Factors Influencing Overland Mobility

of Cryptosporidium Oocysts

A dissertation submitted by

Christine E. Kaucner

In fulfilment for the requirements of the degree of

Master of Science (MSc.)

Centre for Water and Waste Technology School of Civil and Environmental Engineering The University of New South Wales

January 2007 ABSTRACT

The mechanisms responsible for overland transport of faecal pathogens, particularly Cryptosporidium oocysts, from animal sources to water bodies are not fully understood. Surface properties of microbes, such as electrostatic charge and hydrophobicity, are thought to contribute to their aggregation and attachment to solid surfaces. There is conflicting evidence that methods used to purify Cryptosporidium oocysts from faecal material may affect the oocyst surface, leading to biased conclusions from transport studies. By studying oocyst surface properties, aggregation and soil attachment, this thesis addressed whether oocyst purification methods influence overland transport studies, and whether oocysts are likely to be associated with particles during transport.

When using the microbial adhesion to hydrocarbon (MATH) assay with octane, oocyst hydrophobicity was shown to be method and isolate dependent, with oocysts displaying moderate to high hydrophobicity in 0.01 M KNO3. There was no observed attachment, however, to the hydrophobic octyl-SepharoseTM bead ligands when using the same suspension solution. Oocyst age did not appear to influence their hydrophobicity. A small but statistically significant proportion of oocysts displayed a net negative surface charge as observed by their attachment to an anion exchange ligand (DEAE). There was no difference in hydrophobicity or surface charge observed between purified oocysts and oocysts that had been extracted without the use of harsh chemicals and solutions with dehydrating properties.

Purified oocysts did not aggregate at pH values between 3.3 and 9.0, nor in solutions lower than 0.59 M in ionic strength at a pH 2.7 which is approaching the reported isoelectric point of oocysts. This finding suggests that oocysts may not form aggregates under general environmental conditions. The association of purified oocysts with soil particles was observed in settling columns. Attachment to soil particles was not conclusive since the settling of the soil particles may have entrained single oocysts. Nonetheless, approximately 27% of oocysts were estimated to be unbound to soil or associated with small soil particles. Hence models for oocyst overland transport should consider a significant fraction as single entities or associated with soil particles less than about 3 m in size.

i ORIGINALITY STATEMENT ‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project’s design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………….

Date ……………………………….

ii ACKNOWEDGMENTS

First and foremost I must acknowledge the support of my supervisors, Professor Nicholas Ashbolt and Dr. Cheryl Davies. I thank them for their guidance, and for not giving up on me when life, or work, overcame my part-time study endeavours.

Next I wish to thank those involved with the AwwaRF-CRCWQT project. There were many who worked on the project, but in particular I’d like to thank Dr. Christobel Ferguson (Ecowise Environmental) who was the project manager. Thanks also go to the staff from the Centre for Water and Waste Technology who supported this project, in particular Dr. Gautam Chattopadhyay, Lynette Menzies and Robbie Smith.

I am thankful to the Cooperative Research Centre for Water Quality and Treatment for the support they have provided me. Including me under the umbrella of their organisation and providing financial support for overseas travel is gratefully acknowledged.

From BTF Decisive Microbiology I would like to thank Dr. Graham Vesey for supplying the Iowa strain of Cryptosporidium oocysts, and also allowing me the use of a flow cytometer. Thanks go to Jin Chung (also a UNSW PhD student) for her help with flow cytometry. Thanks also to Associate Professor Justin Brookes (Adelaide University) who answered many questions about Stoke’s Law, Stephen Burgun who allowed me access to Arthursleigh Farm for soil collection, the Leppington Pastoral Company for permitting collection of faecal material from their calves and Dr. Michael Storey (Sydney Water) for proof reading this thesis.

To my parents; I thank them for all their support over the years, both financial and otherwise. Thanks also to Gabriel and Julie Dayeh who have become my surrogate Sydney family. And a million thanks go to Andrew for his patience and understanding while I completed this work.

And finally, I’d like to dedicate this thesis to my sister who was always proud of me and my achievements. I know she would have also been proud of this work.

iii

iv TABLE OF CONTENTS

Abstract...... i Acknowedgements...... iii Table of Contents ...... v Index of Figures...... x Index of Tables ...... xiii Abbreviations ...... xiv

Chapter One Introduction...... 1

Chapter Two Background...... 4

2.1 Introduction...... 4

2.2 Cryptosporidium Background...... 7 2.2.1 Brief history of Cryptosporidium...... 7 2.2.2 Cryptosporidium in the environment ...... 9 2.2.3 Cryptosporidium oocyst viability and infectivity ...... 13 2.2.4 Cryptosporidium oocyst survival ...... 15

2.3 Oocyst Isolation from Faecal Material...... 16

2.4 Surface Chemistry...... 19 2.4.1 Colloidal chemistry ...... 21 2.4.1.1 DLVO theory...... 23 2.4.1.2 The van der Waals interaction...... 24 2.4.1.3 The electrical double layer interaction...... 25 2.4.2 Measurement of surface chemistry ...... 26 2.4.2.1 Microbial adhesion to hydrocarbons (MATH) ...... 28 2.4.2.2 Hydrophobic interaction chromatography (HIC)...... 29 2.4.2.3 Contact angle...... 31 2.4.2.4 Atomic force microscopy...... 31 2.4.2.5 Salting-out...... 31 2.4.2.6 Surface charge...... 32

v 2.4.3 Cryptosporidium surface properties...... 33 2.4.3.1 Oocyst hydrophobicity...... 33 2.4.3.2 Oocyst surface charge ...... 34

2.5 Aggregation and Attachment...... 36 2.5.1 Particle aggregation...... 36 2.5.2 Oocyst attachment and aggregation ...... 37

2.6 Sedimentation Kinetics...... 39 2.6.1 Gravitational settling...... 39 2.6.2 Oocyst sedimentation...... 41

2.7 Oocyst Transport ...... 42

2.8 Summary of Current Knowledge ...... 44

Chapter Three Cryptosporidium Oocyst Preparations ...... 47

3.1 Introduction...... 47

3.2 Materials and Methods...... 47 3.2.1 Particle sizing...... 47 3.2.2 Extraction of Cryptosporidium oocysts from calf faeces...... 49 3.2.2.1 Collection of calf faeces...... 49 3.2.2.2 Removal of faecal lipids ...... 49 3.2.2.3 Sucrose flotation...... 50 3.2.2.4 Salt flotation...... 50 3.2.2.5 SephadexTM G-50 gel filtration...... 51 3.2.2.6 Octyl-SepharoseTM ...... 51 3.2.2.7 Flow cytometry sorting ...... 52

3.3 Results ...... 53 3.3.1 Diethylether defatting and sucrose flotation ...... 53 3.3.2 SephadexTM G-50 column...... 53 3.3.3 Flow cytometry...... 54

3.4 Discussion...... 60

3.5 Conclusions...... 61

vi

Chapter Four Cryptosporidium Oocyst Surface Properties...... 63

4.1 Introduction...... 63

4.2 Materials and methods ...... 64 4.2.1 Cryptosporidium oocysts ...... 64 4.2.2 Hydrophobic interaction chromatography (HIC) columns ...... 64 4.2.2.1 Pre-cast HIC columns ...... 64 4.2.2.2 In-house HIC columns ...... 65 4.2.3 Interactions with SepharoseTM beads in suspension...... 67 4.2.4 Microbial adhesion to hydrocarbons (MATH) ...... 68 4.2.5 Statistical analysis...... 69

4.3 Results ...... 69 4.3.1 Hydrophobic interaction chromatography (HIC) columns ...... 69 4.3.1.1 Pre-cast HIC columns ...... 69 4.3.1.2 In-house HIC columns ...... 71 4.3.2 Interactions with SepharoseTM beads in suspension...... 72 4.3.3 Microbial adhesion to hydrocarbons (MATH) ...... 74

4.4 Discussion...... 78 4.4.1 Hydrophobic interaction chromatography (HIC) columns ...... 78 4.4.2 Interactions with SepharoseTM beads in suspension...... 79 4.4.3 Microbial adhesion to hydrocarbons (MATH) ...... 80 4.4.4 Correlation of results from different methods...... 83

4.5 Conclusions ...... 84

Chapter Five Cryptosporidium Oocyst Aggregation...... 85

5.1 Introduction...... 85

5.2 Materials and Methods...... 85 5.2.1 Cryptosporidium oocysts ...... 85 5.2.2 Particle size distributions...... 85 5.2.3 pH experiments...... 86 5.2.4 Ionic strength experiments ...... 86

vii 5.2.5 Effect of stirring on aggregation...... 87

5.3 Results ...... 87 5.3.1 Effect of pH on aggregation...... 87 5.3.2 Aggregation with changing ionic strength...... 88 5.3.3 Effect of stirring on the particle size profiles...... 91

5.4 Discussion...... 97

5.5 Conclusions...... 100

Chapter Six Cryptosporidium Attachment To Soil...... 101

6.1 Introduction...... 101

6.2 Materials and Methods...... 101 6.2.1 Soil characteristics...... 101 6.2.2 Settling columns...... 102 6.2.3 Particle size distribution profiles...... 103 6.2.4 Quantification of Cryptosporidium oocysts...... 104 6.2.5 Collection of fractions method trial ...... 104 6.2.6 Temperature monitoring ...... 105 6.2.7 Statistical analysis...... 105

6.3 Results ...... 105 6.3.1 Sample collection protocol development...... 105 6.3.2 Gravity settling conditions...... 107 6.3.3 Soil settling...... 107 6.3.4 Cryptosporidium oocyst settling...... 108

6.4 Discussion...... 111

6.5 Conclusions...... 114

Chapter Seven Discussion ...... 115

Chapter Eight Future Research ...... 124

References...... 126

viii

Appendix A Raw particle size data (range 0.5 to 600 m, results as percentage of total solids by volume) for Cryptosporidium oocysts suspended in HEPES buffer at pH values ranging from 3.3 to 9.0 ...... 148 Appendix B Raw particle size data (range 0.5 to 600 m, results as percentage of total solids by volume) for a Cryptosporidium oocyst suspension at pH 6.8 with ionic strengths varying from 0.025 to 0.46 M...... 149 Appendix C Raw particle size data (range 0.5 to 600 m, results as percentage of total solids by volume) measured in duplicate for Cryptosporidium oocysts suspended at pH 3.4 in ionic strengths varying from 0.002 to 3.57 M...... 151 Appendix D Raw particle size data (range 0.2 – 180 m, results as percentage of total solids by volume) for Cryptosporidium oocysts at pH 2.7 in various ionic strength solutions with and without stirring of the oocyst suspension ...... 155 Appendix E Raw particle size data (range 0.5 to 600 m, results as percentage of total solids by volume) measured in duplicate for fractions collected from five settling columns containing soil. Column A was sampled at each of five time points (0, 26, 88, 300 and 900 minutes) and columns B to E were sacrifical columns sampled at one time point only...... 161 Appendix F Raw particle size data (range 0.5 to 600 m, results as percentage of total solids) in duplicate for fractions collected from settling columns 10 cm below the surface. Each of the three columns were sampled on five occasions (0, 26, 88, 300 and 900 minutes) at various times between 0 and 900 minutes...... 164

ix INDEX OF FIGURES

Figure 2.1 Hofmeister series reproduced from Xia et al. (2004)……………..……….30 Figure 3.1 Flow cytometry and sort gate for diluted calf faecal slurry. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts……………...…….55 Figure 3.2 Flow cytometry and sort gate for oocyst suspension prepared using SephadexTM G-50 column. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts…………………………………………………...…56 Figure 3.3 Flow cytometry and sort gate for oocyst suspension prepared using SephadexTM G-50 column followed by removal of lipids using octyl- SepharoseTM. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts ………………………………………………………………..…...…57 Figure 3.4 Flow cytometry and sort gate for oocysts extracted using diethylether and sucrose. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts ………………………………………………………………………………..58 Figure 3.5 Particle size distributions of oocyst suspensions prepared using one and two rounds of diethyl-ether defatting and sucrose flotation……………...... 59 Figure 4.1 Hydrophobic interaction chromatography columns: a) Sample being forced through pre-cast HIC octyl-SepharoseTM column; b) samples filtering by gravity through octyl-SepharoseTM columns made in-house with Pasteur pipettes………………………………………………..……………………....67 Figure 4.2 SepharoseTM beads after settling: a) SepharoseTM without side-chains, b) DEAE-Sepharose and c) octyl-SepharoseTM…….……………….………...... 68 Figure 4.3 MATH assay flow diagram……………………………………………...…69 Figure 4.4 Percentage of spiked oocysts removed from pre-cast HIC columns using eluents of decreasing ionic strengths of NaCl solutions (4 M, 2 M, 1 M, 0.5 M and 0.1 M) followed by up to 3 reagent water washes (MQ1 - MQ3)…..…...70

x Figure 4.5 Percentage removal of oocysts remaining in the HIC column prior to each elute passed through the columns…………………………………………....71 Figure 4.6 Comparison of SephadexTM G-50 and diethylether/sucrose extracted oocysts in suspension with DEAE- and octyl-SepharoseTM and SepharoseTM beads in

0.01 M KNO3, pH 5.8, error bars are +1 SD for triplicate oocyst suspensions ………………………………………………………………………………..73 Figure 4.7 Comparison of SephadexTM G-50 and diethylether/sucrose extracted oocysts in suspension with DEAE and octyl-SepharoseTM and SepharoseTM beads in

0.01 M KNO3 and 1 M NaCl, pH 5.8, error bars are + 1 SD for triplicate oocyst suspensions………….……………………………………………...... 74 Figure 4.8 Removal efficiency of five different isolates of Cryptosporidium oocysts from the aqueous layer of the MATH assay. Calf isolates 1 to 3 were extracted using diethylether treatment followed by and salt flotation, isolate Calf 4 was extracted using diethylether treatment followed by sucrose flotation, and the Iowa was a commercial suspension………………………..76 Figure 5.1 Particle size distribution profiles of a Cryptosporidium oocyst suspension in

