Interactive effects of oyster and seaweed on seawater dissolved inorganic carbon systems : implications for integrated multi-trophic aquaculture

We examined the separate effect of Portuguese oyster Crassostrea angulata and the interactive effects of oyster and red seaweed Gracilaria lemaneiformis on seawater dissolved in organic carbon (DIC) systems and the air−sea CO2 flux (FCO2) in Daya Bay, southern China. Re spiration and calcification rates of oysters were measured and the effects of oyster aquaculture on marine DIC systems were evaluated. The interactive effects on seawater DIC and air−sea FCO2 were examined using mesocosms containing oyster and seaweed assemblages. Results showed populations of C. angulata cultured in Daya Bay sequestered ca. 258 g C m−2 yr−1 for shell formation, whereas the CO2 released due to respiration and calcification was 349 and 153 g C m−2 yr−1, respectively. This indicates that oyster cultivation in Daya Bay is a CO2 generator, favoring the escape of CO2 into the atmosphere. DIC, HCO3 and CO2 concentrations and the partial pressure of CO2 in oyster−seaweed co-cultured mesocosms were significantly lower than the oyster monoculture mesocosm. These results indicated that G. lemaneiformis effectively absorbs the CO2 released by oysters. The negative values of air−sea FCO2 in the co-culture groups represent a CO2 sink from the atmosphere to the sea. These results demonstrated that there could be an interspecies mutual benefit for both C. angulata and G. lemaneiformis in the integrated culture system. Considering that photosynthesis of seaweed is carbon limited, we suggest that the 2 species are co-cultured at a ratio of ca. 4:1 (based on fresh weight) for efficient utilization of DIC in seawater by G. lemaneiformis, and further to increase the ocean CO2 sink.


INTRODUCTION
Calcification of aquatic animals such as shellfish (e.g.oyster, scallop, and clam) is also a source of CO 2 (Chauvaud et al. 2003, Martin et al. 2006, Mistri & Munari 2013, Munari et al. 2013, Jiang et al. 2014).Shellfish utilize carbon in 2 ways.First, they consume organic carbon to sustain their growth and meta -bolism, following the reaction CH 2 O + O 2 → CO 2 + H 2 O. Second, they use HCO 3 − from seawater to generate CaCO 3 shells, based on the reaction Ca 2+ + 2HCO 3 − ↔ CaCO 3 + CO 2 + H 2 O.These 2 processes both lead to net CO 2 production in ocean waters.Second, shellfish secrete calcium carbonate (CaCO 3 ) to form their skeletal material.This process acts as a marine biological pump by removing CO 2 from circu-lation and storing carbon in the ocean (Lerman & Mackenzie 2005).In fact, the ratio of released CO 2 /precipitated CaCO 3 is largely dependent on the buffering capacity of the surrounding seawater, such that in some marine ecosystems, the ratio could be ca.0.6 (Frankignoulle et al. 1995).It is reasonable that the buffering capacity of seawater might vary significantly among different waters with variations in pH, alkalinity, salinity, and temperature (Millero 1995, Lerman & Mackenzie 2005, Dickson 2010, Mackenzie & Andersson 2013) Seaweeds (e.g.Saccharina, Gracilaria) are intricately involved as primary producers in coastal ecosystems.They assimilate inorganic carbon either via diffusion (for CO 2 ), or active uptake of HCO 3 − using carbon-concentration mechanisms.During photosynthesis, these mechanisms result in an increase in seawater pH and a drop in seawater CO 2 partial pressure (pCO 2 ) (Han et al. 2013).Therefore, seaweed could induce a significant shift in seawater dissolved inorganic carbon (DIC) systems according to the following formula: . Therefore, seaweed may exert a significant impact on the DIC buffering capacity of seawater.
As mentioned above, both shellfish and seaweed can change the seawater DIC system and the buffering capacity.One implication is a complex interspecies interaction between shellfish and seaweed in co-cultured systems.For example, CO 2 or HCO 3 − can become a major limiting factor affecting the photosynthetic rates and aquaculture production of seaweed, particularly when they are grown under conditions of high biomass densities and reduced seawater motion (Zou et al. 2004).Likewise, the alteration in DIC speciation can cause responses in calcifying organisms (e.g.oyster), thereby potentially affecting their growth and physiological functions (Ho & Carpenter 2017, Scanes et al. 2017).Thus, the interaction between shellfish and seaweed and their combined effect on DIC partitioning and cycling still needs to be investigated using an ecosystem approach.Similarly, few studies have been conducted to elucidate the influence of integrated aquaculture of shellfish and seaweed on variations in DIC systems, as well as the air−sea CO 2 flux.Furthermore, the optimum culture ratio in co-culture systems for obtaining the largest CO 2 sink is not known.Shellfish and seaweed mariculture in the coastal waters of China has been growing rapidly over the past 3 decades; they are by far the largest and most well-known aquaculture industries in the world, with an annual production of ca.13.6 × 10 6 t and 2.1 × 10 6 t, accounting for ca.72.4% and 11.1% of the total mariculture production in China, respectively (China Bureau of Fisheries 2016).In most coastal waters, shellfish and seaweed are co-cultured, using suspended longlines as the main cultivation method.In fact, they usually dominate an entire bay, such as in Sanggou Bay (Fang et al. 2016) and in Daya Bay (Yu et al. 2014) where this study was conducted.
In the present study, we conducted an in situ mesocosm experiment to measure the calcification and respiration rates of the Portuguese oyster Crassostrea angulata to evaluate its effect on marine DIC systems.In addition, the role of the red seaweed Gracilaria lemaneiformis was assayed for elimination of CO 2 in seawater.Subsequently, we investigated the impact of co-culture interactions on DIC systems and the air−sea CO 2 flux using different ratios.The results from this study will be useful in evaluating the effects of large-scale coastal aquaculture of oysters and seaweed on the marine CO 2 budget with the hope of finding methods of carbon removal from coastal waters.

