Sediment Properties and Benthic Fauna Associated with Stock Enhancement and Farming of Marine Bivalve Populations in Xiangshan Bay, China

,


Introduction
Shellfsh play an important ecological role in coastal ecosystems.Many studies have demonstrated that coastal flterfeeding bivalve populations, such as oyster reefs, mussel beds, and clams, can regulate coastal nutrient fuxes, sedimentation, and primary productivity [1,2].By removing particulate suspended solids from the water, bivalves can decrease phytoplankton abundance and/or chlorophyll concentration, and enhance the water clarity [3].Because bivalves can assimilate land-derived nitrogen and phosphorus from the ambient environment, they are also used to remediate water eutrophication [4][5][6][7].Moreover, through the calcifcation process, flter-feeding bivalves can convert bicarbonate into carbonate and thus play an important role in the global carbon cycle [8].
Filter-feeding bivalve farming is growing rapidly worldwide [9].China generally leads the world in aquaculture production of shellfsh but it has continued to increase, reaching 1.46 million metric tons of annual production in 2019 [10].Bivalve farming usually occurs in coastal bays and lagoons due to ease of access and high primary production, which results in faster growth and increased production [11,12].Bivalve shellfsh aquaculture can be viewed as a disturbance that modifes the coastal ecosystem [13], but the impact of this disturbance can be positive or negative and difers depending on the system.Concentrated bivalve biomass with high fltration rates can generate large amounts of fne particulate, feces, and pseudofeces that are enriched with organic matter, resulting in intense biodeposition on sediments [12,14,15].Large accumulations of such organic matter beneath bivalve farms can change the benthic environment, including oxygen depletion, sulfde accumulation, and nitrifcation inhibition [12,14,16].Due to their sedentary habits and weak mobility, benthic faunas are sensitive indicators of environmental change [17,18].Te accumulation of organic-rich biodeposits may infuence the abundance, biomass, and diversity of benthic fauna communities [19][20][21].Sediment properties afect the distribution of heavy metals.It has been found that the heavy metal content in sediments is negatively correlated with the sediment particle size and positively correlated with the organic matter content [22,23].Due to the higher specifc surface area and organic matter content, the large amounts of fne particulate generated by flterfeeding shellfsh might absorb more heavy metals.Conversely, some studies have found that shellfsh farming has little impact on sediment properties or benthic communities [18,24,25].
Due to anthropogenic pressures such as overfshing, coast reclamation works, and habitat loss, the natural resources of shellfsh in China have been severely depleted, resulting in decreased annual harvests of wild shellfsh [10,26].Releasing cultured organisms into natural coastal systems has been shown to be an efective way to enhance, conserve, or restore fsheries [27].In China, to solve the problem of natural resource depletion, corresponding measures, including summer fshing moratorium and seed enhancement, have also been adopted [28].From 2006 to 2015, China invested approximately 0.87 billion dollars in the release of 300 billion various aquatic fries and shellfsh juveniles [29].Numerous studies have acknowledged the benefcial role of shellfsh stock enhancement [30][31][32].Currently, the assessment of the efectiveness of shellfsh stock enhancement is focused on productivity, and social and economic benefts [33], while few studies have investigated its environmental efects.
Xiangshan Bay is a long and narrow bay in northern Zhejiang Province, China, around 70 km long and 4 km wide.At its center, the average depth is 7 m, and there are large areas of mudfat [34].Xiangshan Bay is a traditional fshing area and provides a habitat for many marine fsh and shellfsh species [35].Te ark shell, Scapharca subcrenata, was one of the most widely distributed native shellfsh species in Xiangshan Bay, China.According to the survey results in 1995, the distribution area of S. subcrenata in Xiangshan Bay is about 4,433.3hectares, mainly in Tiegang inner bay and Huangdungang inner bay, with an average density of 20.95 g/m 2 (2.095 individuals/m 2 ) and an estimated biomass of 927 tons [36].However, the wild fshery resources, which this species relies upon, have declined due to overfshing and land reclamation.In 2011, the highest biomass (ash-free dry mass) of S. subcrenata in Tiegang inner bay was only 1.51 g/m 2 [37].In response, the local government has initiated stock enhancement of S. subcrenata through the release of hatchery juveniles.Xiangshan Bay also serves as an important shellfsh mariculture zone, where the suspended culture of oyster (primarily Crassostrea sikamea and Ostrea plicatula) has been conducted since the 1990s.Te area farmed has recently increased from 3.7 in 2006 to 7.5 km 2 in 2016 [18].A previous study showed that raft-farmed oysters in Xiangshan Bay signifcantly reduced phytoplankton abundance and chlorophyll a [38], implying a strong biodeposition.
Xiangshan Bay is an ideal place to study and compare the ecological efects of shellfsh stock enhancement and artifcial culturing.In this study, two species of bivalve molluscs, S. subcrenata and C. sikamea, which inhabits in the benthic zone and are farmed on rafts at the surface of the water in Xiangshan Bay, respectively, were selected.Te potential efects of large-scale marine bivalve enhancement and suspended oyster farming on benthic fauna and substrate properties were studied.

