Influence of groundwater characteristics on the survival of enteric viruses
S. Toze, Private Bag No. 5, PO Wembley WA, Australia 6913 (e-mail: firstname.lastname@example.org).
Aims: This study was undertaken to further understand the processes affecting the persistence of enteric viruses in groundwater.
Methods and Results: Varying temperature, oxygen and nutrient levels were tested in the presence and absence of groundwater micro-organisms to determine which of the factors tested had dominant influence on the decay of Escherichia coli, the bacteriophage MS2, poliovirus and coxsackievirus. The results indicated that the most influential factor affecting the decay of the viruses and E. coli was the presence of groundwater micro-organisms. The results also implied that temperature, the presence of oxygen and nutrient levels indirectly influence viruses and E. coli decay by influencing the activity of the groundwater micro-organisms.
Conclusions:E. coli and the viruses displayed maximum decay under aerobic conditions, at 28°C without the addition of nutrients in the presence of groundwater micro-organisms.
Significance and Impact of the Study: The results suggest that if the mode of action of the groundwater micro-organisms could be determined then the decay of viral pathogens in recharged waters may be more easily predicted.
Selected aquifers have potential use for the storage of surplus water for later recovery. This can be an important part of improving and enhancing urban water supplies. As communities continue to expand and dependence on water sources such as groundwater and reclaimed water grows, potential hazards involved in using a resource such as reclaimed water must be better understood. Amongst the most important and ubiquitous of these risks are water-borne pathogens, whose persistence in aquifers is still poorly understood.
There are two major advantages involved in water reclamation – the additional water supply and reducing environmental pollution (Dillon et al. 1999). One issue influencing the use of reclaimed water is storage during periods of low demand. One potential means of short- to long-term storage is artificial recharge of reclaimed waters into suitable aquifers. Artificial recharge into aquifers is achieved by either infiltrating water directly from the surface; by direct injection into an aquifer via injection wells; or by altering conditions of the surface so as to increase the natural filtration of water. The principal reason for using artificial recharge is to store excess water for later reclamation. However, artificial recharge also has the advantage in that it affords a means of improving the quality of the recharged water through its storage in situ (Dillon et al. 1999).
It is currently known and accepted that microbial pathogens such as enteric viruses lose viability in groundwater (Keswick et al. 1982; Yates et al. 1985). However, the factors that are most influential in the decay of the pathogens, particularly viruses, are not well understood. The factors influencing viral decay in groundwater are temperature (Melnick and Gerba 1980; Yates et al. 1990) and the type of pathogenic organism (Yates and Yates 1988; Jansons et al. 1989). Other abiotic factors include dissolved oxygen (Jansons et al. 1989), water chemistry such as pH, electrical conductivity and organic carbon concentrations (Katzenelson 1978), water source (Melnick and Gerba 1980) and the presence of indigenous micro-organisms (Yates and Yates 1988; van Leeuwen 1996).
The purpose of this study was to compare the influence of several factors implicated in pathogen decay in groundwater and to determine which was the most influential on the decay of enteric viruses, male-specific bacteriophage and E. coli.
Groundwater samples were collected from the superficial aquifer located on the Swan Coastal Plain Perth (Western Australia). Anoxic groundwater was collected from a well using a submersible pump. Groundwater was pumped for 30 min to purge the bore prior to collection of the samples. Samples were then collected in sterile 1-l borosilicate bottles. When anoxic groundwater was required for experiments, the bottles were flushed with nitrogen and sealed with a silicon septum prior to sterilization to prevent contact with atmospheric oxygen. Groundwater, tested at the beginning of the study for background levels of E. coli, coliphages such as MS2 and enterovirus, was found to be free of these micro-organisms on all occasions.
Viruses and indicator micro-organisms
The micro-organisms tested were poliovirus type 1 (PL 0692), coxsackievirus B1 (ATCC VR-28), Escherichia coli (ACM 1803) and the male-specific bacteriophage MS2 (ATCC 15597-B1).
