- Email: [email protected]
Balancing carbon, nitrogen and phosphorus concentration in seawater as a strategy to prevent accelerated membrane biofouling
Water Research 165 (2019) 114978
Contents lists available at ScienceDirect
Water Research journal homepage: www.elsevier.com/locate/watres
Balancing carbon, nitrogen and phosphorus concentration in seawater as a strategy to prevent accelerated membrane biofouling Siqian Huang a, 1, Nikolay Voutchkov b, Sunny Jiang a, * a b
Civil and Environmental Engineering, University of California, Irvine, USA Water Globe Consultants, LLC, USA
a r t i c l e i n f o
a b s t r a c t :
Article history: Received 18 April 2019 Received in revised form 22 July 2019 Accepted 12 August 2019 Available online 13 August 2019
Membrane biofouling remains a significant challenge in seawater reverse osmosis desalination for drinking water production. This study investigated nutrient imbalance as the cause of biofouling in labscale experiments and carried out a year-long field-testing at a seawater desalination pilot plant. Lab experiments showed that growth medium with excess of organic carbon (C) but with low nitrogen (N) and phosphorus (P) accelerated the formation of bacterial biofilm. Balancing C to N and P ratios by adding N and P to growth medium increased the proliferation of free-living cells but reduced attached form of bacteria as biofilm. The cell excretion of excess C in the form of extracellular polysaccharides (EPS) was considered as a strategy for nutrient storage for future use. Cell enzyme activity assays indicated some of the bacteria had enhanced enzyme activities to degrade polysaccharides in the absence of organic C in growth medium, possibly using EPS in the biofilm. A year-long field study indicated that accelerated biofouling of seawater reverse osmosis (SWRO) membranes was associated with the elevated content of total organic carbon (TOC) in the intake seawater. Adding N and P to the intake seawater to balance the increase of TOC resulted in reduction of membrane biofouling. Microbial community analysis of the biofouling layer using 16S rRNA gene sequencing indicated biofouling communities varied with seasonal changes. Dosing of N and P did not induce dramatic changes in the fouling microbial community growing on the membrane surface. The outcome of this work implies that membrane biofouling associated with the elevated concentration of TOC in intake seawater is caused by imbalance of C:N:P in the source seawater which occurs often during algal blooms. Addition of N and P to rebalance the nutrients can prevent accelerated SWRO membrane biofouling. © 2019 Elsevier Ltd. All rights reserved.
Keywords: Membrane desalination Reverse osmosis Nutrient imbalance Biofouling
1. Introduction Reverse osmosis (RO) membrane fouling is described as the Achilles heel of the membrane processes for water purification and seawater membrane desalination (Flemming et al., 1997). Although inorganic-, organic- and bio-fouling all contribute to the retardation of the water production and elevation of energy consumption, biofouling and associated organic fouling are the most difficult to predict and control in seawater desalination plants (Ridgway, 1991; Ridgway and Flamming, 1996). Biofouling is caused by biofilm formation on the RO membrane surface by bacteria that have bypassed the pretreatment. The membrane biofilm formation is
* Corresponding author. 844 Engineering Tower, Irvine, CA, 92697, USA. E-mail address: [email protected] (S. Jiang). 1 Current affiliation, Department of Biology, University of Minnesota Duluth. https://doi.org/10.1016/j.watres.2019.114978 0043-1354/© 2019 Elsevier Ltd. All rights reserved.
initiated by organic conditioning of the membrane surface, followed by the attachment of individual bacteria and establishment of micro-colonies before developing into continuous layers of biofilm (Bold and District, 1998; Davies et al., 1993). The availability of organic nutrients, specifically organic carbon (C), seems to be the most important factor that influences the formation rate of biofilm on membranes (Ridgway and Flamming, 1996; Schneider et al., 2005; Sim et al., 2011; Tsai et al., 2004; Vrouwenvelder and Kooij, 2002). In our previous study (Huang et al., 2013), we investigated the influence of different environmental and water quality parameters on membrane biofilm formation using a flat-sheet membrane system fed with a side-stream of ultrafiltration pretreated seawater. The tests have shown that the increase of organic C concentration in the natural feed seawater (from ~0.5 mg/L to above 1.5 mg/L) has greatly affected the number of attached cells and the biofilm thickness on RO membranes over the one-year study period (Huang et al., 2013). Yet, the organic C
2
S. Huang et al. / Water Research 165 (2019) 114978
concentration that causes accelerated biofouling in seawater (~1e1.5 mg/L) is only ~10% of that commonly found in treated sewage effluent (~8e15 mg/L) used for reclamation by RO membrane process (OCSD, 2016). Biofouling of RO membranes for wastewater reclamation under such condition is not accelerated. These results suggest that in addition to the increase of organic C concentration, other factors also play a role in accelerated biofilm formation on RO membranes for desalinating seawater. Contrary to the nutrient enrichment observation in the membrane biofilm formation in the water treatment industries, the long-standing theory of biofilm formation in clinically important microorganisms is that biofilm formation is triggered by unfavorable conditions including nutrient starvation (O'Toole et al., 2000; Wai et al., 1998). The disparity between such seemingly opposite observations may be due to the nutrient composition required for bacterial growth and biofilm production. Specifically, the ratio of organic C to nitrogen (N) and to phosphorus (P) may be important in biofilm formation. Biofilm is an organic film composed of live and dead microorganisms embedded in a polymer matrix, known as extracellular polymeric substances (EPS) (Characklis and Marshall, 1990; Wingender et al., 1999). Although proteins and lipids are commonly observed in biofilm retrieved from wastewater treatment facilities and membrane bioreactors, polysaccharides are the main component of marine biofilms (Decho, 2000). The growth of marine bacteria depends on the availability of dissolved organic C as an energy and structural C source, N and P for replication of nucleic acid, synthesis proteins and cell wall for cell division (Thompson et al., 2006). The production and secretion of EPS, which plays an important role at the initial stage of biofilm formation, are at the expense of cellular energy. EPS production also increases the demand of bioavailable organic C in competition for essential cellular functions and organelle synthesis. Therefore, from a microbial ecology point of view, EPS production may be a mechanism to store excess organic C that cannot be utilized by the cell for replication due to the lack of N and P (limiting nutrients) to support cell proliferation. This postulation was supported by work by Stuart et al. who showed that cyanobacteria could reuse extracellular organic C in microbial mats and suggested EPS as a pool of extracellular nutrient resources (Stuart et al., 2016). This function may be more common in marine environments, where N is limited even during the post algal bloom periods (Romanı, 2009). Previous studies on bacterial biofilm growth have indirectly identified the C:N:P ratio as one of the key factors for EPS production and biofilm formation (Thompson et al., 2006; Allan et al., 2002; Fang et al., 2009). For example, Huang et al. have reported that polysaccharide secretion by biofilm forming E. coli DH5a increased significantly at high C:N ratio (10:1) (Huang et al., 1994). Thompson et al. have tested the effects of the C:N:P ratio on biofilm formation using Enterobacter and Citrobacter. They have found that higher C:N:P ratios (334:28:5.6) have resulted in significantly higher numbers of attached cells in both bacterial genera (Thompson et al., 2006). The adherence property of the plant pathogen Agrobacterium tumefaciens was enhanced under P limitation, despite the lower cell growth rate (Danhorn et al., 2004). The low P concentration was shown to trigger the two-component regulatory system PhoR-PhoB, which promoted the biofilm formation (Danhorn et al., 2004). These studies suggest that lower C:N:P ratios may reduce EPS production and the biofilm formation by bacteria. Past studies have attempted to reduce organic C loading to desalination RO feed seawater to achieve low C:N:P ratios in order to prevent membrane biofouling (Huang et al., 2013). However, the existing technologies for desalination pretreatment, including coagulation, flocculation, microfiltration, and ultrafiltration are only effective at removing particulates in water but are inefficient
to remove dissolved organic C. In fact, pressure-driven filtrations, such as ultrafiltration, have been shown to break algal cells during algal blooms to release additional dissolved organic C to support the excess biofilm production (Jiang and Voutchkov, 2014). Thus, the alternative approach to balance organic C:N:P ratio is to increase N and P concentration in the treated seawater to match the C content without removing it from the feed seawater. In this study, we have first examined bacterial isolates for biofilm production under different organic C:N:P composition. We then investigated bacterial physiological changes through upregulation of enzyme activities triggered by changes of the C:N:P composition in growth media. In order to develop a practical methodology for modifying organic C:N:P ratio for desalination membrane fouling prevention, we performed a year-long study at a desalination pilot facility, which included dosing of N and P to the pretreated intake seawater to match the content of organic C in the water and thereby rebalance the C:N:P ratio.
