Monitoring microbial quality, contaminant sources and bacterial communities of bathing water in Finland

(1)

Dissertations in Forestry and Natural Sciences

ANANDA TIWARI

Monitoring Microbial Quality, Contaminant Sources and

Bacterial Communities of Bathing Water in Finland

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND

(2)

(3)

Dissertations in Forestry and Natural Sciences No 385

MONITORING MICROBIAL QUALITY, CONTAMINANT SOURCES AND BACTERIAL COMMUNITIES OF BATHING

WATER IN FINLAND

(4)

(5)

Ananda Tiwari

MONITORING MICROBIAL QUALITY, CONTAMINANT SOURCES AND BACTERIAL COMMUNITIES OF BATHING

WATER IN FINLAND

Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences

No 385

University of Eastern Finland Kuopio

2020

Academic dissertation

To be presented by permission of the Faculty of Science and Forestry for public examination in the Auditorium SN201 in the Snellmania Building at

the University of Eastern Finland, Kuopio, on October, 30, 2020, at 12 o’clock noon

(6)

Grano Oy Jyväskylä, 2020

Editors: Pertti Pasanen, Matti Vornanen, Jukka Tuomela, Matti Tedre

Distribution: University of Eastern Finland / Sales of publications www.uef.fi/kirjasto

ISBN: 978-952-61-3458-1 (Print) ISBN: 978-952-61-3459-8 (PDF)

ISSNL: 1798-5668 ISSN: 1798-5668 ISSN: 1798-5676 (PDF)

(7)

Author’s address: Ananda Tiwari

Finnish Institute for Health and Welfare Department of Health Security

P.O. Box 95

70101 KUOPIO, FINLAND email: ananda.tiwari@thl.fi Supervisors: Docent Tarja Pitkänen, Ph.D.

Finnish Institute for Health and Welfare Department of Health Security

P.O. Box 95

70101 KUOPIO, FINLAND email: tarja.pitkanen@thl.fi Docent Eila Torvinen, Ph.D.

University of Eastern Finland

Department of Environmental and Biological Sciences P.O. Box 1627

70211 KUOPIO, FINLAND email: eila.torvinen@uef.fi

Professor Jukka Pumpanen, Ph.D.

University of Eastern Finland

Department of Environmental and Biological Sciences P.O. Box 1627

70211 KUOPIO, FINLAND email: jukka.pumpanen@uef.fi

Reviewers: Dr. Muruleedhara Byappanahalli, Ph.D.

United States Geological Survey Great Lakes Science Center

Lake Michigan Ecological Research Station 1574 Kemil Road N 300 E

CHESTERTON, IN 46304, USA email: byappan@usgs.gov

(8)

Docent Isabel Douterelo Soler, Ph.D.

The University of Sheffield

Department of Civil and Structural Engineering Sir Frederick Mappin Building

Mappin Street, Sheffield, S1 3JD SHEFFIELD, UK email: i.douterelo@sheffield.ac.uk

Opponent: Docent Rauni Kivistö, Ph.D.

University of Helsinki

Department of Food Hygiene and Environmental Health

PL 66 (Agnes Sjöberginkatu 2) 00014 HELSINKI, FINLAND email: rauni.kivisto@helsinki.fi

(9)

Tiwari, Ananda

Monitoring Microbial Quality, Contaminant Sources and Bacterial Communities of Bathing Water in Finland

Kuopio: University of Eastern Finland, 2020 Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences 2020; 385 ISBN: 978-952-61-3458-1 (Print)

ISSNL: 1798-5668 ISSN: 1798-5668

ISBN: 978-952-61-3459-8 (PDF) ISSN: 1798-5676 (PDF)

ABSTRACT

Visiting natural bathing sites, such as those located by lakes and coastal areas, are major summertime activities and part of domestic and international tourism. It may indeed provide plenty of human health benefits, however the microbial contamination of such bathing sites poses a considerable human health risk. Therefore, the microbial contamination status of such areas is regularly monitored by enumerating Escherichia coli and intestinal enterococci as faecal indicator bacteria (FIB), in order to protect the health of beach users. This doctoral study has tested the ap- plicability of various microbial water quality monitoring methods towards providing tools and knowledge for the effective monitoring of bathing water.

The membrane filtration based method (ISO 9308-1:2000) and the miniaturised most probable number (MMPN) method (ISO 9308-3:1998) are recommended methods for enumerating E. coli by the European Union bathing water directive (EUBWD). However, the ISO 9308-1 method has proved to be unsuitable for monitoring environmental water due to highly sensitive culture media, which is unable to suppress the growth of background flora. The ISO 9308-3 method intended for polluted waters, such as wastewater, is not practical in the monitoring of bathing water. This study demonstrated that E. coli enumerated with the Colilert-18 Quanti- Tray (ISO 9308-2:2012) method as an alternative method gives equivalent results with the MMPN method. The Colilert-18 method provides relatively faster (18 hours) results, compared to the time requirement of the reference methods (48-72 hours). Therefore, Colilert-18 can be used for the monitoring of bathing water quality for regulatory purposes.

Membrane filtration-based method (ISO 7899-2:2000) is an official method for enumerating intestinal enterococci, according to EUBWD. This study confirmed that the performance of this method, utilizing Slanetz and Bartley medium, is suffi- ciently sensitive, specific, selective, and efficient for enumerating intestinal enterococci. This study calculated a 7% false-positive rate and a 20.9% false-negative rate of the ISO 7899-2:2000 method. However, the performance is affected by the num-

(10)

ber of presumptive bacterial colonies on the membrane. The high number of presumptive colonies (>100) per membrane increased the false-negative rate and reduced the false-positive rate. Contrary to that, a lower number of presumptive colonies (<10) per membrane increased the false-positive rate and reduced the false- negative rate.

Current bathing water monitoring methods cannot differentiate the source of pollution and, thus, assign equal risk levels for FIB from all the sources. Especially in cases where environmental sources of FIB occur, the monitoring result gives a false alarm of the human health risk. This study evaluated the role of aquatic vegetation on the decay of a FIB (E. faecalis), a viral indicator (MS2 coliphage), and environmental pathogen (Vibrio cholerae). It demonstrated different decay patterns and rates of tested targets in coastal bathing water, sediment, and vegetation.

Further, the bacterial communities from pollutant sources may affect the alloch- thonous community in ambient surface water. This study recorded the distinct bacterial communities and diversities from ambient surface water, municipal and in- dustrial effluent, as well as drinking water (DW) production, by using the bioin- formatic analysis of high-throughput sequencing of the 16S rRNA gene. Bacterial phyla Proteobacteria and Bacteroidetes were omnipresent. Phyla Firmicutes and Fusobacteria were solely detected from sewage. The share of potential health- related bacterial reads was higher in municipal effluent and surface water than in DW processing samples.

In summary, this doctoral thesis has evaluated the performance of current microbial bathing water monitoring techniques. It has enhanced our knowledge about the use of FIB for monitoring the microbial quality of bathing water and human health risk assessment. It concluded that molecular methods, such as real-time polymerase chain reaction and high-throughput sequencing, can be used for FIB enumeration, pathogen enumeration, and microbial source tracking.

(11)

ACKNOWLEDGMENTS

At first, I would like to express my cordial thanks to my main supervisor, Dr. Tarja Pitkänen, for providing me with the opportunity for conducting the doctoral study with her. The completion of doctoral study would not have been possible without her teaching, timely advice, and all other practical supports. My sincere thanks go to my supervisors, Dr. Eila Torvinen and Professor Jukka Pumpanen, for their advice and cooperation during the study. I express special thanks to Dr. Outi Zacheus, a bathing water expert of THL, for her valuable comments on the dissertation. My sincere thanks go to the leading researcher, Dr. Ilkka Miettinen, for providing the learning opportunities in the water microbiology laboratory of THL’s Expert Mi- crobiology Unit.

