Content uploaded by Alexander Kouzminov
All content in this area was uploaded by Alexander Kouzminov on Apr 03, 2018
Content may be subject to copyright.
EVALUATION OF THE H2S PAPER STRIP TEST
– A FIELD TEST FOR ASSESSING THE
MICROBIOLOGICAL QUALITY OF WATER
T. Wnorowski, Watercare Services Limited, Auckland, New Zealand
A Kouzminov, Ministry of Health, Wellington, New Zealand
Epidemics arising from drinking water contaminated with faecal matter are a global health problem. Standard
methods currently used for routine microbiological testing of drinking water have many limitations such as the
requirement of short transport time, need to employ specialist personnel and equipment, high costs of testing
materials. Manja et al (1982) developed a simple and inexpensive field test that could be particularly useful for
screening of water supplies in remote and rural areas.
The present paper describes validation of the H2S strip method. The new method was compared with the
standard microbiological tests including total coliforms, E.coli, heterotrophic plate count, sulphite reducing
clostridia spores count, aerobic spores count and F-specific bacteriophage count. A limited number of samples
were also tested in parallel for pathogenic bacteria (Campylobacter and Salmonella) and viruses (Adenoviruses,
Enteroviruses, Noroviruses and Hepatitis A virus).
A total of 312 samples were studied. Different types of water samples, including both uncontaminated drinking
water and contaminated environmental sources (tank waters, bore waters, surface waters, geothermal waters)
were tested. Geothermal groundwater and high manganese water were specifically included to determine
whether presence of hydrogen sulphide and manganese salts in water may cause interferences. Sensitivity of the
test was assessed by spiking the samples with low and high levels of bacteria known to produce H2S. Both
reference and wild strains isolated from the H2S-positive samples were used.
A range of temperatures (ambient , 20 oC, 25 oC, 30 oC and 35 oC) and incubation times (24 and 48 hrs) were
tested. Although the H2S method could be used at the temperature range from ambient to 35 o
between 30-35 oC produced the results faster.
The results indicate that the H2S paper strip method and the Total Coliform/E.coli Colilert test were equally
effective in the detection of faecal contamination in water samples. Especially, excellent agreement (97.3%) was
found between the H2S method and E.coli by the Colilert test. The H2S method does not detect viruses but
detects microorganisms other than coliforms that are associated with faecal contamination, including Clostridium
perfringens, Salmonella and bacteriophages.
The H2S paper strip test appears to be a sensitive, simple and inexpensive procedure for screening of water
supplies towards potential contamination. It eliminates complicated procedures and costly chemicals and lab
equipment, including an incubator in subtropical and tropical regions.
Hydrogen sulphide paper strip test, H2S-metabolising bacteria, drinking water standards, E.coli
An important aspect of the protection of public health is the provision of safe and reliable drinking water. This
means that the water used for domestic purposes should be free of pathogenic microorganisms and other
substances that may present a health risk. It is impractical to test water supplies for all potential pathogens for a
number of reasons such as the time required to carry out tests, the cost of testing and the inability to test for some
organisms. For this reason a system of testing for organisms that indicate faecal contamination (indicator
organisms) is used to ensure the safety of drinking water.
Indicator organism tests, such as those for faecal coliforms and E. coli, have limitations. They require trained
staff and expensive materials and equipment for their execution. In addition, the monitoring of water supplies in
remote areas is hindered by the requirement for samples to be tested within 24 hrs of sample collection, for
results to be valid. If no resources are available locally, remoteness and a lack of funds may make it impractical
to adequately monitor drinking water.
The H2S paper strip test is simple and inexpensive. This method was developed by Manja et al (1982) and it is
particularly suitable for developing countries with ambient temperatures of 25-44 0C. The method has been
extensively evaluated during last 20 years. Four commercially produced brands of the H2S medium are available
The purpose of this trial is:
• To validate the H2S paper strip method against the reference coliform/E. coli most probable number
(MPN) method (Colilert® MPN 9223 APHA, 1998) using both naturally contaminated and spiked
• To test possible limitations and sources of misinterpretation in the H2S paper strip test
• To determine the sensitivity of the H2S paper strip method against the reference coliform/E. coli method
• To determine the sensitivity of the H2S paper strip method against methods for detection/ enumeration of
other faecal indicator microorganisms and pathogens (e.g. Salmonella, Campylobacter, viruses)
• To determine whether presence of sulphide-containing groundwaters and high manganese waters cause
• To provide cost estimates for the manufacture of the H2S paper strip test.
2 MATERIALS AND METHODS
2.1 SAMPLE COLLECTION
Samples were collected from various natural sources such as shallow and deep bore wells, roof water supplies,
surface waters including rivers, dams, streams, geothermal wells and treated drinking water supplies.
A minimum 800mL of water was collected into sterile 1L bottle for each sample. For the pathogen study, larger
volumes, up to 100L were collected in sterile plastic containers. For chlorinated waters, sodium thiosulphate
was added to sample bottles. All samples were processed in the laboratory within 24 hours of collection and
were maintained at low temperature during transport.
2.2 HYDROGEN SULPHIDE STRIP TEST METHOD
2.2.1 MEDIA (STOCK SOLUTION)
The concentrated medium (final volume of 100mL) was prepared using the composition below:
Bacteriological peptone (e.g. Difco® Bacto Peptone) 40.0g
Dipotassium hydrogen phosphate 3.0g
Ferric ammonium citrate 1.5g
Sodium thiosulphate 2.0g
Sodium dodecyl sulphate (SDS) 0.4g
Distilled water 100mL
The H2S medium was prepared according to Manja et al (1982). In the original recipe 2mL of Teepol 610 was
used as a surfactant. This product is not available anymore. Sodium dodecyl sulphate (SDS) was used as a
substitute surfactant for Teepol, according to Jangi at al (2001). The concentration of 0.02% SDS after the
addition of water sample gave the closest results to the Teepol preparation.
