83 ISJ 15: 83-93, 2018 ISSN 1824-307X RESEARCH REPORT Combined effects of temperature and salinity on growth, survival, gill morphology, and antioxidant capabilities in the horse mussel Modiolus modiolus Y Zhan, M Yang, D Cui, J Li, J Sun, J Ning, Z Hao, W Zhang, Y Chang Key Laboratory of Mariculture & Stock Enhancement in North China’s Sea, Ministry of Agriculture, Dalian Ocean University, Dalian, Liaoning, 116023, P. R. China Accepted March 04, 2018 Abstract Temperature and salinity are critical to the reproduction, growth, and survival of shellfish. We investigated the combined effects of temperature (15 °C, 20 °C, and 25 °C) and salinity [20, 30, and 40 practical salinity units (PSUs)] on the horse mussel Modiolus modiolus, which inhabits the northern Chinese coast. The specific growth rate of M. modiolus increased with salinity at a low temperature (15 °C). The interaction between temperature and salinity with respect to M. modiolus survival was significant (Two-Way ANOVA; p < 0.01). Across all temperatures tested, M. modiolus specimens kept at low salinity (20 PSU) had narrower gill filament lumens than those kept at moderate salinity (30 PSU); the gill filament lumens of specimens kept at high salinity (40 PSU) were wider than either of these. The combined effects of temperature and salinity on superoxide dismutase (SOD) activity, total antioxidant capacity (T-AOC), and glutathione peroxidase (GPx) activity in the gills of M. modiolus were significant (Two-Way ANOVA; p < 0.05). Our results suggested that temperature and salinity significantly affect the survival, gill morphology, and antioxidant capabilities of M. modiolus. Our findings not only enrich our knowledge of the interactions between temperature and salinity with respect to shellfish but also provide a framework for M. modiolus aquaculture development. Key Words: horse mussel; gill lumen; SOD; T-AOC; GPx; superoxide dismutase; total antioxidant capacity; glutathione peroxidase; SGR Introduction The horse mussel, Modiolus modiolus (Mollusca: Bivalvia), is a benthic, sessile species distributed worldwide in coastal areas (Hutchison et al., 2016). In China, M. modiolus is mainly found in the subtidal zones of the Yellow and Bohai Seas (Ning et al., 2015). M. modiolus modifies its habitat via bio-deposition, forming beds on various substrata (Fariñas‐Franco et al., 2016). It is therefore believed that M. modiolus has an important impact on the ecological equilibrium of the sedimentary environments of the shallow sublittoral (Ragnarsson and Burgos, 1984). M. modiolus is highly edible and grows rapidly to a large size (Ning et al., 2016). This species has thus recently been proposed as a potential new aquaculture species in China (Ning et al., 2015). ___________________________________________________________________________ Corresponding author: Yaqing Chang Key Laboratory of Mariculture & Stock Enhancement in North China’s Sea Ministry of Agriculture Dalian Ocean University Dalian, Liaoning, 116023, P. R. China E-mail: yqkeylab@hotmail.com Temperature and salinity are the two primary environmental factors affecting the reproduction, growth, and survival of shellfish (Meryem et al., 2015). For example, increased water temperature and water quality were thought to contribute to mortality in the oyster Crassostrea gigas (Shelaghk et al., 2009), while high temperature significantly affected survival, oxygen consumption, behavior, ammonia-N excretion, and related immune indicators in the Japanese scallop Mizuhopecten yessoensis (Hao et al., 2014). In addition, the restoration of normal antioxidant enzyme activity in the clam Cyclina sinensis was increased after exposure to abnormal salinities; C. sinensis had a greater tolerance for low salinity than for high salinity (Li et al., 2012). Recently, the combined effects of temperature and salinity on shellfish have received more attention (Samuel et al, 2015). For example, Wang et al. (2017) showed that a temperature of 26 °C and a salinity of 30 practical salinity units (PSU) were optimal for the culture of juvenile ark shells (Anadara broughtonii). Unfortunately, studies of the combined influence of temperature and salinity on shellfish remain scarce. mailto:yqkeylab@hotmail.com 84 Fig. 1 Effects of temperature and salinity on SGR