I would like to express my sincere thanks to all the co-authors of the manu- scripts, namely Dr. Ari Kauppinen, Ms. Anna-Maria Hokajärvi, Mr. Asko Vepsäläinen, Mr. Balamuralikrishna Jayaprakash and Ms. Sallamaari Siponen from THL; Dr. Jorge W Santo Domingo, Dr. Hodon Ryu and Mr. Michael Elk from USEPA; Dr. Aki Artimo and Mr. Osmo Puurunen from Turku Region Water; Dr.

Noora Perkola and Prof. Timo Huttula from SYKE; Dr. Jarkko Rapala from Ministry of Social Affairs and Health; Dr. Seija Kalso from Metropolilab and Dr. Seppo I.

Niemelä from Helsinki for their help during the manuscript production process.

My gratitude goes to all THL staff, namely Dr. Jaana Kusnetsov, Ms. Pia Räsänen, Ms. Anna Pursiainen, Ms. Ulla Usvalinna, Ms. Piia Airaksinen, Ms. Tiina Heiskanen, Ms. Tarja Rahkonen, Ms. Marjo Tiittanen, and Ms. Iina Laaksonen for their support during the doctoral study.

I would like to express my special thanks to the local health authorities who participated in different experiments and provided samples. My sincere thanks go to different projects of THL, Kaute Foundation (grant number 20190366), and Faculty of Science and Forestry of University of Eastern Finland (grant design 63/2020) that provided stipends and grants during my doctoral study. My sincere thanks go to my wife Toranta Kumari Banjara, son Aditya Tiwari and daughter Akriti Tiwari for their support during the doctoral study.

Kuopio, 11^th May 2020 Ananda Tiwari

(12)

LIST OF ABBREVIATIONS

AGR Artificial Groundwater Recharge ANOSIM Analysis of Similarity

ANOVA Analysis of Variance

APHA American Public Health Association BAV Beach Action Values

BEA Bile Aesculin Azide Agar BWD Bathing Water Directive CFU Colony Forming Unit cENT Culturable Enterococci DNA Deoxyribonucleic Acid

DW Drinking Water

DST Defined Substrate Technology DWPP Drinking Water Production Process

EC European Commission

EEA European Environment Agency ENT Enterococci

EU European Union

FIB Faecal Indicator Bacteria

GC Gene Copy

GI Gastrointestinal

GM Geometric Mean

GW Groundwater

HTS High-Throughput Sequencing iENT Intestinal Enterococci

ISO International Organization for Standardization LOD Limit of Detection

LOQ Limit of Quantification MF Membrane Filtration

MMPN Miniaturised Most Probable Number MPN Most Probable Number

MST Microbial Source Tracking MTF Multiple Tube Fermentation

NCBI National Center for Biotechnology Information

NEEAR National Epidemiological and Environmental Assessment of Recrea- tion

NMDS Nonmetric Multidimensional Scaling OTU Operational Taxonomic Unit

PCR Polymerase Chain Reaction PFU Plaque Forming Unit

PHRB Potential Health-Related Bacteria

(13)

QMRA Quantitative Microbial Risk Assessment qPCR Quantitative Polymerase Chain Reaction rDNA Ribosomal Deoxyribonucleic Acid RNA Ribonucleic Acid

rRNA Ribosomal Ribonucleic Acid

RT-qPCR Reverse Transcription- Quantitative Polymerase Chain Reaction S&B Slanetz & Bartley

Spp. Species

THL Finnish Institute for Health and Welfare TSA Tryptone Soya Agar

USEPA United State Environmental Protection Agency

UV Ultraviolet

VBNC Viable but not Culturable WHO World Health Organization WWTP Wastewater Treatment Plant

(14)

(15)

LIST OF ORIGINAL PUBLICATIONS

This thesis is based on data presented in the following articles, referred to by the Roman Numerals I-IV.

I Tiwari A, Niemelä SI, Rapala J, Kalso S, Pitkänen T (2016) Comparison of Colilert-18 with miniaturized most probable number method for monitoring of Escherichia coli in bathing water. J Water Health; 14(1): 121-31. doi:

10.2166/wh.2015.071

II Tiwari A, Hokajärvi A-M, Santo Domingo JW, Kauppinen A, Elk M, Ryu H, Jayaprakash B, Pitkänen T (2018). Categorical performance characteristics of method ISO 7899-2 and indicator value of intestinal enterococci for bathing water quality monitoring. J Water Health; 16(5): wh2018293.

doi:10.2166/wh.2018.293

III Tiwari A, Kauppinen A, Pitkänen T (2019). Decay of Enterococcus faecalis, Vib- rio cholerae and MS2 coliphage in a laboratory mesocosm under brackish beach conditions. Front. Public Health; 7:269. doi:10.3389/fpubh.2019.00269 IV Tiwari A, Hokajärvi A-M, Santo Domingo JW, Elk M, Jayaprakash B, Ryu H,

Siponen S, Vepsäläinen A, Kauppinen A, Puurunen O, Artimo A, Perkola N, Huttula T, Miettinen IT, Pitkänen T. Bacterial diversity and predicted enzymatic function in a multipurpose surface water system – from wastewater effluent discharges to drinking water production. Appl. Environ. Microbiol.

(Submitted)

(16)

AUTHOR'S CONTRIBUTION

I) The author participated in data analysis and manuscript writing, together with Tarja Pitkänen. All authors were involved in the review of the manuscript draft and approved the final version.

II) The author was responsible for data analysis and interpretation of the result.

He drafted the manuscript, together with Tarja Pitkänen. All authors were involved in the review of the manuscript draft and approved the final version.

III) The author participated in the planning of the experiments, together with Ari Kauppinen and Tarja Pitkänen. He was responsible for the execution of the experiments, sample collection, and data analysis. He prepared the first draft of the manuscript, together with the co-authors. All authors approved the final version of the manuscript.

IV) The author analyzed amplicon data, together with Balamuralikrishna Jaya- prakash. He also analyzed all metadata and interpreted the results. He drafted the first version of the manuscript together with Tarja Pitkänen.

All other co-authors commented on the draft and approved the final version.

(17)

1 INTRODUCTION

Visiting bathing sites located in lakes, rivers, estuaries, and coastal areas are major summertime activities (Tanaka 2009). About 80% of global tourism and 50% of all international tourism are located around coastal areas (UN 2017). The contribution is even more when the tourism contribution around freshwater lakes and rivers is taken into consideration (Steinman et al. 2017). The ocean-economy contributes between US$ 3-6 trillion each year i.e. employment and ecosystem services (UN 2017).

Finland is rich in surface water resources, as 10% of the nation is covered by lakes, ponds, and rivers (FWRF 2020). The livelihood of Finnish people is also close- ly tied to water activities. Mainly during summer, they enjoy it passionately around open beaches by swimming, sailing, boating, canoeing, cruising, and fishing. Every year, Finnish bathing sites serve thousands of foreign tourists, in addition to its population.

According to the European Environment Agency, (EEA 2018), Finland has listed 301 official bathing sites, of which 75% are inland and 25% are in the coastal area.

Besides these, there are numerous smaller bathing sites all over Finland available for recreation. Currently, more than 91 % of the large bathing sites in Finland have reported good or excellent bathing water quality (EEA 2018). In the future, changing environmental conditions, including the rise in the bathing water temperatures due to global warming, more shallow water depths at beaches, and increased run- offs from catchment areas, may lead to an increase in water contamination cases (Sinisi and Aertgeerts 2011). Besides serving as bathing sites, the lakes and rivers are an important source for the production of drinking water and agricultural water use.