2.2.2 PREPARATION OF H2S PAPERS
H2S paper strips were prepared by pouring of 3.2mL aliquots of stock solution onto 1 absorbent paper pad from
the membrane filtration apparatus (Sartorius®, 47mm diameter, 1.5 mm thickness). A 100mL of water sample
required 3.2 mL of media (1 thick Sartorius® absorbent pad). The pads were dried in an oven at 550C, placed in
steriliser paper bags and autoclaved for 15 minutes at 1210C. These reagent-impregnated pads can be stored dry
(in their steriliser bags) for several months – until ready to use.
When ready to conduct the test, two pads were placed into a 250mL sterile Schott® bottle (or a sterile plastic
container) and a volume of 100mL of water sample was added.
2.3 TOTAL COLIFORMS AND E.COLI
Total coliforms and E. coli densities were estimated using the Colilert® MPN method (Colilert® MPN 9223
APHA, 1998). For this test, the 97-well Colilert® Quanti-trays were incubated at 35 ± 0.50C for 18 hours.
2.4 IDENTIFICATION OF BACTERIA
Positive cultures were streaked onto plates of Tryptose Soy agar (TSA) for aerobic bacteria and TSC agar (for
Clostridium spp.). They were incubated at 350C aerobically (TSA) and anaerobically (TSC). The bacterial
isolates were identified by using API® kits (API20E®, API20NE® etc).
2.5 ALTERNATIVE FAECAL INDICATORS
F-specific Bacteriophages enumeration was performed according to ISO 1075–1E (1995).
For the heterotrophic plate count, samples were plated onto Plate Count Agar and incubated at 35 0C for 48
hours (pour plate method).
Clostridia spores were enumerated by heating the 100 mL of the sample at 750C for 15 minutes, filtering the
heat-treated sample through 0.45 µm filters and incubating the filters anaerobically on the TSC agar.
Aerobic mesophilic spores count (Bacillus spore count) was performed according to the British Standard 4285
(1986). Samples were heated at 750C for 15 minutes, plated onto Plate Count Agar with starch and incubated for
3 days at 300C. Bacillus isolates were identified with the API 50Ch® kit.
The volume of water sample used varied with the test. For Salmonella and Campylobacter enumeration, 1L
volumes were filtered through membrane filters, while 100L samples were filtered through hollow fibre filters
Salmonella analysis consisted of a number of steps: filtration through 0.45µm filters, pre-enrichment in Buffered
Peptone Water, enrichment in selective broths (RVS and Mannitol Selenite broths), growth and detection on
selective solid media (XLD and modified Brilliant Green agars) and confirmation of Salmonella isolates.
For Campylobacter enumeration, 1L samples were filtered through 0.22µm filters, enriched in Bolton
Enrichment Broth and plated onto Campylobacter Isolation Blood-free agar plates. The plates were incubated in
microaerophilic conditions at 410C. Both Salmonella and Campylobacter were enumerated using MPN methods.
For virus analysis, material retained on a hollow fibre filter was eluted and concentrated. Disposable filters (one
per each sample) were used for virus concentration. A negative control (20L) was pumped through a filter before
concentrating of each sample. The virus concentrates were divided into 3 portions.
Two were tested for enumeration of culturable adenoviruses and enteroviruses grown on appropriate cell lines.
Enteroviruses were tested by the suspended plaque assay on the African Green Monkey kidney (BGM) cell line.
Adenoviruses were tested by the MPN method using two cell lines (Hep 2 Clone B and 293 N3S) and confirmed
with the Adenovirus Direct Immunofluorescent Assay kit. ESR, Wellington using the PCR method, tested the
third portion for Noroviruses and Hepatitis A viruses.
3 STUDY 1: COMPARISON OF H2S PAPER STRIP METHOD WITH THE
The purpose of this study is to compare the H2S paper strip method with a referee method. The enzyme substrate
coliform test (E. coli) APHA 9223B, 1998) is cited as a referee method in Drinking-water Standards for New
Zealand 2000 (MoH, 2000). The Colilert® test has been used for this study.
3.2 EXPERIMENTAL PROCEDURE
All samples were tested by the H2S paper strip test (in quintriplicate) and the Colilert® MPN method. To study
the temperature range at which the H2S paper strip method was effective, 20, 25, 30, 350C and ambient
temperature were tested. The ambient temperature varied between 17-230C during the period at which the
experiments were conducted.
100 mL of water sample to be tested was placed in a sterile bottle containing 2 absorbent pads impregnated with
the H2S medium, allowed to stand for 5 minutes, shaken and stored in appropriate incubators (20, 25, 30 and
350C). One bottle (from each sample) was left on the bench at room temperature.
All bottles were examined after one hour of incubation to check for sulphide already present in the sample (e.g.
from sediments). The rapid reaction of iron with sulphide present in a water sample could result in darkening of
the H2S test almost immediately upon addition of the sample. For this reason, it is very important that the test
procedure includes a visual inspection after one hour of incubation to exclude samples producing quick or early
positive reactions. Any sample exhibiting such rapid discoloration indicates that it is contaminated by sulphides
and should be treated as a false positive.