in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. Bars show the means of 3 replicates. Error bars show SD. Table 1 One-way and two-way ANOVA of the effects of different temperatures and salinities on SGR in M. modiolus. Values are means ± SD (n = 3). Temperature (°C) Salinity (PSU) p-value Duncan test Temperature Salinity Interaction One-way ANOVA (factor: Temperature) 15, 20, 25 20 0.85ns -- -- 15 °C = 20 °C = 25 °C 30 0.96ns 15 °C = 20 °C = 25 °C 40 0.04* 15 °C > 20 °C = 25 °C One-way ANOVA (factor: Salinity) 15 20, 30, 40 -- 0.02* -- 40 PSU > 30 PSU = 20 PSU 20 0.84ns 20 PSU = 30 PSU = 40 PSU 25 0.91ns 20 PSU = 30 PSU = 40 PSU Two-way ANOVA (Fixed factors: Temperature & Salinity) 15, 20, 25 20, 30, 40 0.15ns 0.16ns 0.11ns 15 °C = 20 °C = 25 °C 20 PSU = 30 PSU = 40 PSU *, value is significant based on the Duncan tests (p < 0.05); ns, value is not significant based on the Duncan test. Although M. modiolus tolerates a wide range of environmental conditions (Lesser and Kruse, 2004), it has previously been shown that alterations in a single environmental factor (such as temperature, salinity, or pH) affect growth, survival, and feeding (Lesser and Kruse, 2004; Ning et al., 2012). Before large-scale farming of M. modiolus begins, it is critical to identify the optimal conditions for such farms. We therefore explored the combined impact of temperature and salinity on M. modiolus, specifically aiming to 1) identify the effects of temperature and salinity on growth and survival in M. modiolus; 2) identify morphological changes in the gills of M. modiolus in response to temperature and salinity; 3) identify alterations in superoxide dismutase (SOD) levels, total antioxidant capacity (T-AOC), and glutathione peroxidase (GPx) levels in the gills of M. modiolus in response to changes in temperature and salinity. Our results will help to determine the ideal nursery conditions and farming locations for M. modiolus. 85 Materials and Methods Animal collection and treatments In November 2015, we collected 220 healthy adult Modiolus modiolus specimens (mean shell height = 42.71 ± 3.08 mm; mean shell length = 96.09 ± 4.65 mm; mean shell width = 49.82 ± 3.21 mm; and mean mass = 13.39 ± 2.32 g) from the near-shore sublittoral zone of the North China Sea near the Ministry of Agriculture Key Laboratory of Mariculture & Stock Enhancement at Dalian Ocean University, Dalian, China (121°33′47″ E, 38°51′55″ N). All specimens were kept in ~60 L recirculating seawater tanks; each tank was fitted with an automatic temperature control and monitoring system (Dalian Huixin Titanium Equipment Development Co., Ltd, Liaoning, China). Seawater was sand filtered and continuously aerated. Specimens were kept in natural light. All specimens were supplied with 0.25 g spirulina powder (Dalian Golden Bay Technology Co., Ltd, Liaoning, China) daily. The amount of spirulina powder supplied exceeded the daily requirements of the specimens (Ning et al., 2015) to ensure that our results were not biased by insufficient nutrition. We also judged that the spirulina was supplied in excess because extra powder remained in the water after a day of circulation. All specimens were acclimated to default laboratory conditions (13.95 ± 1.32 °C and 31.69 ± 0.35 PSU) for one week before experiments began. Experiments were conducted between November 2015 and January 2016. After acclimation, all specimens were dried with a paper towel and weighed with digital balance (0.01 g sensitivity; AL204; Mettler Toledo, Shanghai, China) to obtain initial mass (W1). We then divided mussels into 27 groups of three specimens each (three temperatures × three salinities with three replicates of each combination; Table S1). Each group was housed in a separate tank. To reach the desired temperature/salinity combinations, we removed half of the seawater from each tank every day, and replaced it with seawater at a different temperature/salinity. We changed the temperature of the new seawater such that the temperature of the entire tank did not increase by more than 1 °C per day; this was based on previous study on M. modiolus (Ning et al., 2015) and on field survey data of the coastal waters of the Yellow Sea (Zhang et al., 2014). We altered the salinity of the replacement seawater either by adding fresh filtered water or by adding