Chlorine Tolerance and Inactivation of Escherichia coli recovered from Wastewater Treatment Plants in the Eastern Cape, South Africa

This study investigated the survival of Escherichia coli (E. coli) recovered from secondary effluents of two wastewater treatment plants in the Eastern Cape Province, South Africa, in the presence of different chlorine concentrations. The bacterial survival, chlorine lethal dose and inactivation kinetics at lethal doses were examined. The bacterial isolates were identified by 16S rRNA gene sequencing. Comparison of the nucleotide sequences of 16S rRNA gene of bacteria with known taxa in the GenBank revealed the bacterial isolates to belong to Escherichia coli. At the recommended free chlorine of 0.5 mg/L, reduction of E. coli isolates (n = 20) initial bacterial concentration of 8.35–8.75 log was within a range of 3.88–6.0 log at chlorine residuals of 0.14–0.44 mg/L after 30 min. At higher doses, a marked reduction (p < 0.05) in the viability of E. coli isolates was achieved with a greater than 7.3 log inactivation of the bacterial population. Inactivation kinetics revealed a high rate of bacterial kill over time (R2 > 0.9) at chlorine dose of 1.5 mg/L. This study indicates poor removal of bacteria at free chlorine at 0.5 mg/L and a greater efficacy of 1.5 mg/L in checking E. coli tolerance.


Introduction
The growing demand for water for industrial, agricultural, environmental, municipal and domestic requirements has extended the requirements for an improvement in water treatment processes [1]. Water quality is a topical issue in public health due to concerns emanating from the indiscriminate discharge of inadequately treated sewage into water bodies, which is deleterious to human health and the environment [2,3]. The lack of access to good quality water increasingly impaired by the presence of waterborne pathogens continues to be a major contributor to the disease burden, morbidity, the retardation of economic growth and the well-being of the populace in many developing countries [4].
The use of an efficient water treatment system that relies on technologically compatible, cost-effective disinfectants that minimizes the production of disinfectant by-products [5] offers a safe margin for wastewater reuse. Commonly used disinfectants for water treatment include ultraviolet (UV) irradiation frequently used in large water and wastewater treatment plants which directly impairs the intracellular functions of microbial cells leading to growth inhibition and death [6]. UV irradiation does not produce disinfectant by-products [7], it is cost-intensive and requires large amounts of energy and frequent maintenance, including replacement of the UV lamps [8,9].
Ozone is highly efficient in the inactivation of viruses, bacteria and protozoa but decomposes rapidly. It also, requires special operation, and potentially forms bromate as a by-product in waters containing bromide [9,10]. Chlorine is a potent oxidant which causes the destruction of nucleic acids and cell membranes and it is an attractive option for disinfection due to its ease of handling, low capital cost and production of residual chlorine [11,12]. However, limitations in the use of chlorine include the production of off-tastes and odours, and potential formation of disinfection by-products (DBPs) such as trihalomethanes (THMs), Haloactic acids (HAAs), Haloacetylnitrile (HAN) and nitrosodimethylamine (NDMA) [13][14][15]. In addition, recovery of chlorine tolerant microbial pathogens at low chlorine dosages has been reported [16][17][18][19].
The Ct concept (disinfection concentration C multiplied by the contact time t over defined time intervals) has been used to model the effieicncy of many disinfection systems [20][21][22]. The detailed impact of disinfectant concentrations and exposure times required to sufficiently control microbial pathogens of public health hazards are useful in developing regulations and treatment strategies [6] and have been described by the Ct concept. Published data on Ct values include 21.7, 18.5 and 52.6 mg/L·min free chlorine for 3 log 10 inactivation of total coliforms, Enterococcus and Salmonella species [23] and 48.99-194.7 mg/L·min free chlorine for strains of Aspergillus and Penicillium [22].
Water regulatory bodies have prescribed guidelines for effluent discharge into water bodies [24]. For example, the South Africa water regulation stipulates a maximum contaminant level (MCL) of zero for total coliforms including E. coli and a residual chlorine concentration of 0.25 mg/L for effluent discharge into water bodies [25]. Microbiological parameters for routine monitoring of water do not regulate for all pathogens but rather stipulate indicator organisms, including E. coli, that specify a contamination problem with a disinfection step [1,16,21]. Strains of E. coli such as enterohemorrhagic E. coli O157:H7 have been associated with various outbreaks of waterborne and foodborne diseases such as bloody diarrhoea, haemolytic-uremic syndrome and sometimes death in immune-compromised individuals [26].
Previous studies have demonstrated the susceptibility of E. coli to chlorine. For example, a bench-scale inactivation study examined the inactivation of pathogenic and wild strains E. coli 0157:H7 after 1 min exposure to 1.1 mg/L free chlorine and 1.2 mg/L total chlorine and found significant removal of bacterial strains [27]. Likewise, greater than 99.9% inactivation of E. coli was achieved at the free chlorine concentration of 0.2 mg/L at contact time of 0.50 min [28]. Zhou et al. [29] also reported total sensitivity of bacterial pathogens including E. coli O157:H7 to free chlorine concentrations above 3.66 mg/L in wash water disinfection process.
However, studies have reported the survival of E. coli survival in chlorinated effluents [30,31]. In the study by [32] the tolerance of E. coli CGMCC to 10 mg/L chlorine dose was observed with only 4.0 log reduction achieved in a bacterial population of approximately 10 7 colony forming units (CFU/mL). In South Africa, studies have demonstrated a strong correlation between E. coli survival and low chlorine residuals [33][34][35]. For example, in the Eastern Cape, South Africa, [36] recovered E. coli strains in wastewater effluents containing low chlorine residuals ranged between 0.05 and 0.24 mg/L. However, there is paucity of information on chlorine tolerance limits and inactivation kinetics of E. coli in the Eastern Cape, South Africa thus necessitating the investigation on the efficacy of chlorine in the inactivation of E. coli isolated from secondary effluents of two wastewater treatment plants in the Eastern Cape Province, South Africa. Chlorine tolerance limits of E. coli strains were evaluated at varying chlorine concentrations. Data were fitted to Chick-Watson model and Hom model to decscribe inactivation kinetics.

