Effect of Salinity on Cr(VI) Bioremediation by Algal-Bacterial Aerobic Granular Sludge Treating Synthetic Wastewater

: Heavy metal-containing wastewater with high salinity challenges wastewater treatment plants (WWTPs) where the conventional activated sludge process is widely applied. Bioremediation has been proven to be an effective, economical, and eco-friendly technique to remove heavy metals from various wastewaters. The newly developed algal-bacterial aerobic granular sludge (AGS) has emerged as a promising biosorbent for treating wastewater containing heavy metals, especially Cr(VI). In this study, two identical cylindrical sequencing batch reactors (SBRs), i.e., R1 (Control) and R2 (with 1% additional salinity), were used to cultivate algal-bacterial AGS and then to evaluate the effect of salinity on the performance of the two SBRs. The results reﬂected that less ﬁlamentation and a rougher surface could be observed on algal-bacterial AGS when exposed to 1% salinity, which showed little inﬂuence on organics removal. However, the removals of total inorganic nitrogen (TIN) and total phosphorus (TP) were noticeably impacted at the 1% salinity condition, and were further decreased with the co-existence of 2 mg/L Cr(VI). The Cr(VI) removal efﬁciency, on the other hand, was 31–51% by R1 and 28–48% by R2, respectively, indicating that salinity exposure may slightly inﬂuence Cr(VI) bioremediation. In addition, salinity exposure stimulated more polysaccharides excretion from algal-bacterial AGS while Cr(VI) exposure promoted proteins excretion.


Introduction
The rapidly increasing global population, along with the development of industrialization, has triggered various environmental issues in many countries worldwide. Heavy metal (HM) in wastewater has emerged as a global environmental concern because of its adverse effects on human beings and ecosystems [1]. Among the HMs commonly found in wastewater, hexavalent chromium (Cr(VI)) is considered as one of the most harmful and toxic metal forms because of its high carcinogenic and mutagenic properties [2,3]. More specifically, exposure to Cr(VI) through inhalation, digestion, or direct contact can negatively affect human health, leading to liver and kidney damage, internal hemorrhage, and respiratory disorders [4]. Furthermore, Cr(VI)-contaminated wastewater is gaining more attention due to its diverse applications and higher concentrations in water contributed by various industrial processes such as metallurgy, refractory materials, textile dyeing, leather tanning, electroplating, and chemicals production [5,6].
To date, many techniques have been developed and applied to remove Cr(VI) from wastewater, such as chemical precipitation, reverse osmosis, ion exchange, and adsorp-

Experimental Design and Operation
In this study, two identical cylindrical SBRs with a working volume of 250 mL (D × H = 43 mm × 300 mm) each, namely R1 (control, no additional salinity addition) and R2 (with 1% additional salinity), were used. The reactors were operated at room temperature (25 ± 2 • C) under a 6 h cycle consisting of 7 min of feeding, 60 min of nonaeration, 282 min of aeration, 2 min of settling, 8 min of decanting, and 1 min of idling. The volume exchange rate (VER) was set at 50%, resulting in a hydraulic retention time (HRT) of 12 h. Air was introduced into each reactor from the bottom by an air pump (AK-40, KOSHIN, Japan), and the superficial uplift air velocity was controlled at 0.81 cm/s. The reactors were operated under 24 h illumination, which was provided by two artificial lights (lb055cw, RIKAKEN, Nagoya, Japan) attached to the outside wall of the reactors at an illuminance of around 6000 lux. During the experimental period, the sludge retention time (SRT) was not controlled, which varied between 30-40 days for the two SBRs according to the biomass concentration in the reactors and their effluent biomass concentrations. The whole experiment and test consisted of four stages as listed in Table 1.

Synthetic Wastewater and Seed Mature Algal-Bacterial Granules
Synthetic wastewater was prepared and fed to the two SBRs, which consisted of the main components as follows: 300 mg chemical oxygen demand (COD)/L (with NaAc as the sole carbon source), 30 mg NH 4 + -N/L (NH 4 Cl), 5 mg PO 4 -P/L (KH 2 PO 4 ), 10 mg Ca 2+ /L (CaCl 2 ·2H 2 O), 5 mg Fe 2+ /L (FeSO 4 ·7H 2 O), 5 mg Mg 2+ /L (MgSO 4 ·7H 2 O), and 1 mL/L of trace elements solution. The specific composition of the trace elements solution was described in a previous study [18]. In addition, 50 mg/L NaHCO 3 was added to maintain the influent pH. In our previous study [15], the algal-bacterial AGS system was found to tolerate relatively low salinity <5 g/L (or 0.5%) during the 6 h Cr(VI) biosorption, which would be negatively affected when exposed to higher salinity levels. As the major purpose of this study was to examine the effect of the co-existing Cr(VI) and salinity on algal-bacterial AGS during a long-term operation, approximately 1% salinity was initially applied in the experiments. The results from the operation under higher salinity levels or real wastewaters will be included in the follow-up reports. The 1% salinity (with conductivity of about 17,000 µS/cm) in wastewater was generated by adding NaCl at a concentration of 10 g/L. The 2 mg/L Cr(VI) in the influent wastewater was obtained by diluting a 1000 mg/L stock Cr(VI) solution prepared from K 2 Cr 2 O 7 . Mature algal-bacterial AGS was used in this study, which was developed from bacterial AGS. More specifically, the bacterial AGS was first sampled from the long-term bacterial AGS system operated by Wang et al. [19]. Then, the bacterial AGS was cultivated in another lab-scale SBR (D × H = 200 mm × 800 mm with a working volume of 15.7 L) under an artificial light illumination system of approximately 15k lux on the water surface and 40 k lux on the reactor wall (light on/off = 12/12 h). This lab-scale SBR was stably operated for 6 months at a mixed liquor suspended solids (MLSS) concentration of 3.4 g/L with sludge volume index in 30 min (SVI 30 ) of about 70 mL/g. Moreover, prior to the 4-stage operation in this study, the algal-bacterial AGS was again cultivated in R1 and R2 for about 14 days by feeding the same influent wastewater without supplementation of salt and Cr(VI). Both achieved the same stable system performance before starting Stage I.