2. Materials and methods 2.1. Impact of the C:N:P ratio on bacterial growth and biofilm production Six bacteria (B1 to B6) isolated from biofouled RO membrane and cartridge filters from the Carlsbad Desalination Pilot Plant were used in this study. They were identified using 16S rRNA gene sequencing in a previous study (Zhang et al., 2012). B1 and B4 were identified as Shewanella sp.; B2 and B3 were nearly identical in 16S rRNA gene sequence and were matched to Alteromonas sp. B5 was identified as Cellulophaga sp. and B6 matched Vibrio sp. in the RDP database (Zhang et al., 2012). The bacteria were isolated on artificial seawater medium formulated by John Paul (ASWJP) (Paul, 1982) supplemented with 1 mg/L peptone and 0.5 mg/L yeast extract (PY). The bacteria were grown overnight at room temperature in ASWJP þ PY with shaking at 30 rpm, harvested by centrifugation at 10000 rpm for 1 min. The cell pellet was washed with ASWJP first to remove any remaining organic nutrient before it was used for biofilm production experiments. To characterize bacterial growth and biofilm formation under preset C:N:P conditions, sodium acetate (NaAc) was used as the single organic C source in the biofilm formation experiments. The inorganic N and P concentrations in the ASWJP were modified to match the typical concentrations found in the Pacific coast seawater (called base concentration in this study). The C:N:P concentrations were varied based on the desired ratio, as shown in Table 1, in two sets of experiments. The highest organic C concentration chosen (8 mg/L) in the experiments represents the high total organic carbon (TOC) loading condition in the Southern California Pacific coast seawater (e.g. post algal bloom or after a heavy storm that carries pollution from urban runoff). The first set of lab experiments focused on holding the N and P Table 1 Defined substrate nutrient concentration and nutrient ratio tested. C:N:P ratio
C (NaAc-C) mg/L
Experiment I 1:1:1 0.4 10:1:1 4 20:1:1 8 Experiment II 20:1:1 8 20:5:5 8 20:10:10 8 20:20:20 8
N (NH4NO3eN) mg/L
P (Na2HPO4eP) mg/L
0.2 0.2 0.2
0.05 0.05 0.05
0.2 1 2 4
0.05 0.25 0.5 1
S. Huang et al. / Water Research 165 (2019) 114978
concentrations at a constant value, while increasing the organic C concentration from 0.4 mg/L (1x) to 4 mg/L (10x) and to 8 mg/L (20x). Experiment II focused on holding the organic C concentration at 8 mg/L (biofouling condition) while increasing the concentration of N and P from the base level of 0.2 mg/L and 0.05 mg/L, respectively, to 5x, 10x and 20x the base concentration (Table 1). The cell growth and biofilm biomass assays were performed in clear 96-well microplates (round bottom, 330 ml/well, VWR International Corp. Cat # 734e1546). Briefly, washed cell pellets were resuspended and diluted in 1:10 ratio in each modified ASWJP medium as shown in Table 1. Each individual bacterial suspension was inoculated into 8 wells (8 replicates) of the microplate. The 8 wells in the last column of the plate were inoculated with medium only and were used as the negative control. Two identical plates were prepared: one was used for cell growth measurement by optical density using absorbance wavelength of 550 nm, and the other was used for biofilm quantification using the crystal violet assay method (Alhadidi et al., 2012). Crystal violet biofilm biomass assay was performed following the protocol of Stephanovic et al. (Stepanovic et al., 2000). Briefly, bacterial biofilm on the plate was fixed with methanol, stained with crystal violet (2% Hucker crystal violet used for Gram stain), and then released with the bound dye using 33% glacial acetic acid. The optical density of the released solution was measured at absorbance wavelength of 590 nm. Microplates were incubated at 24 C and were read using Spectramax Plus 384 (Molecular Devices, Sunnyvale, CA, USA). The average value of eight control wells was used to normalize the average value of each set of 8 wells for individual bacterial strain.
2.2. Impact of the C:N:P ratio on bacterial enzyme activities Bacterial enzyme activity assays were performed to examine the physiological responses of biofouling bacteria as a result of changing the C:N:P ratio. Enzyme assays were carried out using fluorometric assay protocol adapted from (Saiya-Corka et al., 2002) in black 96-well microplates (round bottom, 330ml/well, Fisher Scientific, Cat # 07-200-627) for b-glucosidase (BG), b-N-
3
acetylglucosaminidase (NAG), leucine aminopeptidase (LAP), and alkaline phosphatase (AP). Briefly, cells were grown in 96 well plates in ASWJP þ PY for 48 h under the same condition as for growth and biofilm assay. The growth medium together with suspended cells was then carefully removed from each well, leaving only the biofilm attached to the plate. The wells with biofilm were carefully washed twice using ASWJP to remove residual nutrient medium before they were filled with ASWJP without adding any organic C. The plates were incubated for two more days at room temperature. The enzyme activities were first measured in bacterial biofilm grown in rich complex organic nutrient medium immediately after the removal of unattached cells and nutrient medium (time zero), and then again at 24 h and 48 h when ASWJP was used as the only source of nutrients without any addition of organic C. The parallel clear 96-well microplates managed under the same condition were used to quantify cell density and biofilm thickness at each time point when enzyme assay were performed.
2.3. Pilot-scale testing of C:N:P ratio on membrane biofouling A year-long field study was set up at the Carlsbad Desalination Pilot facility. Two flat sheet bio-monitoring systems were operated in parallel with the spiral wound RO membrane train of the pilot plant using ultrafiltration-pretreated RO feed. Fig. 1 shows the schematic set up of the testing system. The system was designed to change the C:N:P ratio through dosing of N and P rather than reducing natural level of organic C in feed water. Unit 1 is a control membrane without dosing of N and P; while unit 2 is dosed with NH4NO3 and Na2HPO4 through a pump to final concentrations of 10 mg/L and 2 mg/L, respectively. No additional organic C was added to the feed seawater, thus the TOC concentration in this water reflected the natural variability of TOC in the source seawater. The TOC of the RO intake water was measured using the GE Sievers 5310 C On-Line TOC Analyzer (GE Instruments, Boulder, Colorado) as described by (Huang et al., 2013). The bio-monitor consists of a 5 20 cm flat sheet SWC5 membrane (Hydranautics) placed directly on top of a plastic
Fig. 1. Schematics of bio-monitor system at Carlsbad desalination pilot plant.
4
S. Huang et al. / Water Research 165 (2019) 114978
support; the feed channel was separated by a single piece of 0.87 mm thick nylon spacer. The cross section area of the feed channel is 44 mm2. The system was run at cross flow rate of 0.023 m3 h1 using 42 kPa feed pressure without permeation. The flat sheet bio-monitor was designed to simulate the cross flow velocity of the seven spiral wound RO membrane elements in the pressure vessels of the pilot plant. The membrane was removed from the monitor for analysis every two weeks and was replaced with a new membrane each time. After each removal, the system was cleaned and rinsed with clean tap water. The system was run at the Carlsbad desalination pilot facility from June 2010 to July 2011. 2.4. Determination of membrane biofilm density and thickness The retrieved flat sheet membranes from the bio-monitor were examined for biofilm thickness and total bacterial cell counts on the membrane surface by confocal laser scanning microscopy (CLSM, Zeiss LSM 510 META). Briefly, membranes were cut into 1 cm 1 cm squares and stained using SYTO 9 green-fluorescent nucleic acid stain for detecting live bacteria and propidium iodide red-fluorescent nucleic acid stain (FilmTracer™ LIVE/DEAD Biofilm Viability Kit, Invitrogen, Carlsbad, CA) for detecting dead bacteria. The stained membranes were then mounted onto a glass slide and observed under the CLSM. Excitation/emission wavelengths of 488nm/500 nm were used for SYTO 9 and 510nm/635 nm were used for propidium iodide, respectively. Images captured at each wavelength were composited into one final image. The Z sectioning method was used to determine the thickness of the biofilm. The thickness was calculated from 10 images captured cross the diagonal section of the membrane by locating the highest point above each pixel in the bottom layer containing biomass (Heydorn et al., 2000). Bacterial colonization was evaluated by visually counting the number of cells attached to the membranes surface, and was determined by the average counts from 10 images for each sample. 2.5. Analysis of bacterial community on the membranes To identify the bacteria colonizing the membrane surfaces, four sets membrane samples (4 controls, and 4 dosed with N and P) were collected each in autumn, winter, spring, and summer season and were used for bacterial community analysis. Approximately 2 2cm area of each membrane surface was swabbed for biofilm biomass. The genomic DNA was extracted and used for PCR of 16S rRNA gene using primers targeting the V7eV9 (939F-1492R) region. Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) (Roche, NJ, USA) was performed by the Research and Testing Laboratory (TX, USA). The sequences from pyrosequencing were analyzed using QIIME (Quantitative Insights Into Microbial Ecology) Pipeline (Caporaso et al., 2010). Low-quality sequences (<25) and short sequences (<200bp) were filtered and removed. Sequences with >97% similarity were clustered and defined as an Operational Taxonomic Unit (OTU). Representative sequences from each OTU were aligned and assigned to taxonomic identity using RDP classifier (Cole et al., 2005). The microbial communities were compared using the taxonomic and phylogenetic assignments. Beta diversity was plotted in Principal Coordinate Analysis (PCoA) to identify the community diversity between membrane samples. 3. Results 3.1. Impact of C:N:P ratio on bacterial growth and biofilm production The cell density and biofilm biomass in response to increased organic C in the growth medium are shown in Fig. 2. Changes in cell
and biofilm densities are expressed as the ratio of the measurements at a specific nutrient condition to those at the base nutrient concentration of C:N:P (1:1:1), which is set as 1. With the increase of C (by adding NaAc) concentration from 1x (0.4 mg/L) to 20x (8 mg/L) in growth medium, the cell density (OD550 ratio) had minimal changes (p > 0.01). This result showed although bacterial isolates could use NaAc as the single organic C source, the cell growth was limited by other nutrients. However, the biofilm biomass (OD590 ratio) increased significantly (p < 0.01) in bacteria B1 and B2 when organic C in the growth medium increased from 0.4 mg/L to 8 mg/L (20x) (Fig. 2). This result indicated these two strains actively converted organic C to biofilm content (EPS) in the presence of high concentration of organic C in the medium. Fig. 3 shows the results from experiment II, in which N and P concentrations increased from 1x (0.2 mg/L NH4NO3, 0.05 mg/L Na2HPO4, respectively) to 20x with the fixed NaAc concentration at 20x (8 mg/L). Increases in P and N concentrations resulted in increases in bacterial growth. The cell density was about 30% higher at 20x N, P concentration than those at base N, P concentration, suggesting N, P additions promoted the proliferation of free living cells. However, the biofilm biomass (OD590) decreased significantly (p < 0.01) for four of six bacteria tested (strain B1, B2, B4 and B5) at the higher N, P concentrations. There was a dramatic drop (more than 50%) in biofilm biomass once the N, P reached 5x of the base concentration for bacteria B1, B4 and B5. Bacteria strain B3 and B6 generated low level of biofilm at all N, P concentrations.
