Bacteriophage tracer experiments in groundwater

Journal of Applied Bacteriology 1988,65387-39s 2748/02/88

Bacteriophage tracer experiments in groundwater

HELEN SKILTON & D. WHEELER* Microbiology Department, University of Surrey and *Robens Institute, University of Surrey, Guildford, Surrey GU2 5 X H , U K

Received 29 February 1988 revised 6 July 1988 and accepted 9 July 1988

SKILTON, H.E . & WHEELER, D. 1988. Bacteriophage tracer experiments in groundwater. Journal of Applied Bacteriology 65,387-395.

Three tracer experiments employing three different bacteriophage were performed at one groundwater site near Beverley, Humberside. In two of the experiments the bacteriophage were injected into the aquifer by a borehole at a distance of 366 m from the pumping borehole. In the other experiment they were injected at a distance of 122 m. Regular samples were taken of water abstracted at the pumping boreholes as well as from the injection boreholes. The objectives were to: (1) investigate the pattern of bacteriophage recovery from the aquifer; (2) calculate the total number of bacteriophage recovered and the rate of their migration; and (3) detect any differences in bacteriophage behaviour which could be directly related to the morphology of the three bacteriophage. In all experiments the pattern of recovery was similar, exhibiting a peak of high numbers reaching the pumping borehole soon after injection. The highest percentage of original inoculum recovered was 1.9%. In the majority of cases, however, recovery was usually one log,, lower than this. The fastest migration rates were very rapid, reaching 2.8 cm/s in one experiment. No variation in percentage recovery or transit time could be directly attributed to morphology of bacteriophage. The most important factor governing the pattern of migration was undoubtedly the hydrogeological conditions.

The purpose of using tracers in groundwater is to investigate the direction and velocity of the groundwater or the behaviour of any potential contaminant within it.

The principal materials used as tracers of groundwater movement within aquifers have been salts, radio-isotopes and dyes. Radio- isotopes have not been widely used because of obvious public health and environmental con- cerns. The most commonly used radioactive tracer in hydrogeology has been 'H (Davis et al. 1980) because it forms part of the water mol- ecule and travels with the groundwater. If it were not hazardous it would be an ideal tracer. Nevertheless, radioactive agents have still been used for tracing non-potable water, for short- term tests of water movement within the wells themselves, and for carefully controlled labor-

* Correspondence to: H.E. Skilton, Division of Bio- technology, PHLS, CAMR, Porton, Salisbury SP4 OJG.

atory experiments. The use of salts (primarily chloride and bromide) as tracers has been limited by techniques for measuring the amounts of salt in water samples and differenti- ating the introduced salts from those already in the system. Organic dyes, such as sodium fluorescein, rhodamine WT, lissamine FF and amino G acid, are sorbed rather easily on solid materials, have low to high toxicity, and have been found to suffer larger losses in groundwater when compared with bacteria (Pyle 1979; Rahe et al. 1979).

Micro-organisms have also been used in groundwater mainly as models of contaminant transport, in particular, as models of faecal bacteria migration. Faecal bacteria are especially important, as their presence is universally recog- nized as an indicator of risk of faecal-oral infec- tion and their absence is often used as a key requirement for potable water. Some microorganisms, however, possess properties which may allow them to be tracers of groundwater

388 Helen Skilton and D . Wheeler itself as well as models for important microbial contaminants.

The three main classes of micro-organisms used as tracers in groundwater are yeasts, bacteria. and bacteriophage. Their use has been extensively reviewed by Keswick et al. (1982).

Yeasts are the least sensitive tracers because they are relatively large, their volume being 1000-times greater than that of a representative bacterium. Thus fewer organisms/ml of tracer inoculum can be prepared in the laboratory. They may also be naturally present in high numbers in polluted waters making it difficult to distinguish the tracer organism. They have been used in one sand and gravel aquifer site by Wood & Ehrlich (1978), where they were observed to penetrate more than 7m in less than 48 h.

Bacteria are more frequently used as tracers in groundwater studies. The main disadvantage of bacteria as tracers or model systems is the difficulty in selecting an organism which can be easily distinguished from the many different types present in highly contaminated waters. The types of selective characteristics most commonly used are antibiotic resistance, pigmentation and unusual biotypes for vegetative bacteria, and pigmentation and elevated tem- perature for bacterial spores. This requirement necessarily restricts the number and type of bacteria which can be used. Furthermore, the wisdom of using antibiotic-resistant organisms in groundwaters and the possibility of trans- ferring resistance to other naturally occurring bacteria cannot be discounted. Nevertheless, bacteria have often been used successfully as tracers in groundwater. In one case, Bacillus stearothermophilus and an H,S-producing strains of Escherichia coli, were observed to travel over a distance of 920m through a gravel aquifer at a rate of 200m/d (Sinton 1980).