Sample Collection and Processing
Secondary effluent samples were collected from the clarifier of two wastewater treatment plants in the Eastern Cape Province, South Africa and was processed for the isolation of E. coli following standard procedures [37,38] with some modifications. Twenty milliliters of wastewater samples were filtered through sterile cellulose-nitrate membrane filter (0.45 µm pore size, 47 mm diameter, Millipore filters) under partial vacuum in five replicates. The membrane filters were immediately placed in Petri dishes containing E. coli Chromogenic agar (Conada Pronadisa, Madrid, Spain) with sterile forceps. The agar plates were incubated at 37 • C for 24 h. After incubation, isolates showing dark blue colonies (typical of E. coli) were recovered, purified and stored in 20% glycerol at −80 • C.

Preparation of Stock Chlorine Solution
A stock chlorine solution (1% w/v) containing 7000 mg/L of free chlorine was prepared from calcium hypochlorite granules. This stock solution was diluted to a final free chlorine concentration of 0.5 mg/L and then quantified by the N,N-diethyl-ρ-phenylenediamine (DPD) method [38]. The free chlorine concentration was measured using free chlorine kit (Hanna instruments Inc. Woonsocket, RI, USA). The photometer has an accuracy of measurement of ±3% of reading at 25 • C and a sensitivity of 0-5.0 mg/L.

DNA Extraction
The genomic DNA of the presumptive E. coli isolates was extracted by the boiling method as described by [39,40] with modifications. Briefly, previously stored glycerol stocks of E. coli were resuscitated in Tryptic Soy Broth (TSB) at 37 • C for 24 h. The isolates were then purified and cultured on nutrient agar at 37 • C for 24 h. Single colonies from presumptive E. coli isolates and a positive control of E. coli (ATCC 3695) were inoculated into 200 µL of sterilized distilled water in sterile Eppendorf tubes. Cells were lysed at 100 • C for 10 min and centrifuged at 15,000 rpm for 10 min at 4 • C. The supernatant was collected and stored at −20 • C for further use.

