The biomass growth and productivities are shown in Table 1. Due to availability of abundant organic carbon, the biomass productivity was 122.5 mg/l/d in the reactor, which is higher than an earlier report (Chinnasamy et al., 2010) of 57 mg/l/d (in raceway pond) and 70 mg/l/d (mixed algal species cultured in poly bags). Compared to these, Chlorella minutissima grown in an optimised media showed a biomass productivity of 1780 mg/l/d ( Li et al., 2011) in suspended algal systems, which seems quite high and rare.

TOC reduced from 219 ± 14 to 31 ± 12 mg/L in 12 days with a removal efficiency of ~86% due to an initial heterotrophic mechanism (with the abundant organic load). This was comparable to 86% carbon removal with Chlorella vulgaris (Feng et al., 2011), 98% carbon removal in mixed algal culture ( Prathima Devi et al., 2012) and ~36–57% carbon removal in diluted piggery wastewater (Wang and Lan, 2011). Incomplete removal of TOC could be attributed to growth inhibition due to either light limitation or presence of refractory carbon/algal polysaccharide secretions. Due to the filtration of cultured broth, visible suspended matter was found to be less (absorbance = 0.04) with ample light penetration during initial stages. Higher biomass densities (absorbance = 2.9) after the day 10 have contributed to a light restricted environment (with 5–7% light penetration) ceasing algal growth and consequent TOC removal. Specific growth rates differed from 0.38/d (initial 4 days) to 0.26/d (second phase corresponding to day 6 to day 12) that correlated with corresponding TOC removal of 219 to 111 mg/l (r = 0.9, p < 0.01) during the initial period and 77–33 mg/L (r = 0.99, p < 0.01) during the second phase of the exponential growth during algal culture.

Decline of TN from 64.6 ± 4.2 to 6.66 ± 2.1 mg/l (Table 2; Fig. 1b) during the growth period indicate ~90% of nitrogen assimilation (in the form of ammonium, ~80%) by growing algal cells in the presence of higher quantum of organic carbon. Nitrogen removal of 34.5% was observed during the first 2 days (heterotrophic phase) followed by 32.5% from day 6 to day 8 (mixotrophic phase) and overall removal of 82% was observed during all the 8 days (Fig. 1d). These were comparable to an earlier report (Cho et al., 2011) and they were also relatively higher compared to 78% TN removal of diluted primary wastewater and 54–75% of digested manure (Wang et al., 2010).

NH₄-N had dropped from 51.5 ± 5.5 to 5.76 ± 2 mg/l (89% removal) during 12 days of batch experiment (Table 2; Fig. 1b). This was lower compared to the values in the earlier studies carried out with Chlorella vulgaris in artificial wastewater (97% NH₄-N removal) conducted in heterotrophic/mixotrophic mode (Feng et al., 2011), and 96% removal in mixed algal culture grown in municipal wastewaters (Woertz et al., 2009). The incomplete removal of NH₄-N could have been due to carbon limitation or light interception as dense algal cultures were floating (of large and filamentous algae Spirulina and Phormidium, respectively) thereby interrupting light penetration. However, the current report of NH₄-N removal is better than the reported NH₄-N removal (69%) by Auxenochlorella protothecoides UMN280 grown in municipal wastewaters (Zhou et al., 2012b).

The average C/N and N/P ratio of the media were 3.77 ± 1.14 and 1.69 ± 0.67 respectively that favoured nitrogen removal. The decrease in nitrogen values correlated with the increased biomass (r = −0.9; p < 0.005) indicating nitrogen assimilation during algal growth. Highest NH₄-N removal was observed during the first 2 days (25.04%) and from day 6 to 8 (29.51%), which were similar to nitrogen removal ( Fig. 1d).

NH₄-N removal happens due to ammonia volatilisation under high pH and temperature. However, in the present study the major fraction of NH₄-N removal was due to assimilation by algae as pH and temperature were moderate (pH was <8.5 and temperature was 26 ± 4 °C) in the reactor. Fig. 1b shows a gradual increase of nitrate-N concentrations from1.07 ± 0.15 to 1.54 ± 0.28 mg/l due to nitrification. This was lower than 67% of nitrate-N removal in wastewater through mixed algal cultures isolated from wastewaters (Prathima Devi et al., 2012). On the other hand, Nitrite-N decreased slightly from 0.32 ± 0.05 to 0.21 ± 0.05 mg/l due to rapid assimilation by algae and slower nitrification process.

TP values listed in Table 2 showed a decrease from 28 ± 4.3 to 8.26 ± 2.4 mg/l (70% removal) compared to ~72% (Wang et al., 2010), 31–77% (Wang and Lan, 2011), ~84% removal (Cho et al., 2011); ~90% (Feng et al., 2011, Zhou et al., 2012a and Zhou et al., 2012b) and ~95% (Yujie et al., 2011). Phosphorus concentration in wastewater was relatively higher (Fig. 1c) than that of carbon and nitrogen (C:N:P = 7.8:2.3:1). Algae used only a small fraction of phosphorus that was available in the system, as a result of which ~30% of phosphorus remained residual in the culture. This fraction was unutilised as algal growth reached saturation due to carbon and nitrogen limitations as well as light limitations in the dense algal culture. Higher phosphorus removal on day 4 (30%) coincided with carbon assimilation that had resulted in a higher biomass growth (Fig. 1d). Excess phosphorus was also revealed by a lower N/P ratio (range 0.8–2.6) of 1.69 ± 0.67, which had not affected either biomass growth or productivity. Ortho phosphate removal (Fig. 1c) efficiency was 78% (22.2 ± 2.4 to 5.13 ± 1.67 mg/l) that was relatively higher than phosphates removal (40–65%) by mixed algal cultures grown in wastewater (Prathima Devi et al., 2012). Higher OP removal (33.15%) had occurred on day 4 (Fig. 1d). Organic phosphates varied from 7.6 to 1.78 mg/l, which constituted to about 16% of TP. Fig. 1c highlights this decline of TP that reached a minimum of ~2 mg/L (on day 8, after which, it stabilised to ~3 mg/l). The conversion of organic phosphorus into inorganic forms aided in the assimilation of phosphorous by algae.