In evaluating the use of bacteria it should be noted that: (1) some bacteria are potential pathogens; (2) some are capable of multiplication within the environment and so may give false results; and (3) they are large enough to be filtered by some aquifer material and so are only representative of bacteria and not necessarily the groundwater itself or of viruses.

In contrast, bacteriophages have never been known to cause disease in humans, animals or plants. They are very easily distinguishable from each other because of their specificity for bac-

terial hosts and therefore several may be used at the same time without necessitating the selec- tion of a specific property. They are unlikely to be present in a normal groundwater systems unless it is highly polluted with domestic sewage, and they are probably incapable of multiplication within these environments and thus will not suffer from background inter- ference. They are also very small, which means they are less susceptible to removal by the aquifer constituents. Their small size and physical composition permit them to be a potential model of human enteric viral behaviour. According to Wimpenny (1977), microbial tracers are not soluble in water, but because of their small size they will not separate from the liquid phase. Their buoyant densities are quite close to that of water and so in addition to pro- viding a model of contaminant transport, they can be considered in solution for the purposes of water tracing. Finally, high titres of the bacteriophage can be prepared inexpensively and in very small volumes. This, together with the sensitivity of detection (one plaque forming unit may be detected per 1OOOml of sample (Purdy et al. 1984)) allows bacteriophage to undergo a high degree of dilution (theoretically up to 17 log,, units) and still be detected. Bacteriophage are by far the most sensitive microbial tracers and possibly more sensitive than any of the chemical tracers including the radio-isotopes.

Previous experiments using bacteriophage at groundwater sites have been very successful either as models of contaminant transport or of the groundwater itself. For example Martin & Thomas (1974) performed a tracer experiment at a spoil heap to investigate the origin of water which was flooding one of the spoil tips, in order to take preventative action to avoid tip collapse. An Aerobacter aerogenes phage was used and successfully sustained original deduc- tions about the downslope movement of groundwater beneath the tip. Noonan & McNabb (1979) used two different bacteriophage to trace the possible migration of viruses from beneath a land application site. The coliphage T4 travelled through 920m of a gravel aquifer within 96 h. Finally, the York- shire Water Authority (Watkins, J. personal communication) used a Serratia marcescens bacteriophage to investigate the source of a water supply pollutant and was able to help identify the most likely source, a leaky sewer.

Bacteriophage in groundwater studies 389 TOP view

Pumping borehole no 4

Pumping borehole no. 2 Observation borehole 4 e Observation borehole 3e

0 0 4 122 m

Side view

Pumping boreholes numbers 2 and 4

In these studies, bacteriophage were used primarily to investigate aspects of groundwater systems rather than the behaviour of the phage within these systems.

The object of this investigation, however, was to investigate the behaviour of three bacteriophage at one groundwater site and to deter- mine such features as: (1) the pattern of recovery from the groundwater site; (2) the percentage of original inoculum of bacteriophage recovered ; (3) the rate of bacteriophage migration; and (4) to provide a tentative exploration of the location of the displaced phage within the aquifer. Furthermore, we were concerned to observe any differences in behaviour exhibited by three phage each belonging to a different morphological group.

The investigation was done at Etton pumping station which is near Beverley, Humberside. The source is an unconfined chalk aquifer. It is a highly uniform and exceedingly fine-grained sequence of pure white limestones with moder- ate porosity (14-20%) yet extremely low inter-

granular permeability (< m/d (Foster & Milton 1974). It is therefore essentially a fissure- flow aquifer, i.e. transmission of the bulk of water is by the physical discontinuities of the rock mass (joints and fissures, solution open- ings, fractures and cavities).

The section of Etton pumping station involved in this investigation consisted (Fig. 1) of a pumping station in command of pumping boreholes 2 and 4, and observation boreholes 3e and 4e. Both pumping boreholes are 90 cm in diameter and 70m deep. They are 17m apart. Observation borehole 3e is lOcm in diameter, 70m deep and 122m ENE from pump 2. Obser- vation borehole 4e is 10cm in diameter, 76m deep and 366 m ENE from pump 2.