Molecular Identification of Presumptive E. coli Isolates
The identities of the presumptive E. coli isolates were confirmed by amplification of the uidA gene in combination with the cultural characteristic of the isolate as previously reported by [41]. A reference strain of E. coli ATCC 3695 was used as positive control. The uidA gene amplification was carried out in a final volume of 25 µL mixture containing 12 µL of TaqMan DNA polymerase Master Mix (New England Biolabs Inc., Ipswich, MA, USA), 6 µL of nuclease free water, 1 µL of each primer UidA F:5 (AAAACGGCAAGAAAAAGCAG) and uidA R: 5 (ACGCGTGGTTAACAGTCTTGCG) and 5 µL of DNA template. The PCR mixture was subjected to a 5-min denaturation step at 94 • C, followed by 35 cycles of 30 s at 95 • C, 60 s for 58 • C and 60 s at 72 • C and a final elongation step for 8 min at 72 • C. The samples were kept at −20 • C until analysis. PCR products were confirmed in 1% Agarose gel electrophoresis in 1 × Tris-Borate-EDTA (TBE) buffer at 100 V for 60 min, visualized after staining with Ethidium bromide in ALLIANCE 4.7 UV transilluminator and photographed. Thereafter, three E. coli isolates with the highest bacterial counts were sequenced using the 16S rRNA gene analysis. The nucleotides sequences were compared to known sequences in the GenBank and were submitted to the Basic Local Alignment Search tools (BLAST) search engine at the National Centre for Biotechnological Information (NCBI) GenBank.

Bacterial Survival at the Recommended Free Chlorine Concentration (0.5 mg/L)
A chlorine disinfection assay was carried out for E. coli isolates (n = 20) in a bench-scale inactivation study to determine free chlorine residuals and bacterial survival after 30 min exposure. Prior to the disinfection experiment, the glycerol stock of the isolates was resuscitated in 5 mL of Tryptic Soy Broth and incubated at 37 • C for 24 h. Subsequently, a loop of culture streaked on nutrient agar and grown overnight at 37 • C for 18 h were aseptically transferred to sterile tubes. Cells were harvested by centrifugation at 6000 g for 5 min, washed twice with sterilized phosphate buffer (PBS), and re-suspended in sterile PBS at pH 7 and 22 • C to obtain a concentration of approximately 2.26 × 10 8 CFU/mL as the initial bacterial concentration. Bacterial concentrations in suspensions were measured at an absorbance of 600 nm, using a spectrophotometer (Helios Epsilon, Kinesis Inc., Berlin Township, NJ, USA) and confirmed by plating 0.1 mL portions of dilutions on E. coli chromogenic agar (Merck). Plates were incubated at 37 • C for 24 h.
The disinfection experiment was carried out by the addition of Calcium hypochlorite (1% w/v) containing 0.5 mg/L free chlorine to 100 mL of PBS containing E. coli isolates in 250 mL sterile bottles at pH 7.0 and 22 • C for 30 min. The reaction mixture was stirred continuously at 160 rpm on a magnetic stir plate. After 30 min, duplicates of 10 mL of treated bacterial suspensions were withdrawn and analysed for the residual chlorine concentration by the DPD method [38]. Chlorine residues were neutralized by the addition of 25 µL sodium thiosulfate (3% w/v). After series of dilutions, triplicates of one hundred microliters from the final dilutions were spread on E. coli chromogenic agar and plates were incubated at 37 • C for 24 h. A disinfection assay was repeated twice for each set of experimental conditions and triplicates samples of untreated water were processed as disinfection control.

Determination of Lethal Dose of Chlorine
The disinfection efficacy at higher doses of chlorine was assessed against three E. coli isolates following the method described by [28]. Three E. coli isolates, E. coli SAMRC-1, E. coli SAMRC-2, E. coli SAMRC-3 were selected based on the high bacterial survival at 0.5 mg/L chlorine concentration. At initial bacterial concentration of approximately 1.6-1.7 × 10 8 CFU/mL, Calcium hypochlorite (1% w/v) was added to 100 mL of 0.1 M phosphate-buffered suspension of E. coli isolates contained in 250 mL sterile bottles at pH 7 and at 22 • C to achieve chlorine concentrations ranging between 0.75 and 1.5 mg/L. The mixture was stirred continuously at 160 rpm on a magnetic stir plate for 30 min. After 30 min exposure, the reaction was immediately terminated by the addition of sodium thiosulfate (3% w/v). After a series of dilutions of the treated suspension, triplicates of one hundred microliters of the final dilution were transferred to E. coli Chromogenic agar and incubated at 37 • C for 24 h. The means of residual chlorine concentrations were also determined in duplicates and the disinfection assay was repeated twice.

