Assessing potential benthic impacts of harvesting the Pacific geoduck clam Panopea generosa in intertidal and subtidal sites in British Columbia, Canada.
KEY WORDS: benthic impact, eelgrass, geoduck, clam, harvest, Panopea generosa
The Pacific geoduck clam Panopea generosa (Gould, 1850) is distributed from Alaska to Baja California (58-28[degrees] N) (Bernard 1983). It lives in the low intertidal zone and subtidally as deep as 110 m, buried in sand, silt, and gravel (Goodwin & Pease 1989, Bureau et al. 2002, Zhang & Hand 2006). It is the largest infaunal clam in the world, growing up to 3.25 kg whole weight and living up to a meter below the sediment surface (Goodwin & Pease 1987). This species is also long-lived, the oldest individual on record being approximately 168-y old (Bureau et al. 2002).
The Pacific geoduck clam currently supports the most valuable dive fishery on the west coast of North America, 1,963 metric tons (MT), worth US$36.2 million, being landed in Washington state (WA) in 2010 (Washington Department of Fish and Wildlife 2012); 1600 MT, worth C$40.9 million, in British Columbia (BC), Canada in the same year (BC Ministry of Agriculture 2010); and 312 MT, worth US$3.8 million, in Alaska in the 2009/2010 fishing season (Alaska Department of Fish and Game 2014). In addition, a fishery for two species of geoduck (Panopea generosa and Panopea globosa) in Mexico has grown rapidly since the early 2000s to landings of 2,000 MT in 2011, worth US$30.0 million (Aragon-Noriega et al. 2012). Two relatively underdeveloped fisheries (annual harvest less than 20 MT) for smaller species of Panopea occur in Argentina with Panopea abbreviata (Morsan & Ciocco 2004, Morsan et al. 2010) and New Zealand with Panopea zelandica and Panopea smithae (Breen et al. 1991, Gribben & Creese 2005, New Zealand Fisheries 2013).
Aquaculture production of geoducks started intertidally in WA in the mid-1990s and has increased quickly; approximately 613 MT of cultured clams, worth US$18.5 million, were harvested in 2010 (Washington Department of Fish and Wildlife 2012). There has been widespread interest in the culture of geoducks in BC for many years, but commercial-scale development has been hindered until fairly recently by a lack of governmental policy/legislation and concerns of how geoduck culture impacts the environment. Nevertheless 52 MT of farmed geoduck, worth C$1.1 million, were harvested in 2010 in BC (BC Ministry of Agriculture 2012, BC aquaculture production statistics from BC Ministry of Agriculture, unpublished data received August 2012). Environmental concerns with geoduck aquaculture are usually focused on the harvest process since pressurized water jets (or "stingers" in industry vernacular) are used to dislodge the soft-bottom substrates around the clams to extract them. This procedure is not used just by aquaculturists, it is also the typical harvest technique used in the wild fisheries of all Panopea species and so similar environmental concerns surrounding harvest practices apply to the various Panopea fisheries as well.
Geoduck harvesting by water jets may be highly disruptive of the substrates (Goodwin 1978, Breen & Shields 1983). During the harvest, disturbed sediments are suspended in the water column. While large particles settle fairly rapidly in the immediate vicinity, finer ones will be carried away by water currents, forming turbid plumes, and subsequently redeposited some distance away (Short & Walton 1992). After a geoduck is removed, a shallow hole about 0.5 m in diameter, partially filled with an emulsion of loose substrate and water, is created (Goodwin 1978, Breen & Shields 1983). The ecological implications of harvesting, however, may extend far beyond such purely physical, sedimentary effects. As the substrate is disturbed, both abiotic and biotic attributes of the benthos may also be altered. Geoduck harvesting may impact the benthic environment in a number of ways: (1) alteration of sediment grain size due to loss of fine particles and loose compaction of redeposited substrate that is more susceptible to removal by water currents (Goodwin 1978); (2) loss of organic matter, minerals, and heavy metals associated with loss of fine particles, as the fines (<63 [micro]m) tend to accumulate or bond with such materials more than other grain-size fractions (Horowitz & Elrick 1987, Tam & Wong 2000); (3) exposure of anoxic sediments and oxygenation of sediment pore water, affecting sediment chemistry (Palazzi et al. 2001, Straus et al. 2008); (4) release of materials back into the water column, including nutrients, eggs or cysts, and contaminants (Pilskaln et al. 1998, Tengberg et al. 2003, Straus et al. 2008), subsequently affecting water quality and animal and plant growth; (5) reduction in infaunal abundance due to damage, burial, and exposure to currents and predators (Goodwin 1978, Breen & Shields 1983, Currie & Parry 1996); and (6) impact on nearby aquatic communities in areas outside the immediate harvest bed due to creation of turbid plumes and deposition of materials from such plumes (Short & Walton 1992). The areas nearby harvest plots may be important near-shore marine habitats such as open sand/mud flats and eelgrass (seagrass) meadows, both hosting diverse animal and plant communities (Cain & Bradbury 1996, Short & Wyllie-Echeverria 1996, Vermaat et al. 1997, Chambers et al. 1999. Rossi et al. 2007). Deposition of materials from turbid harvest plumes onto nearby areas may lead to changes in sediment grain size and infaunal communities through burying, smothering, crushing, and alteration of the benthic chemical microenvironment (Miller et al. 2002, Airoldi 2003). Further, decreased light levels due to shading, as a result of increased turbidity from sediment plumes and deposition of sediments on leaf surfaces, may reduce eelgrass growth and survival (Moore et al. 1997, Cabello-Pasini et al. 2002, Tamaki et al. 2002).
The potential impact of geoduck harvesting on benthic environments appears to be minimal for commercial subtidal fisheries in both WA and BC. Goodwin (1978) reported that (1) harvesting did not significantly affect sediment grain size distribution in harvest plots as a whole; (2) harvesting did not create dramatic decreases in the major infaunal species present; and (3) holes created during a disturbance had disappeared completely 7 mo after the harvest. There were, however, significant decreases in the percentage of fine and coarse sediments within the harvest holes immediately after harvesting (Goodwin 1978). Breen and Shields (1983) did not find any significant difference in sediment grain size distribution or simple relationship in changes in infaunal community structure between harvested and nonharvested plots, but did report an increase in species diversity in the disturbed one. Using a modeling approach, Short and Walton (1992) concluded that most suspended materials settled within 1 m of the holes created and that transport and deposition of suspended sediments associated with geoduck harvesting would have minimal impacts on the physical environment of the harvest bed and adjacent area. No studies, however, have examined the potential effect of harvesting subtidally cultured or enhanced geoduck populations, where clam densities would more likely be higher than in the wild.
Theoretically, intertidal areas may be more resilient to disruptions than subtidal ones as they are subject to more frequent and intense natural disturbance. Thus, impacts of intertidal harvesting might be expected to be even less than those observed in the subtidal wild fisheries. This supposition is tempered by the fact that clam densities (and hence level of disturbance) in the former will be much higher than those in the latter. But recent studies have confirmed the lack of impact. Sauchyn et al. (2013) reported that impact of small-scale intertidal harvests on various sedimentary variables was limited in terms of scale and duration. Price (2011) concluded that commercial-scale harvesting did not cause any distinct response in infaunal communities within harvest plots and that effects on infauna were within the range of natural variation experienced by the community and not of long-term ecological significance. Regarding infaunal community structure. Price (2011) also found that harvesting did not cause any "spillover" effects in areas adjacent (up to 60 m outside) to the plots. The objective of the present study was to evaluate the spatial and temporal extent of the potential impact of large-scale subtidal and intertidal geoduck harvests on the benthic environment.
MATERIALS AND METHODS
Study Sites and Sampling Locations
The study was conducted between October 2008 and October 2010 at two sites in the Strait of Georgia, BC, both comprising a harvest plot, a nearby nonharvest reference area, and an eelgrass bed (Fig. 1). The Cortes Island (CI) site (50[degrees] 02' N, 124[degrees] 58' W, approximate) was located in the northern part of the Strait, on a subtidal sandy strip 3.5-7.8 m below chart datum on a portion of a geoduck fisheries bed. The harvest plot (100 X 60 m) was a geoduck fisheries enhancement area placed within the commercial bed, previously seeded between 1999 and 2000 and ready for harvesting during the course of the present study. The area nearby the enhanced plot had never been seeded or harvested.
