Microbiological Water Testing

The need for microbiological analysis in the field is a debated subject. There are those that argue that it an essential tool to assess the degree of faecal pollution, and there are others that say that the test results are not so meaningful, and that in any case, it is easy to say if a source is faecally polluted or not through other assessment methods such as sanitary surveys. However, the recommendation is that testing for E. coli or other indicators of pathogens is beneficial and should be carried out if possible. However, it is not essential if the water is going to be treated. A consideration will be cost of the equipment, as well as the availability of proper expertise in its use.

Indicator or direct testing

Tests on pathogen removal using sand filtration tend to be in one of two main categories.

  1. Tests for actual pathogenic organisms that are found in the filtrate. Raw water in these cases is usually seeded with the pathogens of interest and the tests are more complicated and expensive.
  2. Tests for indicator organisms that are usually associated with pathogenic organisms, but which may not be that harmful themselves. These tests are carried out more easily with less expensive equipment.

When testing in the field it is very unlikely that direct testing will be possible as there are too many pathogen types to be identified in the field in practice (e.g. – species of helminths, protozoa and bacteria). Indicator bacteria in comparison are relatively easy to culture in the field and are relatively good at indicating other faecal pollution.

Indicator organisms

Testing for indicator organisms is one of the most common types of field test. The presence of indicator bacteria (such as E. coli) is always linked to faecal pollution because they originate in the human gut. Since most water-related disease is also faecally related (with the exception of Guinea Worm), where you find these indicator bacteria, there is a likelihood that you will also have other pathogens. In the same way, where there are no bacteria the likelihood of other pathogens being present is reduced.

Acceptable levels of indicator bacteria

Acceptable levels of indicator bacteria vary, depending on whether the situation is an emergency, long term or short term. For short periods of time, up to 1000 coliforms/100 ml can be tolerated, but for longer term measures, the maximum number of coliforms per 100 ml should not exceed 10 (House and Reed (1997 [ref.01]Ref.01: House, S.; Reed, R. (1997). Emergency Water Sources. WEDC, Loughborough, UK, p.171. Available online.). This variance may seem odd, but it reflects something called the infective dose. Pathogens (disease-causing organisms) have to be ingested by humans in a sufficient quantity to cause illness – this is called the infective dose. For some pathogens, the infective dose is high (meaning you have to ingest a lot to become ill), while for others it is low. For example, the ingestion of around 105 Salmonella cells is required to cause illness, while for some pathogenic protozoa and viruses, the ingestion of as little as one unit results in infection and illness (Kothary and Babu, 2001 [ref.08]Ref.08: Kothary, M.H. and Babu, U.S. (2001) Infective dose of foodborne pathogens in volunteers: a review. Journal of Food Safety 21(1), 49-73.; Leclerc et al., 2004 [ref.10]Ref.10: Leclerc, H., Schwartzbrod, L. and Dei-Cas, E. (2004) Microbial agents associated with waterbourne diseases. In: Cloete, T.E., Rose, J., Nel, L.H., Ford, T. (Editors) Microbial waterbourne pathogens. IWA publishing, London.).

Reliability of indicator bacteria

It is important to understand that indicator bacteria are indicative only. This means that there is always the possibility that where there are few indicator bacteria, pathogens may still be present, particularly those which are more resistant, and therefore survive longer, than coliform bacteria. Studies of the presence of pathogens in surface waters and pathogen removal through treatment processes have shown that correlations between coliform indicator bacteria and specific protozoan and viral pathogens are typically weak (Brookes et al., 2005 [ref.03]Ref.03: Brookes, J.D., Hipsey, M.R., Burch, M.D., Regel, R.H., Linden, L.G., Ferguson, C.M. and Antenucci, J.P. (2005) Relative value of surrogate indicators for detecting pathogens in lakes and reservoirs. Environmental Science and Technology 39, 8614-8621.; Harwood et al., 2005 [ref.06]Ref.06: Harwood, V.J., Levine, A.D., Scott, T.M., Chivukula, V., Lukasik, J., Farrah, S.R. and Rose, J.B. (2005) Validity of the indicator organism paradigm for pathogen reduction in reclaimed water and public health protection. Applied and Environmental Microbiology 71(6), 3163-3170.)