Inactivation Kinetics of E. coli Isolates
Inactivation kinetics was carried out by subjecting the E. coli isolates to chlorine lethal dose of 1.5 mg/L at intervals of 10 min over 30 min exposure. At an initial bacterial concentration of approximately 1.6-1.7 × 10 8 CFU/mL, thirty-seven microliters of 1% (w/v) Calcium hypochlorite was added to 100 mL of bacterial suspension. The mixture was continuously stirred at 160 rpm on a magnetic stir plate for 30 min. At intervals of 10 min, aliquots of 10 mL were withdrawn from the mixture and analysed for a residual chlorine concentration. Aliquots were also withdrawn from the mixture at 10 min interval and instantaneously neutralized by the addition of 3% w/v sodium thiosulfate. An enumeration of the bacteria was carried out by spreading one hundred microliters of the final dilution of each isolate on E. coli chromogenic agar and plates which were incubated at 37 • C for 24 h. Residual chlorine concentrations were determined in duplicates while bacterial survival was carried out in triplicates.

Data Analysis
The removal efficiency by the chlorine concentrations were expressed as log 10 reduction of bacterial population as described by [42]: LR = log reduction of bacteria count at time t; N 0 = initial bacterial concentration at time 0, N t = final bacterial concentrations after a treatment time t.
The disinfection kinetic parameters of E. coli at higher chlorine concentrations were determined by fitting inactivation data to the Chick-Watson empirical model [43]. The empirical logarithm equation was expressed as: where, N = bacterial concentration at time t, N 0 = initial bacterial concentration at time 0, C = free chlorine concentration (mg/L), T= contact time (min), k = inactivation rate constant, n = coefficient of dilution. For log reduction/time plots for survivors, inactivation data was fitted to the empirical model of [43,44]. Data were log 10 transformed and statistical analysis was performed by linear regression using Origin Pro 2017.
The Hom model is expressed as: where N = bacterial concentration at time t, N 0 = initial bacterial concentration at time 0, C = residual free chlorine concentration (mg/L) at time t, T = contact time (min), k = inactivation rate constant, n = chlorine exponent (Hom model), m= time exponent Using multiple linear regression analysis, the constants were evaluated as: Statistically significant differences were determined between untreated and chlorine treated E. coli strains. Statistically significant differences in inactivation were also determined among the chlorine treated E. coli strains by one-way analysis of variance (ANOVA) and Turkey Post Hoc comparison test using SPSS (IBMSPSS Statistics 23). Differences between means were considered significant at p < 0.05.

Molecular Confirmation of Presumptive E. coli Isolates
Molecular identification showed the presumptive bacterial isolates to belong to E. coli. The three isolates with the highest chlorine tolerant profiles were identified by 16S ribosomal RNA gene sequencing and BLAST results revealed them to have 99% similarities to E. coli. The nucleotide sequences of the three E. coli isolates were deposited in GenBank as E. coli SAMRC-1 (accession number KX874327), E. coli SAMRC-2 (accession number KX874328) and E. coli SAMRC-3 (accession number KX874329).

