North Carolina; Evaluation of the H2S Method for Detection of Fecal Contamination of Drinking Water

 

WHO/SDE/WSH/02.08 Distr.: Limited English only

Evaluation of the H 2S Method for  Detection of Fecal Contamination of  Drinking Water 

Water, Sanitation and Health Department of Protection and the Human Environment World Health Organization Geneva

 

© World Health Organization 2002

The illustration of the cover page is extracted from Rescue Mission: Planet Earth, International 1994; used by permission

©  Peace Child

This document is not issued to the general public and all rights are reserved by the World Health Organization. The document may not be reviewed, abstracted, quoted, reproduced or translated, in part or in whole, without the prior written permission of WHO. No part of this document may be stored in a retrieval system or transmitted in any form or by any means – electronic, mechanical or other without the prior written permission of WHO. The views expressed in documents by named authors are solely the responsibility of those authors.

 

WHO/SDE/WSH/02.08 Distr.: Limited English only

Evaluation of the H 2S Method for  Detection of Fecal Contamination of  Drinking Water 

Prepared by

Mark D. Sobsey and Frederic K. Pfaender  Department of Environmental Sciences and Engineering, School of Public Health University of North Carolina, Chapel Hill, NC 27599

 

TABLE OF CONTENTS  

Page

Foreword ..................................................................................................................................................... i 1.

Introduction.... Introduction....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ .... 1

2.

Feca Fecall In Indi dica cator tor De Dete tect ctio ion n and and Testi Testing ng iin n Dr Drin inki king ng Wate Water: r: C Con onte text xt,, Pu Purp rpos oses, es, Ne Need edss and Criteria ................................................................................................................................................. 2 2.1 Drin Drinkin king g Wat Water er Safety Safety P Plans lans and Detec Detectin ting g fecal fecal indic indicato ators rs in drinki drinking ng water water ... ..... .... ..... ..... ..... ..... ..... ..... .... .. 2 2.2 Purp Purposes oses and Needs Needs for Detect Detecting ing Indica Indicator torss o off Feca Fecall C Conta ontamin minati ation on in Drinkin Drinking g Water............................................................................................................................................... 2 2.3 Crit Criteri eriaa for detec detectin ting g fecal fecal contami contaminat nation ion of drink drinking ing water water using using indic indicato ators rs .. ..... ..... .... ..... ..... ..... ..... ..... ..... .... .. 3 2.4 Mic Microb robiol iologi ogical cal presen presencece-abse absence nce tests tests aand nd tthei heirr use use iin n dete detecti cting ng and quanti quantifyi fying ng fecal contamination........................................................................................................................ 4 2.5 Adva Advanta ntages ges,, disa disadva dvanta ntages ges and limita limitatio tions ns of tes tests ts for bacteri bacterial al indica indicator torss o off ffeca ecall contamination................................................................................................................................. 5 2.6 Need Needss for for and benefit benefitss of of alter alternat native ive test testss tto o dete detect ct fec fecal al contami contaminat nation ion of dri drinkin nking g water................................................................................................................................................ 5

3.

Basis a an nd Hi Historical De Development H2 S Tests .............................................................. ............................... .......................................................... ........................... 7 3.1 What H2S Tests Measure and How They Measure It ............... ...................... ............... ............... ............... ............... .............. .............. ....... 7 3.2 Hydrogen Hydrogen Sulfide Sulfide and the the Biogeoch Biogeochemist emistry ry of Sulfur......... Sulfur............. ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ .... 8

3.3 3.4 3.4 3.5 3.6 3.7 3.8 3.9

Su Sullfate rreeducing b baacteria aan nd H2 S tests ....................................................................................... 14 Other Other possib possible le limi limita tati tion onss or sourc sources es o off m misi isint nter erpr pret etat atio ion n in th thee H2S test ................................ 15 H2S Test Procedures: Media, Formats and Test Conditions ............... ....................... ............... .............. ............... ............... ......... .. 16 Comparison of H2S Tests to Other Tests for Detecting D etecting Fecal Contamination of Water .......... .......... 20 Determination if H2S Tests Meet the Criteria of an Ideal or Preferred Indicator of  Fecal Contamination ..................................................................................................................... 25 Po Pottenti ntial mo modifications o off the the H2S test to improve specificity for H2S-producing  bacteria of fecal origin........ origin............ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ....... ... 27 Costs of H2S Tests ......................................................................................................................... 27

4.

Summary Summary and Conclusions Conclusions .......... ............. ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ....... ....... ........ ........ ....... ..... .. 29 29

5.

Literature Literature Cited .... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ........ ........ ....... ....... ........ ....... ....... ....... ... 32

Tables

Table 1 Table 2 Tablee 3 Tabl Table Table 4 Table 5

Microbial sources of hydrogen sulfide in water and other other envi ronmental media.............. media...................... ............... ............... ........ 9 Microorganisms capable of producing hydrogen sulfide ............ ................... ............... ............... .............. ............... ............... ............... ............... .......... ... 12 Comparisons Comparisons of Media, Media, Materials Materials and Methods Methods for tthe he H2S Test in Different Reported Studies....... Studies ........... .... 17 Com Compar pariso ison n of H2S and Standard Bacteriological Tests Applied to Various Samples ................. ........................ .......... ... 21 Fecal Indicator Criteria for Fecal Contamination and the Extent to Which they are Met in Reported Repor ted Studi Studies es of H2S Tests .............................. ............................................................. .............................................................. ..................................................... ...................... 26

 

FOREWORD Around 2.2 million die of basic hygiene related diseases, like diarrhoea, every year. The great majority are children in developing countries. Interventions in hygiene, sanitation and water supply make proven contributors to controlling this disease  burden. For decades, deca des, universal uni versal access to safe sa fe water and sanitation has been b een promoted p romoted as an essential step in reducing this preventable disease burden  Nevertheless the target of “universal access” to improved water sources and basic sanitation remains elusive. The “Millenium Declaration” established the lesser but still ambitious goal of halving the proportion of people without access to safe water   by 2015. The provision of drinking water of acceptable microbiological quality and low infectious disease risk requires a number of essential elements within a Water Safety Plan. Within any water safety plan emphasis is placed on controlling and detecting fecal contamination of drinking water and its sources. Traditionally, this measure of  fecal contamination has been a bacterium or group of bacteria considered indicative of  fecal contamination. The measurement of such indicator bacteria of fecal contamination requires trained analysts, media and other supporting materials and facilities available only in a microbiology laboratory or the use of a water  microbiology field analysis kit. Lack of access to laboratories or field analysis kits is an obstacle to the provision of  microbiologically safe drinking water to many communities and people worldwide. In an effort to overcome this problem, a number of alternative indicators and tests to detect fecal contamination of drinking water have been proposed and developed. Some of these proposed fecal indicators and their tests are simple, low cost and do not require a microbiology laboratory or bacteriological field test kit. Some of these simple, low cost fecal indicator tests have come into use in actual drinking water  supply practice. Prominent among these is the so-called hydrogen sulfide sulfid e or H2S test, which is intended to detect or quantify hydrogen sulfide-producing bacteria, considered to be associated with fecal contamination. The purpose of this report is to review the basis of the hydrogen sulfide test as a measure of fecal contamination of drinking water and the available scientific and empirical evidence for and against the test as a valid, useful and reliable measure of  fecal contamination and drinking water quality. The report addresses the fundamental microbiological considerations of the test, including its chemical and biochemical  basis, what organisms it detects and how it detects and quantifies them and the reported experiences with its practical application to assessing water quality. In developing this report many sources of data and supporting information were generously provided by developers and users of the test and others who also have attempted to modify, improve, validate and apply it. We are grateful to these many individual and organizations for their assistance. In particular, we gratefully thank the following for providing reports and other supporting information: Rajiv Gandhi National Drinking Water Mission, Ministry of Rural Development, Department of Drinking Water Supply, Government of India UNICEF, New Delhi

6

 

Ms. Bettina Genthe, Division of Water Environment and Forestry, CSIR, Stellenbosch, South Africa This report has been prepared as part of a programme of activity towards the updating of WHO’s Guidelines for Drinking-water Quality. Following a process of  development and review it is released in draft form. This document represents “work in progress” and further information concerning the H2S test and experience with its application would be welcome. Such information should be forwarded to: Dr Jamie Jamie Bartram Bartr am Coordinator  Water, Sanitation and Health Programme World Health Organization 20 avenue Appia 1211 Geneva 27 Switzerland

7

 

1. INTRODUCTIO INTRODUCTION N AND BACKGROUND This report critically reviews the scientific basis, validity, available data and other  evidence for and against H 2S tests as measures or indicators of fecal contamination in drinking water. The test was originally developed to detect in a volume of water the  production of H2S by enteric bacteria associated with fecal contamination by the formation of a black precipitate from the reaction of the H2S with iron in the medium. This relatively simple, low cost test has been studied, modified in various ways, tested and used to some extent in many parts of the world as an indicator of fecal contamination of drinking water. Since the initial report of the H2S test 20 years ago (Manja et al., 1982), many versions or modifications of it have been described in the published literature. Hence, there are now many versions of the H 2S test and these differ in medium composition, the preparation of the medium and supporting materials, test format and sample volumes, incubation time, incubation temperature and scoring of results. The test is not standardized worldwide and only some versions of it have been subjected to collaborative testing or comparison with other bacteriological tests for fecal contamination. The plethora of different H2S test versions and the variety of ways they have been evaluated in laboratory and filed studies make it difficult to compare them. The essential criteria of any test to detect and quantify fecal contamination of  drinking water and other waters are used as the basis for evaluating the validity and  performance of various v arious versions of the H2S test and the quality of the data available for evaluation and validation purposes.

8

 

2.

FECAL FECAL INDIC INDICAT ATOR OR DETEC DETECTI TION ON AND AND TESTI TESTING NG IN DRIN DRINKI KING NG WATER: CONTEXT, PURPOSES, NEEDS AND CRITERIA

2.1 Drinking Drinking Wate Waterr Safety Safety Plans Plans and and Detect Detecting ing fecal fecal indica indicator torss in drinkin drinking g water

An essential goal for the provision of safe drinking water is that it be essentially free of (at low risk of containing) disease-causing microorganisms.   Since the beginning of  the 20th   century, the detection of fecal indicator bacteria in drinking water has been used as the basis of criteria, guidelines and standards for acceptable limits of fecal contamination and as the basis for judging or predicting the possible presence or  absence of pathogenic (disease-causing) microorganisms. The WHO WHO Guidelines for  Drinking-water Quality (GDWQ) and many other authorities continue to support the use of bacterial indicator levels and their measurement as a basis for judging and verifying drinking water quality. However, such fecal indicator analysis of drinking water as a measure of end-product quality and determinant of microbial disease risk is only one of many measures and activities in an overall system for providing safe drinking water. In the updating of the WHO GDWQ, the goal of providing safe safe drinking water will be promoted through the development and use of a Water Safety Plan. This plan includes risk assessment, water system assessment and process control that examines all aspects of drinking water from its source, through treatment and distribution (or collection and storage) to the consumer. uses a management  plan that incorporates Hazard Analysis-Critical Control Points It(HACCP). In such a  plan the measurement of fecal indicator presence or levels in water is only one of  several management tools and not always among the most critical ones for process control or overall collection, production and delivery of drinking water of acceptable microbial quality. Nevertheless, measures or indicators of water quality, particularly those measuring or indicating fecal contamination, are useful if not essential tools in the provision of safe drinking water. The ability to easily, rapidly and affordably detect fecal contamination in drinking water is still a desirable goal and worthy endeavor in the overall effort to provide microbiologically safe (low risk) drinking water. 2.2

Purposes Purposes and Needs Needs for Detec Detecting ting Indicato Indicators rs of Fecal Fecal Contamin Contaminatio ation n in Drinking Water