It is also important to be aware that high numbers of heterotrophic bacteria (total bacteria) in a water sample can sometimes give false positives with some tests for indicator bacteria. Some environmental species of bacteria may also appear on tests for faecal indicator bacteria, particularly faecal coliforms. In addition, it has been reported that E. coli are actually able to grow in soil in the tropics, in response to nutrient availability (Byappanahalli and Fujioka, 1998 [ref.05]Ref.05: Byappanahalli, M.N. and Fujioka, R.S. (1998) Evidence that tropical soil environment can support the growth of Escherichia coli. Water Science and Technology 38(12), 171-174.). In order to minimize the occurrence of false positives it is best to a) perform repeat tests and/or b) use an additional indicator, such as faecal Streptococci/Enterococci.

Field testing for indicator organisms

There are several types of tests for indicator organisms. These include the Most Probable Number (MPN) method, membrane filtration, and hydrogen sulphide tests. Here we deal with membrane filtration, with the example of the Delagua incubator, which is often used in the field.

Overview of membrane filtration

Coliform bacteria are a group of bacteria, which are broadly defined as gram-negative, lactose fermenting bacteria (Leclerc et al., 2001 [ref.09]Ref.09: Leclerc, H., Mossel, D.A.A., Edberg, S.C. and Struijk, C.B. (2001) Advances in the bacteriology of the coliform group: their suitability as markers of microbial water safety. Annual Review of Microbiology 55, 201-234.). The ability of coliform bacteria to ferment lactose and the resulting production of acid is exploited to allow relatively simple, visual testing of coliform bacteria. A sub-population of coliform bacteria is the thermotolerant coliforms, more commonly referred to as ‘faecal coliforms’, which multiply at 44.5°C. A significant proportion of these thermotolerant coliforms are typically Escherichia coli so sometimes the terms ‘faecal coliforms’ and ‘E. coli’ are interchanged although this is not strictly correct.

In any natural water source there exist general bacteria that are not from the coliform group of bacteria. Along with coliform bacteria, many of these non-coliform bacteria will multiply at 37°C in a standard incubator. In the Delagua membrane filtration method, the Membrane Lauryl Sulphate Broth (MLSB) media inhibits the growth of many of these bacteria, while acting as a food source for the coliform bacteria. The higher temperatures used in the Delagua (44.5°C) means that only thermotolerant coliform bacteria and some thermotolerant non-coliform bacteria will grow on the membrane filter. Thermotolerant/faecal coliform bacteria colonies are identified on the membrane filter by their yellow colour, produced from the original red broth (as a result of the fermentation of lactose). These yellow colonies are counted and indicate faecal pollution of the water source. Non-coliform bacteria typically appear red, pink, or transparent and are not counted.

Overview of Hydrogen Sulphide Tests (Presence/Absence Testing & Most Probable Number)

This experiment is a test for H2S. In order to check for the presence of these “sentinel” or “indicator” bacteria in water, we put them in contact with a nutritive substance (food for bacteria) plus a colour indicator that turns black when it comes in contact with hydrogen sulphide. If the test turns black, it means that H2S was produced, which in turn means that bacteria of faecal origin are likely present in the water sample.

Hydrogen sulphide producing bacteria are generally found in faeces. Some of these bacteria are unique to faeces, however, many are not. Therefore this test does not indicate total coliform bacteria but measures the number of bacteria that produce hydrogen sulphide under test conditions. A study by Sobsey and Pfaender (2002) showed good correlation between hydrogen sulphide test results and other indicator tests, however they do not unequivocally recommend them.

In assessing the applicability of the H2S test in presence/absence format, the limitations of Presence/Absence (P/A) testing should be considered. P/A testing is applicable where most tests provide a negative result. Where a significant proportion of tests provide a positive reaction, quantitative testing (e.g. membrane filtration or MPN) is preferred. In contrast to P/A testing, quantitative testing allows determination of the relative health risk and therefore the relative priority of need for improved treatment or for finding higher quality source water for supply.

Another problem encountered with the H2S test is that false positive results are possible from either naturally occurring H2S-producing bacteria that are sometimes found in water, or from H2S that is already present in the water. The former will show up only after several hours of sample/medium incubation, whereas the latter will show up quickly after the sample and medium are mixed together (minutes or tens of minutes). The possibility of false positives has been a concern about this type of test.