The bottles were incubated at various temperatures and examined after 24 hours and 48 hours to determine the
extent of blackening in the bottles due to the reduction of the ferric ammonium citrate by any hydrogen sulphide
gas produced by microorganisms. The date and time of observation was recorded on the report form and the
observations were recorded with: (-) = no blackening, no growth, (-G) = no blackening, growth (turbid), (+) =
trace of blackening, (++) = the paper strip was partially back, up to half of the bottle turned black, (+++) the strip
and entire bottle was dark black.
Isolates from all positive H2S bottles were identified.
A set of the QA/AC controls (blanks, positive and negative controls) was processed with each batch of samples.
Citrobacter freundii and Salmonella typhimurium in low concentrations (5-10 cfu/100mL) were used as positive
control. E.coli was used as negative control (growth but no blackening). Sterile distilled water was used as a
blank. The blank was used as a benchmark to compare the extent of colour change in test samples and to ensure
that sample bottles and H2S paper strips had been properly sterilised prior to use.
3.3 RESULTS AND DISCUSSION
A total of 244 samples were included in the assessment of the H2S paper strip medium. Various types of water
samples, ranging from uncontaminated drinking water samples to contaminated environmental samples were
collected and tested using the H2S paper strip medium. The sample range included the following: treated water
(20) surface waters (71), bore well waters (72), roof (tank) water samples (70) and geothermal well waters (11).
A summary of results of tests performed on these samples are shown in Table 1.
Table 1: Comparison of the H2S strip method and the Colilert® MPN method
Colilert® Method H2S Method – positive samples
Positive samples Negative
TC* E.coli 24
71 71 54 0 6 61 3 58 20 61 56 64 57 62
Bore Water 72 55 29 17 0 37 0 34 13 42 35 45 39 42
70 50 29 20 3 17 0 14 2 30 20 40 28 35
20 2 0 18 0 0 0 0 0 0 0 2 0 1
11 5 1 6 0 0 0 0 0 0 5 5 5 5
TOTAL 244 183 113 61 9 115 3 106 35 133 116 156 129 145
% (all) 100 75 46 25 4 47 1 43 14 55 48 64 53 59
100 61 0 5 63 1 57 19 73 64 85 71 79
* TC = total coliforms
Approximately 75% of the samples analysed in this study (183 samples) contained total coliforms, and nearly a
third of them contained high concentrations (>100cfu/100mL). Of the 244 samples analysed, approximately half
(113 samples) contained E.coli bacteria so the water collected from untreated water sources would not comply
with the NZ Drinking Water Standards 2000 (MoH, 2000).
Out of 244 samples tested, 157 samples were positive by both the H2S and total coliforms test and 60 samples
were negative by both techniques. There were 26 samples, which were positive for total coliforms but were
negative for the H2S test and 1 sample, which was coliform-negative, but the H2S-positive (Tables 1 and 2).
Table 2: Comparison of H2S method and Colilert® MPN method
Positive samples Negative samples Total
Colilert H2S method
Colilert H2S method
48 hrs 24 hrs 48 hrs
N° of Samples 113 103 110 183 129 156 61 60 244 181 210
91 97 71 85 102 71 86
88 % of samples which were H2S-negative, but total coliform-positive, contained low concentrations of
coliforms (1-20 MPN/100mL) and 3 H2S-negative samples contained 30-55 MPN/100mL, see Table 3. 86 % of
the samples tested showed similar positive and negative responses in the H2S paper strip test and the MPN
Colilert® method, see Table 2. If the H2S paper strip test was to be used alone, 84.8% of the contaminated water
samples would be identified, in comparison to 99.4% if the total coliforms MPN procedure was used.
When the average total coliforms was more than 20 MPN/100mL, the H2S test showed 98% agreement with total
coliforms results. Similar results were obtained by various investigators who had tested the H2S method in
different tropical and temperate regions, including Indonesia, Peru, India, Chile, Malaysia, Nepal and South
Africa (Ratto et al., 1989; Kromoredjo and Fujioka, 1991; Rijal and Fujioka, 1998; Manja et al., 1982; Manja et
al., 2001; Kaspar et al., 1992; Castillo et al., 1994; Venkobachar et al., 1994; Martins et al., 1997; Genthe and
Table 3: Negative results by the H2S method but positive by the MPN Colilert® Method
Number of Samples Colilert® - Total Coliform Ranges
1-10 11-20 30-55
26 17 6 3
100% 65% 23% 12%
Excellent agreement (97%) was found between the H2S test and E. coli MPN Colilert® method. Most of E .coli
positive samples (91.1%) turned black in less than 24 hours incubation at 30-35 0C.
As illustrated in Figure 1, 48 hours incubation period was found to be optimal for H2S test. After 24 hours
incubation at lower ranges of temperatures (17-25 0C), only 1.7-19.6% of the contaminated samples were
showing a positive H2S reaction (usually slightly black). At higher temperatures (30-35 0C), 62.3-69.7% of the
contaminated samples were H2S-positive after 24 hours incubation.
Figure 1: Comparison of numbers of positive samples for the Colilert® test and the H2S strip test at
different incubation times and temperatures
Colilert 18hr 35ºC Am bient 17-23ºC 20ºC 25ºC 30ºC 35ºC
Number of positive samples
The number of H2S positive reactions significantly increased (up to 84.8% at 300C) when the incubation period
was extended to 48 hours. This indicates that at lower temperatures growth of H2S producers was slower and
therefore H2S production was delayed. This observation was more pronounced at lower concentration of
coliforms. They seemed to require a longer period to obtain H2S positive results. At all temperatures, the
incubation period required to display a positive reaction increased with a decrease in the coliform concentration
(Figure 2). At lower concentrations and at lower temperatures the black colour only appeared as small patches at
the bottom of the bottles and did not extend to the entire water sample. This could be due to the restricted
growth of the H2S producing bacteria because of low numbers or low temperatures.