Determining the microbial quality of drinking water by measuring the presence, absence or concentrations of indicator bacteria continues to be widely practiced worldwide to: (1) meet water quality standards and guidelines, (2) to determine source water quality, treatment system efficacy and distribution system integrity, and (3) to inform Water Safety Plans, risk assessments and management systems. In some countries and regions and for international commerce (commercial bottled water) and transportation (airplane, rail and other travel conveyances), such analysis of drinking water may be required by law or governance. In addition to these purposes and needs, measuring the microbial quality of water for   presence of fecal contamination can be and is now being done for other beneficial  purposes. One such purpose is for f or community involvement and empowerment e mpowerment in the  provision, andaremonitoring of drinking water, local including its sources treatment. management Great efforts being made to encourage participation in and the

9

 

 provision of safe drinking water and in the oversight or o r monitoring of its provision by other responsible parties (governments, privatized water companies, water supply contractors, water vendors, etc). The ability to test drinking water for fecal contamination is a powerful and empowering tool for these purposes. Another useful purpose for measuring the microbial quality of water is educational. Teaching people about the microbial quality of water and the fundamentals of germ theory within the context of education and outreach programs for water, sanitation and hygiene at the individual, household, community and regional levels is a continuing and long-term goal in the global health initiative. In delivering these educational messages the availability of simple, practical, accessible and affordable tests for fecal contamination of drinking water are extremely useful and potentially  powerful tools. In some situations the best tests to accomplish these goals are those that are the simplest to use, understand, visualize and interpret. This is because such tests can be widely disseminated both directly by the primary educators and then subsequently via communications within households, families, schools and communities and by other means (educational materials such as leaflets, signs and labels). For these purposes the H2S tests and other simple and affordable tests have great value and even greater potential use for drinking water supply management and health education in the water and sanitation sectors. However, needs andIfpurposes served only testsprovide are reliable and  provide themeeting correctthese information. they areisunreliable andif ifthethey incorrect information (false positives and false negatives), their potential uses and benefits will  be undermined. undermined . Therefore, H2S tests and other simple tests to detect and quantify fecal contamination of drinking water must be evaluated and judged on the basis of  their reliability and predictability as well as their accessibility, practicality and affordability. 2.3 Criteria Criteria for for detec detecting ting ffeca ecall contamin contaminatio ation n of drinki drinking ng water water using using indicators

For more than 100 years, bacteriological tests have been used to detect fecal contamination of drinking water, other waters and other media, such as wastewater  and foods. During this time, there has been an evolution in the bacterial indicators used and the articulation of the criteria for an ideal or reliable indicator of fecal contamination in drinking water and other waters (Olivieri, 1983; Sinton et al., 1998). The current criteria of an ideal or preferred indicator of fecal contamination have been defined and stated by WHO and other authorities. According to these authorities the essential criteria of a fecal indicator are the following (WHO, 2002): • 

The indicator should be absent in unpolluted water and present when the source of pathogenic microorganisms of concern (fecal contamination) is present. •  The indicator should be present in greater numbers than the pathogenic microorganisms. •  The indicator should respond to natural environmental conditions and water  treatment processes in a manner similar to the pathogens of concern. •  The indicator should be easy to isolate, identify and enumerate. • 

The test should be inexpensive thereby permitting numerous analyses to be taken.

10

 

• 

The indicator should not be a pathogenic microorganism (to minimise the health risk to analysts).

Authorities have identified at least one additional criterion for indicators of fecal contamination (as distinct from indicators of other sources of microbial contamination) (Olivieri, 1983). This fecal indicator criterion is: • 

The indicator should not multiply in the environment

The rationale for this last criterion is that the presence and concentration of fecal indicators should be in proportion to the level of fecal contamination. Hence, microbial proliferation in the environment could result in the microbe being present at high concentrations when no fecal contamination (and its pathogens) or very low levels of fecal contamination are actually present. Therefore, any indicator of fecal contamination of drinking water and its sources is  best judged according to these essential criteria and perhaps other oth er criteria that may be relevant for or specific to a particular indicator and type of test for it. 2.4

Microbio Microbiologi logical cal prese presence nce-abs -absenc encee tests tests and and their their use in in detect detecting ing and and quantifying fecal contamination

In addition toof water the criteria described above, (P) some microbial tests fecal contamination are based on the presence or absence (A) of the for microbial indicator in a specified volume of water, a so-called P-A test. In the application of PA tests to detecting bacterial indicators of fecal contamination in drinking water the tested sample volume is 100 ml. According to some standards and guidelines, the fecal indicator is expected or required to be absent (A) in all of (zero tolerance) or  most of (e.g., 95%) the 100-ml sample volumes successively tested over time. In other (i.e. non – P/A) formats of fecal indicator analysis of water, the concentration of  the target microbe is determined by using multiple and varying sample volumes, each of which is scored as positive or negative for the test microbe. These data are then used to estimate microbial concentration by a quantal method typically the Most Probable Number (a maximum likelihood analysis method). Alternatively, the water  is analyzed for the fecal indicator microbe or microbe group by an enumerative method in which the concentration of bacteria per unit volume of water can be expressed as colony forming units (CFU) per unit volume. The use of the P-A concept and P-A tests for fecal indicator bacteria, primarily coliforms, fecal coliforms and  E. coli , has a history of development that goes back  more than four decades (Clark, 1968). Considerable effort in the form of expert analysis and judgement went into the development and implementation of P-A tests for these microbes in drinking water. Much of this effort included consideration of  the wealth of available historical data on the occurrence of these indicator bacteria in municipal drinking water, based on the frequency of positive results (fecal indicator   presence) in 100-ml volumes of drinking water and the acceptability (or risk) of  drinking waters based on these observed frequencies. These analyses led to current guidelines and standards for the microbial quality of drinking water based on positive P-A test results. There appears to be no corresponding analyses and expert judgement 2S that went thecompared development of H2S test a P-A fecal test. indicator Instead, results of P-A tests haveinto been to results for as various microbe testsHto

11

 

determine the extent to which the results give the same outcome when scored as  positive or negative results. In the initial report of the H2S test by Manja et al. (1982) the test was applied to various drinking water samples of several cities in India. Water samples containing 10 or more coliform bacteria by MPN were subjected to the new H2S test using a 20-ml sample volume in a P-A format. On this basis, positive H2S tests were considered unsatisfactory as drinking water because they contained 10 or more total coliforms and were positive for H2S. Since then, others have compared H2S tests to tests for fecal indictor bacteria using either these or other criteria. The use of this comparative approach in determining the validity of the H2S test has never been subject to review of its scientific merit and validity. Considering the differences in the target bacteria being detected, absent any consideration of pathogen  presence in water, and without formal efforts to determine how well they fulfill the essential criteria of an ideal or acceptable indicator of fecal contamination, the validity of H2S tests, the meaning and reliability of interpretation of their results, and their ability to predict microbial health risks is a matter of concern. Because the basis of the WHO GDWQ are now becoming risk-based, the absence of a microbial risk   basiss for H2S tests raises concerns about their validity and interpretation in judging  basi the acceptability of drinking water quality. Hence, the application of P-A criteria or  various quantitative criteria to H2S tests as fecal indicator tests is an important but still unresolved issue. There are no fully articulated public health or risk-based criteria for  specifying the volume of water to be tested in a P-A test format or the acceptable (or  unacceptable) concentrations of H2S bacteria as measured in a quantitative test. In this report, the use of P-A tests and criteria for H 2S tests will be considered in terms of  the validity and reliability of this test format to detect and quality fecal contamination in drinking water and its sources. 2.5

Advantag Advantages, es, disadva disadvantag ntages es and limitat limitations ions of of tests tests for for bacter bacterial ial indicat indicators ors of fecal contamination

Fecal indicator bacteria generally are present in much higher numbers than the frank  waterborne pathogens of fecal origin, they are easier to detect relatively rapidly by standard culture methods and the costs of analysis are far less than the costs of  analyses for pathogens. A number of well-developed and extensively tested methods are widely available for the detection of various fecal indicator bacteria such as thermotolerant (fecal) coliforms and  E. coli   in drinking water and its sources. The historical basis, uses and interpretations of these tests are described in detail elsewhere (WHO, 2002). Suffice it to say that testing for these bacterial indicators of  fecal contamination of drinking water is still encouraged and widely used worldwide. Despite their advantages and value, the use of bacterial indicators of fecal contamination and the methods for them have limitations. In addition, because of the  previously noted lack of accessibility accessib ility or availability in many settings in many parts of  the world, it has become apparent that there are other limitations to and problems with the use of the usual bacterial indicators of fecal contamination of drinking water  (thermotolerant coliforms and  E. coli ). It has been well documented that waters considered bacteriologically safe (less than 1 bacterial fecal indicator per 100 ml), can contain sufficient pathogenic enteric viruses and protozoans to cause disease outbreaks andmicrobes, Noton, 1976; and Gunn, 1979;ofMacKenzie al., 1994). Other fecal(Berry indicator such Craun as enterococci, spores Clostridiumetperfringens

12

 

and coliphages, can be detected in drinking water when the usual coliform bacteria (total or thermotolerant) or  E. coli  are not detectable. Furthermore, there is some evidence that coliforms possibly including  E. coli c oli can proliferate in tropical and subtropical waters. Warmer water temperatures may contribute to the growth of  coliforms, thermotolerant coliforms and  E. coli  and the greater survival of some enteric bacteria, notably Salmonella,  compared to coliforms (Hazen, 1988; Iverson and Fleay, Fleay, 1991; Jimenez et al., 1989; Townsend, 1992). For these reasons, coliforms, thermotolerant coliforms and  E. coli  are not ideal indicators of fecal contamination and alternative indicators of fecal contamination continue to be considered, evaluated and applied (Sinton, 1998; Sobsey, 2001; WHO 2002). 2.6

Needs Needs for and benefit benefitss of alter alternativ nativee tests tests to detect detect fecal fecal contam contaminat ination ion of  drinking water

The requirements for laboratory resources or field analysis kits for standard  bacteriological tests for fecal contamination of drinking water are major barriers to their accessibility in many parts of the world. The need for sterilized bacteriological materials (media, sample bottles, sterile diluent, culture tubes, bottle or plates, membrane filters, pipettes or other volumetric dispensing devices, etc), controlled temperature incubators, the required use of aseptic technique by trained individuals, and relatively high costs make it difficult, impractical or impossible to perform these tests in many places. The resources and infrastructure are simply not available to allow for routine bacteriological testing of drinking water using the standardized methods for fecal indicator bacteria analysis. The lack of availability of standard bacteriological tests for drinking water quality highlights the great need for a rapid, simple, inexpensive test for the microbial quality of drinking water. This need is especially great for small community and household water supplies that lack access to and can not afford conventional bacteriological testing of drinking water. On-site testing using portable equipment and use of  simplified tests, such as the H2S tests, may both contribute to overcoming these constraints. H2S tests deserve evaluation as accessible alternatives to conventional bacteriological tests for fecal contamination of drinking water. Therefore, the potential merits and  beneficial uses of H2S tests deserve consideration, as does the determination of their  reliability and predictability in detecting fecal contamination of drinking water. Key issues to be addressed are whether H2S tests are sufficiently reliable and adequately developed as tests of fecal contamination of drinking water to be recommended for  widespread and routine use, and if, so what caveats and cautions should be applied and under what conditions. This report addresses the key aspects of H2S tests as indicators of fecal contamination of drinking water. The aspects to be addressed are as follows: • 

what organisms the tests actually measure; •  the extent to which test measurements are or are not indicative of fecal contamination; and under what conditions; • 

the basis for and likelihood of false positive and negative results;

13

 

• 

the current state of the methodology with respect to reliability, uniformity,  practicality, availability and cost and •  the extent to which the tests fulfill the ideal criteria of an indicator of fecal contamination; and recommendations for future actions and their directions.