Swimming has plenty of human health benefits as an aerobic exercise that boosts one’s metabolism and supports an active lifestyle (Tanaka 2009, Mestre- Alfaro et al. 2011, Sigwalt et al. 2011). It also provides a medium for relaxation, a psychological benefit that improves the mood of people (Sigwalt et al. 2011, Goto et al. 2018). However, a waterborne disease caught due to swimming in water having poor microbial quality can counterbalance the health and wellbeing benefits of swimming (Fewtrell and Kay 2015). Therefore, it is important to evaluate the microbial quality of bathing water in order to protect the health of beach users. Faecal indicator bacteria (FIB) are globally used for monitoring the microbial quality of bathing water with culture-based methods. This study contributes to improving microbial methods for the monitoring of bathing water quality levels.

(20)

18

2 MICROBIAL QUALITY OF BATHING WATER

2.1 INLAND AND COASTAL BATHING SITES

The general principles of hydrology, microbiology and public health concern in inland and coastal bathing sites are similar. The ingestion of a certain dose of viable pathogens can cause illness in both bathing site types. However, various environmental stress factors affecting the decay rate of enteric microbes, such as sunlight, water temperature, salinity, starvation, predation, particle interaction, and deposi- tion (Anderson et al. 2005, Byappanahalli et al. 2012, Lutz et al. 2013), may vary between inland and coastal waters.

Within a sea, the coastal waters are quite homogeneous in terms of temperature, salinity and nutrient levels, as all are physically connected. Such ecological parame- ters can vary more in the inland sites, i.e. from lake to lake, as these are physically separate (Dorevitch et al. 2010). Therefore, to some extent, the coastal water charac- ters are more predictable than inland bathing sites (Dorevitch et al. 2010). In general, lakes have less water volume for microbial dilution than coastal bathing sites, so bathers themselves and the surrounding land areas can have a high contribution to poor bathing water quality in inland sites. The global trend of human settlements is closer to the coastal region, but relatively more agricultural farms, pastures, and forests are nearby the inland water bodies. The land-use pattern of a catchment area influences the microbial quality of surface water (Ding et al. 2015). Rivers are a part of freshwater, but are different from lakes in terms of their size, shape, and movement of water. Generally, a river has cold water and a high sediment flow (Dorevitch et al. 2010).

The risk of waterborne illness is different between inland and coastal water. The illness rate of swimmers in seawater is estimated to be about two times greater than in freshwater (Dufour 1984, as cited in WHO 2003). The chance of gastrointestinal (GI) illness and infection of the ear and urinary tract can be higher in inland bathing sites, but the chance of skin rashes can be higher in coastal bathing sites, even on an equal level of the FIB count (Collier et al. 2015). Of pathogens, certain viruses and bacteria, such as Shigella spp., can be more common in inland bathing water than in coastal bathing sites (Wyn-Jones et al. 2011), and the occurrence of Vibrio spp. is common only in coastal bathing sites (Lutz et al. 2013). In some studies, the relation between pathogens and FIB is higher in coastal sites than in inland sites (Wu et al.

2011, Korajkic et al. 2018).

(21)

19

2.2 ETIOLOGICAL AGENTS CAUSING WATERBORNE ILLNESS

The severity of the bathing water-related health risk to bathers is affected by many factors:

(a) source of contamination; human faecal contamination poses the highest health risk due to having a chance of the involvement of enteric viruses, and the second-highest health risk is from bovine faecal contamination due to having a chance of the involvement of a wide range of zoonotic agents (i.e. originating from animal sources) like Campylobacter jejuni, E. coli O157: H7, Salmonella spp., Giardia lamblia, and Cryptosporidium parvum (USEPA 2009, Soller et al. 2010).

(b) the health status of the defecating host; the etiological agents are found only in faecal material of infected hosts (Savichtcheva and Okabe 2006),

(c) infectious dose; different pathogens have different infection doses. There is a higher chance of infection from an agent having a lower infection dose than having a higher infection dose (WHO 2003),

(d) viability of the pathogen; a more viable agent has higher chances of causing infection than unviable agents. Treated sewage and hospital effluents can have lower numbers of viable pathogens than raw sewage because the use of antibiot- ics and disinfectants may damage the microbes (WHO 2003),

(e) movement of water; for example, pathogens can remain longer in stagnant water than in moving water (Dorevitch et al. 2010),

(f) the duration of water contact of bathers; the longer the bathing duration, the higher the chance of microbial infection (Wade et al. 2010). Many bathers, mainly children (accidentally) swallow several times more water than adults. The fre- quency of the ingestion of water can be high if the duration of bathing is longer.

(g) the immunity of the bathers

(h) mode of contact, different pathogens affect different body parts; for example, norovirus infection occurs only after the ingestion of virus particles (Pond 2005).

The recreational exposure to polluted water increases the risk of waterborne diseases, such as diarrhoea, fever, nausea, ear and eye infections, and rashes on the skin (WHO 2003, Colford et al. 2007, Wade et al. 2010, Collier et al. 2015). The majori- ty of bathing water diseases are self-limiting or even asymptomatic, but in some cases, the illness can be life-threatening, such as typhoid, cholerae, meningitis, poli- omyelitis, hepatitis, myocarditis, meningoencephalitis, and schistosomiasis (WHO 2003).

The etiological agents of bathing related illnesses fall under wide taxonomic ranges of bacteria, viruses, protozoa, eukaryotes and harmful algal blooms (WHO 2003, WHO 2009, Collier et al. 2015). Human faecal contamination through wastewater effluent outfall, an accidental release of raw sewage and urban runoff during rain and snowmelt are major sources of enteric pathogens. Further, the high number of infected bathers may also deteriorate the bathing water quality (Elmir et al. 2007). Bathers can have different levels of pathogenic shedding with them, such

(22)

20

as those who currently suffer from a contagious disease or have recently recovered from one or shedding infectious agents on asymptomatic conditions (Collier et al.

2015). Further, faecal contamination from farm animals through agricultural runoff and from wild animals through rural runoff can increase the risk of zoonotic pathogens (USEPA 2009, Pitkänen 2013).

2.2.1 Bacteria

Several bacteria, (as summarised in Table 1) like Campylobacter spp., Salmonella spp., pathogenic E. coli, Shigella spp., Vibrio spp. and Yersinia enterocolitica poses public health risks in bathing sites (Schönberg-Norio et al. 2004, Hokajärvi et al. 2013, Baker-Austin et al. 2016). A surveillance study reported Campylobacter jejuni, E. coli O157: H7 and Shigella sonnei as the major bacterial agents of bathing water outbreaks in the USA during the year 2009-2010 (Hlavsa et al. 2014). Many of the bacterial agents; C. jejuni, E. coli O157:H7 and Salmonella spp. have zoonotic sources. Bac- terial group (E. coli O157: H7 and E. coli O111 and Shigella sonnei) accounted for one- third of the outbreaks (n=21), 16% of cases (n=479) and 82% of hospitalized cases (n=22) related to recreation outbreak in the USA during 2011-2012 (Hlavsa et al.

2015). While most of the bacterial pathogens cause gastroenteritis, the members of Mycobacterium avium complex, Vibrio vulnificus, and other Vibrio spp. increase the risk of skin rash and infection of wounds and soft tissues (WHO 2003, Baker-Austin et al. 2016). In addition to known sources of human pathogens, certain bacterial pathogens, such as Campylobacter, Salmonella, and Shigella, are also reported in environmental matrices (e.g., co-occurring with nuisance algae Cladophora) (Ishii et al., 2006, Byappanahalli et al., 2009), presumably unrelated to known fecal sources.

However, epidemiological studies to demonstrate the pathogenicity of such environmental strains are difficult to conduct.