Analytical Methods
All water samples were first collected at the end of the third operation cycle (19:00) every day, then filtered through 0.22 µm membrane filters before further analysis. Ammonia nitrogen (NH 4 + -N), nitrite nitrogen (NO 2 − -N), nitrate nitrogen (NO 3 − -N), phosphate phosphorus (PO 4 3− -P), mixed liquor (volatile) suspended solids (ML(V)SS), and sludge volume index (SVI) were measured according to the standard methods [20]. In this study, total inorganic nitrogen (TIN) was calculated as the sum of NH 4 + -N, NO 2 − -N, and NO 3 − -N. Dissolved organic carbon (DOC), dissolved inorganic carbon (DIC), and dissolved total carbon (DTC) in the effluent samples were analyzed by a total organic carbon (TOC) analyzer (TOC-VCSN, Shimadzu, Kyoto, Japan). The concentration of Cr(VI) was measured by a UV-vis spectrophotometer (UV 1800, Shimadzu, Japan), which has been described in detail elsewhere [14]. Chlorophyll a (Chl-a) content in the granules was extracted and analyzed by the method described previously [21]. All the other assays, such as the granule size and distribution, granule strength (expressed as integrity coefficient or IC), extraction Processes 2021, 9, 1400 4 of 15 of extracellular polymeric substances (EPS) (including loosely bound EPS (LB-EPS) and tightly bound EPS (TB-EPS)), and determination of proteins (PN) and polysaccharides (PS), were the same as in a previous study [22].

Statistical Analysis
All the analyses were performed in triplicate and the results were expressed as mean or mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was conducted to compare the performance of the two SBRs and test stages using Microsoft Excel 2010. Statistically significant difference was assumed if p < 0.05. In this study, changes in morphology and size of the granules from the two reactors were observed and recorded during the four stages' operation ( Figure 1). Initially, the mature algal-bacterial AGS from both reactors was relatively regular in shape and dark green in color, demonstrating a dense and compact structure. After Stage I, there was not much change in the morphology of the granules because the salinity and Cr(VI) were not supplemented to the influent of the two reactors. Starting from Stage II, when 1% salinity was added to the influent wastewater of R2, some difference was clearly observed in the granules from the two reactors, especially the growth of filamentous bacteria outside the granules. Specifically, under the 1% salinity condition (R2), bacterial growth may have been inhibited, and further inhibition was observed under the prolonged exposure to salinity (to the end of Stage IV). In contrast, the granules in R1 (control) demonstrated the rapid growth of filamentous bacteria on their surface along with the operation. In addition, the presence of Cr(VI) in the influent to the two reactors (in Stage III) led to the appearance of specific colored precipitates (in gray and bright yellow) surrounding the surface of the granules from R1 and R2, respectively. Processes 2021, 9, x FOR PEER REVIEW 4 of 15 detail elsewhere [14]. Chlorophyll a (Chl-a) content in the granules was extracted and analyzed by the method described previously [21]. All the other assays, such as the granule size and distribution, granule strength (expressed as integrity coefficient or IC), extraction of extracellular polymeric substances (EPS) (including loosely bound EPS (LB-EPS) and tightly bound EPS (TB-EPS)), and determination of proteins (PN) and polysaccharides (PS), were the same as in a previous study [22].

Statistical Analysis
All the analyses were performed in triplicate and the results were expressed as mean or mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was conducted to compare the performance of the two SBRs and test stages using Microsoft Excel 2010. Statistically significant difference was assumed if p < 0.05.