The Nanoose Bay (NB) site(49[degrees] 16'05.68" N, 124[degrees] 10'43.74" W, center of harvest plot) was located on a shellfish (Pacific oyster and Manila clam) tenure on an intertidal sand flat (3.6-5.1 m above chart datum at high tide). The entire study site, including the harvest plot (30 x 15 m), had not been used for aquaculture for many years before this study and no geoduck clams or other cultured bivalves were present, although clams and oysters were being commercially cultured nearby. At the time of this study there were no commercial-scale intertidal geoduck farms within BC that were ready for harvesting, hence a mock harvest was conducted (i.e., the ground was disturbed as if a harvest were occurring, but no geoducks were actually present). While the mock harvest mimicked the physical disturbance that occurs during such an event, the study was not able to assess the potential effects of releasing/suspending certain biological components associated with geoduck culture (e.g., faeces, pseudofaeces). It should be noted that there was a small eelgrass-bed intrusion in the northeast corner of the harvest plot at NB (Fig. 1).
At the start of the project, current profiles were determined at both sites using an Acoustic Doppler Current Profiler (Teledyne RD Instruments, San Diego, CA) set centrally in the plots destined for harvesting. Current direction and velocity were recorded every 10 min for a period of 6 days at CI and 7 days at NB. Data from three depth bins (0.3, 2.8, and 5.7 m above sea bed at CI; 0.2, 0.6, and 1.1 m above sea bed at NB) were extracted and averaged to determine major current directions and velocities. The data were then used to determine transect lines through the centers of the harvest plots and parallel to the major current direction. The nearby reference areas were positioned in the predominant down-current direction of the harvest plots at both sites as this was the area predicted to be most likely impacted by suspended sediments. Nearby eelgrass beds were in the direction paralleling the current (CI and NB) and up-/down-current of the harvest plot (NB) (Fig. 1). Typical current speed was 6-18 cm/s at CI and 0-12 cm/s at NB.
In the nearby areas, five sampling distances were allocated at CI and six at NB along the transect lines: 5, 10, 25, 50, and 75 m from the edge of the harvest plot at CI and 1, 5, 10, 25, 50, and 75 m at NB. The harvest plot was considered as 0 m for both study sites. The gradient sampling design assumed that maximum impact occurred at or adjacent to the harvest plot with impact intensity decreasing with distance, dropping to nil at a certain distance from the area of harvest (Borja et al. 2009). Previous small-scale, intertidal research revealed that impacts of harvest were localized to within 10 m of the harvest zone (Sauchyn et al. 2013). We therefore considered the 50 and 75 m distances to be likely control locations. For the eelgrass bed at CI, four sampling distances from the edge of the harvest plot (5, 10, 25, and 50 m) were assigned (Fig. 1). The eelgrass bed at NB was located in two directions from the harvest plot (shoreward and seaward), three sampling distances (1,5, and 10 m) being used for each direction (Fig. 1). The maximum eelgrass-bed sampling distances approximated the eelgrass boundary or the access limit during low tides (i.e., the seaward direction at NB).
Sampling Schedules and Variables
Samples were taken in the harvest plot, nearby area, and eelgrass bed over a 2-y period, ranging from 1 y prior to harvest to 1 y post harvest for CI and immediately prior to harvest to 2 y post harvest for NB (Table 1). The different sampling schedules at the two sites reflected a trade-off between longer-term, preharvest sampling, which allowed documentation of the natural variability prior to disturbance, and longer-term, postharvest sampling, which allowed assessment of the potential rate of recovery of impacted variables over multiple seasons/years. At each time, samples were taken at each sampling distance in the nearby area and eelgrass bed from five points (n = 5), which were spaced approximately evenly across the length or width of the harvest plot (Fig. 1). Five random samples were also taken within the harvest plot at each sampling time at both study sites (Fig. 1). Within the harvest plot and nearby area, samples were collected to determine sediment grain size, percent organics, total nitrogen, total organic carbon, sulfide content, redox potential, infaunal community structure, and sedimentation during harvesting. Within the eelgrass bed, samples were collected to examine sediment grain size, infaunal community structure, eelgrass shoot length, eelgrass shoot density, eelgrass biomass, and sedimentation during harvesting. It should be noted that eelgrass samples were not taken immediately post harvest at either study site since harvesting was not conducted directly on the eelgrass beds and no direct physical damage to the eelgrass populations would be expected from the disturbance so soon after harvest. Indirect harvest effects on the eelgrass, due to siltation and/or release of dissolved compounds from the sediments, would unlikely be detected for some time after the disturbance. Additional samplings were undertaken to monitor seasonal eelgrass variations.
Sample Analysis: Sediment Physics and Chemistry
At each sampling point and time, the top 2-cm layer of sediments was collected using a sample corer (6.5-cm diameter X 20-cm height), transported to the laboratory on ice, and frozen at -20[degrees]C. After samples were thawed and overlying seawater removed, subsamples were freeze-dried for later determination of percent organics, total nitrogen, and total organic carbon. The remaining portion of the samples was dried at 60[degrees]C to constant weight for later determination of sediment grain size. Organic content was expressed as a percentage of sample dryweight loss after combustion at 500[degrees]C for 5 h. Total nitrogen and organic carbon were determined by high-temperature combustion in a Carlo Erba CHN analyzer (NA-1500) and expressed as percentages of sample dry weight. Sediment grain size was determined by sifting samples through a series of nested 203-mm diameter sieves on a sediment shaker. Particle compositions were calculated as percentages of total sample dry weight for gravel (>2,000 [micro]m), very coarse/coarse sand (2,000-500 [micro]m), medium sand (500-250 [micro]m), fine/very fine sand (250-63 [micro]m), and silt/clay (<63 [micro]m), according to the Wentworth (1922) scale.
Sulfide content and redox potential of sediments collected at 2- and 6-cm depth were measured. At CI, a sample corer (6.5-cm diameter X 20-cm height) with two small holes (1.7-cm diameter, 4 cm apart vertically) was pushed into the seabed at each sampling point to position the two holes at 2- and 6-cm depths. A sediment sample was then taken from each hole using a 10-ml cut-off plastic syringe. The syringe was sealed air-tight, stored on ice, and transported to the laboratory. At NB, a sample corer (as above, but with the two holes sealed with duct tape) was pushed into the seabed at each sampling point. The whole corer, with contained sediments, was then sealed air-tight at the two ends and taken on ice to the laboratory, as the presence of gravels made it difficult to apply the syringes on site given the time available. Samples were analyzed within a few hours after collection. Prior to analysis, samples were left at room temperature in the dark for 1 h. Sulfide content was measured with a silver/sulfide electrode and redox potential with a platinum redox electrode, after the method of Wildish et al. (1999). Redox potential readings were corrected to a standard hydrogen reference electrode.
Sample Analysis: Infaunal Community
A sediment core (6.5-cm diameter X 10-cm height) was collected at each sampling point and time. During preliminary samplings at both study sites we rarely encountered larger species such as bivalves and gastropods that could not be taken by a corer 6.5 cm in diameter. After overnight storage at 4[degrees]C, the cores were washed on a 0.5-mm sieve and the materials retained were preserved in 8% phosphate-buffered formalin for 1-2 wk and then transferred to 70% ethanol for longer-term storage. Prior to identification, the materials were washed on a 1-mm sieve. We chose this mesh size based on (1) protocols established for the US Environmental Protection Agency for sampling subtidal benthic macroinvertebrate assemblages in Puget Sound (Puget Sound Water Quality Authority 1987) and (2) preliminary observations on our own 0.5-mm fraction which revealed that it was relatively clean in comparison with the 1-mm component, being made up of mostly smaller individuals or juveniles of what was present in the 1-mm fraction. While many studies on infaunal populations have used either 0.5 or 1-mm mesh sizes for screening, some recent work has examined the meiofaunal (45-1,000 [micro]m) community (e.g., Gallucci et al. 2012), which can be an important component of soft-sediment environments. The present study did not assess potential impacts on meiofaunal populations, which would include recruiting larvae. All retained organisms were classified to the lowest taxonomic level by one designated infaunal taxonomy specialist. The numbers of species and individuals and Shannon-Wiener index were calculated for each core (Crawford et al. 2003, Borja et al. 2009).
Sample Analysis: Eelgrass
Eelgrass samples were taken from a 40 x 40-cm quadrat at each eelgrass-bed sampling point and time. All above-ground shoots in the quadrats were severed and stored at -20[degrees]C until examination. The thawed samples were sorted to determine maximum shoot length and shoot density (CI), cleaned of any visible epifauna, and dried at 60[degrees]C to constant weight to determine per-quadrat dry biomass (CI and NB) for each sampling point.