Figure 1: Relationship between coliform concentrations and numbers of H2S positive results
a. 24 hour incubation
Total coliforms by Colilert
H2S Ambient (17-23ºC) H2S 20ºC H2S 25ºC H2S 30ºC H2S 35ºC
Number of positive samples
H2S Ambient (17-23ºC) H2S 20ºC H2S 25ºC H2S 30ºC H2S 35ºCTotal coliforms by
Colilert 18hr 35ºC
1-50 MPN/100ml 51-250 MPN/100ml >250 MPN/100ml Total positives
1-50 MPN/100ml 51-250 MP N/100ml >250 MPN/100ml Total positives
b. 48 hour incubation
Total coliforms by Colilert
H2S Ambient (17-23ºC) H2S 20ºC H2S 25ºC H2S 30ºC H2S 35ºC
Number of positive samples
H2S Ambient (17-23ºC) H2S 20ºC H2S 25ºC H2S 30ºC H2S 35ºCTotal coliforms by
Colilert 18hr 35ºC
Pillai et al. (1999) noted that if the temperature was between 28-440C, blackening could be obtained within 48
hrs while at 220C it could take up to 90 hours for the same sample. Their results are in good agreement with this
study. However, Genthe and Franck (1999) reported that the results of the H2S strip test best correlated with
indicator organisms when left to incubate for 48 hrs at 220C. Other workers (Castillo et al., 1994; Ratto et al.,
1989 and Kasper et al., 1992) concluded that effectiveness of the method was independent of the temperature but
most of them incubated their samples at higher temperatures (25-350C).
A total of 37 bacterial species were isolated from the positive H2S bottles (Appendix 1f). Most of them belonged
to the Enterobacteriaceae family. E. coli (58 isolates), Klebsiella spp (64), Enterobacter spp. (32), Proteus spp.
(29), Citrobacter freundii (22) and Serratia spp. (23) were the most common coliforms detected.
58 of the contaminated samples (37%) were also positive for Clostridium spp. It would appear that clostridia
might play a role equivalent to that of the coliform organisms in producing a positive H2S paper strip reaction.
More than half of Clostridium spp. isolates was found in surface water samples.
A small number of pathogenic bacteria such us Salmonella spp. (8 samples) and Yersinia spp. (3 samples) were
isolated from the positive H2S bottles. This indicates that the H2S paper strip test may be an indicator of
bacterial pathogen contamination.
Another significant group of bacteria isolated from the positive H2S bottles were Gram negative oxidase positive
bacteria: Aeromonas hydrophila (35 isolates) and Pseudomonas spp.
Several investigators (Castillo et al., 1994; Ratto et al., 1989; Nagaraju and Sastri, 1999) found a large variety of
bacteria, primary Enterobacteriaceae and Clostridium perfringens, in samples giving a positive reaction in the
H2S test: Enterobacter, clostridia, Citrobacter freundii, Klebsiella, Escherichia, Salmonella, Acinetobacter,
Aeromonas, Morganella, Proteus, Hafnia etc.
4 STUDY 2: COMPARISON OF H2S PAPER STRIP TEST WITH A
SELECTION OF ALTERNATIVE FAECAL INDICATORS
Following a recent trend towards the use of ‘alternative’ indicator organisms the study was extended to evaluate
a wider range of faecal indicators including sulphite reducing clostridia, aerobic spores (Bacillus spores),
heterotrophic bacteria and F-specific bacteriophages, when compared with the H2S paper strip method. As for
Study 1, H2S strips were incubated at various temperatures but for 48 hours only. In addition, the samples were
analysed for the following tests: total coliforms and E.coli (MPN Colilert® method), sulphite reducing clostridia,
aerobic spore count, heterotrophic plate count at 35oC and F-specific bacteriophages.
4.2 EXPERIMENTAL PROCEDURE
A total of 58 samples, comprised of surface water (15), tank water (16), bore water (17) and geothermal water
samples (10) were analysed. The summary of the results is presented in Table 4. The number of chemical tests
were also performed on those 10 geothermal water samples.
4.3 RESULTS AND DISCUSSION
Out of 58 samples tested, 46 were positive in both H2S and total coliform tests and 8 were negative in both tests.
There were also 4 samples that were positive for total coliforms but negative in the H2S method. Even when the
total coliforms numbers were low (1-20 MPN/100mL) the H2S test displayed a very good correlation with the
total coliform test.
Table 4: Summary of results of comparison of H2S strip test with alternative faecal indicators
Sample Type Result
Surface water Tank water Bore water Geothermal
Number of samples 15 16 17 10 58
H2S test +ve 15 13 11 7 46
-ve 0 3 6 3 12
Total coliforms +ve 15 15 12 8 50
MPN/100 mL -ve 0 1 5 2 8
E.coli +ve 11 8 7 7 33
MPN/100 mL -ve 4 8 10 3 25
Sulphite reducing clostridia 1-50 4 5 2 4 15
Cfu/100mL 51-330 10 4 0 4 18
<1 1 7 15 2 25
Bacillus spores cfu/100mL 1-1000 4 13 5 3 25
1001-4900 11 3 8 7 29
<1 0 0 4 0 4
Bacteriophages 10-100 7 6 12 5 30
Pfu/100L 101-960 8 0 0 1 9
<10 0 10 5 4 19
Heterotrophic Plate 1-200 3 8 14 5 30
Count cfu/mL 201-26000 12 8 2 4 26
<1 0 0 1 1 2
Excellent agreement was found between the H2S test and E. coli MPN Colilert® method as well as between the
H2S test and sulphite reducing clostridia method. Nearly all E .coli and/or sulphite reducing clostridia-positive
samples were also positive in the H2S method. All these samples turned black after only 24 hrs of incubation at
30-35oC even when the samples contained very low numbers (1-20 per 100mL) of E. coli or clostridia.