The E. coli cells were grown overnight in nutrient broth at 37°C with shaking, then washed three times in sterile groundwater and resuspended in sterile groundwater. Cell numbers in the groundwater suspension were determined using the most probable number (MPN) method in nutrient broth. The MS2 bacteriophage was grown using the E. coli host HS(pFamp)R (ATCC 700881). Bacteriophages were collected by scraping the E. coli growth plus bacteriophage from the agar plate and resuspending in sterile groundwater. The suspension was centrifuged at 2000 × g for 2 min to pellet the E. coli cells and cell debris, and the supernatant transferred to a fresh tube. The viral suspension was then washed three times in sterile groundwater to remove excess nutrients carried over from the E. coli lawn using the JumboSepTM with 100K MWCO filters (Pall, Melbourne, Australia). The bacteriophages were then suspended in sterile groundwater to ca 1010 plaque forming units (PFU) ml−1 and the exact number of phage particles determined by plating serial dilutions using the double agar layer plaque method for detection of F-RNA bacteriophage (Ferguson et al. 1996).
Poliovirus and coxsackievirus were grown in cell culture of Vero cell lines (African Green Monkey Kidney cells) by the Pathology Centre (Pullman, WA, USA) and then harvested from the cultures and frozen as a crude cell extract at −80°C until needed. The initial number of infective viral particles in the viral suspensions was determined by the Pathology Centre using the MPN method in fresh cell culture. The titre for each virus was determined to be 109 PFU ml−1 for poliovirus and 108 PFU ml−1 for coxsackievirus. The detection limit for both viruses using reverse transcriptase-polymerase chain reaction (RT-PCR) was determined by making serial 10-fold dilutions of extracted viral RNA and determining the lowest detectable dilution. The lowest detectable dilution was 10−6 for both viruses which equated to a detection limit of ca 10 or less viral RNA molecules per PCR reaction (based on original MPN titre). Prior to use in the survival experiments, volumes of the viral suspension were centrifuged at 2000 × g to remove the cell debris and then washed three times in sterile groundwater using the JumboSepTM with 100 K MWCO filters (Pall) to remove excess cell culture medium. The viral suspensions were finally resuspended in sterile groundwater to their original viral titre (assessed by comparing RT-PCR results with non-washed viral suspensions).
All experiments were undertaken by seeding groundwater with known numbers of one of the viruses or E. coli cells. Experiments were undertaken under anoxic conditions in an anaerobic glovebox using bottles that had been purged with nitrogen and sealed with butyl rubber stoppers prior to sterilization.
The groundwater samples (1 l) were divided into two 500-ml aliquots. One 500-ml aliquot was passed through a 0·2 μm filter to remove the indigenous groundwater micro-organisms (termed ‘filtered’) while the second portion was left unfiltered (termed ‘unfiltered’). From each of the 500 ml non-filtered and filtered aliquots, three 100 ml subsamples were taken and placed in sterile bottles. The 100 ml samples were then inoculated with poliovirus, coxsackievirus, E. coli or MS2 bacteriophage to a final concentration of ca 106 viral particles or cells ml−1. Three 10 ml replicates were then taken from each inoculated 100 ml subsample and placed into 15 ml polypropylene centrifuge tubes. All tubes were incubated under conditions described by the following ‘temperature’, ‘oxygen’ or ‘nutrient’ experiments.
Temperature experiments were carried out at 15 and 28°C. The tubes of all replicates of each sample were incubated under aerobic conditions in static incubators.
Oxygen experiments were undertaken in the presence and absence of oxygen. All tubes used in this test were incubated at 28°C. Tubes incubated in the presence of oxygen were placed in a shaking water bath while the tubes tested under anoxic conditions were incubated in a heated anaerobic chamber.
Nutrient experiments were undertaken by the addition to groundwater samples of either peptone (0·1% final concentration) as a protein-based nutrient source or glucose (0·01% final concentration) as a simple, non-protein carbon nutrient source. Concentrated sterilized peptone or glucose in groundwater were filter-sterilized and added to the collected 1-l groundwater samples prior to the groundwater being divided into 500 ml filtered and non-filtered aliquots. The resulting final replicates of all samples were incubated at 28°C, under aerobic conditions.