synthetic sea salt (Aqua Salt; Dalian Salt Industry Co., Ltd, Liaoning, China). We altered the salinity of the new seawater such that salinity in the entire tank did not change by more than 2 PSU per day (Ning et al., 2015, Hamer et al., 2008). We monitored the temperature and salinity in each tank both using the automatic temperature control, and with a water quality monitor (A329 Portable Meter; Thermo Scientific Orion Star, Beijing, China). Growth and survival in each treatment During the 30-day experimental period, dead M. modiolus [defined as loss of motility and lack of external stimulus response (Ning et al., 2015)], were noted and removed immediately. At the end of the experiment, all living specimens were again dried and weighed to obtain a final mass (W2). We then calculated the specific growth rate (SGR; defined as the growth rate per day) and survival rate (SR) for each group as follows: SGR (% × day-1) = 100 × (lnW2 - lnW1)/t SR (%) = 100 × (N2/N1) where N1 is the initial number of live specimens; N2 is the final number of live specimens; and t is the duration of the experiment in days. Gill filament morphology and antioxidant analyses The gills of all living M. modiolus specimens were carefully removed at the end of the experimental period. Gills were divided into two parts: one for morphological analysis, and the other for detection of SOD, T-AOC, and GPx. To example gill filament morphology, we fixed gill tissue samples for 24 h in Bonn Fluid (75 ml saturated picric acid, 25 ml 40 % formaldehyde, and 5 ml acetic acid), following the procedures of Wu et al. (2007). We then transferred the fixed samples into different concentrations of ethanol for gradient dehydration (Han et al.,2014). Dehydrated fixed tissues were embedded in paraffin wax, and 5 µm sections were stained with hematoxylin and eosin (Qu et al., 2013). We then examined gill filament morphology and photographed the gill filament lumens using an optical microscope with attached camera (50i; Nikon, Japan). Using DN-2 (Dalian Zhonghe Company, Liaoning, China) with our gill photographs, we measured the width of each gill filament lumen for each specimen, as it has been shown that lumen width is affected by changes in temperature and salinity (Wang et al., 2016). To determine SOD, T-AOC, and GPx, gill samples were dried with paper towels and homogenized immediately in an ice bath. Gill homogenates were centrifuged at 3,500 rpm for 10 min at 4 °C. Supernatants were stored at -80 ºC. We used commercial kits to determine SOD, T-AOC, and GPx (Nanjing Jiancheng Company, China; kit model used for SOD: A001-1; for T-AOC: A015-1; for GPx: A005), following the manufacturer's instructions. SOD activity, T-AOC level, and GPx activity were calculated in units per mg protein (U × mg protein-1) as follows: SOD = (A2 - A3)/0.5 × (V1/V0)/C0 T-AOC = (A3 - A2)/0.01/30 × (V1/V0)/C0 GPx = (A2 - A3)/(A1 - A0) × C1 × 5/5/(V0 × C0) where A0 is the light absorption (O.D.) of the blank group; A1 is the light absorption (O.D.) of the standard group; A2 is the light absorption (O.D.) of the control group; A3 is the light absorption (O.D.) of the given experimental group; C0 is the protein concentration of the homogenate; C1 is the protein concentration of the standard (20 μg/L); V0 is the volume of the sample; and V1 is the volume of the total reaction. 86 The protein concentration of each gill homogenate was determined with the Bradford method, using bovine serum albumin as the standard, on a multimode reader (SpectraMax i3x; Molecular Devices, Shanghai, China). Data analysis All statistical analyses were performed with SPSS 16.0 (IBM, Shanghai, China). We first confirmed that our data were normally distributed and homogeneous with the Shapiro-Wilk test and with Levene’s test. We then compared differences in SGR, survival rate, gill filament lumen size, SOD activity, T-AOC level, and GPx activity among treatments with one-way ANOVA (factor: temperature or salinity) and with two-way ANOVAs (fixed factors: temperature and salinity). We considered p < 0.05 statistically significant. Significant differences between pairs of treatments were identified with Duncan’s multiple range test. Results Effects of temperature and salinity on growth and survival Only M. modiolus specimens cultured at 15 °C in high salinity (40 PSU) had an increased SGR (Fig. 1; Table 1, S2). We found that M. modiolus cultured in moderate salinity (30 PSU) had relatively higher survival rates as compared to other salinity treatments at the same temperature (Fig. 2). High survival rates (> 90%) were also observed at temperatures below 20 °C and salinities between 30 and 40 PSU. M. modiolus specimens kept at 20 °C in low salinity (20 PSU) had lower survival rates (68.75 ± 0.08 %). For M. modiolus specimens maintained at 25 °C, survival rate decreased as salinity increased (Fig. 2). The interaction between temperature and salinity was significant with respect to survival (Two-way ANOVA, p < 0.01), and salinity had a more pronounced effect Fig. 2 Effects of temperature and salinity on survival in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. Bars show the means of 3 replicates. Error bars show SD. Table 2. One-way and two-way ANOVA of the effects of different temperatures and salinities on survival rate in M. modiolus. Values are means ± SD (n = 3). Temperature (°C) Salinity (PSU) p-value Duncan test Temperature Salinity Interaction One-way ANOVA (factor: Temperature) 15, 20, 25 20 0.06ns -- -- 15 °C > 25 °C > 20 °C 30 0.15ns 15 °C = 20 °C = 25 °C 40 0.004** 20 °C = 15 °C > 25 °C One-way ANOVA (factor: Salinity) 15 20, 30, 40 -- 0.19ns -- 20 PSU = 30 PSU = 40 PSU 20 0.01* 40 PSU = 30 PSU > 20 PSU 25 0.14ns 20 PSU = 30 PSU = 40 PSU Two-way ANOVA (Fixed factors: Temperature & Salinity) 15, 20, 25 20, 30, 40 0.02* 2.05×10-4**** 0.50×10-2* 15 °C = 20 °C > 25 °C 30 PSU > 20 PSU = 40 PSU *, value is significant based on the Duncan tests (*p < 0.05, **p < 0.01, ****p < 0.0001); ns, value is not significant based on the Duncan test. 87 Fig. 3 Effects of temperature and salinity on gill filament morphology in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. on M. modiolus survival than the temperature (Two-way ANOVA, p < 0.01). It is interesting to note that the negative effects of low salinity (20 PSU) on M. modiolus were somewhat alleviated at the low temperature (i.e., 15 °C) and the high temperature (25 °C) (One-way ANOVA, p > 0.05; Table 2). Effects of temperature and salinity on gill filament morphology As salinity and/or temperature increased, the gill filament lumens of M. modiolus specimens widened (Fig. 3). Relatively narrow gill filament lumens were observed in M. modiolus specimens cultured at 20 °C in low salinity (20 PSU). At moderate salinity (30 PSU), gill filament lumen size did not vary obviously with incubation temperature. Wider gill filament lumens were observed in M. modiolus specimens cultured at 15 °C in high salinity (40 PSU). At 25 °C, gill filament lumens were significantly smaller in low salinity (20 PSU) than in high salinity (40 PSU; Fig. 4; Table 3). The interaction between temperature and salinity with respect to the width of gill lumen was significant (Two-way ANOVA, p < 0.01), and salinity had a more pronounced effect on gill filament morphology than did temperature (Two-way ANOVA, p < 0.01). High temperatures (25 °C) partially alleviated the negative effects of high salinity on the gill filament morphology of M. modiolus (One-way ANOVA, p < 0.05; Table 3). 88 Fig. 4 Effects of temperature and salinity on gill filament lumen width in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. Bars show the means of 3 replicates. Error bars show SD. Table 3 One-way and two-way ANOVA of the effects of different temperatures and salinities on gill filament lumen size in M. modiolus. Values are means ± SD (n = 3). Temperature (°C) Salinity (PSU) p-value Duncan test Temperature Salinity Interaction One-way ANOVA (factor: Temperature) 15, 20, 25 20 3.54×10-11**** -- -- 15 °C = 25 °C > 20 °C 30 1.51×10-4**** 15 °C = 20 °C > 25 °C 40 4.60×10-11**** 15 °C > 20 °C > 25 °C One-way ANOVA (factor: Salinity) 15 20, 30, 40 -- 1.93×10-9**** -- 40 PSU > 30 PSU > 20 PSU 20 4.43×10-31**** 40 PSU > 30 PSU > 20 PSU 25 0.10×10-2** 40 PSU = 30 PSU > 20 PSU Two-way ANOVA (Fixed factors: Temperature & Salinity) 15, 20, 25 20, 30, 40 3.27×10-13**** 2.09×10-40**** 1.57×10-10**** 15 °C > 20 °C = 25 °C 40 PSU > 30 PSU > 20 PSU *, value is significant based on the Duncan tests (**p < 0.01, ****p < 0.0001); ns, value is not significant based on the Duncan test. Effects of temperature and