Although etiological agents originating from faecal contamination are a major focus of this thesis; still, autochthonous agents, such as toxin-producing Cyanobacte- ria spp. and pathogenic Vibrio spp., may also pose a health risk to bathers. Hlavsa et al. (2014) demonstrated cyanobacterial toxins as being responsible for 11 out of 24 recreational and bathing water outbreaks during 2009-2010 in the USA. Several Vibrio species with pronounced pathogenic potential (non-O1/non-O139 serotype of Vibrio cholerae and Vibrio vulnificus, Vibrio parahaemolyticus and Vibrio alganolyticus) are reported from Baltic Sea (Baker-Austin et al. 2012, Lutz et al. 2013, Baker-Austin et al. 2016). In recent years, Vibrio infection cases have increased around Finnish coastal areas, due to heatwaves, and have shown that a rise in surface water temperatures, together with elevated nutrient levels in water, may exacerbate more Vibrio cases in the future in Finnish bathing sites (Baker-Austin et al. 2016, THL 2018). Typically, Vibrio species are autochthonous, yet contamination from infected hosts, mainly during an outbreak, may enlarge their numbers in ambient water (Islam et al. 2020).

(23)

21 2.2.2 Protozoa

Pathogenic protozoa, mainly the species Giardia lamblia and Cryptosporidium parvum, are commonly reported during bathing water outbreaks (Hlavsa et al. 2014). These were also reported from Finnish surface water (Hörman et al. 2004). Parasites (Trematoda causing cercarial dermatitis, Cryptosporidium spp., Giardia intestinalis) accounted for 19% of total outbreaks (n=21), 15% of total cases (n=479) and none of the infections were severe as there was no hospitalization (n=22) from recreational water outbreaks in the USA during 2011-2012 (Hlavsa et al. 2015). These enteric protozoa are obligate parasites and their cysts and oocysts are highly resistant to environmental stress factors. Other known protozoan parasites at bathing sites are Entamoeba histolytica, Cyclospora cayetanensis, Isospora belli, Microsporidia, Ballantidium coli, Toxoplasma gondii and Naegleria fowleri (WHO 2003, WHO 2009). Protozoan parasites originate from both human and animal faecal sources. The most common illness to bathers caused by these parasites is gastroenteritis (Table 1).

2.2.3 Viruses

Viruses are major causes of bathing related illness (Sinclair et al. 2009, Wyn-Jones et al. 2011, Soller et al. 2015, Griffith et al. 2016). A study analyzed a total of 1410 bath- ing water samples collected from both inland and coastal bathing sites in eight EU countries and reported 40% of them as being positive for viral pathogens (Wyn- Jones et al. 2011).

Viruses are mostly host-specific and, thus, human infecting viruses originate from human faecal contamination (Harwood et al. 2014) (Table 1). As enteric viruses are adapted to the hostile gut environment, even an effective wastewater treatment may not be able to destroy them completely. Once reaching ambient water, the viruses can tolerate wide environmental stresses and persist long in the water and sediment (Thurston-Enriquez et al. 2003), although the persistency, pathogenicity, and infectivity of viruses vary over different virus types (Sinclair et al. 2009).

Adenovirus and norovirus are the most frequently reported during bathing water outbreaks (Sinclair et al. 2009, Fewtrell and Kay 2015, Hlavsa et al. 2014, Kaup- pinen et al. 2017). The contribution of adenovirus and norovirus on recreational water outbreaks accounted for 14% of the total outbreaks (n=21), 18% of total cases (n=479) and 9% hospitalized cases (n = 22) in the USA during 2011-2012 (Hlavsa et al. 2015). Adenovirus infections are mostly mild and self-limiting gastrointestinal cases (GI) or respiratory illnesses (Sinclair et al. 2009). Norovirus causes vomiting, diarrhoea, a slight fever and headache (Sinclair et al. 2009). Enterovirus (poliovirus, coxsackievirus, and echovirus) are another group of viruses that may be found in bathing sites, transmitted through the faecal-oral route (Sinclair et al. 2009). These may cause conjunctivitis, respiratory or GI illness, meningitis, paralysis, myocarditis, head-foot and mouth disease (Sinclair et al. 2009). Further, viruses like echovi- rus, coxsackieviruses, astroviruses, and hepatitis A (Table 1, Sinclair et al. 2009,

(24)

22

Kauppinen et al. 2016), sapprovirus, paraechovirus, enterovirus 69-91, reovirus, hepatitis E virus, rotavirus, astrovirus, picobirnavirus, and coronavirus may also be found in bathing sites (WHO 2003). Among them, hepatitis E virus is a zoonotic pathogen (WHO 2003).

Table 1. The bathing water-related etiological agents, their probable sources and transmission pathways. GI = gastrointestinal. (WHO 2003, Pond 2005).

Agent Illness Probable source Transmission pathway

Bacteria Campylobacter

spp. Gastroenteritis, fever Human and animals Ingestion Enteropatho-

genic E. coli Bloody diarrhoea, abdominal

cramp Human and animals Ingestion

Helicobacter

pylori Gastritis, abdominal pain Human and animals Ingestion Legionella spp. Pneumonia, gastroenteritis Natural Inhalation Leptospira spp. Fever, headache, vomiting,

jaundice Natural and animals

Salmonella spp. Gastroenteritis, fever, pain Human and animals Ingestion Mycobacterium

avium Respiratory disease Environmental Inhalation/contact Vibrio vulnificus Infection in pre-existent open

wound Environmental Wound infection

Shigella spp. Bacillary dysentery, ab-

dominal pain Human Ingestion

Viruses

Adenovirus Gastroenteritis, respiratory

disease Human Ingestion, inhalation

Noroviruses Gastroenteritis Human Ingestion

Rotaviruses Gastroenteritis Human Ingestion

Coxsackievirus Mild febrile illness to myocarditis and other more serious diseases

Human Ingestion

Enteroviruses The central nervous system, ocular and respiratory infections

Human Ingestion

Echovirus Diarrhoea, secretions

from the eyes or throat Human Ingestion Hepatitis A

virus Liver disease Human Ingestion

Hepatitis E

virus Liver disease Human and animals Ingestion

Protozoa / trematodes Cryptosporidi-

um Diarrhoea, abdominal pain,

fever Human and animals Ingestion

Giardia Diarrhoea, abdominal cramp Human and animals Ingestion Microsporidia GI illness, diarrhoea Human and animals Ingestion Naegeria

fowleri Meningoencephalitis Natural Contact

Schistosoma

spp. GI illness, haematuria Human Ingestion, Contact

Entamoeba

histolytica Amoebic dysentery Human Ingestion

(25)

23

2.3 FAECAL INDICATOR BACTERIA

Faecal indicator bacteria (FIB) are a group of bacteria mostly present in the gut of warm-blooded animals including human beings and are used for predicting faecal contamination and the presence of enteric pathogens. The abundance of pathogens in water is usually low and each pathogen has a different method of detection (Field and Samadpour 2007, Harwood et al. 2014). Monitoring each of them can be time-consuming, expensive, and inefficient (WHO 2009, Valente et al. 2010); so, the routine monitoring of pathogens is impractical for the monitoring of bathing water quality levels. Therefore, FIB are used worldwide as surrogates for faecal contamination and human health risks due to poor hygiene (Field and Samadpour 2007).

An ideal faecal indicator microbe would have particular characteristics, as presented in Table 2.

Table 2. Characteristics of ideal faecal indicator (Dukta 1973, USEPA 2015).