0.01 M KNO3 at pH values between 3.3 and 9.0…………………….……….88 Figure 5.2 Particle size distributions of a Cryptosporidium oocyst suspension at pH 6.8 with ionic strengths varying from 0.025 to 0.46 M with NaCl……………....89 Figure 5.3 Particle size distributions of a Cryptosporidium oocyst suspension at pH 3.4

with ionic strengths varying from 0.002 to 3.6 M using MgCl2…….….….....90 Figure 5.4 Particle size profiles of a stirred Cryptosporidium oocyst suspension in acidified reagent water (pH 2.7) measured at regular intervals over a two hour period…………………………………………………………………………92 Figure 5.5 Particle size profiles of an unstirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 measured at 15 minute intervals over a two hour period…………………………………………………………………....92 Figure 5.6 Particle size profiles of a stirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 and an ionic strength of 0.59 M measured at regular intervals over a two hour period…………….………………………..94 Figure 5.7 Particle size profiles of an Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 and an ionic strength of 0.59 M measured at 15 minute intervals over a two hour period……...……………………………………....94

xi Figure 5.8 Particle size profiles of a stirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 and an ionic strength of 0.59 M measured at 15 minute intervals over a two hour period, with the 0, 5 and 10 minute profiles removed…………...…….………………………...………………....95 Figure 5.9 Stirred and unstirred Cryptosporidium oocyst suspension at an ionic strength of 1.6 M and pH 2.7…………………...…………………….………………..96 Figure 5.10 Particle size profiles of a Cryptosporidium oocyst suspension at pH 2.7 in different ionic strengths after two hours with and without stirring…………..97 Figure 6.1 Settling columns after settling for 100 minutes. Each column originally contained 20 g soil (wet weight), 1 L artificial rain water and ~105 Cryptosporidium oocysts……………………………………..……………..103 Figure 6.2 Particle size profiles for samples collected from five settling columns containing soil. Column A was sampled at each of the five time points, columns B to E were sacrificial columns and were sampled at one time point only………………………………………………………………………….106 Figure 6.3 Particle size distributions of fractions collected from 10 cm below the surface of triplicate settling columns at various times between 0 and 900 minutes, error bars ± 1 SD ……………………………………...…………..110 Figure 6.4 Oocyst counts from each sampled fraction of triplicate settling columns, error bars are ±1 SD…………………………………………………..……..110

xii INDEX OF TABLES

Table 3.1: Number of oocysts in 100 sorted particles from various oocyst preparations...... 59 Table 4.1: Mean percentage of three Cryptosporidium oocyst isolates passing directly through the octyl-SepharoseTM HIC columns compared to control SepharoseTM HIC columns………………………………………………………...………..72 Table 4.2: Percentage of oocysts remaining unbound to the hydrophobic component of the assay; comparison of results from three hydrophobicity assays and two Cryptosporidium isolates………………………………………………..……77 Table 6.1: Estimated and measured particle size fractions in three settling columns associated with each size fraction…………………………………………...111 Table 6.2: Estimated distance of single oocyst settlement using high and low reported oocyst density in Stoke’s equation, compared to the number of oocysts counted in each fraction and their statistical SNK ranking…………………111

xiii ABBREVIATIONS

Degree of statistical significance ANOVA Analysis of variance AwwaRF American Water Works Association Research Foundation C. andersoni Cryptosporidium andersoni C. hominis Cryptosporidium hominis C. muris Crytposporidium muris C. parvum Cryptosporidium parvum C. serpentis Cryptosporidium serpentis CRCWQT Cooperative Research Centre for Water Quality and Treatment cm Centimetre d(50) Median grain size d(60)/d(10) Uniformity coefficient DAPI 4’,6-diamidino-2-phenylindole DEAE Diethylaminoethanol DLVO theory Derjaguin-Landau-Verwey-Overbeek theory FISH Fluorescence in-situ hybridisation FSC Forward scatter g Grams g Gravitational force HEPES N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid HIC Hydrophobic interaction chromatography IMS Immunomagnetic separation kg Kilogram km Kilometer m Metre M Molarity MATH Microbial adhesion to hydrocarbons mg Milligram min Minute mL Millilitre mm Millimetre

xiv mM Millimolar MQ MilliQ reagent water mV Millivolts N Number of observations n/a Not applicable nd Not done nm Nanometre P Probability PBS Phosphate buffered saline PC2 Physical containment level two PCR Polymerase chain reaction PI Propidium iodide SD Standard deviation SNK Student-Newman-Keuls test RI Refractive index RNA Ribonucleic acid rRNA Ribosomal ribonucleic acid UK United Kingdom L Microlitre m Micrometre S Microsiemens s Second SSC Side scatter w/v Weight per volume % Percentage ‰ Per mille ºC Degrees celcius ~ Approximately < Less than > Greater than

xv CHAPTER ONE

INTRODUCTION

Waterborne outbreaks of illnesses occur worldwide. Between 1991 and 1998 there were 109 outbreaks associated with drinking water in the United States. A viral, bacterial or protozoan agent was identified in 39% of the outbreaks, and in a further 45% an infectious agent was suspected but not identified (Craun et al., 2003). A single outbreak of cryptosporidiosis in Milwaukee in 1993, the largest outbreak reported in the United States since consistent records began in 1920, was responsible for 403 000 cases of illness (MacKenzie et al., 1994) and an estimated 54 deaths (Hoxie et al., 1997). Most waterborne disease is associated with faecal pollution of water sources, with overland runoff from rainfall events being implicated in the faecal contamination of surface waters (Atherholt et al., 1998; Curriero et al., 2001; Tyrrel and Quinton, 2003).

In 1998, Cryptosporidium oocysts and Giardia cysts were detected in the drinking water distribution system of the city of Sydney, Australia. There were three such events identified between July and September that resulted in the issuing of three boil water recommendations for consumers (McClellan, 1998). An inquiry set up to investigate the cause of the contamination, the events and their management, the efficiency of the treatment plants, regulation, and treatment process, determined that the source of the contamination was most likely from animals in the catchment, although some human faecal contamination was also identified. Heavy rainfall during and prior to the events, and the formation of a thermocline and short-circuiting in the main raw water reservoir (Lake Burragorang) (Hawkins et al., 2001), were identified as some of the events that led to high numbers of oocysts and cysts being detected in the finished drinking water (McClellan, 1998).

A project jointly funded by the American Water Works Association Research Foundation (AwwaRF) and the Cooperative Research Centre for Water Quality and Treatment (CRCWQT) was undertaken during 2001-2004 (Davies et al., 2005a). This study was designed to advance the state of knowledge regarding the sources, fate and transport of pathogens in catchments. A systematic approach was implemented for

1 identifying the industry’s research priorities relating to pathogen fate and transport in catchments that were relevant to Australia and North America. Knowledge gaps were identified (Ferguson et al., 2003), and the research priorities determined by combining the magnitude of the gaps, the benefits of having them narrowed and the resources required to reduce the size of the knowledge gap. Research priority areas included quantification of pathogen sources in catchments, quantification of their attenuation as a function of organism characteristics and catchment-specific features, identification and quantification of the principal factors affecting their viability in faeces and soil, and quantification of the principal factors affecting overland transport of mobile pathogens to local surface water.

The AwwaRF-CRCWQT project focused on the fate and transport of three key model microorganisms: Cryptosporidium oocysts, bacteriophages and Escherichia coli. Defined quantities of pure suspensions of each of the microorganisms at high concentrations were required for the transport studies. Suspensions containing high concentrations of the bacteriophages and E. coli could easily be grown in the laboratory. However, high concentrations of Cryptosporidium oocysts could only be harvested from the intestines of an animal host. To obtain suspensions containing high concentrations of oocysts, therefore, large numbers were extracted from the faeces of infected calves less than 21 days old (Davies et al., 2005a). The extraction method used to purify the oocysts involved the use of a chemical and a high ionic strength solution treatment to obtain a pure oocyst suspension. The impact on any surface characteristics of the oocysts, such as surface charge and hydrophobicity, from the use of purification treatments was largely unknown. There was, however, some evidence that the surface charge of Cryptosporidium oocysts could be affected by the extraction methodology used (Brush et al., 1998). If the surfaces of Cryptosporidium oocysts were altered as a result of the purification process, then interactions that oocysts may have with soil particles would potentially be affected. This in turn has the potential to influence the movement of oocysts during overland transportation studies. Therefore the effect of various purification methods on the surface properties of Cryptosporidium oocysts was identified as part of the AwwaRF-CRCWQT transport studies as an important area of research.

2 The association of microorganisms with particles is an important consideration for the modelling of pathogen transport during periods of rainfall. The attachment or non- attachment of microorganisms to soil particles will influence their movement by changing the overall effective size of the microorganisms transported. Organisms that are attached to soil particles are less likely to travel as far, and are more likely to settle out and become trapped on vegetation, than those floating freely in suspension (Tyrrel and Quinton, 2003). Understanding and quantifying the mechanisms for pathogen fate and transport within drinking water supply catchments is a fundamental requirement to the reduction and management of pathogen risks to drinking water quality. Elucidation of transport mechanisms for key pathogens would enable modelling and prediction of water quality, and better management of environmental factors that govern pathogen transport.

The overall objective of this study was to determine whether the surface properties of Cryptosporidium oocysts influence their overland transport to surface water. Specific objectives were to examine whether Cryptosporidium oocysts were likely to attach to soil particles or to each other and hence move as a larger entity during rainfall events, and whether purification methods used were likely to have an impact on the results obtained for overland transport studies.

Chapter two provides a literature review of laboratory methods and current knowledge in the areas of colloidal chemistry, microbial surface chemistry, Cryptosporidium oocyst aggregation and attachment to solid surfaces, and aquatic transport. Different extraction techniques used to produce purified oocyst suspensions are presented in Chapter three, and in Chapter four results from various assays to study oocyst surface properties, including changes due to the use of different extraction methods, are presented. In Chapter five, the likelihood of Cryptosporidium oocyst aggregation under normal environmental conditions is explored, while Chapter six provides results from experiments on the attachment of purified oocysts to soil particles using settling columns. A final discussion is presented in Chapter seven which relates the results of this study to outcomes from the overland transport studies of the AwwaRF-CRCWQT project.

3 CHAPTER TWO

BACKGROUND

2.1 Introduction

Many waterborne pathogens are shed in the faeces of infected persons or animals. These pathogens may enter surface water by way of direct defecation, discharge of treated and untreated sewage, septic seepage, and runoff from agricultural lands and land where wastes from infected hosts are disposed or dispersed on soils. Soil macropores, soil fractures, or coarse granular particles can all promote the rapid transport of pathogens through soil and into groundwaters that contribute to local surface waters. Regardless of the mechanism of microbial movement, it is well recognised that pathogens are present in rivers and streams at higher concentrations and significantly higher loads after rainfall (Atherholt et al., 1998; Walter et al., 2001; Ashbolt and Roser, 2003; Ferguson et al., 2003), with heavy rainfall correlating with a higher incidence of illness in the community (Curriero et al., 2001). A study of 548 reported disease outbreaks in the United States from 1948 to 1994 found a significant relationship between extreme rainfall and illness, with the strongest association occurring when the rain fell in the month prior to the outbreak for surface waters or two months prior for ground water sources (Curriero et al., 2001).

The processes involved in overland transport of pathogens from the faecal source to the water include the release of the pathogens from the faecal deposits, possible attachment of freely suspended pathogens to soil particles, and sedimentation and resuspension of free and attached pathogens. The survival of microorganisms should also be considered in transport studies. This is particularly important when studying pathogens such as Cryptosporidium because, when using the methods currently available for routine water analysis (e.g. US-EPA 2005), intact oocysts are detected in water supplies regardless of their viability.

The initial stage of pathogen surface transport in a catchment is microbial release from faecal deposits. This is relevant to fields spread with both manure and deposited faecal pats. The mechanical effect of falling rain droplets has been identified as an important

4 mechanism in the release of Cryptosporidium oocysts from cow and calf faeces (Schijven et al., 2004). However, with increased water ponding in the field there is a corresponding decrease in the ability of raindrops to disturb and release pathogens from faecal material (Dai, 2003). When ponding occurs, therefore, pathogen release by erosion from overland flow is likely to be more important than disturbance induced by the energy from raindrops.

Immobilisation of freely suspended pathogens on soil particles, vegetation or formation of aggregates can restrict their transport. The attachment of pathogens to particles effectively changes physical properties such as diameter and net buoyant density. This in turn leads to changes in both the settling velocity and entrainment energy needed for transport (Harvey and Garabedian, 1991; Walker and Montemagno, 1999). The degree of change of their physical properties is dependent upon the type and number of attached particles. Muirhead et al. (2005) showed that 92% of Escherichia coli released from cow faecal pats were not attached to other particles and would therefore be readily transported upon release. Other studies have shown that microbes can attach to environmental particles (Stenström, 1989; Loveland et al,. 1996; Chattopadhyay and Puls, 1999; Gantzer et al., 2001; Searcy et al., 2005), although attachment can also be reversed by the frictional force of flowing water (Le Berre et al., 1998; Walker and Montemagno, 1999).

Macropores and fractures in subsurface layers are important for relatively fast, unrestricted transport of colloids (Rahe et al., 1978; Jacobsen et al., 1997). Macropores arise in agricultural soils for a number of reasons, including worm holes, channels created by plant roots that have since died and withered away, cracks in dry clay soil, and inter-aggregate spaces which become water-filled in wet soil conditions (Ghodrati et al., 1999; Capowiez and Bérard, 2006). Some researchers have suggested that macropores are the main means by which suspended matter can pass through the unsaturated zone, since such particles would otherwise be efficiently retained by physical filtration processes when moving through the soil matrix pores (McDowell- Boyer et al., 1986; McGechan and Lewis, 2002).