salinity on SOD activity, T-AOC level, and GPx activity SOD activity in the gills of M. modiolus specimens kept at 15°C declined significantly with increased salinity (p < 0.001); in specimens kept at 25°C, SOD activity increased with increased salinity (Fig. 5; Table 4). At 20 °C, SOD activity was significantly higher in specimens cultured in low salinity (20 PSU) and in those cultured in high salinity (40 PSU), as compared to those cultured in moderate salinity (30 PSU). The interaction between temperature and salinity with respect to SOD activity was significant (Two-way ANOVA, p < 0.001). The T-AOC of the M. modiolus gill was significantly affected by both temperature (One-way ANOVA, p < 0.05) and salinity (One-way ANOVA, p < 0.05; Fig. 6; Table 5). At each of the tested temperatures, the gills of M. modiolus specimens cultured in moderate salinity (30 PSU) had lower T-AOC than those cultured in low (20 PSU) and high salinities (40 PSU; Fig. 6). At each of the tested salinities, the gills of M. modiolus specimens kept at 89 Fig. 5 Effects of temperature and salinity on gill SOD level in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. Bars show the means of 3 replicates. Error bars show SD. Table 4 One-way and two-way ANOVA of the effects of different temperatures and salinities on SOD levels in M. modiolus. Values are means ± SD (n = 3). Temperature (°C) Salinity (PSU) p-value Duncan test Temperature Salinity Interaction One-way ANOVA (factor: Temperature) 15, 20, 25 20 8.49×10-6**** -- -- 15 °C = 25 °C > 20 °C 30 0.10×10-2** 20 °C > 25 °C > 15 °C 40 0.30×10-2** 20 °C = 25 °C > 15 °C One-way ANOVA (factor: Salinity) 15 20, 30, 40 -- 2.53×10-6**** -- 20 PSU > 30 PSU > 40 PSU 20 3.03×10-5**** 20 PSU > 30 PSU = 40 PSU 25 0.10×10-2** 40 PSU > 30 PSU = 20 PSU Two-way ANOVA (Fixed factors: Temperature & Salinity) 15, 20, 25 20, 30, 40 1.59×10-12**** 3.00×10-11**** 9.96×10-12**** 15 °C > 20 °C = 25 °C 40 PSU > 30 PSU > 20 PSU *, value is significant based on the Duncan tests (**p < 0.01, ****p < 0.0001); ns, value is not significant based on the Duncan test. a moderate temperature (20 °C) had higher T-AOC than those maintained at low (15 °C) or high temperatures (25 °C; Fig. 6). The interaction between temperature and salinity with respect to T-AOC in the M. modiolus gill was significant (Two-way ANOVA; p < 0.01). GPx activity in the gills of M. modiolus was significantly affected by both temperature (One-way ANOVA; p < 0.0001) and salinity (One-way ANOVA; p < 0.0001; Fig. 7; Table 6). At each of the tested temperatures, specimens kept at both low (20 PSU) and high salinities (40 PSU) had significantly less GPx activity than those kept at a moderate salinity (30 PSU). At each of the tested salinities, the gills of M. modiolus specimens cultured at a moderate temperature (20 °C) had relatively higher GPx activity than those cultured at low (15 °C) and high (25 °C) temperatures (Fig. 7). The interaction between temperature and salinity with respect to GPx activity in the M. modiolus gill was significant (Two-way ANOVA; p < 0.0001). Discussion Previous studies have indicated that shellfish growth is affected by both temperature and salinity (Hao et al., 2014; Çelik et al., 2015). Hao et al. (2014) showed that exposing M. yessoensis to a low temperature (15 °C) for 30 days increased SGR significantly, while Bashevkin and Pechenik (2015) 90 Fig. 6 Effects of temperature and salinity on gill T-AOC in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. Bars show the means of 3 replicates. Error bars show SD. Table 5 One-way and two-way ANOVA of the effects of different temperatures and salinities on T-AOC in M. modiolus. Values are means ± SD (n = 3). Temperature (°C) Salinity (PSU) p-value Duncan test Temperature Salinity Interaction One-way ANOVA (factor: Temperature) 15, 20, 25 20 0.1×10-1* -- 20 °C > 15 °C = 25 °C 30 0.1×10-2** 20 °C >25 °C > 15 °C 40 2.32×10-6**** 20 °C > 25 °C > 15 °C One-way ANOVA (factor: Salinity) 15 20, 30, 40 -- 0.1×10-2** -- 20 PSU > 40 PSU > 30 PSU 20 4.26×10-4**** 40 PSU > 20 PSU = 30 PSU 25 0.60×10-2** 40 PSU > 20 PSU = 30 PSU Two-way ANOVA (Fixed factors: Temperature & Salinity) 15, 20, 25 20, 30, 40 1.95×10-8**** 6.24×10-11**** 9.71×10-6**** 20 °C > 15 °C = 25 °C 40 PSU > 20 PSU > 30 PSU *, value is significant based on