1. Ubiquitous on the faeces of a warm-blooded animal, including human beings 2. Have a higher number than faecal pathogens in a gut environment

3. Have an easy, rapid and inexpensive enumeration method

4. Have enough epidemiological data showing correlation to human health risks 5. Have a similar decay rate as faecal pathogens in different stress factors 6. Have good correlation with faecal pathogens

7. Do not grow in the environment 8. Non-pathogenic

9. Identify the source of contamination

FIBs are higher in numbers in the gut environment and transmitted through the same transmission route as pathogens (Korajkic et al. 2018). The positive relation between GI illness and other bathing illness cases with FIB counts, E. coli counts mainly in freshwater and enterococci (ENT) counts in both fresh and marine water have been reported (Cabelli et al. 1982, Wade et al. 2003, Wade et al. 2008). Wade et al. (2003) calculated that a log10 increase in ENT count was associated with 1.34 times (95% confidence intervals: 1-1.75) increases in relative risk of infection in marine waters and a log¹⁰ increase in E. coli count was associated with a 2.12 times (95% confidence intervals: 0.925-4.85) increase in relative infection risk in freshwa- ters. However, mainly during the non-point sources of contamination, the poor relation between FIB counts with bathing water-related illness cases has been reported (Boehm et al. 2009, Colford et al. 2007). Such a poor correlation makes the FIB count interpretation and assessing the human health risk, based on the FIB counts, difficult (Fujioka and Byappanahalli 2003, Colford et al. 2007, Fujioka et al. 2015).

(26)

24

2.3.1 Relationships between FIB and pathogens

FIB are common in the intestine of all warm-blooded animals (healthy, asymptomatic carriers of some pathogens and infected hosts), but pathogens are common only to the infected host (symptomatic, asymptomatic and recently recovered).

Further, the host ranges of FIB are wide, FIB are released from all warm-blooded animals, including human beings, but certain pathogens, like human-infecting viruses, are released mainly from infected humans. Therefore, FIB are detected in higher numbers in water bodies than pathogens. A good correlation between enteric viruses and FIB has been reported at bathing sites with human faecal contamination and a good correlation between zoonotic pathogens and FIB can be at bathing sites with animal faecal contamination (Harwood et al. 2014). The illnesses from autochthonous microbes, such as Vibrio spp. and toxins produced by Cyanobacteria, may have poor relationships with FIB counts (Hlavsa et al. 2014, Islam et al. 2020).

Various other reasons for poor relationships can be found, such as different decay rates and transport characteristics, that may affect the relationship between pathogens and FIB (Wu et al. 2011).

The persistence or probable growth of FIB in environmental habitats, such as soil, sediments, and vegetation, has been reported (Whitman et al. 2003, Badgley et al. 2011, Byappanahalli et al. 2012). Such environmental habitats may work as a source and sink of FIB (Ishii and Sadowsky 2008, Byappanahalli and Ishii 2011, Byappanahalli et al. 2012). The release of FIB from the environmental source has poor relation with enteric pathogens and poses a false positive alarm concerning the human health risk (Badgley et al. 2011). Environmental disturbances due to tides, winds, animal movements, and anthropogenic disturbance, such as bathing and water sports, can stir FIB from the sediment and aquatic vegetation and may increase FIB counts in water (Boehm et al. 2009). In contrast, the long persistence of enteric pathogens over FIB in an environmental habitat also weakens the relation between FIB and pathogens. Due to a false-negative FIB alarm, pathogens may jeopardize the health of beach users (Anderson et al. 2005, Badgley et al. 2010). Fur- ther, currently used FIB cannot differentiate the source of contamination. Therefore, the criteria for ideal FIB cannot yet be fulfilled completely.

The pathogens, mostly bacteria, viruses, and protozoa, have a wide variation in the taxonomic range, cell structure, morphology, and physiology and they can have different decay rates than FIB (Anderson et al. 2005, Byappanahalli et al. 2012, Lutz et al. 2013). Therefore, these wide taxonomic ranges of microbes might have a dif- ferent response towards environmental stresses, such as unfavourable pH values, solar radiation, salinity, predation and temperature and nutrients concentration (Anderson et al. 2005, Korajkic et al. 2018). FIB can be more sensitive to inactivation in the wastewater treatment process and by sunlight, than viruses and protozoan parasites (Sinclair et al. 2009). However, in fresh faecal contamination, there can be a high correlation between FIB with a human enteric virus and parasitic protozoa (Hartel 2011).

(27)

25 2.3.2 Coliform bacteria and E. coli

Coliforms are a group of bacteria that express the enzyme β-D-galactosidase (ISO 9308-2 2012). In the ISO 9308-2 method, coliform positive wells produce a yellow colour and are recognized by visual inspection. These are a Gram-negative, non- spore-forming, oxidase negative, rod-shaped bacteria that can grow in aerobic or facultatively anaerobic conditions in the presence of bile salt (Chao 2006, ISO 9308-2 2012). The earlier definition, based on lactose fermentation, defines coliform as a group of bacteria having the capacity to ferment lactose into gas and acid within 48 hours at 32-35°C (APHA 1989, Chao 2006). These bacteria belong to different genera within the Enterobacteriaceae family, namely: Escherichia, Klebsiella, Citrobacter, Haf- nia, and Enterobacter. This group of coliform bacteria was used for a long time as FIB for the regulatory monitoring of bathing water quality (EC 1976). However, the use of total coliforms as FIB has some limitations, such as their ability to grow in a natural environment (Carrillo et al. 1985, Byappanahalli et al. 2006) and a lack of correlation with faecal pathogens and numbers of bathing water illness cases (Wade et al.

2003, Korajkic et al. 2018). The current bathing water monitoring method (EC 2006) does not include this group of bacteria for regulatory monitoring.

Faecal coliforms are a subgroup of total coliforms that are more specific to the faecal origin. The members of this group of bacteria are capable of growing in the presence of bile salts. These are oxidase negative, and produce acid and gas from lactose within 48 hours at 44 ± 0.5°C (APHA 1989). Faecal coliform bacteria can tolerate higher temperatures than other members of coliforms, so faecal coliforms are also called thermotolerant coliform bacteria. Some members of faecal coliform bacteria, like Klebsiella, can persist for a long time in environmental water.

E. coli is a major species of faecal coliform group, which is the most reliable among the coliform group as FIB (Korajkic et al. 2018). E. coli expresses both β-D- galactosidase and β-D-glucuronidase enzymes (ISO 9308-2 2012). The β-D- glucuronidase activity can be measured with fluorescence produced under ultraviolet light (365 nm; ISO 9308-2:2012). Further, E. coli is a species among the group of coliform bacteria which can produce indole from tryptophan within (21±3) h at (44±0.5) °C. It provides a positive result in the methyl red test and can decarbox- ylate 1-glutamic acid, but is not able to produce acetyl methyl carbinol, utilize citrate as the sole source of carbon or grow in KCN broth. However, some strains of E.

coli, such as E. coli O157:H7 express only β-D-galactosidase activity, but do not have β-D-glucuronidase activity. E. coli ferments lactose at 44 °C and produces indole from tryptophane (Chao 2006, ISO 9308-2 2012). It is a primary indicator of choice of inland bathing water quality monitoring (Wade et al. 2003) and is also applied to coastal bathing water monitoring. E. coli has been used as a FIB for regulating the monitoring of bathing and recreational water quality (EC 2006, USEPA 2012).

(28)

26

2.3.3 Enterococci and intestinal enterococci

Enterococci (ENT) are Gram-positive, non-spore-forming, obligatory fermentative, chemoorganotrophic coccoidal bacteria, belonging to the genus Enterococcus (Boehm and Sassoubre 2014). This group of bacteria grows at a temperature range of 10°C - 45°C in 6.5% NaCl at pH 9.6 and survives for 30 minutes at 60°C, hydro- lyzes aesculin, is resistant to sodium azide and reduces triphenyl tetrazolium chloride (Byappanahalli et al. 2012). ENT are a primary indicator of choice for the monitoring of coastal bathing water (Wade et al. 2003) and are also applied to the monitoring of inland bathing water. ENT have been used as FIB for regulating the monitoring of bathing and recreational water quality (EC 2006, USEPA 2012).