Sample collection was undertaken on days 0, 2, 5, 7 and then every 7 days until day 45 for the majority of replicate tubes from all experiments. However, because of time constraints, some experiments were sampled on days 0, 4, 7, 14 and 21 only. All the collected samples were processed within 2 h of collection.
Analysis of viable E. coli and MS2 and detectable viral RNA
All analyses for each micro-organism were performed in triplicate. The numbers of E. coli and the bacteriophage MS2 at each time interval were detected by direct culture while poliovirus and coxsackievirus numbers were determined by detecting the number of detectable copies of the RNA genome using RT-PCR.
For E. coli and MS2, prior to plating, samples were diluted serially (one in 10) in autoclaved groundwater to obtain between 30 and 100 colonies or plaques per plate. The final dilutions plated varied depending on the expected number of viable cells or infective phage present in the collected groundwater samples. This was determined from the results of the previous sampling occasion. Normally two or three sequential dilutions were plated on each sampling occasion to ensure that a countable result was obtained.
Escherichia coli was detected by spread plating 100 μl of appropriate dilutions on ChromocultTM agar (Merck, Melbourne, Australia). Inoculated plates were incubated at 37°C overnight and then characteristic E. coli colonies were counted to determine the average colony forming units (CFU) ml−1 in the original sample. ChromocultTM agar was used as it prevents the growth of the majority of non-E. coli groundwater bacteria. These few groundwater micro-organisms that are capable of growing on the ChromocultTM medium are easily distinguished from E. coli by the blue–violet colour of E. coli on this medium.
Enumeration of MS2 was achieved by plating three 100 μl replicates of appropriate dilutions using the double agar layer plaque method, with E. coli HS(pFamp)R (ATCC 700881) as the host (Ferguson et al. 1996). All inoculated plates were incubated for 24 h at 37°C. Following incubation, the number of MS2 plaques on each plate were counted and used to determine the number of PFU ml−1 in the original sample.
Changes in detectable poliovirus and coxsackievirus RNA numbers were determined using quantitative real-time RT-PCR. Enterovirus RNA was extracted from 250 μl aliquots using the High PureTM Viral RNA Extraction Kit (Roche Diagnostics, Melbourne, Australia). Quantitative RT-PCR was performed using the TitanTM One-Step RT-PCR kit (Roche Diagnostics) and the Ent-up and Ent-down primers (Abbaszadegan and DeLeon 1997). The RT-PCR protocol described by Roche Diagnostics was amended with the addition of Sybr Green (final concentration of 0·1×). Quantitation was undertaken using an iCycler Thermocycler (Bio-Rad, Sydney, Australia). The temperature cycle was as follows: 30 min at 50°C, 45 cycles at 95°C for 30 s, 55°C for 20 s and 68°C for 20 s (with 5-s increases of the extension step in each cycle after the 20th cycle). The final cycle had a 5-min extension at 68°C. A melt curve was also added after the PCR step to distinguish the product from primer dimers and eliminate the potential for false-positive detection. The melt curve was achieved using 80 cycles starting at 55°C for 10 s, with 0·5°C increase with each cycle to a final temperature of 95°C. The Tm for each amplicon was determined using the iCycler software (Bio-Rad).
Mean viral copy numbers were calculated from a standard curve generated during the RT-PCR using the iCycler software. The standard curve was constructed from 100-fold serial dilutions of poliovirus or coxsackievirus RNA using RNA extracted from the original washed suspensions. Four dilutions of the standard RNA were used to construct the standard curve starting with the initial extracted RNA solution being used as the first dilution (100). Aliquots of the same set of RNA standards were used for all experiments for comparative purposes.
A 1 log10 removal time (T90) τ was determined between each sampling occasion using the following equation.
where Ct is the final number at day t, C0 is the number at day 0 and t is the time interval.