the Duncan tests (*p < 0.05, **p < 0.01, ****p < 0.0001); ns, value is not significant based on the Duncan test. found that exposure to low temperature and low salinity significantly reduced both the larval and juvenile growth rates of the gastropod, Crepidula fornicata. It has also been shown that salinity had the greatest effect on the growth rate of A. broughtonii (Wang et al., 2017). Here, we found that the SGR of adult M. modiolus specimens increased with salinity at a low temperature (15 °C). We therefore hypothesize that response to temperature and salinity in marine mollusks varies with species. Adult M. modiolus specimens survived ten days at salinities of 20 and 40 PSU and temperatures between 10 °C and 12 °C; at moderate salinity (~31 PSU) and temperatures above 25 °C, survival decreased (Ning et al., 2015). Our results were consistent with this previous study. We also found that salinity alone, especially low salinity, was the main cause of reduced survival in M. modiolus. Therefore, our data support the hypothesis that salinity maintenance should be the first concern of shellfish aquaculture (Wang et al., 2017). As low salinity significantly reduced the feeding rate of C. fornicata juveniles (Samuel and Pechenik, 2015), we postulate that the reduced survival rate of M. modiolus might be due to a dramatic decrease in feeding rate. With respect to aquaculture management, it was worth noting that the negative effects of low salinity were somewhat alleviated by decreased or increased temperatures. 91 Fig. 7 Effects of temperature and salinity on gill GPx levels in M. modiolus. S20, salinity of 20 PSU; S30, salinity of 30 PSU; S40, salinity of 40 PSU. Bars show the means of 3 replicates. Error bars show SD. Table 6 One-way and two-way ANOVA of the effects of different temperatures and salinities on GPx levels in M. modiolus. Values are means ± SD (n = 3). Temperature (°C) Salinity (PSU) p-value Duncan test Temperature Salinity Interaction One-way ANOVA (factor: Temperature) 15, 20, 25 20 1.31×10-10**** -- -- 20 °C > 25 °C > 15 °C 30 3.79×10-7**** 20 °C > 25 °C > 15 °C 40 2.33×10-4**** 20 °C > 25 °C = 15 °C One-way ANOVA (factor: Salinity) 15 20, 30, 40 -- 5.59×10-8**** -- 30 PSU > 20 PSU > 40 PSU 20 1.40×10-7**** 30 PSU > 20 PSU > 40 PSU 25 1.20×10-5**** 30 PSU > 20 PSU > 40 PSU Two-way ANOVA (Fixed factors: Temperature & Salinity) 15, 20, 25 20, 30, 40 1.89×10-17**** 1.30×10-18**** 1.17×10-6**** 20 °C > 25 °C > 15 °C; 30 PSU > 20 PSU > 40 PSU *, value is significant based on the Duncan tests (****p < 0.0001); ns, value is not significant based on the Duncan test. The adverse effects of temperature and salinity on M. modiolus were also clear in our examination of gill morphology. Gills are the primary organs of both respiration and defense in bivalves (Fiddy et al., 2016). Here, abnormal temperatures and salinities damaged gill filaments and altered gill filament lumen width. Salinity affects the respiration and growth of larval and juvenile C. fornicata (Samuel et al., 2015); thus, it is possible that the reduction in M. modiolus survival observed here might have been due to the effects of temperature and salinity on respiration, instead of feeding rate. Alternatively, it has been shown that changes in gill filament lumen size indirectly demonstrate an imbalance in osmotic pressure (Han et al., 2014). Therefore, it might be that temperature and salinity affect the survival of M. modiolus by disrupting endogenous osmotic pressure. Further work is necessary to assess the combined impact of temperature and salinity on respiration and osmotic pressure in M. modiolus. The generation of reactive oxygen species (ROS) is a typical physiological phenomenon in aerobic organisms. It is clear that excessive accumulation of ROS (such as O2, H2O2, and OH -) can cause oxidative damage to lipids, proteins, carbohydrates, and nucleic acids (Khazri et al., 2016; Shenaitirodkar et al., 2017). The maintenance of optimal ROS levels is thus vital for endo-redox homeostasis and for the survival of aerobic species. 