P > 0.05
SNK test

                       Swan Quarter 1

                     MS      F       P

Me                  0.01     0.1   0.750   Me
De                  0.11     1.6   0.235   Bl (Me)
Bl (Me x De)        0.07    10.5   0.000   Res
Me x De             0.01     0.1   0.877
Res                 0.01
Cochran test        C = 0.30, P > 0.05
SNK test

                          Newport

               df    MS      F       P

Me              1   17.5    59.5   0.000
De             32    0.3     1.9   0.013
Bl (Me x De)   68    0.2
Me x De
Res
Cochran test        C = 0.12, P > 0.05
SNK test            Me R > B

Me = method (2 levels; rack [R], bottom rack [B]), De = depth (3
levels; shallow [S], medium [M], deep [D]), Bl = block [3 levels])
Res = residual, n = 3.