3.2. Enzymatic activities in the presence of low organic carbon Bacterial enzymatic activities were measured to determine bacteria's ability to use biofilm as nutrient when organic C was removed from the seawater (growth medium). The bacterial enzyme activities following the removal of organic C are presented in Fig. 4. Up-regulation of all four degradation enzymes was observed for bacterial B2. B3 also displayed enhanced activities for b-glucosidase (BG) and leucine aminopeptidase (LAP). Enzyme activities decreased in bacterial B4, and B5 only had elevation in leucine aminopeptidase (LAP) activity. These results indicated that some bacteria could utilize the nutrients stored in the biofilm for cell activities but not all bacteria could.
3.3. Field experiments with bio-monitoring system The year-long monitoring of water quality and biofouling potential at the Carlsbad desalination pilot plant showed that TOC of the feed seawater varied between 0.03 and 2 mg/L (30 ppb and 2000 ppb). Increases in TOC concentration in the intake raw seawater were observed several times over the year. These increases were mainly associated with land runoffs after rainfall events (Fig. 5). Higher biofilm thickness and cell counts were observed on the control (U1) membranes that were collected on the dates with elevated organic C in the feed water as indicated by the upward trend in fitted lines in Fig. 5. In comparison, membranes recovered from bio-monitor unit 2 (U2), where N and P were dosed into the feed water, did not experience dramatic increases in biofilm thickness or total cell counts as indicated by the relatively flat fitted line (Fig. 5). Both the biofilm thickness and cell counts on U2 membranes were fairly constant over the one-year study period. CLSM images (Fig. 6) also revealed that there was an observable difference between the flat sheet RO membrane retrieved from the control unit U1 (Fig. 6A) and the one from the N, P dosed unit U2 (Fig. 6B) during the period of high TOC measurements in feed water (Jan. 6, 2011).
S. Huang et al. / Water Research 165 (2019) 114978
5
Fig. 2. Changes in total bacterial cell density and biofilm biomass following the changes of organic carbon concentration in culture medium. The base concentrations of C:N:P (1:1:1) were set at 0.4 mg/L NaAc, 0.2 mg/L NH4NO3 and 0.05 mg/L Na2HPO4, respectively. The cell density and biofilm biomass at 10x (4 mg/L) and 20x (8 mg/L) C concentration were compared with the base level measurements (Cellnew/Cellbase and Biofilmnew/Biofilmbase). Error bars represent the standard deviation among eight experimental replicates.
3.4. Analysis of biofilm bacteria community on RO membranes by 16S rRNA gene The relationship between membrane biofouling, the bacterial community and the change of water quality (seasonality) was determined using 16S rRNA gene sequence analyses (Fig. 7). A total of 18728 qualified pyrosequencing reads of the V7eV9 regions of the 16S rRNA gene were obtained from the eight membrane samples collected between September 2010 and June 2011. Sequences were clustered into a total of 1778 OTUs by complete linkage clustering. Among them, >95% OTUs matched RDP database and were assigned the taxonomic identities (Table S1). The bacterial composition varied among samples retrieved from different seasons. Of the 1206 OTUs that were assigned taxonomy data, only 15 of them were observed on all membranes, and 396 of them were presented in a single set of samples. Taxonomy assignment for the top 10% most abundant classes of microorganism is shown in Fig. 7. Samples are labeled using membrane retrieval dates. For example, U1.9.16.10 is the biomass from membrane unit 1 (untreated control unit) retrieved on Sept. 16, 2010 and U2.9.16.10 is the biomass from membrane unit 2 (treated by N, P dosing) retrieved on the same date. Alpha-,
Gammaproteobacteria and Flavobacteria dominated membrane biofilm communities, with the combined abundance of 60e80% in all samples. The most severe membrane fouling was observed on U1.3.2.11 that was retrieved during the period of high TOC concentration in the feed water. The abundance of Gammaproteobacteria increased significantly on this membrane in comparison with other membranes. PCoA plot of beta diversity showed microbial biofilm composition was unique for each season (Fig. 8). Biofouling communities collected during autumn, winter, spring and summer seasons were spread across the four different coordinates of the PCoA plot. However, samples collected from control unit (U1) and treatment unit (U2) were clustered together, indicating nutrient dosing did not change the biofilm composition significantly but reduced the level of biofouling. 4. Discussion The results from bench-top experiments using variable C:N:P composition media (seawater) for cultivating marine biofouling bacteria support the hypothesis that excess organic C to N and P promotes the formation of microbial biofilm by marine bacteria. On
6
S. Huang et al. / Water Research 165 (2019) 114978
Fig. 3. Changes in total bacterial cell density and biofilm biomass following the changes of nitrogen and phosphorous concentration in culture medium. The cell density and biofilm biomass at base concentrations of C:N:P (1:1:1) were used as the comparison with measurements taken at the 20x base level of C (8 mg/L) and increasing concentration of N and P from the base to 20x the base concentration (Cellnew/Cellbase and Biofilmnew/Biofilmbase). Error bars represent the standard deviation among eight experimental replicates.
the other hand, increasing N, P concentration in the medium in the presence of bioavailable organic C promotes cell proliferation but reduces biofilm formation. We attribute the reduction of biofilm to the utilization of organic C in support of cell growth in the presence of N and P (for synthesis of proteins, lipids and the other cell components) based on our knowledge on nutrient compositions and bacterial growth (Touratier et al., 1999). In the absence of sufficient N and/or P (limiting nutrients) to support protein synthesis and cell proliferation, we observed bacterial excretion of the excess organic C as biofilm in both the bench-scale experiments and the pilot-scale membrane fouling studies. The bacterial enzyme activity assay results in part support the hypothesis that the biofilm formation may be a nutrient storage mechanism for future use in low-carbon conditions. This result agrees with a previous study demonstrating trace isotopically labeled EPS uptake into single cells of cyanobacteria (Stuart et al., 2016). Although we were not able to directly trace the uptake of C from biofilm EPS into cell biomass, the results of enzyme activity assays indicated that at least some of the bacteria could utilize nutrients in biofilm for active growth. Future research, such as isotopically labeled C tracer studies, is necessary to directly assess the role of biofilm as potential nutrient source for bacterial metabolic function under different environmental conditions.
The field experiments using two parallel tests with RO feed stream from a desalination pilot plant further confirmed the importance of the C:N:P ratio in formation of membrane biofilm. The year-long study confirmed that TOC increase in source water was the cause to trigger accelerated biofouling. This observation agrees with previous studies showing that elevation of organic C is the cause of biofilm production and a major culprit for membrane biofouling (Ridgway and Flamming, 1996; Schneider et al., 2005; Sim et al., 2011; Tsai et al., 2004; Vrouwenvelder and Kooij, 2002). The field experiments in the pilot plant demonstrated for the first time that N and P dosing in the feed stream could reduce biofouling on membrane surface. The biofilm formation on membrane surface under the shear of flow should be viewed as a dynamic process, in which biofilm Accumulation ¼ Production e Detachment þ Attachment (Hunt et al., 2004). The rate of direct bacterial attachment on the membrane was expected to be low because the bacterial concentration in the UF pretreated feed water was low (<103/L) (Huang et al., 2013). Thus the lower density of biofilm observed in the membrane unit with balanced C:N:P seawater feed is attributed to the slower rate of biofilm formation. Our observations showed that the addition of N and P to balance the C content in the source seawater reduced the available organic
S. Huang et al. / Water Research 165 (2019) 114978
7
Fig. 4. Enzyme activity of bacteria in biofilm 24 h (a) and 48 h (b) after removal of organic C from the culture medium. BG: b-glucosidase; NAG: b-N-acetylglucosaminidase; AP: alkaline phosphatase; LAP: leucine aminopeptidase. Enzyme activity is expressed as the ratio of measured activity at times to the initiate activity immediately after the removal of organic carbon (Et:E0) from the culture medium. Values greater than 1 indicate up-regulation of enzyme activity (values in the blue box).
Fig. 5. Plots of biofilm thickness (a) and total bacterial cell counts (b) on the surface of the RO membrane vs. TOC concentration in raw RO feed water. Membrane U1 (control unit) and U2 (membrane treated with N, P dosing) were collected from flat sheet biomonitor systems.
C for EPS production (as also demonstrated in the bench scale experiments). Accelerated biofilm detachment or sloughing under balanced C:N:P concentration could also play a part in the change of biofilm quantity, which was documented in a study of nutrientinduced P. aeruginosa biofilm dispersion (Sauer et al., 2004). This study showed that the increased concentration of inorganic N triggered up-regulation of flagella gene expression, which led to enhanced biofilm detachment under flow shear (Sauer et al., 2004). Overall, the reduced biofilm thickness and cell counts on N, P treated membrane unit in the field experiment coincided with the results observed in the laboratory tests of bacterial growth and biofilm production in defined C:N:P nutrient media. Both results suggest balancing the ratio of total organic C with the addition of N and P can be a strategy to reduce membrane biofouling during periods with excess organic loading. However, the practical application of such approach is seriously challenging because the excess N and P in the feed stream will end up in the brine reject and eventually be discharged back to the ocean. The excess nutrient loading could trigger new algal blooms and other negative environmental consequences. A nutrient trapping technology in brine reject, e.g. using a biofiltration system, will be needed before N and P dosing can be considered for membrane fouling mitigation. The RO membrane biofouling microbial community analyses
results agree with previous work indicating that Gammaproteobacteria and Alphaproteobacteria are the most dominant bacteria classes of membrane surface biofilm (Zhang et al., 2012). Seasonal shifts of biofouling microbial communities on membrane surface reflect the well-known seasonal succession of marine bacterial plankton (Bunse and Pinhassi, 2017). Complex physical, chemical, and biological factors influence marine microbial communities in time and space. Differential adaptation or sensitivity to variability in these factors is likely to determine the dynamics of the functional and taxonomic/phylogenetic diversity of bacterioplankton. Manual dosing of N and P in the feed water can also influence the microbial composition. For example, Ovreas et al. have shown that the addition of organic C, N and P in seawater lead to the increase in population of Gammaproteobacteria and a smaller increase in Alphaproteobacteria (Ovreas et al., 2003). The results of the current study indicate the changes of microbial composition due to dosing of N, P are less significant than the seasonal shift of natural microbial community in feed water. The diverse microbial composition recovered from the membrane surface also indicates that there is no single or a few bacterial species that are responsible for the biofilm formation. The reduction of membrane biofilm using N, P dosing implies that such management action has a broad effect in suppressing the biofilm
8
S. Huang et al. / Water Research 165 (2019) 114978
Fig. 6. Confocal laser scanning microscopy images of flat sheet membranes collected from bio-monitor. A: 2-D image of membrane in Unit 1 (control) collected on 1/6/2011; B: 2-D image of membrane in Unit 2 (dosed with N and P) collected on 1/6/2011.
formation by a diverse biofouling community. Further investigation, such as gene expression analysis, could shed new lights on the molecular mechanism of the N, P dosing on biofilm formation. 5. Conclusions
Fig. 7. The class level taxonomy assignments of biomass retrieved from control (U1) and treatment membrane (U2) in bio-monitors. Membranes are labeled based on the date of collection.