5 Soil type is a major factor influencing microbial transport due to differences in the adsorptive properties of different soil components. Whilst the solid phase of soil is comprised of organic matter, sand, silt and clay, it is the inorganic matter and clay particles that have the greatest effect on movement as a result of microbial adsorption to their surfaces (Mawdsley et al., 1995). This adsorption is believed to be largely responsible for the retention of viruses in the soil. Filtration and sedimentation processes are also important, particularly with bacteria and protozoa that are orders of magnitude larger than viruses (Mawdsley et al., 1995). In addition to particle attachment influencing microbial movement, attachment has been shown to offer some additional protection to microbes and extend their field-life (Sherer et al., 1992; Karim et al., 2004).

High stocking densities of domestic livestock, in particular cattle, will have an impact on the physical, chemical and bacteriological properties of stream water more than most other catchment uses (Meehan and Platts, 1978). On highlands, heavy grazing compacts the soil, reduces infiltration, increases runoff, and increases erosion and sediment yield. Light and moderate grazing, however, have much less impact (Gary et al., 1983). Grazing in riparian zones decreases resistance to erosion by reducing vegetation and exposing more vulnerable substrate, and trampling directly erodes banks. This increases turbulence in the stream and consequently increases erosion (Trimble and Mendel, 1995).

Vegetated buffer strips have been suggested as a means of reducing runoff pollution from agricultural land (Walker et al., 1990; Abu-Zreig et al., 2001; Atwill et al., 2002). Researchers have shown that vegetation significantly reduces the amount of runoff compared to bare soil (Davies et al., 2004; Ferguson et al., 2006), and that the type of ground cover does not affect the removal rates of faecal coliforms (Entry et al., 2000). Although vegetated buffer strips have been shown to remove significant numbers of microorganisms (Coyne et al., 1995; Davies et al., 2004; Ferguson et al., 2006), and are an effective means of controlling erosion (Wynn and Mostaghimi, 2006; Zaimes et al., 2006), the use of vegetation alone is insufficient to meet water quality goals for control of faecal coliforms (Walker et al., 1990; Coyne et al., 1995).

6 Modelling of pathogen transport is a useful tool to help predict the impact of land management strategies on faecal contamination of rural streams. From a risk assessment point of view, it is also a highly effective way of predicting pathogen loads for assessment of the associated risks. Collins and Rutherford (2004) developed a model to predict E. coli concentrations in streams draining from pastoral land. The model was shown to broadly reproduce observed E. coli concentrations in a catchment grazed by sheep and beef cattle. Park and Huck (2003) developed a framework for modelling Cryptosporidium oocysts that included equations for oocyst settling rate, runoff calculations, overland transport, detachment, source quantification from point and non-point sources, oocyst flux based on mass balance, water channel transport, stream confluence and lake transport. Ferguson et al. (2003, 2005) developed a conceptual model of a catchment and identified knowledge gaps with regard to pathogen movement. They identified riparian entrapment and host prevalence as the areas where the least information was available. Faecal disintegration and dispersion, manure treatment, wetland retention, retention pond entrapment, soil retention and environmental inactivation were also identified as areas requiring further research and understanding.

Pathogen transport is well recognised as being responsible for rainfall event-related increases in the concentrations and loads of waterborne pathogens in rivers and streams. However, there are significant knowledge gaps concerning the precise mechanisms of overland transport. Current state of knowledge and factors involved in attachment of microbes to soil particles, which can impact on pathogen movement and survival, are reviewed in this chapter.

2.2 Cryptosporidium Background

2.2.1 Brief history of Cryptosporidium

The first report of a Cryptosporidium infection was in 1907 by Ernst Edward Tyzzer. His report of Cryptosporidium in mice (Tyzzer, 1907) was followed by a publication in 1910 in which he proposed Cryptosporidium as a new genus and C. muris as the type strain (Tyzzer, 1910). In 1912 Tyzzer proposed the existence of another

7 Cryptosporidium species, C. parvum (Tyzzer, 1912). Infections of Cryptosporidium in other animals were reported sporadically over the next 60 years (Triffit, 1925; Slavin, 1955; Jervis et al., 1966; Panciera et al., 1971; Barker and Carbonell, 1974; Proctor and Kemp, 1974), during which time cryptosporidiosis was thought to be a disease of veterinary importance only. The first cases of cryptosporidiosis in humans were published in 1976 (Meisel et al., 1976; Nime et al., 1976). By the time cryptosporidiosis was reported to cause the death of an AIDS patient in 1984, the protozoan parasite had become accepted as a significant zoonotic pathogen warranting scientific research (Current et al., 1983).

Cryptosporidium was identified as being capable of causing waterborne outbreaks of human disease as early as 1985 (D'Antonio et al., 1985; Hayes et al., 1989), with researchers detecting Cryptosporidium oocysts in waters from the late 1980s (Rose et al., 1986; Musial et al., 1987; Rose et al., 1988; Hayes et al., 1989). Cryptosporidium became widely recognised as a waterborne pathogen of human public health significance due largely to the largest recorded modern outbreak of waterborne disease in the United States in 1993 in Milwaukee, Wisconsin. MacKenzie et al. (1994) estimated that 403 000 people contracted watery diarrhoea attributable to the outbreak, although a later report by Hunter and Syed (2001) suggested that the number was an overestimate and the actual community illness was 1 to 10% of that initially reported. Based on death certificate records for two year prior and two years following the outbreak, Cryptosporidium associated deaths of Milwaukee residents were reported to have significantly increased from 4 to 54 (Hoxie et al., 1997). The presence of Cryptosporidium in the drinking water that led to the outbreak was attributed to a treatment plant failure during a period where creeks and rivers had swelled from springtime rainfall (MacKenzie et al., 1994). Overland runoff and drainage from an abattoir into the source water lake were originally suspected as sources of the Cryptosporidium oocysts, but later molecular characterisation of oocysts isolated from humans during the outbreak suggested that the source was from human faeces (Peng et al., 1997).

After the Milwaukee outbreak of cryptosporidiosis, Cryptosporidium rapidly became recognised as one of the most serious and difficult waterborne pathogens to control

8 (Ramirez et al., 2004). It is now known that Cryptosporidium oocysts are ubiquitous in the environment, including surface waters (Hörman et al., 2004; Joachim, 2004), groundwater (Hancock et al., 1998) and wastewater (Rose, 1997; Gennaccaro et al., 2003), and that cryptosporidiosis has been described in over 170 vertebrate species (O'Donoghue, 1995; Fayer et al., 2000a). Cryptosporidium has been responsible for documented waterborne outbreaks of disease in the United States (MacKenzie et al., 1994; Rose et al., 2002) the United Kingdom and Japan (Smith and Rose, 1998), and Europe (Rose et al., 2002). In many instances, rainfall has been strongly associated with drinking water outbreaks (Rose et al., 2002). However, to date there have not been any outbreaks of cryptosporidiosis in Australia linked to municipal drinking water. Risk factors of sporadic cryptosporidiosis in two major cities in Australia have been identified as swimming in public pools and contact with a person with diarrhoea (Robertson et al., 2002).

Cryptosporidiosis is now recognised as one of the most common causes of diarrhoea in humans and livestock worldwide (Casemore et al., 1997). Cryptosporidium parvum was determined to be the most common species that caused illness in humans, and within the species two genotypes were previously distinguished (Peng et al., 1997; Morgan et al., 1998). Although the two genotypes were morphologically identical, they could be separated on the basis of distinct genetic differences, animal infectivity and pathogenicity. Using these criteria, and specifically differences at the 18S rRNA gene level, the previously named genotype 1 (human genotype) of C. parvum was renamed Cryptosporidium hominis (Morgan-Ryan et al., 2002), while genotype 2 (bovine genotype) retained the species name of C. parvum. Šlapeta (2006a), however, has recently proposed that the strain formally known as genotype 2 be renamed to C. pestis. This proposal is currently a topic of debate (Šlapeta, 2006b; Xiao et al., 2006a).

2.2.2 Cryptosporidium in the environment

There are currently 15 species of Cryptosporidium that are considered valid species. Furthermore, morphologically distinct species of Cryptosporidium have been found in fish, reptiles, birds and mammals but have not yet been named (Xiao et al., 2004). Most species of Cryptosporidium infect the small intestine, whereas C. muris, C. andersoni

9 and C. serpentis primarily infect the stomach (Aydin and Özkul, 1996; Lindsay et al., 2000). Molecular studies have suggested that some species and genotypes of Cryptosporidium are adapted to specific hosts (Morgan et al., 1999b).

Recognised potential sources of Cryptosporidium oocysts that could contaminate surface waters include the faeces from infected domestic animals such as cattle, cats, mice, pigs and horses (Olson et al., 1997; Guselle et al., 2003; McGlade et al., 2003), wildlife (Perz and Le Blancq, 2001; Warren et al., 2003; Power et al., 2004; Cox et al., 2005), and sewage discharges (Rose, 1997). In order to quantify the potential loading of Cryptosporidium oocysts in catchments, and therefore the associated risk of Cryptosporidium release from the farms into surface waters, various researchers have studied oocyst excretion by animals. Calves can excrete up to 1011 oocysts per animal during the latent period of 6 to 8 days (Fayer et al., 1998a) and, according to Atwill et al. (2003), the estimated daily environmental loading rates of Cryptosporidium parvum oocysts for adult beef cattle are 3900 to 9200 oocysts per animal per day. In contrast, Davies et al. (2005a) reported that the mean number of oocysts excreted by apparently healthy adult cattle was 331 oocysts per gram (dry weight) which corresponds to the excretion of approximately 106 oocysts per animal per day. Starkey et al. (2005) determined that the average number of oocysts detected in calves was 1.3 x 105 per gram.

The species of Cryptosporidium excreted in cattle varies with age. Pre-weaned calves tend to have significantly higher prevalence of C. parvum than older animals (Sischo et al., 2000; Wade et al., 2000; Atwill et al., 2003) and tend to become infected within 1 to 3 weeks of age (Uga et al., 2000; Becher et al., 2004; Starkey et al., 2005). They also tend to have a significantly higher prevalence of zoonotic genotypes than do older diary cattle (Santín et al., 2004; Fayer et al., 2006). Santín et al. (2004) found that 50% of pre-weaned and 19% of post-weaned calves on 15 dairy farms covering 7 states of the United States were infected with Cryptosporidium, and although C. parvum constituted 85% of infections in the pre-weaned calves, it caused only 1% of infections in post- weaned calves. Fayer et al. (2006) reported that a variety of Cryptosporidium species and genotypes, including C. suis, C. parvum, Cryptosporidium deer-like genotype, C.

10 bovis and C. andersoni, were isolated from 1 to 2 year old dairy cattle. C. hominis has also been isolated from cattle (Smith et al., 2005).

Sischo et al. (2000) determined that the risk factors for calves shedding Cryptosporidium were contact between calves and frequent bedding changes. This was explained by the process of bedding removal used, which had no form of biological control between calf pens. In this study the personnel and equipment became the means of spreading the infection.

Cryptosporidiosis is also a recognised disease in other farmed animals (Morgan et al., 1999a; Guselle et al., 2003; Ryan et al., 2005b) and, potentially, companion animals (Lloyd and Smith, 1997; McGlade et al., 2003). Novel genotypes have been reported in sheep (Chalmers et al., 2002), though recently Ryan et al. (2005b) suggested that sheep may not be an important zoonotic reservoir for Cryptosporidium, even though they have been known to harbour C. parvum (Morgan et al., 1997; Smith et al., 2005) and C. hominis (Ryan et al., 2005b) at a low incidence rate. Cryptosporidium has also been reported in pigs (Guselle et al., 2003) with the mean age of oocyst detection being 45.2 days post-weaning. The duration of infection averaged 28.7 days, and infections were asymptomatic (Guselle et al., 2003). Experimentally infected two day old piglets infected with 2.5 x 105 oocysts showed symptoms of diarrhoea and excreted oocysts for 5 to 16 days (Enemark et al., 2003). Morgan et al. (1999a) found the porcine and bovine Cryptosporidium genotypes in pigs, showing that pigs could act as a reservoir for human infectious Cryptosporidium.

Native animals, including migratory birds, are also recognised as a reservoir of Cryptosporidium (Bodley-Tickell et al., 2002; Zhou et al., 2004), and have been shown to be capable of harbouring C. parvum (Perz and Le Blancq, 2001). In Australia, a host- adapted Cryptosporidium ‘marsupial’ genotype (Morgan et al., 1997) has been detected in nine native marsupial species (O'Donoghue, 1995; Warren et al., 2003; Power et al., 2004). The Eastern Grey kangaroo (Macropus giganteus) has been found to excrete the marsupial subtype (Power et al., 2005), as well as a second novel genotype (Power et al., 2004), and to share the same (unknown) genotype with nearby sheep (Blasdall et al., 2003). Another Australian marsupial, the bilby (Macrotis lagotis), has been reported

11 with a C. muris infection (Warren et al., 2003). In a recent survey of Cryptosporidium in various animals in Sydney’s drinking water catchment, oocysts were detected in the faecal material of kangaroos and opossums but not in that of antechinus (Antechinus stuartii), wood duck (Chenonetta jubata), platypus (Ornithorhynchus anatinus), native rats (Rattus fuscipes) or wombats (Vombatus arsinus). With respect to feral wildlife, Cryptosporidium was also found in faecal material from one deer (Odocoileus virginianus) and one rabbit (Oryctolagus cuniculus) but not in those of carp (Cyprinus carpio), goats (Capra aegagrus hircus), feral cats (Felis catus) or feral pigs (Sus scrofa) (Cox et al., 2005). The survey, though, was not extensive with only one to eleven animals from each group being tested for Cryptosporidium oocysts.