The average T90 on each sampling occasion was determined from the replicates of each organism. The removal times for each of the conditions over the entire sampling period were then averaged to obtain a mean T90 and standard deviation for each micro-organism. Where the decay for one of the micro-organisms was clearly second order non-linear decay, two T90 values were recorded, one for the initial rapid removal and the other for the slower second stage of decay.
Statistical significance of the results for individual organism under each of the conditions was determined by comparing the difference between the regression slopes obtained for the log of each organism under different conditions (Fowler et al. 1998). The null hypothesis (H0) was that there was no significant difference between the regression slopes of the two conditions being compared. As a result of differences in detection methods, comparison of statistical significances was carried out only for individual, not different, micro-organisms within different treatments.
Effect of temperature on decay
The results show that neither 15 nor 28°C had much influence on the decay of E. coli or the viruses in the absence of indigenous micro-organisms (filtered groundwater samples) with very slow or no decay observed at either temperature. A slow continuous decay was observed in the absence of indigenous groundwater micro-organisms, except for E. coli where a sharp increase in the T90 time was observed after day 21 at both temperatures. The T90 times of all four micro-organisms in the absence of groundwater micro-organisms were slower at 15°C than 28°C (Table 1). However, this difference was not statistically significant. When indigenous groundwater micro-organisms were present in the groundwater samples, much faster T90 times were observed for all four micro-organisms.
Table 1. Mean removal times for viruses and Escherichia coli at different temperatures
|Non-filtered||1·4||1·0||5·0; 10·0||10·5; 165·8|
For E. coli, the observed decay was much greater in the presence, than in the absence, of groundwater micro-organisms with T90 times in the filtered samples being >95 days up to day 21 while T90 in the unfiltered samples were <2 days. The difference in the T90 time for E. coli between the two temperatures in the presence or absence of groundwater micro-organisms was not statistically significant.
MS2 had a similar behaviour to E. coli in that there was a much greater removal time in the presence of groundwater micro-organisms than in their absence (Fig. 1). T90 times >180 days in the absence of groundwater micro-organisms and <3 days in their presence were observed (Table 1). Again, there was no statistically significant difference between the T90 times at 15 or 28°C when groundwater micro-organisms were present or absent. The only significant differences observed were between the presence and absence of the groundwater micro-organisms.
In the presence of groundwater micro-organisms the influence of temperature was more pronounced for the enteroviruses. Under both temperatures the decay of poliovirus was two staged with a rapid loss of virus over the first short stage and a much slower loss over the second stage (Fig. 1). The decay was faster for both stages at 28°C with the first rapid stage of decay finished within 2 days at 28°C compared with 7 days at 15°C, and the detection limits reached by 21 days at 28°C and 35 days at 15°C. This difference in the decay of poliovirus in the presence of groundwater micro-organisms at 28 and 15°C was found to be statistically significant. In the absence of groundwater micro-organisms no significant difference was detected between the two temperatures.
In contrast, there no significant difference was detected between coxsackievirus incubated in the presence of groundwater micro-organisms at either 28 or 15°C with a T90 time of 10·2 days at 28°C and 10·5 days changing to 165·8 days after day 21 at 15°C (Table 1). While this difference was not significant the t-ratio obtained was approaching significance at P = 0·05, indicating that there was some significance in the results. Similar to poliovirus there was no difference in the T90 time in the absence of groundwater micro-organisms and a statistically significant difference was detected only between the samples with and without groundwater micro-organisms.
Effect of oxygen on decay
In line with the influence of temperature, the only observed significant decay of the introduced viruses and E. coli in the presence and absence of oxygen occurred when groundwater micro-organisms were also present (Fig. 2). However, the presence of oxygen in the groundwater containing micro-organisms had a greater influence on the decay of each of the four micro-organisms compared with the removal times observed in the absence of oxygen (Table 2).
Table 2. Mean removal times for viruses and Escherichia coli in the presence and absence of oxygen
|Non-filtered||1·1||2·7||1·0; 29·0|| 10·2|
Escherichia coli had similar T90 times in the presence and absence of oxygen, although the removal time was significantly faster in the presence of oxygen. As observed for E. coli in the temperature experiments, there was little decay in the absence of groundwater micro-organisms with little difference between the observed T90 times in the presence or absence of oxygen (Table 2).