Results of laboratory studies showed that dosing N and P in the presence of high organic C promoted replication of marine bacteria but reduced biofilm formation. When organic C was removed from the culture medium (source seawater), some of the marine bacteria could utilize nutrients in the biofilm for growth (as indicated by elevated enzyme activities). A year-long field study demonstrated that dosing N and P to feed seawater could reduce the thickness of biofilm on the membrane surface during conditions of elevated TOC concentration in the feed water. Analysis of biofilm bacterial community revealed that the biofouling bacteria were influenced by the seasonal changes in water quality. N and P dosing did not change the fouling microbial composition significantly but suppressed biofilm production.
Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgement Financial support for this work was provided by WateReuse Foundation award WRF-08-19. Authors thank the following organizations and individuals for their support to this project: Daniel Marler and Steve LePage at Carlsbad Desalination Pilot facility; Steven Peck at Hydranautics. Technical assistance from Matthew Linder for bacterial enzyme activity assay is acknowledged. Appendix A. Supplementary data Fig. 8. 2D Principal Coordinate Analysis (PCoA) plot of microbial biofilm community on eight membrane samples.
Supplementary data to this article can be found online at https://doi.org/10.1016/j.watres.2019.114978.
S. Huang et al. / Water Research 165 (2019) 114978
References Alhadidi, A., Kemperman, A.J.B., Schurer, R., Schippers, J.C., Wessling, M., van der Meer, W.G.J., 2012. Using SDI, SDIþ and MFI to evaluate fouling in a UF/RO desalination pilot plant. Desalination 285, 153e162. Allan, V.J.M., Callow, M.E., Macaskie, L.E., Paterson-Beedle, M., 2002. Effect of nutrient limitation on biofilm formation and phosphatase activity of a Citrobacter sp. Microbiology 148 (1), 277e288. Bold, R.M., District, O.C.W., 1998. Influence of Molecular Conditioning Films on I Microbial Colonization of Synthetic Membranes-Determined by Internal Reflection Spectrometry. NWRI. Bunse, C., Pinhassi, J., 2017. Marine bacterioplankton seasonal succession dynamics. Review Special Series: Micro Commun. 25 (6), 494e505. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., Fierer, N., Pena, A.G., Goodrich, J.K., Gordon, J.I., 2010. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7 (5), 335e336. Characklis, W.G., Marshall, K.C., 1990. Biofilms. Wiley, New York. Cole, J.R., Chai, B., Farris, R.J., Wang, Q., Kulam, S., McGarrell, D., Garrity, G.M., Tiedje, J.M., 2005. The Ribosomal Database Project (RDP-II): sequences and tools for high-throughput rRNA analysis. Nucleic Acids Res. 33 (Suppl. 1), D294eD296. Danhorn, T., Hentzer, M., Givskov, M., Parsek, M.R., Fuqua, C., 2004. Phosphorus limitation enhances biofilm formation of the plant pathogen Agrobacterium tumefaciens through the PhoR-PhoB regulatory system. J. Bacteriol. 186 (14), 4492e4501. Davies, D.G., Chakrabarty, A.M., Geesey, G.G., 1993. Exopolysaccharide production in biofilms - substratum activation of alginate gene-expression by pseudomonasaeruginosa. Appl. Environ. Microbiol. 59 (4), 1181e1186. Decho, A.W., 2000. Microbial biofilms in intertidal systems: an overview. Cont. Shelf Res. 20 (10e11), 1257e1273. Fang, W., Hu, J., Ong, S., 2009. Influence of phosphorus on biofilm formation in model drinking water distribution systems. J. Appl. Microbiol. 106 (4), 1328e1335. Flemming, H., Schaule, G., Griebe, T., Schmitt, J., Tamachkiarowa, A., 1997. Biofouling the Achilles heel of membrane processes. Desalination 113, 215e225. Heydorn, A., Nielsen, A.T., Hentzer, M., Sternberg, C., Givskov, M., Ersboll, B.K., Molin, S., 2000. Quantification of biofilm structures by the novel computer program COMSTAT. Microbiology 146 (Pt 10), 2395e2407. Huang, C.T., Peretti, S.W., Bryers, J.D., 1994. Effects of medium carbon-to-nitrogen ratio on biofilm formation and plasmid stability. Biotechnol. Bioeng. 44 (3), 329e336. Huang, S., Voutchkov, N., Jiang, S.C., 2013. Investigation of environmental influences on membrane biofouling in a Southern California desalination pilot plant. Desalination 319, 1e9. Hunt, S.M., Werner, E.M., Huang, B., Hamilton, M.A., Stewart, P.S., 2004. Hypothesis for the role of nutrient starvation in biofilm detachment. Appl. Environ. Microbiol. 70 (12), 7418e7425. Jiang, S.C., Voutchkov, N., 2014. Investigation of Desalination Membrane Biofouling. Water Reuse Foundation. O'Toole, G., Kaplan, H.B., Kolter, R., 2000. Biofilm formation as microbial development. Annu. Rev. Microbiol. 54 (1), 49e79.
9
OCSD, 2016. GWRS Final Expansion - Final Implementation Plan. Project No. SP-173, Effluent Reuse Study. https://www.ocwd.com/media/5119/sp-173-vol1.pdf. Ovreas, L., Bourne, D., Sandaa, R.A., Casamayor, E.O., Benlloch, S., Goddard, V., Smerdon, G., Heldal, M., Thingstad, T.F., 2003. Response of bacterial and viral communities to nutrient manipulations in seawater mesocosms. Aquat. Microb. Ecol. 31 (2), 109e121. Paul, J.H., 1982. Use of Hoechst dyes 33258 and 33342 for enumeration of attached and planktonic bacteria. Appl. Environ. Microbiol. 43 (4), 939e944. Ridgway, H.F., 1991. Bacteria and membranes - ending a bad relationship. Desalination 83 (1e3), 53-53. Ridgway, H., Flamming, H.-C., 1996. Water Treatment Membrane Processes. McGraw-Hill, New York. Romanı, A.M., 2009. Freshwater biofilms. Biofouling 137. Saiya-Corka, K.R., Sinsabaugha, R.L., Zakb, D.R., 2002. The effects of long term nitrogen deposition on extracellular enzyme activity in an Acer saccharum forest soil. Soil Biol. Biochem. 34, 1309e1315. Sauer, K., Cullen, M., Rickard, A., Zeef, L., Davies, D., Gilbert, P., 2004. Characterization of nutrient-induced dispersion in Pseudomonas aeruginosa PAO1 biofilm. J. Bacteriol. 186 (21), 7312e7326. Schneider, R., Ferreira, L., Binder, P., Bejarano, E., Goes, K., Slongo, E., Machado, C., Rosa, G., 2005. Dynamics of organic carbon and of bacterial populations in a conventional pretreatment train of a reverse osmosis unit experiencing severe biofouling. J. Membr. Sci. 266 (1), 18e29. Sim, L.N., Ye, Y., Chen, V., Fane, A.G., 2011. Investigations of the coupled effect of cake-enhanced osmotic pressure and colloidal fouling in RO using crossflow sampler-modified fouling index ultrafiltration. Desalination 273 (1), 184e196. Stepanovic, S., Vukovic, D., Dakic, I., Savic, B., Svabic-Vlahovic, M., 2000. A modified microtiter-plate test for quantification of staphylococcal biofilm formation. J. Microbiol. Methods 40 (2), 175e179. Stuart, R.K., Mayali, X., Lee, J.Z., Everroad, R.C., Hwang, M., Bebout, B.M., Weber, P.K., Pett-Ridge, J., Thelen, M.P., 2016. Cyanobacterial reuse of extracellular organic carbon in microbial mats. ISME J. 10, 1240e1251. Thompson, L., Gray, V., Lindsay, D., Von Holy, A., 2006. Carbon: nitrogen: phosphorus ratios influence biofilm formation by Enterobacter cloacae and Citrobacter freundii. J. Appl. Microbiol. 101 (5), 1105e1113. Touratier, F., Ledendre, L., Vezina, A., 1999. Model of bacterial growth influenced by substrate C:N ratio and concentration. Aquat. Microb. Ecol. 19, 105e118. Tsai, Y., Pai, T., Qiu, J., 2004. The impacts of the AOC concentration on biofilm formation under higher shear force condition. J. Biotechnol. 111 (2), 155e167. Vrouwenvelder, J.S., Kooij, D.V.D., 2002. Diagnosis of fouling problems of NF and RO membrane installations by a quick scan. Desalination 153, 121e124. Wai, S.N., Mizunoe, Y., Takade, A., Kawabata, S.I., Yoshida, S.I., 1998. Vibrio cholerae O1 strain TSI-4 produces the exopolysaccharide materials that determine colony morphology, stress resistance, and biofilm formation. Appl. Environ. Microbiol. 64 (10), 3648e3655. Wingender, J., Neu, T.R., Flemming, H.C., 1999. Microbial Extracellular Polymeric Substances: Characterization, Structure, and Function. Springer Verlag. Zhang, M.Y., Zhang, K.S., De Gusseme, B., Verstraete, W., 2012. Biogenic silver nanoparticles (bio-Ag-0) decrease biofouling of bio-Ag-0/PES nanocomposite membranes. Water Res. 46 (7), 2077e2087.