Morphology and Granule Size
In this study, changes in morphology and size of the granules from the two reactors were observed and recorded during the four stages' operation ( Figure 1). Initially, the mature algal-bacterial AGS from both reactors was relatively regular in shape and dark green in color, demonstrating a dense and compact structure. After Stage I, there was not much change in the morphology of the granules because the salinity and Cr(VI) were not supplemented to the influent of the two reactors. Starting from Stage II, when 1% salinity was added to the influent wastewater of R2, some difference was clearly observed in the granules from the two reactors, especially the growth of filamentous bacteria outside the granules. Specifically, under the 1% salinity condition (R2), bacterial growth may have been inhibited, and further inhibition was observed under the prolonged exposure to salinity (to the end of Stage IV). In contrast, the granules in R1 (control) demonstrated the rapid growth of filamentous bacteria on their surface along with the operation. In addition, the presence of Cr(VI) in the influent to the two reactors (in Stage III) led to the appearance of specific colored precipitates (in gray and bright yellow) surrounding the surface of the granules from R1 and R2, respectively.    Figure 2 demonstrates the changes in granular size and its distribution in the two reactors (R1 and R2). Generally, granules ranging between 1.5-2 mm and 2-3 mm were predominant in both R1 and R2 during the test period. In addition, an opposite trend was observed in the occupying ratio of the smallest (≤0.5 mm, decreased) and largest (>3 mm, increased) granules in both reactors. This trend was more obvious in R2, possibly due to the salinity exposure. Regarding the average diameter of granules, there was a gradual increase over time in both reactors, from the initial 1.92 ± 0.87 (R1) and 1.90 ± 0.87 (R2) mm to 2.58 ± 1.39 (R1) and 2.62 ± 1.11 (R2) mm at the end of test, respectively. predominant in both R1 and R2 during the test period. In addition, an opposite trend w observed in the occupying ratio of the smallest (≤0.5 mm, decreased) and largest (>3 m increased) granules in both reactors. This trend was more obvious in R2, possibly due the salinity exposure. Regarding the average diameter of granules, there was a gradu increase over time in both reactors, from the initial 1.92 ± 0.87 (R1) and 1.90 ± 0.87 (R2) m to 2.58 ± 1.39 (R1) and 2.62 ± 1.11 (R2) mm at the end of test, respectively.  Figure 3a shows the changes of the MLSS and MLVSS/MLSS ratio in the two reacto during the four operation stages. Obviously, the MLSS concentration in the two reacto increased gradually over time, from the initial 3.40 to 6.67 g/L in R1 and 9.08 g/L in R respectively. A larger increase in MLSS was detected in R2, which is probably attributab to the effect of salinity exposure. The larger increase in biomass under the salinity exp sure condition (R2 in this study) is consistent with previous studies [23,24], possibly d to the increased accumulation of inorganic materials into the granules. In addition, t presence of Cr(VI) in wastewater (Stage III) also triggered the increase in biomass grow in R1. In terms of MLVSS/MLSS, this ratio in R1 fluctuated between 80-92% during t experiment, while that in R2 decreased from 86% (Stage I) to 52% after exposure to 1 NaCl (Stage II), and then stabilized around 89% during the rest period of test.

Granular Biomass Growth and Settleability
On the other hand, salinity exposure also enhanced the granule settleability (Figu 3b), as the SVI5 of granules from R2 decreased from 62.0 mL/g (Stage I) to 26.4 mL/g (Sta IV). This finding agreed well with other previous works [23,25]. In addition, Dong et [25] attributed the excellent settleability of algal-bacterial AGS under salinity conditio to its inhibitory effect on the growth of filamentous bacteria.  Figure 3a shows the changes of the MLSS and MLVSS/MLSS ratio in the two reactors during the four operation stages. Obviously, the MLSS concentration in the two reactors increased gradually over time, from the initial 3.40 to 6.67 g/L in R1 and 9.08 g/L in R2, respectively. A larger increase in MLSS was detected in R2, which is probably attributable to the effect of salinity exposure. The larger increase in biomass under the salinity exposure condition (R2 in this study) is consistent with previous studies [23,24], possibly due to the increased accumulation of inorganic materials into the granules. In addition, the presence of Cr(VI) in wastewater (Stage III) also triggered the increase in biomass growth in R1. In terms of MLVSS/MLSS, this ratio in R1 fluctuated between 80-92% during the experiment, while that in R2 decreased from 86% (Stage I) to 52% after exposure to 1% NaCl (Stage II), and then stabilized around 89% during the rest period of test.

Granular Biomass Growth and Settleability
On the other hand, salinity exposure also enhanced the granule settleability (Figure 3b), as the SVI 5 of granules from R2 decreased from 62.0 mL/g (Stage I) to 26.4 mL/g (Stage IV). This finding agreed well with other previous works [23,25]. In addition, Dong et al. [25] attributed the excellent settleability of algal-bacterial AGS under salinity conditions to its inhibitory effect on the growth of filamentous bacteria.