Harvesting and Sedimentation During Harvesting
At CI, the harvest plot had a surveyed geoduck density of 1.58 ind/[m.sup.2] prior to harvest in 2008. A total of 1,554 geoducks (mean weight: 0.82 kg) were collected in two work days by a commercial dive crew using standard harvest practices. This represented a harvest intensity of 0.26 ind/[m.sup.2] on the 6,000-[m.sup.2] plot, which is substantially higher than what would be occurring in the wild fishery. The range of densities on wild geoduck beds in the vicinity of the harvest plot was 0.03-0.32 ind/[m.sup.2], but the wild fishery operates on a 3-y rotation at a harvest rate of only 1.8% estimated biomass per year or a maximum of 5.4% estimated biomass every 3 y (Fisheries and Oceans Canada 2012). The wild fishery would, therefore, target an overall removal rate of 0.02 ind/[m.sup.2] every 3 y for wild geoduck beds near the study site, illustrating how potential impacts from cultured/enhanced geoduck harvesting may be amplified compared with the wild fishery. Individual clams were identified by their "shows" (siphon tips protruding from the sediments) and harvested one by one.
At NB, a mock harvest was carried out by inserting a pressurized water jet, powered by a 5.5-hp water pump, repeatedly into the substrate across the 450-[m.sup.2] harvest plot during a low tide, creating approximately 9 holes/[m.sup.2] (essentially the whole plot being disturbed). There was a small eelgrass-bed intrusion in the northeast corner of the harvest plot and it was also affected (Fig. 1). Typically, intertidal tenures in WA are planted with 20.000 predator-protection tubes per 0.5 acre with two to three juveniles placed in each tube (Davis 2004), which amounts to a seeding density of 20-30 ind/[m.sup.2]. Assuming one geoduck per tube survives (Davis 2004), then a harvest density would be approximately 10 ind/[m.sup.2], which is very similar to our 9 holes/[m.sup.2]. Typically, intertidal harvests in WA are carried out using the swath technique whereby the entire culture plot is stung out, moving from one end of the plot to the other in a systematic fashion.
Deposition of suspended materials created by harvesting was determined using sediment traps. At both study sites, three sediment traps were used in the harvest plot (along the central line perpendicular to the transect line) and at each sampling distance in the nearby area and in the eelgrass bed (Fig. 1). Each trap was 40 cm high and 7.7 cm in diameter, with an aspect ratio >5 (Ongley 2006). Prior to harvesting, the traps were deployed for 2 days to collect background suspended sediment data and then redeployed just before harvesting and collected 2-3 days later when the harvest was completed. It should be noted that the subtidal traps collected both sediments created during harvesting and those redeposited by water currents after it was completed. The intertidal traps, however, only collected sediments redeposited by water currents after harvesting was completed as the tide came in. It should also be noted that, at both study sites, it was quite windy before harvesting, but very calm during and after.
At each sampling point, the trap was placed in a larger polyvinyl chloride (PVC) pipe, embedded in the seabed, to minimize disturbance of the surrounding sediments during the set-up and removal of the traps. At NB, sediments inside the larger PVC pipes had been carefully dug out, so that the openings of the traps placed inside were 15 cm above the seabed, to increase the submersion time of the traps as the tide came in. No sediments in the larger PVC pipes were removed at CI and the openings of the traps were 40 cm above the seabed.
After recovery, the traps were kept in the dark for at least 12 h to allow suspended material to settle and overlying seawater was then siphoned off. The trapped materials were transferred into preweighed 50-ml plastic tubes and centrifuged for 10 min at 1,509 g. The resultant solids were washed with distilled water, centrifuged again under the same conditions, and dried at 60[degrees]C to constant weight. Sedimentation rates were determined as dry sediment weight collected per trap per day (g/trap/day) at each sampling point.
Sedimentation from Additional Sampling
Suspended sediments were collected at CI during a winter storm in 2011. Six sediment traps (three in the nearby area and three in the eelgrass bed) were deployed just before the storm (February 11) and retrieved after the storm (February 16). Background data on suspended sediments for a calm sea were not collected until March 20-24, 2011 since sporadic storms passed through the area for a long time.
Sampling during winter storms was not possible at NB as a tide low enough to facilitate sample collection during a storm event never occurred during the study period. Instead, annual sedimentation rate was monitored at NB every 2-3 mo for 1 y (April 2009 to April 2010). At each sampling time, nine sediment traps (three in the nearby area and three in both directions of the eelgrass bed) were deployed for 11-14 days during a tidal cycle. At both study sites, the set-up of sediment traps and processing of sediment samples were the same as previously described.
Statistical analysis was facilitated using the software PERMANOVA + for Primer (Clarke & Gorley 2006, Anderson et al. 2008). This software was used to examine the main effects of sampling distance and time, and their interaction, on various data sets from CI and NB (Tables 2 and 3), the harvest plot/nearby area and eelgrass bed of each study site (the two directions of eelgrass beds at NB) being analyzed separately. Infaunal data were fourth-root transformed and other environmental data were standardized, with the Bray-Curtis similarity and the Euclidean distance measures being used, respectively, to generate similarity matrices. Data in the text are presented as ranges from the lowest to the highest means observed across the different distances over the study period for each variable examined, unless otherwise specified. The false discovery-rate control procedure of Benjamini and Hochberg (1995) was used to control excessive type I error in the PERMANOVA tests with an overall significance level set at P < 0.05. A total of 19 PERMANOVA tests were conducted for CI and NB combined. This resulted in an adjusted significance level of P < 0.005.
Interpretations of potential harvesting effects in the present study are based on concepts of the BACI design (Green 1979, Stewart-Oaten et al. 1986, 1992, Goldberg et al. 2012). If the interaction between sampling distance and time is nonsignificant, this suggests that each distance (including the harvest plot and control location) shows the same pattern of variation in response to time, therefore indicating that the harvest effect is likely none. This is irrespective of the main effects and likely due to heterogeneity across space and considerable natural variability over time. If the interaction is significant, however, it does not necessarily mean that the harvest effect is significant as other factors may also be contributing to spatial and temporal variation. Attention was paid to consistent patterns in the data, if any, to elucidate if the significant interaction terms were more likely due to a harvest effect rather than natural variability.
In addition, two-way analysis of variance (ANOVA) was used to examine sedimentation before/during harvesting and one-way ANOVA used to examine sedimentation from additional sampling. Newman-Keuls (NK) analyses were used to identify the occurrence of significant pair-wise differences. Analysis of variance and NK were conducted using the software NCSS 2007 (Kaysville, UT). All data in the ANOVA analyses were log-transformed to satisfy normality and homogeneity, as confirmed by Kolmogorov-Smirnov and Levene's tests, respectively. The level of significance of all ANOVA tests was set at P < 0.05.
Harvest Plot and Nearby Area: Sediment Physics and Chemistry
Sediments of the harvest plot and nearby area at CI were composed mainly of medium sand (48.0%-58.8%), followed by very coarse/coarse and fine/very fine sands (17.5%-26.5% and 18.9%-26.9%, respectively). Silt/clay accounted for only <0.3% of the sediments and no gravel was encountered (Fig. 2). Percent organics ranged from 0.42% to 0.64%, total nitrogen from 0.015% to 0.025%, and total organic carbon from 0.078% to 0.169%. Sulfide contents were 12.5-326.4 [micro]M at 2-cm depth and 45.4-273.0 [micro]M at 6-cm depth. Redox potential at the respective depths was 188.5-334.8 mV and 186.5-323.7 mV (Fig. 3). The software PERMANOVA did not reveal any significant interactions between sampling distance and time for any of these data sets except for sulfide content and redox potential at 2-cm depth (Table 2). There was no consistent pattern, however, to relate the significance to the harvest (Fig. 3).
Sediments of the harvest plot and nearby area at NB were composed mainly of fine/very fine sand (41.8%-82.2%). The site was also characterized by a wide range of gravels (0.1%-36.5%), suggesting a heterogeneous sediment composition. Percents of very coarse/coarse sand, medium sand, and silt/clay were low (2.8%13.3%, 8.7%-25.6%, and 2.5%-7.5%, respectively) (Fig. 4). The interaction between sampling distance and time was significant for sediment grain size (Table 3). This significance appeared to be related to the +18 sampling (April 30, 2010), when a recent landwater runoff swept away finer sediments at 50 and 75 m, but had the opposite effect at the other distances (Fig. 4). Percent organics at NB ranged from 0.80% to 1.54%, total nitrogen from 0.034% to 0.074%, and total organic carbon from 0.27% to 0.56% (Fig. 5). Sulfide contents were 34.7-445.7 and 152.9 192.5 [micro]M at 2- and 6-cm depths, respectively, and redox potential 120.3-262.9 and 91.1-257.0 mV, respectively (Fig. 5). The interaction between sampling distance and time was nonsignificant for each set of sediment chemistry variables examined at NB (Table 3).