14

 

3. BASIS AND HISTORICAL DEVELOPMENT H2S TESTS

In 1982 Manja et al. reported the development of a simple method for detecting evidence of fecal contamination in drinking water. The test was intended to meet the "…need for a simple, reliable field test for use by village public health workers" to detect fecal contamination in drinking water. They observed that the presence of  coliform bacteria in drinking water was consistently associated with organisms that  produce hydrogen hydro gen sulfide (H2S). The test is based on the readily observable formation of an iron sulfide precipitate on a paper strip (or in the water sample liquid) in a bottle or test tube, as a result of the reaction of H 2S with iron. The test is intended to detect  bacteria associated with fecal contamination due to the activity of these microorganisms in reducing organic sulfur to the sulfide oxidation state (as H2S gas) which then reacts rapidly with iron to form a black, iron sulfide precipitate (Allen and Geldreich, 1975). The advantage of the method is its simplicity, low cost and ability to be performed in the absence of a typical microbiology laboratory or field laboratory test kit. Tubes or other containers holding the test materials are prepared in a central laboratory to be used in the field by minimally trained personnel. Over the last two decades, various investigators have tested this method and various modifications of it in different tropic and temperate regions, including Indonesia, Peru, Paraguay, and Chile, Nepal, and South Africa (Ratto et. al., 1989; Kromoredjo and Fujioka, 1991, Kaspar et al., 1992; Castillo et. al., 1994; Venkobachar et al., 1994; Martins et. al., 1997; Rijal and Fujioka, 1998; Genthe and Franck, 1999), and compared it to traditional bacterial indicators of fecal contamination of water. As will  be described describ ed in more detail in a later section sectio n of this report, report , the results of o f these studies generally indicate that the method gives results comparable to the test for traditional  bacterial indicators of fecal contamination contaminatio n and is sometimes superior to these test in detecting fecal contamination, based on other criteria for evidence of fecal contamination. Furthermore, some studies indicate method worked well as a  presumptive test for the detection of Salmonella (Gawthorne et. al., 1996). 3.1

What H2S Tests Measure and How They Measure It

The H2S method also, commonly called the paper strip method, and the various modifications of it do not consistently measure the presence of either total coliform  bacteria, specific groups of fecal bacteria (e.g., fecal coliforms) or a specific fecal  bacterium  bacter ium  E.  (E. coli ). The test is based on measuring bacteria that produce hydrogen sulfide under the test conditions employed. However, some coliform bacteria (e.g., Citrobacter  spp.),   spp.), some other enteric bacteria (e.g., Clostridium perfringens) as well as many other types of bacteria produce H2S. The test measures the production (actually, the presence) of H2S by its reaction with iron to form an insoluble, black   precipitate of iron sulfide. Given the low solubility product of iron sulfide, the test can detect even small amounts of sulfide formation or presence. Any source of H2S in the sample can lead to a positive result. Sulfides also can be formed by abiotic chemical reactions. Many different bacteria, from a variety of habitats, including many of enteric origin, can release sulfide from proteins, amino acids and other  reduced sulfur compounds by reduction reactions. Therefore, there are many possible sources of a positive result in the H2S test.

15

 

In evaluations of the H2S test, several investigators have attempted to identify by speciation, the bacteria present in positive H2S tests. Castillo et al., (1994) (1994) found a large variety of bacteria, primarily various  Enterobacteriaceae   and Clostridium  perfringens , in samples giving positive reactions in the H 2S test:  Enterobacte  Enterobacter, r, clostridia , Klebsiella,  Escherichia, Salmonella, Acinetobacter, Aeromonas,  Morganella. Ratto et al. (1989) found Citrobacter   to be a common organism in  positi  pos itive ve H2S tests. This suggests that while the test organisms many not be all coliforms they are organisms typically associated with the intestinal tracts of warm blooded animals. Because some of these microbes may arise from fecal contamination of non-human origin, the test is not specific for human fecal contamination. In many settings domestic and agricultural animals and humans live in intimate contact and therefore, distinguishing the contributions of human and animal sources to fecal contamination of water is not possible and perhaps not necessary. Because animal fecal contamination also contains a variety of human  pathogens, such contamination c ontamination also poses pos es risks to human health if present presen t in drinking water. Although the intent of H2S tests is to detect bacteria associated with fecal contamination, there is considerable concern that the test also may detect bacteria not associated with fecal contamination and its attendant pathogens. Therefore, an examination of the sources, sinks and transformation of sulfur and the role of  microbes in its cycling is important to understanding the applicability of this test. 3.2

Hydrog Hydrogen en Sulfi Sulfide de and and the Biog Biogeoc eochem hemist istry ry of Sulf Sulfur ur

The sulfur cycle and the role of H2S and bacteria in it.  Sulfur is one of the ten most abundant elements on earth and its various elemental, oxidized and reduced forms is driven by a biogeochemical cycle, the sulfur cycle, involving bacteria and other microbes. The key pathways and constituents of this biogeochemical cycle are shown in Figure 1.

16

 

As shown in Figure 1, hydrogen sulfide is a key compound in the sulfur cycle and one of the most abundant forms of sulfur in the environment. Four fundamental types of  reactions are involved in the sulfur cycle: (a) mineralization or decomposition of  organic sulfur (from living cells or of synthetic origin), (b) microbial assimilation of  simple sulfur compounds into biomass, (c) oxidation of elemental sulfur and inorganic compounds such as sulfides and thiosulfate and (d) reduction of sulfate and other  anions to sulfide. H2S is a direct intermediate in three of these reactions: mineralization, sulfur oxidation and sulfate reduction, all of which can be mediated by various microbes. Of particular interest for H2S production are the reductive sulfur reactions mediated  by various anaerobic bacteria and the decomposition reactions on sulfur amino acids and other forms of sulfur in biomass. As shown in Table 1, elemental sulfur can be anaerobically reduced by bacteria growing on acetate, such as  Desulfuromonas acetoxidans, which occurs in anaerobic sediments rich in sulfide and elemental sulfur. Sulfate reduction occurs in many anaerobic environments by the activity of   Desulfovibrio , other strict anaerobes and by some  Bacillu  Bacilluss and  Pseudomonas species. H2S also is produced by sulfur respiration with molecular hydrogen, and this reaction occurs in surface and subsurface geohydrothermal environments (e.g., hot springs, subsurface and submarine hydrothermal vents, etc.). H 2S also is produced by mineralization or decomposition of amino acids and other organic forms of sulfur in  biomass. This is a widespread wid espread phenomenon phe nomenon in many environments environme nts and is produced produce d by many different kinds of bacteria. Table 1. Microbial sources of hydrogen sulfide in water water and other environmental media Re a c t i o n Reduction of  elemental sulfur  Sulfur respiration with molecular  hydrogen Sulfate reduction

Conditions Anaerobic

Typical Locations Anaerobic se sediments

Anaerobic

Anaerobic

Decomposition of  Anaerobic or  aerobic organic sulfur in  biomass

Submarine hydrothermal vents

Bacteria  Desulfuromonas acetoxidans Thermophilic archaebacteria

Comments Occurs in fresh waters Occurs in fresh waters

Many: wa water logged soils, sediments and

 Desulfovib rio,  Desulfovibrio,  Desulfomonile,  Desulfomo nile,

Occurs in fresh waters

other anaerobic conditions where sulfate is present

 Desulfovibrio,  Desulfovib rio,  Desulfobacter,  Desulfuromonas an d others, including some Bacillus and  Pseudomonas species Many; includes desulfhydration of  cystiene to yield H2S, ammonia and  pyruvic acid

under a wide range of  conditions

Many but wherever  there are dead plants and animals

Occurs in fresh waters under a wide range of  conditions

As indicated above, many bacteria can produce H2S in water and in media (soils, sediments, etc.) in contact with and containing water under a variety of environmental conditions unrelated to the presence of fecal contamination of water. Therefore, it is

17

 

 possible that t hat false positive results re sults for fecal contamination cont amination in H2S tests can come from a number of sources, including the presence of sulfides of non-biological origin and from the activities of microbes of non-fecal origin. Because of these possibilities, care must be taken in interpreting positive results from H2S tests. The water source and its environmental conditions, especially geohydrological conditions, must be considered carefully in order to reliably interpret a positive result as being indicative of fecal contamination. Table 2 represents an attempt to list and summarize the properties of bacteria capable of producing H2S. This list is, which was compiled in part from the 8th   edition Bergey's Manual of Determinative Microbiology, is not comprehensive because hydrogen sulfide production information is not available for all genera or species within genera (Buchanan and Gibbons, Gibbons, 1974). It is also clear that these cultured organisms represent a very small fraction of all the microorganisms that occur in natural waters, soil and in association with plants and animals. They may represent as little as 1% of the total bacterial population. Because it is well known and generally accepted that only a small portion of the bacteria in water and other media have actually been isolated and characterized, it is likely that many other H2S producers exist that have not been described (Altas and Bartha, 1993). Indeed, new H 2S  producers continue to be discovered, characterized and taxonomized (Fukui et al., 1999). The list in Table 2 is based on genera that contain members that can produce hydrogen sulfide. In most cases not all species in a genera have this ability. However, it makes little sense to expand the list to species because there is still great confusion about species names and relationships within genera. Furthermore, there continue to be issues in bacterial taxonomy based on classifications using phenotypic criteria and those using genetic (nucleotide sequence) data. Additionally, only a small fraction of the species in most genera has been described. Despite these limitations in the criteria for listing them, it is clear that many organisms have the potential to give a  positive response in the H2S test.

18

 

Table 2. Microorganisms capable capable of producing hydrogen sulfide Name

Common Source or Habitat

Pathogens

Capable of Giving Positive 1 Test

Water- strict anaerobic

None reported

Unlikely- slow growth

Water, soil Mammal oral saprophytes Mammal oral saprophytes

None reported None reported None reported

Possible Possible Possible

Soil, water

None reported

Unlikely- slow growth

Animals Fres h and salt water Man and animals

Several None reported Several

Unlikely- fastidious Possible Possible

Soil, water Soil, plant tissues Wa t e r

Plant pathogens Plant saprophytes and pathogens None reported

  Halococcus- Archebacteria

Wa t e r

None reported

 

Animal bodies

Many human and animal

Wa t e r

Human and animal parasites

Possible Possible Not possible - require 12%  NaCl Not possible – require 12%  NaCl Unlikely- require CO2 to grow Unlikely- fastidious

Water, human & animal intestines Water, food, animal feces, urine Water, food, animals Soil, water, animal feces Human and animal intestines Widely distributed Plant Animals Beer, fermenting fruit Wa t e r Animals Animal membranes and tissues Human nose and throat Mammal intestines

Possible None reported Many pathogens Possible, none reported Yes Unlikely Plant pathogens Possible None reported Rodent pathogen Many animal pathogens Animal pathogens Possible Animal parasites and pathogens

Likely + Likely + Likely + Likely + Likely + Likely + Unlikely- rarely in water Likely + Possible Possible- H2S rare Unlikely- fastidious Likely + Unlikely- fastidious Unlikely- fastidious

Intestinal, oral cavity animals

Rumen bacteria- unlikely

Unlikely- fastidious

Phototrophic Bacteria Gram negative Rhodospirillum Myxobacteria Gram negative   Flexibacteria   Simonsiella   Alysiella Budding or Appendaged- Gram negative   Hy p o m i c r o b i u m Spirochaetes and Spirals- Gram negative   Treponema   S pi ri l l um   Campylobacter Gram negative Rod and Cocci   Xanthomonas   Agrobacterium   Halobacterium- Archebacteria  

Brucella

  Francisella Gram negative Facultative Rods- Enterics   Edwardsiella   Citrobacter   Salmonella   P rot e us   Yersinia   Klebsiella   Erwinia   Aeromonas   Zymomonas   Flavobacterium   Pasteurella   Actinobacillus   Cardiobacterium   Streptobacillus Gram negative Anaerobic   Bacteroides

 

 

1

 Fastidious means microbes require blood, serum, sterols, etc; slow growth = not able to produce response in the incubation period of the test; no substrates means the needed nutrients are not in the test medium 19

 

   