Ethical Statement.
Tis study was performed in accordance with the standard operating procedures for the use of experimental animals of the East China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences.Tese feld studies did not involve endangered or protected species.

Experimental Area.
Te survey was conducted in Tiegang inner bay, located at the head of Xiangshan Bay, China (Figure 1).Samples were taken from an area where ark shell stock enhancement (the ark shell enhancement area, AEA) occurred which was about 1.0 km 2 , with an average water depth of 3-4 m.Te hatchery juveniles (average shell length � 3.71 ± 0.70 mm) were cultivated in a local nursery farm and then released in the study area on December 5, 2018, at densities ranging from 50 to 100 individuals/m 2 .An oyster (C.sikamea) farm located in the northern part of Tiegang inner bay was selected for sampling (the oyster farming area, OFA), which was approximately 1.5 km 2 with an average water depth of 6-7 m.Suspended (long-line) oyster culture has been conducted in Tiegang inner bay since the 1990s [18].Te oyster farm has been operated for more than 10 years.Oyster seeds were collected using bicycle tires as a metamorphosis attachment substrate in an estuary close to the farm, from July to August each year.When the oysters attached to the bicycle tires grew to around 0.5 cm in shell length, the tires were transported to the farm and hung on the underside of the rafts situated at the surface of the water for a 2-year farming period.Each raft contained 300-350 tires.Te distance between the rafts was 5 m.Oysters were farmed throughout the year.An area with no ark shell and oyster farming was selected as the control area (CA).Actually, both the control area and the enhancement area were formerly S. subcrenata habitat, but the resources have declined rapidly in recent years due to overfshing.

Sample Collection.
In each treatment area, 4 sampling sites were selected.Te sampling sites were evenly distributed, with an average distance between each station of about 200 m.At each sampling site, four benthic samples were collected with a 0.1 m 2 Van Veen grab sampler on October 20, 2019.For benthic fauna communities, the sediment 2 Aquaculture Research samples were sieved through a 500 μm mesh, and all organisms retained were preserved in a 5% formaldehyde solution.Ark shells in each sample were counted, shell length was measured with electronic Vernier calipers, and wet weight was assessed using an electronic scale.After 319 days of the enhancement, the mean shell length of the ark shell increased from 3.71 ± 0.70 mm to 24.01 ± 2.98 mm, and the wet weight increased from 0.007 g to 3.02 ± 0.57 g.Te densities of ark shells from the four sampling sites were 16.60, 89.03, 84.73, and 9.27 individuals/m 2 .No ark shell was found in the oyster farming area.In the control area, the average concentration of the ark shell was 0.29 individuals/ m 2 .Benthic fauna was sorted and identifed to the lowest possible taxonomic order, generally species, and enumerated.Te nomenclature was referenced using the World Register of Marine Species (WoRMS: https://www.marinespecies.org).For physical and chemical property analysis, the central part of each sediment sample was taken and immediately capped and frozen using carbon dioxide ice.All samples were then transferred to the laboratory in iceboxes and stored at −20 °C.