Granular Stability and Chl-a Content
In this study, the integrity coefficient (IC) and Chl-a represent the granular stability (or granular strength) and algal content in granules, respectively. From Figure 3c, during Stage I with no salinity exposure and Cr(VI) addition, the IC values of granules from R1 and R2 were similar at 2.01% and 2.04%, respectively. However, in the presence of 10 g/L of NaCl (during Stage II), the IC of algal-bacterial AGS in R2 was consistently higher than that in R1 (no salinity exposure), indicating the decrease in granular strength in R2. The negative effect of salinity on granular strength in this study is consistent with previous studies [25,26]. Interestingly, when 2 mg/L Cr(VI) (Stage III) was co-existing, the IC values of granules in both reactors slightly decreased in comparison to their previous stage, and then slightly increased to 1.82% (R1) and 4.82% (R2), respectively, when Cr(VI) exposure ceased (Stage IV), implying a positive effect of Cr(VI) on granule stability. As a whole, a very slight change in IC value was observed in the granules from both reactors during the test period, probably due to the very compact structure of algal-bacterial AGS and the low concentrations of NaCl and/or Cr(VI) tested.
As seen from Figure 3c, after the exposure to 1% salinity (Stage II), the Chl-a content in the granules from R2 was consistently lower than that from R1, implying the negative influence of salinity on the growth of microalgae. On the other hand, in the presence of Cr(VI), the Chl-a contents tended to decrease in the granules from both reactors, which continued to decline even after Cr(VI) exposure ceased (Stage IV), to 6.65 mg/g-VSS and 5.30 mg/g-VSS in the granules from R1 and R2, respectively.

Organics Removal
The changes in effluent DIC, DOC concentrations, and DOC removal efficiency are shown in Figure 4. As seen, little impact was observed on organics removal under the salinity exposure and/or Cr(VI) presence, achieving, on average, >93% of DOC removal by the two reactors (p = 0.152 >> 0.05). Specifically, the effluent DOC concentrations from the two reactors were consistently lower than 18 mg/L, with an average DOC removal efficiency of 96.7 ± 1.1% by R1 and 96.5 ± 0.8% by R2 during Stage I, respectively. After the 1% salinity exposure (Stage II), the effluent DOC concentration from R2 began to increase and exceeded that from R1 (no salinity exposure) by 10.52 mg/L and 6.70 mg/L, respectively. Then, both DOC concentrations tended to increase after the exposure to 2 mg/L Cr(VI) (Stage III). When the exposure to Cr(VI) ceased during Stage IV, a slight decrease in DOC was detected in the effluents from both reactors, suggesting that both reactors could recover their DOC removal capacity. The above results imply the limited negative effect of salinity and Cr(VI) on algal-bacterial AGS with respect to DOC removal performance. The relatively low salinity level (1%) may lead to a result of "no significant difference" in DOC removal between the two reactors.
Compared with the effluent DIC concentration from R1, that from R2 was slightly higher. This phenomenon may be associated with the less algae growth in R2, which can be clearly reflected from the lower Chl-a content in the granules from R2 after salinity exposure (from Stage II to the end of test, Figure 3c). Thus, salinity exposure may partially contribute to the increase of effluent DIC concentration. For instance, according to Dong et al. [25], high salinity (1% and 3%) conditions can negatively affect the algal growth in algal-bacterial AGS system, resulting in high DIC concentrations in the effluent of SBR. Interestingly, after exposure to 2 mg/L Cr(VI) (Stage III), both reactors produced a lower DIC concentration effluent, and then became relatively stable when the Cr(VI) exposure ceased.  Compared with the effluent DIC concentration from R1, that from R2 was slightly higher. This phenomenon may be associated with the less algae growth in R2, which can be clearly reflected from the lower Chl-a content in the granules from R2 after salinity exposure (from Stage II to the end of test, Figure 3c). Thus, salinity exposure may partially contribute to the increase of effluent DIC concentration. For instance, according to Dong et al. [25], high salinity (1% and 3%) conditions can negatively affect the algal growth in algal-bacterial AGS system, resulting in high DIC concentrations in the effluent of SBR. Interestingly, after exposure to 2 mg/L Cr(VI) (Stage III), both reactors produced a lower DIC concentration effluent, and then became relatively stable when the Cr(VI) exposure ceased.

Nitrogen Removal
The changes in effluent N species (including NH4 + -N, NO2 − -N, and NO3 − -N) and total inorganic nitrogen (TIN) removal are shown in Figure 5. Among the variations of the three N species, the concentrations of effluent NH4 + -N from the two reactors showed the most apparent fluctuations. More specifically, the effluent NH4 + -N concentration from R2 gradually increased after the exposure to 1% salinity began (Stage II). This increasing trend continued when the granules were exposed to 2 mg/L Cr(VI) (Stage III), while the NH4 + -N concentration steadily decreased when the exposure ceased from Stage IV. In contrast, the effluent NH4 + -N concentration from R1 (no salinity exposure) gradually increased from the beginning of exposure to Cr(VI) (Stage III), but decreased during the later period of Stage III. Being different from the apparent fluctuation in effluent NH4 + -N concentrations, the effluent NO2 − -N concentrations remained at a relatively constant low level, averagely 0.61 mg/L and 1.22 mg/L from R1 and R2, respectively. As for NO3 − -N, Cr(VI) exposure did not exert a noticeable effect on the granules in the two reactors. In addition, as can be seen from Figure 5, the effluent NO3 --N concentrations from R2 were generally lower than those from R1 during most of the test period, to some extent reflecting the negative effect of salinity exposure on nitrification.
Based on the changes of TIN removal during the 28 days' operation of R1 and R2, the main findings can be summarized as follows: (i) In the test SBR (R2), nitrification process