During the experiments only one of the two pumps was in operation. The average abstraction rate was 4.2 x 106ml/min. During all three experiments the abstracted water was pumped normally into the supply system and this requirement determined the pumping regimes. The abstraction rate was not necessar-

Helen Skilton and D . Wheeler ily constant throughout each experiment and at times there was no pumping. The bacterio- logical quality was consistently good and the pH remained around 7.5.

Materials and Metbuds

B A C T E R I O P H A G E

Three bacteriophage were used. marcescens bacteriophage; (2) cloacae bacteriophage; and (3) K coli bacteriophage.

(1) Serratia Enterobacter

2 Escherichia

The S. marcescens phage was isolated from seawater by J. Watkins of Yorkshire Water Authority (host strain NCIB 10644). It has a symmetrical polyhedral head, approximately 50 nm in diameter, a very short tail and belongs to morphology group I11 (Tikhonenko 1970).

The En[ . cloacae phage was isolated from Holdenhurst Sewage Works by D. Wheeler (the host was a wild strain isolated from seawater at Bournemouth). It has a large polyhedral head, approximately IOOnm x 70nm, an 80nm tail, a contractile sheath and belongs to morphology group V.

The K12 E. coli phage is a wild strain phage (host strain NTCC 9481 (W1485)). It has a symmetrical polyhedral head, about 50nm in diameter, with a long thin non-contractile tail of 150nm. It is in morphology group IV.

Tracers were prepared and donated by the Yorkshire Water Authority.

Technicai notes

All water samples were assayed with the soft agar overlay technique of Adams (1959). Adap- tations were made to the method to allow as many samples to be examined in as short a time as possible.

The bacterial host was grown overnight at 37°C in rich nutrient broth containing (g/l distilled water): Brain Heart Infusion (Oxoid), 20.0; casein hydrolysate acid (Oxoid), 20.0; KH2P0,, 5.0; MgSO, .7H,O, 1.0; yeast extract (Difco), 1.0; and 20ml glycerol (Analar grade). The pH was adjusted to 7.2 before auto- claving. A 0.1 or 1 ml amount or dilution of the sample was added to 0.1 ml of the bacterial host on a plate of Blood Agar Base (Oxoid). A 3ml volume of molten soft agar (kept at 44°C) containing (g/l distilled water): purified agar (Oxoid), 5 . 5 ; Nutrient Broth (Oxoid), 11.2;

NaCl (M&B Pronalys A.R.), 7.0; was pipetted on to the mixture. The plate was rotated vigor- ously to allow mixing and was left with the lid partly open until the agar had set. The lid was replaced and the plate inverted and incubated overnight at 37°C.

The number of areas of lysis or plaque- forming units (pfu) were counted.

Method of sampling

At Etton pumping station there is a sample tap for each pumping borehole which allows a small volume of the water pumped from the aquifer to be collected for examination. By allowing the tap to run into a container of known volume with an outlet at the top, a surveyor sampler (Warren Jones Engineering, Bicester) could be connected to the container and automatically set to take samples at chosen intervals.

Water samples from the observation boreholes were taken using Flow-Through Depth Samplers. All samples were stored in sterile 25 ml screw-capped bottles at 4°C before examination.

Analysis of results

An estimate of the total numbers of phage migrating through the chalk aquifer as well as the rate of migration and pattern of recovery could be made by calculating the number of pfu/ml and the overall volume of water extracted by the pumping station.

The total number of phage remaining in the boreholes could also be made by calculating the volume of water within the observation boreholes and relating this to the number of pfu/ml at different depths.

SITE D E S C R I P T I O N S A N D E X P E R I M E N T A L P R O C E D U R E

Experiment 1 (6.1 2.84-10.12.84)

The three phage were injected into observation borehole 3e (122m from the abstraction point) while pump 4 was in operation. The tracer was poured through a funnel into the borehole, followed by approximately 2 1 of groundwater. Water samples were taken hourly for 120 h from the sample tap for pump 4.

Bacteriophage in groundwater studies 391 before injection and at regular intervals there- Experiment 2 (18.12.8425.12.84)

1.0-

0.01

after. Samples were also taken at four different depths from within borehole 3e (intermediate

to investigate further the spread of the contaminant plume. Samples were taken from the pumping station hourly for 45 h.

The three phage were injected into observation

while pump 4 was in operation. The method of injection and water sampling were as above. Samples were taken hourly for 168 h.

borehole 4e (366m from the abstraction point) between borehole & and the pumping station)

-

Experiment 3 (26.1 1.85-28.11.85) Results

The three phage were injected into observation borehole 4e (366m from the abstraction point) while pump 2 was in operation. They were injected through a 40m length of a hose pipe (25 mm diameter) which allowed the phage to be inoculated directly into the borehole at a chosen depth. Again this was followed by 21 of groundwater.