Bacterial
Survival at Free Chlorine Concentration of 0.5 mg/L for 30 min Table 1 shows results of bacterial reductions and free chlorine residuals obtained from E. coli suspensions (n = 20) after initial screening at 0.5 mg/L chlorine concentration. At initial bacterial concentration of 8.5-8.8 log, bacterial reductions ranged from 3.88-6.0 log after 30 min and residual chlorine concentrations were 0.29-0.44 mg/L. The lowest bacterial reductions were observed for three strains of E. coli (E. coli SAMRC-1, E. coli SAMRC-2 and E. coli SAMRC-3) at 4.84 log (3.23 × 10 3 CFU/mL), 4.37 log (9.67 × 10 3 CFU/mL) and 3.87 log (3.03 × 10 4 CFU/mL), respectively. Although residual chlorine concentrations were high and within the recommended free chlorine of 0.25 mg/L free chlorine for South African wastewater effluent discharge [25], many isolates still showed high bacterial counts at these chlorine concentrations. The poor removal efficiency of 0.5 mg/L in eliminating E. coli suggests tolerance of the bacteria to 0.5 mg/L chlorine concentration. In a related study, poor removal efficiency of 0.5 mg/L free chlorine was observed for Citrobacter sp. where only 4-5 log reduction (initial bacterial population of 8.2 log) was achieved in chlorinated bacterial suspension [19]. Another study also reported recovery of high counts of E. coli ranging between 1.24 and 4.95 log 10 CFU/mL in wastewater effluent samples containing chlorine residuals of 0.24-0.44 mg/L [45]. Similarly, [42] also documented significant regrowth and reactivation of E. coli and Salmonella species in reclaimed water samples with chlorine levels of 0.2 and 0.5 mg/L thus indicating the inefficiency of 0.5 mg/L chlorine dosing in removing E. coli as observed in this study.
The recommended Ct values for chlorination corresponds to 15 mg/L·min at pH of less than 7.5 and temperature above 10 • C [21, 29,46]. In this study, a Ct value of 15 mg/L·min achieved 3-4 log inactivation of E. coli at pH 7.0 and 22 • C. However, this Ct value was lower than the Ct value obtained from a previous study by [23] where Ct for 3-log inactivation for total coliforms was estimated at 21.7 mg/L·min. The differences in the Ct values may be due to reclaimed water samples used for chlorine disinfection compared to chlorine demand-free water used in this study.
The survival of the test E. coli isolates at the recommended free chlorine concentration of 0.5 mg/L indicates a compromise of water quality as the survival of E. coli in chlorine treated water signifies the existence of other microbial pathogens [21] and is a threat to public health and environmental safety.

Effect of Lethal Dose of Chlorine on E. coli Survival
To investigate the efficacy of increasing chlorine doses on the inactivation of the test E. coli isolates, three E. coli strains in suspensions (initial bacterial concentration of 8.2-8.23 log) were exposed to chlorine concentrations of 0.75-1.5 mg/L for 30 min. The survival of the three E. coli isolates at higher free chlorine concentrations were compared with survival at 0.5 mg/L (Figure 1).