Study site
Daya Bay, located in Guangdong Province, southern China, is a 600 km 2 semi-enclosed embayment in the northeast South China Sea (Fig. 1).The average water depth is 10 m (range: 6-20 m).The annual mean air temperature is 22°C.The minimum sea surface temperature occurs in winter (15°C) and the maximum in summer and fall (30°C).The bay is one of the most intensive culture areas in China.The Portu guese oyster Crassostrea angulata is the main cultured bivalve species.Suspension aquaculture of C. angulata has been practiced for over 3 decades, with an estimated standing stock of 6.6 × 10 4 t in 2016.The seaweed Gracilaria lemaneiformis is another important cultured species, with a production of ca.27 × 10 4 t in China in 2015 (China Bureau of Fisheries 2016).

Estimation of calcification and respiration rate of C. angulata
Experimental C. angulata were collected from Daya Bay in April 2016 and taken to the laboratory in a temperature-controlled case within 1 h.After arrival, animals were disinfected, and any visible fouling organisms on shell surfaces were cleaned by washing with filtered seawater.Approximately 30 oysters of similar sizes (7−8 cm shell height) were acclimatized to laboratory conditions for 1 wk in a 50 l tank with aerated seawater.During acclimation, seawater was changed once per day and oysters were fed daily with 2 × 10 4 cell ml −1 Chaetoceros sp. at 08:00 h.During the trial, temperature (T ), salinity (S), and pH were 22 ± 1°C, 30.8 ± 0.1, and 8.03 ± 0.02, respectively.
At the end of the acclimatization period, the oysters were placed in closed 20 l transparent polyethylene plastic mesocosms filled with seawater.The mesocosms were hung from a suspended longline so that the experimental oysters were at a depth of ca. 2 m, corresponding to the routine culture depth for oysters.A factorial design was used to test the effects of 3 stocking densities: i.e. low, medium, and high (ca. 1, 5, and 10 g (fresh weight, FW) oyster l −1 , respectively) on water pH, dissolved oxygen (DO), total alkalinity (TA), DIC, and carbonate ion (CO 3 2− ) concentrations.Water samples were taken at 0 and 4 h.T and S were measured using a multi-parameter water quality meter (YSI Professional Plus 6600, Yellow Springs Instrument Company).pH was measured using a pH meter (Thermo Scientific Orion 320P-01, Thermo Electron Corporation) calibrated on the US National Bureau of Standards scale.The precision of pH measurements was ± 0.01 pH units.Oxygen concentrations were determined by the Winkler method (Strickland & Parson 1972).TA was measured using an 848 Titrino plus automatic titrator (Metrohm) on 100 ml GF/F filtered samples.The accuracy of measurements was ± 2 μmol l −1 .DIC and CO 3 2− concentrations were computed from T, S, pH, and TA using the CO 2 _SYS_XLS calculation program (Pierrot et al. 2006).The dissociation constants for carbonic acid (K 1 , K 2 ) were from Mehrbach et al. (1973) as refitted by Dickson & Millero (1987), the dissociation constant for bisulfate ion (K HSO4 ) was obtained from Dickson (1990), and the dissociation constant for boric acid (K B ) was from Uppstrom (1974).
CO 2 respiratory rate (R; μmol FW g −1 h −1 ) can be expressed as follows: where ΔDIC is the net change in DIC concentration (μmol l −1 ), which was caused by the interaction between calcification and respiration.DIC i and DIC f were the initial and final DIC concentration (μmol l −1 ), respectively.V is the incubation chamber volume (l), t is the experimental time (h), and M is the fresh weight of experimental oyster (g).TA i and TA f were the initial and final TA concentration (μmol l -1 ), respectively.
