Extracellular polymeric substance production in high rate algal oxidation ponds

Integrated algal pond systems (IAPSs) are an established wastewater treatment technology and exploit the mutualistic relationship between bacteria and algae (Oswald 1995; Green et al. 1995; Munoz & Guieysse 2006; Park et al. 2011; Mambo et al. 2014; Craggs et al. 2015). During wastewater treatment by microalgal-bacterial (MaB) consortia, biomass accumulates due to CO₂ assimilation concomitant with in situ O₂ production. Heterotrophic bacteria consume O₂ for the aerobic biodegradation of organic compounds to CO₂, which is then assimilated by the microalgae. The continuous exchange of O₂ and CO₂ between microalgae and bacteria facilitates formation of microalgal-bacterial flocs (MaB-flocs) (Gutzeit et al. 2005; Medina & Neis 2007; Su et al. 2011). MaB-flocs typically consist of a consortium of microalgae, cyanobacteria and bacteria and may include a number of rotifers, ciliates and precipitates (Van Den Hende et al. 2011). Paddlewheel driven high rate oxidation ponds maintain these MaB-flocs in suspension as biological aggregates, presumably as components of a biofilm. Even so, IAPS treated water is characterized by elevated COD/TSS which led to recommendations that additional treatment of the outflow from algal settler ponds (ASPs) is required to meet specific discharge standards (Craggs et al. 2012; Mambo et al. 2014).

Extracellular polymeric substances (EPSs) are high-molecular weight compounds secreted into the environment and are responsible for the functional and structural integrity of biofilms (Staudt et al. 2004). Furthermore, EPS are considered the fundamental component that determines the physicochemical properties of a biofilm. In aquatic systems, EPS accumulate in the water when cells lyse or following secretion of high molecular weight compounds by microorganisms and, with the addition of organic matter, form matrices that serve as flocculants for biofilm formation (Sheng et al. 2010; More et al. 2014). Recent studies have elaborated on the downstream uses of these MaB-flocs, which emphasizes the value of this resource (Czaczyk & Myszka 2007; Natrah et al. 2013; Essam et al. 2013; Wieczorek et al. 2015; Arcila & Buitrón 2016; Coppens et al. 2016; Van Den Hende et al. 2016). The influence of EPS in wastewater treatment includes biofilm formation, adherence of cells to surfaces, enhanced settleability, and protection against harm from toxic substances and dewatering (Sutherland 2001; Sheng et al. 2010; More et al. 2014). Both soluble and bound EPS occur in MaB-flocs-containing wastewater treatment systems. Soluble EPS are loosely attached to cells or dissolved while bound EPS are closely associated with cells. These two forms of EPS can apparently be separated by centrifugation with the bound EPS forming a pellet and the soluble EPS remaining in the supernatant (Sheng et al. 2010). Although the composition of EPS varies depending on culture medium, population dynamic, growth phase of the organisms, and the extraction method adopted, the matrix of compounds in wastewater typically includes carbohydrates (in particular polysaccharides), proteins, humic substances, lipids, and uronic and nucleic acids (Wang et al. 2014). Consequently, these materials are of biotechnological importance and potential application in bioremediation, and the food and pharmaceutical industries, has been documented (Mishra & Jha 2009).

Previously, it was shown that effluent from the Belmont Valley IAPS had elevated levels of total suspended solids (TSS) and chemical oxygen demand (COD), which was considered to be the result of microalgal programmed cell death (Mambo et al. 2014). The present study describes the MaB-flocs formed in the high rate algal oxidation ponds (HRAOPs) of this IAPS and shows that loss of biomass is due to passive settling. Settleability and high effluent COD are attributed to formation of a soluble EPS associated with MaB-flocs generated in HRAOP of this IAPS.