Contents lists available at ScienceDirect
Water Research journal homepage: www.elsevier.com/locate/watres
Balancing carbon, nitrogen and phosphorus concentration in seawater as a strategy to prevent accelerated membrane biofouling Siqian Huang a, 1, Nikolay Voutchkov b, Sunny Jiang a, * a b
Civil and Environmental Engineering, University of California, Irvine, USA Water Globe Consultants, LLC, USA
a r t i c l e i n f o
a b s t r a c t :
Article history: Received 18 April 2019 Received in revised form 22 July 2019 Accepted 12 August 2019 Available online 13 August 2019
Membrane biofouling remains a significant challenge in seawater reverse osmosis desalination for drinking water production. This study investigated nutrient imbalance as the cause of biofouling in labscale experiments and carried out a year-long field-testing at a seawater desalination pilot plant. Lab experiments showed that growth medium with excess of organic carbon (C) but with low nitrogen (N) and phosphorus (P) accelerated the formation of bacterial biofilm. Balancing C to N and P ratios by adding N and P to growth medium increased the proliferation of free-living cells but reduced attached form of bacteria as biofilm. The cell excretion of excess C in the form of extracellular polysaccharides (EPS) was considered as a strategy for nutrient storage for future use. Cell enzyme activity assays indicated some of the bacteria had enhanced enzyme activities to degrade polysaccharides in the absence of organic C in growth medium, possibly using EPS in the biofilm. A year-long field study indicated that accelerated biofouling of seawater reverse osmosis (SWRO) membranes was associated with the elevated content of total organic carbon (TOC) in the intake seawater. Adding N and P to the intake seawater to balance the increase of TOC resulted in reduction of membrane biofouling. Microbial community analysis of the biofouling layer using 16S rRNA gene sequencing indicated biofouling communities varied with seasonal changes. Dosing of N and P did not induce dramatic changes in the fouling microbial community growing on the membrane surface. The outcome of this work implies that membrane biofouling associated with the elevated concentration of TOC in intake seawater is caused by imbalance of C:N:P in the source seawater which occurs often during algal blooms. Addition of N and P to rebalance the nutrients can prevent accelerated SWRO membrane biofouling. © 2019 Elsevier Ltd. All rights reserved.
Keywords: Membrane desalination Reverse osmosis Nutrient imbalance Biofouling
1. Introduction Reverse osmosis (RO) membrane fouling is described as the Achilles heel of the membrane processes for water purification and seawater membrane desalination (Flemming et al., 1997). Although inorganic-, organic- and bio-fouling all contribute to the retardation of the water production and elevation of energy consumption, biofouling and associated organic fouling are the most difficult to predict and control in seawater desalination plants (Ridgway, 1991; Ridgway and Flamming, 1996). Biofouling is caused by biofilm formation on the RO membrane surface by bacteria that have bypassed the pretreatment. The membrane biofilm formation is
* Corresponding author. 844 Engineering Tower, Irvine, CA, 92697, USA. E-mail address: [email protected] (S. Jiang). 1 Current affiliation, Department of Biology, University of Minnesota Duluth. https://doi.org/10.1016/j.watres.2019.114978 0043-1354/© 2019 Elsevier Ltd. All rights reserved.
initiated by organic conditioning of the membrane surface, followed by the attachment of individual bacteria and establishment of micro-colonies before developing into continuous layers of biofilm (Bold and District, 1998; Davies et al., 1993). The availability of organic nutrients, specifically organic carbon (C), seems to be the most important factor that influences the formation rate of biofilm on membranes (Ridgway and Flamming, 1996; Schneider et al., 2005; Sim et al., 2011; Tsai et al., 2004; Vrouwenvelder and Kooij, 2002). In our previous study (Huang et al., 2013), we investigated the influence of different environmental and water quality parameters on membrane biofilm formation using a flat-sheet membrane system fed with a side-stream of ultrafiltration pretreated seawater. The tests have shown that the increase of organic C concentration in the natural feed seawater (from ~0.5 mg/L to above 1.5 mg/L) has greatly affected the number of attached cells and the biofilm thickness on RO membranes over the one-year study period (Huang et al., 2013). Yet, the organic C
2
S. Huang et al. / Water Research 165 (2019) 114978
concentration that causes accelerated biofouling in seawater (~1e1.5 mg/L) is only ~10% of that commonly found in treated sewage effluent (~8e15 mg/L) used for reclamation by RO membrane process (OCSD, 2016). Biofouling of RO membranes for wastewater reclamation under such condition is not accelerated. These results suggest that in addition to the increase of organic C concentration, other factors also play a role in accelerated biofilm formation on RO membranes for desalinating seawater. Contrary to the nutrient enrichment observation in the membrane biofilm formation in the water treatment industries, the long-standing theory of biofilm formation in clinically important microorganisms is that biofilm formation is triggered by unfavorable conditions including nutrient starvation (O'Toole et al., 2000; Wai et al., 1998). The disparity between such seemingly opposite observations may be due to the nutrient composition required for bacterial growth and biofilm production. Specifically, the ratio of organic C to nitrogen (N) and to phosphorus (P) may be important in biofilm formation. Biofilm is an organic film composed of live and dead microorganisms embedded in a polymer matrix, known as extracellular polymeric substances (EPS) (Characklis and Marshall, 1990; Wingender et al., 1999). Although proteins and lipids are commonly observed in biofilm retrieved from wastewater treatment facilities and membrane bioreactors, polysaccharides are the main component of marine biofilms (Decho, 2000). The growth of marine bacteria depends on the availability of dissolved organic C as an energy and structural C source, N and P for replication of nucleic acid, synthesis proteins and cell wall for cell division (Thompson et al., 2006). The production and secretion of EPS, which plays an important role at the initial stage of biofilm formation, are at the expense of cellular energy. EPS production also increases the demand of bioavailable organic C in competition for essential cellular functions and organelle synthesis. Therefore, from a microbial ecology point of view, EPS production may be a mechanism to store excess organic C that cannot be utilized by the cell for replication due to the lack of N and P (limiting nutrients) to support cell proliferation. This postulation was supported by work by Stuart et al. who showed that cyanobacteria could reuse extracellular organic C in microbial mats and suggested EPS as a pool of extracellular nutrient resources (Stuart et al., 2016). This function may be more common in marine environments, where N is limited even during the post algal bloom periods (Romanı, 2009). Previous studies on bacterial biofilm growth have indirectly identified the C:N:P ratio as one of the key factors for EPS production and biofilm formation (Thompson et al., 2006; Allan et al., 2002; Fang et al., 2009). For example, Huang et al. have reported that polysaccharide secretion by biofilm forming E. coli DH5a increased significantly at high C:N ratio (10:1) (Huang et al., 1994). Thompson et al. have tested the effects of the C:N:P ratio on biofilm formation using Enterobacter and Citrobacter. They have found that higher C:N:P ratios (334:28:5.6) have resulted in significantly higher numbers of attached cells in both bacterial genera (Thompson et al., 2006). The adherence property of the plant pathogen Agrobacterium tumefaciens was enhanced under P limitation, despite the lower cell growth rate (Danhorn et al., 2004). The low P concentration was shown to trigger the two-component regulatory system PhoR-PhoB, which promoted the biofilm formation (Danhorn et al., 2004). These studies suggest that lower C:N:P ratios may reduce EPS production and the biofilm formation by bacteria. Past studies have attempted to reduce organic C loading to desalination RO feed seawater to achieve low C:N:P ratios in order to prevent membrane biofouling (Huang et al., 2013). However, the existing technologies for desalination pretreatment, including coagulation, flocculation, microfiltration, and ultrafiltration are only effective at removing particulates in water but are inefficient
to remove dissolved organic C. In fact, pressure-driven filtrations, such as ultrafiltration, have been shown to break algal cells during algal blooms to release additional dissolved organic C to support the excess biofilm production (Jiang and Voutchkov, 2014). Thus, the alternative approach to balance organic C:N:P ratio is to increase N and P concentration in the treated seawater to match the C content without removing it from the feed seawater. In this study, we have first examined bacterial isolates for biofilm production under different organic C:N:P composition. We then investigated bacterial physiological changes through upregulation of enzyme activities triggered by changes of the C:N:P composition in growth media. In order to develop a practical methodology for modifying organic C:N:P ratio for desalination membrane fouling prevention, we performed a year-long study at a desalination pilot facility, which included dosing of N and P to the pretreated intake seawater to match the content of organic C in the water and thereby rebalance the C:N:P ratio.