MS2 also had a larger difference in decay between the non-filtered aerobic and anoxic groundwater samples with T90 times of 2·7 and 8·2 days, respectively. The only non-significant difference observed between T90 times was between the aerobic and anoxic groundwater samples where the groundwater micro-organisms had been removed.
Both poliovirus and coxsackievirus had highT90 times under anoxic conditions in the presence of groundwater micro-organisms compared with the T90 times obtained in the unfiltered groundwater in the presence of oxygen. The T90 times in the unfiltered groundwater under anoxic conditions for both of these viruses were not statistically different from any of the filtered samples that had the groundwater micro-organisms removed.
Influence of nutrient concentrations
When nutrients were added to groundwater, as in the temperature and oxygen experiments, a significant decay was observed only when the groundwater micro-organisms were present (Fig. 3), where the addition of 0·01% glucose or 0·1% peptone groundwater significantly decreased the decay of E. coli, poliovirus and coxsackievirus when compared with the non-amended groundwater (Fig. 3). The exception was coxsackievirus in glucose-amended groundwater where the T90 time (4·4 days) was only marginally slower than the non-amended groundwater (3·1 days) (Table 3).
Table 3. Mean removal times for viruses and Escherichia coli with increased nutrients
|Non-filtered|| 0·7|| 4·2||3·1|
|Glucose||Filtered|| 20·0; 3·3||ND||103·6|
|Non-filtered|| 15·9|| 9·5||4·4|
|Non-filtered||ND; 6·0|| 13·0||25·0|
When groundwater micro-organisms were present, E. coli decayed the slowest in peptone-amended groundwater with an initial increase in cell numbers for the first 7 days followed by a gradual decay over the remaining time of the experiment. The T90 time for E. coli in glucose-amended groundwater was significantly faster than for peptone-amended groundwater but the fastest T90 time occurred when the groundwater remained unamended.
No statistical significance was found between the T90 times for poliovirus in non-filtered groundwater amended with either glucose or peptone. Both these T90 times were significantly slower than decay in the non-filtered non-amended groundwater. In the presence, rather than the absence, of groundwater micro-organisms all T90 times were much faster (Table 3).
There was no significant difference between the decay of coxsackievirus in the filtered groundwater samples and the non-filtered groundwater amended with 0·1% peptone until day 28, after which coxsackievirus numbers began to decrease (Fig. 3). The T90 time for coxsackievirus in the non-filtered, peptone-amended groundwater was significantly less than in the glucose-amended groundwater or in the non-amended groundwater. As with poliovirus and E. coli the greatest T90 time was observed in unamended groundwater containing indigenous micro-organisms. However, no statistically significant difference was observed between this decay rate and that for groundwater amended with 0·01% glucose.
This study involved determining the dominant processes influencing the survival of enterovirus pathogens and indicator micro-organisms in groundwater under different conditions. Poliovirus and coxsackievirus were used as known human pathogens that can be transmitted via water, while MS2 and E. coli were included in the study as these micro-organisms are often used as indicator organisms for the presence of faecal contamination or as a model virus particularly for the enterovirus group (Yates et al. 1985). While viral pathogens decay in groundwater (Gerba 1999), the dominant processes involved in inactivation are not well understood.
Understanding the dominant processes involved in the removal of enteric pathogens in aquifers is fundamental for the management of artificial recharge systems, particularly those that use non-potable water sources. The injection of water into an aquifer during artificial recharge can cause a range of perturbations including changes in temperature, or changes in oxygen and nutrient availability. These changes have the potential to influence the survival potential of microbial pathogens introduced with the recharged water. Toze and Hanna (2002) demonstrated that pathogens do decay in aquifers during artificial recharge. If this decay process can be more clearly understood then issues such as management, residence or storage times, and pre- or post-treatment requirements can be predicted and possibly modelled.