The ability to characterise Cryptosporidium oocysts using molecular techniques has enabled the tracing and tracking of Cryptosporidium species and genotypes. Xiao et al. (2001) reported that C. parvum and C. hominis dominated in raw surface waters from sites where there was potential contamination from humans and animals, and intriguingly C. andersoni dominated the wastewater. The cattle species, C. andersoni, most likely made its way into the wastewater from cattle slaughterhouses that discharged their waste directly into the sewerage system studied. In contrast, the stormwaters tested yielded a high percentage of Cryptosporidium of the wildlife genotype, and none of the genotypes found in stormwater matched those of humans, farm animals or companion animals (Xiao et al., 2000). A study of Cryptosporidium oocysts detected in Sydney’s main drinking water reservoir showed a variety of genotypes including C. parvum (cattle genotype), C. suis, pig genotype II, the cervid genotype and a novel goat genotype (Ryan et al., 2005a).

The various species and genotypes of Cryptosporidium cannot be differentiated on the basis of morphology (Fall et al., 2003). Although molecular analysis provides an enormous amount of information regarding the species and genotype of any isolates, it is a specialised research tool and is not used in testing laboratories during routine analysis of samples for the presence and enumeration of Cryptosporidium oocysts. As more isolates of Cryptosporidium are genotyped, it is becoming increasingly clear that some species are host adapted and may be of little human significance (Monis and Thompson, 2003; Ryan et al., 2005b; Ruecker and Neumann, 2006). This has lead to

12 the suggestion that, with the exception of C. parvum, Cryptosporidium may not be a zoonotic parasite (Monis and Thompson, 2003). If the only species of Cryptosporidium that are important to human health are C. hominis and C. parvum and zoonosis only occurs with a couple of species, then any risk assessments based on the identification of Cryptosporidium oocysts in water samples should be considered as very conservative unless a sanitary survey supports a predominantly human and/or calf source (Xiao et al., 2006b).

2.2.3 Cryptosporidium oocyst viability and infectivity

Viability and infectivity are two terms used to describe active Cryptosporidium oocysts, with viability being used to describe oocysts that are alive, and infectivity used to describe oocysts that are capable of causing an infection. It is possible, however, for oocysts to be viable but not infectious.

Excystation is considered to be the reference method for determining the viability of oocysts. In this method oocysts are subjected to acidified conditions that imitate the passage of oocysts through the stomach, with the sporozoites being subsequently released from the oocyst (Robertson et al., 1993). Oocysts and sporozoites are generally enumerated using light microscopy (Uguen et al., 1997), though in an effort to simplify the method Vesey et al. (1997) described the use of flow cytometry for oocyst and sporozoite quantification. Even so, for excystation results to be reliable, a Cryptosporidium oocyst suspension must be relatively clean and concentrated. This makes excystation an unsuitable method for assessing the viability of individual or small numbers of oocysts that are typically isolated from water samples (Campbell et al., 1992).

Vital dye staining, using the fluorogenic dyes 4’,6-diamidino-2-phenylindole (DAPI) and propidium iodide (PI) has been one of the most commonly used methods for easily and quickly determining the viability of low numbers of oocysts (Campbell et al., 1992; Anguish and Ghiorse, 1997; Jenkins et al., 1997; Bukhari et al., 1999). After treatment with an acidified buffer, DAPI stains the nuclei of sporozoites regardless of their viability, but only oocysts containing sporozoites with disrupted or broken membranes

13 stain with PI (Campbell et al., 1992). Dead oocysts that no longer contain sporozoites may also be described as ‘ghosts’ since there is no nucleic acid material available for staining (Campbell et al., 1992).

Other methods have been developed to determine oocyst viability, such as different nucleic acid stains (Belosevic et al., 1997), fluorescence in-situ hybridisation (FISH) (Smith et al., 2004), and polymerase chain reaction (PCR) (Stinear et al., 1996; Deng et al., 1997; Rochelle et al., 1997). FISH utilises a fluorescent nucleic acid probe so that species recognition is possible as well as viability assessment. Similarly, PCR-based methods have been shown to detect viable oocysts from water samples and have the capability of determining the species or genotypes present (Rochelle et al., 1997; Kaucner and Stinear, 1998).

The ‘gold standard’ method for determining the infectivity of oocysts is by neonatal mice infectivity assays (Korich et al., 2000). Due to host specificities of different Cryptosporidium species, as well as requirements for upkeep of an animal house and ethics approval for experiments, mice cannot always be used by researchers. The use of cell culture has been developed as a quicker, easier, less costly and more widely available method of determining oocyst infectivity. Some researchers combine cell culture techniques with PCR to make the method more rapid and sensitive (LeChevallier et al., 2000)

Comparisons between viability and infectivity methods often do not correlate well. This may be due to various methods targeting different components of the oocyst (such as DNA targeted nucleic acid stains, RNA based FISH methods, different gene targets for PCR) or infection (availability of different cell lines and mouse models). Comparisons may also be problematic due to the high numbers of oocysts generally required for infectivity assays compared to the lower numbers typically needed for testing viability. When used as a surrogate for infectivity, viability assays tend to overestimate the number of infectious oocysts (Bukhari et al., 2000; Schets et al., 2005). Cell culture, however, has been shown to be equivalent to mouse infectivity (Rochelle et al., 2002; Jenkins et al., 2003).

14 2.2.4 Cryptosporidium oocyst survival

Although the life cycle of Cryptosporidium is complex (Hijjawi et al., 2002; Barta and Thompson, 2006), with numerous stages between infection and release, thick-walled oocysts are the only stage that may persist outside the host, while thin-walled oocysts are only capable of reinfecting the host. Thick-walled oocysts have a tough outer shell that protects four potentially infective sporozoites (Fayer, 2004). The oocyst wall protects the sporozoites from a variety of environmental stresses and also chlorine disinfection processes in conventional drinking water treatment. Thus, oocysts are well designed to survive outside the host body until ingested, at which time stomach acids stimulate excystation and the release of the sporozoites into the intestine of the new host.

Under favourable conditions, oocysts can survive in the environment for approximately six months (Robertson et al., 1992) but after that the infectivity rapidly decreases (Fayer et al., 1998b). Neither moderate freezing nor heating to 50 ºC completely inactivates oocysts (Robertson et al., 1992; Olson et al., 1999). However, dessication has been shown to be lethal (Robertson et al., 1992).

Davies et al. (2005a) demonstrated a four log10 inactivation of oocysts in bovine faeces at 35 ºC within 56 days. The high temperature was chosen as the maximum temperature recorded at the surface of bovine faecal pats in an Australian field in summer when the air temperature reached 30 ºC. In faecal microcosms in which the factors of temperature, moisture and biotic status were tested, temperature was the most influential in Cryptosporidium inactivation (Davies et al., 2005a). Biota in samples, however, can also have an impact on oocyst survival, with oocysts in soil or faeces showing more rapid degradation than those in water (Jenkins et al., 1999; Olson et al., 1999). Heisz (1997) showed that the level of heterotrophic bacteria in natural waters influenced oocyst survival. Several species of rotifers, found ubiquitously in moist environments from soil to puddles, ponds, rivers and lakes, have been shown to ingest and excrete oocysts (Fayer et al., 2000b; Stott et al., 2001), but the ultimate survival of the oocysts have not been reported. Infectivity of Cryptosporidium oocysts have also been shown to decrease over time with a high environmental salinity (35‰), though after 35 days

15 the inactivation may not be complete (Robertson et al., 1992; Freire-Santos et al., 1999).

Oocysts also come in contact with many types of soil. Davies et al. (2005b) found that both temperature and soil type were significant factors influencing Cryptosporidium oocyst inactivation. Soils samples collected from 37 farms in New York state and analysed for the presence of Cryptosporidium oocysts revealed that the risk factor of oocysts in soils of neutral and basic pH was lower than those of low pH (Barwick et al., 2003). This finding was inconsistent with the report by Robertson et al. (1992) that described how oocysts are unable to survive low extremes of pH.

Cryptosporidium oocysts are also resistant to chlorine and most conventional disinfectants (Korich et al., 1990), although ozone and UV irradiation have been shown to be capable of causing inactivation (Korich et al., 1990; Quinn et al., 1996; Rochelle et al., 2005). Since oocysts have shown such resistance to chlorine inactivation, removal of oocysts typically depends on physical treatment processes (Dai and Hozalski, 2002). A review of Cryptosporidium ecology and public health implications found that conventional treatment (coagulation, sedimentation, filtration) of raw drinking water typically removes, on average 99%, of Cryptosporidium oocysts (Rose, 1997). Even so, a study of 82 surface water treatment plants showed that viable Cryptosporidium oocysts can still be detected in filtered drinking water (Aboytes et al., 2004). The authors of that study concluded that an additional treatment barrier, such as UV disinfection, should be used to meet health risk goals of one infection per 10 000 people per year (Aboytes et al., 2004).

2.3 Oocyst Isolation from Faecal Material

Oocysts used for research are generally obtained from infected animals and must be purified to remove faecal matter and debris before use. Faeces may contain mucus, lipids, bacteria, fungi, insects, pollen and partially digested foodstuffs (Zarlenga and Trout, 2004). Oocyst suspensions are often prepared by first removing the lipids from the faecal material using ethyl acetate or diethylether prior to isolation by centrifugation in solutions of various densities (Weber et al., 1992; Clavel et al., 1996; Upton, 1997),

16 though some authors report efficient oocyst extraction without the use of defatting (Kuczynska and Shelton, 1999; Entrala et al., 2000). Flotation is the simplest form of density centrifugation. In this procedure an aqueous sample is mixed with, or layered over, a dense solution of a predetermined specific gravity and subjected to centrifugation (Zarlenga and Trout, 2004). Stepwise and gradient centrifugation systems are also used by researchers for oocyst purification (Zarlenga and Trout, 2004).

The most common form of flotation is on saturated NaCl or sucrose with a specific density of 1.20 (Upton, 1997), though other flotation media used in oocyst purification ® can include solutions such as zinc sulphate (ZnSO4), Percoll or cesium chloride (CsCl) (Ryley et al., 1976; Waldman et al., 1986; Kuczynska and Shelton, 1999). Ryley et al. (1976) treated Eimeria oocysts with salt, sucrose or zinc sulphate, for periods of one to seven days. They observed that with salt treatment the oocysts would deform or collapse. These effects, however, were reversible upon washing without loss of viability. In contrast, those oocysts treated with sucrose and zinc sulphate had a significantly reduced viability. Increasing the length of treatment in each solution increased oocyst inactivation. Oocysts that underwent flotation purification steps were in contact with the hyper-osmotic solutions for much shorter periods of time, and this presumably had little effect on the viability or morphology of the oocysts (Ryley et al., 1976). Freire-Santos et al. (1999) demonstrated that mouse infectivity of Cryptosporidium oocysts decreased with increasing salinity and increasing contact time with the saline suspensions. Distorted and collapsed oocysts were observed when subjected to storage conditions of salinity at 35‰ for 40 days.

Recovery of oocysts can be influenced by oocyst viability. Only water impermeable oocysts can be recovered using a hyperosmotic gradient, so density centrifugation often acts as an enrichment, or selective concentration, for viable oocysts (Bukhari and Smith, 1995; Entrala et al., 2000). Percoll® gradient centrifugation has been reported to preferentially concentrate viable oocysts compared to other methods such as sucrose flotation followed by dialysis (Suresh and Rehg, 1996). At the same time, however, the recovery efficiency using that method was less than that obtained for some other methods. Bukhari and Smith (1995) found that a method which involved ether defatting

17 followed by washing of the resulting pellet in water gave a higher recovery of Cryptosporidium oocysts compared to zinc-sulfate and sucrose flotation methods.

The effect on viability and infectivity of oocysts purified using different isolation methods was studied by Slifko et al. (2000) , who concluded that defatting agents, immunomagnetic separation (IMS) and bleach treatment had no detrimental effects on oocyst infectivity. Oocysts purified using CsCl alone showed the most consistent cell culture infection compared to other methods including defatting prior to CsCl purification and flotation on Sheather’s solution (a mixture of sucrose, phenol and Tween 80 (Suresh and Rehg, 1996)).

Most of the purification methods described above often display significant losses associated with the procedures, and a vast amount of effort has been put into improving recovery efficiencies (Weber et al., 1992; Clavel et al., 1996; Suresh and Rehg, 1996; Kuczynska and Shelton, 1999; Entrala et al., 2000). In situations where there are high concentrations of oocysts in faecal material, some oocyst losses would be expected and can often be tolerated. In contrast, when low concentrations of oocysts in faecal material require quantification, then IMS is another procedure that has been used successfully for quantitative oocyst extraction and purification (Davies et al., 2003). IMS, however, is not suitable for purification of high oocyst concentrations due to a limited number of oocyst binding sites on the beads.

After purification, the oocyst suspensions used by researchers are generally stored in solutions such as distilled water or phosphate buffered saline (PBS). Potassium dichromate was a commonly used preservative for oocyst suspensions in the 1990s. Its use in the water industry, however, has reduced in recent years. Campbell et al. (1993) found that the use of the preservatives potassium dichromate or formalin reduced the viability of Cryptosporidium oocysts as determined by vital dye staining and excystation. Dufour et al. (1999) published criteria for evaluating protozoa detection methods in water specifying that oocysts should not be stored in preservatives. Instead, they recommended that oocysts be stored in a medium for decontamination, such as water containing antibiotics. Storage of oocysts in antimicrobial solutions to inhibit bacterial and fungal growth, and surface sterilisation of oocysts using hypochlorite,

18 have thus been more recently used by researchers as alternative means of maintaining pure suspensions. Researchers, however, have reported infection of dichromate preserved oocysts in mice and cell lines (Vergara-Castiblanco et al., 2000; Pokorny et al., 2002; Yang et al., 2006). Recently Inoue et al. (2006) showed that a storage period of 100 days in potassium dichromate had little effect on oocyst morphological features, size, specific gravity, or hydrophobicity.