For CaCO 3 production, 30 oysters were sampled, and the dry weight (DW) of oyster shell was determined by drying at 80°C till constant weight (± 0.01 g).Bivalve shells largely consist (95%) of CaCO 3 , and the remaining 5% are made up by magnesium, β-chitin, and various glycoproteins (Goulletquer & Wolowicz 1989).Shell DWs were corrected accordingly.The calcimass (g CaCO 3 m −2 ) was estimated by the shell DWs per m 2 and its CaCO 3 content.Dry tissue weight was calculated for each individual using the ash-free dry weight (AFDW) method: oyster soft tissue was dried at 80°C (72 h) and then ashed at 500°C (4 h), with tissue weight computed as the difference between the 2 weights.
The monthly ratio of CO 2 released to CaCO 3 precipitated (Ψ) was estimated as a function of the water temperature, measured with a YSI meter.CO 2 fluxes due to calcification were calculated using a Ψ value

Deployment of in situ mesocosm experiment
The in situ mesocosm experiment was carried out from 20 to 22 April 2016 in Daya Bay (22°34' N, 114°32' E) (Fig. 1).The cylindrical mesocosms (1 m diameter × 1.5 m height) were made from transparent polyethylene plastic and were hung on suspended longlines with the top ca.2.0 m below the water surface.Fifteen cylindrical mesocosms (450 l) were deployed over 24 h periods and consisted of 5 treatments each with 3 replicates (Table 1).One treatment with only seawater served as the control (C), the second treatment contained only oysters (oyster only, O), and the other 3 treatments were coculture systems, with 3 oyster:seaweed ratios, i.e. 8:1, 4:1, and 2:1 (based on FW of oyster and seaweed, referred to as OS_8:1, OS_4:1 and OS_2:1, respectively).After filling with natural seawater, the mouths of the mesocosms were tied using ropes.A pipe was placed in the tied site and maintained for 24 h to keep DIC concentration in equilibrium with air due to water mixing with air.Oysters and seaweed were coiled into 100-cm-long ropes and placed in the mesocosms.The ropes were suspended using thin ropes and tied to the mouth of the mesocosm, such that the seaweed thalli were positioned vertically around the oysters.The mesocosms were immobilized using a set of ropes connected top-side to a float and submerged under the water surface.
The experiment began at 10:00 h and lasted for 24 h.Water T, S, DO, pH, and TA were measured at the beginning and end of the experiment.Parameters for the seawater DIC system and aqueous pCO 2 were calculated by the CO2_SYS_XLS calculation program (Pierrot et al. 2006).
The sea−air CO 2 fluxes (F CO2 ) were calculated based on the following equation: where k (cm h −1 ) is the gas exchange coefficient of CO 2 .We computed k using the parameterization given by Wanninkhof & McGillis (1999) that uses short-term winds k = 0.0283u 10 3 (Sc/660) −1/2 .u 10 stands for the wind speed at a 10 m height from the water surface level (m s −1 ) and Sc is the Schmidt number calculated according to the relationship proposed by Wanninkhof (1992).α (mol kg −1 atm −1 ) is the solubility coefficient of CO 2 calculated after Weiss (1974).ΔpCO 2 is the pCO 2 difference between surface seawater and the atmosphere.In this study, the value of atmospheric pCO 2 was downloaded from www.cmdl.noaa.gov(National Oceanographic and Atmospheric Administration, NOAA, Climate and Meteorological Diagnostics Laboratory) and corrected for water vapor pressure (Takahashi et al. 2002).Positive magnitudes of F CO2 indicate a flux from water to air and vice versa.
The net oxygen production rate by G. lemaneiformis in co-culture systems was determined based on the DO concentrations in experimental mesocosms, as: where C os and C o are the DO concentrations (μmol l −1 ) of oyster−seaweed co-culture mesocosms and oyster-only mesocosm, respectively, at the end of the experiment.V is the volume of mesocosms (l), FW is the fresh weight of G. lemaneiformis (g), and t is the duration of the experiment (h).