During this investigation, samples were taken from within borehole 4e at four different depths

E X P E R I M E N T 1

All three phage showed similar recovery patterns (Fig. 2) producing a peak of high numbers decreasing to a tail of low numbers.

The rate of migration of the phage across the 122m to the pumping station was the same for all three tracers. The first positive sample was detected within 2.5 h, giving a fastest migration rate of 1.35 cm/s (Table 1).

Table 1. Fastest migration rates observed for the bacteriophage

Exp. no. I Exp. no. 2 Exp. no. 3 (122m) (366m) (366 m)

Bacteriophage of m/s cm/s 4 s Smratia marcescens 1.35 2.0 2.8 Escherichia coli 1.35 1.85 2.8 Enterobactm cloacae 1.35 1.7 2.8

Fig. 2. Recovery patterns of the three bacteriophage inoculated into borehole 3e, 122m from pumping borehole no. 4. ~ , Serratia marcescens bacteriophage; 0, Enterobacter cloacae bacteriophage; 0, K12 coliphage.

392 Helen Skilton and D . Wheeler Table 2. Percentage of original inoculum of bacteriophage recov-

ered within 45 h.

Exp. no. 1 Exp. no. 2 Exp. no. 3 (122m) (366 m) (366 m)

Bacteriophage of Yo Yo %

Smratia marcescens 0.49 0.23 0.10 Eschmichia coli 0.48 1.90 0.17 Enterobacter cloacae 0.12 0.34 0.18

The percentage of original inoculum recovered from the aquifer for the S . marcescens phage was 0.49 representing 1.99 x 10” pfu. A similar percentage was found for the K12 coliphage (0.48) but only 0.12% of the Ent. cloacae phage was recovered (Table 2.).

E X P E R I M E N T 2

A peak of higher numbers was observed as in the first experiment but detectable numbers per- sisted for a longer time, producing a wider spread, even after considering the influence of the pumping regimes which, as in all the experiments, elongated the spread of significant numbers. Two of the tracers, the S. marcescens and Ent. cloacae

W

D D

C aJ c

Y a

phages, exhibited similar

recovery patterns (Fig. 3). The K12 coliphage behaved slightly differently producing several peaks of recovered phage within a wide spread of significant numbers.

A different migration rate was observed for the three phage. A slightly longer time passed before the first positive sample was detected (between 5 and 6 h after addition of the tracers) but this constitutes a faster migrational rate than in the first experiment: 1.7 to 2.Ocm/s as compared with 1.35 cm/s (Table 1).

The percentage of original inoculum recovered for each phage differed (Table 2). Only 0.23% of the S . marcescens phage was recovered (less than half of that recovered in the previous experiment); 0.34% of the Ent. cloacae phage was recovered (more than twice as much as that detected in the first experiment); and finally,

Pump off

1

0 O O O I I 0 10

I I I 20 30

Time after inoculation ( h )

Fig. 3. Recovery patterns of the three bacteriophage inoculated into borehole 4, 366m from pumping borehole no.4. ___ , Serratia marcescens bacteriophage; 0, Enterobacter cloacae bacteriophage; 0, K12 coliphage.

Bacteriophage in groundwater studies 393 1.9% of the K12 coliphage was recovered, representing the highest percentage recovered for any of the phage in these experiments.

EXPERIMENT 3

The recovery patterns for all three phage showed greater similarity to each other in this experiment than in the previous two (Fig. 4), producing a short peak of high numbers similar to that observed in the first experiment. The pumping regime had again influenced the behaviour pattern.

The migration rate was the same for all three phage and was faster than in the previous experiments, i.e. 2.8 cm/s (Table 1).

The percentage of phage recovered from the aquifer was small in comparison with experiments 1 and 2 but again some variation was observed between the three (Table 2): 0.1% of the S. marcescens phage was detected compared with 0.18% of the Ent. cloacae phage and 0.17% of the K12 coliphage.

OBSERVATION BOREHOLES

The number of phage remaining within the injection observation borehole 4e was assayed

N

'0 x I

-0 E

E > V

- $ 3 V

c

aJ 5- 0 L Q

.-

+-

a Tn 0

a +

E a

1.

0.c

0.000

Time after inoculation ( h )

Fig. 4. Recovery patterns of the three bacteriophage inoculated into borehole 4e, 366m from pumping borehole no. 2. ~ , Smatia marcescens bacteriophage; 0, Enterobacter cloacae bacteriophage; 0, K12 coliphage.