In this study, changes in MS2 and E. coli numbers were determined by culture techniques while poliovirus and coxsackievirus numbers were determined using real-time quantitative RT-PCR. As RT-PCR is not a direct measure of loss of infectivity, instead being a measure of the loss of viral RNA, it almost underestimates the removal times of the enteroviruses. As such, the use of different detection methods prevented direct comparison between the results of E. coli, the bacteriophage MS2, and the enteroviruses poliovirus and coxsackievirus. Instead, only the influence of different conditions was assessed for each individual microorganism.
Temperature has been rated in the literature as one of the most significant factors affecting pathogen inactivation (Melnick and Gerba 1980; Yates et al. 1990). While the exact mechanism behind this inactivation is uncertain, increased temperature could cause thermal degradation of the viral capsid (Yates et al. 1985). In addition, temperature has an indirect influence on virus inactivation, by controlling the activity of other inactivation mechanisms (Kapuscinski and Mitchell 1980). In this study, 15 and 28°C were used to represent temperatures commonly found in aquifers, and are consistent with temperature ranges tested in other studies (Yates et al. 1990; Toze and Hanna 2002).
The results in this study show an indirect influence of temperature on the decay of the viruses and E. coli. This is demonstrated by the significant differences observed between the removal times of the samples containing or lacking indigenous groundwater micro-organisms with the faster removal times occurring in the presence of the groundwater micro-organisms. Lower temperatures (4 and 12°C) decrease decay rates of viruses compared with the decay of the same viruses in groundwater at 23°C (Yates et al. 1985). These lower temperatures were not tested in this study, as they do not represent temperatures found in Australian groundwater.
It is possible that at higher temperatures the indigenous groundwater micro-organisms are more metabolically active, thereby degrading the pathogens at a faster rate. Jansons et al. (1989) observed a difference in the decay of poliovirus in groundwater bores in the same aquifer that had groundwater at different ambient temperatures.
Similar to the results obtained for the temperature experiments, the presence of oxygen appeared to have had an indirect influence on pathogen decay. Previous studies on the influence of oxygen on the survival of pathogens in groundwater have been limited. The findings of the few studies that considered the impact of oxygen observed that an increase in dissolved oxygen increased virus inactivation (Keswick and Gerba 1980; Jansons et al. 1989). A possible explanation for this observation may be that the increased oxygen levels have a direct influence on inactivation by increasing the oxidation of the viral capsid (Jansons et al. 1989). An alternative suggestion is that oxygen may operate indirectly by increasing the activity of indigenous micro-organisms in the groundwater, which would subsequently impact on virus inactivation rates (Yates and Yates 1988).
The results in this study show a definite relationship between the decay of E. coli and the viruses, the presence of oxygen, and the presence of indigenous groundwater micro-organisms. The general pattern observed was that, in the presence of oxygen and the indigenous groundwater micro-organisms, the T90 times of the enteroviruses and indicator micro-organisms were significantly increased. These observations support the conclusions made for the influence of temperature on T90 times – that the presence of oxygen influenced the activity of the indigenous micro-organisms, which in turn, increased the removal times of the viruses and E. coli.
Organic carbon is another factor that has been implicated in influencing the decay of pathogens in groundwater (Sobsey et al. 1995). Organic carbon can enter groundwater as a result of pollution, natural or intentional recharge. In particular, the intentional recharge of treated effluents could result in organic carbon being introduced into groundwater along with microbial pathogens. Any additional nutrients entering groundwater would increase the metabolic activity and change the population dynamics of the indigenous groundwater micro-organisms. Based on the observations that temperature and oxygen influence the decay of the viruses and E. coli by increasing the activity of the indigenous groundwater micro-organisms, an increase in nutrients would also increase the decay of the tested micro-organisms by increasing the metabolic rate of the groundwater micro-organisms. In this study peptone was used as a protein-based nutrient while glucose was selected as a simple carbohydrate-based nutrient source.