2. Materials and methods 2.1. Impact of the C:N:P ratio on bacterial growth and biofilm production Six bacteria (B1 to B6) isolated from biofouled RO membrane and cartridge filters from the Carlsbad Desalination Pilot Plant were used in this study. They were identified using 16S rRNA gene sequencing in a previous study (Zhang et al., 2012). B1 and B4 were identified as Shewanella sp.; B2 and B3 were nearly identical in 16S rRNA gene sequence and were matched to Alteromonas sp. B5 was identified as Cellulophaga sp. and B6 matched Vibrio sp. in the RDP database (Zhang et al., 2012). The bacteria were isolated on artificial seawater medium formulated by John Paul (ASWJP) (Paul, 1982) supplemented with 1 mg/L peptone and 0.5 mg/L yeast extract (PY). The bacteria were grown overnight at room temperature in ASWJP þ PY with shaking at 30 rpm, harvested by centrifugation at 10000 rpm for 1 min. The cell pellet was washed with ASWJP first to remove any remaining organic nutrient before it was used for biofilm production experiments. To characterize bacterial growth and biofilm formation under preset C:N:P conditions, sodium acetate (NaAc) was used as the single organic C source in the biofilm formation experiments. The inorganic N and P concentrations in the ASWJP were modified to match the typical concentrations found in the Pacific coast seawater (called base concentration in this study). The C:N:P concentrations were varied based on the desired ratio, as shown in Table 1, in two sets of experiments. The highest organic C concentration chosen (8 mg/L) in the experiments represents the high total organic carbon (TOC) loading condition in the Southern California Pacific coast seawater (e.g. post algal bloom or after a heavy storm that carries pollution from urban runoff). The first set of lab experiments focused on holding the N and P Table 1 Defined substrate nutrient concentration and nutrient ratio tested. C:N:P ratio
C (NaAc-C) mg/L
Experiment I 1:1:1 0.4 10:1:1 4 20:1:1 8 Experiment II 20:1:1 8 20:5:5 8 20:10:10 8 20:20:20 8
N (NH4NO3eN) mg/L
P (Na2HPO4eP) mg/L
0.2 0.2 0.2
0.05 0.05 0.05
0.2 1 2 4
0.05 0.25 0.5 1
S. Huang et al. / Water Research 165 (2019) 114978
concentrations at a constant value, while increasing the organic C concentration from 0.4 mg/L (1x) to 4 mg/L (10x) and to 8 mg/L (20x). Experiment II focused on holding the organic C concentration at 8 mg/L (biofouling condition) while increasing the concentration of N and P from the base level of 0.2 mg/L and 0.05 mg/L, respectively, to 5x, 10x and 20x the base concentration (Table 1). The cell growth and biofilm biomass assays were performed in clear 96-well microplates (round bottom, 330 ml/well, VWR International Corp. Cat # 734e1546). Briefly, washed cell pellets were resuspended and diluted in 1:10 ratio in each modified ASWJP medium as shown in Table 1. Each individual bacterial suspension was inoculated into 8 wells (8 replicates) of the microplate. The 8 wells in the last column of the plate were inoculated with medium only and were used as the negative control. Two identical plates were prepared: one was used for cell growth measurement by optical density using absorbance wavelength of 550 nm, and the other was used for biofilm quantification using the crystal violet assay method (Alhadidi et al., 2012). Crystal violet biofilm biomass assay was performed following the protocol of Stephanovic et al. (Stepanovic et al., 2000). Briefly, bacterial biofilm on the plate was fixed with methanol, stained with crystal violet (2% Hucker crystal violet used for Gram stain), and then released with the bound dye using 33% glacial acetic acid. The optical density of the released solution was measured at absorbance wavelength of 590 nm. Microplates were incubated at 24 C and were read using Spectramax Plus 384 (Molecular Devices, Sunnyvale, CA, USA). The average value of eight control wells was used to normalize the average value of each set of 8 wells for individual bacterial strain.
2.2. Impact of the C:N:P ratio on bacterial enzyme activities Bacterial enzyme activity assays were performed to examine the physiological responses of biofouling bacteria as a result of changing the C:N:P ratio. Enzyme assays were carried out using fluorometric assay protocol adapted from (Saiya-Corka et al., 2002) in black 96-well microplates (round bottom, 330ml/well, Fisher Scientific, Cat # 07-200-627) for b-glucosidase (BG), b-N-
3
acetylglucosaminidase (NAG), leucine aminopeptidase (LAP), and alkaline phosphatase (AP). Briefly, cells were grown in 96 well plates in ASWJP þ PY for 48 h under the same condition as for growth and biofilm assay. The growth medium together with suspended cells was then carefully removed from each well, leaving only the biofilm attached to the plate. The wells with biofilm were carefully washed twice using ASWJP to remove residual nutrient medium before they were filled with ASWJP without adding any organic C. The plates were incubated for two more days at room temperature. The enzyme activities were first measured in bacterial biofilm grown in rich complex organic nutrient medium immediately after the removal of unattached cells and nutrient medium (time zero), and then again at 24 h and 48 h when ASWJP was used as the only source of nutrients without any addition of organic C. The parallel clear 96-well microplates managed under the same condition were used to quantify cell density and biofilm thickness at each time point when enzyme assay were performed.
2.3. Pilot-scale testing of C:N:P ratio on membrane biofouling A year-long field study was set up at the Carlsbad Desalination Pilot facility. Two flat sheet bio-monitoring systems were operated in parallel with the spiral wound RO membrane train of the pilot plant using ultrafiltration-pretreated RO feed. Fig. 1 shows the schematic set up of the testing system. The system was designed to change the C:N:P ratio through dosing of N and P rather than reducing natural level of organic C in feed water. Unit 1 is a control membrane without dosing of N and P; while unit 2 is dosed with NH4NO3 and Na2HPO4 through a pump to final concentrations of 10 mg/L and 2 mg/L, respectively. No additional organic C was added to the feed seawater, thus the TOC concentration in this water reflected the natural variability of TOC in the source seawater. The TOC of the RO intake water was measured using the GE Sievers 5310 C On-Line TOC Analyzer (GE Instruments, Boulder, Colorado) as described by (Huang et al., 2013). The bio-monitor consists of a 5 20 cm flat sheet SWC5 membrane (Hydranautics) placed directly on top of a plastic
Fig. 1. Schematics of bio-monitor system at Carlsbad desalination pilot plant.
4
S. Huang et al. / Water Research 165 (2019) 114978
support; the feed channel was separated by a single piece of 0.87 mm thick nylon spacer. The cross section area of the feed channel is 44 mm2. The system was run at cross flow rate of 0.023 m3 h1 using 42 kPa feed pressure without permeation. The flat sheet bio-monitor was designed to simulate the cross flow velocity of the seven spiral wound RO membrane elements in the pressure vessels of the pilot plant. The membrane was removed from the monitor for analysis every two weeks and was replaced with a new membrane each time. After each removal, the system was cleaned and rinsed with clean tap water. The system was run at the Carlsbad desalination pilot facility from June 2010 to July 2011. 2.4. Determination of membrane biofilm density and thickness The retrieved flat sheet membranes from the bio-monitor were examined for biofilm thickness and total bacterial cell counts on the membrane surface by confocal laser scanning microscopy (CLSM, Zeiss LSM 510 META). Briefly, membranes were cut into 1 cm 1 cm squares and stained using SYTO 9 green-fluorescent nucleic acid stain for detecting live bacteria and propidium iodide red-fluorescent nucleic acid stain (FilmTracer™ LIVE/DEAD Biofilm Viability Kit, Invitrogen, Carlsbad, CA) for detecting dead bacteria. The stained membranes were then mounted onto a glass slide and observed under the CLSM. Excitation/emission wavelengths of 488nm/500 nm were used for SYTO 9 and 510nm/635 nm were used for propidium iodide, respectively. Images captured at each wavelength were composited into one final image. The Z sectioning method was used to determine the thickness of the biofilm. The thickness was calculated from 10 images captured cross the diagonal section of the membrane by locating the highest point above each pixel in the bottom layer containing biomass (Heydorn et al., 2000). Bacterial colonization was evaluated by visually counting the number of cells attached to the membranes surface, and was determined by the average counts from 10 images for each sample. 2.5. Analysis of bacterial community on the membranes To identify the bacteria colonizing the membrane surfaces, four sets membrane samples (4 controls, and 4 dosed with N and P) were collected each in autumn, winter, spring, and summer season and were used for bacterial community analysis. Approximately 2 2cm area of each membrane surface was swabbed for biofilm biomass. The genomic DNA was extracted and used for PCR of 16S rRNA gene using primers targeting the V7eV9 (939F-1492R) region. Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) (Roche, NJ, USA) was performed by the Research and Testing Laboratory (TX, USA). The sequences from pyrosequencing were analyzed using QIIME (Quantitative Insights Into Microbial Ecology) Pipeline (Caporaso et al., 2010). Low-quality sequences (<25) and short sequences (<200bp) were filtered and removed. Sequences with >97% similarity were clustered and defined as an Operational Taxonomic Unit (OTU). Representative sequences from each OTU were aligned and assigned to taxonomic identity using RDP classifier (Cole et al., 2005). The microbial communities were compared using the taxonomic and phylogenetic assignments. Beta diversity was plotted in Principal Coordinate Analysis (PCoA) to identify the community diversity between membrane samples. 3. Results 3.1. Impact of C:N:P ratio on bacterial growth and biofilm production The cell density and biofilm biomass in response to increased organic C in the growth medium are shown in Fig. 2. Changes in cell
and biofilm densities are expressed as the ratio of the measurements at a specific nutrient condition to those at the base nutrient concentration of C:N:P (1:1:1), which is set as 1. With the increase of C (by adding NaAc) concentration from 1x (0.4 mg/L) to 20x (8 mg/L) in growth medium, the cell density (OD550 ratio) had minimal changes (p > 0.01). This result showed although bacterial isolates could use NaAc as the single organic C source, the cell growth was limited by other nutrients. However, the biofilm biomass (OD590 ratio) increased significantly (p < 0.01) in bacteria B1 and B2 when organic C in the growth medium increased from 0.4 mg/L to 8 mg/L (20x) (Fig. 2). This result indicated these two strains actively converted organic C to biofilm content (EPS) in the presence of high concentration of organic C in the medium. Fig. 3 shows the results from experiment II, in which N and P concentrations increased from 1x (0.2 mg/L NH4NO3, 0.05 mg/L Na2HPO4, respectively) to 20x with the fixed NaAc concentration at 20x (8 mg/L). Increases in P and N concentrations resulted in increases in bacterial growth. The cell density was about 30% higher at 20x N, P concentration than those at base N, P concentration, suggesting N, P additions promoted the proliferation of free living cells. However, the biofilm biomass (OD590) decreased significantly (p < 0.01) for four of six bacteria tested (strain B1, B2, B4 and B5) at the higher N, P concentrations. There was a dramatic drop (more than 50%) in biofilm biomass once the N, P reached 5x of the base concentration for bacteria B1, B4 and B5. Bacteria strain B3 and B6 generated low level of biofilm at all N, P concentrations.
3.2. Enzymatic activities in the presence of low organic carbon Bacterial enzymatic activities were measured to determine bacteria's ability to use biofilm as nutrient when organic C was removed from the seawater (growth medium). The bacterial enzyme activities following the removal of organic C are presented in Fig. 4. Up-regulation of all four degradation enzymes was observed for bacterial B2. B3 also displayed enhanced activities for b-glucosidase (BG) and leucine aminopeptidase (LAP). Enzyme activities decreased in bacterial B4, and B5 only had elevation in leucine aminopeptidase (LAP) activity. These results indicated that some bacteria could utilize the nutrients stored in the biofilm for cell activities but not all bacteria could.
