This paper was presented at the International Workshop "Microorganisms - Water and Aquifers", organized by the Zuckerberg Institute For Water Research, J. Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, held on September 20-21, 2005 in the Sede Boqer Campus, Israel.

Introduction

The survival of non-soil bacteria, introduced to groundwater by anthropogenic activities, is of great importance to drinking water quality. Although groundwater scientists know that indicator bacteria and pathogens do survive to a certain extent in groundwater (Engelbrecht, 2004) the overwhelming majority of water managers, engineers and the general public believes otherwise. Groundwater has long been considered to be of excellent quality because of the soil barrier providing effective isolation of groundwater from surface pollutants. However, if microorganisms survive the passage through the vadose zone, they will contaminate groundwater and may emerge in springs and boreholes used as drinking water supplies (Powelson and Gerba, 1995).

The number and variety of the microorganisms in natural waters vary greatly in different places and under different conditions. Bacteria are washed into the water from the air, the soil and from almost every conceivable object. The faeces of animals contain vast numbers of bacteria and may enter natural water systems. The sizes of laboratory grown bacterial cells are in the order of one micrometer, but the smallest survival forms in soil may be only a few tenths of a micrometer in diameter (Ghiorse and Wilson, 1988). The size of openings in subsurface material can be assumed to be variable and are generally not measured, but porosity and permeability measurements on aquifer sediments indicate that adequate spaces for bacteria exist in many sediment types, even in some rather dense porous rocks (McNabb and Dunlap, 1975). The interstices of shallow aquifer sediments can easily accommodate bacteria and probably protozoa and fungi as well. Larger organisms will be excluded from most subsurface formations, except for gravelly and cavernous aquifers.

Aim

To demonstrate that the survival of non-soil bacteria occurs in real life situations at a specific site, in this case Atlantis, where enough background information is available to verify the results in a fully integrative project.

Methodology

During phase one, a laboratory study on the movement of Escherichia coli through sand columns was conducted. During phase two, groundwater samples were collected in an area where groundwater is artificially recharged using treated sewage (Tredoux and Cave, 2002).

During phase one, sterile water containing Escherichia coli was filtered through columns filled with soil of different grades. The filtrate was sampled and analysed for Escherichia coli, by membrane filtration using mFC agar and MUG supplement. Four 90 litre cylinders, with three 15 mm PVC taps connected, 250 mm apart from top to bottom, were used. Four water permeable membranes are placed in 15 L plastic buckets, which in turn are placed in an 80 L bucket. One cylinder each is then placed in each of the 15 L buckets on top of the four water permeable membranes and filled with sand of a specific characteristic.

Sterile tap water was used to simulate rainwater and was distributed onto the sand in the cylinders with a 15 L plastic bucket with 1300 1-mm holes in the bottom. The Escherichia coli were cultured in tryptone water. The bacterial count in the stock solution before dosage was 1.63 x 1014 cfu per 100 mL. The four cylinders were filled with sand with different grain sizes: No. 1 = 0.42 to 0.47 mm; No. 2 = 0.60 to 0.65 mm; No. 3 = 250 to 355 mm; and No. 4 = 250 to 275 mm. The total rainfall, 25.5 L, as calculated for a cylinder was ”rained” in succession onto each of the four cylinders as follows: first watering, 10 L of sterile water containing 10 mL inoculant; second watering, 8 L of sterile water; and third watering, 7.5 L of sterile water. The cylinders were again “rained” on with sterile water (without inoculant) on day, 14, 20, 21, 29, 34, 36, 42 and 43. The experiment was prematurely stopped due to unexpected fieldwork.

Phase 2 was conducted at a wellfield in Atlantis, where artificial recharge has been practised since 1982. The water used for infiltration, flows from a collection basin by pipeline into a settling basin. Out of this basin it flows over a flow-measuring device into an infiltration basin called "Pond 7". From this pond the water naturally enters the subsurface to finally reach the aquifer from where it is abstracted for treatment and use. A handheld auger was used to drill two holes to the water table. The augerholes, Auger 1 and Auger 2, are 5 m and 100 m from Pond 7, respectively. The water table below ground level is at 1.0 m in Auger 1 and at 1.5 m in Auger 2. The closest borehole, G33134, is 50 m away and "downstream" of Pond 7. This is an observation borehole and therefore not used for abstraction. The borehole is 33 metres deep and the water level below surface is 0.3 m. Moles have created "flowpaths" through the western weir of Pond 7, resulting in a semi-flooded area around borehole G33134 causing the high water table. Eight water samples were collected from the inflow, outflow, surface water, auger water and borehole water. The water was analysed for faecal coliform bacteria including Escherichia coli.

Results and discussion

The flow through the cylinders during phase one was observed and the following differences were noted.