Knowledge of the influence of purification methods on oocyst surface chemistry is limited (Brush et al., 1998), yet oocysts are generally still studied without taking into account this potentially crucial factor (Butkus et al., 2003). There is conflicting evidence that some purification methods may influence the surface properties of Cryptosporidium ooycsts, and also some evidence that the surface charge can be influenced by the medium in which the oocysts are stored. This is an area of ongoing research, made more complicated by the inability of obtaining pure suspensions of oocysts that have not undergone chemical defatting or hyperosmotic gradient centrifugation.

2.4 Surface Chemistry

Microbial adhesion to solid surfaces is a crucial step in many processes such as biofilm formation (Fang et al., 2000), and is considered to be instrumental in the behaviour of microorganisms in the water treatment process (Hsu et al., 2001). Attachment of microbes to soil particles has also been recognised as a process with the potential to affect overland transport of pathogens (Tyrrel and Quinton, 2003). Similarly to colloidal attachment and aggregation, the surface chemistry of microorganisms is thought to play a major role in their attachment to solid surfaces and aggregation.

Hydrophobicity is the repulsion between a non-polar compounds and a polar environment such as water. Cell surface hydrophobicity of bacteria is generally considered to be determined mainly by the type and amount of extracellular polymers, such as polysaccharides, lipopolysaccharides, glycoproteins and proteins on their surfaces. There are numerous reports of an increase in hydrophobicity when the

19 oligosaccharides are removed from a cell surface (Rosenberg et al., 1980; Hermansson et al., 1982; Kuznar and Elimelech, 2006).

Hydrophobic interactions have been reported to contribute to adhesion processes involving viruses (Gerba, 1984; van Voorthuizen et al., 2001), bacteria (Dahlbäck et al., 1981; Stenström, 1989) and protozoa (Capizzi-Banas et al., 2002). Hydrophobicity is generally accepted as one of the determinant factors of microbial adhesion to surfaces and, according to some researchers, it dominates in the adhesion of hydrophobic organisms regardless of their surface charge (van Loosdrecht et al., 1987a; Gilbert et al., 1991). Other researchers, however, have concluded that interactions other than hydrophobicity may play a role in microbial partitioning (van Loosdrecht et al., 1987b; Ahimou et al., 2001).

Some other researchers claim that the heterogeneity in cell surface macromolecules could explain the lack of consensus regarding adhesion characteristics. Using atomic force microscopy, Camesano and Abu-Lail (2002) showed substantial heterogeneity in the adhesion affinities of different bacteria, and also at different areas of the same bacterial cell. They concluded that the heterogeneity of the macromolecules on bacterial surfaces should be incorporated into models of adhesion, since these properties were more important than previously thought. In contrast, Jones et al. (2003) showed that adhesion in E. coli K-12 was mostly associated with the ends of the bacteria, which were rod-shaped, and that lipopolysaccharide length did not affect adhesion.

The role of microbial surface charge during adhesion processes is also unclear, with some reports advocating, and others rejecting, a primary role for charge interactions (Chattopadhyay and Puls, 2000), particularly in low ionic strength or organic environments (Gardner et al., 1998a; Dai and Boll, 2003). Under different environmental conditions the governing force of interaction between the microbes and solid surfaces can change between electrostatic force and hydrophobicity (Brush et al., 1998), complicating the role of hydrophobicity and surface charge in transport studies. The surface charge, wettability, surface texture and diversity of different types of soil particles may also influence the adhesion process, making it difficult to predict microbial adhesion in soil ecosystems (Stenström, 1989).

20

The electrostatic charges on microbial surfaces are caused by dissociation of various organic functional groups, like carboxyl and amino groups, located on the outer surface and situated in deeper layers of the cell wall (Stevik et al., 2004). Electrostatic charges have been shown to influence the attraction of bacteria (Gilbert et al., 1991) and Cryptosporidium oocysts (Gitis et al., 2002) to solids. Jones et al. (2003) showed that surface non-uniformities in bacteria play an important role in adhesion. Such non- uniformaties could result from bacterial polarity, and charged non-uniformity has been shown to affect adhesion in colloidal systems (Feick and Velegol, 2002). Nanodomains on bacterial ends are important for adhesion and the time scale for irreversible adhesion is short (Jones et al., 2003). Some researchers have shown that surface charge is the most important factor in adhesion, dominating any hydrophobic effects (Dai et al., 2004). Others, however, suggest that surface charge does not correlate with the transport of bacteria (Gannon et al., 1991b). Researchers are beginning to concede that the physical and chemical systems of particle aggregation and environmental conditions are not separable (Gardner et al., 1998a), and that more than one characteristic of a microorganism will determine its attachment to, or transport through, soil (Gannon et al., 1991b).

2.4.1 Colloidal chemistry

When microorganisms are included in overland transport models they are often considered as colloidal material (Harvey and Garabedian, 1991; Harter et al., 2000). In colloidal chemistry, the designation “colloid” is used for particles that are of some small dimension that cannot be seen with the naked eye, and cannot pass through a membrane with a pore size of one micron. Colloid particles may be spherical, but in some cases one dimension can be much larger than the other two. The size of colloidal particles may therefore range from 0.01 to 10 m (Birdi, 1997).

Due to their small size and high surface area to volume ratio, colloids are primarily influenced by surface and hydrodynamic forces, and are not significantly influenced by external forces such as gravity (Bergendahl and Grasso, 1999; Bergendahl and Grasso, 2000). Electrostatic surface charge and hydrophobicity are surface properties that are

21 important to the adhesion mechanisms between two colloidal particles. Like-charged particles tend to repel, while oppositely-charged surfaces are attracted and will adhere. Compared to hydrophilic surfaces, hydrophobic surfaces have a lower surface energy and are not likely to form hydrogen bonds and so are less reactive in aqueous environments. Suspended hydrophobic particles will therefore generally adhere to hydrophobic surfaces in solutions since both are repelled by the hydrogen bonds in water.

Solid surfaces in aquatic environments generally carry electrostatic charges. These charges may be permanent and are generally negative, or they may be dependent on the pH of the surrounding electrolyte solution (Gannon et al., 1991a). Similarly, hydrophobicity is a function of pH and ionic strength (Hsu and Huang, 2002). The initial interactions, as microbes and colloid particles approach, are governed by long and medium range forces. These are primarily van der Waals and electrostatic forces (Razatos et al., 1998), and they depend on the physicochemical properties of the particle and the microbial surfaces, such as hydrophobicity (Gannon et al., 1991a), free energy (Busscher et al., 1984), and surface charge (Gannon et al., 1991a).

The term stability is used to describe the tendency of particles to resist aggregation. The stability of particles is also a function of repulsive and attractive forces between the particles. Factors that affect particle stability in water include pH, ionic strength, surface functional groups, and adsorption of inorganic and organic matter from solution. Mechanisms of particle destabilisation and aggregation include compression of the double layer, charge neutralisation, precipitate enmeshment, and inter-particle bridging (Butkus and Grasso, 1998).

The comprehension of intermolecular forces is far from complete, and qualitative results have been obtained only for simple and idealised models of the real matter (Birdi, 1997). Furthermore, the quantitative relations that link intermolecular forces to macroscopic thermodynamical properties by using statistical mechanics are also at present limited to simple and idealised cases. Therefore, the theory of intermolecular forces gives no more than a semi-quantitative basis to interpret and generalise the existing experimental data (Birdi, 1997).

22

2.4.1.1 DLVO theory An approach commonly used to understand the forces of interactions between colloids in liquids has been to build on the work of Derjaguin, Landau, Verwey and Overbeek (DLVO) (Butkus and Grasso, 1998). In the 1940s, as two independent research teams, these researchers reported that the force of interaction between surfaces in liquids is the sum of the van der Waals and electrical double layer interactions (van Loosdrecht et al., 1987a; Butkus and Grasso, 1998).

The DLVO theory is often applied to describe microbial adherence to particles and other surfaces. As two similarly charged bodies approach each other they are subjected to both attractive and repulsive forces, which are additive in effect (Busscher et al., 1990). Typically, at long distances of separation of around 10 nm, attractive forces that include hydrophobic interactions between hydrophobic areas on the surface of the microbial cell and the particles, predominate and serve to form a reversible attractive interaction (Jones et al., 1996). At shorter distances the overlap of adsorbed counter ion clouds on each surface, which account for the apparent surface charge, serves as a repulsion barrier between the cells and the particles. Finally, irreversible adhesion occurs via specific interactions between stereochemically complimentary molecules on the cell surface and the inert substratum (Busscher et al., 1990; Bunt et al., 1993; Jones et al., 1996).

DLVO theory has several shortcomings in predicting attachment and detachment of colloids. It has generally been shown that under unfavourable conditions for attachment, DLVO theory underestimates attachment by many orders of magnitude (Schijven and Hassanizadeh, 2000). DLVO theory was also formulated for smooth bodies with ideal and uniform properties. In practice, real particles are irregular, and their surfaces are rough and likely to be heterogeneous in composition and charge (Swanton, 1995). Therefore, the DLVO theory has since been extended to include molecular-level hydrophobic interactions, Brownian motion forces and Lewis acid-base interactions (Butkus and Grasso, 1998).

23 The incorporation of additional interactions into the DLVO theory is often called the extended DLVO theory. These interactions may include hydration pressure, hydrogen bonding forces, hydrophobic effects, solvation forces, disjoining pressure, structural forces or Lewis acid-base forces (Bergendahl and Grasso, 1999; Smets et al., 1999). Bergendahl and Grasso (1999) showed that the extended DLVO theory gave a better prediction of colloid detachment from glass beads when suspended in solutions of various pH values and ionic strengths.

Both the traditional and extended DLVO theories have been used successfully to describe microbial surface thermodynamics (Grasso et al., 1996; Smets et al., 1999; Chen and Strevett, 2001). DLVO theory is largely applied for the prediction of the stability of many colloidal systems, including the natural ones. Despite this, numerous phenomena of colloidal behaviour cannot be explained quantitatively or qualitatively by DLVO theory (Dong, 2002; McBride and Baveye, 2002). There are many possible causes for the total energy of interactions to be different from that predicted from purely van der Waals and electrical double layer interactions (Missana and Adell, 2000). An alternative description of the fundamental forces involved in the formation of dispersions and gels was outlined in 1938 by Langmuir (1938), and more recently developed further (McBride and Baveye, 2002). This model, unlike DLVO theory, hypothesizes that the long range Coulombic attractive force counters osmotic repulsion. Such forces could become important in situations of multi-particle interactions in conditions of low electrolyte concentration and high particle charge (McBride and Baveye, 2002).

2.4.1.2 The van der Waals interaction Long-range forces, such as dispersion, dipole orientation, and induction, collectively known as van der Waals forces, have interaction energies that decay as the inverse sixth power of intermolecular distance (Birdi, 1997). Dipole interactions arise from the electric field produced by one dipole acting on a second dipole, and generally lead to a negative potential energy (i.e. attraction) between the interacting atoms or molecules. The van der Waals forces consist mainly of Keesom dipole-orientation, Debye induction and London dispersion forces (Chattopadhyay and Puls, 1999).

24

The van der Waals interactions are linearly dependent on the value of the Hamaker constant, which depends on the nature of the interacting materials (Swanton, 1995). The Hamaker constants of most forms of organic matter are similar to that of water, hence the van der Waals interactions between organic colloids are weak. In contrast, inorganic matter tends to have large Hamaker constants (Schijven and Hassanizadeh, 2000). The direct measurement of this constant is extremely difficult, and its determination has been the subject of considerable research efforts. Generally, the Hamaker values used in equations for natural colloids are those estimated for silica colloids.

2.4.1.3 The electrical double layer interaction In an aqueous medium the electrostatic potential on microbial cell surfaces influences the distribution of charge in the surrounding medium and ions of opposite charge are attracted toward the surface. The charging of surfaces in aqueous media can arise from dissociation of surface groups, by the adsorption of ions onto the surface, or can be due to intrinsic surface charge. When a charged surface exists in an aqueous solution, counter (oppositely charged) ions will associate next to the surface. An electrical double layer can thus be considered as a layer of charge at the surface, and a diffuse layer of ions decaying away from the surface (Gardner and Theis, 1996).

When the energy of ion attraction is greater than their Brownian motion energy, a fraction of the counter ions will be bound directly to the surface. The larger the energy of attraction, the greater the number of counter ions attached, with the thickness of this diffuse layer depending on the ionic strength of the solution and the valencies of the counter ions (Gannon et al., 1991a; Busscher and Norde, 2000). The electrical interactions between particles in solution are governed by the extension of the diffuse layer. Increasing the salt concentration results in a decrease in electrical interactions between two like charged particles, and the effect of the electrokinetic potential will increase with decreasing hydrophobicity (van Loosdrecht et al., 1987a).

25 Therefore, as a result of increasing ionic strength reducing the thickness of the diffuse layer of counter ions, bacterial cells may be brought close enough to the media surfaces for the van der Waals attraction energies to overcome the repulsion barrier. In water, both protein and ligand surfaces are surrounded by water molecules. In this situation the hydrophobic interaction between the two surfaces is shielded since the hydrophobic parts of both surfaces are hidden beneath the hydrating layers of water. Some salts, called kosmotropic salts, form ions that have high polarity and bind water strongly. This induces the exclusion of water from the protein and ligand surfaces, allowing the surfaces to adsorb more strongly. In contrast, chaotropic salts have less polarity and bind water loosely, which induces inclusion of water on the protein and ligand surfaces. Neutral salts lie in between kosmotropic and chaotropic salts (Arakawa and Timasheff, 1982; Arakawa and Timasheff, 1984; Xia et al., 2004).