Effect of Lethal Dose of Chlorine on E. coli Survival
To investigate the efficacy of increasing chlorine doses on the inactivation of the test E. coli isolates, three E. coli strains in suspensions (initial bacterial concentration of 8.2-8.23 log) were exposed to chlorine concentrations of 0.75-1.5 mg/L for 30 min. The survival of the three E. coli isolates at higher free chlorine concentrations were compared with survival at 0.5 mg/L (Figure 1). Water guidelines recommend a maximum contaminant level (MCL) of zero for total coliforms including E. coli for effluent discharge into water bodies [26] and a greater removal efficiency of E. coli isolates was achieved at higher chlorine doses of 0.75-1.5 mg/L compared to log inactivation at 0.5 mg/L in 30 min. This might be due to the presence of higher chlorine species reacting with bacteria cells which causes greater inactivation of bacteria. It has been hypothesized that solutions with higher oxidizing species have higher oxidation potential to inactivate bacteria [28].
A study also reported greater than 5 log reductions of Shiga-toxigenic E. coli serotypes 026 and 0103 after treatment with a chlorine dose of 1.0 mg/L in 5 min [47]. High sensitivity of E. coli O157:H7 and Listeria monocytogenes to free chlorine concentrations equal to or higher than 1.0 mg/L was recorded with no detection of bacteria after chlorine treatment [48]. The Minimum Inhibitory Concentration (MIC) of planktonic bacterial isolates was also achieved at free chlorine concentrations ranging between 1.0 and 2.0 mg/L [16]. Low percentages of intact cells of coliforms and E. coli were found in drinking water samples after treatment with free chlorine concentrations above 0.5 mg/L, while higher percentages of intact cells were observed at free chlorine concentrations below 0.5 mg/L [49].  At 0.75-1.25 mg/L free chlorine, the removal efficiency of E. coli SAMRC-1 was 6.41-6.45 log and virtually 100% inactivation was achieved at free chlorine concentration of 1.5 mg/L within 30 min at residual chlorine concentrations of (0.29 ± 0.02)-(0.9 ± 0.05) mg/L. In comparison, chlorine concentration of 0.5 mg/L achieved the lowest log removal of 4.1 log. For E. coli SAMRC-2, a removal efficiency of 6.2-7.9 log was achieved at 0.75-1.25 mg/L free chlorine and complete inactivation of the E. coli population at chlorine concentration of 1.5 mg/L at residual chlorine concentrations of (0.55 ± 0.05)-(1.23 ± 0.04) mg/L. In contrast, chlorine concentration of 0.5 mg/L showed a lower removal efficiency of 4.2 log. The removal efficiency for E. coli SAMRC-3 ranged between 6.2 and 6.9 log at chlorine concentrations of 0.75-1.25 mg/L while the highest removal of 7.3 log units was achieved at chlorine concentration of 1.5 mg/L. Residual chlorine concentrations after 30 min were (0.42 ± 0.08)-(1.22 ± 0.01) mg/L. In contrast, 0.5 mg/L achieved disinfection efficiency of 3.7 log.
Water guidelines recommend a maximum contaminant level (MCL) of zero for total coliforms including E. coli for effluent discharge into water bodies [26] and a greater removal efficiency of E. coli isolates was achieved at higher chlorine doses of 0.75-1.5 mg/L compared to log inactivation at 0.5 mg/L in 30 min. This might be due to the presence of higher chlorine species reacting with bacteria cells which causes greater inactivation of bacteria. It has been hypothesized that solutions with higher oxidizing species have higher oxidation potential to inactivate bacteria [28].
A study also reported greater than 5 log reductions of Shiga-toxigenic E. coli serotypes 026 and 0103 after treatment with a chlorine dose of 1.0 mg/L in 5 min [47]. High sensitivity of E. coli O157:H7 and Listeria monocytogenes to free chlorine concentrations equal to or higher than 1.0 mg/L was recorded with no detection of bacteria after chlorine treatment [48]. The Minimum Inhibitory Concentration (MIC) of planktonic bacterial isolates was also achieved at free chlorine concentrations ranging between 1.0 and 2.0 mg/L [16]. Low percentages of intact cells of coliforms and E. coli were found in drinking water samples after treatment with free chlorine concentrations above 0.5 mg/L, while higher percentages of intact cells were observed at free chlorine concentrations below 0.5 mg/L [49].
The difference between the control and chlorine treated E. coli strains was not statistically significant (p < 0.05) at all the treatment levels and there were no statistically significant differences (p < 0.05 in inactivation among the chlorine treated E. coli isolates at 0.75-1.5 mg/L. However, a chlorine concentration of 1.5 mg/L was found to be optimal for E. coli inactivation.

Inactivation of E. coli at Free Chlorine Dose of 1.5 mg/L after 30 min Exposure
The inactivation of the three E. coli strains initial bacterial density of approximately 8.2 log by chlorine disinfectant was assessed at the initial chlorine concentration of 1.5 mg/L. Bacterial survival was progressively monitored at 10 min interval over 30 min exposure (Figures 2-4). The difference between the control and chlorine treated E. coli strains was not statistically significant (p < 0.05) at all the treatment levels and there were no statistically significant differences (p < 0.05 in inactivation among the chlorine treated E. coli isolates at 0.75-1.5 mg/L. However, a chlorine concentration of 1.5 mg/L was found to be optimal for E. coli inactivation.

Inactivation of E. coli at Free Chlorine Dose of 1.5 mg/L after 30 min Exposure
The inactivation of the three E. coli strains initial bacterial density of approximately 8.2 log by chlorine disinfectant was assessed at the initial chlorine concentration of 1.5 mg/L. Bacterial survival was progressively monitored at 10 min interval over 30 min exposure (Figures 2-4).   The difference between the control and chlorine treated E. coli strains was not statistically significant (p < 0.05) at all the treatment levels and there were no statistically significant differences (p < 0.05 in inactivation among the chlorine treated E. coli isolates at 0.75-1.5 mg/L. However, a chlorine concentration of 1.5 mg/L was found to be optimal for E. coli inactivation.