In cylinder no.1, ponding and flow were equal; fingering occurred fast with the wetting front close-up and spreading through the whole sand column almost immediately. The water flows through the sand in the cylinder, passes through the four water permeable membranes, filling the 15 L bucket from the bottom overflowing into the 80 L bucket. The total outflow measured was 20.5 L. Therefore, 5 L water was contained in the sand column. Three samples were collected and analysed.

In cylinder no.2, ponding on top of the sand occurred and fingering was slow with no outflow. Due to the ponding, overflow over the top of the cylinder into the 15 L bucket happened, contaminating possible outflow. Therefore no samples were collected from this cylinder.

In cylinder no.3, ponding with very slow fingering occurred, with the wetting front only halfway through the sand column after the first watering. Outflow occurred only after the third watering. The total outflow measured was 13.5 L and thus 12 L water was contained in the sand column.

In cylinder no.4, ponding and very slow fingering occurred. The wetting front was only one quarter through the sand column after the first watering and slow outflow occurred 10 minutes after the third watering. The total outflow measured was 13.5 L with 12 L of water contained in the sand column.

In order to determine the die-off or survival of Escherichia coli in the sand columns, the cylinders were left alone for a period of 14 days. Watering and sampling schedules with results are set out in Table 1. The results were a bit confusing. It was assumed that with the fast flow through cylinder 1, the bacterial count in the outflow water would have been high. However, the opposite was found with low numbers in the fast flow cylinder and high numbers in the slow flow cylinders. Due to the contamination of the receiving bucket in cylinder No. 2, this cylinder was only used from Day 34. At that stage no Escherichia coli was found.

The Escherichia coli count in the first sampling is always lower than the count in the second sample. Comparing the 1st sample with the 2nd sample it appears that although the bacteria are spread through the sand column, higher numbers occur higher-up in the sand column.

From the results it is also evident that the counts are getting lower over time although there are higher counts in between at later days. Why the bacteria disappeared in the outflow water from cylinder No. 1 after 15 days is still mystery. They are either locked in a biomass in the cylinders, or washed out after sampling or most of them died during the race through the sand column.

Table 1: Escherichia coli counts (cfu/100 mL) in sampled water from cylinders on different days as indicated

The results from phase 2 are presented in Table 2.

Table 2: Microbiological analyses results (per 100 mL) of Atlantis samples

Escherichia coli were found in all the samples collected. With this limited information it is not possible to evaluate or to scrutinise the results in depth. However, one can hypothesise that Escherichia coli can survive in the environment and move through the soil-water matrix. The water used for infiltration is collected from stormwater and secondary treated domestic sewage. The Escherichia coli count in the inflow is rather low and it is assumed that the die-off occurs in the collection basins before it reaches Pond 7. It is also true in the case of the outflow sample. Flow between the inflow and outflow is controlled by the settling basin. Birds and other animal life in and around Pond 7 obviously add Escherichia coli to Pond 7. The difference in Escherichia coli numbers in Pond 7 is due to many factors. The dilution factor created by the volume of water in Pond 7 during the rainy season may be a main factor. From the augerhole results it is further assumed that there are more Escherichia coli in the saturated soil than in the water from these augered holes. At Augerhole 1, Escherichia coli numbers varied by 1-log when the water was allowed to clear. The Escherichia coli in the borehole may be because of the fact that the water is standing and the borehole is not pumped.

Conclusions

It is obvious from this two-phase study that there is a decline in Escherichia coli numbers over time. However, the Escherichia coli survived for more than 40 days in the laboratory study. Escherichia coli does occur in groundwater in the natural environment. It was also established that the Escherichia coli numbers in the natural soil is higher than in the subsurface water and Escherichia coli can move from surface water to groundwater.

References

1. Engelbrecht, J.F.P. (2004). The survival of Escherichia coli in groundwater. CSIR Report ENV-S 2004-008, Environmentek, Stellenbosch.
2. Ghiorse W.C. and Wilson J.T. (1988). Microbial ecology of the terrestrial sub-surface. Advances in Applied Microbiology, 33.
3. McNabb J.F. and Dunlap W.J. (1975). Subsurface biological activity in relation to groundwater pollution. Ground Water, 13(1).
4. Powelson D.K. and Gerba C.P. (1995). Fate and transport of micro-organisms in the vadose zone. In: Handbook of Vadose zone characterisation & monitoring. L.G. Wilson, L.G. Everett and S.J. Cullen (eds), Geraghty & Miller, Environmental Science and Engineering Series.
5. Tredoux,G and Cavé L.C. (2002). Atlantis Aquifer: A status report on 20 years of groundwater management at Atlantis. CSIR Report ENV-C-S 2002-069, Environmentek, Stellenbosch.
6. Website: http://www.atcc.org/ (April 2004)
   Contact Us for subscription Information