The IAPS used in this study is located at the Belmont Valley Municipal WWTW, Grahamstown, South Africa (33° 19′ 07″ South, 26° 33′ 25″ East) and has a 75 m³ d⁻¹ design capacity equivalent to 500 persons. The system is in continuous operation and comprises of an in-pond digester (IPD), advanced facultative pond (AFP), two HRAOPs, and two ASPs configured in series. The typical process flow is: 3 d in the IPD → 20 d AFP → 2 d HRAOP A → 0.5 d ASP A → 4 d HRAOP B → 0.5 d ASP A → return to Belmont Valley WWTW inlet works. Although details on system configuration, process flow, and water treatment efficiency have been previously reported (Mambo et al. 2014) a flow diagram of the functioning of this IAPS is presented in Figure 1.

Yield of MaB-flocs in HRAOP B was estimated by measuring mixed liquor suspended solids (MLSS) and by determining overall productivity as described by Banat et al. (1990). To achieve this, samples of mixed liquor were sourced directly from the front of the paddlewheel and transferred to the laboratory immediately. Both morning (09:00) and evening (16:30) samples were collected from August to November 2015 for seven consecutive 4 or 5 d periods. Thoroughly stirred samples were filtered under vacuum using pre-dried (105 ^oC × 1 h) glass fibre discs (Whatman; 47 mm, 0.45 μm) that had been placed in a desiccator for at least 30 min prior to use. After filtration of the final rinse (∼3 min), discs were removed and dried to a constant weight at 105 °C (1 h), cooled in a desiccator for 30 min, and the weight recorded. MLSS was calculated using the formula: MLSS (mg L^−l) = [(A – B) ×1,000] ÷ [Volume of sample in mL]. Where: A is the sample + filter weight; and B the weight of the filter only.

Pond productivity (g m⁻² d⁻¹) was calculated by including the function C_s, which is the algal biomass in mg L⁻¹, determined by multiplying the MLSS (or TSS; APHA 1998) by a factor, n (i.e. the algal ratio in the TSS). Thus, HRAOP productivity was quantified per unit surface using the expression: P = 10^d/_t·n·MLSS. Where: P = MaB-flocs productivity (g m⁻² d⁻¹); d = pond depth (m); t = detention time (d); SS= TSS (mg); and n= factor expressing the algal ratio in the suspended solids which, for near pure cultures is 0.9–1.0 (Al-Shayji et al. 1994).

Aliquots of wastewater (500 mL) were taken from HRAOP B and contained a mixed liquor comprising algae, diatoms, bacteria and other unidentified microorganisms. Flasks were placed in a controlled environment on a rotary shaker either under cool white fluorescent light or in total darkness at 25 °C. Aliquots were removed immediately (t = 0) and at 2, 3, 4, 6, 8, and 9 d intervals for quantification of EPS and biochemical analyses.

Only EPS released into the medium (i.e. soluble EPS) was extracted and, quantified by gravimetric analysis. Extraction was according to the method described by Ahmed et al. (2014). In brief, samples (100 mL) were centrifuged at 5,000 × g for 20 min, the supernatant was collected, MLSS removed by filtration (0.45 μm), and the aqueous filtrate filtered (0.22 μm), flash frozen using liquid nitrogen, freeze dried at −80 °C, weighed, and stored in a desiccator for further analysis.

FT-IR analysis of sub-samples of the extracted EPS (∼1 mg) was carried out using a PerkinElmer Spectrum 100 instrument (PerkinElmer, Waltham, MA, USA) with attenuated total reflectance (ATR) accessory eliminating the need for mixing of samples with potassium bromide (KBr). The ATR accessory, fitted with a diamond top-plate, has spectral range of 25,000–100 cm⁻¹, refractive index of 2.4, and 2.01 μ depth of penetration. FT-IR spectra were recorded in the range of 4,000–650 cm⁻¹.