Fusobacterium Desulfovibrio

Animal mucus membranes Soil, water, sediment

Several None reported

Likely + but rare in water Unlikely- no substrates

   

Desulfomonas Desulfobacter

Human intestinal tract Soil, water, sediment

None reported

Unlikely – fastidious Possible

So oiill,, w S waatteerr,, sseed diim meen ntt Soil, water, sediment Rumen Animal intestines

None reported None reported None reported

Po ossssiib bllee P Possible Unlikely- fastidious Possible

Animals- many tissues Mouth, intestinal track of animals Rumen, sheep intestine

M a ny Possible None reported

Unlikely- fastidious Unlikely- slow growth Unlikely- slow growth

Human skin, membranes, air, dust Animals- mostly humans Animal respiratory and UG tract

M a ny Possible Several

Likely + Likely + Possible

Soil, water, sediment Soil, water

Some None reported

Possible Unlikely- no substrates

Soil, water, fish, animals

Some

Likely +

Cavities of man and animals Soil, water Soil, humans, animals Soil, animals Plants, soil, animal skin Soil Soil Soil- antibiotic producers Soil, animal wastes

Some Both man and animals Some None reported None reported None reported None reported None reported None reported

Possible Unlikely- slow growth Likely + Unlikely- slow growth Unlikely- slow growth Possible Possible Unlikely- slow growth Unlikely- Thermophilic

Parasites of man

Several

Unlikely- fastidious

Deessuullffuorcoom ccounsas    D   Desulfosarcina   Butyrivibrio   Selenomonas Gram negative Cocci   Neisseria   Veillonella   Megasphaera Gram positive Cocci   Staphylococcus   Peptococcus   Peptostreptococcus Gram positive Endospore Formers

  Clostridium   Desulfotomaculum Gram positive Non-Spore Formers   Erysipelothrix Actinomycetes-- Gram positive Actinomycetes   Eubacterium   Actinomycetes   Arachnia   Rothia   Actinoplanes   Planobispora   Dactylosporangium   Streptoverticillium   Thermomonospora Mycoplasmas Gram negative   Mycoplasma

20

 

 

It is highly likely that almost any water will contain a mixture of organisms of both fecal and nonfecal origin that can, under some set of conditions, give a positive result in H 2S tests. There appears to be no reasonable way to preclude all of them from being the sources of false positive reactions. Organisms of fecal origin tend to grow more rapidly than many environmental isolates and are therefore favored under the incubation conditions of the test. The longer incubation times required for the sulfate reducers to be important and for most non-fecal heterotrophs to grow would argue for  care to be exercised to keep the incubation times as short as possible, probably no more than 48 hours. However, even shorter incubation times may not prevent the growth and detection of these  bacteria in H2S tests, as will be further documented below. Additional suggestions are made below for modifications to the H2S media and incubation procedures that may reduce the likelihood of false  positive results from environmental H 2S-forming organisms. 3. 3.3 3

Sulf Sulfat atee red reduc ucin ing g bac bacte teri ria a and and H2S tests

Hydrogen sulfide producing bacteria To adequately evaluate the usefulness of H 2S tests for fecal contamination it is necessary to examine which bacteria are likely to be important because they produce hydrogen sulfide. As shown in T Table able 2, many bacteria are capable of producing hydrogen sulfide from organic materials. Some of these are unique to or strongly associated with fecal contamination and many others are not. A major  group of environmental bacteria producing H2S is the sulfate reducing bacteria group. Sulfate reducing bacteria (SRB) are key players in the global sulfur cycle. They represent a heterogeneous group of Bacteria and Archaea physiologically unified by their ability to perform dissimilatory

sulfate reduction for energy-generating processes. contrast sulfide to assimilatory sulfate reduction use of sulfate as electron acceptor and its reduction toIn hydrogen is restricted to this group.the These  bacteria are a re ubiquitous ubiquito us and occur in a variety of habitats, including incl uding marine and freshwaters and their  sediments, soils, biofilms, microbial mats, intestinal contents, termite guts, walls of "black smokers" and in association with marine worms. Based on 16S rRNA rRNA sequences the SRB can be  phylogenetically divided into five distinct lineages: (1) Gram-negative mesophilic SRB (deltaProteobacteria), (2) Gram-positive sporeforming SRP (Low G+C Gram-positive Bacteria), (3) thermophilic bacterial SRB (Nitrospira pyhlum), (4) thermophilic bacterial SRB (Thermodesulfobacterium group), and (5) thermophilic archaeal SRB (Euryarchaeota). It might be assumed that anaerobic sulfate reducers are unlikely to be able to produce a positive reaction in this test. Sulfate reducers occur in freshwater environments, but they are rarely present in high numbers and usually are associated with the sediments and with geothermal and hydrothermal environments (Widdel, 1988). However, at least some SRBs SRBs are microaerophilic and not strictly anaerobic, and these bacteria are now known to be capable of degrading a wide variety of substrates, including saturated hydrocarbons and a variety of aromatic substrates. If the concentrations of  sulfate in freshwater are very low, there will be little substrate for these bacteria to use (Wetzel, 2001). However, where sulfate concentrations in water are high, such as geothermal environments, sulfate-reducing bacteria could give a positive results in H 2S tests. It might be assumed that because sulfate reducers do not metabolize complex organic compounds, such as those included as substrates in the test medium, instead requiring short chain organic acids, and other products of fermentation, they would not grow and give positive results in H2S tests (Kelly and Smith, 1990, Widdel, 1988). For a positive reaction to occur there would need to be time the system to become highly anaerobic, time for fermentative bacteria to produce these products and then allow for the growth of the sulfate reducers. It might be assumed that these conditions are not likely to be achieved in the incubation times typically employed in H 2S tests (1-2 days). However, it is likely that they could be achieved in longer in incubation times of 5 to 7 days, which have been employed in some versions of H2S tests (Widdel, 1988).

21

 

Studies on non-fecal, environmental bacteria in an H2S test

Experiments were conducted in the laboratory of Mark Sobsey to determine if mixed populations of  sulfate reducing bacteria (SRBs) would give a positive result in an H 2S test. Sixteen samples of SRBs were obtained from the microbial ecology laboratory of Terry Hazen, Lawrence Berkeley Laboratory. The bacteria were soil isolates from a pristine geothermal region in Kanchatka, Russia, where studies are being done to isolate bacteria of potential value for commercial biogeochemical enterprises. These bacteria were grown in SIM medium and then tested using a commercial H2S test, the HACH Pathoscreen Field Kit, which is designed to detect hydrogen sulfide-producing enteric  bacteria. The HACH test cultures were incubated at 29° to 32°, and examined after 24 hours and 48 hours for darkening of the medium or formation formation of a black precipitate. Of the sixteen isolates, four   produced hydrogen sulfide by the HACH H2S assay, which is a false positive rate of 25%. These results suggest that H2S tests may give false positive results for fecal contamination by give positive results for H2S-producing bacteria (primarily sulfate reducing bacteria) unrelated to and not originating from fecal contamination. 3.4

Other Other possibl possiblee limitati limitations ons or or source sourcess of misinterp misinterpreta retation tion in the the H2S test

Another issue to consider in relation to H 2S tests is the use of thiosulfate and cystine in the medium as a possible dechlorinating agent. Thiosulfate, like sulfate, could serve as a source of sulfur for  microbial reduction and H2S production, and cystine can be degraded, releasing H 2S. The same considerations discussed for sulfate reducers also apply to these obligate anaerobes that degrade amino acids ands other carbon compounds containing sulfur. Only some of the research conducted to date on the H2S test has addressed groundwater specifically, and when it has, apparently false positive results have been observed (Kaspar et al., 1992). In ground waters, particularly those contaminated with human or animal wastes, fecal or otherwise, or those containing reduced sulfur from natural or anthropogenic sources, there is a high potential for  anaerobic aquifers and the formation of sulfides by bacteria of non-human or non-animal origin. In many rural areas small-scale industry, animal husbandry, and human dwellings are all contiguous, which offers the potential for sulfide formation from sediment-derived degradation of organic wastes from these sources, only some of which are fecal sources.. The rapid reaction of the iron with sulfide already present in a water sample could produce a darkening in an H 2S test almost immediately upon addition of the sample. For this reason, it is very important that the test procedure include visual checking for a quick or early positive reaction, after perhaps a few minutes to one-hour of  incubation. A positive result so quickly is likely to mean that the sample already contained sulfides. Such a result is not readily interpretable as either positive or negative for fecal contamination  because it is not useable as evidence of microbially mediated H2S activity likely to be associated with fecal bacteria. Because a black precipitate is the experimental end point of the H2S test, there may be concern for  formation of other dark colored metal salts from constituents present in a water sample. The Handbook of Chemistry and Physics identifies relatively few iron salts that are black besides FeS. Some of the oxides of iron can form black precipitates but they are unlikely to form in water samples as a result of microbial activity and should be visible immediately at the onset of the test if present. Other metals can react with sulfide to produce a black or dark precipitate, but such precipitates, if   produced after a time period consistent with microbial activity also would be evidence of H2S  production. Microbially induced asand a source of H2S-producing bacteria in water Another concern with corrosion the H 2S test its susceptibility to detecting organisms of non-fecal origin is microbially-induced corrosion of iron, steel and other metals associated with water sources, 22

 

treatment systems, conveyances and storage facilities. Microbially-induced Microbially-induc ed metal corrosion is caused by a number of naturally occurring bacteria and fungi in microbial communities that include sulfate reducing bacteria, acid producing bacteria and other types of bacteria involved in the corrosion process (Pope and Morris, 1995; Yasushi, 1998). Microbially induced corrosion as a source of H2S and H2S producing (sulfate reducing) bacteria is a widespread problem in drinking water supplies. In terms of vulnerability, small, rural groundwater supplies may be at particular risk   because of their construction and materials, the presence of iron, steel and perhaps other metals at air-water interfaces, and the presence of risers that are particularly prone to bacterial colonization and corrosion (Tyrell et al., 1996). 3.5

H2S Test Procedures: Media, Formats and Test Conditions

In the initial development of the H2S test by Manja et al. (1982) the test was applied to various drinking water samples of several cities in India . Water samples containing 10 or more coliform  bacteria by MPN were subjected to the th e new ne w H2S test using a 20-ml sample volume in a P-A format. On this basis positive H2S tests were considered unsatisfactory as drinking water. Since this original description of the H2S test, several investigators have reported modifications of the test intended to improve its performance. Such modifications have included: test medium, medium preparation (dried at elevated temperature, lyophilized, autoclaved only, etc.) sample volume (20 ml,, 100 ml, etc.), paper use, paper type and paper size to which the medium is absorbed, incubation times and temperatures, and test formats (presence-absence, quantitative MPN and membrane filter  enumeration). The H2S test conditions used in the original study and some of the various modifications thetested. literature are summarized in Table 3 below, including some information on reported how they in were

23

 

Table 3. Comparisons of Media, Materials and Methods for the H2S Test in Different Reported Studies Medium C Co omposition

Paper and Size

Prepar ati on

Dried?/ Conditions

Format/Use Procedures

Sampl es Tested and Other Tests

Comments

Reference

1) 20g peptone, 1.5 g dipotassium hydrogen  phosphate, 0.75g ferric ammonium citrate, 1g sodium thiosulfate, 1 ml Teepol, 50 ml water  2) Ditto 1) above

Tissue paper, 80 cm2 , folded

Absorb 1 ml,  place in bottle; sterilize

Yes/50o C

P-A, 20 ml sample; Ambient (30-37o C) incubation; 12-18 hrs.