The H2S test detects bacteria other than coliforms associated with faecal contamination, such as Clostridium
perfringens which is one of the more resistant indicators of faecal contamination and can still be found when
coliforms are no longer present (Sobsey and Pfaender, 2002). Grant and Ziel (1996) have found a strong
agreement between the H2S paper strip method and clostridia spore enumeration. The H2S test produced about
10% more positive samples than the coliform test because it included samples that were positive only for
clostridia (Castillo et al., 1994).
Bacillus spore count, heterotrophic plate count and F- specific bacteriophage count were found to be of no use as
indicators of faecal contamination in water. Their correlation with the H2S method was poor. A satisfactory
agreement between the H2S method and these three tests was found only for heavily contaminated samples.
However, the samples testing positive also contained high numbers of total coliforms, E.coli and/or sulphite
reducing clostridia and the positive reaction was most probably caused by these organisms. Similar results were
reported by Genthe and Franck (1999). They observed poor correlations between the heterotrophic plate count
and the H2S test.
A number of samples that contained relatively high numbers of Bacillus spores and heterotrophic bacteria were
found to be negative for total coliforms, E .coli, sulphite reducing clostridia and in the H2S method. 97% of
samples tested positive for heterotrophic bacteria and 93% for Bacillus spores while only 79% samples were
found positive by the H2S method.
A good agreement was found between F-specific bacteriophages and the H2S method for surface water and bore
water samples. A comparison with tank water was not feasible as only a few tank water samples contained
bacteriophages (and at very low numbers). Most of the tank water samples that tested positive in the H2S
method were negative for bacteriophages.
The studies of Martins et al., 1997 and Castillo et al., 1994 have indicated that both the H2S paper strip test and
bacteriophage test are viable indicators of potable water quality and potable water treatment.
5 STUDY 3: COMPARISON OF H2S PAPER STRIP METHOD WITH THE
PRESENCE OF PATHOGENS
The purpose of testing for indicator organisms in drinking water is to detect the risk of disease to the consumers.
For this reason a limited study was carried out to relate the presence of a number of pathogens to the results from
the H2S strip test.
5.2 EXPERIMENTAL PROCEDURE
10 samples were analysed to compare the performance of the H2S paper strip method with the presence of
A total of 10 samples, comprised of surface water (4), tank water (3) and bore water (3) were tested tested for
bacterial (Campylobacter, Salmonella) and viral (Enteroviruses, Adenoviruses, Noroviruses, Hepatitis A)
pathogens as well as the alternative faecal indicator organisms tested in Study 2 and the H2S paper strip method.
. Only samples, which were expected to have high numbers of microorganisms, were used.
5.3 RESULTS AND DISCUSSION
All four surface water samples contained high numbers of total coliforms, E .coli, sulphite reducing clostridia,
Bacillus spores, heterotrophic bacteria and bacteriophages. They were also highly positive in the H2S method.
Three of the four surface samples were found to be positive for Noroviruses, Adenoviruses, Enteroviruses,
and/or Campylobacter and Salmonella. The fourth sample contained high numbers of faecal indicator bacteria
but was found to be free of pathogenic organisms. All 4 samples were collected from small dams which are used
as raw water supplies for producing treated drinking water for small towns.
Results of Gawthorne et al. (1996) show that H2S test can indicate the presence of salmonellae. Their
recommended length of incubation for a negative result is 48 hours at 35oC to exclude the possibility of slow-
growing or injured bacteria in the sample.
However, three samples that were supposed to be very “clean” because they contained very low numbers or were
totally free of faecal indicator bacteria and also were negative (or weakly positive) in the H2S test, were found to
harbour pathogens. Particularly worrying was to find Noroviruses and Salmonella in bore and tank water
samples that tested negative in both E. coli and H2S tests. Two of these Norovirus positive samples were
collected from the tanks of people who complained about chronic unexplainable diarrhoea. Both these water
samples were E. coli free so they were meeting the criteria of NZ Drinking Water Standards.
The results of this trial confirm the known fact that high levels of faecal bacteria may increase the chances of
finding pathogens but it cannot be guaranteed that water free of faecal indicator bacteria is also free of pathogens
In summary, the H2S method has shown good potential to indicate the presence of pathogenic bacteria and
viruses - six out of seven H2S-positive samples were also found to be positive for one or more pathogens.
6 STUDY 4: EVALUATION OF POTENTIAL INTERFERENCES WITH THE
H2S STRIP TEST
Geothermal waters contain hydrogen sulphide and other chemicals, which might interfere with the H2S, strip test
by producing false positive or negative results. Manganese was also considered to be a possible source of
6.2 EXPERIMENTAL PROCEDURE
6.2.1 INITIAL STUDY
A total of eleven samples were collected. Ten samples were collected from geothermal springs, bores, shallow
geothermal pools, lakes or streams. All had a strong or very strong smell of H2S. One treated water sample was
collected from the reticulation system in a geothermal region, to serve as a control.
Five of the samples were collected from the hot springs or bores with water temperature approximately40-700C,
and 6 samples from their outlets (temperature approximately 30-400C) or from shallow open pools (at ambient
temperature). Additional samples were collected in separate bottles (with preservative) for sulphide analysis.