Sample Analysis. Acid volatile sulfde (AVS) contents
were assayed according to the method of Allen et al. [39].
Te sediment sample (about 3 g) was placed into a reaction fask and sparged with continuous pure N 2 fow and was then acidifed for 20 minutes by adding 20 mL of 6 M HCl at room temperature.Te generated H 2 S was collected in a NaOH solution, and the concentration of AVS was determined spectrophotometrically at a wavelength of 670 nm.All reagents used during the assay were of Merck analytical grade, and all glass vessels were acid-cleaned before use.
For heavy metal analysis, sediment samples were digested for heavy metal analysis using HNO 3 and HClO 4 .According to Sun et al. [40]; Cu, Pb, Zn, and Cd were the main heavy metal contaminants in the sediments in Xiangshan Bay, and concentrations of Cu, Pb, Zn, and Cd were determined using atomic absorption spectroscopy (AAS SOLAAR M6, Termo Electron Corp., Watham, MA, USA) according to the method of China National Standard GB17378.5-2007, which used a fame atomic absorption method for the analysis of Cu, Zn, and Pb.Te graphite furnace atomic absorption method was used for the analysis of Cd concentrations.A duplicate sample analysis was run for 10% of the total samples, and national standard reference materials (GBW07314) were used to ensure the precision of the analysis.When conducting the analysis, the relative standard deviation of duplicate samples was less than 5%, and the recovery rates of the standard reference were around 90-110%.
Sediment samples were treated with a 0.5 M sodium hexametaphosphate solution for 24 h, and the particle size distribution was measured using a particle size analyzer (Malvern, Britain).Te samples were dispersed and homogenized using an ultrasonic vibrator before analysis.Grain size parameters, including median particle diameter (MD), were calculated according to the matrix method [41].Te moisture content of the sediments was determined by calculating the weight loss after drying at 105 °C for 24 h.For total organic carbon (TOC) analysis, the inorganic carbon (mainly in the form of carbonate) was removed with 10% diluted hydrochloric acid.Te samples were then washed repeatedly with deionized water, freeze-dried, and carefully crimp-sealed in tin capsules.Carbon analysis was conducted using an elemental analyzer (TOC-5000A, Shimadzu, Kyoto, Japan) with a measuring accuracy of 0.1%.

Statistical Analysis.
Te seawater environmental parameters, sediment moisture content, TOC, particle size, sulfde, and heavy metal data were analyzed using ANOVA (SPSS ® , v20.0).Diferences in mean values were assessed using a post hoc least-signifcant diference (LSD) test at the 5% level of signifcance (p < 0.05).Percentage data were square arcsine-transformed prior to analysis to meet the assumptions of equal variance.
Benthic fauna community data were analyzed using multivariate statistical analysis, and the Shannon-Wiener diversity index was calculated using PRIMER 6.0 [42].A hierarchical cluster analysis was conducted, and a similarity matrix was constructed based on the Bray-Curtis index on log-transformed (ln (x + 1)) abundance data.Diferences in species abundance between the ark shell enhancement, oyster farming area, and control area were examined using the analysis of similarity (ANOSIM).If global R-statistics were statistically signifcant, ANOSIM was then used to examine paired diferences between the treatment areas.

Results
3.1.Sediment Properties.Sediment moisture content was signifcantly higher in the oyster farming area (OFA) than in the ark shell enhancement area (AEA) (p < 0.05, F � 10.887, n � 11) but neither difered from the control area (CA) (Figure 2).Total organic carbon (TOC) was signifcantly lower in the AEA than that in the CA (p < 0.05, F � 3.307, n � 11) but there was no diference between that in OFA and CA.No signifcant diference was found between the mean grain size (MD) present in each of the three areas, though the mean MD value in the OFA was the lowest.Acid volatile sulfde (AVS) contents in the sediments were >10 times higher in the OFA than in the AEA and CA (p < 0.01, F � 37.435, n � 11).Te oyster farming area also had signifcantly higher Cu, Zn, Pb, and Cd concentrations, but no signifcant diference in these metal concentrations was observed between the AEA and CA (p < 0.01) (Figure 3).

Benthic Fauna Communities.
In this study, 16 faunal taxa were identifed.Te main taxonomic groups were as follows: Annelida (8 taxa), Echinodermata (3 taxa), Mollusc (2 taxa), and others (3 taxa) (Table 1).Te dominant taxa were diferent across the three treatment areas.Echinodermata, crustaceans, and molluscs were collected in the ark shell enhancement and control areas, but not in the oyster farming area.Nemerteans and the annelids Aglaophamus dibranchis and Sternaspisscutata were sampled only in the OFA.Ruditapes philippinarum was the most abundant species, followed by Glycera onomichiensis and Musculus senhousia.No signifcant diference was found between species richness in the diferent treatment areas.Te densities of benthic fauna in the AEA and CA were signifcantly higher than those in the OFA (p < 0.01, F � 12.298, n � 11), resulting in no signifcant diference in the Shannon-Wiener diversity index amongst treatment areas.CA is the control area; AEA is the ark shell enhancement area; and OFA is the oyster farming area.