When two charged surfaces approach one another, consideration must be given to the overlap of the respective electrical double layers that occurs. The ion concentration within the overlapping region will be significantly higher than in the bulk solution, and this will result in a repulsive osmotic pressure between the surfaces. Therefore, as two identically charged surfaces approach one another, both the pressure and the direct electrical stresses will contribute to the force of interaction (Swanton, 1995).

2.4.2 Measurement of surface chemistry

There are numerous methods available for testing the surface properties of microbes. Techniques commonly employed for characterising the physicochemical nature of microbial cell surfaces are bacterial adhesion to hydrocarbons (Rosenberg et al., 1980), electrophoretic mobility (Brush et al., 1998), retention on chromatographic resins and adhesion to inanimate materials (Hazen and Hazen, 1987; Zita and Hermansson, 1997; Nielsen et al., 2001), atomic force microscopy (Butt, 1991; Ducker et al., 1991), hydrophobic interaction chromatography (Stenström, 1989), water contact angle measurement (Busscher et al. 1984), and salting out aggregation (Lindahl et al., 1981).

Any measurement of microbial hydrophobicity has inherent technical difficulties. It is not uncommon for a high proportion of test microorganisms to adhere to the walls of the

26 experimental vessel, which emphasises the need to use clean, standard vessels. Since most of the assays use aqueous suspensions, attention must be paid to the properties of the aqueous phase. Culturing conditions, and in some cases, sequential subculturing of isolates, can result in a gradual disappearance of hydrophobic properties, and bacteria with hydrophobic surface properties may have a tendency to adhere to one another, resulting in aggregation (Rosenberg and Doyle, 1990).

The lack of consensus concerning the contribution of hydrophobicity to microbial adherence to solid surfaces could, in part, be due to the methods employed to measure the cell surface properties. Variation between results from different hydrophobicity methods that should measure similar properties of cell surfaces is a common problem. One potential factor is the inconsistency of experimental conditions from one test to another (Bunt et al., 1993; Bunt et al., 1995). Failure to control parameters such as the ionic strength of the suspension buffer may result in a lack of correlation between two or more methods. Furthermore, different methods may not necessarily provide significant inter-method correlations when a microbe contains both hydrophobic and hydrophilic regions (Jones et al., 1996).

Different methods for surface property measurements have provided different results, with some researchers reporting no correlation between different hydrophobicity tests. Overcoming this problem is not easy, however, in an effort to provide improved correlations between methods, Pembrey et al. (1999) recommend suspending cells in a buffer that resembles the composition of the environment of the microbe as closely as possible. Environmental conditions can, however, vary considerably. Jones et al. (1996) therefore suggested that, for consistency of results, the buffer systems should remain constant when using different methods. Using this principle, Jones et al. (1996) obtained a significant correlation between the microbial adhesion to hydrocarbons (MATH) assay and hydrophobic interaction chromatography (HIC). Furthermore, Ahimou et al. (2001) reported that the MATH assay and HIC were better correlated with each other than with the water contact angle assay.

Researchers have reported that cell surface hydrophobicity determination using both MATH and hydrophobic interaction columns is dependent upon the buffer system

27 employed (Bunt et al., 1993; Bunt et al., 1995). Small variations in the experimental test conditions, such as the diameter of test tubes in the MATH assay, washing cells in different ionic strength solutions, or storage in different solutions, can also significantly alter the results (Lichtenberg et al., 1985; Butkus et al., 2003).

Some of the more commonly used methods for hydrophobicity testing, and the basis for each of the tests, with their limitations, are described below.

2.4.2.1 Microbial adhesion to hydrocarbons (MATH) The MATH assay has been used to measure the hydrophobicity of both bacteria (Rosenberg et al., 1980; Lichtenberg et al., 1985; van der Mei et al., 1995; Pembrey et al., 1999) and parasites (Drozd and Schwartzbrod, 1996; Capizzi-Banas et al., 2002), and is the most commonly used method to determine microbial cell surface hydrophobicity (Geertsema-Doornbusch et al., 1993). In this test a microbial cell suspension is mixed with a hydrocarbon for a period of time to allow optimal interaction of the microbes with the hydrocarbon phase. Cells may remain in the liquid phase or partition either into the liquid-hydrocarbon interface or into the hydrocarbon phase, depending on their hydrophobicity (Rosenberg et al., 1980).

Many investigators have modified the original MATH test and, as noted above, found that seemingly small variations in experimental conditions, such as the diameter of the test tubes, the pH of the suspension medium, and the volume of hydrocarbon used, can significantly alter the results (Lichtenberg et al., 1985; Bunt et al., 1993; Geertsema- Doornbusch et al., 1993). Lichtenberg et al. (1985) reported that the adherence of cells as a function of mixing time was exponential, and that increasing the hydrocarbon to water ratios also increased attachment.

Aliphatic hydrocarbons have often been used in MATH tests because of reports that aromatic hydrocarbons cause lysis of some bacterial species (Pembrey et al., 1999). Van der Mei et al. (1995) showed that adhesion to hydrocarbons was pH dependent, and significant adhesion to hydrocarbon only occurs in the absence of electrostatic

28 repulsion. They conclude that their observations disqualify MATH as a hydrophobicity assay.

Busscher et al. (1995) reported that the most commonly used hydrocarbons in the MATH assay (octane, hexadecane, xylene and toluene) are highly negatively charged in solutions in which MATH is often carried out. Zeta potentials of the aliphatic hydrocarbon droplets were generally more negative than those of the aromatic hydrocarbons, with values up to -60 mV at pH 7 (Busscher et al., 1995). Geertsema- Doornbusche et al. (1993) determined that the MATH assay does not measure the cell surface hydrophobicity, but instead measures the interplay of hydrophobicity and electrostatic interactions. Both authors concede, however, that the test could be used if the effect of the electrostatic interactions can be reduced by performing the test under ionic or pH conditions in which either the cells or the hydrocarbon droplets (or both) are uncharged. In addition, van der Mei et al. (1995) stressed that even though the MATH assay may not qualify as a hydrophobicity assay, it is useful for studying the adhesion of microorganisms to a hydrophobic surface.

2.4.2.2 Hydrophobic interaction chromatography (HIC) Hydrophobic interaction chromatography (HIC) is a separation technique that has been used for the successful purification of proteins, enzymes and removal of viruses from human plasma. In HIC, interactions between hydrophobic ligands on the adsorbate (i.e. biomolecule) and an insoluble immobilised hydrophobic ligand on the support effect separations. HIC stationary phases are manufactured by attaching hydrophobic functional groups to an agarose or polymer backbone (Queiroz et al., 2001), and similarly to MATH, it has been suggested that HIC is influenced by electrostatic interactions (Ahimou et al., 2001).

The most widely used ligands for HIC are the linear chain alkanes with or without a terminal amino group. At a constant degree of substitution on the matrix, the n-alkane ligands constitute a homologous series in a hydrophobicity scale (Tanford, 1972). Thus the relative hydrophobicity of alkane ligands is methyl < ethyl < propyl < butyl < pentyl < hexyl < heptyl < octyl. The hydrophobicity and strength of interaction increases with

29 the increase in n-alkyl chain length, though the selectivity of adsorption may decrease (Queiroz et al., 2001).

Hydrophobic adsorption of surface proteins to the HIC systems is a process that is driven by the release of water molecules from the solute and stationary phase surface (Arakawa and Timasheff, 1984; Esquibel-King et al., 1999). Salts can be used in HIC systems to either increase or decrease solute binding (Xia et al., 2004). The influence of different salts on hydrophobic interactions follows the Hofmeister (lyotropic) series (Figure 2.1) for the precipitation of proteins from aqueous solutions (Queiroz et al., 2001; Xia et al., 2004). The use of salts to increase or stabilise adsorption between two hydrophobic surfaces is called ‘salting out’. Conversely, the use of chaotropic salts can destabilise hydrophobic interactions in a process called ‘salting in’.

Increasing salting-out (kosmotropic/lyotropic) effect

Anions : PO3- , SO 2- , CH COO- , Cl- , Br - , NO- , ClO- , I- , SCN - 4 4 3 3 4       2 2 2 Cations : NH 4 , Rb , K , Na , Cs , Li , Mg , Ca , Ba

Increasing salting-in (chaotropic) effect

Figure 2.1 Hofmeister series reproduced from Xia et al. (2004)

Thus, in the presence of these salts the hydrophobic amino acids will interact with the functional groups on the stationary phase surface, forming a protein-ligand complex. After the complex has been formed, the water around the complex will redistribute. Due to the decrease of the hydrophobic exposed surface area, water will be released during the adsorption process. A chaotropic salt is less polar, hence it binds water loosely. Thus, water stays around the protein and stationary phase surfaces, thereby reducing the chance of exposing the hydrophobic surfaces. Accordingly, solute will have less chance to bind to stationary phases. The neutral salt has an intermediate influence on protein binding in HIC systems (Queiroz et al., 2001; Xia et al., 2004).

30 2.4.2.3 Contact angle To measure the contact angle, a layer of bacterial cells is deposited onto a membrane filter, and the contact angle of a drop of a diagnostic liquid on the bacterial filter cake is measured with a goniometer. In this system the values of water contact angles depend on the degree of dehydration of the cells in the filter cake. Contact angles change continuously as the filter cake dries until the level of dehydration allows the angle at the liquid-surface interface to remain stable for three to twelve seconds (Pembrey et al., 1999). The contact angle is a relative measurement of the hydrophobicity of the surface (van Loosdrecht et al., 1987b).

Mozes and Rouxhet (1987) report that contact angle of water is a significant measure of hydrophobicity. Due to certain limitations, however, they recommended that it should be supplemented with a measurement using another method. They compared contact angle to HIC, salt aggregation and a MATH test with toluene, and found that all of the methods had drawbacks.

2.4.2.4 Atomic force microscopy Atomic force microscopy produces a topographical image of the surface of a microbial cell. This is achieved utilising a microprobe (tip) mounted on a flexible cantilever scanning across the surface (Fang et al., 2000). Atomic force microscopy has been used to probe the surface macromolecules of bacteria (Camesano and Abu-Lail, 2002) and Cryptosporidium oocysts (Considine et al., 2002). In addition, it is a powerful tool for investigating nanometric physicochemical and mechanical properties of the cell surface. A force-distance curve provides valuable information on the tip-cell interaction, which is sensitive to the chemical nature of both the tip and the cell surface (Considine et al., 2002).

2.4.2.5 Salting-out The use of salting-out, as a measure of relative surface hydrophobicity of microorganisms, was first described by Lindahl et al. (1981). Similarly to the HIC method, it is a method based on the precipitation of cells by salts, with the most

31 hydrophobic cells being the first to precipitate at low salt concentrations. The results are dependent on temperature, time, pH and bacterial cell concentration (Lindahl et al., 1981).

2.4.2.6 Surface charge The determination of surface charge, which is usually measured as electrophoretic mobility and expressed as zeta potential, can be obtained by a variety of methods. The conventional laser Doppler electrophoresis apparatus determines electrophoretic mobility from a shift in the scattered light frequency by an amount dependent upon the speed of the particles (McNeil-Watson et al., 1998). This system has been superceded by an instrument based on the principles of phase analysis light scattering. This technique is reported to be useful in the study of dispersions in non-polar media, and may measure mobilities at least two orders of magnitude lower than conventional laser Coppler electrophoresis (McNeil-Watson et al., 1998). Several authors have used the Rank Bros particle microelectrophoresis apparatus to study electrophoretic mobility of Cryptosporidium oocysts (Karaman et al., 1999, Considine et al., 2002), and there are other proprietary apparatus now available, such as those used in the studies of Brush et al. (1998) and Glynn et al. (1998).

Other methods that have been used for determination of the relative surface charge of different microorganisms have been based on methods other than electrophoresis. Diethylaminoethanol (DEAE) is a weak anion exchanger, and SepharoseTM beads with DEAE ligands have been immobilised in HIC columns and used to study relative surface charges (Schiemann et al., 1987). Capizzi-Banas et al. (2002) used hydrophilic beads averaging two m in size, and measured the number attaching to Ascaris eggs to determine hydrophilic regions on the egg surface. This method was successful due to the large size of the eggs (55 x 40 m). Few comparisons between methods available for surface charge analysis exist, but Jones et al. (1996) reported a high correlation between results obtained from Zeta potential measurements and an anion exchange column for bacterial cells originally obtained from a catheter biofilm.

32 2.4.3 Cryptosporidium surface properties

The oocyst wall is thought to be around 40 nm thick, consist of three distinct layers (Harris and Petry, 1999), and is known to contain cysteine, proline and histidine (Ranucci et al., 1993). Llovo et al. (1993) determined that the Cryptosporidium wall contains N-actyl glucosamine as part of the surface glycoproteins which could have a function in invasion (Barnes et al., 1998) and, using microelectrophoresis, Karaman et al. (1999) calculated 1 - 1.5 x 106 negatively-charged sites on gamma-irradiated oocysts. Furthermore, they estimated the pKa of the surface sites was around 2.5, which suggested that the ionisable groups contributing to the oocyst surface potential could be either carboxylate groups in proteins, phosphate groups associated with phospholipids, or a combination of the two.

The theory of Karaman et al. (1999) is consistent with other researchers who suggest that the surface of the Cryptosporidium oocyst consists of a carbohydrate matrix on the outer bilayers that is characteristic of a glycocalyx (Nanduri et al., 1999), or a mucin- like glycoprotein (Barnes et al., 1998). Butkus et al. (2003) used diffuse reflectance spectra from freeze-dried oocysts, to identify alkyl C-H stretches and carbonyl functional groups on oocyst surfaces, and Considine et al. (2002) reported the presence of a ‘brush-like’ conformation of proteins extending from the oocyst surface.