Statistical analysis
Data were analyzed by 1-way ANOVA.All data were graphically assessed for normality and homogeneity of residuals (Faraway 2002).When overall differences were significant at the 0.05 level, Tukey's HSD multiple range test was used to compare the mean values of individual groups.Data are reported as means ± SE (n = 3).All statistical tests were performed using SPSS 17.0 for Windows.

Crassostrea angulata
As shown in Fig. 2, after a 4 h incubation, the seawater pH, TA, and CO 3 2− concentrations in the closed mesocosms gradually decreased with increasing oyster stocking density.Values in the highest stocking density group were significantly lower than the other groups (p < 0.05) (Fig. 2a−c).The pCO 2 followed the converse general pattern, and values were signifi-cantly different between each treatment (p < 0.05) (Fig. 2d).
The lowest and highest respiration rates of oysters were found in the high-and medium-density groups, respectively, and were significantly different (p < 0.05) (Fig. 3a).Calcification rates of oysters decreased with increasing stocking density, but no significant difference occurred (p > 0.05) (Fig. 3b).
The total CaCO 3 production by the C. angulata population in Daya Bay was estimated to be ca.2150 g CaCO 3 m −2 yr −1 .C. angulata sequestered 258 g C m −2 yr −1 for shell formation.The ratio of CO 2 released to CaCO 3 precipitated ( Ψ) ranged from 0.54 to 0.65 and varied monthly with water temperature variation in Daya Bay (Table 2).CO 2 released due to calcification and respiration was 153 and 349 g C m −2 yr −1 , respectively.

Variations of seawater pH and TA in different mesocosms
As shown in Fig. 4a, the seawater pH differed significantly among different mesocosms (p < 0.05).pH was lowest in the oyster-only treatment, significantly lower than that of the control and the oyster−seaweed co-culture groups (p < 0.05).pH gradually increased with increasing seaweed density in co-culture treatments, and was significantly higher than that of the control (p < 0.05).TA was highest in the control, significantly higher than that of the other groups (p < 0.05) (Fig. 4b).TA values in OS_4:1 and OS_2:1 coculture groups were significantly lower than that of the other groups (p < 0.05) (Fig. 4b).

Variations of seawater DIC systems and pCO 2 in different mesocosms
After 24 h incubation, DIC, HCO 3 − , and CO 2 concentrations and pCO 2 showed similar trends among treatments (Fig. 5a,b,d,e).The highest values were found in the oyster-only group, and then continuously decreased with increasing seaweed density in co-culture groups.Co-culture with seaweed lead to a significant decrease of CO 2 concentration and pCO 2 (p < 0.05).The degree of reduction was positively correlated with the seaweed density (Fig. 5d,e).CO 3 2− concentration followed the converse general pattern to HCO 3 − concentrations (Fig. 5c).