Overview of ColiPlate tests (MPN)

An alternative to the presence/absence test is the ColiPlate test, which quantifies coliforms and E. coli bacteria ranging from 5 to 5,000 colony forming-units (cfu) per 100 mL sample. Greater numbers of bacteria can be determined with dilutions. A coliform positive test results in a distinctive blue colour and an E. coli positive test produces fluorescence. The colour distinction enables analysis of brownish, turbid or rust-filled water.  The ColiPlate assay is based on: (i) the ability of coliform bacteria to utilize specialized nutrients and reagents to form a distinctive blue colour, and (ii) the ability of E. coli to form a fluorescent substance under the assay conditions. Quantification of coliforms and E. coli is based on the principle of the Most Probable Number (MPN) of colony forming-units per 100 mL.

Details of membrane filtration method

The Delagua is probably the easiest field incubator to use for membrane filtration. The kit, together with a full manual is available from the Robens Centre for Public and Environmental Health [ref.04]Ref.04: Robens Centre for Public and Environmental Health AW02, University of Surrey, Guildford, GU2 5XH, United Kingdom. Visit the website..

Preparation of culture

  1. Wash plastic bottles in warm rainwater. Note: pH of rainwater needs to be in range of 6.8 – 8.2. If not, find other source.
  2. Add a tub of broth powder (38.1 grams) to 500ml distilled water to a pan. Note: use all broth at once as it deteriorates in humidity.
  3. Heat so it dissolves but don’t boil
  4. Pour into plastic polypropylene bottles and screw on lids. Note: lids should not be tight but slightly loose while being sterilised; the quantity of broth used per sample is 2.5ml, and once a bottle is opened, it should be used within one day (by the book) or 2-3 days (in practice) – so depending on the number of tests planned per day, there should not be too much broth per bottle, otherwise you will end up throwing away a lot of broth. It is better to put a small quantity of broth in many bottles. Delagua supplies 10 bottles but it is better to have at least 20. Bottles should be of heat-resistant plastic – don’t use film canisters as these melt.
  5. Sterilise bottles on a rack in a pressure cooker at full pressure (112 degrees) for 15 mins. If there is an autoclave (121 degrees), sterilise for only 10 mins. Notes: some pressure cookers come with a built-in rack – this is very convenient. The pressure cooker should be set aside for this sterilisation and should not be used for cooking. You can sterilise the bottles in an open pan, but this will be only at 100 degrees and the broth has to be used within 24 hours – therefore it’s not a good idea and should be avoided.
  6. Allow the bottles to cool, making sure that the caps are still loose, then tighten caps when they are completely cool. Note: don’t overtighten as it can cause leakage, and don’t tighten while warm as the bottles will deform. If they do deform, loosen lid slightly and reform the bottle while it is still warm.
  7. Store in a cool dark place. Notes: the broth is stable for several months – if it gets cloudy or yellow, then it must be thrown away. However, cloudiness resulting from having been stored in a fridge is different and is the result of temperature, and will disappear once the broth is again at room temperature.

Sampling/testing

For method of sampling, see Delagua manual that comes with the kit.