S and G, MPN coliform tests for samples with 10 or  more TC/100 ml

First reported development and use of the H2S test

Manja et al., 1982

Ditto 1) above

Ditto 1) above

Ditto 1) above

P-A, 20-ml sample; incubate at 22 and 35o C

Potable water samples, Lima, Peru

Ratto et al., 1989

3) Ditto 1) above

Ditto 1 above

Ditto 1 above

Not reported

S (Rx and Cl2); Colilert and LTB-MUG

4) Ditto 1) above, except specified only certain  peptones 5) Per Liter: 400 g peptone, 30 g K 2 HPO4 , 15 g ferric

 None

2 ml medium in  bottle; sterilize

 No/lyophilize instead of  heat, after autoclaving

P-A; 20-ml sample; Ambient 26-30oC incubation; 12-15 hrs. P-A, 20 ml sample; Ambient (22-37oC)

Compared to MPN and P-A tests for coliforms and fecal coliforms Compared to E. coli  tests in municipal water 

 1 microliter  added to folded

 Yes/at 50o C for 3-4 hours, after autoclaving

Q, 5 bottles of 20-ml sample volumes each

Slightly modified medium and its preparation and test format (no tissue paper) Improved test using cystine in medium and a paper strip

Kaspar et al., 1992

Same as above

S , G and mineral waters; compare to TC and FC MPN tests Various, S and G, sometimes diluted; MPN coliforms and fecal coliform tests

in a quantitative, 5-bottle (MPN) format

Not reported

tissue paper, dried, added to  bottle, autoclaved Not reported

Not reported

P-A, 100 ml

Various treated and untreated waters; total coliforms and coliphages

Castillo et al., 1994

Ditto 1) above

No

P-A, 20 ml sample,

8) Ditto 1) above

Paper towel, 80 cm2 , folded Not reported

Not reported

Not reported

P-A, 100 ml

Distilled water seeded with Salmonella Various S and G; total coliform bacteria and coliphages

9) Ditto 5) above

Paper towel 50 cm2

 No

P-A, 20 ml sample

Various S and G; Total and fecal coliforms by MF methods

10) Ditto 1) above

Ditto 1) above

1 ml aliquots added to folded  paper towel, autoclaved;  placed in sterile sterile  plastic 40-ml sample bottles Ditto 1) above

Applied to Chilean waters; isolated and speciated  bacteria from positive H2S tests Optimized as presumptive Salmonella  Test Tested for comparative detection in treated and untreated Chilean drinking waters Tested for sensitivity and specificity with pos. and neg. controls and in presence of  high levels of other bacteria; also field tested

Ditto 1) above

P-A, 20 ml

Ground waters

11) Tested 3 media: ditto 1) above; ditto 1) above + 0.125g; ditto 1) above+5 g yeast extract and only 15g  peptone 12) 10 g peptone, 1.5 g dipotassium hydrogen  phosphate, 1.5 g ferric

s ame as above

Add 5 ml medium, place in bottle, autoclave

 No

P-A, 100 ml sample, multiple temperatures o from 0-47 C, incubate up to 5 days

Feces diluted in sterile distilled water; 100-ml H 2S test vs. fecal coliforms

Isolated and speciated  bacteria from pos. samples optimize medium, incubation time and temperature on samples of feces, but not water 

 Naraju and Sastri, 1999 Pillai et al., 1999

 None

Prepare agar  medium, autoclave and

 No

Filter sample thru 47mm diameter membrane filter; place on agar 

Cistern, ground and stream waters; total and fecal coliforms and E. coli

Compared H 2S MPN to H2 S MF to TC and E. coli.

Rijal et al., 2000

Compare original liquid and commercial spray dried media on Citrobacter and Salmonella

Manja et al, 2001, 2001

ammonium citrate, 20 g sodium thiosulfate, 20 mL Teepol and 0.25 g/L Lcystine, pH = 6.9 6) Ditto 1) above

7) Ditto 1) above

Kromoredjo and Fujioka, 1991

Venkobachar et al., 1994

Gawthorne et al., 1996 Martins et al., 1997

Genthe and Franck, 1999

24

 

ammonium citrate, 1 g sodium thiosulfate,, 15 g agar and 1 L deionized water, pH 7 13) ditto 1) above; ditto 1) above, except 15g peptone and 1g yeast extract; "1above except 250mg L-

 pour into 60x15 mm plates

"1)above

"1)above; or   single-strength  powder media, radiation

medium; incubate anaerobically at RT (2530o C) , 24 hrs. Yes for original medium/60o C for 2 days.

P-A, 20 ml sample; 35oC; 18, 24 and 48 hour incubations times

cystiene, "1)above except sterilization; no 15g peptone and 250mg Ltissue cystiene Abbreviations: G = groundwater, S = surface water, P-A = presence-absence test, Q = quantitative test, TC = total coliforms, FC = fecal or thermotolerant coliforms, Ec = E. coli

25

 

As shown by the data summarized in Table 3 above, various modifications of the H 2S test have been reported. Besides the use of different modifications of the original test in the various studies, the modifications were developed and evaluated using different samples in the various studies. Some studies used diluted feces, others used laboratory cultures of specific bacteria, such as Salmonella  and others used field samples of water of varying quality. Pillai et al. (1999) used feces diluted in distilled water to evaluate medium composition, incubation time, incubation temperature and fecal coliform bacteria concentration to optimize conditions for a 100-ml sample volume H2S test. They found that lower concentrations of bacteria required longer  incubation times and higher temperatures for H2S detection, results did not require an incubator if room temperature was between 20-44oC, and the presence of L-cystine in the medium improved detection. The test was not applied to field samples of water. Gawthorne et al. (1996) evaluated the H 2S test for Salmonella  detection using four species grown in the laboratory and then seeded into water. They found that detection of as little 5 CFU/100 ml was  possible, longer incubation times (48 hours) increased detection of low Salmonella  levels, and the  presence of other bacteria has no effect on Salmonella   detection. The H2S test was recommended as a  presumptive test for Salmonella  in drinking water in conjunction with coliform testing. Venkobachar and colleagues (1994) incorporated cystine into the H 2S medium and compared the original H2S test to the one with the modified medium containing cystine using different water sources. Correlation analyses indicated that the cystine-modified test was more sensitive and less timeconsuming than the original test. Rijal and colleagues (2001) developed and evaluated two modifications of the H2S test: (1) a MPN version using replicate sample volumes of 1, 10 and 100 ml and an enumerative version for H2S colonies on membrane filters in using an agar medium. medium. When both H2S tests were compared to each other and to coliforms and  E. coli   in rainwater cisterns of drinking water, both H2S methods gave results comparable to E. coli . In studies by Manja et al. (2001) the following media were compared for the H 2S test: (1) original H2S medium, (2) original medium with 250mg L-cystiene, (3) original medium with decreased peptone at 15g and added yeast extract at 2 g, and (4) medium 3 with 250 mg L-cystiene and a the lower peptone concentration of 15 g Based on detecting low levels of Citrobacter freundii   and Salmonella typhimurium   type strains seeded into sterile distilled water at about 5 CFU per sample, medium formulation 3 (original medium plus 250 mg/L cystiene) was judged to give the best results based on the numbers of positive samples obtained. In addition, the commercial powder form of the medium gave better results than the strip medium (liquid medium applied to paper and dried in the lab). It was concluded that as few as 1 Salmonella  was detectable in 20 ml of sample. As summarized in Table 3, the results of several studies indicate that various forms of H 2S test have  been evaluated evalu ated and are being used. The various forms of the H2S test differ in the following ways: the medium and its preparation procedures, media format (dried onto paper strip, use as powder, and agar  medium), test format (presence-absence, MPN, and membrane filter), sample volumes, incubation times and incubation temperatures. If the results of these studies are taken together, it appears that the

27

 

addition of cystine or cystiene, longer incubation times (24-48 hours) and incubation temperatures in the range of 25-35oC give the best results in terms of detecting low levels of H2S producing bacteria. However, there has been no systematic comparison of the various forms of H 2S tests used by different investigators and no effort to achieve a standard test procedure. Until recently, all of the H2S tests required the use of media formulated from scratch and applied to paper strips manually. Therefore, the test media or materials were were was not readily available from commercial sources. In India, there have  been efforts to have the medium made commercially and to implement performance criteria for the commercially prepared medium. In the United States States of America one Company has marketed an H2S test kit for use by small labs and consumers. However, this commercial test is probably too complex, too inaccessible and too costly to be used in the developing world in response to the constraints in section 2.6. Research and development studies have been done in India by UNICE UNICEF F (United Nations Children's Fund)-India and its partners (Rajiv Gandhi National Drinking Water Mission, Department of  Drinking Water Supply, Ministry of Rural Development, Government of India) to develop, evaluate and disseminate the specifications for a H2S test and field kit for use in drinking water (Manja et al., 2001). The test is not advocated as a replacement for conventional coliform and other bacteriological testing. It is recommended for use by community workers to monitor water supply sources. At the  present time there remain r emain considerable obstacles ob stacles to the widespread widespr ead use of the H 2S tests because of their  lack of uniformity and lack of availability in a ready-to-use form. Greater efforts to determine the optimum properties for and test conditions of H2S tests are recommended. Also recommended are further efforts to evaluate their validity, and predictability as fecal indicators before widespread production, dissemination and usereliability of either commercial or made-from-scratch H2S tests. 3 .6

Comparison of H2S Tests to Other Tests for Detecting Fecal Contamination of Water

In evaluating the performance of the H 2S test for detecting bacteriological evidence of fecal contamination of water, it is necessary to have a basis or reference point ("benchmark") for determining efficacy. In some studies the basis for determining efficacy has been the use of sterile water seeded with known quantities of specific H2S-producing bacteria or mixed populations of H2S bacteria in the form of diluted feces. However, such testing does not address the variability of water matrices in terms terms of their chemical and microbiological quality. Testing of natural waters containing fecal contamination also is necessary. Therefore, many evaluations of the H2S test have been based on performance comparisons with other, conventional bacteriological tests for fecal contamination of drinking water  when applied concurrently to field samples of water. Tests for fecal contamination of water against which the H2S test has been compared includes: coliforms, thermotolerant (fecal) coliforms,  E. coli , Clostridium perfringens, Salmonella and coliphages. Because none of these other tests are ideal at detecting fecal contamination of water, the results of such comparisons are open to interpretation. However, most investigators assume that if the H2S test gives positive results similar to or greater than the reference test, its performance is acceptable. The results of a number of comparative studies of the H2S test against other bacteriological tests for fecal contamination are summarized in Table 4 below.

28

 

Table 4. Comparisons of H2S and Standard Bacteriological Tests Applied to Various   Samples W at ate r S a am mp pll es es

Bacterial Indicator(s) Compared to H2S test

Bacterial Results Suitable/Unsuitable

H2S test Results Suitable/Unsuitable

Agreement, %

% D is isagr ee ee me me nt nt

Refe re re nc nce

Various S and

Coliforms (+ E.