Total carbohydrate, total protein, and total α-amino nitrogen of the extracted EPS were determined spectrophotometrically using the phenol sulphuric acid (Dubios et al. 1956), protein dye-binding (Bradford 1976), and Ninhydrin (Lie 1973) methods, respectively. For total carbohydrate, 2 mg aliquots of extracted EPS were resuspended in 0.5 mL distilled water to which was added, 0.5 mL of phenol solution (4% phenol in distilled water) followed immediately by 2.5 mL concentrated sulphuric acid (reagent grade). The mixture was vortexed and cooled to room temperature. Absorbance at 490 nm was determined using a UV-Vis spectrophotometer with distilled water and reagents used as background. Carbohydrate concentration was determined by interpolation from a standard curve for D-glucose (Dubios et al. 1956).

For total protein, 3 mg aliquots of extracted EPS were reconstituted in 0.5 mL of distilled water and 5 mL Bradford reagent (prepared by dissolving 100 mg of Coomassie Brilliant blue in 50 mL 95% ethanol and after addition of 100 mL 85% phosphoric acid, the solution was diluted to 1 L and filtered) added and, mixed thoroughly (Bradford 1976). Absorbance was measured at 595 nm after incubation for 10 min at room temperature and protein concentration determined by interpolation from a standard curve for bovine serum albumin.

For estimation of free α-amino nitrogen, 2 mg aliquots of EPS were dissolved in 2 mL distilled water and 1 mL of reagent (prepared by dissolving 100 g Na₂HPO₄.12H₂O, 60 g anhydrous KH₂PO₄, 5 g Ninhydrin, and 3 g fructose in 1 L distilled water, pH 6.7) added and placed in boiling water for 16 min and thereafter cooled to 20 °C (20 min). Dilution reagent (5 mL; prepared by dissolving 2 g KIO₃ in 600 mL distilled water which was then made to 1 L with 96% ethanol) was then added, samples thoroughly mixed, and absorbance at 570 nm determined spectrophotometrically within 30 min (Lie 1973). The concentration of α-amino nitrogen was determined by interpolation from a standard curve for glycine.

An example of the settleability and composition of the MaB-flocs produced in the HRAOPs of the pilot-scale IAPS treating municipal sewage during the course of this study are shown in Figure 2.

MaB-flocs appear as discrete entities and the bulk of these flocs settle readily within 2 h (Figure 2(a) and 2(b)). Light microscope analysis of these MaB-flocs revealed recruitment of microalgae, diatoms, cyanobacteria, and bacteria presumably facilitated by production of EPSs (Figure 2(c)). Among the more prominent species were the chlorophytes, Pediastrum sp., Chlorella sp., Closterium sp. and Scenedesmus sp. and the diatoms Cyclotella sp., Nitzschia sp. and Navicula sp.

In an effort to derive information about the formation of MaB-flocs two approaches were adopted. First, HRAOP productivity and the change in MaB-flocs concentration between morning (MLSS_AM) and evening (MLSS_PM) were monitored. Second, the substructure of the MaB-floc matrix was examined by a combination of biochemical analysis and FT-IR spectroscopy. As expected, production of MaB-flocs in HRAOPs, i.e. productivity measured as MLSS, gradually increased coincident with a rise in water temperature from ∼10 g m⁻² d⁻¹ in early August (winter) to >20 g m⁻² d⁻¹ in November (Figure 3). This gradual increase in productivity with increasing water temperature was associated with a diurnal fluctuation in the accumulation of MLSS (Figure 3). As shown in Table 1, the loss of biomass between early evening, i.e. MLSS_PM and the following morning (MLSS_AM) was due to continuous operation of the system and passive settling of the MaB-flocs in the ASP. Indeed, based on influent flow rate and hydraulic retention in the HRAOPs, the estimated loss of MLSS to the ASP between early evening and the following morning was 29.57 ± 1.60 mg L⁻¹ which, using a t test (SigmaPlot Ver. 11; Systat Software, Inc., San Jose, CA, USA), was not significantly different (P = 0.14) from the measured loss of 36.70 ± 4.95 mg L⁻¹.

