Flocculants or flocculating agents are chemicals that promote flocculation by aggregation of colloids and other suspended particles, forming a floc [1]. Flocculants are widely applied in various industrial processes, including wastewater treatment, downstream processing, food and fermentation processes [2,3,4]. Although chemical flocculants have been widely used due to their effective flocculating activity and low cost, some synthetic flocculants are known to be hazardous to the environment [5,6,7]. In recent years, utilization of microbial flocculants has been promoted due to their biodegradability and their environmentally inert nature [8,9]. Screening new microorganisms for bioflocculant activity, therefore, has become a subject of intensive investigations globally.

Marine bacteria are among the most economically and biotechnologically valuable prokaryotes, responsible for the production of about 50% of all discovered bioactive secondary metabolites [10]. The rate of discovery of new compounds from terrestrial bacteria has decreased, whereas the rate of re-isolation of known compounds has increased [11], suggesting the need for exploring underexploited habitats such as the marine environment as sources of novel bioactive compounds including bioflocculants. In this paper, we report a novel bioflocculant produced by a marine bacterium belonging to the genus Cobetia, isolated from sediment samples of Algoa Bay in the Eastern Cape Province of South Africa. To the best of our knowledge, this is the first time that the genus has been implicated in bioflocculant production.

Amplification of the 16S rRNA gene of the bacteria resulted in an amplicon of expected size (approximately 1.5 kb). Basic Local Alignment Search Tool (BLAST) analysis of the nucleotide sequence of the 16S rDNA revealed 99% similarity to that of Cobetia sp. L222 and the nucleotide sequence was deposited in GenBank as Cobetia sp. OAUIFE with accession number JF799092.

The result of the effect of inoculum size on bioflocculant production is as shown in Figure 1. Inoculum sizes of 0.5 mL, 1.0 mL, 1.5 mL and 2.0 mL in 50 mL production medium were investigated, representing 1%, 2%, 3% and 4% (v/v) respectively. Inoculum size of 2% (v/v) resulted in the highest flocculating activity (86.4%). Further increases in inoculum size resulted in decreases in flocculating activities. As a result, an inoculum size of 2% was used for all subsequent experiments.

Figure 2 shows the effects of various carbon (20 g/L) sources on bioflocculant production. The bacteria effectively utilized all organic carbon sources for bioflocculant production with flocculating activities of more than 70% except for xylose. The bacteria poorly utilized all the inorganic carbon sources except for sodium carbonate, which resulted in flocculating activity of 62.5%. Glucose and maltose were favourable carbon sources for the test bacteria, with flocculating activities of 92.2% and 90.5% respectively. Glucose was used for all subsequent cultures.

The bacteria utilized various nitrogen sources (1.2 g/L) resulting in production of bioflocculant with flocculating activities ranging from 14.8% to 51.9% (Figure 3). Organic nitrogen sources were poorly utilized with virtually all resulting in flocculating activities of less than 50% except for casein which along with the inorganic nitrogen sources ammonium chloride and sodium nitrate resulted in flocculating activity of about 50%.

The effect of initial medium pH on bioflocculant production was investigated and reported as shown in Figure 4. Bioflocculant was produced at pH ranges of 3–7, with the highest flocculating activity (90.5%) observed at a weakly acidic pH of 6. Alkaline pH was unfavourable for the production of bioflocculant by the bacteria.

All the metal ions significantly stimulated bioflocculant production resulting in very high flocculating activities, except for Fe³⁺ which completely inhibited bioflocculant production as evident from the zero flocculating activity. The extent of preference of the metal ion for bioflocculant production followed the order Mn²⁺>K⁺>Na⁺>Ca²⁺>Li⁺>Mg²⁺>Al³⁺ (Table 1).

Flocculating activities with different letters are significantly different (p < 0.05) from each other.