The term enterococci and intestinal enterococci (iENT) have been used inter- changeably (Leclerc et al 1996, Byappanahalli et al. 2012). However, iENT are a subgroup of ENT majorly belonging to four species Enterococcus faecalis, Enterococcus faecium, Enterococcus durans and Enterococcus hirae (WHO 2003). These species grow at 44 °C at aerobic conditions, hydrolyze 4-methylumbelliferyl-b-D-glucoside in the presence of thallium acetate, nalidixic acid and 2,3,5-triphenyl-tetrazolium chloride (ISO 7899-2 2000). Among them, E. faecium and E. faecalis have a high prevalence in human faeces (Moore et al. 2006; Layton et al. 2010). Enterococcus is a large genus, and not all species originate from faeces (Byappanahalli et al. 2012). For example, Enterococcus mundtii, Enterococcus casseliflavus, Enterococcus aquimarinus, and Entero- coccus sulfureus are reported from vegetation sources (Mundt and Hinkle 1976, Moore et al. 2006, Byappanahalli et al. 2012).

2.4 ALTERNATE FIB

Bacteria, protozoa, and viruses of a wide taxonomic variation have a different cell structure, size, and defence mechanism against external stress factors (Field and Samadpour 2007, Griffith et al. 2016), so their decay rates in various environmental stress conditions are different (Griffith et al. 2016, Korajkic et al. 2018). So, FIB used for monitoring the microbial quality of surface water does not always relate to pathogens and parasites (Savichtcheva and Okabe 2006). That is why, different alternate indicators, such as F+ RNA coliphages, somatic coliphages, and bacterial genera Bacteroides, Prevotella, Catellicoccus, Clostridium, Bifidobacterium, Staphylococcus and Brevibacterium, have been used for the monitoring of water quality (Savichtche- va and Okabe 2006, USEPA 2012, Fujioka et al. 2015, Korajkic et al. 2018).

Viruses generally have longer survival rates in ambient environments than FIB (Thurston-Enriquez et al. 2003). Viruses can be detected in surface water, even when the FIB numbers are below the safe limit according to current monitoring protocols (Sinclair et al. 2009, Kauppinen et al. 2017). Further, a stronger correlation between F-specific coliphages and somatic coliphages with GI illness than bacterial indicators has been demonstrated (Wade et al. 2003, Griffith et al. 2016). Such findings

(29)

27 justify the need for a separate viral indicator for the regulatory monitoring of bathing water (USEPA 2015, WHO 2018). Currently, F-specific and somatic coliphages have been used as an alternate feacal indicator (Griffith et al. 2016). Coliphages have similar physical structure, composition, morphology and survival characteristics in the environment; and these are suggested as a potential viral indicator for faecal contamination for monitoring bathing water (USEPA 2015). F-specific coliphage is one of the most used virus indicators for water quality testing (USEPA 2015). Its detection and quantification methods are simple, reliable, rapid, and inexpensive (USEPA 2015). Somatic coliphages are more persistent against environmental stress factors and have a higher concentration in faeces than F-specific coliphages (USEPA 2015). Norovirus resembles more F-specific coliphages than somatic coliphages due to having single-strand RNA, while adenovirus resembles more somatic than F- specific coliphages due to having double-stranded DNA (USEPA 2015).

However, the current FIB cannot differentiate the source of faecal contamination (Stoeckel and Harwood 2007). Strict anaerobic gut bacteria genera, such as Bac- teroides, Catellicoccus, Brevibacterium and Prevotella that are highly host-specific, are used as alternative faecal indicators and for microbial source tracking (McLellan and Eren 2014). Also the bacterial species Clostridium perfringens is highly resistant to environmental stress factors and have a limited potential to proliferate in the ambient environment. This species has been used as an alternate faecal indicator for the monitoring of bathing water (Fujioka and Shizumura 1985, Okabe and Shimazu 2007, Fujioka et al. 2015).

During bathing seasons, a high number of bathers may visit bathing sites in a single day. There can be a high chance of a cross-contamination of pathogens among bathers. However, the currently used FIB cannot indicate such cross con- taminations. The bacterial species Staphylococcus aureus has been studied as an indicator of direct contamination from bathers to bathers (Elmir et al. 2007). The species has a high survival rate in coastal water. It is more resistant to chlorine and salinity than bacteria from the total coliform group (Elmir et al. 2007). Further, a positive correlation between the numbers of S. aureus in water with skin, ear, and respiratory tract illness was demonstrated at a coastal bathing site (Elmir et al. 2007). In addition to all the above-mentioned microbial groups as an alternate indicator, direct enumeration of pathogens, mainly Salmonella spp. and enteric viruses, are also con- sidered as a sign of faecal contamination (Savichtcheva and Okabe 2006).

(30)

28

3 METHODS FOR MICROBIOLOGICAL

BATHING WATER QUALITY CHARACTERIZA- TION

The microbial quality of bathing water has been monitored for decades by measur- ing indicator bacteria that are common in the gut of warm-blooded animals, including human beings. Historically, microbial monitoring was started with microscopic techniques, but now culture-based methods and molecular methods are the two most used water quality monitoring methods. The strengths and weaknesses of variants of these two most commonly used methods are listed in Table 3.

Table 3. Strengths and weaknesses of different microbiological monitoring methods. MF= membrane filtration, MPN = most probable number, PCR = polymerase chain reaction, qPCR= quantitative polymerase chain reaction, RT-qPCR= reverse transcription-quantitative polymerase chain reaction.

Method Strengths Weaknesses

Culture MF methods

Widely accepted gold standard, econom- ically cheap, highly standardized, and easy to operate and interpret

Long incubation time, not suitable for unculti- vable microbes, not suitable for samples having high suspended particles

Culture MPN methods

Widely accepted gold standard, econom- ically cheap, highly standardized, and easy to operate and interpret

Long incubation time, not suitable for unculti- vable microbes, sometimes identification can be subjective due to the phenotypic method

PCR Highly specific Not quantitative

qPCR (DNA target)

Fast, highly specific, can enumerate different microbes by changing target primers.

May detect dead and non-viable microbes, high installation and operation costs, even a small laboratory mistake can alter the result RT-qPCR

(RNA target)

Count only viable cells, no need to incu- bate, highly sensitive, can differentiate the source of faecal contamination

Cell quantification is not possible

High- throughput sequencing

Provides information on a large group of microbes at the same time, gives an idea about the microbial community structure

Resource intensive. Not fully developed yet.

The taxonomic genes of some species are highly conserved, so cannot identify the genus and species level

3.1 CULTURE-BASED METHODS

Culture-based methods are the accepted gold standard for enumerating FIB and many pathogens. These methods are relatively inexpensive, highly standardized, and easy to operate and interpret. The culture-based methods can be different types, but membrane filtration based, the most probable number (MPN) method and chromogenic substrate methods are most widely used. The targeted indicators are selectively isolated and incubated in nutrient-rich media (Niemelä et al. 2003, Pitkänen et al. 2007). However, these methods are often criticized for enumerating

(31)

29 only the viable organism. Even within the viable group, not all of them are culturable (Pitkänen et al. 2013). Further, these methods require a relatively long incubation time, ~18-48 hours at the minimum, depending on the target FIB. These methods rely on biochemical or immunological methods of identification. Sometimes, such identification can be subjective and have personal biases (Niemelä et al. 2003, Pitkänen et al. 2007).

Membrane filtration (MF) is a widely used water sample concentration tech- nique utilized before the culture-based quantitation of microbial targets (ISO 7899-2 2000). By using this method, indicator bacteria from water samples are initially concentrated into a membrane filter with a help of filtration, and then the membrane is transferred onto a petri dish having a solid selective growth medium specific for the target organism (ISO 7899-2 2000). Then the petri dish containing the culture medium with the membrane filter is incubated for about 35-45 ⁰C, depending on the medium used. The number of colonies grown on the membrane filter is counted and expressed as the colony-forming unit CFU/100 ml sample (ISO 7899-2 2000). The multiple tube fermentation (MPN) method is the next common culture- based method used for microbial enumeration. As the MF method is not suitable for the water samples having high turbidity, the MPN method overcomes the limitation. In this method, water samples are poured into a liquid medium, and the microbial counts are expressed in MPN/100ml (ISO 9308-2 2012). Viruses and bacte- riophages are concentrated from water samples by using electropositive filters, electronegative filters, and ultrafilters (Cashdollar and Wymer 2013). Host cells are needed for culturing viruses and the viruses can be isolated from the resulting plaques (Cashdollar and Wymer 2013).