92 SOD and GPx, two important enzymes in the antioxidant defense system, also serve as ROS scavengers (Levent et al., 2005); T-AOC typically reflects the balance of oxidants and antioxidants in aerobic organisms (Fan et al., 2004). Decreases in temperature (from 26 ºC to 15 ºC) significantly decreased SOD activity in M. yessoensis at normal salinity (~31 PSU; Hao et al., 2014). In adult C. gigas, SOD activity decreased and then increased as temperature increased from 24 ºC to 32 ºC, but in adult C. brasiliana, SOD activity decreased steadily with the same temperature increase (Moreira et al., 2017). Here, we found that SOD activity increased and then decreased as temperature rose from 15 ºC to 25 ºC. These results suggested that the effects of temperature on SOD activity vary among bivalves; there may be species-specific mechanisms that manage temperature variations. At 15 °C, SOD activity decreased significantly as salinity increased, but at 25 °C, SOD activity increased significantly as salinity increased. This interesting result indicated that salinity might have a more profound impact on SOD activity in the gill of M. modiolus than temperature. It was reported that elevated temperature increased the activity of GPx in the gills of the brown mussel Perna perna (Almeida et al.,2007). In contrast, GPx activity in the gills of M. modiolus at both low and high temperatures (15 °C and 25 °C) was lower at all tested salinities than GPx activity at 20 °C. This discrepancy might be due to differences in GPx function among species. In addition, the decreased gill GPx activity at both low and high salinities (20 and 40 PSU) observed in M. modiolus might indicate that increased salinity interrupts the antioxidant system of this species by reducing the catalytic capacity of GPx. T-AOC levels in M. modiolus increased at both low and high salinities, suggesting that M. modiolus might compensate for the increased numbers of ROS induced by abnormal temperatures and salinities by increasing T-AOC. Further research is necessary to clarify the mechanism by which M. modiolus increases T-AOC. Our results suggested that M. modiolus might be suitably cultured between 15 °C and 20 °C, and between 30 and 40 PSU. Our study provides a framework for the selection of suitable locations for M. modiolus aquaculture, as well as for the development of sustainable aquaculture methods. Our data also increase our understanding of the impact of temperature and salinity on molluscan growth and survival. Future studies should focus on identifying and testing additional ROS markers to further clarify the effects of temperature and salinity on endo-redox homeostasis in M. modiolus. In addition, molecular studies (including differential gene expression, transcriptome comparison, and microRNA analysis) would help to further clarify the mechanisms affected by temperature and salinity in adult M. modiolus. Acknowledgments This work was supported by a Grant for Chinese Outstanding Talents in Agricultural Scientific Research (for Yaqing Chang). References Almeida EAD, Bainy ACD, Loureiro APDM, Glaucia R M, Sayuriet M, Janice O, et al. Oxidative stress in Perna perna, and other bivalves as indicators of environmental stress in the Brazilian marine environment: Antioxidants, lipid peroxidation and DNA damage. Comp Biochem Physiol A Mol Integr Physiol. 146(4): 588-600, 2007. Fan XH, Li JK, Zhang Q, Shi JZ. Effect of walnut oil on the activities of antioxidase and total antioxidation capacity (T-AOC) in vivo of mice. Jour. of Northwest Sci-Tech Univ. of Agri. and For (Nat. Sci. Ed.) 32(11): 122-124, 2004. Fariñas-Franco JM, Sanderson WG, Roberts D. Phenotypic differences may limit the potential for habitat restoration involving species translocation: a case study of shape ecophenotypes in different populations of Modiolus modiolus (Mollusca: Bivalvia). Aquatic Conserv: Mar Freshw Ecosyst. 26(1): 76-94, 2016. Fiddy SP, Comeau LA, Gastineau R, Decottignies P, Cognie B, Morançais M, et al. Effect of marennine produced by the blue diatom Haslea ostrearia on behavioral, physiological and biochemical traits of juvenile Mytilus edulis and Crassostrea virginica. Aquaculture 467: 138-148, 2016. Hamer B, Jaksić Z, Pavicićhamer D, Perić L, Medaković D, Ivanković D, et al. Effect of hypoosmotic stress by low salinity acclimation of Mediterranean mussels Mytilus galloprovincialis on biological parameters used for pollution assessment. Aquat. Toxicol. 