332/337

332/337

88.3

13% and 13%

Manja et al.,

coli  in one set of  G Waters; samples) India Tap water, Total and fecal Peru coliforms Tap (Rx S), Coliforms and Banjarmasin,  E. coli Indonesia Various S and Total and fecal G, India coliform MPN and original H2S test

(coliforms) 11/9 (TC); 11/9 FC 1/24 Colilert 3/24 LTB-MUG

1982 11/9

95

5%

2/24

96

4

11

Undisinfected water, Chile

49/170 (TC)

36-40/179-183*

TC: TC: 89 FC: 91 Orig H2S: 94 90-92

Disinfected water, Chile Raw waters, Chile Treated water, Chile Mysore, India

290/113 42/12

2 54-278/125149* 42/12

7177% 100

0

50/4

43/9

81

19

Not Done

Not Done

14/37

 No data

Total and fecal coliform MF Fecal co coliform

189/224

150/263

No Data 82 %

Tested b bu ut reported results not quantified 0/25 (TC), 8/17 (E. coli)

Tested but reported results not quantified 8/17 H2S agar  9/16 H2S MPN

Various, South Africa Feces diluted in distilled water  Cistern rainwater 

Municipal drinking water  supplies

Fecal coliform MPN

 Not Reporte Reported; d;  Nor reporte reported; d; see % agreement see % an and d agreement and disagreement disagreement

294/392

296/390

9 6

10-12%

Castillo et al., 1994

23-29%

Castillo et al., 1994 Martins et al., 1997 Martins et al., 1997 Nagara Nagaraju ju an and d Sastri, 1999 Genthe and Franck, 1999 Pillai et al., al., 1999

18 %

 Not  Not reported reported report ed 68 0 TC TC 32  E. ccoli oli 100  E. coli 90 % (92% for  (+)iv e; 88% for  (-)ive

Ratto et al., 1989 Kromoredjo and Fujioka, 1991 Venkobachar  et al., 1992

10 %

Rijal et al., 2000

Manja et al., 2001

As shown in Table 4, when applied to water samples having 10 or more total coliform MPN per 100ml, Manja et al. (1982) scored the same numbers of samples suitable and suitable by both MPN coliform tests (#10/100 ml = suitable >10/100 ml = unsuitable) and the H 2S test (not black in 20 ml = suitable and black in 20 ml = unsuitable). When the water water samples were divided into ranges of 

29

 

coliform concentrations, the H 2S test gave positive reactions for all samples with 41 or more coliforms  per 100 ml, 25/34 25 /34 H2S positive for samples with 21-40 coliforms per 100 ml, and 37/44 positive with 11-20 coliforms per 100 ml. H2S positive-samples contained Citrobacter freundii  (23 samples), Salmonella  species (6 samples), Proteus mirabilis  (2 samples), Arizona species (2 samples) , Klebsiella (1 sample) and  E. coli   (3 sample). Only 1 type of H2S-producing organisms was isolated from each separate sample and the methods of isolation were not specified. No tests were done for the presence of  absence of other, environmental H 2S-producing bacteria. No specific tests were done to determine the  presence of viral, bacterial or parasitic pathogens, although althou gh Salmonella  were detected in some samples. The authors considered the test reliable, simple to perform and useful for screening purposes where resources, time, manpower and laboratory facilities are limited. Ratto et al. (1989) evaluated the original H2S test at incubation temperatures of 22 and 35°C and compared it to MPN and P-A total coliform (TC) and fecal coliform (FC) tests on 20 potable water  samples from Lima, Peru. The frequency of positive (unsuitable) samples was similar but not identical for all tests: 9/20 by P-A, 9/20 by H 2S at 35oC, 6/20 by H2S at 22oC, 8/20 by TC MPN and 6/20 by FC MPN. It was concluded that the H2S test was an equally or more sensitive test than TC and FC tests and was an ideal procedure for isolated water supplies where laboratory facilities do not exist. Kaspar et al. (1992) evaluated a modified version of the original H 2S test (no tissue paper and lyophilizingthat rather thanwas heatnot drying of for the control medium) applied to 101 samples. concluded the test suitable of and surface waterit and dug water well water due toThey the frequent presence of non-fecal (total) coliforms presumed to arise from degradation of plant tissues and  poikilothermic animals. Dug wells nearly always gave positive results in the H2S test and in the coliform test, but not fecal coliforms. They concluded that the test was useful for qualitative screening of piped or treated water systems. However, it was concluded that positive H2S test results must be confirmed by standard bacteriological tests. The test was considered valuable as an educational and motivational tool for improved water sanitation, because of the color change and foul smell from  positive samples. sample s. Venkobachar et al. (1994) developed a modified H2S test that included cystine in the medium and was used in an MPN test with five 20-ml samples. samples. The modified test reduced the test time from 23 to 17 hours, was more sensitive than the original H 2S test, and was well correlated with total coliform (89%) and fecal coliform (91%) tests when applied to 101 water samples. It was concluded to be simple, requiring little laboratory support and well suited for routine quality assessment of rural water sources. Sivaborvorn (1988) tested 705 samples from a variety of waters in Thailand (shallow and deep wells, rainwater, pond water) by the original H2S test and by coliform MPN. Based on agreement between a  positive H 2S test and 10 MPN/100 ml as a coliform or fecal coliform positive, the two tests agreed 85% and 88% of the time, respectively. It was concluded that the H 2S test can be used to screen water for  fecal contamination in the field where laboratory facilities are limited. Castillo et al. (1994) reported that for 622 water samples tested by the H2S and coliform tests, 168 samples were positive by both tests and 179 samples were negative by both tests. The H2S test  produced about 10% more positive samples than the coliform test but included samples that were  positive for Clostridium  spp. The H2S test gave similar results at both 32 and 35°C, indicating that temperatures in this range range are not critical. Bacteria detected detected from H2S positive samples included

30

 

 Klebsiella  spp., Enteroba  Enterobacter  cter  spp.,  spp., E. coli, Citrobacter  spp.,  spp.,  Aeromona  Aeromonass spp., Clostridium spp.,  Hafnia spp., Salmonella  spp.,  Acinetobacter  spp.,   spp.,  Morganella  spp.. It was concluded that the simplicity and low cost of the H2S test was applicable to tropical and subtropical potable waters.

In studies of 54 complete conventionally treated drinking waters and their corresponding raw source waters Martins et al. (1997) found 100% agreement between total coliform and H2S results for raw waters and 81% agreement for treated waters (Table 3). The H2S test was modified by increasing the sample size to 100 ml. In treated waters more samples were positive by the H2S test (9) than by the coliform test (7), which was attributed to the presence of Clostridium perfringens in the H2S-positive  but TC-negative TC-negativ e samples. In treated waters the H2S and TC results were significantly positively correlated (P< 0.0001) Spearman rank correlation test) but in raw waters they were significantly negatively correlated (P = 0.0008). The authors concluded that the H2S test was a suitable indicator of   potable water quality and treatment trea tment and provided greater prot protection ection than the total coliform test test..  Nagaraju and Sastri (1999) tested ground water wells of Mysore city, India for H2S bacteria using the methods of Manja et al. (1982) and 24-hour incubation at 37 oC. 37 of 51 ground water water samples were  positive. From these thes e H 2S-positive samples the following 63 bacteria were isolated:  Proteus mirabilis (19),  Proteus vulgaris (14), Citrobacter freundi  (13), Salmonella spp. (8),  Klebsiella  pneumoniae (5) and Klebsiella ozaene  (4) were isolated. Genthe and Franck (1999) evaluated the H 2S test of Venkobachar et al. (1994) for specificity, sensitivity and interference by non-target bacteria using seeded positive and negative samples and reported favorable results. When applied to 413 water samples from various sources, including ground and surface water, the H2S test showed 82% and 86% agreement with fecal coliform results when applied to higher quality (so-called level 1) waters with test incubation temperatures of 35 and 22 oC, respectively. It was concluded that the H2S test was sensitive and correlated with traditional indicator   bacteria, especially fecal coliforms. The test was considered useful for on-site field use, light, easy to use and portable. The test was recommended for use in addition to current water quality indicators of  microbial quality, and especially where testing would otherwise not normally be done. Pillai et al. (1999) evaluated various modifications of H2S tests for detection of fecal contamination using 100-ml volumes of feces diluted in distilled water to contain different levels of fecal coliform  bacteria. The presence of cystine in the medium and higher incubation temperatures 28-44oC vs. 22oC) improved detection, with lower levels fecal contamination (fewer fecal coliforms) detected faster. Rijal et al. (2000) compared two versions of the H2S test, a paper strip MPN and a membrane filter  enumeration on agar medium, to each other and to the occurrence of total coliforms and   E. coli  in samples of cistern rainwater, ground water and stream water. Similar detection of bacterial contamination was achieved by the MPN, MF version of the H2S test and  E. coli, although total coliforms were detected in more samples than were either  E. coli   or H2S bacteria. The H2S test was compared to total and fecal coliform and  E. coli  tests to determine efficacy of a solar disinfection system. Similar results for indicator reductions were achieved by all fecal indicator tests used. The authors concluded that the H 2S test was a reliable method to measure the hygienic quality of water. Manja et al. (2001) compared the H2S test (with cystiene in the medium, different sample volumes, different incubation times and incubation at different temperatures) to MPN tests for coliforms for 

31

 

detecting fecal contamination in 686 water samples in India. The H2S test gave results comparable to the MPN test (not significantly different), with concordance in 620 (90%) samples, negative H2S test and positive MPN test (false negative) in 34 samples (4.9%), and positive H2S test and negative MPN test (false positive) in 32 samples (4.7%). However, 21 of 23 "false positive" (negative coliform MPN) samples had coliforms in H2S bottles. Agreement of H2S-positive and coliform- positive samples increased from 91% at 48 hours to 95% at 72 hours. The H 2S test results were comparable (not significantly different) for sample volumes of 20, 55 and 100 ml. Positive H2S results were generally obtained in 18-48 hours of incubation at 25-44oC. Use of an incubation temperature below 25oC was not recommended. As shown in Table 4,  The H2S method has been extensively studied by a number of investigators in different parts of the world. Such studies include evaluations of the original method, studies on modifications of the method and field testing, usually with side-by-side comparison to other water  quality tests. In some of these comparison studies the data are limited or have not been subjected to rigorous statistical analysis. However, the results of most studies suggest that the H2S method detects fecally contaminated water with about the same frequency and magnitude as the traditional methods to which it was compared. In general, the sensitivity of the H2S test appears about the same as other tests for fecal contamination of water, although as already noted, this aspect of the test has not been rigorously tested in some of the reported studies. Testing conditions and format, sample size, incubation temperature and incubation influences sensitivity. conditions have differed among the different studies time reported in thetest literature, it isBecause difficultthese to make consistent comparisons and draw overall conclusions. However when when comparisons with other methods of  detecting fecal contamination were done, the H2S method appeared to have sensitivity similar to the other methods, based on finding contaminated samples. In most comparative studies there were always samples that yield positive results for other  microbiological tests and negative H 2S tests, and vice versa. However, such results are not unexpected. For one, the various tests measure different things and do not always employ the same sample volumes. Furthermore, when the levels of microbial contamination are low, it is statistically possible for one sample volume to contain bacteria of interest and for another to not contain them. Where study data were subjected to statistic analysis most studies found high associations (e.g., correlation) between fecal indicator bacteria (e.g., E. coli) and positive H2S results. Given the previously discussed ability of a large variety of heterotrophic bacteria to produce a positive H 2S test and therefore a false positive the observed correlations suggest that in most natural and treated waters the majority of the H 2S  producers come from organisms associated with the human or animal digestive tract. A false fa lse positive is less likely to lead to a risk of disease because it would result in the suspect water either not being used or subject to additional testing. Of great concern with the H2S test as with other fecal indicator tests is the potential for false negatives; that is, not detecting fecal contamination when it is present. In this case the test does not identify water  water  that is unsafe and the water could be consumed, leading to pathogen ingestion and to disease. The method, as with the various bacteriological tests, does not detect viruses or parasites. Testing of  drinking water for the many viruses and parasites of concern is still impractical and unaffordable and still not done on a routine basis in most countries and regions. However, the H2S test detects bacteria other than coliforms that are associated with fecal contamination, including Clostridium perfringens. Clostridium perfringens is one of the more resistant indicators of fecal contamination and can be found in drinking waters when no coliforms can be found. Therefore, it is possible for the H2S test to give a

32

 

 positive result when fecal contamination contaminatio n is present even if no coliforms are present. Such findings finding s have  been observed in some comparative studies stud ies between the H 2S test and other bacteriological tests. 3.7 3.7

Det Deter ermi mina nati tion on if H2S Tests Meet the Criteria of an Ideal or Preferred Indicator of Fecal Contamination

In Table 5 below are listed the essential criteria of an ideal or preferred indicator of fecal contamination of drinking water and other waters and the extent to which these criteria were addressed and fulfilled in  previously published studies on the H2S test as a method to detect fecal contamination. It is apparent from this compilation of data that most of the key criteria for fecal indicators of water quality were not investigated in the studies reported in the literature to date. This lack of data on the extent to which which H2S tests fulfill the essential criteria of an indicator is a major concern. This is because the test has  been in existence for two decades, it has been repeatedly modified, tested and field applied in many  parts of the world, it is now widely promoted by some scientists and other authorities, and yet it has never been subjected to critical testing for its ability to fulfill or meet the essential criteria of a fecal indicator of drinking water quality.