Aquaculture Research
Cluster analysis of the community data showed that the average similarity between the treatment areas was less than 5% (Figure 4).Communities in the CA and AEA were generally similar but distinct from those in the OFA.
ANOSIM pairwise comparisons demonstrated similar groupings with signifcant diferences between the OFA and both that in the CA and AEA treatments (ANOSIM: R � 0.64, p < 0.05) (Table 2).

Discussion
Te present study indicates that the ark shell enhancement had less impact on the benthic environment in Xiangshan Bay, while the raft-farmed oysters had signifcant efects on sediment properties, heavy metal accumulation, and benthic fauna communities after more than ten years of farming.

Efects of Ark Shell Enhancement and Oyster Farming on
Sediment Properties.In this study, sediments in the oyster farming area studied here had a signifcantly higher moisture content, acid volatile sulfde, and heavy metal concentrations, while the physical and chemical properties of sediments in the ark shell enhancement area and control area were similar.
Moisture content was lower in the oyster farming area, a sediment characteristic that has been reported to be correlated with other components in sediments, including organic matter, sediment-water permeability, and grain size and is usually inversely proportional to the grain size [43].Biologically mediated sedimentation processes can enhance the deposition of fne sediments in estuaries and coastal areas [44].Suspension-feeding bivalves can promote the sedimentation of particles via biodeposition [45].Te hanging ropes and rafts of shellfsh long-line culture systems can reduce the water current velocity [46], which can also promote the process of deposition.Te deposit rate in mussel farming areas were around 2-3 times higher than that in areas without mussels, and the deposits in shellfsh farming areas usually displayed a fner structure, lower density, and higher moisture content [47][48][49].Tis may explain the lower MD value and signifcantly higher moisture content in the oyster farming area in the present study.
Organic carbon has been used as an indicator of organic matter enrichment in sediments.Increased suspensionfeeding bivalve production can result in a proportional increase in organic matter biodeposition [16,50].In the River Exe estuary in England, oyster (C.gigas) farming rafts signifcantly reduced the water currents, which doubled the sedimentation rate and increased the organic content of the sediments [51].No signifcant diference was found in TOC between the treatment areas in this study, suggesting that the organic matter in biodeposits was not retained in the sediments.Researchers examining sediments at another oyster farm in Xiangshan Bay also found that TOC in sediments was not signifcantly diferent between treatment areas [18].In South St. Simon Bay, Canada, the organic content in sediments collected from an oyster farming area was not signifcantly higher than the reference area, although the oysters may contribute to the high deposition rate [52].Crawford et al. [53] found that there was no signifcant diference in the organic carbon content between sites inside and outside of oyster farms.Highly hydrodynamic environments can reduce the impact on benthic sediments derived from shellfsh culture [54][55][56].In this study, the biodeposits in the ark shell enhancement area might be washed away by tidal currents, while in the oyster farming area, biodeposits may be lost due to organic matter degradation or other mineralization processes [57].Generally, organic matter in marine sediments is transformed into dissolved inorganic carbon (DIC) with the absence of microbial communities, which can then be released to the upper water layer, or can form carbonate by combining with calcium and magnesium ions [58,59].Te mineralization processes include oxidation, denitrifcation, Mn (IV)-oxide reduction, Fe (III)-oxide reduction, and sulfate reduction [57].In this study, the average water depth in the oyster farming area was 6-7 m, increasing to 10 m at high tide.In an anaerobic environment, the accumulation of organic matter can result in sulfate reduction and increase the level of sulfde [47].Tis could explain the signifcantly higher AVS values found in this study in the oyster farming area, suggesting that biodeposition from farming oysters formed a large component of organic enrichment in sediments.Yan et al. [60] also reported increased sediment AVS in oyster (C.plicatula) farms in a diferent area of Xiangshan Bay, China.Sulfate reduction rates at a long-line Pacifc oysters (C.gigas) farm in South Korea were 2.4-5.2 times higher than those at the control site [12].In this study, the Cu, Zn, Pb, and Cd concentrations in the oyster farming area were signifcantly higher than those in the other two treatment areas.Heavy metals are highly persistent and toxic to humans due to the potential for bioaccumulation throughout the food chain.Many molluscs have been employed as contamination indicators due to their ability to accumulate heavy metals [61,62].Te common mussel Mytilus edulis has been found to have heavy metal concentrations from 10 3 to 10 6 times higher than the concentrations in the surrounding water [63].Oysters have also been deployed as biomonitors and shown to uptake a variety of heavy metals [61,64,65].Adiyiah et al. [66] reported that the sediment with the fnest grain size (<45 μm) had the highest concentrations of heavy metals due to having a larger surface area and higher adsorption capacity.Tis might be one of the reasons for the signifcantly higher Cu, Zn, Pb, and Cd concentrations in the oyster farming area in this study.Alternatively, oyster farming may accelerate heavy metal accumulation in sediments through biodeposition.In Hailing Bay, China, concentrations of Cr, Ni, Cu, Zn, As, Pb, and Cd, in the surface sediments from aquafarming areas, were signifcantly higher than those from nonaquafarming areas [67].In this study, we found no signifcant diference in heavy metals between the ark shell enhancement area and the control area, suggesting that the enhancement has a minimal impact on the substrate environment.