Nitrogen Removal
The changes in effluent N species (including NH 4 + -N, NO 2 − -N, and NO 3 − -N) and total inorganic nitrogen (TIN) removal are shown in Figure 5. Among the variations of the three N species, the concentrations of effluent NH 4 + -N from the two reactors showed the most apparent fluctuations. More specifically, the effluent NH 4 + -N concentration from R2 gradually increased after the exposure to 1% salinity began (Stage II). This increasing trend continued when the granules were exposed to 2 mg/L Cr(VI) (Stage III), while the NH 4 + -N concentration steadily decreased when the exposure ceased from Stage IV. In contrast, the effluent NH 4 + -N concentration from R1 (no salinity exposure) gradually increased from the beginning of exposure to Cr(VI) (Stage III), but decreased during the later period of Stage III. Being different from the apparent fluctuation in effluent NH 4 + -N concentrations, the effluent NO 2 − -N concentrations remained at a relatively constant low level, averagely 0.61 mg/L and 1.22 mg/L from R1 and R2, respectively. As for NO 3 − -N, Cr(VI) exposure did not exert a noticeable effect on the granules in the two reactors. In addition, as can be seen from Figure 5, the effluent NO 3 − -N concentrations from R2 were generally lower than those from R1 during most of the test period, to some extent reflecting the negative effect of salinity exposure on nitrification.
Based on the changes of TIN removal during the 28 days' operation of R1 and R2, the main findings can be summarized as follows: (i) In the test SBR (R2), nitrification process was inhibited when exposed to 1% salinity, which became worse after the granules were exposed to 2 mg/L Cr(VI) together. The granules could be recovered when the exposure to Cr(VI) ceased. (ii) In the control SBR (R1), Cr(VI) exposure also exhibited an inhibitory effect on nitrification, and the granules could be acclimated and recovered along with the operation. Once the Cr(VI) exposure ceased, the granules could quickly be recovered to the initial level. the initial level.
The above results indicate that exposure to 1% salinity and/or 2 mg/L Cr(VI) may negatively affect nitrogen removal performance to different extents. This observation is probably associated with the inhibitory effects of the above harsh environment on the growth of algae or bacteria that are responsible for nitrogen uptake and conversion, in agreement with previous works [25,27].  Figure 6 shows the effluent TP concentration and TP removal efficiency of the two SBRs during the 28 days of operation. At Stage I with no exposure to 1% salinity and 2 mg/L Cr(VI), the TP removal efficiency was averagely kept at about 68.8% by R1 and 80.8% by R2, respectively, and this difference was mainly due to their different SRTs (no SRT control). After that, an obvious increase in the effluent TP concentration from R2 was detected, ranging from 0.97 mg/L to 3.59 mg/L when the granules were exposed to 1% salinity (Stage II), which led to a dramatic reduction in TP removal to the lowest level of 28.1%. However, this P removal performance was quickly recovered over the operation time. When the granules were exposed 2 mg/L Cr(VI) (Stage III), the effluent TP concentrations from R1 and R2 increased from 0.73 mg/L and 1.26 mg/L to 2.34 mg/L and 1.99 mg/L, corresponding to the reductions in TP removal efficiency to 53.3% and 60.3%, respectively. Even so, their TP removal efficiencies could be rapidly recovered, reaching 97.2% and 78.1% by R1 and R2 when the exposure to Cr(VI) ceased (Stage IV). This means that the algal-bacterial granules had a good TP removal capability and recoverability under fluctuated wastewater quality and variable environmental conditions, which could quickly recover from the short-term inhibition of 1% salinity and/or 2 mg/L Cr(VI) exposure. This observation is different from a previous study [25], i.e., the salt-stressed algal- The above results indicate that exposure to 1% salinity and/or 2 mg/L Cr(VI) may negatively affect nitrogen removal performance to different extents. This observation is probably associated with the inhibitory effects of the above harsh environment on the growth of algae or bacteria that are responsible for nitrogen uptake and conversion, in agreement with previous works [25,27]. Figure 6 shows the effluent TP concentration and TP removal efficiency of the two SBRs during the 28 days of operation. At Stage I with no exposure to 1% salinity and 2 mg/L Cr(VI), the TP removal efficiency was averagely kept at about 68.8% by R1 and 80.8% by R2, respectively, and this difference was mainly due to their different SRTs (no SRT control). After that, an obvious increase in the effluent TP concentration from R2 was detected, ranging from 0.97 mg/L to 3.59 mg/L when the granules were exposed to 1% salinity (Stage II), which led to a dramatic reduction in TP removal to the lowest level of 28.1%. However, this P removal performance was quickly recovered over the operation time. When the granules were exposed 2 mg/L Cr(VI) (Stage III), the effluent TP concentrations from R1 and R2 increased from 0.73 mg/L and 1.26 mg/L to 2.34 mg/L and 1.99 mg/L, corresponding to the reductions in TP removal efficiency to 53.3% and 60.3%, respectively. Even so, their TP removal efficiencies could be rapidly recovered, reaching 97.2% and 78.1% by R1 and R2 when the exposure to Cr(VI) ceased (Stage IV). This means that the algal-bacterial granules had a good TP removal capability and recoverability under fluctuated wastewater quality and variable environmental conditions, which could quickly recover from the short-term inhibition of 1% salinity and/or 2 mg/L Cr(VI) exposure. This observation is different from a previous study [25], i.e., the salt-stressed algal-bacterial AGS (under sudden salinity disturbance conditions) could not recover its original TP removal rate. The different response of algal-bacterial AGS in these two works are most probably attributable to the differences in SBR operating conditions (dimension and size, cycle time, aeration time, operating duration, etc.), light condition, salinity exposure procedure, and the different characteristics of the initial granules used.