3.1.1 E. coli monitoring

The membrane filtration-based ISO 9308-1 2000 method and the miniaturised most probable number (MMPN) based ISO 9308-3 1998 method are official methods for enumerating E. coli; according to bathing water directive (BWD EC 2006). However, the ISO 9308-1 (2000) method has been modified completely to ISO 9308-1 (2014).

The earlier version of the method operated on two steps and used TTC Tergitol® 7 agar and rapid test using TSA/TBA agar. The recent version of ISO 9308-1 (2014) of the method uses Chromogenic Coliform Agar (CCA) media. The ISO 9308-1 (2000) method has been criticized as not being suitable for monitoring environmental water samples having a high background flora (Jozić et al. 2018, WHO 2018). The recent version is intended mainly for monitoring the high quality of drinking water (Jozić et al. 2018, WHO 2018). The next official method for enumerating E. coli, ISO 9308-3 1998 is not used in laboratories serving local health protection authorities of Finland. Both methods need about 48-72 hours to confirm the results.

Colilert-18 Quanti-Tray (ISO 9308-2 2012) is relatively more rapid than the reference methods and gives a result within 18 hours. It is an accepted method for use in monitoring the quality of drinking water in many countries (Niemelä et al. 2003, Pitkänen et al. 2007). In both Colilert-18 and MMPN methods, the detection of E. coli

(32)

30

is based on the fluorogenic reaction (positive for β-glucuronidase) (Lebaron et al.

2005, Valente et al. 2010). The Colilert-18 method also detects coliform bacteria, based on chromogenic reaction, and the E. coli result is recorded by both chromo- genic and a fluorogenic reaction to occur and are detected in the same well (Nie- melä et al. 2003, Lebaron et al. 2005, Valente et al. 2010). The MMPN method enu- merates E. coli, but no other coliform bacteria groups (Lebaron et al. 2005, Valente et al. 2010).

3.1.2 Enterococci monitoring

Different varieties of techniques are available for ENT enumeration (Domig et al.

2003). Most of the available methods are based on Slanetz and Bartley (S&B) agar (Slanetz and Bartley 1957) and the Kanamycin Aesculin Azide (KAA) agar (Domig et al. 2003). EUBWD has assigned two methods; ISO 7899-1: 1998 and ISO 7899-2:

2000 for the selective isolation and enumeration of ENT from bathing water. The first one (ISO 7899-1 1998) is based on the MPN method, using the miniaturized 96- well system to enhance precision. The Miniaturized MPN enumerates iENT on a basis to their capacity to grow at 44 ± 0.5 °C and of hydrolyzing 4- methylumbelliferyl-b-D-glucoside in the presence of thallium acetate, nalidixic acid, and 2,3,5-triphenyltetrazolium chloride, in the liquid medium. The presence of iENT is visualized by the emission of fluorescence in 36-72 h. The second method (ISO 7899-2 2000) is based on membrane filtration and confirms iENT in two steps.

At first, the bacteria retained on the membrane filter are incubated for 44 ± 4 h at 36

±2 °C on a S&B medium. The triphenyltrazolium chloride (TTC) in S&B medium is reduced to formazan and forms red colonies. All new red or maroon coloured colonies are accepted as presumptive ENT. The presence of iENT is then confirmed on bile esculin azide (BEA) agar (incubating for 2 h at 44 ± 0.5 °C). ENT hydrolyze esculin to esculetin, react with ferric citrate in the medium to produce a black phenol- ic iron-complex giving esculinase-positive colonies a brown-black halo. The iENT is confirmed, based on dark brown to black colonies produced on a BEA agar medium.

3.2 MOLECULAR METHODS

Methods, such as polymerase chain reaction (PCR), quantitative polymerase chain reaction (qPCR), reverse transcription-quantitative polymerase chain reaction (RT- qPCR), next-generation sequencing (NSG), nested PCR, digital PCR (dPCR), multiplex PCR, and microarrays, are popular molecular tools in health-related water microbiology (Vierheilig et al. 2015, Zhang and Liu 2019). The use of molecular methods has made it possible to enumerate the large range of pathogens directly from surface waters without culturing them. Monitoring all microbes with the culture-based method is a great challenge as most microbes in the natural environments are difficult to culture.

(33)

31 Molecular methods, mostly qPCR and RT-qPCR, ensure more rapid results than culture-based methods, due to a lack of a microbial multiplication phase.

Rapid methods have high importance for the monitoring of bathing water quality, as the microbial quality of bathing sites may change drastically within 24-48 hours, i.e. during incubating and enumerating FIB with the culture-based methods. In the United States, U.S. EPA has approved a qPCR-based enterococci gene copy enumeration method for the monitoring of regulatory bathing water (USEPA 2012).

The Entero1 assay is the recommended qPCR assay for the purpose (USEPA 2012).

However, the Entero1 primers targets all known species of Enterococcus genera (Ludwig and Schleifer 2000, Haugland et al. 2005), which is equivalent to the total enterococci enumerated with membrane-Enterococcus Indoxyl-β-D-Glucoside (mEI) agar method (USEPA 2012). Earlier studies reported that molecular markers can predict bathing water-related human health risks and the presence of a human virus (mainly norovirus) better than the culture-based methods used in the USA (Wade et al. 2008, Schoen et al. 2011). A strong correlation between culturable cells of E. coli with its qPCR markers was recorded earlier (Shrestha and Dorevitch 2019).

In comparison to culture-based methods using phenotypic characterization, the molecular qPCR methods can have high specificity, as they use genotypic characterization (Savichtcheva and Okabe 2006). However, the partial confirmation may cause biases during reading the positive cases (Niemelä et al. 2003, Pitkänen et al.

2007) in both the phenotypic and genotypic confirmation. Most of the molecular tools (PCR, qPCR, and amplicon-based NGS) use the 16S rRNA gene for the identification of microbes. These genes are highly conserved among bacteria and archaea and are useful for taxonomic identification of microbes. Instead of this gene, different functional genes can also be used for the identification of microbes.

Also molecular methods have some shortcomings. For example, mainly in some genera, the hypervariable regions of the 16S rRNA gene are relatively more conserved, so the base pairs between two neighbouring species are more than 99%

similar, even on the full 16S rRNA gene. Therefore, in some genera, this approach may not be enough to distinguish the variation on the lower taxonomic level. Fur- ther, it has biases during PCR amplification and variation on rRNA gene copy numbers on different microorganisms. The detection of a microbial target is affected by the detection limit of the assay, volume of the filtered water sample, efficien- cy of nucleic acid extraction, and the presence of inhibitory compounds in PCR reaction. The qPCR method may overestimate the microbial counts, compared to a culture-based method, for example, enterococci, counts with the qPCR methods were reported higher than with the culture-based method (Haugland et al. 2005).

The reason for the difference can be that the qPCR method targeting DNA enumerates total DNA gene copies that can come from viable, viable but non-culturable, or even from dead cells (Pitkänen et al. 2013). The use of RNA as a PCR template after the reverse transcription step can overcome the limitation of enumerating dead cells with qPCR methods (Pitkänen et al. 2013). As the RNA is only found in viable cells, its abundance is affected by the metabolic activity of cells (Pitkänen et al.