With regards to the effect of heat of the crude bioflocculant, heating of the bioflocculant culture broth at 100 °C for 25 min revealed a residual flocculating activity of 78%, while heating for the same period at 50 °C and 80 °C resulted in residual flocculating activities of 87% and 82%, respectively (Figure 5).

Figure 6 reveals the profile of the time course experiment. During the exponential growth phase, flocculating activity increased steadily, until it peaked after 72 hour (with a flocculating activity of 92.2%) beyond which flocculating activity started to decline. It was also observed that the optical density of the culture with time ran almost parallel with the flocculating activity curve.

The biochemical analysis of the purified bioflocculant (Table 2) showed that it was mainly composed of uronic acid (93%), protein (5%) and neutral sugar (1.8%).

This study revealed that an inoculum size of 2% (v/v) of the pre-culture resulted in the highest flocculating activity (86.4%) and further increases in inoculum size resulted in decreases in flocculating activities. A similar observation was made by Salehizadel and Shojaosadati [4]. As a result, an inoculum size of 2% was used for all subsequent experiments. Inoculum size affects not only the cell growth of microorganisms but also their product formation [25,26]. Some studies have reported correlations between inoculum size and bioflocculant production by some organisms. Zhang et al. [13] obtained maximum flocculating activity with an inoculum size of 1% (v/v), while Jang et al. [27] and Wang et al. [17] observed best flocculation with an inoculum size of 5% (v/v).

With respect to the effect of carbon sources on bioflocculant production, the bacteria effectively utilized all organic carbon sources for bioflocculant production with flocculating activities of more than 70%, except for xylose. The bacteria poorly utilized all the inorganic carbon sources except for sodium carbonate, which resulted in flocculating activity of 62.5%. Glucose and maltose were favourable carbon sources for the test bacterium, with flocculating activities of 92.2% and 90.5%, respectively. Glucose was the most preferred and least expensive carbon source, and its preference for bioflocculant production has been reported [9,14,18,28,29,30].

It has been well documented that nitrogen source(s) are important nutrient factor(s) that enhance bioflocculant production [9,14,29]. Most microorganism utilise either organic or inorganic nitrogen sources, or both, to produce bioflocculant. Liu et al. [9] reported that Chryseobacterium daeguense W6 produces bioflocculant by using organic nitrogen sources for which tryptone was the most preferred, resulting in flocculating activity of more than 90%, while all inorganic nitrogen sources (sodium nitrate, ammonium sulphate and ammonium nitrate) resulted in poor flocculating activity. In another study, Piyo et al. [15] observed that Bacillus sp. Gilbert utilised an inorganic nitrogen source—ammonium chloride effectively, to produce bioflocculant with a flocculating activity of 91%, while organic nitrogen sources like urea and peptone were poorly utilised. For this current study all the tested organic and inorganic nitrogen sources were poorly utilised individually by the test bacterium as sole nitrogen sources, but the combination of urea, yeast extract and ammonium sulphate, as composed in the production medium resulted in high flocculating activity of 92.2% (Figure 2).

Documented reports revealed that initial pH values required for bioflocculant production differ for different organisms. For example Bacillus sp. PY 90 and Streptomycetes griseus produced bioflocculant under acidic conditions [31,32], while Virgibacillus sp. Rob and Bacillus licheniformis X14 optimally produced bioflocculant under alkaline pH [8,14], and Paecilomyces sp. [33], and Halomonas sp. OKOH [2], produce bioflocculant optimally under neutral pH. In the current study, the bacteria produced bioflocculant well in the range of pH 3–7, with the highest flocculating activity (90.5%) observed at a weak acidic pH of 6. Alkaline pH was unfavourable for the production of bioflocculant by the bacterium. Similar observations were reported for bioflocculant production by Chryseobacterium daeguense W6 [9], Klebsiella mobilis [17], and Serratia ficaria [18].