Efects of Ark Shell Enhancement and Oyster Farming on
Benthic Fauna Communities.Benthic fauna has been widely used to assess the impact of anthropogenic activities on marine environments such as environmental pollution and mariculture in the marine environment due to their environmental sensitivity, sedentary habits, and ease of access [18,21,68].In this study, the flter-feeding bivalve R. philippinarum was the most abundant species found in  4: Cluster analysis of benthic fauna abundance in diferent treatment areas using the Bray-Curtis similarity index.CA is the control area; AEA is the ark shell enhancement area; and OFA is the oyster farming area.6 Aquaculture Research the ark shell enhancement area and control area, and the cluster analysis showed that the benthic communities in both areas were similar.Tis suggests that the enhancement of ark shell has less impact than oyster farming on the benthic community.Similarly, Mantovani et al. [24] reported that in the Po River Delta, Italy, the presence and density of cultivated Manila clams (Tapes philippinarum) had little impact on faunal community abundance and functional group composition.Open-sea mussel culture in the Western Adriatic Sea appears to have few detrimental efects on zoobenthic communities [25].However, the clear community separation between the oyster farming area and the other two treatment areas in this study indicates that oyster farming may have a larger impact on the benthic community.Tis might be demonstrated from another study where macrobenthos species richness and abundance beneath an oyster farm in Xiangshan Bay increased signifcantly 3 years after oyster farming ceased, and more than 75% of dominant taxa were re-established [15].Costa and Nalesso [69] revealed that the sediment microbenthic community in long-line mussel farms in Anchieta, Southeast Brazil, showed signifcantly higher diversity and richness.Dubois et al. [70] found that in the Bay of Veys in France, oyster culture structures had minor efects on macrofauna density, but had profound efects on the composition of microbenthic assemblages.Tey also found that suspension feeders were not found beneath oyster tables, which is consistent with the fndings of this study.Besides the temporal cumulative efects of farming over 10 years, the signifcant change in benthic communities could be due to the reduction of current velocity and accumulation of organic matter via biodeposition.Surprisingly, Liao et al. [18] reported that in another oyster (O.plicatula) farm in Xiangshan Bay, the oyster culture had little impact on the microbenthic community, contrary to our results, suggesting that the impact of oyster culture on benthic fauna may difer between farms, and could be regulated by a variety of factors, such as sediment type, hydrodynamics, water depth, or the history and density of cultured oysters.

Conclusions
In conclusion, this study indicates that enhancement of ark shells had less impact on the benthic environment than oyster farming, which had profound efects on the sediment structure, geochemical processes, heavy metal accumulation, and benthic fauna communities in Xiangshan Bay, China.Te diferences observed between the impacts of ark shell enhancement and oyster farming could be attributed to the slowing of water fow caused by oyster rafts, which in turn promotes the sedimentation of particles and an increase in the organic matter mineralization rate in sediments.Tis study presents useful information for monitoring the environmental impact of clam enhancement and oyster farming to ensure the sustainable development of shellfsh culture and resource restoration in the future.

Figure 1 :
Figure 1: Map showing the study site in Xiangshan Bay, China.CA is the control area; AEA is the ark shell enhancement area; OFA is the oyster farming area.

Figure 2 :
Figure 2: Properties of sediments in diferent treatment areas.Note.Bars denoted with diferent letters are statistically diferent (p < 0.05).CA is the control area; AEA is the ark shell enhancement area; and OFA is the oyster farming area.

Figure 3 :
Figure 3: Heavy metal concentrations in diferent treatment areas.Note.Bars denoted with diferent letters are statistically diferent (p < 0.05).CA is the control area; AEA is the ark shell enhancement area; and OFA is the oyster farming area.

Table 1 :
Mean abundance of species of benthic fauna in diferent treatment areas (individuals/m 2 ).