3.3. Field experiments with bio-monitoring system The year-long monitoring of water quality and biofouling potential at the Carlsbad desalination pilot plant showed that TOC of the feed seawater varied between 0.03 and 2 mg/L (30 ppb and 2000 ppb). Increases in TOC concentration in the intake raw seawater were observed several times over the year. These increases were mainly associated with land runoffs after rainfall events (Fig. 5). Higher biofilm thickness and cell counts were observed on the control (U1) membranes that were collected on the dates with elevated organic C in the feed water as indicated by the upward trend in fitted lines in Fig. 5. In comparison, membranes recovered from bio-monitor unit 2 (U2), where N and P were dosed into the feed water, did not experience dramatic increases in biofilm thickness or total cell counts as indicated by the relatively flat fitted line (Fig. 5). Both the biofilm thickness and cell counts on U2 membranes were fairly constant over the one-year study period. CLSM images (Fig. 6) also revealed that there was an observable difference between the flat sheet RO membrane retrieved from the control unit U1 (Fig. 6A) and the one from the N, P dosed unit U2 (Fig. 6B) during the period of high TOC measurements in feed water (Jan. 6, 2011).
S. Huang et al. / Water Research 165 (2019) 114978
5
Fig. 2. Changes in total bacterial cell density and biofilm biomass following the changes of organic carbon concentration in culture medium. The base concentrations of C:N:P (1:1:1) were set at 0.4 mg/L NaAc, 0.2 mg/L NH4NO3 and 0.05 mg/L Na2HPO4, respectively. The cell density and biofilm biomass at 10x (4 mg/L) and 20x (8 mg/L) C concentration were compared with the base level measurements (Cellnew/Cellbase and Biofilmnew/Biofilmbase). Error bars represent the standard deviation among eight experimental replicates.
3.4. Analysis of biofilm bacteria community on RO membranes by 16S rRNA gene The relationship between membrane biofouling, the bacterial community and the change of water quality (seasonality) was determined using 16S rRNA gene sequence analyses (Fig. 7). A total of 18728 qualified pyrosequencing reads of the V7eV9 regions of the 16S rRNA gene were obtained from the eight membrane samples collected between September 2010 and June 2011. Sequences were clustered into a total of 1778 OTUs by complete linkage clustering. Among them, >95% OTUs matched RDP database and were assigned the taxonomic identities (Table S1). The bacterial composition varied among samples retrieved from different seasons. Of the 1206 OTUs that were assigned taxonomy data, only 15 of them were observed on all membranes, and 396 of them were presented in a single set of samples. Taxonomy assignment for the top 10% most abundant classes of microorganism is shown in Fig. 7. Samples are labeled using membrane retrieval dates. For example, U1.9.16.10 is the biomass from membrane unit 1 (untreated control unit) retrieved on Sept. 16, 2010 and U2.9.16.10 is the biomass from membrane unit 2 (treated by N, P dosing) retrieved on the same date. Alpha-,
Gammaproteobacteria and Flavobacteria dominated membrane biofilm communities, with the combined abundance of 60e80% in all samples. The most severe membrane fouling was observed on U1.3.2.11 that was retrieved during the period of high TOC concentration in the feed water. The abundance of Gammaproteobacteria increased significantly on this membrane in comparison with other membranes. PCoA plot of beta diversity showed microbial biofilm composition was unique for each season (Fig. 8). Biofouling communities collected during autumn, winter, spring and summer seasons were spread across the four different coordinates of the PCoA plot. However, samples collected from control unit (U1) and treatment unit (U2) were clustered together, indicating nutrient dosing did not change the biofilm composition significantly but reduced the level of biofouling. 4. Discussion The results from bench-top experiments using variable C:N:P composition media (seawater) for cultivating marine biofouling bacteria support the hypothesis that excess organic C to N and P promotes the formation of microbial biofilm by marine bacteria. On
6
S. Huang et al. / Water Research 165 (2019) 114978
Fig. 3. Changes in total bacterial cell density and biofilm biomass following the changes of nitrogen and phosphorous concentration in culture medium. The cell density and biofilm biomass at base concentrations of C:N:P (1:1:1) were used as the comparison with measurements taken at the 20x base level of C (8 mg/L) and increasing concentration of N and P from the base to 20x the base concentration (Cellnew/Cellbase and Biofilmnew/Biofilmbase). Error bars represent the standard deviation among eight experimental replicates.
the other hand, increasing N, P concentration in the medium in the presence of bioavailable organic C promotes cell proliferation but reduces biofilm formation. We attribute the reduction of biofilm to the utilization of organic C in support of cell growth in the presence of N and P (for synthesis of proteins, lipids and the other cell components) based on our knowledge on nutrient compositions and bacterial growth (Touratier et al., 1999). In the absence of sufficient N and/or P (limiting nutrients) to support protein synthesis and cell proliferation, we observed bacterial excretion of the excess organic C as biofilm in both the bench-scale experiments and the pilot-scale membrane fouling studies. The bacterial enzyme activity assay results in part support the hypothesis that the biofilm formation may be a nutrient storage mechanism for future use in low-carbon conditions. This result agrees with a previous study demonstrating trace isotopically labeled EPS uptake into single cells of cyanobacteria (Stuart et al., 2016). Although we were not able to directly trace the uptake of C from biofilm EPS into cell biomass, the results of enzyme activity assays indicated that at least some of the bacteria could utilize nutrients in biofilm for active growth. Future research, such as isotopically labeled C tracer studies, is necessary to directly assess the role of biofilm as potential nutrient source for bacterial metabolic function under different environmental conditions.
The field experiments using two parallel tests with RO feed stream from a desalination pilot plant further confirmed the importance of the C:N:P ratio in formation of membrane biofilm. The year-long study confirmed that TOC increase in source water was the cause to trigger accelerated biofouling. This observation agrees with previous studies showing that elevation of organic C is the cause of biofilm production and a major culprit for membrane biofouling (Ridgway and Flamming, 1996; Schneider et al., 2005; Sim et al., 2011; Tsai et al., 2004; Vrouwenvelder and Kooij, 2002). The field experiments in the pilot plant demonstrated for the first time that N and P dosing in the feed stream could reduce biofouling on membrane surface. The biofilm formation on membrane surface under the shear of flow should be viewed as a dynamic process, in which biofilm Accumulation ¼ Production e Detachment þ Attachment (Hunt et al., 2004). The rate of direct bacterial attachment on the membrane was expected to be low because the bacterial concentration in the UF pretreated feed water was low (<103/L) (Huang et al., 2013). Thus the lower density of biofilm observed in the membrane unit with balanced C:N:P seawater feed is attributed to the slower rate of biofilm formation. Our observations showed that the addition of N and P to balance the C content in the source seawater reduced the available organic
S. Huang et al. / Water Research 165 (2019) 114978
7
Fig. 4. Enzyme activity of bacteria in biofilm 24 h (a) and 48 h (b) after removal of organic C from the culture medium. BG: b-glucosidase; NAG: b-N-acetylglucosaminidase; AP: alkaline phosphatase; LAP: leucine aminopeptidase. Enzyme activity is expressed as the ratio of measured activity at times to the initiate activity immediately after the removal of organic carbon (Et:E0) from the culture medium. Values greater than 1 indicate up-regulation of enzyme activity (values in the blue box).
Fig. 5. Plots of biofilm thickness (a) and total bacterial cell counts (b) on the surface of the RO membrane vs. TOC concentration in raw RO feed water. Membrane U1 (control unit) and U2 (membrane treated with N, P dosing) were collected from flat sheet biomonitor systems.
C for EPS production (as also demonstrated in the bench scale experiments). Accelerated biofilm detachment or sloughing under balanced C:N:P concentration could also play a part in the change of biofilm quantity, which was documented in a study of nutrientinduced P. aeruginosa biofilm dispersion (Sauer et al., 2004). This study showed that the increased concentration of inorganic N triggered up-regulation of flagella gene expression, which led to enhanced biofilm detachment under flow shear (Sauer et al., 2004). Overall, the reduced biofilm thickness and cell counts on N, P treated membrane unit in the field experiment coincided with the results observed in the laboratory tests of bacterial growth and biofilm production in defined C:N:P nutrient media. Both results suggest balancing the ratio of total organic C with the addition of N and P can be a strategy to reduce membrane biofouling during periods with excess organic loading. However, the practical application of such approach is seriously challenging because the excess N and P in the feed stream will end up in the brine reject and eventually be discharged back to the ocean. The excess nutrient loading could trigger new algal blooms and other negative environmental consequences. A nutrient trapping technology in brine reject, e.g. using a biofiltration system, will be needed before N and P dosing can be considered for membrane fouling mitigation. The RO membrane biofouling microbial community analyses
results agree with previous work indicating that Gammaproteobacteria and Alphaproteobacteria are the most dominant bacteria classes of membrane surface biofilm (Zhang et al., 2012). Seasonal shifts of biofouling microbial communities on membrane surface reflect the well-known seasonal succession of marine bacterial plankton (Bunse and Pinhassi, 2017). Complex physical, chemical, and biological factors influence marine microbial communities in time and space. Differential adaptation or sensitivity to variability in these factors is likely to determine the dynamics of the functional and taxonomic/phylogenetic diversity of bacterioplankton. Manual dosing of N and P in the feed water can also influence the microbial composition. For example, Ovreas et al. have shown that the addition of organic C, N and P in seawater lead to the increase in population of Gammaproteobacteria and a smaller increase in Alphaproteobacteria (Ovreas et al., 2003). The results of the current study indicate the changes of microbial composition due to dosing of N, P are less significant than the seasonal shift of natural microbial community in feed water. The diverse microbial composition recovered from the membrane surface also indicates that there is no single or a few bacterial species that are responsible for the biofilm formation. The reduction of membrane biofilm using N, P dosing implies that such management action has a broad effect in suppressing the biofilm
8
S. Huang et al. / Water Research 165 (2019) 114978
Fig. 6. Confocal laser scanning microscopy images of flat sheet membranes collected from bio-monitor. A: 2-D image of membrane in Unit 1 (control) collected on 1/6/2011; B: 2-D image of membrane in Unit 2 (dosed with N and P) collected on 1/6/2011.
formation by a diverse biofouling community. Further investigation, such as gene expression analysis, could shed new lights on the molecular mechanism of the N, P dosing on biofilm formation. 5. Conclusions
Fig. 7. The class level taxonomy assignments of biomass retrieved from control (U1) and treatment membrane (U2) in bio-monitors. Membranes are labeled based on the date of collection.