Harvest Plot and Nearby Area: Infaunal Community
The number of species per core at CI ranged from 7.6 to 25.2, the number of individuals from 11.2 to 61.6, and the Shannon-Wiener index from 1.6 to 2.8 (Fig. 6). The interaction between sampling distance and time was nonsignificant for infaunal community structure (Table 2). At each sampling time, annelids, arthropods, and molluscs (predominately bivalves) were the most common infauna, accounting for 20.0%-14.3%, 20.4%-49.7%, and 12.0%-16.4% of the respective total individuals enumerated over the entire harvest plot and nearby area.
At NB, the numbers of species and individuals per core were 5.2-16.6 and 10.2-98.0, respectively. The Shannon-Wiener index ranged from 1.0 to 2.2 (Fig. 7). The interaction between sampling distance and time was nonsignificant for infaunal community structure (Table 3). Annelids, arthropods, and molluscs (predominately bivalves) were the most abundant fauna observed at each sampling time, accounting for 38.1%-59.6%, 17.7%-50.4%, and 6.3%-20.8%, respectively, of the total individuals counted in the entire harvest plot and nearby area.
Harvest Plot and Nearby Area: Sedimentation During Harvesting
At CI, sediments collected at each distance (0-75 m) ranged from 0.22 to 0.69 g/trap/day before the harvest, but were lower (0.04-0.09 g/trap/day) during the harvest except for the harvest plot (0.88 g/trap/day) and the 5-m distance (5.72 g/trap/day) (Fig. 8). The much higher value at 5 m was caused by one large replicate value (16.86 g/trap/day), which was likely due to direct "spill" from the harvest. After that large value was removed from the analysis, two-way ANOVA showed that the interaction between sampling distance and time was significant ([F.sub.(5,23)] = 4.38, P = 0.006). An NK test revealed that there was no significant difference among all distances in the background before-harvest data. During harvesting, sediment levels in the harvest plot (0 m) were significantly higher than those at all the other distances except for 5 m, yet comparable to those before the harvest. When compared with the before-harvest data, although generally less sediment was collected at each distance from 5 to 75 m during the harvest, the differences were significant only for 75 m.
At NB, sediments collected at each distance (0-75 m) ranged between 0.78 and 1.47 g/trap/day before the harvest, but were lower (0.09-0.62 g/trap/day) during the harvest (Fig. 8). Two-way ANOVA showed that the interaction between sampling distance and time was significant ([F.sub.(6.28)] = 5.14. P = 0.001). An NK. test revealed that significantly less sediment was collected during the harvest than before the harvest at each distance except for the harvest plot (0 m) and 5 m.
Eelgrass Bed: Sediment Physics
At CI, sediment composition of the eelgrass bed was similar to that of the harvest plot and nearby area, being 13.1%-28.2% for very coarse/coarse sand, 43.3%-58.5% for medium sand, 18.9%-10.7% for very fine/fine sand, and <0.5% for silt/clay (data not shown). The interaction between sampling distance and time was nonsignificant for sediment grain size (Table 2).
Sediment compositions of the eelgrass beds at NB were predominately fine/very fine sand (63.5%-84.6% and 71.1%-88.3% for the seaward and shoreward beds, respectively), followed by medium sand (7.2%-18.6% and 6.5%-18.0%), very coarse/coarse sand (3.3%-12.2% and 1.0%-5.8%), and silt/clay (2.6%-6.8% and 2.0%-9.4%). Gravel content was low for both beds (<4.0%) (data not shown). The interactions between sampling distance and time were nonsignificant for sediment grain size in both eelgrass beds at NB (Table 3).
Eelgrass Bed: Infaunal Community
At CI, the number of species, the number of individuals, and the Shannon-Wiener index were 6.6-20.2, 13.4-95.0, and 1.4-2.6 per core, respectively (data not shown). The interaction between sampling distance and time was nonsignificant for infaunal community structure (Table 2). At each sampling time, molluscs (bivalves) were the more observed infaunal group, accounting for 37.5%-63.7% of the total number of individuals counted over the entire eelgrass bed, followed by annelids and arthropods (13.6%-30.7% and 16.1%^12.2%, respectively).
Infaunal community structures of the seaward and shoreward eelgrass beds at NB-number of species per core: 7.2-17.0 and 6.2-15.6; number of individuals per core: 14.0-85.2 and 13.4-80.8; Shannon-Wiener index: 1.7-2.3 and 1.4-2.4--were similar (data not shown). The interactions between sampling distance and time were nonsignificant for infaunal community structure for both eelgrass beds (Table 3). At each sampling time, annelids, arthropods, and molluscs (predominately bivalves) were the most common taxa, accounting for 30.5%62.8%, 3.1%-44.4%, and 11.3%-41.1%, respectively, of the total number of individuals enumerated over the entire eelgrass beds.
Eelgrass Bed: Eelgrass Parameters
At CI, the maximum shoot length of eelgrass ranged from 45.4 to 76.8 mm, shoot density from 3.5 to 16.5 per quadrat, and biomass from 1.28 to 7.83 g/quadrat (Fig. 9). The eelgrass was exclusively Zostera marina.
At NB, the eelgrass biomass ranged from 0.57 to 9.23 g/quadrat for the seaward bed and from 0.97 to 12.58 g/quadrat for the shoreward bed (Fig. 9). The eelgrass species present at NB were Zostera marina and Zostera japonica. The inconsistent distribution of the two eelgrass species over space and time made it difficult to compare such variables as shoot length and density. Neither of the interactions between sampling distance and time were significant for eelgrass parameters at CI and NB (Tables 2 and 3).
Eelgrass Bed: Sedimentation During Harvesting
At CI, the amounts of suspended sediments collected at each distance (0-50 m) were 0.28-0.83 g/trap/day before the harvest. Lower amounts of sediment were collected at each distance during the harvest (0.02-0.04 g/trap/day), except for the harvest plot (0 m) (0.88 g/trap/day) (Fig. 8). Two-way ANOVA showed that the effects of sampling distance, time, and their interaction were all significant (distance: [F.sub.(4, 20)] = 6.23, P = 0.002; time: [F.sub.(1, 20)] = 68.1, P < 0.0001; interaction: [F.sub.(4, 20)] = 15.1, P < 0.0001). An NK test revealed that significantly more sediment was collected in the harvest plot (0 m) than at all the other distances during the harvest and that significantly less sediment was collected during than before the harvest at each distance (0-50 m) except for the harvest plot (0 m).
At NB, the amounts of sediments collected at each distance (0-10 m) before harvesting were 0.65-1.08 g/trap/day in the seaward bed and 1.12-4.34 g/trap/day in the shoreward bed. During harvesting, the amounts were lower than before harvesting at 1 and 10 m (0.26 and 0.59 g/trap/day) of the seaward bed, 5 and 10 m (0.36 and 0.26 g/trap/day) of the shoreward bed, and the harvest plot (0 m) as well (0.45 g/trap/day) (Fig. 8). Higher amounts of sediments were observed during than before harvesting at 5 m in the seaward bed (2.92 g/trap/day) and at 1 m of the shoreward bed (2.22 g/trap/day). Two-way ANOVA results, however, did not reveal any significance for sampling distance, time, or their interaction ([F.sub.(3, 16)] = 2.37, 116) = 1.73, and [F.sub.(3,16)] = 0.95, respectively, al; P > 0.05) for the seaward bed. For the shoreward bed, two-way ANOVA revealed that significantly less sediment was collected during than before harvesting [F.sub.(1,16)] = 12.34, P > 0.003) and that the effects of sampling distance and interaction between distance and time were not significant ([F.sub.(3, 16)] = 1.31 and [F.sub,(3, 16)] = 2.36, respectively, both P > 0.05).
Sedimentation from Additional Sampling
The recorded wind speed was 9.8/20 km/h (average/maximum hourly) on February 11; 19.7/33 km/h on February 12; 13.4/28 km/h on February 13; 20.7/35 km/h on February 14; 7.0/19 km/h on February 15; and 6.3/15 km/h on February 16, at the closest weather station at Campbell River, BC (Climate ID: 1021261; Meteorological Service of Canada 2012). The wind was mostly from the southeast, which would have had higher impact at CI. The amount of sediments collected during the winter storm event at CI was 0.36 [+ or -] 0.02 g/trap/day (mean [+ or -] SE, n = 6) which was significantly (one-way ANOVA, [F.sub.(1, 10)] = 69.95, P < 0.01) greater than that collected during a calm sea (0.02 [+ or -] 0.00 g/trap/day).