To gain further insight into the possible role for EPS in formation of MaB-flocs, the carbohydrate, protein and α-amino nitrogen contents were determined. In addition, the EPS from HRAOPs, and from MaB-flocs incubated in continuous light and total darkness were analysed by FT-IR spectroscopy. Results from two independent experiments revealed that the soluble EPS obtained after freeze-drying appeared as a white wispy-like material and that its production by MaB-flocs from HRAOPs was indeed enhanced by exposure to continuous illumination (Figure 4). Thus, soluble EPS increased from ∼100 to >450 mg L⁻¹ and from ∼100 to ≥200 mg L⁻¹ in light- and dark-incubated MaB-flocs, respectively.

The increase in recovery of soluble EPS over time was taken to indicate that production of EPS occurred concomitant with an increase in MLSS (Figure S1, Supplementary data, available with the online version of this paper). Even so, accumulation of this ‘new’ EPS by both light- and dark-incubated MaB-flocs was associated with a decline in carbohydrate, protein, and α-amino nitrogen content of the original material and, this decline was more pronounced for dark-incubated MaB-flocs (Figure 5(a) and 5(b)). Taken together, these results suggested that dark-incubated MaB-flocs transitioned from phototrophic to heterotrophic growth and that in the absence of light, both carbon and nitrogen were sourced mainly from already synthesized EPS. To elucidate this aspect further, the soluble EPS from HRAOPs and from MaB-flocs incubated in either continuous light or total darkness were extracted and analysed by FT-IR spectroscopy. Since the frequency at which a given vibration occurs is determined by the strength of the bonds involved and the mass of the component atoms, it was rationalised that EPS from dark-incubated MaB-flocs growing heterotrophically will differ in intensity of the characteristic functional groups from EPS formed in the light.

Comparative FT-IR spectra of the EPSs obtained after removal of MLSS from the HRAOP and from MaB-flocs incubated in either continuous light or total darkness are shown in Figure 6. A representative spectrum of EPS from HRAOP reveals characteristic functional groups (Figure 6(a)). The medium stretches and bend of frequency in the spectral ranges 3,500–3,300 and 1,640–1,560 cm⁻¹ were assigned to O-H (H-bonded) and N-H stretching. Weak absorbance at 2,940–2,920 cm⁻¹ is attributed to assymetric C-H stretching of aliphatic methyl groups. The region 2,250–2,100 cm⁻¹ is assigned to C≡C stretching of alkynes while the stretch, albeit weak, in the region of 1,660–1,600 cm⁻¹ is assigned to C = C in alkenes but may be due either to CO₂ adsorption (Nabiev et al. 1976) or asymmetric stretching of –N = C = O– (Panda & Sadafule 1996). The sharp bend between 1,460 and 1,380 cm⁻¹ is assigned to C-C of aromatics and, the region 1,300–1,000 cm⁻¹ is assigned to stretching of C-O-C, C-O and corresponds to the presence of carbohydrates (Bremer & Geesey 1991; Bramhachari & Dubey 2006; Mishra & Jha 2009). The bending pattern in the region 950–650 cm⁻¹ is assigned to sp² C-H of alkenes and aromatics. Taken together, the results of FT-IR analysis of EPS from HRAOPs confirmed the presence of primary amines, aromatic compounds, aliphatic alkyl groups, and carbohydrates. Similar spectra were derived from FT-IR analysis of EPSs from light-incubated MaB-flocs (Figure 6(b)). However, the IR spectrum of EPS from dark-incubated MaB-flocs revealed a substantial increase in intensity of absorbance in the vibrational ranges 3,500–2,800, 1,460–1,380, 1,300–1,000, and 950–650 cm⁻¹ corresponding in assignment to O-H (H-bonded), N-H and sp C-H stretching, C-C bending of aliphatics and aromatics, O-C stretching and C-O-O bending of carbohydrates, and C-H bending and ring puckering of aromatics, respectively (Figure 6(c)). Thus, EPSs from dark-incubated MaB-flocs show increased frequency of vibration in aliphatic and aromatic functionalities relative to carbohydrates, presumably as a consequence of the transition to heterotrophic metabolism.