2.4.3.1 Oocyst hydrophobicity The hydrophobicity of Cryptosporidium oocysts has been characterised differently by different research groups. Drozd and Schwartzbrod (1996) used the MATH assay to measure the hydrophobicity of oocysts that had been purified using formaldehyde and ethyl acetate and stored in potassium dichromate. The percentage of oocysts that adhered to octane was at a minimum at pH 7 and increased with both increasing and decreasing pH. Adherence to octane also increased with increased conductivity. Since the range of adhesion was only 10 to 45%, the authors concluded that oocysts were of low hydrophobicity. These results were similar to those obtained by Inoue et al. (2006) from oocysts stored in potassium dichromate for up to 100 days, but differed from the results of Hsu and Huang (2002) who reported that purified oocysts showed marked

33 hydrophobicity using the MATH assay with octane, and that the hydrophobicity significantly increased as the pH decreased from 11 to 2.4.

Brush et al. (1998) estimated hydrophobicity by attachment of oocysts, purified by sucrose flotation, to polystyrene as a function of oocyst age and ionic strength of the suspending medium. Hydrophobicity of the oocyst in their study ranged between 10 and 90%. Fresh (two week old) oocysts showed strong adhesion (greater than 80%) in low ionic strength (0 to 20 mM) solutions, while only 10 to 20% adhered under conditions of higher ionic strength (20 to 95 mM). When they used aged (> 8 week old) oocysts approximately 60 to 80% of the oocysts adhered to the polystyrene surface in the ionic strengths tested. Similar results were obtained by Dai et al. (2004) with 65% of their oocysts attaching to polystyrene, though details of oocyst extraction and age were not presented for their study.

Hydrophobicity is affected by environmental conditions such as pH and ionic strength. In addition, natural organic matter has the ability to coat particles and change their properties. Dai and Hozalski (2002) studied the effect of natural organic matter on the surface properties of oocysts and found that a coating of natural organic matter can increase the hydrophobicity of Cryptosporidium oocysts. The authors, however, indicated that the influence of such increased hydrophobicity was small. Kuznar and Elimelech (2005) showed that inactivation of oocysts by heat and formalin treatment increased the hydrophobicity of the oocyst surface, which they explained as likely due to alterations in the surface protein structures due to the treatments. They further showed that with the removal of surface proteins from the oocysts (as may occur during environmental ageing) there was a corresponding significant increase in the amount of attachment to surfaces (Kuznar and Elimelech, 2006).

2.4.3.2 Oocyst surface charge Using electrophoretic mobility as a measure of surface charge, the net surface charge of Cryptosporidium oocysts has been measured in a number of studies. Drozd and Schwartzbrod (1996), Ongerth and Pecoraro (1996), Karaman et al. (1999), Bustamante et al. (2001) and Hsu and Huang (2002) measured a negative surface charge for purified

34 oocysts in distilled water, river water and NaCl solutions, with the charge remaining negative over a range of pH values. In contrast, Brush et al. (1998), who measured the electrophoretic mobility of oocysts that had been purified by sucrose flotation, detected no net surface charge at pH values between 2 and 10. Oocysts purified using ethyl acetate and Percoll®, which had been used variously in the aforementioned studies, exhibited a negative surface charge. They concluded that purification methods employing chemicals such as ethyl acetate, formalin and Percoll® may alter the surface characteristics of oocysts (Brush et al., 1998).

Oocyst storage conditions, and even the solutions in which oocysts are washed prior to analysis, have been shown to have a significant impact on their electrophoretic mobility. Butkus et al. (2003), using oocysts extracted by chemical defatting, sucrose and Percoll® density gradient centrifugation and stored in phosphate buffer with antibiotics at 4 °C, found that washing the oocysts in distilled water can impart a negative charge to a fraction of oocysts in the suspension. Brush et al. (1998) found that there was no statistical difference in electrophoretic mobilities of oocysts stored in deionised water with or without antibiotics, though oocysts stored with antibiotics had less variability in the results obtained (i.e. a lower standard deviation).

Butkus et al. (2003) also determined that inactivation of oocysts with formalin did not influence their electrophoretic mobility, although Brush et al. (1998) reported an increase in electrophoretic mobility with formalin inactivated oocysts. Regardless, Butkus et al. (2003) felt that their results explained the lack of consensus from other researchers regarding oocyst surface charges, although the washing procedures used in other studies were not always clearly identified. They concluded it was possible that ionised functional groups on the oocyst are coated with up to 20 - 30 nm of carbohydrate-rich glycocalyx (Nanduri et al., 1999), and washing oocysts in low-ionic- strength solvents may result in at least partial removal of the glycocalyx and cause an increase in the electrophoretic mobility. Such a result is consistent with those reported by Pembrey et al. (1999) for other microorganisms.

Thomas et al. (2001), using oocysts extracted by sucrose flotation and then stored in potassium dichromate, reported that oocysts washed in PBS had near zero

35 electrophoretic mobility (zeta potential of -5 mV). They concluded that this low mobility was due to compression of the double layer by the high ionic strength of the PBS solution (0.172 M). Dilution of the PBS with deionised water by 10- and 100- fold resulted in a shift in the zeta potential to -23 and -33 mV respectively, correlating more closely with the results of other researchers (Drozd and Schwartzbrod, 1996; Ongerth and Pecoraro, 1996; Karaman et al., 1999; Bustamante et al., 2001; Hsu and Huang, 2002). Thomas et al. (2001) also obtained a similar effect from 10-fold dilutions of -1 -3 -2 NaCl from 10 to 10 M. The zeta potential of oocysts suspended in 10 M CaCl2 , however, was close to zero, at -2 mV.

Environmental factors other than pH and ionic strength also have the ability to change the surface characteristics of oocysts. The zeta potential of oocysts can be substantially modified in the presence of fulvic acid and other natural organics in water (Dai and Hozalski, 2002).

2.5 Aggregation and Attachment

2.5.1 Particle aggregation

The aggregation of particles is a complex process, and is not well understood or described accurately in mathematical terms (Gardner et al., 1998a). However, for the process of aggregation of colloidal particles to occur, the particles must first be brought into close proximity by a transport mechanism to give rise to a collision. An aggregate will then be formed if the net inter-particle force is attractive and strong enough to overcome thermal agitation and hydrodynamic drag (Vanni and Baldi, 2002).

There are three recognised mechanisms by which particles in suspension collide. The first is particle movement due to random Brownian motion, which is a function of thermal energy. It dominates interactions only at early times when a monomodal size distribution exists, and is only usually appreciable for sub-micron sized particles (Gardner et al., 1998a). The second mode of particle collision is by fluid shear, where particles collide due to spatial velocity gradients in the fluid, and the third mechanism is collision due to differential settling velocities of particles (Gardner and Theis, 1996;

36 Gardner et al., 1998b). Aggregate restructuring is also an important mechanism in explaining aggregation kinetics (Thill et al., 2001).

Gardner and Theis (1996) developed a model that described the kinetics of particle aggregation by a numerical solution of the von Smoluchowski equation which describes the kinetics of coagulation. This model was developed to interpret laboratory data. It was shown that the attachment distance has a profound influence on the predicted aggregation kinetics, as did the choice of solid/solution interfacial model (Gardner and Theis, 1996). The same research group also reported that large colloidal hematite aggregates react rapidly with sub-micron fractions under conditions of applied shear stress (Gardner et al., 1998a), and that the fractal dimension of aggregates varies depending on whether particles have a low or high collision efficiency (Gardner et al., 1998b). Furthermore, cluster-cluster aggregation was shown to be a significant mechanism in aggregate formation (Gardner et al., 1998b). These latter findings help explain why models that replace real aggregates with impermeable spheres of a size chosen such that the overall permeability remains unchanged have been shown to successfully approximate the hydrodynamic behaviour of aggregated particles (Gmachowski, 1996).

2.5.2 Oocyst attachment and aggregation

Some studies have noted that Cryptosporidium oocysts may not occur as discrete particles in the environment, but may have a tendency to stick to other particles or occur in clumps (Bukhari and Smith, 1995; Searcy et al., 2005). There are also many reports, however, which conclude that oocysts do not readily attach to solids or aggregate (Butkus et al., 2003; Kuznar and Elimelech, 2005). Any tendency for oocysts to adhere to other particles is most likely conferred by the surface chemistry of the oocyst.

It has been hypothesised that proteins can extend from the oocyst surface due to charge repulsion between ionisable surface groups, thus giving the oocyst a brush-like conformation (Considine et al., 2002). This may give rise to steric forces that promote oocyst stabilisation (Kuznar and Elimelech, 2006). In addition, electrostatic repulsive forces between negatively-charged oocysts and suspended sediments may hinder

37 oocyst-particle attachment. Electrostatic interactions are highly dependent on both solution and surface chemical conditions, so oocyst-particle aggregation can be favoured under specific solution chemical conditions or with particular types of particles.

Using direct microscopy, Searcy et al. (2005) observed oocysts attached to suspended sediments and showed that oocysts settle at a much higher rate when associated with sediments. In contrast, Dai and Boll (2003) studied particle attachment to oocysts using a combination of flow cytometry, immunofluorescence staining and confocal microscopy. Their oocysts did not attach to soil particles at any pH value tested (4.5, 7.1 and 10.5) at a soil concentration of 2 mg.mL-1. Altering the ionic strength of the suspending solution did not affect attachment. Their lack of attachment may be a function of low collision rates with the soil due to the low soil concentration in suspension rather than due to oocyst stability.

Kuczynska et al. (2005) investigated the effect of bovine manure on the attachment of Cryptosporidium oocysts to sandy loam and clay loam soils. They showed more than 72% of oocysts attached to their soils, but prior incubation of the oocysts with manure enhanced the attachment. The maximal attachment was found when the manure was diluted from 1% to 0.1% in the final soil-manure-oocyst suspension. Considine et al. (2002) noted an increasing negative zeta potential of oocysts by the addition of total organic carbon, which is consistent with the adsorption of negatively-charged dissolved organic carbon molecules to the oocyst surface, thereby increasing the surface charge density.

Medema et al. (1998) reported that a significant proportion of oocysts attached to particles in secondary effluent, with 35% attaching almost instantly, with up to 70% attachment after 24 hours. The sedimentation of the oocyst-particle complexes were determined by the sedimentation characteristics of the individual particles, so they concluded that sedimentation is probably a significant factor in the environmental ecology of Cryptosporidium oocysts (Medema et al., 1998).

38 In purified oocyst suspensions from various suppliers, Butkus et al. (2003) measured no aggregation when oocysts were suspended in 0.5 M NaCl. In contrast, Thomas et al. (2001) noted that 0.15 M NaCl caused irreversible adhesion of oocysts to synthetic filtration media such as polycarbonate, cellulose-ester and polyethersulfone. Dai et al. (2004) obtained little, if any, adhesion of oocysts to glass beads with and without a coating of fluorosiloxane or aminosiloxane (both substances were negatively charged and hydrophobic but glass beads without the coating were negatively charged and hydrophilic). The only appreciable attachment they achieved was 18% with glass beads coated with a cationic polymer.

Some researchers have reported that surface charge is more important than hydrophobicity in oocyst attachment to solid surfaces (Dai et al., 2004), while others suggest that charge-based interactions with colloids may be most pronounced when the oocyst lose their infectivity (Walker and Montemagno, 1999). Various researchers have observed that non-viable oocysts are more likely to aggregate or adhere to environmental particles than their viable counterparts (Bukhari and Smith, 1995; Anguish and Ghiorse, 1997; Kuznar and Elimelech, 2006).

2.6 Sedimentation Kinetics

2.6.1 Gravitational settling

For any given settling time and at a given distance from the surface, there is a critical cutoff size where all particles larger than this size will have settled further down the column and will not be present at the specified depth, even if they started from the surface. Those particles smaller than the critical size will also settle, but those that fall below the specified depth will be replaced by those from above. Therefore, the density of this fraction of particles does not change. Different particle size fractions within a settling column can be collected by exploiting this size-dependent settling velocity of the particles (Palmer and Troeh, 1995).

Stoke’s Law describes the relationship of settling time to particle diameter. It specifies that the velocity of settling is proportional to the square of the diameter of the particles,

39 assuming that particles are rigid, smooth, and spherical, and is expressed as Stoke’s equation: 2 Vs = g d (p-w) 18μ

Where Vs = velocity of fall of particle in m/sec g = acceleration due to gravity in m/sec2 d = the particle diameter in m 3 p = density of the particles in kg/m 3 w= density of the water in kg/m and μ = dynamic viscosity of water in Ns/m2

Stoke’s equation can be used to describe the rate of settling of particles in the size range 0.5 to 50 m. Stoke’s equation can be modified by the addition of a term in the denominator which accounts for a microorganism’s non-spherical shape, but the relationship between morphology and sedimentation velocity is not fully understood (Harvey et al., 1997). For particles that are spherical in shape, this additional term is generally omitted or assumed to be one.

The particle concentration in settling columns, theoretically, should be low enough that every particle has sufficient space to settle independently, but high enough that it can accurately be detected (Irani and Callis, 1963). A soil concentration of 0.5% (w/v) has been reported to give acceptable results. However, often these two restrictions cannot be met simultaneously and in those cases a compromise must be made (Irani and Callis, 1963).

Stoke’s Law has been used to calculate the theoretical settling rate of microorganisms including bacteria and Cryptosporidium oocysts (Pedrós-Alió et al., 1989; Young and Komisar, 2005b; Dai and Boll, 2006). Some researchers have shown that Stoke’s Law adequately describes the settling velocity of free settling microbes (Pedrós-Alió et al., 1989; Dai and Boll, 2006), while forces of interaction not included in Stoke’s law dominate settling in other systems (Pedrós-Alió et al., 1989; Kulkarni et al., 2004).