33

 

Table 5. Fecal Indicator Criteria for Fecal Contamination and The Extent to Which they are Met in Reported Studies of H2S Tests Reference

Absent in non-fecally contaminated water

Manj a et al al .,., 1 19 982

comparison to other indicat indicators ors of fe T/ b by y cco ompar is is on on tto oo ott h heer   T/ by comparison indicators of fecal contamination T /Y /Yes , by compar is is o on n to T/Yes, other indicators of fecal By comparison to other indicators of  fecal contamination contamination T/ by comparison to other indicators of  T, Yes, by comparison to fecal contamination other indicators of fecal contamination T / b y co m mp p aarr iso ison n t o o th th eerr   T/Yes, by comparison to other  indicators of fecal contamination indicators of fecal contamination T/Yes, by comparison to other  T /Y /Y ees, s, b y c o om mp paa ri ri so so n t o indicators of fecal contamination other indicators of fecal contamination

R at at to to et al., 1989

Kromoredjo and Fujioka, 1991 K asp aspaa r e t a l.l. , 1 9 99 92

V en en k ko ob baa cch h ar ar , 1 9 99 94

Cast Castii llll o e t a l.l. , 1 9 99 94

Present in fecally contaminated water

Outnumber pathogens

Detectss Non-pat Detect Non-patho-g ho-genic enic Bacter Bacteria ia

Respond Respond to eenvir nvirononmental condi-tions like patho-gens

Respond to treatment like pathogens

Ease of  Use

Cost (Inex-pensive?)

No environmenTal multiPlicaTiom

 NT

T/Some detec detected ted are non pathogenic  pathogenic

 NT

NT

NT

NT

NT/ M for some

 NT

T/Some detec detected ted are non pathogenic  pathogenic

 NT

NT

NT

NT

NT

 NT

NT

NT

NT

NT

 NT

 NT

T/Some are

NT

NT

T/ Yes*

T/ by cost compariso n  NT

 NT

T/Some dete detect-ed ct-ed are n nonon pathogenic bacteri  pathogenic bacteriaa

 NT

NT

T/ Yes

 NT

NT

NT/

T /Y /Y ees, s, b y c o om mp paa ri ri so so n t o other indicators of fecal contamination NT T// Lab studies on seeded water  T /Y /Y ees, s, b y c o om mp paa ri ri so so n t o other indicators of fecal contamination

T/Yes, by comparison to other  indicators of fecal contamination

 NT

NT

NT

NT

NT

NT

NT

T/Yes, by comparison to levels of  Salmonella in seeded waters

 NT

NT

NT

NT

NT

NT

NT

T/Yes, By comparison to other indicators of  fecal contamination

 NT

NT

NT

 NT

NT

NT

Ge Gent nthe he and and Fr Fran anck ck,, 1 199 999 9

T/ Yes, by comparison to other indicators of fecal contamination

T/ Yes, by comparison to other indicators of fecal contamination

 NT

NT

NT

Y, Indirect T; results compared for  Rx'd and UnRx'ed water  NT

T/ Yes

 NT

 Naga  Nagaraju raju aand nd Sastri, Sastri, 1999

NT

NT

NT

Rijal eett aall., 2 20 000

T/yes, by by co comparison to to other indicators of fecal contamination

T/yes, by comparison to other  indicators of fecal contamination

 NT

T/ Yes T/Some detec detected ted are non pathogenic  pathog enic

T/ Ye Yess  NT

Manja et al., 2001

T/ Yes, comparison to other  indicators of fecal contamination

T/ Yes, by comparison to other indicators of fecal contamination

 NT

NT

NT

Gawth o orrne eett aall .,., 1 19 997

M ar ar titi ns ns et et al al .,., 19 19 9 97 7

 NY

Yes

T/ Yes, by cost compariso n NT

T/ Yes, compared H2S, TC and FC tests on solar   pasteurized  pasteuri zed waters NT

 NT

NT

NY

NT

NT

NT

Abbreviations: T = tested; NT = not tested by objective or specified measures or methods

34

 

3.8

Potential modifications modificati ons of the H2S test to improve specificity for H2Sproducing bacteria of fecal origin

There are several modifications to the H 2S test that could be considered in an effort to make it more specific for organisms of fecal origin and to reduce the probability of  organisms of non-fecal origin giving a positive response. These modifications fall into two main categories: modifications of the medium itself and modifications modifications of the incubation conditions. The inclusion of bile salts such as sodium deoxycholate or taurocholate in the medium is a common method method for inhibiting the growth of many microorganisms. microorganisms. These surfactants can cause the lysis of sensitive cells. Because human and animal intestinal flora has to tolerate these materials in the intestine they tend to be less sensitive to them. Indeed, the use of bile salts is common in bacteriological media to detect Enterobacteriaece Enterobacteriaeceae. Archebacteria and many eubacteria sensitive, Salmonella  and  E.ae. c oli  are coli not (Gerhardt et al., 1994; K Kamekura amekuraare et al., 1988). while Gram negative organisms are generally insensitive, while gram positives other than Group D streptococci are sensitive. Therefore, the inclusion of bile salts at 0.5% would eliminate the bacilli, the archebacteria and most other soil organisms and prevent them from producing a positive result. However, this modification also may decrease the detection of Clostridium perfringens, which can be of fecal origin and is detectable by current versions of the medium. The extent to which the addition of a  bile salt would inhibit C. perfringens detection by its H2S production is uncertain and would have to be investigated.

NT

Several modifications of the incubation procedure are suggested in an effort to increase specificity for bacteria of fecal origin. One modification to consider is aeration of the medium by vigorous shaking prior to incubation in an effort to make the system aerobic. This would slow or preclude the growth of many anaerobic or  microaerophilic organisms, some of which (e.g., the sulfate reducing bacteria) may give false positives. Additionally, elevated incubation temperatures would be advantageous in reducing the growth of some soil and water organisms of natural origin. Some of these environmental bacteria generally do not grow as well at temperature above 30o C, in contrast to bacteria of human or animal origins. However, elevated incubation temperature may not be an effective way to control the growth of thermotolerant or thermophilic bacteria of natural origin, such as those from geohydrothermal environments because these bacteria are adapted to higher  temperatures. Furthermore, higher incubation temperatures such as 35o  C will make the test potentially more difficult to apply in the field and in settings other than microbiology laboratories because an incubator may be required. 3.9

Costs of H2S Tests

It was not possible to find detailed breakdowns of estimated costs for H 2S tests, but all reports indicate lower costs than conventional bacteriological tests for fecal contamination. Several investigators have listed the costs of H2S tests and compared them to the costs of standard bacteriological tests for fecal contamination of water. Kromoredjo and Fujioka (1991) reported that the cost to analyze one sample by a 5tube MPN test was: US$6.50 by commercially available, defined substrate

35

 

technology test, US$1.60 by fecal coliform test using lauryl tryptose broth (LTB) plus 4-methylumbelliferryl-Beta-D-glucuronide (MUG) and $US0.62 using the H2S test. They noted that the relative costs for the commercially available tests would be even higher because its high shipping costs were not included and the tubs would be discarded after their use, while the shipping costs for LTB-MUG fecal coliform and H2S tests would be comparatively low and the tubes used in these tests would be reused up to five times. Kaspar et al. (1992) indicated that the cost of the H2S test was much lower than the costs of common microbiological tests, such as those for  coliforms, but detailed costs estimates and comparisons were not provided. However, they indicated that the costs of materials for an H2S test were about US$0.25 per unit. Genthe and Franck (1999) stated that the cost of the H 2S test test was was inexpensive. inexpensive. They listed the approximate costs costs of the H2S test test at <5.00 South African African Rand or ZAR  <US$0.44). The estimated costs of materials for the membrane filter (MF) fecal coliform test were about 7.00 ZAR or about US$0.61 and for the defined substrate technology they were aboutpreparation 30-50 ZAR or and US$2.60-4.35). costscoliform for the H2S test didtests, not include material time the costs for The the fecal MF test did not include media preparation and carrying our filtration. Overall, these data indicate that the H2S test costs are relatively low compared to those for standard bacteriological tests, either prepared in one's lab or commercially  purchased as ready rea dy to use. However, a more formal analysis of H 2S test costs listing all of the cost elements and methods of cost calculation is recommended. It should be recalled that the costs of analyses may represent a small fraction of total costs if staff  time and travel costs are taken into account.

36

 

4. SUMMARY AND CONCLUSIONS

The H2S method in various modifications has been tested in many places in different waters and produced results reported as indicating it to be a reasonable approach for  testing treated and untreated waters for fecal contamination. It offers advantages including low cost (estimated at 20% of the cost of coliform assays), simplicity and ease of application to environmental samples. It has not been suggested as a replacement for other testing procedures for fecal contamination of water. Because it has not been adequately tested in regions with temperate and cold climates nothing can be said about its applicability in those regions. Because it offers the potential for testing water in places where other testing methods are not feasible, its promotion, dissemination and use have been encouraged by many developers and evaluators. However, as is apparent from the review and analyses  presented here, H2S tests have not been evaluated and judged according to the generally accepted criteria of an indicator of fecal contamination, except perhaps indirectly and by comparison. That is, no systematic efforts have been made to determine directly if H2S tests fulfill the essential criteria for an indicator of fecal contamination in treated and untreated drinking water and its sources. Instead,  previous studies have attempted to validate and evaluate the test against on the basis

of the detection of established fecal indicator bacteria and certain pathogens, notably Salmonella , in experimentally seeded and in field samples of treated and untreated water. If there is adequate agreement (correlation, concordance, etc.) in the classification or quality determination of water based on a standard or generally accepted fecal indicator and an H2S test, the H2S test is taken to be an acceptable indicator of fecal contamination. In some studies these measures of agreement have  been further supported by the isolation and identification of bacteria of fecal origin from H2S-positive cultures. The criteria for determining agreement between outcomes for H2S and fecal indicator bacteria tests in comparison studies are not uniform among the various studies reported and often have not been explicitly stated. In some studies these comparisons are supported by statistical analyses for  correlations or other associations and in other studies they are not. In some studies, the comparative criteria include efforts to determine sensitivity (lower limit of  detection) and specificity (ability to detect specific bacteria or bacterial groups). In all of the reported studies, no efforts have been made to determine if the H2S tests detect non-fecal bacteria capable of producing hydrogen sulfide, such as sulfate reducing  bacteria. A preliminary study conducted in the laboratory of one of us (Mark Sobsey) showed that a standard, commercially available H2S test detects sulfate-reducing  bacteria of non-fecal origin origi n in 25% of the samples tested. Because of these deficiencies, it is not possible to widely and unequivocally recommend H2S tests for the determination of fecal contamination in drinking water. There remain too many uncertainties about the reliability, specificity and sensitivity of  the test for detecting fecal contamination of drinking water and its sources. Despite this lack of formal analytical support for and validation or verification of H2S tests, there are many studies reporting relatively good agreement in the classification