Metal ions either stimulate or inhibit bioflocculant production [8,9,34]. Among the mechanisms proposed for stimulation are: (1) neutralization and stabilization of the residual charge of functional group on the bioflocculant by the metal ions [35], and (2) increase in ionic strength of the suspension solution as a result of addition of metal ion; thereby, decreasing electrostatic forces of the suspended particles [36].

Bioflocculant production was stimulated by Na⁺, K⁺, Li⁺, Ca²⁺, Mg²⁺ and Mn²⁺ resulting in flocculating activities of over 90%, while Fe³⁺ inhibited bioflocculant production completely. Optimal flocculating activity (95%) was observed with Mn²⁺. In a similar report, Feng and Xu [34], reported that bioflocculant production by Bacillus sp. BF3-3 was stimulated by Al³⁺, Mg²⁺ , Ca²⁺, K⁺ and Na⁺ but inhibited by Fe³⁺. In another report, bioflocculant production was stimulated in the presence of Mg²⁺, Ca²⁺, K⁺ , Mn²⁺, but was inhibited by Ba²⁺, Fe³⁺, Al³⁺ [9].

Heating up to 100 °C for 25 min does not appear to significantly affect the flocculating activity of the bioflocculant, suggesting it to be very thermostable. The bioflocculant appears to be more stable than most similar compounds reported in literature. He et al. [37] reported that heating bioflocculant REA-11 (a purified bioflocculant) to 100 °C causes it to completely lose activity. Another bioflocculant (As-101) lost 50% of its flocculating activity when heated at 100 °C for 15 min [38]. Yokoi et al. [31] also reported that almost all residual flocculating activity was lost when the bioflocculant of Bacillus PY-90 was heated at 100 °C for 40 min. The thermal stability of this bioflocculant might be a result of the polysaccharide backbone in its structure, as is suggested elsewhere [39]. From the analysis of the chemical composition of the purified bioflocculant (2.1 g/L), the high content of uronic acid in the bioflocculant will provide large quantities of carboxyl and hydroxyl functional groups in the polysaccharide backbone, which will interact with each other to create a considerable amount of hydrogen bonds within the molecule, necessary for thermal stability of the bioflocculant.

During the exponential growth phase, flocculating activity increased steadily, until it peaked after 72 h (flocculating activity of 92.2%), beyond which flocculating activity started to decline. This trend suggests that the bioflocculant was produced by biosynthesis, and the decline after 72 h could be due to cell autolysis and/or presence of a bioflocculant-degrading enzyme [19,40]. A similar finding was reported for bioflocculant production by Serratia ficaria [18]; however, several studies reported different culture times for maximum bioflocculant production by different bacteria. For example, in Bacillus firmus, bioflocculant production peaked after 33 h [41]. In Agrobacterium sp. M-503, production peaked after 48 h [42]; while, for Vagococcus sp. W31, production peaked after 60 h [19]. Also, in Phormidium strain J-1, production peaked after 96 h [43]; while, for Halomonas sp. OKOH [2], and Bacillus sp. Gilbert [15], maximum flocculating activities were attained in 135 h and 240 h respectively. It was also observed that the bacterial growth profile ran almost parallel with bioflocculant production (Figure 6), suggesting that the bioflocculant production was by biosynthesis during the growth phase of the bacteria.

This study has shown Cobetia sp. OAUIFE to be a source of new thermostable acidic polysaccharide bioflocculant. The bioflocculant showed excellent flocculating activity of kaolin suspensions, and was optimally produced with glucose as a carbon source and urea, yeast extract and ammonium sulphate combined as nitrogen sources, in the presence of Mn²⁺ with weak acidic pH 6 . This is the first report implicating the Cobetia genus in bioflocculant production, the outcomes of which forms the basis of current in-depth ongoing studies on the organism, and we propose that Cobetia sp. OAUIFE has immense potential as a producer of bioflocculant that could be of relevance as alternatives to harmful inorganic and synthetic flocculants.