' r

- 9 9 0 10 20 30 40 50 60

Time (h )

Fig. 5. Etton experiment 3. Smatia marcescens bacteriophage remaining within injection borehole 4e, after injection of bacteriophage. x , 25m; 0, 40m; A, 55 m; 0 , 6 9 m.

during experiment number 3. The behaviour of the phage was similar for all three and at all depths (Fig. 5 shows only S. marcescens phage). The numbers decreased substantially at all levels in the borehole within 4-5 h after adding the tracers, although there appears to be a slight tendency for the numbers to persist longer at the lower depths. The density remaining after 4.5 h within the borehole, however, represented an extremely small fraction of the original, in all cases 4 to 5 log,, units lower in titre.

Samples taken from the intermediate borehole, 3e, contained a few phage throughout the duration of the experiment, but gave no significant peak. This suggests that the bulk of the phage plume did not take a route to the pumping station which traversed that particular observation borehole.

The borehole samples taken during experiment 3 (26.11.85) from borehole 4e and 3e both contained low numbers (Figs 6 & 7) of pfu before addition of the tracer. Tracer experiments 1 and 2 were performed almost 1 year earlier (6.12.84 and 18.12.84 respectively). This shows the stability of these phage, which survived under these chalk groundwater conditions for at least 12 months. It also shows the possible effects of adsorptionfdesorption phenomenom on the phage by the chalk, plus irregular dis- persal of water within the hole, which allowed the retention of small numbers of phage within this area even while pumping was in operation throughout the year.

Helen Skilton and D . Wheeler 60 x-

I \ 50 t \

I \

’ \ 40 \

3epm ‘m i

Fig. 6. Etton experiment 3. Bacteriophage remaining within borehole 3e, before injection of bacteriophage. 0, Serratia marcescens bacteriophage; A, Entero- bacter cloacae bacteriophage; x , K12 coliphage.

25 40 55 69 DFDttl (m)

Fig. 7. Etton experiment 3. Bacteriophage remaining within borehole 4e, before injection of bacteriophage. 0, Serratia marcescens bacteriophage; A, Entero- bacter cloacae bacteriophage; x , K12 coliphage.

Discussion

It is known that the chalk aquifer at this site depends primarily on fissure-flow, and the extremely short migration times observed by the fastest migrating bacteriophage in all experiments sustain this fact.

The recovery patterns observed for the bacteriophage was similar in all experiments. A peak of high numbers of bacteriophage reached the pumping station soon after injection followed by a tail of lower but detectable numbers. The asymmetrical trace produced suggests that some

factor other than dispersion is delaying the bacteriophage in its migrational route to the well. Simple dispersion alone would produce a symmetrical shape with lower numbers of bacteriophage at either side of the peak. This tail is highly significant and suggests the existence of diffusion and/or adsorption.

Once bacteriophage reaches the groundwater it will be drawn to the fastest part of the moving water, i.e. the centre of one or more fissures, and thereafter will travel to the pumping station very quickly, having little contact with the rock matrix. Phage that are slower in reaching their destination will have been affected by dispersion by currents within the fissures as well as diffusion across the chalk matrix to the relatively immobile pore water. This will delay the average groundwater velocity of the bacteriophage as well as expose it to a large area of rock mass where adsorption may come into

In experiments 1 and 2 (Figs 2 & 3) the same pumping borehole was in operation but the site of injection was different. A faster, migrational rate was observed for the second experiment even though the bacteriophage had more than twice as far to travel (Table 1). There was a longer spread of significant numbers observed in experiment 2, however, implying the influence of a greater number of interactions affecting the tracers over the distance. The percentage of original inoculum in these experiments varied from 1.9 to 0.12% (Table 2). Much of this variation may be caused by the method of injection rather than any inherent differences caused by variations in distance or the morphological type of the bacteriophage. Experiment 3 emphasizes the importance of the method of injection (Fig. 4). A close resemblance in behaviour (per cent of original inoculum, fastest migrational rate and shape of recovery) was demonstrated by the three bacteriophage in this experiment. This also suggests that this method of injection (flushing down a pipe of predetermined length) may have allowed the bacteriophage to gain rapid access to the main stream.

The lower percentages of bacteriophage recovered in experiment 3 suggest that the different hydrogeological conditions to which these bacteriophage were subject (a different pumping borehole was in operation, producing different directional and velocity properties of the groundwater flow etc) did have a significant

play.