Phosphorus Removal
nal TP removal rate. The different response of algal-bacterial AGS in these two works are most probably attributable to the differences in SBR operating conditions (dimension and size, cycle time, aeration time, operating duration, etc.), light condition, salinity exposure procedure, and the different characteristics of the initial granules used.
Further studies are still necessary on the effect of salinity and/or Cr(VI) exposure on the anaerobic P release rate and aerobic P uptake rate, thereby shedding light on the mechanism regarding the varied P removal performance of the algal-bacterial AGS system.

Chromium Removal
In order to assess the effect of salinity on Cr(VI) removal efficiency by the algal-bacterial AGS system, the effluent Cr(VI) and Cr(III) concentrations, as well as the Cr(VI) removal efficiency, were monitored during Stage III (Figure 7). It is noteworthy that the Cr(VI) removal rate of R2 was generally lower than that of R1 in Stage III, possibly due to the slightly negative effect of salinity exposure. Specifically, the Cr(VI) removal efficiencies of R1 and R2 fluctuated between 31-51% (37.3% on average) and 28-48% (34.8% on average), respectively (p = 0.005 < 0.05). This observation might be attributable to the inhibitory impact of salinity on microbial activity in the AGS system [28], leading to the decline in Cr(VI) removal efficiency. In addition, the increase in ionic strength under the 1% salinity condition (R2) might have some shielding effect on the surface electrostatic field, thus suppressing the adsorption of chromium species on the active sites of algalbacterial AGS [29].
The effluent Cr(III) concentrations from the two reactors could be obtained by subtracting the effluent Cr(VI) concentration from the effluent total Cr concentration. During Stage III, the effluent Cr(III) concentrations from R1 and R2 were very low, fluctuating between 0.01-0.04 mg/L and 0.01-0.06 mg/L, respectively. Still, it is necessary to further analyze the total Cr (Cr(III) + Cr(VI)) content in the granules, as well as in the EPS samples, to evaluate the effect of salinity on the Cr accumulation capability by the algal-bacterial granules. Further studies are still necessary on the effect of salinity and/or Cr(VI) exposure on the anaerobic P release rate and aerobic P uptake rate, thereby shedding light on the mechanism regarding the varied P removal performance of the algal-bacterial AGS system.