(34)

32

2013). The reverse transcription (RT) is a process of changing RNA into comple- mentary DNA (cDNA). The RT-qPCR is the only option for all non-culturable RNA viruses for enumerating, due to a lack of DNA with them. Multiplex PCR can moni- tor multiple targets simultaneously on a single PCR tube.

3.3 MICROBIAL SOURCE TRACKING

Microbial source tracking (MST) is a process of the identification and quantification of the dominant source of contamination in surface waters (Stoeckel and Harwood 2007). MST method can be culture-based or culture-independent (Santo Domingo et al. 2007). The basic assumption of MST is that certain faecal microorganisms have a strong association with a specific host and their presence in environmental water is taken as evidence of faecal contamination of that host. MST helps for the direct prediction of potential pathogens and actual human health risk assessment.

The FIB based current bathing water monitoring method assigns an equal level of human health risk for all types of contamination. However, FIB from different sources have different risk levels and decay rates in bathing sites (Soller et al. 2010).

Some pathogens, mainly enteric viruses, are host-specific and most of the human- infecting enteric viruses only originate from human faecal contamination (Colford et al. 2007). Non-human faecal contamination from bovine, gull, poultry, swine, and wild animal sources contribute to zoonotic pathogens and parasites like Salmonella spp., Campylobacter spp., pathogenic E. coli, Cryptosporidium, Giardia, Leptospira, and Brucella spp. (USEPA 2009, Pitkänen 2013, Soller et al. 2014). Many zoonotic patho- gens can also have certain host patterns (Pitkänen 2013, Soller et al. 2014). Further, the correlation between the bathing related health risk and FIB counts are also affected by the source of contamination (Fewtrell and Kay 2015).

MST is needed for the identifying the contaminating source, mainly for un- known, non-point and multiple sources, such as contamination from domestic animals or wildlife, and agricultural or urban runoff (Santo-Domingo et al. 2004, Boehm and Sassoubre 2014). Further, MST could be useful in confirming the presence of accidental release of sewage influent or bather shedding. The combination of the MST approach with the current standardized FIB approach makes the interpretation of colony counts and microbial risk assessment easier (Harwood et al.

2014, Zhang et al. 2019).

The use of host-specific primers in the qPCR method is the most common MST method (Harwood et al. 2014). For example, human-specific Bacteroides, gull-specific Catellicoccus, and poultry-specific Brevibacterium have been used for MST (Okabe and Shimazu 2007, Haugland et al. 2010, Harwood et al. 2014). The qPCR-based MST approach is sensitive, specific, rapid, and cost-effective (Harwood et al. 2014).

However, the performance characteristics of each qPCR assay vary and, thus, require careful validation at each geographical location as the feeding patterns and herd characteristics of host animals may vary in different geographical locations

(35)

33 (Okabe and Shimazu 2007). In certain situations, qPCR based MST methods have to be used with caution, as the results might suggest past contamination rather than recent events.

3.4 SEQUENCING METHODS

Sequencing methods categorize the position of nucleic acid bases (adenine, gua- nine, cytosine, and thymine) of gene or genome partially or completely. It has been used in microbiology as a culture-independent tool for identifying the wide varieties of microbes (Shendure et al. 2017). Sanger sequencing is one of the first technologies to be developed for sequencing (Sanger et al. 1977). This technology is highly accurate (99.99%), but it is useful only for short sequence reads. The currently available version of the Sanger sequencer can sequence 96 samples simultaneously on a single sequencing run (Shendure et al. 2017). Shotgun sequencing is the next available sequencing technology. In it, long DNA sequences are randomly broken into smaller pieces and are amplified with the PCR method. Sequencing is then done as in Sanger sequencing. Then each sequence fragments are assembled with an algo- rithm into the long original DNA sequence. This technology can be used for sequencing an entire genome (Shokralla et al. 2012).

Next-generation sequencing (NGS) was developed in 2005 (Margulies et al. 2005). The first version of NGS was based on 454 pyrosequencing. NGS methods can sequence a high number of samples simultaneously on a single sequence run (Shokralla et al. 2012). Most currently available NGS technologies are based on the sequencing of PCR amplicons, such as Illumina, Roche 545, Ion Torrent sequencing, and pyrosequencing platforms. These PCR based methods can sequence less than 800 bases. Illumina MiSeq is one of the most widely used technologies and has a capacity of sequencing up to 300-600 bases (Illumina 2020). Still, the technology is continuously being updated, so the sequencing of longer bases can be available in the future. Further, Pacific Biosciences System is a single-molecule sequencing- based technology that operates without PCR amplification (Shokralla et al. 2012).

This method can sequence the longer sequences (> 1500 bases) in less than an hour (Shokralla et al. 2012).

NGS of the 16S rRNA gene is one of the most widely used methods for prokary- otic community analysis, mainly for bacterial diversity and taxonomic composition analysis. It can also be used for seeking potential health-related bacteria and to predict the enzymatic function of bacterial communities (Ye and Zhang 2011, Langil- le et al. 2013, Koo et al. 2017, Wu et al. 2019, Jin et al. 2018). The NGS method of the 16S rRNA gene was demonstrated to be more sensitive and identified more bacteria per sample than the traditional culture-based method (Gupta et al. 2019). This method enumerates the read counts (both absolute and relative) on different taxonomic levels and evaluates the diversity of the microbial community. In this method, DNA from the study sample is extracted, a specific hypervariable region of the

(36)

34

16S rRNA gene is amplified, sequenced, and then the base position is identified (Gupta et al. 2019). Based on the base position of the sequence, the microbial community composition is identified on different taxonomic levels with publicly available databases, such as SILVA, RDP, NCBI or Greengenes (Bonk et al. 2018). The method allows for parallel sequencing of hundreds of samples simultaneously and such obtained reads are analyzed with bioinformatics tools. The major shortcomings of the method are that the result is affected by various sequencing factors, such as variable regions used for sequencing, sequencing platforms, and sequence analysis pipelines (Gupta et al. 2019). Further, on the lower taxonomic level, mainly on the species-level, the method may not be equally effective for identifying microbes.

For example, the V4 region of two species, S. aureus and S. epidermidis, were 100%

identical (Gupta et al. 2019). Besides, most of the taxonomic annotations with amplicon sequencing approach results often unidentified species or genera, largely because these have not been cultured, identified, and reported earlier in the data- base (Gupta et al. 2019). However, the available databases are gradually expanding to better identify microbial communities to desired taxonomic levels in natural eco- systems.

Monitoring microbial quality, contaminant sources and bacterial communities of bathing water in Finland

Dissertations in Forestry and Natural Sciences

ANANDA TIWARI

Monitoring Microbial Quality, Contaminant Sources and

Bacterial Communities of Bathing Water in Finland

MONITORING MICROBIAL QUALITY, CONTAMINANT SOURCES AND BACTERIAL COMMUNITIES OF BATHING

WATER IN FINLAND

Ananda Tiwari

MONITORING MICROBIAL QUALITY, CONTAMINANT SOURCES AND BACTERIAL COMMUNITIES OF BATHING

WATER IN FINLAND

ABSTRACT

ACKNOWLEDGMENTS

LIST OF ABBREVIATIONS

LIST OF ORIGINAL PUBLICATIONS

AUTHOR'S CONTRIBUTION

CONTENTS

1 INTRODUCTION

2 MICROBIAL QUALITY OF BATHING WATER

2.1 INLAND AND COASTAL BATHING SITES

2.2 ETIOLOGICAL AGENTS CAUSING WATERBORNE ILLNESS

2.3 FAECAL INDICATOR BACTERIA

2.4 ALTERNATE FIB

3 METHODS FOR MICROBIOLOGICAL

BATHING WATER QUALITY CHARACTERIZA- TION

3.1 CULTURE-BASED METHODS

3.2 MOLECULAR METHODS

3.3 MICROBIAL SOURCE TRACKING

3.4 SEQUENCING METHODS