Inactivation of E. coli at Free Chlorine Dose of 1.5 mg/L after 30 min Exposure
The inactivation of the three E. coli strains initial bacterial density of approximately 8.2 log by chlorine disinfectant was assessed at the initial chlorine concentration of 1.5 mg/L. Bacterial survival was progressively monitored at 10 min interval over 30 min exposure (Figures 2-4).    Bacterial inactivation was rapid within the first 10 min while inactivation rate slowly declined afterwards with increase in contact time. Previous studies have shown that the highest reduction of bacterial population occurred within 10-15 min of exposure [50,51].
A tailing-off effect observed for E. coli strains is shown in Figure 2. The tailing-off implies a shielding phenomenon in which a certain portion of the bacterial population, after inactivation by chlorine, returns the bacteria suspension into its initial state, but with a lower number of viable cells. Further reaction between the chlorine species and bacteria then results in inactivation of smaller portions of viable bacteria due to shielding by previously inactivated cells [52]. Similar observations of tailing-off effects have been reported in some inactivation experiments. For example, [53] reported a tailing phase for total coliforms in raw grey water treated with 1-5 mg/L free chlorine for 120 min. Similarly, [54] reported a tailing-off for E. coli O157 in the chlorine-demand free buffer and [55]   Bacterial inactivation was rapid within the first 10 min while inactivation rate slowly declined afterwards with increase in contact time. Previous studies have shown that the highest reduction of bacterial population occurred within 10-15 min of exposure [50,51].
A tailing-off effect observed for E. coli strains is shown in Figure 2. The tailing-off implies a shielding phenomenon in which a certain portion of the bacterial population, after inactivation by chlorine, returns the bacteria suspension into its initial state, but with a lower number of viable cells. Further reaction between the chlorine species and bacteria then results in inactivation of smaller portions of viable bacteria due to shielding by previously inactivated cells [52]. Similar observations of tailing-off effects have been reported in some inactivation experiments. For example, Winward et. al. [53] reported a tailing phase for total coliforms in raw grey water treated with 1-5 mg/L free chlorine for 120 min. Similarly, Van Haute et. al. [54] reported a tailing-off for E. coli O157 in the chlorine-demand free buffer and [55]   These Ct values were higher than compared to the Ct values for 7 log inactivation of E. coli by free chlorine obtained in previous studies. For example, Ref. [56] observed greater than 7-log inactivation of E. coli O157:H7 at a Ct value of 0.25 mg/L·min. The differences in inactivation may be due to higher initial bacterial density (10 8 CFU/mL) of E. coli used in this study compared to initial bacterial density (10 6 CFU/mL) used in the previous study.

Inactivation Kinetics of Bacteria by Chlorine at 30 min Contact Time
In order to evaluate the kinetics of E. coli inactivation at higher chlorine concentrations (0.75-1.5 mg/L) after 30 min exposure, inactivation data were fitted to the Chick-Watson empirical model (Table 2) to determine disinfection kinetic parameters, R 2 (coefficient of determination), inactivation rate constant, k, and coefficient of dilution, n. From the linear regression, R 2 values ranged between 0.64 and 0.77. Significant differences (p < 0.5) in inactivation rate constant k were observed for each tested E. coli strain at −1.99 ± 1.43, −1.33 ± 1.36 and −1.21 ± 0.08 mg/L for E. coli SAMRC-1, E. coli SAMRC-2 and E. coli SAMRC-3, respectively. The inactivation rate constant k, describes the sensitivity of bacteria to inactivation by disinfectants [57]. The highest inactivation rate constant k (−1.99 ± 1.43 mg/L) observed for E. coli SAMRC-1 suggests higher sensitivity of E. coli SAMRC-1 to free chlorine while E. coli SAMRC-3 with These Ct values were higher than compared to the Ct values for 7 log inactivation of E. coli by free chlorine obtained in previous studies. For example, Zhao et al. [56] observed greater than 7-log inactivation of E. coli O157:H7 at a Ct value of 0.25 mg/L·min. The differences in inactivation may be due to higher initial bacterial density (10 8 CFU/mL) of E. coli used in this study compared to initial bacterial density (10 6 CFU/mL) used in the previous study.