Bacteriophage in groundwater studies 395 effect, either by causing delay (hence reducing the number detected because of dilution) or removal.

Etton experiment 3 included extensive sampling of the injection borehole, in order to deter- mine actual input of bacteriophage into the aquifer (Fig. 5). As can be seen, although the majority of bacteriophage left the borehole rapidly within 30min (the first sample), it was not a simple one-point injection of the aquifer, but effectively a multiple step inoculation with a small fraction remaining and gradually reducing to very low numbers within 4-5h. This too, would extend the tail end of the recovery pattern.

Results from observation borehole 3e, the intermediate borehole, suggest that either: (1) the route of the bacteriophage was not direct from the injection site of the pumping borehole; or (2) the observation borehole does not ade- quately monitor groundwater flow due to local factors.

In none of the three experiments were differences detected in recovery pattern or percentages recovered which could be directly attributed to the different morphological groups of the bacteriophage.

It is a reflection of bacteriophage properties that such huge dilution of the original inoculum can occur, and yet significant quantities (up to 2%) may still be detected. Furthermore, the stability of bacteriophage within groundwater enables it to be detected after a considerable period, up to 12 months within observation boreholes as in experiment 3 (Figs 6 & 7).

These experiments reaffirm that bacteriophage are promising tracers at groundwater sites.

This work was funded by a grant from the Natural Environmental Research Council. Special thanks are due to Dr S Foster and Mr L. Bridge from the British Geological Survey for their assistance and to Mr J. Watkins and Mr. T. Crease of Yorkshire Water Authority for invaluable advice, supply of bacteriophage, and the permission to experiment at the Etton groundwater site. The manuscript is based on a

presentation at the Society’s 1986 Summer Con- ference on Tracers.

References ADAMS, M.H. 1959 Bacteriophages pp. 45M51 New

York: Interscience Publishers. DAVIS, S.N., THOMPSON, G.M., BENTLEY, H.W. and

STILES, G. 1980 Groundwater tracers- a short review. Groundwater 18,1423.

KESWICK, B.H., WANG, D.S. & GERBA, C.P. 1982 The use of microorganisms as groundwater tracers: A review. Groundwater 20,142-149.

FOSTER, S.S.D. & MILTON, B.A. 1974 The permeability and storage of an unconfined chalk aquifer. Hydro- logical Sciences-Bulletin-des Sciences Hydrologiques 4,475-500.

MARTIN, R. & THOMAS, A. 1974 An example of the use of bacteriophage as a groundwater tracer. Journal of Hydrology 23, 73-78.

NOONAN, M.J. & MCNABB, J.F. 1979 Contamination of Canterbury groundwater by viruses. In The Quality and Movement of Groundwater in Alluvial Aquifms of New Zealand. Technical Publication No. 2. ed. Noonan, M.J. pp. 195-201. Department of Agricultural Microbiology, Lincoln College. Can- terbury, New Zealand.

PURDY, R.N., Dancer, B.N., DAY, M.J. & STICKLER, D.J. 1984 A note on membrane filtration method for the concentration and enumeration of bacteriophage from water. Journal of Applied Bacte- riology 58,231-233.

PYLE, B.H. 1979 Bacterial movement-experience at Heretauraga. In The Quality and Movement of Groundwaters in Alluoial Aqugers of New Zealand. Technical Publication No. 2. ed: Noonan, M.J. pp. 105-1 15. Department of Agricultural Microbiology Lincoln College, Canterbury, New Zealand.

RAHF, T.M., HAGEWRN, C. & MACCOY, E.L. 1979 A comparison of fluorescein dye and antibiotic resistant Escherichia coli as indicators of pollution in groundwater. Water Air and Soil Pollution 11, 93-104.

SINTON, L.W. 1980 Investigations into the use of bacterial species Bacillus stearothermophilus and Escherichia coli as tracers of groundwater movement. Water and Soil Technical Series No. 17. pp. 12-22. National Water and Soil Conservation Organisation, Wellington, New Zealand.

TIKHONENKO, AS. 1970 Ultrastructure of Bacterial Viruses New York and London: Plenum Press.

WIMPENNY, J.W.T. 1977 Water tracing. In Treatment of Industrial Efluents ed. Callely, A.G. pp. 346373. London, Hodder & Stoughton.

WOOD, W.W. & EHRLICH, G.G. 1978 Use of bakers yeast to trace microbial movement in groundwater. Groundwater 16,398-403.

Documents

Bacteriophage tracer experiments in groundwater