Variations of air−sea CO 2 flux in different mesocosms
The air−sea CO 2 flux (F CO2 ) in the oyster-only treatment group had a high and positive value (110.4 ± 10.5 mmol m −2 d −1 ), representing a CO 2 source to the atmosphere, and was significantly higher than in the other groups (p < 0.05).In contrast, the negative values (from −[8.4 ± 0.7] to −[33.6 ± 4.0] mmol m −2 d −1 ) in the control and co-culture groups represent a CO 2 sink from the atmosphere to the sea, where the degree of CO 2 sink was proportional to seaweed stocking density; there was no significant difference in F CO2 between the oyster−seaweed (OS_8:1) coculture and the control (p < 0.05), but the F CO2 values in OS_4:1 and OS_2:1 groups were significantly lower than that in the control (p > 0.05) (Fig. 6).

Oxygen production rate of Gracilaria lemaneiformis in co-culture mesocosms
In oyster−seaweed co-culture mesocosms, although the DO was mainly produced by G. lemaneiformis, the phytoplankton also produced some oxygen.Therefore, the oxygen concentration in oyster− seaweed co-culture mesocosms (Fig. 7) was refitted by the oxygen concentration in the control and oyster-only mesocosms.The net oxygen production rates in the low-(OS_8:1) and medium-(OS_4:1) seaweeddensity treatments were significantly (p < 0.05) higher than that in the high-seaweed-density group (OS_2:1).

DISCUSSION
The results of the present study indicated that the oyster Crassostrea angulata cultivated in Daya Bay seems to be a CO 2 generator, as pCO 2 increased in oyster-only culture mesocosms.Oyster harvesting sequesters ca.258 g C m −2 yr −1 due to shell formation in Daya Bay.In contrast, the CO 2 fluxes due to respiration and calcification were ca.349 and 153 g C m −2 yr −1 , respectively, accounting for 69.5% and 30.5% of the total CO 2 fluxes (502 g C m −2 yr −1 ), respectively.This result indicated that total carbon fluxes were mainly influenced by respiration, but the contribution of calcification was not negligible.Based on the balance between CaCO 3 sequestration and CO 2 release, the C. angulata populations in Daya Bay increase CO 2 release to the atmosphere in coastal ecosystems.Moreover, our measurements may have underestimated the overall contribution of C. angulata to CO 2 fluxes, since we have not considered the rate of carbonate dissolution of shells that remained in the system after oyster death.
During the 4 h incubation, the DO concentrations in all mesocosms were above 4 mg l −1 .This level was not likely to induce stress to the oysters (Diaz & Rosenberg 2008).The reduced respiration rate by oysters in the high-density group (Fig. 3a) might be a strategy to cope with variability in seawater pH and the ability to adapt to seawater acidification (Guppy & Withers 1999, Langenbuch & Pörtner 2004).
Seawater CO 3 2− can affect the ability of calcifying organisms to precipitate CaCO 3 (Gazeau et al. 2007, Zhang et al. 2011, Dineshram et al. 2013, Li et al. 2013, Mos et al. 2015, McGrath et al. 2016).However, in the present experiment, no significant differences in calcification rate were found among the different density treatments (Fig. 3b), although the CO 3 2− was lower in the high-density group (Fig. 2c).This might indicate that the CO 3 2− deficiency stress was not severe enough to depress calcification.Therefore, further studies with longer incubation times and/or larger biomass of oysters are needed to produce more severe acidification stress.
According to the calcification rate and culture densi ty of oysters, the mean CaCO 3 production by C. angulata population in Daya Bay is ca.2150 g CaCO 3 m −2 yr −1 .This is higher than that of the oyster Crassostrea gigas (134 g CaCO 3 m −2 yr −1 ) in Brest Bay (Lejart et al. 2012).Varying results could be due to species-specific differences.Lejart et al. (2012) studied the natural populations of C. gigas, which inhabit the intertidal zone going through 14 h of underwater respiration and calcification, and 10 h aerial respiration each day with the changing tides.However, C. angulata in Daya Bay were cultured under constant immersion conditions.Therefore, C. angulata has a longer period of calcification to produce higher amounts of CaCO 3 .
CO 2 released during CaCO 3 precipitation of oysters in Daya Bay represented about 30.5% of the total CO 2 production.This result was consistent with previous findings, e.g.30% for Ophiothrix fragilis (Migné et al. 1998), 33% for Potamocorbula amurensis (Chauvaud et al. 2003), and 23−26% for Acroc-nida brachiata (Davoult et al. 2009) in the eastern English Channel.Therefore, although total carbon fluxes were mainly influenced by underwater respiration, there is a contribution from calcification that should not be neglected.
Gracilaria lemaneiformis can use both CO 2 and HCO 3 − for photosynthesis.In the oyster−seaweed coculture mesocosm, the HCO 3 − concentration was significantly higher in the OS_4:1 group than in the OS_2:1 group, but there was no significant difference in CO 2 concentration (Fig. 5).This phenomenon was consistent with the findings of Raven et al. (2014) and Axelsson et al. (2000), who reported that in seawater of pH 8.0 and above, the principal species of DIC in the medium is HCO 3 − , but the active transport of HCO 3 − needs higher energy than passive CO 2 diffusion.Hence, it is reasonable that G. lemaneiformis has a higher affinity for CO 2 than HCO 3