89: 137-151, 2008. Han XL, Gao BQ, Wang HF, Liu P, Chen P, Li H. Effects of low salinity stress on microstructure of gill and hepatopancreas and family survival rate of Portunus trituberculatus. Prog Fish Sci. 35(1): 104-110, 2014. Hao ZL, Tang XJ, Ding J, Yue B, Chang YQ. Effect of high temperature on survival, oxygen consumption, behavior, ammonia-N excretion, and related immune indicators of the Japanese scallop Mizuhopecten yessoensis. Aquacult. Int. 22: 1863-1876, 2014. Hutchison ZL, Hendrick VJ, Burrows MT, et al. Buried Alive: The Behavioural Response of the Mussels, Modiolus modiolus and Mytilus edulis to Sudden Burial by Sediment. Plos One. 11(3): e0151471, 2016. Khazri A, Sellami B, Hanachi A, Dellali M, Eljarrat E, Beyrem H, et al. Neurotoxicity and oxidative stress induced by permethrin in gills of the freshwater mussel Unio ravoisieri. Chem. Ecol. 33 (1): 88-101, 2016. Lesser MP, Kruse VA. Seasonal temperature compensation in the horse mussel, Modiolus modiolus: metabolic enzymes, oxidative stress and heat shock proteins. Comp. Biochem. Phsiol. 137A: 495-504, 2004. Levent C, Yurdakoc K, Yokes B. Antioxidant status of Lobiger serradifalci, and Oxynoe olivacea, (Opisthobranchia, Mollusca). J Exp Mar Biol Ecol. 314(2): 227-235, 2005. 93 Li ZN, Lin TT, Yao ZL, Lai QF, Lu JX, Wang H, et al. Effects of water salinity on the antioxidant enzyme activities and growth of clam Cyclina sinensis. Chinese J. Ecol. 10: 031, 2012. Meryem YÇ, Sedat K, İsmihan K, Bora E, Recep Ö. The effects of environmental factors on survival, growth and biochemical composition of transplanted oysters (Ostreaedulis Linnaeus, 1758) from Aegean Sea to southern Black Sea. Aquac. Res. 46: 959-968, 2015. Moreira A, Figueira E, Pecora IL, Soares AMVM, Freitas, R. Biochemical alterations in native and exotic oyster species in Brazil in response to increasing temperature. Comp. Biochem. Phsiol. 191C: 183-193, 2017. Ning JH, Bai W, Song J, Zhang XS, Chang YQ. Acute toxicity of four kinds of heavy metal ions to clam Modiolus modiolus. J. Dalian Fish Univ. 31(3): 290-294, 2016. Ning JH, Zhang XX, Chang YQ, Song J, Jing CC, Zhang WJ, et al. Effects of several ecological factors on survival in adult mussel Modiolus modiolus. J. Dalian Fish Univ. 30: 285-290, 2015. Qu YJ, Chen SH, Li JE, Su H. Light, Scanning and Transmission Electron Microscopical Observation of Gill of Captive Liza haematocheila on the Coast of Pearl River Estuary. Sichuan Journal of Zoology 32 (4): 550-554, 2013. Ragnarsson SA, Burgos JM. Separating the effects of a habitat modifier, Modiolus modiolus and substrate properties on the associated megafauna, J. Sea Res. 72: 55-63, 1984. Samuel MB, and Pechenik JA. The interactive influence of temperature and salinity on larval and juvenile growth in the gastropod Crepidula fornicata (L.). J. Exp. Mar. Biol. Ecol. 470: 78-91, 2015. Shelaghk M, Elizabeth C, Selena OK, Sharon L, Sarahc C, Jonathanw K, et al. Summer mortality of the Pacific oyster, Crassostrea gigas, in the Irish Sea: The influence of temperature and nutrients on health and survival. Aquaculture 287: 128-138, 2009. Shenai-Tirodkar PS, Gauns MU, Mujawar MW, Ansari ZA. Antioxidant responses in gills and digestive gland of oyster Crassostrea madrasensis (Preston) under lead exposure. Ecotox. Environ. Safe.,142: 87, 2017. Wang Q, Xie X, Zhang M, Teng W, Liang M, Kong N, et al. Effects of temperature and salinity on survival and growth of juvenile ark shell Anadara broughtonii. Fish. Sci. 83: 619-624, 2017. Wang Y, Hu WB, Li JX, Huang XY, Duan LZ, Zhan YY. Effects of Acute Salinity Stress on Gill structure and Four Enzyme Activities in Saxidomus purpurata. J. Agr. Sci. Tech. 18(5): 178-186, 2016. Wu LL, Chen L, Zhang YL, Li JH, Zhao JF. Effects of phenanthrene toxicity on histopathology of Brachydanio rerio gill and liver. Chinese J. Ecol. 26: 688-692, 2007. Zhang H, Jiang ZT, Liu EX. Seasonal variation of water temperature in Yellow Sea and East China Sea. China Water Transport 16: 167-170, 2016. http://xueshu.baidu.com/s?wd=author:(Karay%C3%BCcel,%20%C4%B0smihan)%20Sinop%20University%20Faculty%20of%20Fisheries%20and%20Aquaculture%20Sinop%20Turkey&tn=SE_baiduxueshu_c1gjeupa&ie=utf-8&sc_f_para=sc_hilight=person http://xueshu.baidu.com/s?wd=author:(%C3%96zt%C3%BCrk,%20Recep)%20Sinop%20University%20Faculty%20of%20Fisheries%20and%20Aquaculture%20Sinop%20Turkey&tn=SE_baiduxueshu_c1gjeupa&ie=utf-8&sc_f_para=sc_hilight=person