Inactivation Kinetics of Bacteria by Chlorine at 30 min Contact Time
In order to evaluate the kinetics of E. coli inactivation at higher chlorine concentrations (0.75-1.5 mg/L) after 30 min exposure, inactivation data were fitted to the Chick-Watson empirical model (Table 2) to determine disinfection kinetic parameters, R 2 (coefficient of determination), inactivation rate constant, k, and coefficient of dilution, n. From the linear regression, R 2 values ranged between 0.64 and 0.77. Significant differences (p < 0.5) in inactivation rate constant k were observed for each tested E. coli strain at −1.99 ± 1.43, −1.33 ± 1.36 and −1.21 ± 0.08 mg/L for E. coli SAMRC-1, E. coli SAMRC-2 and E. coli SAMRC-3, respectively. The inactivation rate constant k, describes the sensitivity of bacteria to inactivation by disinfectants [57]. The highest inactivation rate constant k (−1.99 ± 1.43 mg/L) observed for E. coli SAMRC-1 suggests higher sensitivity of E. coli SAMRC-1 to free chlorine while E. coli SAMRC-3 with the lowest value of k, reflects a more resistant bacterial population to free chlorine. The coefficient of dilution, n values were 4.66, 5.07 and 4.67 for E. coli SAMRC-1, E. coli SAMRC-2 and E. coli SAMRC-3, respectively. Differences in n represent the average number of disinfectant molecules in contact with an organism available to cause inactivation [58] and this was reflected in the increased inactivation for E. coli SAMRC-2 (higher value of n = 5.07) compared to the other E. coli strains. Variations in n indicate physiological differences between organisms, inactivation conditions and availability of nutrients in the cultivation media [58]. Difference in the values of k and n for two strains of E. coli, K12 and 0157:H7 exposed to chlorine concentrations of 0.1-8.0 mg/L for 10 min have been reported [59].
Inactivation kinetics of E. coli at optimal chlorine dose of 1.5 mg/L at intervals of 10 min over 30 min exposure were evaluated using the Chick-Watson (Table 3) and Hom model (Table 4) model. Table 3. Inactivation kinetics of E. coli at chlorine concentrations of 1.5 mg/L (Chick-Watson model: log N/N 0 = −kC n T).

Bacterial Strains
Free Chlorine (mg/L) Comparatively, the Hom model gave a good fit for inactivation of E. coli by chlorine disinfectant with R 2 values ranging from 0.9 to 0.98 while Chick-Watson model showed with R 2 values of 0.91-0.95. Overall, these results indicate a strong correlation between bacterial inactivation and chlorine dose over time and bacterial isolates were adequately exposed to reacting chlorine species during the disinfection process. Values obtained for disinfection exponent n ranged between 1.57 and 22.43 while time exponent m had values of −0.23-0.17. For inactivation reactions with coefficient n greater than 1, the chlorine dose has a greater influence on bacterial inactivation than contact time [60]. These results suggest a greater impact of chlorine dose of 1.5 mg/L on bacterial inactivation than the effect of exposure time. A chlorine dose of 1.5 mg/L could serve as an alternative dose for control of E. coli and some bacteria species of public health concern in wastewater treatment plants which will ensure protection of the public and environmental health.

Conclusions
This study evaluated the efficacy of chlorine disinfectant in the inactivation of some E. coli isolates recovered from secondary effluent samples from the clarifier of two selected wastewater treatment plants in the Eastern Cape Province, South Africa. Treatment at 0.5 mg/L free chlorine was ineffective in eliminating bacteria isolates and the survival of E. coli demonstrated a high level of E. coli tolerance to chlorine in all the suspensions at the recommended dose of 0.5 mg/L. Further treatment at higher chlorine concentrations was more effective in inactivating E. coli isolates and complete inactivation of E. coli population was achieved at optimum dose of 1.5 mg/L. For optimization of disinfection processes, the chlorine dose required to achieve sufficient disinfection of wastewater appears to demand a review of current guidelines as observed for 0.5 mg/L free chlorine especially in resource-limited countries dependent on cost-effective and available disinfectant for wastewater treatment. These results demonstrate the potentials of 1.5 mg/L free chlorine obtained in this study as an effective dose to control E. coli tolerance in chlorinated effluents and is hereby proposed.