37

 

of drinking waters as suitable or unsuitable based on the results of H 2S tests in comparison with fecal indicator bacteria tests. Hence, there is substantial empirical evidence showing that H2S tests and fecal indicator bacteria tests may provide information on the suitability or unsuitability of drinking water with respect to fecal contamination. For this reason, there are good reasons to support the further  investigation and use of H2S tests under certain circumstances and in certain settings. In particular, if the alternative to H 2S testing is no water quality testing at all for fecal contamination, the H2S test is recommended for use, with caution. The caution concerns possible false positive results due to H2S presence or formation in water  from sources other than fecal contamination. In addition, H2S testing also is recommended with caution for educational and motivational purposes to promote water sanitation and hygiene education in outreach and dissemination programs. Again, it must be established or verified that the test will give correct results with respect to water classification as suitable or unsuitable when applied to the treated or  untreated drinking waters being tested. Before its adoption for widespread use standardization of the H2S test procedure as well as conditions and precautions for its use and interpretation based on an improved evidence base will be necessary. The studies conducted to date have used a variety of  different media compositions, test strips in some cases, general darkening of the medium in others and even membrane fil filters ters and agar media in yet others. A variety of different incubation times and temperatures have also been used. A number of  investigators have addressed incubation time and temperature issues that can result in false negatives if the appropriate choices are not made (temperatures too low or 

incubation times too short). Some have suggested the addition of cystine or cysteine to the medium and supplementing the medium with yeast extract and deoxycholate. There is likely to be merit in many of these suggestions. Empirically, the test appears to detect mostly organisms of fecal origin (human or  animal), and therefore, it may have value. There is no reasonable way, however, to  preclude a false positive in samples containing other H2S producers, nor is there an easy and expedient way to determine what those organisms may be. The ability to form H2S is too widely distributed within the microbial world for it to be an unequivocal test of fecal contamination in water. It also cannot discriminate between organisms of human or animal origin, as long as they produce H2S. Most of the  problems likely to be encountered with H 2S tests are false positives rather than false negatives. Such misclassification of drinking water errs on the side of safety. However, it can result in the rejection of water that is acceptable with respect to fecal contamination, and it may prompt efforts to seek or provide alternative or further  improved drinking waters. Such efforts in effect increase the costs of providing safe water. If a false positive leads to more testing or the rejection for use of the water for  drinking, than alternative sources must be sought and this may be more expensive. Inadequate attention has been given to the use and reliability of the method for testing ground waters. In many parts of the developing world ground water is the only source of drinking water. In groundwater, there is the strong possibility of sulfides being  present due to natural geohydrological sources and to anthropogenic impacts other  than fecal contamination, both of which are false positive results. The use of the H2S test in ground waters needs to be further assessed, as does modification of the

38

 

 procedure, as suggested suggeste d above, to allow determination of false positives due to sulfide and non-fecal sulfide-producing bacteria in the sample water. In assessing the applicability of the H2S test in presence/absence format, the limitations limitati ons of P/A testing should be recalled. P/A testing was developed for and is applicable where most tests provide a negative result. Where a significant proportion of tests provide a positive reaction quantitative testing is preferred in order to determine relative health risk  and therefore relative priority of need for correction, such as by improved or greater treatment or by finding a higher quality source water for supply. Water quality testing alone is an inadequate response to the challenges of  ensuring water safety. Occasional tests conducted on a water supply may provide a false sense of security as water quality can vary widely and rapidly. For these reasons water quality testing should be accompanied by verification of the state of the source or supply, for instance, by sanitary inspection as described in the WHO Guidelines for Drinking-water Quality Volume 3. As previously indicated, end-product testing is now only one of several key management tools in the provision of safe drinking water under the forthcoming Guidelines from Drinking Water Quality. The development, implementation and use of a Water Safety Plan that considers the quality and overall management of drinking water from source to consumer is the goal for all water supplies. In such a plan, now

including hazard analysis-critical control points (HACCP), endproduct testing is not a critical control point (CCP).

39

 

5. LITERATUR LITERATURE E CITED Allen, M.J., E.E. Geldreich. 1975. Bacteriological criteria for ground water quality. Ground Water 13:5-52. Atlas, R.M. and R. Bartha (1993) Microbial Ecology. Fundamentals and Applications, Benjamin/Cummings Publishing Company, Redwood City Buchanan, R.E. and N.E. Gibbons (1974) Bergey's Manual of Determinative Bacteriology, 8th  edition, Williams and Wilkins, Baltimore. Berry, S.A., and B.G. Noton. 1976. Survival of bacteriophages in seawater. Water   Res. 10:323- 327. Castillo, G., R. Duarte, Z. Ruiz, M.T. Marucic, B. Honorato, R. Mercado, V. Coloma, V. Lorca, M.T. Martins, and B.J. Dutka. 1994. Evaluation of disinfected and untreated drinking water supplies in Chile by the H2S paper strip test. Water   Research 28:1765-1770. Clark, J.A. (1968) A presence-absence (P/A) test providing sensitive and inexpensive detection of coliforms, fecal coliforms and faecal streptococci in municipal drinking water supplies. Can. J. Microbiol ., ., 14:13-18.

Craun, G.F., and R.A. R.A. Gunn. Gunn. 1979. Outbreaks of waterborne disease in the United States 1975-1976. JAWWA  JAWWA..  71:422-428. Fukui M, Teske A, Assmus B, Muyzer G, Widdel F. (1999) Physiology, phylogenetic relationships, and ecology of filamentous sulfate-reducing bacteria (genus Desulfonema)  Archives of Microbiology. 172(4):193-203. Gawthorne, T., R. Gibbs, K Mathew, and G. Ho. H2S papers as presumptive tests for  Salmonella   contamination in tropical drinking water. Water Sci. Technol. 34:187194. Gerhardt, P., R.G.E. Murray, W.A. Woods, and N.R. N.R. Krieg. 1994. Methods for  General and Molecular Bacteriology. ASM Press, Washington, D.C. Genthe, B. and M. Franck (1999) A tool for Assessing Microbial Quality in Small Community Water Supplies: An H2S Strip Test. WRC Report No. 961/1/99, 33  pages. to: Water Research Commission. by: Division of Water, Environment and Forestry Technology, CSIR, Stellenbosch, South Africa. Hazen, T.C. (1988) Fecal coliforms as indicators in topical waters: a review. Tox. Assess., 3:461-477. Iverson, J.B., and B.J Fleay. 1991. Serovars of Salmonella isolated from humans, animals, waters and effluents in natural and disturbed environments in western Australia.  Proceedings of the 14th   Federal Convention, Australian Water and  Wastewater 2:435-441.

40

 

Jimenez, L, I. Munoz, T. Toranzos, and T.C. Hazen. 1989. Survival and activity of  Salmonella typhimurium and  Escherichia coli   in tropical freshwater.  J. Appl.  Bacteriol. 67:61-69. Kaspar, P, I. Guillen, D. Revelli, T. Meza, G. Velazquez, H. Mino de Kaspar, L. Pozolli, C. Nunez and G. Zoulek (1992) Evaluation of a simple screening test for the quality of drinking water systems. Trop. Med. Parasitol ., ., 43: 124-127. Kamekura, M., D. Oesterhelt, R. Wallace, P. Anderson, and D.J. Kushner. 1988. Lysis of halobacteria in Bacto-Peptone by bile acids.  Appl. Environ. Microbiol . 54:990-995.

Kaspar, P, I. Guillen, D. Rivelli, G. Valasquez, H.M. De Kaspar, L. Pozzoli, C. Nunez and G. Zoulek (1992) Evaluation of a simple screening test for the quality of drinking water systems. Tropical Medicine and Parasitology , 43(2): 124-127. Kelly, D.P., and N.A. N.A. Smith. 1988. Organic sulfur compounds in the environment:  biogeochemistry, microbiology, and ecological aspects.  Adv. Microbiol Ecol . 11:345-386. Kromoredjo, P., and R. Fujioka. 1991. Evaluating three simple methods to assess the

microbial quality of drinking water in Indonesia.  Environmental Technol and Water  Quality 6:259-270. MacKenzie, W.R., N.J. Hoxie, M.E. Proctor, M.S. Gradus, K.A.Blair, D.E. Peterson, J.J. Kazmierczak, D.G. Addis, K.R. Fox, J.B. Rose, and J.P. Davis. 1994. A massive outbreak in Milwaukee of Cryptosporidium infection transmitted through the public water supply. The New England J. Medicine 331:161-167. Manja, K.S., M.S. Maurya, and K.M. K.M. Rao. 1982. A simple field test for the detection of faecal pollution in drinking water.  Bulletin of the World Health Organization. 60:797-801. Manja, K.S., R. Sambasiva, K.V. Chandrashekhara, K.J. Nath, S. Dutta, K. Gopal, L. Iyengar S.S. Dhindsa and S.C. Parija (2001) Report of Study on H2S Test for  Drinking Water, 96 Pages, UNICEF, New Delhi. Martins, M., G. Castillo, and B.J. Dutka. 1997. Evaluation of drinking water  treatment plant efficiency in microorganism removal by the coliphage, total coliform and H 2S paper strip tests. Water Sci. Technol . 35:403-407.  Nagaraju. D and J.C.V. Sastri Sastr i (1999) Confirmed faecal pollution to bore well waters of Mysore city. Environmental Geology , 38(4): 322-326. Olivieri, VP. (1983) Measurement of microbial quality. In: Assessment of  Microbiology and Turbidity Standards for Drinking Water. P.S. Berger and Y. Argaman, Eds., EPA Report No. EPA 570-9-83-001. Office of Drinking Water. Washington, D.C.

41

 

Pillai, J. K Mathew, R. Gibbs, and G. Ho. 1999. H2S paper strip method- A  bacteriological test for faecal coliforms in drinking water at various temperatures. Water Sci. Technol. 40:85-90. Popp, D.H. and E.A. Morris (1995) Some experiences with microbially influenced corrosion of pipelines. Materials Performance, pages 23-28, May. Ratto, A., B.J. Dutka, C. Vega, C. Lopez, A. A. El-Shaarawi El-Shaarawi (1989) Potable water  water  safety assessed by coliphage and bacterial tests . Water Research 23:253-255. Rijal, G.K., R.S. Fujioka and C.A. Ziel (2000) Evaluation of the hydrogen sulfide  bacteria test: A simple test to determine the hygienic quality of drinking water. Abstracts of the General Meeting of the American Society for Microbiology. Poster  Presentation (unpublished) Abstract Q354, Page 622. Washington, DC.

Amer. Soc. Microbiol.,

Sinton, L.W., L.W., R.K. Finlay and D.J. Hannah (1998) Distinguishing human from animal faecal contamination in water: A review.  New Zealand Journal of Marine &  Freshwater Research. Resear ch. 32(2): 323-348 Sivaborvorn (1988) On Development of Simple Tests for Bacteriological Quality of  Drinking Water (Water Quality Control Southeast Asia), Department of Sanitary

Engineering, Mahidol University. Thailand Centre File 3 P 83 0317 03. International Development Research Centre, Canada (www.idrc.ca/library/document/008918/ www.idrc.ca/library/document/008918/.. Sivaborvorn, K. and B.J. Dutka (1989) Coliphage presence in various types of potable water in Thailand, Water Pollution Research Journal of Canada, 24(4): 583-588. Sobsey, M.D. (2001) Microbial detection: implications for exposure, health effects, and control. In: Safety of Water Disinfection: Balancing Chemical and Microbial Risks, 2nd  Edition, G.F. Craun (ed.) ILSI Press, Washington, D.C. Townsend. S.A. 1992. The relationships between salmonellas and faecal indicator   bacteria concentrations in two pool in the Australia wet/dry tropics.  J. Appl.  Bacteriol. 73:182-188. Tyrell, S., Gardner, S., Howsam, P. and Carter, R. (1996) Biological removal of iron from well-handpump water supplies. (Manuscript). Venkobachar, C., D. Kumar, K. Talreja, A. Kumar Kum ar strip and I.test Iyengar (1994) Assessment of bacteriological quality using a modified H-2S . Aqua (Oxford  ), 43(6): ), 311-314. Wetzel, R.G. 2001. Limnology, 3 rd Ed., Academic Press, San Diego, CA Widdel, F. 1988. Microbiology and ecology of sulfate- and sulfur-reducing bacteria,  pp 469-586.  In:   Biology of Anaerobic Organisms, A.J.B. Zehnder (ed.), Wiley-Liss,  NY, NY.

42

 

WHO (2002) Indicators (draft document). OECD, World Health Organization, Geneva. Yasushi, K. (1998) Microbially induced corrosion and biodegradation, Corrosion Engineering, 47:909-913.