The annual sedimentation rates at NB were relatively low in April, June, and August (0.48 [+ or -] 0.09, 0.22 [+ or -] 0.06, and 0.10 [+ or -] 0.07 g/trap/day, respectively; mean [+ or -] SE, n = 9), elevated in November (2.07 [+ or -] 1.48 g/trap/day), and peaked in January (9.04 [+ or -] 2.35 g/trap/day), after which the rates decreased (1.92 [+ or -] 0.58 g/trap/day in next April). The amount of sediment collected in January was significantly higher than that at any other time of the year. November to March is typically the heavy precipitation season in the study areas (Environment Canada 2012).
Of the various benthic parameters examined in the harvest plots, nearby areas, and eelgrass beds, the interactions between sampling distance and time were mostly nonsignificant at both study sites, except for sulfide content and redox potential at 2-cm depth at CI and sediment grain size at NB. The former significance seemed not to have been directly related to harvest activities. The latter significance was related to a large land water runoff at the +18 sampling. Overall, these results indicate no significant benthic impacts of harvesting geoduck clams at either site, including the harvest plots, nearby areas, and eelgrass beds. The results will be of relevance not only to the intertidal culture and subtidal enhancement of Panopea generosa, but also to the wild fishery of the species and to the culture and fishery of other Panopea species.
The results of the present study are consistent with previous research on the benthic impacts of subtidal wild fisheries and intertidal aquaculture. Subtidal studies by Goodwin (1978) and Breen and Shields (1983) revealed no dramatic changes in sediment grain-size distribution and no major change or simple relationship in infaunal community structure in harvest plots 7 or 10 mo after the disturbance. Species diversity (Shannon-Wiener index) actually increased as a result of harvesting in the study of Breen and Shields (1983). Similarly, Price (2011) reported that commercial-scale intertidal harvesting did not appear to significantly negatively affect various benthic parameters, including infaunal community structure, over time. In contrast to the present work, some previous studies have observed significant changes in certain benthic characteristics immediately after harvesting, such as sediment composition in harvest plots/holes or in infaunal community structure, but they were short-lived and disappeared within several months (Goodwin 1978, Price 2011, Sauchyn et al. 2013) or did not extend very far outside the area of harvest (<10 m, Sauchyn et al. 2013). Temporal changes in infaunal populations may be short term due to the fact that geoduck harvesting has the potential to displace and yet preserve benthic fauna so that they can recolonize the disturbed areas immediately after harvesting (Price 2011) and because small disturbed patches (resulting from point-source harvesting) can be recolonized more quickly by movement of fauna across sediments due to their higher edge/surface area ratios (Guerra-Garcia et al. 2003).
Table 4 summarizes geoduck harvesting intensities in various subtidal and intertidal studies in WA and BC. Despite these studies varying in harvest intensities (e.g., harvest plot size, harvest duration, and number of holes per unit area) and likely in site-specific conditions (e.g., depth, tidal current, sediment composition, infaunal community structure, and productivity), the collective results suggest that geoduck harvesting has very limited impact on the benthic environment and any significant effect is generally short-lived or near-field. The results seen with geoducks contrast with other commercial shellfish harvesting activities such as suction-dredging cockles, in which a large area could be disturbed intensively within a relatively short time (e.g., a trench of 0.5-1.15 m wide and up to 8 km long per h per boat) causing long-lasting negative effects, up to 8 y, in sediment composition and bivalve stock in the fished area (Piersma et al. 2001). The recovery of the benthic environment after various forms of shellfish harvesting can often take days to months (Hall et al. 1990, Currie & Parry 1996, Kaiser et al. 1996, Ferns at al. 2000, Tuck et al. 2000, Kaiser et al. 2001, Constantino et al. 2009), in extreme cases years (Piersma et al. 2001), depending on the form and intensity of harvesting.
Subtidal geoduck harvesting with water jets places sediments in suspension and may result in effects within a broader area than the point of direct disturbance (ENVIRON International Corporation 2009). Depending on the current speed (0.05-1.00 m/s), small quantities of suspended materials from subtidal harvesting may be deposited up to 100-200 m down-current, but most settle within 1 m of the harvest holes (Short & Walton 1992). Intertidal harvesting at low tide can result in overland flow of water used in the operations, transporting suspended sediments over the exposed intertidal area to the water's edge (Fleece et al. 2004). In both cases, it is the fines (<63 [micro]m) that are the most relevant to transportation by water current and redeposition away from the source substrate, as they settle much more slowly and remain in the water column for longer periods than larger particles (Short & Walton 1992, Palazzi et al. 2001).
Based on a simulation model using a fine content of 8% in sediments, Short and Walton (1992) predicted that deposition of all suspended materials from a subtidal harvest would be 0.4 cm thick (including all grain sizes) in the affected down-current area if 2,500 holes were made per 0.25-acre bed or 2.5 holes/[m.sup.2], typical of high-density geoduck fisheries beds in WA (Palazzi et al. 2001). Palazzi et al. (2001) estimated a layer of 0.2 cm sediments of just fines if 10,000 holes were created per acre with a fine content of 3.5% and if all the fines settled within that acre. At the subtidal site CI, the fines accounted for <0.3% of the sediments. Such a low fine content, usually associated with a high-energy environment, is not uncommon in commercial geoduck fishery beds in BC and would be likely in future geoduck aquaculture tenures. Under such conditions, little fine materials would be available for suspension and subsequent redeposition due to harvesting. This supposition is supported by sedimentation data compiled from sediment traps in the nearby downcurrent area at CI. Suspended sediments collected during the 2-day harvest at 5-75 m were 0.04-0.09 g/trap/day (except for a large replicate value at 5 m), representing a layer of 0.001-0.002 cm thick over the whole nearby area [estimated using a sediment density of 1.84 g/[cm.sup.3] (Short & Walton 1992)]. Even if the present harvest intensity of the subtidal plot was increased 10 times to 2.6 holes/[m.sup.2] within the 6,000-[m.sup.2] harvest plot, the accumulation of suspended sediments would be projected to be 0.01-0.02 cm thick, well below the estimations of Short and Walton (1992) and Palazzi et al. (2001). Further, suspended sediment amounts collected during harvesting at CI were similar to those during a calm sea (0.02 g/trap/day), but much lower than those in a rough sea just before the harvest and during the winter storm at this study site (0.22-0.69 and 0.36 g/trap/day, respectively). In the intertidal study site (NB), the fines accounted for 2.5%-7.5% of the sediments (Fig. 4). The amount of suspended materials collected during the harvest at 1-75 m (except for one large replicate value at 5 m) was 0.09-0.30 g/trap/day, representing a layer 0.002-0.007 cm thick over the one-tidal cycle harvest (estimated as above). The annual sedimentation rates at NB varied in the range of 0.10-9.04 g/trap/day, including those in windy conditions (just before the harvest), and can be much higher than rates during harvesting.
The present study did not examine the phenomenon of overland flow, caused by water used for intertidal harvesting, carrying suspended sediments into the water column. Fleece et al. (2004) and ENVIRON International Corporation (2009) found that increased turbidity from intertidal harvesting was limited to the shore area <25' from shoreline, peaked at 100 [+ or -] 50' downstream of the harvest site, and declined rapidly within a short distance. The distance a turbid plume may travel is dependent on a number of factors including proximity of water edge to the harvest site, strength and direction of near-shore currents, sediment characteristics of the culture beach, and local weather during the harvest. Natural turbidity generated along the shoreline during windy days is generally not distinguishable from that created by harvesting and turbidity generated by harvesting is only visible on calm days (ENVIRON International Corporation 2009). Therefore, it seems probable that any effect of overland flow on the nearby water column by intertidal harvesting would be confined to a limited area close to the site, would not exceed that generated by natural forces, and would dissipate quickly as the tide came in. Note that this limited area potentially affected by the overland flow during harvesting is not the same as the nearby down-current area addressed by the present study. The latter was subject to the redeposition of sediments from the harvested plot after harvesting was completed and the tide came in.