Chromium Removal
In order to assess the effect of salinity on Cr(VI) removal efficiency by the algalbacterial AGS system, the effluent Cr(VI) and Cr(III) concentrations, as well as the Cr(VI) removal efficiency, were monitored during Stage III (Figure 7). It is noteworthy that the Cr(VI) removal rate of R2 was generally lower than that of R1 in Stage III, possibly due to the slightly negative effect of salinity exposure. Specifically, the Cr(VI) removal efficiencies of R1 and R2 fluctuated between 31-51% (37.3% on average) and 28-48% (34.8% on average), respectively (p = 0.005 < 0.05). This observation might be attributable to the inhibitory impact of salinity on microbial activity in the AGS system [28], leading to the decline in Cr(VI) removal efficiency. In addition, the increase in ionic strength under the 1% salinity condition (R2) might have some shielding effect on the surface electrostatic field, thus suppressing the adsorption of chromium species on the active sites of algal-bacterial AGS [29].
The effluent Cr(III) concentrations from the two reactors could be obtained by subtracting the effluent Cr(VI) concentration from the effluent total Cr concentration. During Stage III, the effluent Cr(III) concentrations from R1 and R2 were very low, fluctuating between 0.01-0.04 mg/L and 0.01-0.06 mg/L, respectively. Still, it is necessary to further analyze the total Cr (Cr(III) + Cr(VI)) content in the granules, as well as in the EPS samples, to evaluate the effect of salinity on the Cr accumulation capability by the algal-bacterial granules. Considering the Cr bioremediation mechanisms, the chemical fractionation of Cr in the algal-bacterial AGS (shown in Supplementary Figure S2), as well as surface functional groups (from FTIR spectra in Supplementary Figure S3), were also examined. For the granules in R1 (control, no salinity exposure), organic bound Cr contributed to the largest fraction, about 51.6%, indicating that metal complexation might be the main mechanism involved in Cr(VI) removal by algal-bacterial AGS in this study. This observation is consistent with previous studies [30,31], in which bacterial AGS and algal-bacterial AGS were used to adsorb Cr(III) and Cr(VI), respectively.
As for the granules in R2 (1% salinity exposure), the Fe/Mn oxides bound Cr fraction accounted for the highest percentage, about 52.0%, implying that Fe/Mn oxides in granules play a more important role in binding and adsorption of Cr than other fractions. Compared with the fractionation results of the granules from R1, under 1% salinity, some influence on Cr removal could be discerned, as the granules in R2 exhibited a higher chromium removal and immobilization by Fe/Mn oxides, and lower organic bound and residual fractions. It is possible that under some harsh condition (such as the 1% salinity condition in this study), the two most stable fractions (residual and organic matter bound parts) could be converted to less stable fractions, resulting in the increase in Cr bioavailability. Further studies are necessary on the leaching risk of Cr accumulated in algal-bacterial AGS with prolonged salinity exposure to better understand the effects of Cr-loaded granules on the environment. Moreover, the results from FTIR analysis (Supplementary Figure S3) and the morphological changes on granule surfaces (Supplementary Figure S1) partly confirmed that the functional groups such as carboxyl and phosphate played an important role in the metal complexation and/or chemical precipitation.
Based on the above results, the algal-bacterial AGS cultivated in the SBR system can potentially treat wastewaters containing both Cr(VI) and high salinity during the long operation period. Still, future works are necessary for a clear and insightful explanation of the mechanisms involved in the effects of co-existing Cr(VI) and salinity, which are also included in our follow-up research. For instance, further work on the changes of microbial composition and algae species during the long-term operation of algal-bacterial AGS system could help to clearly elucidate their roles in Cr(VI) removal. Considering the Cr bioremediation mechanisms, the chemical fractionation of Cr in the algal-bacterial AGS (shown in Supplementary Figure S2), as well as surface functional groups (from FTIR spectra in Supplementary Figure S3), were also examined. For the granules in R1 (control, no salinity exposure), organic bound Cr contributed to the largest fraction, about 51.6%, indicating that metal complexation might be the main mechanism involved in Cr(VI) removal by algal-bacterial AGS in this study. This observation is consistent with previous studies [30,31], in which bacterial AGS and algal-bacterial AGS were used to adsorb Cr(III) and Cr(VI), respectively.
As for the granules in R2 (1% salinity exposure), the Fe/Mn oxides bound Cr fraction accounted for the highest percentage, about 52.0%, implying that Fe/Mn oxides in granules play a more important role in binding and adsorption of Cr than other fractions. Compared with the fractionation results of the granules from R1, under 1% salinity, some influence on Cr removal could be discerned, as the granules in R2 exhibited a higher chromium removal and immobilization by Fe/Mn oxides, and lower organic bound and residual fractions. It is possible that under some harsh condition (such as the 1% salinity condition in this study), the two most stable fractions (residual and organic matter bound parts) could be converted to less stable fractions, resulting in the increase in Cr bioavailability. Further studies are necessary on the leaching risk of Cr accumulated in algal-bacterial AGS with prolonged salinity exposure to better understand the effects of Cr-loaded granules on the environment. Moreover, the results from FTIR analysis (Supplementary Figure S3) and the morphological changes on granule surfaces (Supplementary Figure S1) partly confirmed that the functional groups such as carboxyl and phosphate played an important role in the metal complexation and/or chemical precipitation.
Based on the above results, the algal-bacterial AGS cultivated in the SBR system can potentially treat wastewaters containing both Cr(VI) and high salinity during the long operation period. Still, future works are necessary for a clear and insightful explanation of the mechanisms involved in the effects of co-existing Cr(VI) and salinity, which are also included in our follow-up research. For instance, further work on the changes of microbial composition and algae species during the long-term operation of algal-bacterial AGS system could help to clearly elucidate their roles in Cr(VI) removal.