Increasing the activity of the indigenous micro-organisms through the addition of extra nutrients into an aquatic system could increase the inactivation of some viruses (Katzenelson 1978). Unlike the predicted effect, however, the results showed that the addition of peptone and glucose decreased the T90 times of the viruses and E. coli when compared with the unamended groundwater. This is in contrast to the observed increase in decay of the test organisms when the indigenous groundwater micro-organisms were exposed to oxygen and increased temperatures. The results in this study did not indicate the mechanism by which the additional nutrients decreased the decay rate of enteroviruses. However, it has been proposed that nutrients may surround viral particles thus directly protecting them from inactivation (Katzenelson 1978). This may be due to peptone and glucose protecting the introduced micro-organisms from attack from extracellular enzymes until sufficient nutrients are broken down to allow degradation of the introduced micro-organisms. This could have occurred by either the added nutrients acting as an alternative target for the groundwater micro-organisms or by surrounding the micro-organisms preventing direct attack by the enzymes. Further research is required to elucidate the exact mechanism by which the nutrients decrease the rate of decay of the introduced micro-organisms.
The addition of glucose to the groundwater had less of a retardation effect on the decay of the viruses than was observed in peptone amended groundwater. This suggests that the mechanism causing the decay of the viruses may be different in peptone- and glucose-amended groundwater. Based on the observations that the only significant decay of the viruses occurred in the presence of indigenous groundwater micro-organisms, it is more likely that peptone and glucose exert different influences on the groundwater micro-organism activity and population dynamics. This is being investigated further.
The T90 times observed in the decay of E. coli was similar for both glucose- and peptone-amended groundwater except that there was a 7-day lag period in the presence of peptone prior to the commencement of decay. This lag in decay rates suggests that E. coli was responding to the presence of peptone in a manner that did not occur in the glucose-amended groundwater. As suggested for the viruses, peptone may have been surrounding the E. coli cells initially protecting them from attack, until peptone was consumed by the groundwater micro-organisms to a level where there was no more protective effect. Alternatively, the E. coli cells may have been continually degraded from day 1, but were able to replicate in the groundwater up to day 7 when there was insufficient peptone remaining to overcome the degradative effect. The fact that degradation of the E. coli cells commenced from the commencement of the experiments suggests that, unlike peptone, the presence of glucose was either unable to provide this protection to the E. coli cells or could not support sufficient cell growth to counter the inactivation effect.
The outcomes of this study were that the presence of indigenous groundwater micro-organisms had the major influence on the inactivation of enteroviruses and indicator micro-organisms. While the decay of the introduced micro-organisms, in the absence of the groundwater micro-organisms, was generally greater at higher temperature or the presence of oxygen, these effects were secondary to the presence of indigenous groundwater micro-organisms. An additional conclusion is that the pathogen type can also have an impact on the rate of decay. This can be observed in the difference of decay of poliovirus and coxsackievirus, notably under anoxic conditions and nutrient amendment to the groundwater. The influence of pathogen type, while not directly studied in this work, has also been observed in other studies (Keswick et al. 1982; Jansons et al. 1989; Yates et al. 1990). The results indicate that the influence of pathogen type also needs to be examined more closely to better understand the decay processes involved in removing microbial pathogens from groundwater.
The results suggest that if the action of the indigenous groundwater micro-organisms on introduced micro-organisms such as enteric viruses and bacteria are better understood, then the presence of viral pathogens in recharged waters may be more easily predicted. This is already being done where the results of this and other studies have been incorporated into a predictive index (ASRRI) for predicting contaminant attenuation during aquifer storage and recovery (Miller et al. 2002). Further developments in understanding and predictive capability will aid in manipulating artificial recharge schemes to optimize the removal of contaminants such as microbial pathogens by influencing the activity of the indigenous groundwater micro-organisms.
This work has been reported with the permission of the American Water Works Association Research Foundation (AWWARF project No. 2618) and the Steering Committee of the Bolivar ASR Research Project. The project partners are CSIRO, Primary Industries Resources SA, United Water Intl. Pty. Ltd, South Australia Water Corporation, and SA Department of Administrative and Information Services.