No significant changes in sediment grain size, infaunal community, or various plant parameters of the eelgrass beds were detected at either study site in response to harvesting. Although results of the present study might be site specific, some general comments may be made regarding effects of geoduck harvesting on eelgrass beds at other potential culture sites. The depth limit of eelgrass distribution is largely regulated by light availability under water (Duarte 1991), suggesting that beds may not extend below a certain depth contour. For example, surveys in Puget Sound, WA, have shown that eelgrass rarely occurs deeper than the -5.5 m mean lower low-water contour (Palazzi et al. 2001). Similarly, in the present study, the lower boundary of the eelgrass bed at the subtidal CI site occurred at approximately 3.5 m below chart datum. At present, harvesters in the geoduck fishery in BC are not allowed to fish in water shallower than 3.0 m below chart datum, placing them deeper than most eelgrass beds (Fisheries and Oceans Canada 2012). Accordingly, it is likely that future subtidal geoduck culture in BC will be permitted only in areas deeper than where eelgrass beds exist. Indeed, since these beds are considered to be sensitive aquatic vegetation and critical fish habitat in Canada, they are protected from harmful alteration, disruption, and destruction. Future geoduck enhancement/culture plots are unlikely to be allowed in or near eelgrass beds. Since the major near-shore current direction typically parallels shorelines, it is expected that deposition of materials from turbid plumes and increased turbidity from subtidal geoduck harvesting would be minimal in shallower eelgrass beds which would not be subject to the direct down-current influence from harvesting. Findings of the present study at CI are consistent with this notion as sediment amounts collected in the eelgrass bed through harvesting were comparable to those during a calm sea, but much lower than those during a rough sea (just before the harvest) and winter storm at that site.
The shoreward eelgrass bed paralleled the major current direction at the intertidal site NB. Despite the seaward eelgrass bed having been located in the down-current direction, materials available for redeposition from harvesting were first carried in the opposite direction toward the nearby area as the tide came in, leaving less material available for subsequent redeposition on the seaward eelgrass bed during the ebb tide. The amounts of sediments collected in both shoreward and seaward eelgrass beds were much lower during harvesting than during windy conditions (just before the harvest), except for a few large replicate values at 1 m (shoreward) and 5 m (seaward). Therefore, as with the nearby down-current areas, the low levels of sediments caused by harvesting near the eelgrass beds would be inconsequential at both study sites when compared with natural variation. This is consistent with our findings that there were no significant changes in grain size, infaunal community, or eelgrass parameters in the eelgrass beds at either site.
This study identifies little effect of commercial geoduck-harvesting practices beyond short-lived resuspension of sediment on the two harvested plots. More work needs to be done, however, to assess how changes in habitat, size of culture plot, frequency of culture, and seasonal timing of out-planting and harvesting may alter the degree of impact on, and rate of recovery of, the marine environment.
Financial support was provided by the Aquaculture Collaborative Research and Development Program of Fisheries and Oceans Canada, the Underwater Harvesters Association, and the BC Ministry of Forests/BC Timber Sales. We thank Rick Birch and Dave English of ASL Environmental Sciences Inc. for Acoustic Doppler Current Profiler data analyses, Sandy Lipovsky of Columbia Science for infaunal identification and enumeration, Cynthia Durance for eelgrass species identification, Dr. Maureen Soon of the University of British Columbia for total nitrogen and total organic carbon analyses, and Parkes Design for the sitemap drawing. We are grateful to the following persons (in alphabetical order) who assisted in the field and/or laboratory sampling: (1) Dr. Pearce's laboratory: Kalam Azad, John Blackburn, Dan Curtis, Lyanne Curtis, Anya Dunham, Lucie Hannah, Laurie Keddy, Rob Marshall, Haley Matkin, Lindsay Orr, Leah Sauchyn, and Janis Webb; (2) Fisheries and Oceans Canada (Pacific Biological Station): Dominique Bureau, Dan Leus, and Erin Wylie; (3) undergraduate students from Vancouver Island University (Fisheries and Aquaculture Department): Carol Bob. Heather Brown, Luke Devenish, Sara Gordon, Mark Ma, Rianna Martindale, Daniel McNeill, Patrick O'Reilly, Steven Paie, Dave Poter, Kate Rolheiser, and Owen Skipper-Horton; (4) West Coast Geoduck Research Corporation: Lawrence Anderson, Bob Antifave, Mike Atkins, James Austin, Cory Carmen, Bruce Clapp, Kirk Montgomery, Evan Scoffings, Tracy Scott, Greg Sorensen, Kevin White, and Sean Williams. Peter and Susan McLellan allowed the Nanoose Bay study site to be set up on their aquaculture tenure. Scientific advice for the research project was provided by Miriam O from Fisheries and Oceans Canada. We also thank Michelle James (Underwater Harvesters Association) and Kerry Marcus (Fisheries and Oceans Canada) for valuable input on the manuscript.
Airoldi, L. 2003. The effect of sedimentation on rocky coast assemblages. Oceanogr. Mar. Biol. Annu. Rev. 41:161-236.
Alaska Department of Fish and Game. 2014. Geoduck historical harvest information, Southeast Alaska & Yakutat commercial fisheries. Accessed July 30, 2014. Available at: http://www.adfg.alaska.gov/index.cfm?adfg=commercialbyareasoutheast.dive_harvest_geoduck.
Anderson, M. J., R. N. Gorley & K. R. Clarke. 2008. PERMANOVA+ for PRIMER: guide to software and statistical methods. Plymouth. UK: PRIMER-E.
Aragon-Noriega, E. A., E. Alcantara-Razo, L. E. Calderon-Aguilera & R. Sanchez-Fourcade. 2012. Status of geoduck clam fisheries in Mexico. J. Shellfish Res. 31:733-738.
BC Ministry of Agriculture. 2010. BC seafood industry year in review, 2010. Accessed September 10, 2012. Available at: http://bcseafood.ca/PDFs/YIR-2010.pdf. 13 pp.
Benjamini. Y. & Y. Hochberg. 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soe.. B 57:289-300.
Bernard, F. R. 1983. Catalogue of the living Bivalvia of the eastern Pacific Ocean: Bering Strait to Cape Horn. Can. Spec. Publ. Fish. Aquat. Sci. 61:1-102.
Borja, A., J. G. Rodriguez, K. Black, A. Bodoy, C. Emblow, T. F. Fernandes, J. Forte, I. Karakassis, I. Muxika, T. D. Nickell, N. Papageorgiou, F. Pranovi, K. Sevastou, P. Tomassetti & D. Angel. 2009. Assessing the suitability of a range of benthic indices in the evaluation of environmental impact of fin and shellfish aquaculture located in sites across Europe. Aquaculture 293:231-240.
Breen, P. A. & T. L. Shields. 1983. Age and size structure in five populations of geoduck clams (Panope generosa) in British Columbia. Can. Tech. Rep. Fish. Aquat. Sci. 1169:1-62.
Breen, P. A., C. Gabriel & T. Tyson. 1991. Preliminary estimates of age, mortality, growth, and reproduction in the hiatellid clam Panopea zelandica in New Zealand. N. Z. J. Mar. Freshw. Res. 25:231-237.
Bureau, D., W. Hajas, N. W. Surry, C. M. Hand, G. Dovey & A. Campbell. 2002. Age, size structure and growth parameters of geoducks (Panopea abrupta, Conrad 1849) from 34 locations in British Columbia sampled between 1993 and 2000. Can. Tech. Rep. Fish. Aquat. Sci. 2413:1-84.
Cabello-Pasini, A., C. Lara-Turrent & R. C. Zimmerman. 2002. Effect of storms on photosynthesis, carbohydrate content and survival of eelgrass populations from a coastal lagoon and the adjacent open ocean. Aquat. Bol. 74:149-164.
Cain, T. A. & A. Bradbury. 1996. The effect of commercial geoduck (Panopea abrupta) fishing on Dungeness crab (Cancer magister) catch per unit effort in Hood Canal, Washington. Appendix 7 in Appendices to Final Supplemental Environment Impact Statement (S.E.I.S.) for The Puget Sound Commercial Geoduck Fishery. Washington State Department of Natural Resources and Washington State Department of Fish and Wildlife, May 23, 2001. 12 pp.
Chambers, P. A.. R. E. DeWreede, E. A. Irlandi & H. Vandermeulen. 1999. Management issues in aquatic macrophyte ecology: a Canadian perspective. Can. J. Bot. 77:471-487.
Clarke, K. R. & R. N. Gorley. 2006. PRIMER v6: user manual/tutorial. Plymouth, UK: PRIMER-E.
Constantino. R., M. B. Gaspar, F. Pereira, S. Carvalho, J. Curdia. D. Matias & C. C. Monteiro. 2009. Environmental impact of razor clam harvesting using salt in Ria Formosa lagoon (Southern Portugal) and subsequent recovery of associated benthic communities. Aquat. Conserv. Mar. Freshw. Ecosyst. 19:542-553.
Crawford, C. M., C. K. A. Macleod & I. M. Mitchell. 2003. Effects of shellfish farming on the benthic environment. Aquaculture 224:117-140.
Currie, D. R. & G. D. Parry. 1996. Effects of scallop dredging on a soft sediment community: a large scale experimental study. Mar. Ecol. Prog. Ser. 134:131-150.