EPS Secretion from Algal-Bacterial Granules
Algal-bacterial granule samples used for EPS extraction and EPS content determination were obtained from the two SBRs at the end of each operation stage. EPS are mainly composed of PN and PS that are secreted by microorganisms, which play an essential role in granular structure and stability [32,33]. The changes in PN and PS contents under different test stages are illustrated in Figure 8.
treat saline wastewater [24,25,34,35]. For instance, Corsino and coworkers [35] found that the osmotic pressure generated by the increasing salinity can stimulate the production of exopolymers by microorganisms. EPS might mitigate the osmotic pressure on the cell wall of bacteria because of their gelatinous matrix.
As the influent Cr(VI) concentration increased from 0 mg/L to 2 mg/L (Stage III), the total EPS content in granules from both reactors increased. However, the extent of their increase was different. Specifically, after exposure to 2 mg/L Cr(VI), the granules from R1 excreted more EPS, which was remarkably increased to 286.64 mg/g-MLVSS. This increase trend continued, reaching 311.53 mg/g-MLVSS even after Cr(VI) exposure ceased ( Figure  8a). On the other hand, in R2, the granular EPS content rose slightly to 259.39 mg/g-MLVSS and 278.73 mg/g-MLVSS during Stages III and IV, respectively. The significant increase of EPS in granules from R1 during Stage III might help protect microorganisms from harsh external condition (e.g., Cr(VI) presence in the influent) [27,32,36]. The granules from R2 may have better tolerance of Cr(VI) under the co-existence of 1% salinity. From Figure 8a, the TB-EPS content showed a similar trend to the total EPS content during the four operation stages, while the LB-EPS content displayed a somewhat different variation trend. This finding implies that TB-EPS is an essential part affecting the algalbacterial granule structure and stability under Cr(VI) exposure condition. On the contrary, the variation of LB-EPS content better matched with the total EPS change than the TB-EPS content in the granules from R2. Therefore, the LB-EPS played a more important role in algal-bacterial granule structure and stability in the presence of salinity, which is in line with a recent study [25].
As shown in Figure 8b, the PN/PS ratio gradually decreased over time under the presence or absence of Cr(VI), suggesting that 1% salinity exposure may trigger more excretion of PS than PN, somewhat helping the granules maintain a more stable structure to As shown in Figure 8a, the total EPS content in the granules from R1 remained stable during Stages I and II. Meanwhile, for the granules in R2, after exposure to 1% salinity (Stage II), a slight increase in the EPS content, from 213.82 90 mg/g-MLVSS to 243.90 mg/g-MLVSS, was detected (Figure 8b). These results indicate that algal-bacterial AGS can produce more EPS to protect themselves from the increasing osmotic pressure due to the increase of influent salinity, in agreement with previous studies using granular sludge to treat saline wastewater [24,25,34,35]. For instance, Corsino and coworkers [35] found that the osmotic pressure generated by the increasing salinity can stimulate the production of exopolymers by microorganisms. EPS might mitigate the osmotic pressure on the cell wall of bacteria because of their gelatinous matrix.
As the influent Cr(VI) concentration increased from 0 mg/L to 2 mg/L (Stage III), the total EPS content in granules from both reactors increased. However, the extent of their increase was different. Specifically, after exposure to 2 mg/L Cr(VI), the granules from R1 excreted more EPS, which was remarkably increased to 286.64 mg/g-MLVSS. This increase trend continued, reaching 311.53 mg/g-MLVSS even after Cr(VI) exposure ceased (Figure 8a). On the other hand, in R2, the granular EPS content rose slightly to 259.39 mg/g-MLVSS and 278.73 mg/g-MLVSS during Stages III and IV, respectively. The significant increase of EPS in granules from R1 during Stage III might help protect microorganisms from harsh external condition (e.g., Cr(VI) presence in the influent) [27,32,36]. The granules from R2 may have better tolerance of Cr(VI) under the co-existence of 1% salinity.
From Figure 8a, the TB-EPS content showed a similar trend to the total EPS content during the four operation stages, while the LB-EPS content displayed a somewhat different variation trend. This finding implies that TB-EPS is an essential part affecting the algalbacterial granule structure and stability under Cr(VI) exposure condition. On the contrary, the variation of LB-EPS content better matched with the total EPS change than the TB-EPS content in the granules from R2. Therefore, the LB-EPS played a more important role in algal-bacterial granule structure and stability in the presence of salinity, which is in line with a recent study [25].
As shown in Figure 8b, the PN/PS ratio gradually decreased over time under the presence or absence of Cr(VI), suggesting that 1% salinity exposure may trigger more excretion of PS than PN, somewhat helping the granules maintain a more stable structure to resist saline conditions. This observation is consistent with the finding from a previous work [25]. One possible explanation for this phenomenon is that PS contains abundant polar functional groups that are often strongly water-binding and can thicken the hydration layer around the cells, mitigating the adverse effect of osmotic pressure generated by salinity [37]. Meanwhile, for the granules from R1, the PN/PS ratio remained stable at 3.66 and 3.68 during Stages I and II, respectively (Figure 8a). However, after exposure to 2 mg/L Cr(VI) (Stage III), their PN/PS ratio increased to 3.92, which fell to 3.76 when Cr(VI) exposure ceased (Stage IV). More excretion of PN than PS was detected from the granules in R1 during Stage III when compared to Stage II, which is probably associated with some exoenzymes contained in EPS that could reduce heavy metal ions [38] as enzymes are mainly composed of proteins. Interestingly, during the four test stages, the granules from both R1 and R2 contained consistently higher PN content than PS content. To summarize, salinity exposure promoted algal-bacterial AGS to secrete more PS, while Cr(VI) exposure stimulated more PN excretion.

Conclusions
The present study examined the effects of salinity on Cr(VI) bioremediation and nutrients removal performance by algal-bacterial AGS. The results from this work showed that algal-bacterial AGS reflected less filamentous and a rougher surface under salinity exposure, with little influence on organics removal being observed. However, the removals of total inorganic nitrogen and total phosphorus were noticeably impacted at the 1% salinity condition, and were further decreased with the co-existence of 2 mg/L Cr(VI). The Cr(VI) removal efficiency, on the other hand, was 31-51% in the control SBR and 28-48% in the test SBR, respectively, indicating that salinity exposure may slightly influence Cr(VI) bioremediation. Furthermore, in this study, it was found that 1% salinity exposure stimulated more PS excretion from algal-bacterial AGS, while PN excretion was enhanced with the introduction of 2 mg/L Cr(VI).