Samples were tested by the H2S paper strip test and the MPN Colilert® method.
Only samples taken straight from the hot springs contained high amount of sulphide (0.15-0.5mg/L). No
sulphide was found in the samples taken from the bore outlets. Full agreement (100%) was found between total
coliforms and the H2S test results for all 11 samples. High content of sulphide did not interfere with the H2S
method and no false positive results were produced.
6.2.2 CHEMICAL CHARACTERIZATION
In a second part of the study, an additional ten geothermal water samples were collected to determine their
chemical composition. Nine samples were collected from geothermal springs, bores, shallow geothermal pools,
lakes or streams. All had a strong or very strong smell of H2S. One treated water sample was collected from the
reticulation system in a geothermal region to serve as a control.
Samples were analysed for total coliforms and E.coli (MPN Colilert method), sulphite reducing clostridia,
aerobic spore count, heterotrophic plate count at 35oC and F-specific bacteriophages. Chemical analyses were
arsenic, boron, iron, dissolved oxygen, manganese, pH, sulphide and total dissolved solids. In the H2S-strip test,
all samples were incubated for 1 hour at appropriate temperatures and examined for early blackening.
6.2.3 MANGANESE INTERFERENCE
Raw waters with elevated manganese level are rare in New Zealand. Therefore samples with elevated levels of
manganese had to be artificially created. 8 water samples were spiked with 0.5-1.5mg/L of manganese. Two
samples with naturally elevated level of manganese were also tested in this trial.
6.3 RESULTS AND DISCUSSION
Most of the samples contained high levels of one or more of the following chemicals: arsenic, sulphide, boron,
iron and manganese. Three samples had a very low pH (3-3.3).
In the H2S-strip test, all samples were incubated for 1 hour at appropriate temperatures and examined for early
blackening to evaluate the effect of naturally occurring H2S. No samples changed colour to black after 1 hour of
A good comparison was found between the H2S method and the total coliforms, E. coli and sulphite reducing
clostridia tests in all 10 samples.
Ten samples containing high level of manganese in the water were tested by the H2S paper strip and MPN
Colilert® methods. Only two samples were collected from naturally contaminated source, the remaining samples
(8) were spiked with 0.5-1.5 mg/L of manganese. Full agreement (100%) was found between total coliforms/E.
coli tests and the H2S paper strip method. High content of manganese did not interfere with the H2S test results
and no false positive or false negative results were produced (see Appendix 4b).
7 STUDY 1: STUDY 5: SPIKING TRIAL
This study was carried out to determine the response of the H2S strip media to known organisms in
predetermined inoculum ranges. The sensitivity of the method was tested with low level inocula.
7.2 EXPERIMENTAL PROCEDURE
The following 14 reference strains of potential sulphur metabolising bacteria (common in New Zealand) were
tested in this trial: Citrobacter freundii NZRM 982 , Proteus vulgaris NZRM 67, Morganella morganii NZRM
65, Clostridium perfringens NZRM 20, Klebsiella pneumoniae spp. pneumoniae NZRM 482, Klebsiella ozonae
2104, Enterobacter aerogenes NZRM 798, Escherichia coli NZRM 916, Salmonella typhimurium NZRM 1138,
Yersinia enterocolitica NZRM 2603, Campylobacter jejuni NZRM 2397, Acinetobacter lwoffii NZRM 2581,
Aeromonas hydrophila NZRM 804 and Pseudomonas aeruginosa NZRM 918. All bacterial reference strains
were purchased from ESR, Wellington.
For the preparation of the inoculum, most of the bacteria were inoculated onto Tryptose Soy Agar (TSA).
Campylobacter jejuni was streaked onto Campylobacter Isolation Blood-free Agar and Clostridium perfringens
onto Columbia Blood Agar and incubated under microaeophilic (Campylobacter) and anaerobic (Clostridium)
conditions. The TSA agar plates were incubated for 24 hours at 35 ± 0.50C aerobically. Next, the cells were
harvested with 0.1% peptone saline diluent and then diluted with the same liquid. This procedure provided
inocula of approximately 11-30 cfu/100mL (low inoculum level) and 100-210 cfu/100mL (high inoculum level)
for spiking of the 100 mL aliquots of sterile distilled water.
All spiked samples were processed according to the same procedure as naturally contaminated samples.
HACH PathoScreen® Medium was used as a reference method for all spiked samples tested. The same spike
levels were used for inoculation of the HACH® medium bottles. They were processed according to
manufacturer’s instruction, i.e. incubated at 300C for 48 hours.
PathoScreen® Medium is commercially produced by HACH®. Its composition is the same as the H2S paper strip
test prepared for this trial. It is dehydrated, sterilised, and packaged in powder pillows, which are added to the
100mL aliquots of water samples.
In addition to reference strain testing, a small spiking trial was conducted on 20 bacteria strains, which were
isolated from positive samples. These spiked samples were processed according to the same procedure as for the
Another spiking trial was conducted to establish the minimum number of organisms that the strip test was
capable of detecting. Known concentrations (3-120cfu/100mL) of the H2S producers: Citrobacter freundii,
Proteus vulgaris and Salmonella typhimurium were added to samples used in the H2S strip test. These spiked
samples were processed according to the same procedure as for reference strains.
7.3 RESULTS AND DISCUSSION
A spiking study was conducted to determine the minimum number of cells of hydrogen sulphide-producing
bacteria required to create a positive reaction in 100 mL volumes of H2S medium. The inocula of 3-5 cells of
Salmonella typhimurium, Citrobacter freundii and Proteus vulgaris produced a positive reaction in 3 out of 3
replicates by 24 hours at 30-35 0C.