Davis, J. P. 2004. Geoduck culture on intertidal beaches: procedures, expenses and anticipated income for an intermediate-size farm. Accessed July 30, 2014. Available at: http://www.ecy.wa.gov/programs/sea/shellfishcommittee/rpt/rptl_Deliverable3_Baywater.pdf. 11 pp.
Duarte, C. M. 1991. Seagrass depth limits. Aquat. Bot. 40:363-377.
Ellis, J. I. & D. C. Schneider. 1997. Evaluation of a gradient sampling design for environmental impact assessment. Environ. Monit. Assess. 48:157-172.
ENVIRON International Corporation. 2009. Programmatic biological evaluation of potential impacts from new geoduck aquaculture sites to essential fish habitat, endangered species, and forage fish in Puget Sound, Washington. 129 pp.
Environment Canada. 2012. Canadian Climate Normals 1971-2000. Accessed March 20, 2012. Available at: http://climate.weatheroffice. gc.ca/climate_normals/results_e.html?stnID=192&autofwd= 1.
Ferns, P. N., D. M. Rostron & H. Y. Siman. 2000. Effects of mechanical cockle harvesting on intertidal communities. J. Appl. Ecol. 37:464-474.
Fisheries and Oceans Canada. 2012. Integrated fisheries management plan, geoduck and horse clam, January 1 to December 31, 2011. 35 pp.
Fleece, C., D. Waller, J. Fisher, J. Vanderpham & G. Reub. 2004. Programmatic biological evaluation of potential impacts of intertidal geoduck culture facilities to endangered species and essential fish habitat. Draft biological evaluation prepared on October 27, 2004 by Entrix Inc. for Taylor Shellfish, Seattle Shellfish, and Chelsea Farms, Olympia, Washington.
Gallucci, F., P. Hutchings, P. Gribben & G. Fonseca. 2012. Habitat alteration and community-level effects of an invasive ecosystem engineer: a case study along the coast of NSW, Australia. Mar. Ecol. Prog. Ser. 449:95-108.
Goldberg, R., R. Mercaldo-Allen, J. M. Rose, P. Clark, C. Kuropat, S. Meseck & J. Pereira. 2012. Effects of hydraulic shellfish dredging on the ecology of a cultivated clam bed. Aquacult. Environ. Interact. 3:11-21.
Goodwin, L. 1978. Some effects of subtidal geoduck (Panope generosa) harvest on a small experimental plot in Hood Canal, Washington. State of Washington Department of Fisheries Progress Report 66. 21pp.
Goodwin, C. L. & B. Pease. 1987. The distribution of geoduck (Panopea abrupta) size, density and quality in relation to habitat characteristics such as geographic area, water depth, sediment type, and associated flora and fauna in Puget Sound Washington. Wash. Dep. Fish. Tech. Rep. no. 102. 44 pp.
Goodwin, C. L. & B. Pease. 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest)--Pacific geoduck clam. US Fish. Wildl. Serv. Biol. Rep. 82 (11.120). US Army Corps of Engineers, TR EL-82-4. 14 pp.
Green, R. H. 1979. Sampling Design and Statistical Methods for Environmental Biologists. New York: Wiley-Interscience. 257 pp.
Gribben, P. E. & R. G. Creese. 2005. Age, growth, and mortality of the New Zealand geoduck clam, Panopea zelandica (Bivalvia: Hiatellidae) in two north island populations. Bull. Mar. Sci. 77:119-135.
Hall, S. J., D. J. Basford & M. R. Robertson. 1990. The impact of hydraulic dredging for razor clams Ensis sp. on an infaunal community. Neth. J. Sea Res. 27:119-125.
Horowitz, A. J. & K. A. Elrick. 1987. The relation of stream sediment surface area, grain size and composition to trace element chemistry. Appl. Geochem. 2:437-451.
Kaiser, M. J., G. Broad & S. J. Hall. 2001. Disturbance of intertidal soft-sediment benthic communities by cockle hand raking. J. Sea Res. 45:119-130.
Kaiser, M. J., D. B. Edwards & B. E. Spencer. 1996. Infaunal community changes as a result of commercial clam cultivation and harvesting. Aquat. Living Resour. 9:57-63.
Meteorological Service of Canada. 2012. Accessed September 27, 2012. Available at: http://climate.weatheroffice.gc.ca/climateData/hourlydata_e. html?timeframe=l&StationID=145&Year=2012&Month=9&Day=25.
Miller, D. C., C. L. Muir & O. A. Hauser. 2002. Detrimental effects of sedimentation on marine benthos: what can be learned from natural processes and rates? Ecol. Eng. 19:211-232.
Moore, K. A., R. L. Wetzel & R. J. Orth. 1997. Seasonal pulses of turbidity and their relations to eelgrass (Zostera marina L.) survival in an estuary. J. Exp. Mar. Biol. Ecol. 215:115-134.
Morsan, E. & N. F. Ciocco. 2004. Age and growth model for the southern geoduck, Panopea abbreviata, off Puerto Lobos (Patagonia, Argentina). Fish. Res. 69:343-348.
Morsan, E., P. Zaidman, M. Ocampo-Reinaldo & N. F. Ciocco. 2010. Population structure, distribution and harvesting of southern geoduck, Panopea abbreviata. in San Matias Gulf (Patagonia, Argentina). Sci. Mar. 74:763-772.
New Zealand Fisheries. 2013. Fisheries assessment plenary May 2013: stock assessments and yield estimates. Deepwater king clam. Accessed February 3, 2014. Available at: http://fs.fish.govt.nz/Page.aspx?pk=113&dk=23301. pp. 224-227.
Ongley, E. 2006. Sediment measurements. In: Bartram, J. & R. Ballance, editors. Water quality monitoring: a practical guide to the design and implementation of freshwater quality studies and monitoring programmes. London: E & FN Spon, published on behalf of United Nations Environmental Programme and the World Health Organization.
Palazzi, D., L. Goodwin, A. Bradbury & R. Sizemore. 2001. State of Washington commercial geoduck fishery, final supplemental environmental impact statement. Olympia, WA: Washington Department of Natural Resources and Washington Department of Fish and Wildlife. 135 pp.
Piersma, T., A. Koolhaas, A. Dekinga, J. J. Beukema, R. Dekker & K. Essink. 2001. Long-term indirect effects of mechanical cockle-dredging on intertidal bivalve stocks in the Wadden Sea. J. Appl. Ecol. 38:976-990.
Pilskaln, C. H., J. H. Churchill & L. M. Mayer. 1998. Resuspension of sediment by bottom trawling in the Gulf of Maine and potential geochemical consequences. Conserv. Biol. 12:1223-1229.
Price, J. 2011. Quantifying the ecological impacts of geoduck (Panopea generosa) aquaculture harvest practices on benthic infauna. MSc thesis, University of Washington. 136 pp.
Puget Sound Water Quality Authority. 1987. Recommended protocols for sampling and analyzing subtidal benthic macroinvertebrate assemblages in Puget Sound. Report for the US Environmental Protection Agency, January 1987, iv + 34 pp.
Rossi, F., R. M. Forster, F. Montserrat, M. Ponti, A. Terlizzi, T. Ysebaert & J. J. Middelburg. 2007. Human trampling as short-term disturbance on intertidal mudflats: effects on macrofauna biodiversity and population dynamics of bivalves. Mar. Biol 151:2077-2090.
Sauchyn, L., C. M. Pearce, J. Blackburn, L. Keddy & S. Williams. 2013. Assessing potential benthic habitat impacts of small-scale, intertidal aquaculture of the geoduck clam (Panopea generosa). Can. Sci. Advis. Sec. Res. Doc. 2013/001. vi + 34 pp.
Short, F. T. & S. Wyllie-Echeverria. 1996. Natural and human-induced disturbance of seagrasses. Environ. Conserv. 23:17-27.
Short, K. S. & R. Walton. 1992. The transport and fate of suspended sediment plumes associated with commercial geoduck harvesting--final report. Prepared for the State of Washington Department of Natural Resources. 48 pp.
Stewart-Oaten, A. & J. R. Bence. 2001. Temporal and spatial variation in environmental impact assessment. Ecol. Monogr. 71:305-339.
Stewart-Oaten, A., J. R. Bence & C. W. Osenberg. 1992. Assessing effects of unreplicated perturbations: no simple solutions. Ecology 73:1396-1404.
Stewart-Oaten, A., W. M. Murdoch & K. R. Parker. 1986. Environmental impact assessment: "pseudoreplication" in time? Ecology 67:929-940.