The species composition and dynamics are elucidated in Fig. 1e. Euglenophyceae members such as Euglena sp., Phacus sp. etc. dominated (58%) during the early culture period up to the first 4 days that coincided with carbon uptake and higher biomass growth. In high organic environment, euglenoids grow in heterotrophic mode ( Mahapatra et al., 2013a, Mahapatra et al., 2013b and Ramachandra et al., 2013). Chlorophyceae (especially Chlorella sp. and Chlorococum sp.) increased from 22% to 56% (day 4), and then, showed a declining trend. Mixotrophic chlorophycean members (Chlorococcum sp.) are known to grow faster in nutrient rich conditions ( Mahapatra and Ramachandra, 2013).

Bacillariophyceae (Cyclotella sp. and Nitzschia sp.) considered as facultative heterotrophs ( Mahapatra et al., 2013b) increased to 34% (day 6) and showed a gradual decline to 9% (day 12). The species turnover indicated changes in nutrient availability and reactor conditions that was suitable for assimilation and growth of chlorophyceae and bacillariophyceae members resulting in switch over from heterotrophy to mixotrophy. Cyanophyceae (blue green algae as Phormidium sp. (filamentous) and Spirulina sp. (spiral shaped)) showed an increasing trend ( Fig. 1e) from 2% (day 4) to 44% (day 12). These algae are often found as naturally floating (due to dense gas vacuoles) masses in urban lakes and wastewater treatment lagoons ( Mahapatra et al., 2013b). The dense gas vacuoles in filamentous algae help in floatation. This highlighted that algal consortia of varied proportion depending on the nutrient availability enables the treatment of wastewater. Diverse species with differential interactions/competition also contributes to the system stability with enhanced biomass growth and efficient removal of nutrients. This required further investigation, which is being undertaken.

3.2. Mechanisms for optimal harvesting of algae – algal floc formation, P accumulation and elemental composition

SEM-EDXA analysis of suspended algal biomass revealed higher organic content comprising ~55% C; 7% N with a higher P accumulation of ~4% (Fig. S1 provided in the supplementary material). Higher phosphorus accumulation (in the form of Poly-P; Chen et al., 2011) in this biomass was due to algal habitat that stored enormous phosphorus which was coupled with a higher metabolic phosphorus demand to fulfill the cellular phosphorus requirement. Further, phosphorus fractionation was required to understand the metabolic phosphorus demand and phosphorus storage, which would help in estimating phosphorus holding capacity and its relevance to phosphorus recovery and recycling in the environment. In the present case, phosphorus accumulation in the biomass was attributed to higher ortho phosphate concentration remaining in the media (~22%) that had not precipitated out because of low alkaline conditions. In the present study, the suspended matter were mostly in the form of granules (15–45 μm) that comprised of ~50% C and ~33% oxygen with higher levels of Ca (~5%) indicating the formation of CaCO3 salts that get adsorbed over other organic granules or over biogenic Si (from diatoms) in the reactor. The reactor sludge in the final stages showed abundant organic matter with inorganic elements like Si, Al, Mg etc.

The analysis of elemental composition of the algal flocs/scum (helpful in algal harvest), floating on the reactor surface and the surface morphology (Fig. S1 provided in the supplementary material), was done through SEM-EDXA. Smaller flocs (<0.5 cm) mainly composed of green algal cells were flocculated by probable algal derived flocculant (unpublished data) in combination with metal ion. Algal flocculation was due to the negative charges on the cell surfaces during the peak period of algal growth (Chen et al., 2011). Floc formations or auto-flocculation (Salim et al., 2011) have been attributed to pH transitions (Wu et al., 2012) and availability of higher cations (divalent) which lead to mineralisation at the mature stages of algal growth. Filamentous cyanophycean masses as Phormidium sp. (1–4 cm diameter) with gas bubbles float in large masses on the surface. This natural phenomenon of aggregation and floatation was useful for efficient algal harvesting. Harvesting of algae has been a major constraint for its commercial use as existing mechano-chemical processes are energy and also time consuming when compared to bio flocculation (Park et al., 2012). Therefore, there were attempts to harvest algal biomass through bio-flocculation involving algal-bacterial interaction ( Su et al., 2011) and fungal palletisation-assisted bio-flocculation of algal cells ( Zhou et al., 2012a). In such cases, the culture environment plays a major role in phase separations in flocculation and sedimentation.

Table 3 lists the elemental composition of Algal scum (flocs) and other substrates (reactor sludge, naive algal biomass and suspended matter). The elemental analysis showed high concentration of Ca (~26%), Fe (~5%), K (~8%) and Cl (~8%) in the algal surface flocs. However, these elements were low in other substrates as they were primarily organic. The elemental peaks indicated the presence of Ca and P (Fig. S1 provided in supplementary material). Higher froth or algal floc formations had been attributed to Ca, Al, Cl (Chen et al., 2011, Park et al., 2012 and Wu et al., 2012). Phosphorus helps in coagulation while CaCO3 (calcium, carbon and oxygen) and FeCl3 (due to higher chlorine and iron) aid in the formation of flocks in algal scum.