The H2S medium did not display tendency to give false positive reactions when inoculated with cultures not
typically regarded as H2S producers. When inoculated with low and high levels of E .coli, Enterobacter
aerogenes, Klebsiella pneumoniae, Acinetobacter lwoffii, Pseudomonas aeruginosa or Yersinia enterocolitica,
no blackening of the medium occurred during incubation for 48 hours.
A very good agreement (98%) was observed for the Watercare Laboratory Services’ H2S medium and HACH®
PathoScreen H2S medium both for samples spiked with reference cultures and for bacterial isolates from
naturally contaminated samples.
8 STUDY 6: H2S PAPER STRIP KIT PRODUCTION TRIAL
This trial was to determine the cheapest and easiest way of producing the H2S paper strip test kit. The following
production issues were considered in this trial: type of the paper strip, preparation and dispensing of medium
concentrate, practical ways of drying of paper strips saturated with the H2S medium, sterilising and packaging of
dried H2S paper strips, quality controls including sterility, and performance tests. Five types of paper strips
ranging from paper tissues to microbiological filter pads were assessed for their suitability for H2S medium
concentrate saturation, drying and packaging.
8.2 EXPERIMENTAL PROCEDURE
The following issues were considered during the H2S paper strip production trial:
• type of paper used for saturation with the H2S medium concentrate
• preparation of medium concentrate and its dispensing
• drying of saturated paper strips at different temperatures
• packaging, sterilising and storage of produced paper strips
• costs of production of the H2S paper strips in the lab in comparison to commercially produced kits.
Five types of paper strips including ordinary paper towels, filter papers and microbiological absorbent pads (1.5
mm thick - Sartorius® and 0.8 mm thick – Millipore®) were used for saturation with the H2S medium
concentrate. The paper strips saturated with the medium were placed onto trays and dried at room temperature,
at 500C or at 700C.
After drying they were placed separately into the small paper bags and sterilised in an autoclave. They were
then stored at room temperature until used.
All types of paper strips worked well but the most convenient to use were 47 mm diameter thick (1.5 mm)
microbiological absorbent pads from Sartorius®. These pads were chosen and used for the whole project. They
were easy to saturate with the medium and only one pad was needed per 100mL samples. The blackenings of
these pads were brighter and easier to read then on other paper strips especially after the 24 hour incubation
period. The most convenient method of drying was to leave the saturated paper strips on the bench overnight at
room temperature. Next morning the strips were ready for packing and autoclaving. The sterilised saturated
thick pads were stored at room temperature to assess their performance after different storage times. Every two
months, 20 H2S paper strips were used for testing of water samples inoculated with low level (5-20cfu) of H2S-
producing bacteria Salmonella and Citrobacter freundii. The performance of H2S paper strips prepared in the
lab was excellent. Even after 8 months, a 100% of them gave positive results and there were no problems with
The estimated paper strip production cost in the lab was low, below $1.5 per sample. The costs included
materials and labour but not the cost of preparing or buying the sterile bottle. The sterile plastic containers
(120mL) can be purchased from local laboratory suppliers for 50-60 cents each.
8.3 RESULTS AND DISCUSSION
At the moment, there are at least four brands of the H2S kit available on the market. The most often used are two
modified versions of the product: HACH®’ PathoScreen Medium (HACH®, 2002) and LTEK® Bacto-H2S kit
(LTEK®, 2003). In this modified version, instead of absorbing the media on a strip of paper the medium is
contained in a sachet (HACH®) or a hand-breakable glass ampoule, sealed and sterilised. The manufacturers
guarantee at least one year shelf life for the product when stored in a cool, dry place. Both kits are very easy to
use. The HACH® medium used in this trial worked very well. Commercial kits are usually available in two
versions: presence/absence per 100mL kit and MPN per 20mL kit.
In New Zealand, with the fluctuating currency exchange and very low usage of these products, they are more
expensive than the good quality, reliable, rapid test kits such as Colilert® or Readycult® that are commonly used
for testing of drinking water supplies for coliforms and E .coli.
To compete with those easy to use products a reputable local supplier that would guarantee a good quality
product should produce the H2S paper strip kit.
Data summarised above indicate that the H2S paper strip test and coliform MPN Colilert® test were equally
effective in detection of bacterial contamination in water samples.
In particular, an excellent agreement (97.3%) was found between the H2S test and E. coli MPN Colilert® method.
The results of this study indicate that for the H2S test the incubation temperature in the range of 25-350C and the
incubation period of 48 hours are critical. New Zealand ambient temperatures are often lower than 200C so an
incubator is essential to conduct this test.
The method does not detect viruses. However, The H2S test detects bacteria other than coliforms that are
associated with faecal contamination, including Clostridium perfringens and Salmonella.
The H2S test is a sensitive, simple and inexpensive procedure for screening of water supplies for potential faecal
contamination. However, in New Zealand conditions, it does require incubation at a controlled temperature for
48 hours to maximise performance. This means that a controlled temperature incubator with the associated
temperature calibration and monitoring procedures is required. It also means that results would not be available
until 30 hours later than those obtained by Colilert® 18 hour or 24 hours later than those obtained if the
alternative Colilert® 24 hour test was used.