Results of laboratory studies showed that dosing N and P in the presence of high organic C promoted replication of marine bacteria but reduced biofilm formation. When organic C was removed from the culture medium (source seawater), some of the marine bacteria could utilize nutrients in the biofilm for growth (as indicated by elevated enzyme activities). A year-long field study demonstrated that dosing N and P to feed seawater could reduce the thickness of biofilm on the membrane surface during conditions of elevated TOC concentration in the feed water. Analysis of biofilm bacterial community revealed that the biofouling bacteria were influenced by the seasonal changes in water quality. N and P dosing did not change the fouling microbial composition significantly but suppressed biofilm production.
Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgement Financial support for this work was provided by WateReuse Foundation award WRF-08-19. Authors thank the following organizations and individuals for their support to this project: Daniel Marler and Steve LePage at Carlsbad Desalination Pilot facility; Steven Peck at Hydranautics. Technical assistance from Matthew Linder for bacterial enzyme activity assay is acknowledged. Appendix A. Supplementary data Fig. 8. 2D Principal Coordinate Analysis (PCoA) plot of microbial biofilm community on eight membrane samples.
Supplementary data to this article can be found online at https://doi.org/10.1016/j.watres.2019.114978.
S. Huang et al. / Water Research 165 (2019) 114978
References Alhadidi, A., Kemperman, A.J.B., Schurer, R., Schippers, J.C., Wessling, M., van der Meer, W.G.J., 2012. Using SDI, SDIþ and MFI to evaluate fouling in a UF/RO desalination pilot plant. Desalination 285, 153e162. Allan, V.J.M., Callow, M.E., Macaskie, L.E., Paterson-Beedle, M., 2002. Effect of nutrient limitation on biofilm formation and phosphatase activity of a Citrobacter sp. Microbiology 148 (1), 277e288. Bold, R.M., District, O.C.W., 1998. Influence of Molecular Conditioning Films on I Microbial Colonization of Synthetic Membranes-Determined by Internal Reflection Spectrometry. NWRI. Bunse, C., Pinhassi, J., 2017. Marine bacterioplankton seasonal succession dynamics. Review Special Series: Micro Commun. 25 (6), 494e505. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., Fierer, N., Pena, A.G., Goodrich, J.K., Gordon, J.I., 2010. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7 (5), 335e336. Characklis, W.G., Marshall, K.C., 1990. Biofilms. Wiley, New York. Cole, J.R., Chai, B., Farris, R.J., Wang, Q., Kulam, S., McGarrell, D., Garrity, G.M., Tiedje, J.M., 2005. The Ribosomal Database Project (RDP-II): sequences and tools for high-throughput rRNA analysis. Nucleic Acids Res. 33 (Suppl. 1), D294eD296. Danhorn, T., Hentzer, M., Givskov, M., Parsek, M.R., Fuqua, C., 2004. Phosphorus limitation enhances biofilm formation of the plant pathogen Agrobacterium tumefaciens through the PhoR-PhoB regulatory system. J. Bacteriol. 186 (14), 4492e4501. Davies, D.G., Chakrabarty, A.M., Geesey, G.G., 1993. Exopolysaccharide production in biofilms - substratum activation of alginate gene-expression by pseudomonasaeruginosa. Appl. Environ. Microbiol. 59 (4), 1181e1186. Decho, A.W., 2000. Microbial biofilms in intertidal systems: an overview. Cont. Shelf Res. 20 (10e11), 1257e1273. Fang, W., Hu, J., Ong, S., 2009. Influence of phosphorus on biofilm formation in model drinking water distribution systems. J. Appl. Microbiol. 106 (4), 1328e1335. Flemming, H., Schaule, G., Griebe, T., Schmitt, J., Tamachkiarowa, A., 1997. Biofouling the Achilles heel of membrane processes. Desalination 113, 215e225. Heydorn, A., Nielsen, A.T., Hentzer, M., Sternberg, C., Givskov, M., Ersboll, B.K., Molin, S., 2000. Quantification of biofilm structures by the novel computer program COMSTAT. Microbiology 146 (Pt 10), 2395e2407. Huang, C.T., Peretti, S.W., Bryers, J.D., 1994. Effects of medium carbon-to-nitrogen ratio on biofilm formation and plasmid stability. Biotechnol. Bioeng. 44 (3), 329e336. Huang, S., Voutchkov, N., Jiang, S.C., 2013. Investigation of environmental influences on membrane biofouling in a Southern California desalination pilot plant. Desalination 319, 1e9. Hunt, S.M., Werner, E.M., Huang, B., Hamilton, M.A., Stewart, P.S., 2004. Hypothesis for the role of nutrient starvation in biofilm detachment. Appl. Environ. Microbiol. 70 (12), 7418e7425. Jiang, S.C., Voutchkov, N., 2014. Investigation of Desalination Membrane Biofouling. Water Reuse Foundation. O'Toole, G., Kaplan, H.B., Kolter, R., 2000. Biofilm formation as microbial development. Annu. Rev. Microbiol. 54 (1), 49e79.
9
OCSD, 2016. GWRS Final Expansion - Final Implementation Plan. Project No. SP-173, Effluent Reuse Study. https://www.ocwd.com/media/5119/sp-173-vol1.pdf. Ovreas, L., Bourne, D., Sandaa, R.A., Casamayor, E.O., Benlloch, S., Goddard, V., Smerdon, G., Heldal, M., Thingstad, T.F., 2003. Response of bacterial and viral communities to nutrient manipulations in seawater mesocosms. Aquat. Microb. Ecol. 31 (2), 109e121. Paul, J.H., 1982. Use of Hoechst dyes 33258 and 33342 for enumeration of attached and planktonic bacteria. Appl. Environ. Microbiol. 43 (4), 939e944. Ridgway, H.F., 1991. Bacteria and membranes - ending a bad relationship. Desalination 83 (1e3), 53-53. Ridgway, H., Flamming, H.-C., 1996. Water Treatment Membrane Processes. McGraw-Hill, New York. Romanı, A.M., 2009. Freshwater biofilms. Biofouling 137. Saiya-Corka, K.R., Sinsabaugha, R.L., Zakb, D.R., 2002. The effects of long term nitrogen deposition on extracellular enzyme activity in an Acer saccharum forest soil. Soil Biol. Biochem. 34, 1309e1315. Sauer, K., Cullen, M., Rickard, A., Zeef, L., Davies, D., Gilbert, P., 2004. Characterization of nutrient-induced dispersion in Pseudomonas aeruginosa PAO1 biofilm. J. Bacteriol. 186 (21), 7312e7326. Schneider, R., Ferreira, L., Binder, P., Bejarano, E., Goes, K., Slongo, E., Machado, C., Rosa, G., 2005. Dynamics of organic carbon and of bacterial populations in a conventional pretreatment train of a reverse osmosis unit experiencing severe biofouling. J. Membr. Sci. 266 (1), 18e29. Sim, L.N., Ye, Y., Chen, V., Fane, A.G., 2011. Investigations of the coupled effect of cake-enhanced osmotic pressure and colloidal fouling in RO using crossflow sampler-modified fouling index ultrafiltration. Desalination 273 (1), 184e196. Stepanovic, S., Vukovic, D., Dakic, I., Savic, B., Svabic-Vlahovic, M., 2000. A modified microtiter-plate test for quantification of staphylococcal biofilm formation. J. Microbiol. Methods 40 (2), 175e179. Stuart, R.K., Mayali, X., Lee, J.Z., Everroad, R.C., Hwang, M., Bebout, B.M., Weber, P.K., Pett-Ridge, J., Thelen, M.P., 2016. Cyanobacterial reuse of extracellular organic carbon in microbial mats. ISME J. 10, 1240e1251. Thompson, L., Gray, V., Lindsay, D., Von Holy, A., 2006. Carbon: nitrogen: phosphorus ratios influence biofilm formation by Enterobacter cloacae and Citrobacter freundii. J. Appl. Microbiol. 101 (5), 1105e1113. Touratier, F., Ledendre, L., Vezina, A., 1999. Model of bacterial growth influenced by substrate C:N ratio and concentration. Aquat. Microb. Ecol. 19, 105e118. Tsai, Y., Pai, T., Qiu, J., 2004. The impacts of the AOC concentration on biofilm formation under higher shear force condition. J. Biotechnol. 111 (2), 155e167. Vrouwenvelder, J.S., Kooij, D.V.D., 2002. Diagnosis of fouling problems of NF and RO membrane installations by a quick scan. Desalination 153, 121e124. Wai, S.N., Mizunoe, Y., Takade, A., Kawabata, S.I., Yoshida, S.I., 1998. Vibrio cholerae O1 strain TSI-4 produces the exopolysaccharide materials that determine colony morphology, stress resistance, and biofilm formation. Appl. Environ. Microbiol. 64 (10), 3648e3655. Wingender, J., Neu, T.R., Flemming, H.C., 1999. Microbial Extracellular Polymeric Substances: Characterization, Structure, and Function. Springer Verlag. Zhang, M.Y., Zhang, K.S., De Gusseme, B., Verstraete, W., 2012. Biogenic silver nanoparticles (bio-Ag-0) decrease biofouling of bio-Ag-0/PES nanocomposite membranes. Water Res. 46 (7), 2077e2087.