This study set out to determine the source of elevated TSS and COD in water from an IAPS treating domestic sewage. It was previously suggested that microalgal programmed cell death in both HRAOPs and ASPs might be responsible (Mambo et al. 2014). As shown in the present study, MaB-flocs are assemblages of microorganisms produced in HRAOPs as discreet aggregates and form as a result of microbial EPS production (Sheng et al. 2010; More et al. 2014). Formation of MaB-flocs in the HRAOPs increased with increasing water temperature and the transition from winter to summer. Inherent diurnal fluctuation in the concentration of MaB-flocs (measured as MLSS) and to a lesser extent EPS, in HRAOPs was due largely to passive removal of the flocs by the ASPs. Furthermore, formation and accumulation of loosely bound or soluble EPS was stimulated by light suggesting a link between photosynthesis and EPS production. Indeed, while EPS accumulation was observed in dark-incubated MaB-flocs the carbohydrate and protein content of these was markedly reduced relative to EPSs produced in continuous light. It was rationalised that dark incubation of MaB-flocs resulted in a transition from phototrophic to heterotrophic metabolism and that in the absence of photosynthesis, carbon and nitrogen from existing EPS was being recycled to support growth. Taken together, these results suggest that MaB-floc-derived EPS and not microalgal programmed cell death, appears to be the major contributor to elevated TSS and COD in IAPS treated water.

Analysis by FT-IR of EPSs isolated after removal of MLSS from the HRAOP and MaB-flocs incubated in continuous light revealed characteristic carbohydrate enrichment of these polymeric substances (Mishra & Jha 2009). In contrast, FT-IR spectra of EPSs from dark-incubated MaB-flocs appeared to confirm that these polymers contained increased aliphatic and aromatic functionalities relative to carbohydrates. One possibility for these differences is recruitment of humic and/or fulvic substances. As pointed out by More et al. (2014), humic substances are an integral component of EPSs and are adsorbed by the EPS matrix from the natural environment rather than synthesized and secreted by microorganisms. Humics occur naturally as complex ligands and are widely distributed in aquatic systems and, in particular, sewage water and sediments. Since EPSs can adsorb humic acids and do so by a combination of hydrophobic and cationic bridging (Esparza-Soto & Westerhoff 2003), it is proposed that in darkness (or in the absence of photosynthesis as might be expected in ASPs) EPS-adsorbed humics and/or fulvics facilitate access to polymeric substance carbohydrates by MaB-flocs to sustain growth and metabolism without affecting the floc matrix.

Microbial (i.e. MaB-floc) EPS production rather than microalgal programmed cell death appears to be the major contributing factor towards elevated levels of TSS/COD of water from IAPS treating domestic sewage. The EPS from HRAOPs have been isolated and partially characterized. The FT-IR-spectra confirmed the presence of primary amines, aromatic-compounds, aliphatic alkyl groups and polysaccharides, whereas EPSs from dark-incubated MaB-flocs showed increased vibration in aliphatic and aromatic functionalities relative to carbohydrates. These differences, it was interpreted, were due to dark-induced transition from phototrophic to heterotrophic metabolism. Studies are currently underway to elucidate fully the structure of HRAOP EPSs and to evaluate any potential commercial application of these polymers for use in agriculture and industry and, as natural microbial flocculants. Precisely how MaB-floc EPS formation and accumulation impact design and operation of IAPS systems is also under consideration.

We gratefully acknowledge financial support from the Water Research Commission (WRC) of South Africa through WRC Projects (No. 7055 and 7164) awarded to Prof A. Keith Cowan of Rhodes University. Ms Taobat Jimoh acknowledges receipt of a graduate bursary from EBRU. This work was partially supported by a grant from the National Research Foundation, South Africa (IFR1202220169) awarded to Professor A. K. Cowan.