The cost of the Colilert® media is more expensive than preparation of H2S strips and it is necessary to purchase
a uv light to read E. coli. However, there are other enzyme substrate tests which are comparable in price or
perhaps even cheaper than preparing and marketing H2S strips. The advantage of these tests is that a confirmed
result from an internationally accepted method is obtained 24 hours earlier than would be obtained from the H2S
The authors wish to acknowledge the New Zealand Ministry of Health for funding this research, Dr. Michael
Taylor of NZ MoH for initiation of this project, Prof. Michael Sobsey for valuable comments at the initial stage
of the project, Ms Lynette Ronberg from Watercare Services ltd and Prof. Andrew Ball from ESR ,Christchurch
for expert review of the research results. Thanks are also due to Dr Gail Greening from ESR, Wellington for
Norovirus and Hepatitis A virus analysis.
APHA (1998) Standard Methods for the Examination of Water and Wastewater Enzyme substrate coliform test.
Method 9223, APHA 20th Ed., AWWA.
Castillo, G., Duarte, R., Ruiz, Z., Marucic, M.T., Honorato, B., Mercado, R., Coloma, V., Lorca, V., Martins,
M.T. and Dutka, B.J. (1994) ‘Evaluation of disinfected and untreated drinking water supplies in Chile by
the H2S paper strip test’ Wat. Res. 28, 8: 1765-1770.
Desmarchelier, P., Lew, P., Caique, W., Knight, S., Toodayan, W., Isa, A.R. and Barnes, B. (1992) ‘An
evaluation of the hydrogen sulphide water screening test and coliform counts for water assessment in
rural Malaysia’ Trans. R. Soc. Trop. Med. Hyg. 86: 448-450.
Dutka, B.J. and El-Shaarawi, A.H. (1990) ‘Use of a simple inexpensive microbial water quality test: results of
three continent, eight country research project’ IDRC, Report IDRC-MR247e, Jan 2000
Forget, G. (1994) Community-based water quality monitoring for remote communities. IDRC
Gawthorne, T., Gibbs, R.A., Mathew, K. and Ho, G.E. (1996) ‘H2S papers as presumptive tests for Salmonella
contamination in tropical drinking water’ Wat. Sci. Tech., 34, 7-8: 187-194.
Genthe, B. and Franck, M. (2001) ‘A field test for assessing the microbial quality of water: an H2S strip test’
Wat. Sci. Tech., 36, 8:183-190.
Grant, M.A. and Ziel, C.A. (1996) ‘Evaluation of simple screening test for faecal pollution in water’ J.Water
SRT-Aqua, 45, 1:13-18.
IDRC (2001) Testing the waters: Split Lake Study
HACH’s Analytical Procedures (2002) Bacteria: Hydrogen Sulphide Producing. Method 10032, 1-12.
Jangi, M.S., Leong, L.C. and Ho, P.Y.C. (2000) ‘Development of a simple test for the bacteriological quality of
drinking water and water classification’ Malaysian Centre File:3-p-83-0317-02, IDRC
Kaspar, P., Guillen, I., Rivelli, D., Valasquez, G., De Kaspar, H.M., Pozzoli ,L., Numez ,C. and Zoulek, G.
(1992) ‘Evaluation of the simple screening test for the quality of drinking water systems’ Tropical
Medicine and Parasitology, 43, 2: 124-127.
Kromoredjo, P. and Fujioka, R. (1991) ‘Evaluating three simple methods to assess microbial quality of drinking
water in Indonesia’ Environ. Tox. and Wat. Qual. J., 6: 259-270.
LTEK’s Analytical Procedures (2003) Bacto H2S kit.
Manja, K.S. (2001) Report of R&D study on H2S test for drinking water. Rajv Gandhi National Drinking Water
Mission, New Delhi – 110 003, 92pp.
Manja, K.S., Maurya, M.S. and Rao, K.M. (1982) ‘A simple field test for the detection of faecal pollution in
drinking water’ Bull WHO 60:797-801.
Martins, M.T., Castillo, G. and Dutka, D.J. (1997) ‘Evaluation of drinking water treatment plant efficiency in
microorganisms removal by the coliphage, total coliform and H2S paper strip tests’ Wat. Sci.Tech., 35,
MOH (2000) Drinking-water standards for New Zealand 2000. MOH:Wellington
Nagaraju, D. and Nastri, J.C.V. (1999) ‘Confirmed faecal pollution to bore well waters of Mysore city’
Environmental Geology, 38, 4: 322-326.
Pillai, J., Mathew, K., Gibbs, R. and Ho, G.E. (1999) ‘H2S paper strip method – a bacteriological test for faecal
coliforms in drinking water at various temperatures’ Wat. Sci. Tech., 40, 2: 85-90.
RADG Projects (2003) Bacteriological water test method: the Hydrogen Sulphide method (H2S method). RADG.
Ratto, A., Dutka ,B.J., Vega, C., Lopez, C. and El-Shaarawi, A. (1989) ‘Potable water safety assessed by
coliphage and bacterial tests’ Water Research, 23: 253-255.
Rijal, G.K., Fujioka R.S. and Ziel, C.A. (2000) E’valuation of the hydrogen sulphide bacteria test: A simple test
to determine the hygienic quality of drinking water’ Abstracts of the General Meeting of the American
Society of Microbiology, Abstract Q354. page 622. Amer. Soc. Microbiol., Washington, DC.
Sobsey, M.D. and Pfaender, F.K. (2002) ‘Evaluation of the H2S Method for detection of faecal contamination of
drinking water’ Water and Sanitation, WHO/Sde/WSH/02.08, Geneva.
Venkobachar, C., Kumar, D., Talreja, K., Kumar, A. and Iyengar, L. (1994) ‘Assessment of bacteriological water
quality using a modified H2S strip test’ J Water SRT-Aqua, 43, 6: 311-314.