−
, which lead to a preferential CO 2 exhaust over HCO 3 − .In the present study, calcification and respiration by oysters occurred over the duration of 24 h in the closed mesocosm system, while photosynthesis by G. lemaneiformis occurred only during the daytime when there is light.The CO 2 :CO 3 2− ratio and pH of the seawater in the co-culture system would depend on the balance between the photosynthesis rate by the seaweed and the respiration rate and calcification rate of the oysters (Menéndez et al. 2001, Zhang et al. 2012).Seawater pCO 2 and CO 2 :CO 3 2− ratios decreased in all oyster−seaweed co-culture systems, indicating that there was stronger CO 2 uptake by G. lemaneiformis than CO 2 release from C. angulata, leading to a net uptake of CO 2 from the atmosphere into the seawater.Meanwhile, we found that the net oxygen production rate of G. lemaneiformis in the OS_2:1 treatment was significantly decreased compared with that in the OS_8:1 and OS_4:1 groups (Fig. 7).As the primary production of seaweed is carbon limited, the carbon-saturated maximum photosynthesis of G. lemaneiformis would drastically reduce when it was 'starved' of DIC (Han et al. 2013).Since the numbers of oysters among the 3 treatments were almost the same, the decreased oxygen production rate i.e. the photosynthesis rate of G. lemaneiformis in the OS_2:1 group was probably due to a carbon limitation.Thus, there could be an evident interspecies mutual benefit for both C. angulata and G. lemaneiformis in the co-culture system.Based on the results of the present study, we suggest that the 2 species are co-cultured at a ratio of ca.4:1 (fresh weight) for efficient utilization of seawater DIC by G. lemaneiformis, and further to increase the ocean CO 2 sink.
In conclusion, the physiological activities of C. angulata lead to a shift in the seawater DIC system equilibria towards higher CO 2 , lower pH, and lower CO 3 2− concentration, and subsequently are affected by this shift.Seaweed G. lemaneiformis could act as an efficient sink for CO 2 .Incorporation of seaweed into oyster aquaculture can be helpful in eliminating DIC release from C. angulata.There could be complex interspecies effects between C. angulata and G. lemaneiformis.The beneficial effects of an integrated multi-trophic aquaculture system on seawater carbon budget and air−sea CO 2 fluxes should be determined based on an ecosystem approach.

Fig. 1 .
Fig. 1.Location of the study area (D) in Daya Bay

Table 1 .
Overview of the co-culture systems in the 5 treatments, with oysters and seaweed retained in mesocosms (g fresh weight per mesocosm, mean ± SD)

Table 2 .
Mean monthly water temperature (T, °C) and corresponding molar ratio (Ψ) in Daya Bay