  1. Sample bottles should be sterilised by boiling in a pan for 20 mins. Note: the pan used for general sterilisation should not be used for cooking and needs to be suitably marked. The bottles need to be able to hold at least 100ml.
  2. There is a sampling cup and cable with the Delagua. This is useful for sampling wells and off bridges. This sample cup should be sterile, especially when sampling possibly clean sources, such as a protected well.
  3. Sampling can also be done by hand in a scooping motion using the cup/jar, which is facing the flow of the current.
  4. Sample volumes are explained in the Delagua manual. However, in practice, good sample volumes to be filtered through the grid can be the following: for lakes/rivers 10ml, for unprotected wells/springs 50ml, and for protected sources 100ml. The reason for this is twofold – bacteria counts and colour: any more than 200 bacteria colonies are hard to count and also the colonies start to merge, making it inaccurate; if the water is also coloured then it will also be difficult to count yellow colonies as the grid will be stained a yellow/brown colour; also the colour itself may be due to small organic particles that will block the membrane quickly. So for these reasons, when you have a sample of river water, as river water generally has higher numbers of bacteria, more colour and higher organic matter in suspension, that is the reason for filtering less sample such as 10ml. There may be a time when even 10ml of sample will be too dirty, in which case dilutions will have to be made. Note: the actual volume can be measured by pouring the sample into the filter column to the marked points on it – this is not valid though for the 10ml mark as you usually can’t see it very well! For this it is better to use a sterile syringe. Also note that once you pour in water to the marker, it will start to filter through and the level will drop, so don’t keep topping up the level – it should be done once initially and not over some minutes.
  5. Samples should be in the incubator no later than 4-6 hours after sampling. Within this time, they need to be kept in a cool box with ice packs. After this time, bacteria will multiply in the sample water making the result meaningless. Note: there should be 1 hour of recuperation time from the time the last sample was taken and when the incubator is switched on. This is to allow the bacteria to recuperate in a new environment.
  6. Sample bottles should be marked with the date, time and sample name as soon as the sample is taken, otherwise you will forget.
  7. The petri dishes should also be marked in the same way. Eyeliner is good for marking, but pencil also is fine. Don’t use permanent marker as the petri dishes are always re-used.

For actual procedure of using the filter apparatus, refer to the Delagua manual.

Reading results

Switch the incubator on in the afternoon so that you can read results the next day in good light.

The incubation period is for 16-18 hours. If you cannot count the colonies straight away at the time the incubator is switched off, don’t worry as you can also count them a few hours later as long as the incubator is kept shut. 1 CFU = 1 Colony Forming Unit. These look like lumps/dots but represent one bacterium that has multiplied to form a colony using available food. The more bacteria there are, the more competition and therefore the smaller the colonies will look. The reason for the incubation is to produce these colonies that can be visible to the naked eye. Colonies should be counted in good natural light but not bright sunlight. Looking at the colonies from various angles in the light helps to differentiate a transparent colony from a yellow colony, as sometimes this is difficult to make out, especially when there are many colonies and the whole background colour will have changed from red to yellow. Only yellow colonies are to be counted, and the number arrived at should be multiplied by the number of ml used as faecal coliforms are expressed in terms of 100ml (e.g. – if 10ml was filtered, and count is 20, the result is 10 x 20 = 200CFU/100ml).

If the water is contaminated, as a preference to treatment, preventative measures should be looked at to prevent pollution. Possible contamination routes can be identified prior to any treatment via a sanitary inspection.

In very contaminated samples, or in very coloured water or in water with a high organic content, samples may need to be diluted if the results cannot be read when 10ml is passed through the membrane due to a high CFU count or a strong colouration. Indeed, 10ml may not even pass through due to clogging. In these cases, dilution may be necessary rather than using fewer number of ml (due to reduced accuracy). The original sample needs to be diluted with a solution that does not feed the bacteria so that they multiply, but equally does not cause the cell walls to explode – for this reason, distilled water is unsuitable. A decent solution for this is called ‘¼ Ringers’. A sample calculation is made as follows:

Dilution No. CFU Diluted by To get to 100ml No. CFU/100ml
10ml original sample Too many to count Not diluted x 10 Uncounted
1ml original + 9ml ¼ Ringers = 10ml (A) 120 x 10 x 10 12,000
1ml (A) + 9ml ¼ Ringers = 10ml (B) 10 x 100 x 10 10,000
1ml (B) + 9ml ¼ Ringers = 10ml (C) 0 x 1000 x 10 0
Average 11,000

The number of CFU needs to be multiplied up by both the dilution and by the original sample size (in this case 10ml). The average is taken between results, hence 11,000.

The make-up of ¼ ringers is done as follows:

  • Either use one ¼ Ringers tablet to 500ml of distilled water and then sterilise or
  • or use 2.25 grams sodium chloride, 0.105 grams potassium chloride, 0.12 grams calcium chloride and 0.05 grams sodium bicarbonate into separate beakers and then add approx. 100ml distilled water to each and stir until dissolved. Transfer each to 1 litre beaker and rinse each small beaker and add rinsings. Then make up to 1 litre with distilled water. Mix well and stopper.