Straus, K. M., L. M. Crosson & B. Vadopalas. 2008. Effects of geoduck aquaculture on the environment: a synthesis of current knowledge. Washington Sea Grant. 64 pp.
Tam, N. F. Y. & Y. S. Wong. 2000. Spatial variation of heavy metals in surface sediments of Hong Kong mangrove swamps. Environ. Pollut. 110:195-205.
Tamaki, H.. M. Tokuoka, W. Nishijima, T. Terawaki & M. Okada. 2002. Deterioration of eelgrass, Zostera marina L., meadows by water pollution in Seto Inland Sea, Japan. Mar. Pollut. Bull. 44:1253-1258.
Tengberg, A., E. Almroth & P. Hall. 2003. Resuspension and its effects on organic carbon recycling and nutrient exchange in coastal sediments: in situ measurements using new experimental technology. J. Exp. Mar. Biol. Ecol. 285:119-142.
Tuck, I. D., N. Bailey, M. Harding, G. Sangster, T. Howell, N. Graham & M. Breen. 2000. The impact of water jet dredging for razor clams, Ensis spp., in a shallow sandy subtidal environment. J. Sea Res. 43:65-81.
Underwood, A. J. 1997. Experiments in Ecology. Cambridge, UK: Cambridge University Press. 504 pp.
Vermaat, J. E., N. S. R. Agawin, M. D. Fortes, J. S. Uri, C. M. Duarte, N. Marba, S. Enriquez & W. van Vierssen. 1997. The capacity of seagrasses to survive increased turbidity and siltation: the significance of growth form and light use. Ambio 26:499-504.
Washington Department of Fish and Wildlife. 2012. Shellfish aquaculture and harvest production and values, 1970-2011. Washington Department of Fish and Wildlife, Olympia, WA. Unpublished data received August 2012.
Wentworth, C. 1922. A scale of grade and class term for classic sediments. J. Geol. 30:377-392.
Wildish, D. J., H. M. Akagi, N. Hamilton & B. T. Hargrave. 1999. A recommended method for monitoring sediments to detect organic enrichment from mariculture in the Bay of Fundy. Can. Tech. Rep. Fish. Aquat. Sci. 2286:1-31.
Zhang, Z. & C. Hand. 2006. Recruitment patterns and precautionary exploitation rates for geoduck (Panopea abrupta) populations in British Columbia. J. Shellfish Res. 25:445-453.
WENSHAN LIU, (1) CHRISTOPHER M. PEARCE (1) * AND GRANT DOVEY (2)
(1) Fisheries and Oceans Canada, Pacific Biological Station, Nanaimo, British Columbia, Canada V9T 6N7;
(2) West Coast Geoduck Research Corporation, Lady smith, British Columbia, Canada V9G 1T6
* Corresponding author. E-mail: Chris.Pearce@dfo-mpo.gc.ca
TABLE 1. Sampling and harvest schedules at CI and NB. CI Date Time point October 9-10, 2008 -12 February 12-13, 2009 -8 July 6-7, 2009 -3 October 2-3, 2009 -0 October 4-5, 2009 Harvest October 6-7, 2009 +0 February 7-8, 2010 +4 May 4-5, 2010 +7 October 5/27, 2010 +12 NB Date Time point October 16, 2008 -0 October 18, 2008 Harvest October 20, 2008 +0 January 7-8, 2009 +3 March 31-April 1, 2009 +6 November 3, 2009 +13 April 29-30, 2010 +18 October 10, 2010 +24 -, Months before harvest; +, months after harvest; -0, immediately prior to harvest; +0, immediately post harvest. TABLE 2. Results of PERMANOVA of effects of sampling distance, time, and interaction on various sets of variables from CI. Variables Sources df F P Harvest plot Sediment grain Distance 5 5.79 <0.001 and size Time 7 3.79 <0.001 nearby area Distance X time 35 0.73 0.975 Error 192 Percent organics, Distance 5 2.14 0.033 total nitrogen, Time 7 17.31 <0.001 and total Distance X time 35 0.99 0.514 carbon Error 192 Sulfide content Distance 5 1.52 0.137 and redox Time 7 10.78 <0.001 potential 2 cm Distance X time 35 1.64 0.003# Error 192 Sulfide content Distance 5 1.60 1.099 and redox Time 7 6.73 <0.001 potential 6 cm Distance X time 35 1.21 0.130 Error 192 Infaunal Distance 5 0.98 0.434 community Time 7 36.68 <0.001 structure Distance X time 35 0.79 0.802 Error 192 Eelgrass bed Sediment grain Distance 3 13.91 <0.001 size Time 7 7.33 <0.001 Distance x time 21 0.71 0.932 Error 128 Infaunal Distance 3 6.27 <0.001 community Time 7 19.89 <0.001 structure Distance X time 21 1.01 0.461 Error 128 Eelgrass Distance 3 3.22 0.005 parameters Time 8 14.15 0.001 Distance x time 24 0.63 0.973 Error 144 Significance (P < 0.005, adjusted) in interaction is indicated with bold. Note: bold indicated with #. TABLE 3. Results of PERMANOVA of effects of sampling distance, time, and interaction on various sets of variables from NB. Variables Sources df F P Harvest Sediment grain Distance 6 11.36 <0.001 plot and size Time 6 5.29 <0.001 nearby Distance X time 36 1.79 <0.001# area Error 196 Percent organics, Distance 6 5.36 <0.001 total nitrogen, Time 6 6.78 <0.001 and total Distance X time 36 1.03 0.411 carbon Error 196 Sulfide content Distance 6 2.23 0.013 and redox Time 6 10.31 <0.001 potential 2 cm Distance X time 36 1.219 0.141 Error 196 Sulfide content Distance 6 2.13 0.017 and redox Time 6 2.19 <0.001 potential 6 cm Distance X time 36 1.20 0.154 Error 196 Infaunal Distance 6 3.43 0.002 community Time 6 28.92 <0.001 structure Distance X time 36 1.59 0.019 Error 196 Eelgrass Sediment Distance 2 0.46 0.825 bed grain size Time 6 11.58 <0.001 seaward Distance X time 12 0.660 0.921 Error 84 Infaunal Distance 2 2.28 0.100 community Time 6 18.31 <0.001 structure Distance X time 12 1.51 0.122 Error 84 Eelgrass Distance 2 2.44 0.098 parameters Time 8 9.19 <0.001 Distance X time 16 0.42 0.976 Error 108 Eelgrass Sediment Distance 2 24.47 <0.001 bed grain size Time 6 4.80 <0.001 shoreward Distance X time 12 0.47 0.995 Error 84 Infaunal Distance 2 3.94 0.018 community Time 6 9.08 <0.001 structure Distance x time 12 1.71 0.073 Error 84 Eelgrass Distance 2 3.16 0.047 parameters Time 8 13.76 <0.001 Distance x time 16 1.02 0.438 Error 108 Significance (P < 0.005, adjusted) in interaction is indicated with bold. Note: bold indicated with #. TABLE 4. Summary of reported intensities of harvesting of subtidal and intertidal geoduck clams (Panopea generosa) in WA and BC, Canada. Total duration Actual Harvest plot when harvest harvest Number of harvest size ([m.sup.2]) occurred (days) days holes ([m.sup.2]) 90 29 5 4.3 30 6 - 8.4 60 1 1 Swath harvest 2,500-4,500 2-5 (mo) - -* 6,000 2 2 0.26 450 1 1 9 Harvest plot Type of size ([m.sup.2]) harvest Reference 90 S, F Goodwin (1978) 30 S. F Breen and Shields (1983) 60 I, A Sauchyn et al. (2013) 2,500-4,500 I, A Price (2011) 6,000 S, A/F Present study 450 I, A Present study -, Not specified in the study; I, intertidal plot; S, subtidal plot; F, fisheries plot; A, aquaculture plot. * The number of harvest holes per unit area is expected to be relatively higher on these aquaculture plots. Note that an estimation of 2.5 holes/[m.sup.2] is assumed for high-density commercial geoduck fisheries beds in WA (Palazzi et al. 2001).
|Printer friendly Cite/link Email Feedback|
|Author:||Liu, Wenshan; Pearce, Christopher M.; Dovey, Grant|
|Publication:||Journal of Shellfish Research|
|Date:||Dec 1, 2015|
|Previous Article:||First description of symbionts, parasites, and diseases of the Pacific geoduck Panopea generosa from the Pacific Coast of Baja California, Mexico.|
|Next Article:||Annual reproductive cycle and condition index of the New Zealand surf clam Mactra murchisoni (Deshayes, 1854) (Bivalvia: Mactridae).|