Sterilisation issues

One fingertip in the wrong place and it could deposit a whole load of bacteria that will start to breed. Sterilisation procedure is therefore very important if the results are to mean anything.
Things that need to be sterilised are:

  1. Whatever contacts the water sample: sampling cup, sample bottles, internal surface of filter funnel, upper part of filter base, surface of bronze disc.
  2. Things in contact with the broth: internal surfaces of petri dishes, absorbent pads, tweezers.
  3. Things in contact with the membrane/grid: filter apparatus, absorbent pads, tweezers.

Practically therefore, the following procedures need to be followed:

  1. Sterilise sample cup/bottles, syringes and petri dishes in a pan by boiling for 15mins. If a pressure cooker is used, the time can reduce to 10 mins.
  2. Ensure that the absorbent pad dispenser is kept in a closed container, and that the dispenser arm and base are not touched by fingers or the edge of the petri dish itself when dispensing the pad. When initially placing the pad tube on the dispenser, ensure that you don’t touch the inside surface of the base.
  3. The filter apparatus needs to be sterilised between each filtration of a different water source sample. This is done by putting the funnel apparatus into the middle position (loose but not removable- so that all crevices are sterilised), throwing out the filtered water from the cup, drying it with tissue, adding 20 drops of methanol and lighting it away from your face. When the ring of blue flame is almost burnt up, replace the other part of the filter (with the funnel etc) upside down inside the cup. This ensures that the methanol burns without oxygen to form formaldehyde, an effective disinfectant. Theoretically you should wait 15 mins before removing the apparatus for the next sample, but in practice the time is shorter due to the time constraint of 4-6 hours for the samples. Note: the apparatus can be boiled in a pan for 5 mins between filtrations if there is no methanol, but this method takes too long because of the time needed for the equipment to cool down before use.
  4. Tweezers should always be flamed before every use, and placed in the top of the Delagua lid (or at least not on any surface). It is good to always flame the tweezers when you finish an action, so that they are cool and ready for the next work.
  5. Incinerate any used pads and grids. Ensure they burn completely as they are full of bacteria.

References:

Ref.01: House, S.; Reed, R. (1997). Emergency Water Sources. WEDC, Loughborough, UK, p.171. Available online.

Ref.02: Arbelot, A. (Ed.) (1994). Public Health Engineering in Emergency Situation. Médecins Sans Frontières, Paris, France.

Ref.03: Brookes, J.D., Hipsey, M.R., Burch, M.D., Regel, R.H., Linden, L.G., Ferguson, C.M. and Antenucci, J.P. (2005) Relative value of surrogate indicators for detecting pathogens in lakes and reservoirs. Environmental Science and Technology 39, 8614-8621.

Ref.04: Robens Centre for Public and Environmental Health AW02, University of Surrey, Guildford, GU2 5XH, United Kingdom. Visit the website.

Ref.05: Byappanahalli, M.N. and Fujioka, R.S. (1998) Evidence that tropical soil environment can support the growth of Escherichia coli. Water Science and Technology 38(12), 171-174.

Ref.06: Harwood, V.J., Levine, A.D., Scott, T.M., Chivukula, V., Lukasik, J., Farrah, S.R. and Rose, J.B. (2005) Validity of the indicator organism paradigm for pathogen reduction in reclaimed water and public health protection. Applied and Environmental Microbiology 71(6), 3163-3170.

Ref 07: Huisman, L; Wood, W.E. (1974). Slow Sand Filtration. WHO, Geneva, Switzerland. p.44. Available from WHO

Ref.08: Kothary, M.H. and Babu, U.S. (2001) Infective dose of foodborne pathogens in volunteers: a review. Journal of Food Safety 21(1), 49-73.

Ref.09: Leclerc, H., Mossel, D.A.A., Edberg, S.C. and Struijk, C.B. (2001) Advances in the bacteriology of the coliform group: their suitability as markers of microbial water safety. Annual Review of Microbiology 55, 201-234.

Ref.10: Leclerc, H., Schwartzbrod, L. and Dei-Cas, E. (2004) Microbial agents associated with waterbourne diseases. In: Cloete, T.E., Rose, J., Nel, L.H., Ford, T. (Editors) Microbial waterbourne pathogens. IWA publishing, London.

Mr. TMicrobiological