40 2.6.2 Oocyst sedimentation

The settling velocities of Cryptosporidium oocysts have been determined by various researchers, with Stoke’s Law being commonly used to predict their settling velocities. The inputs to Stoke’s equation are particle diameter, particle density, water density, acceleration due to gravity and absolute viscosity of water. Of those inputs, if settling occurs in water then only oocyst density and, to a smaller degree, oocyst diameter will cause any significant change in the calculated settling velocity, since the values for the other parameters are consistent for specified temperatures. Cryptosporidium oocysts are generally regarded as being round to ovoid, 4 – 6 μm in diameter (US-EPA, 2005), and approximately 1050 kg.m-3 in density (Gregory, 1994). However, large variations in oocyst densities, mostly attributed to oocyst age, have been observed (Young and Komisar, 2005b).

Medema et al. (1998) measured a mean oocyst diameter of 4.9 μm with a standard deviation of 0.3 μm. Their oocysts were purified by CsCl density centrifugation and stored in 5% potassium dichromate and were used between two and eight months of age. A geometric mean density of 1045 kg.m-3 was calculated for the oocysts from data fitted to a log-normal distribution (Medema et al., 1998). In contrast, Young and Kosimar (2005a) measured the density of purified oocysts 14 weeks after purification by sucrose flotation, and unpurified oocysts in calf faeces after various storage times in the faeces. They measured oocysts ranging from 1005 - 1108 kg.m-3 over the duration of the study. At 1.4 weeks of age the unpurified oocysts were 1070 – 1073 kg.m-3. However, 11- and 12-week old unpurified oocysts showed a bimodal range of densities, with densities most frequently being 1005 - 1041 kg.m-3 or 1077 - 1108 kg.m-3. Those oocysts with densities above 1077 kg.m-3 were generally viable, while those in the 1005 - 1024 kg.m-3 range were generally non-intact. Oocysts in the range 1024 - 1041 kg.m-3 were generally intact but non-viable. In comparison, the purified 14-week old oocysts were 1067 - 1070 kg.m-3. The mean diameter of Dai and Boll’s (2006) 13-week old oocysts was 6.6 ± 1.1 μm, and the average density was 1009 kg.m-3.

Medema et al. (1998) calculated an initial settling rate of oocysts suspended in Hanks Buffered Saline Solution of 0.35 μm.s-1. Although they saw no variation in oocyst morphology, with 99.3% of oocysts showing internal contents, the range of oocyst

41 densities obtained resulted in slower settling with the less dense oocysts. Dai and Boll (2006) estimated that oocysts settled at 0.36 μm.s-1, which was similar to the theoretical velocity predicted by Stoke’s equation. Settling velocities measured by Young and Komisar (2005b) were also similar to those predicted by Stoke’s Law. Importantly, the older and non-viable oocysts settled significantly more slowly than the fresher and more viable oocysts due to the different oocyst densities (Young and Komisar, 2005a). Young and Komisar (2005b) also obtained settling velocities significantly faster than those predicted by Stoke’s law. This was presumably due to an interaction with the particles that were associated with the unpurified faecal material. In contrast, Kulkarni et al. (2004) obtained setting velocities an order of magnitude lower than those predicted by Stoke’s Law. However, they did not measure their oocyst densities, and it is possible that the oocysts were less dense than assumed.

In combination with settleable solids, Searcy et al. (2005) noted that the rate of oocyst settling depended primarily on the type of sediment with which the oocysts were mixed. Changes in background water conditions had a relatively small impact on the extent of oocyst-particle association and the resulting oocyst settling. The settling velocity of purified oocysts increased from 0.76 μm.s-1 when oocysts were not mixed with sediments to 12.8, 53.3 and 7.9 μm.s-1 when mixed with kaolinite, iron oxide particles, and creek sediment respectively.

2.7 Oocyst Transport

There are numerous publications on the presence and detection of Cryptosporidium oocysts in water bodies (Ionas et al., 1998; Skerrett and Holland, 2000; Buckley and Warnken, 2003). Fewer report on oocyst transport through soils (Mawdsley et al., 1995; Brush et al., 1999; Darnault et al., 2003), yet rapid soil transport has been noted (Harter et al., 2000). In medium to coarse sand, Harter et al. (2000) reported that Cryptosporidium oocysts move 10 – 30% faster than a conservative tracer, presumably due to the more complex pathway taken by the bulk water which moved throughout the soil pores while the larger oocysts would be expected to move largely through macropores. They also showed that the observed late-time oocyst elution can be quantitatively explained by postulating that a significant fraction of oocyst filtration is

42 reversible and subject to time-dependent detachment. Hijnen et al. (2005) reported that the mobility of oocysts in sandy soils was low, with soil type significantly affecting transport. Subsurface transport via preferential flow may create a significant risk of groundwater contamination, even though only a small percentage of applied oocysts may be transported (Darnault et al., 2003).

Rainfall and runoff events are major factors affecting the presence of total microbial load, including Cryptosporidium, in surface waters and drinking water reservoirs (Atherholt et al., 1998; Kistemann et al., 2002; Ashbolt and Roser, 2003). In the United States, agricultural animal waste has been identified as a national threat to water quality because runoff following rainfall events carries nutrients and pathogens into surface waters (Fayer, 2004). Nearly all pre-weaned calves become infected and excrete large number of oocysts in their faeces, and older cattle continue to excrete oocysts, albeit in smaller numbers per gram of faeces. Consequently, cattle are believed to be a major source of waterborne Cryptosporidium.

Although some outbreaks of water-related cryptosporidiosis have been associated with rainfall events (MacKenzie et al., 1994; Rose, 1997), overland transport mechanisms from faecal pat to stream are not well understood (Dai, 2003; Davies et al., 2005a). Publications such as those by Dai and Boll (2003) and Butkus et al. (2003) have suggested that Cryptosporidium oocysts have a low tendency to attach to particles or each other at neutral pH, and yet others indicate that the percentage of attachment is high (Medema et al., 1998). Reliable information about the survival and movement of Cryptosporidium oocysts is required to model oocyst transport in drinking water catchments, for risk assessment, and to develop effective control practices such as vegetated riparian buffer zones (Atwill et al., 2002).

Vegetation significantly reduces overland transport of Cryptosporidium oocysts compared to bare soil (Davies et al., 2004), with up to four log10 reduction of spiked oocysts released from a faecal pat over a distance of one metre of vegetation (Atwill et al., 2002; Davies et al., 2004). Cryptosporidium transport has also been shown to be significantly affected by runoff volume, intensity and duration, vegetation, degree of slope (Davies et al., 2004). Soil type also effects oocyst transport, with sandy loam

43 soils being the least effective at removing oocysts from overland flow relative to silty clay loam and loam soils (Atwill et al., 2002). An investigation into the transport of Cryptosporidium during rainfall from fresh and aged faecal pats showed that higher numbers of oocysts were found down-slope of the pats when they were fresh (Ferguson et al., 2006). A riparian buffer strip for slopes less than 20% and a length of greater than three metres has been recommended to remove 99.9% of Cryptosporidium oocysts from agricultural runoff during mild to moderate rain events (Atwill et al., 2002).

Although pH and ionic strength are recognised as affecting both the surface properties of colloids and colloid collision, there is some evidence that the pH and ionic strengths commonly encountered in soil and fresh water are not sufficient to influence microbial cell attachment (Jewett et al., 1995). Dai et al. (2003) suggested that little energy is required to detach settled oocysts into overland flow, and once detached they will travel freely in the water and not as part of the particulate sediment load.

Cryptosporidium oocysts are often considered to move conservatively downstream and to have little interaction with sediments because of their small size and low specific gravity (Searcy et al., 2006). The low sedimentation velocity has also been suggested to be one of the major factors responsible for the transmission of oocysts via drinking water (Sréter and Széll, 1998). Hyporheic flow is the multidimensional percolating- flow mixing of shallow groundwater and surface flow. The process of hyporheic exchange has been shown to lead to high rates of suspended particle deposition in sediment beds, even when the suspended particles are very small and have no appreciable settling velocity (Ren and Packman, 2002). A recent report into the effects of hyporheic exchange with Cryptosporidium oocysts indicates that these hydrodynamic interactions between an overlying flow and sediment bed caused both free and attached oocysts to accumulate in the sediments, thus reducing their concentrations in the surface water (Searcy et al., 2006).

2.8 Summary of Current Knowledge

Cryptosporidium is ubiquitous in the environment. However, with the exception of C. parvum, there is some evidence that Cryptosporidium is not a zoonotic pathogen.

44 Therefore, many of the oocysts excreted by domestic and wild animals may not pose a significant health risk to humans. Given that routine methods for detection and enumeration of oocysts from water do not differentiate between different species or genotypes of Cryptosporidium, all types are generally included in risk assessments and modelling, making them conservative estimates.

Calves are one of the most well recognised domestic animals that harbour Cryptosporidium that may be infectious to humans, and they can excrete large numbers of oocysts per animal per day. Hence, collection of their faecal material remains one of the easiest ways of obtaining high concentrations of oocysts. The faecal debris, however, often complicates laboratory analyses and therefore most researchers attempt to remove it to obtain purified Cryptosporidium oocysts.

The literature suggests that purification methods may affect the surface properties of Cryptosporidium oocysts, with some researchers avoiding defatting but still using hyper-osmotic density solutions in order to obtain pure oocysts. Reports on the surface properties of the purified oocysts are conflicting, though there generally seems to be a consensus that oocysts purified using sucrose and without defatting have only slightly negative or neutral surface charges. However, storage conditions and age may also play important roles in oocyst surface properties. The measurement of surface properties in the presence of faecal solids has, so far, not been possible.

Determining the surface properties of Cryptosporidium oocysts is complicated by pH and ionic strength effects. There has been little consistency between methods and oocyst suspension solutions used by different researchers, and often technical details of age and oocyst purification methods and storage conditions are not specified in published reports. Similarly, the ability of oocysts to attach to soil and other environmental particles is complicated by the same details.

There is no doubt that the concentrations of Cryptosporidium oocysts in waters increase under the influence of rainfall. Generally the majority of the oocysts are thought to be from overland runoff and sewage treatment plant effluents, although recently the idea of resuspension of oocysts from sediments that had accumulated oocysts from settling and

45 hyporheic flow has been suggested (Searcy et al., 2006). A key question remains, however, as to whether oocysts travel freely in water or are attached to sediment particles during overland flow. This study focused on providing evidence to answer this outstanding issue for catchment managers and for the modelling of oocyst transport.

46 CHAPTER THREE

CRYPTOSPORIDIUM OOCYST PREPARATIONS

3.1 Introduction

As discussed in Chapter two, different preparations of oocysts may yield differences in the surface properties of oocysts (Ongerth and Pecoraro, 1996; Brush et al., 1998; Considine et al., 2002), although Dai and Boll (2003) reported that oocyst treatment had no apparent effect on their studies of attachment to soils. Those who have reported differences between untreated and treated oocysts have defined their ‘untreated’ oocysts as having been purified by sucrose flotation using continuous flow centrifugation (Brush et al., 1998). Sucrose solutions have a high ionic strength, are hyper-osmotic with dehydrating properties (Ryley et al., 1976), and could also potentially affect oocyst surfaces.

Ideally, a method that involves no harsh chemical or strongly ionic solution would be preferred for oocyst purification. In addition to obtaining high quality preparations of Cryptosporidium oocysts, another factor to consider was the need for high oocyst concentrations necessary for some of the fate and transport experiments in the AwwaRF-CRCWQT project (Davies et al., 2005a) and for surface property experiments described in later chapters of this thesis.

The aim of Chapter three was to establish an alternative method to diethyl-ether defatting and flotation steps so as to minimise any impact to oocysts and their surface properties.

3.2 Materials and Methods

3.2.1 Particle sizing

Particle size profiles of samples were obtained using laser interference on a Malvern MastersizerTM E Particle sizing apparatus (Malvern Instruments Ltd., Malvern, Worcestershire, UK), with analysis of samples based on Standard Method 2560D

47 (APHA 1998). Instrument-based data analysis was undertaken using the Malvern MastersizerTM software version 1.2 and the Malvern ‘Standard presentation’ setup (spherical particles, density 1.00 g.mL-1, polydispersed size distribution, refractive index (RI) of 1.54, and water dispersant RI=1.33) using a continually stirred 14.3 mm path length cuvette. With the 300-mm focal length lens, particles were separated into 31 bands (0.5 to 600 μm). The top and bottom size boundaries of each band differed by a factor of 1.2 (Roser et al., 2003). In addition to the 31 bands for light scattering measurement, a centre ring measured the unscattered light energy for determination of the obscuration, which is essentially a measure of the number of particles in the sample. The obscuration range for accurate particle sizing was in the range 10 to 30%. Samples that initially had an obscuration reading higher than 30% were diluted so that they were in the acceptable range. This was to avoid multiple scattering effects as described by Guardani et al. (2002).

To obtain a single particle size profile, the MastersizerTM E was configured to ‘sweep’ the sample 5000 times, with each sweep consisting of a single ‘snapshot’ of all detector values. Each sample was measured three times to obtain three particle size profiles, and the means for each band calculated to give an average particle size profile for the sample. The results were volume based, with the distribution being expressed in terms of volumes of equivalent spheres. When the result lists, for example, 10% of the distribution is in the size category 2 - 2.4 μm, this means that the volume of all particles with diameters in this range represents 10% of the total volume of all measured particles in the distribution.

Data was graphically presented by charting the percentage of the total solids by volume against the 31 bands, with the lower limit of each size range displayed on the horizontal axis of the chart. The scale of the horizontal axis of the particle size charts are not linear since the size boundaries of each size range differ as described above.

48 3.2.2 Extraction of Cryptosporidium oocysts from calf faeces

3.2.2.1 Collection of calf faeces Calf faeces were collected from a dairy farm on the edge of Sydney (Leppington Pastoral Co. Pty. Ltd, Bringelly, Sydney). The faeces were collected from the floor of calf pens which held individual calves at 7 to 21 days of age. The faeces that appeared the freshest were collected and diluted approximately 1:4 with cold 0.01 M PBS (0.138 M NaCl, 0.0027 M KCl, pH 7.4 @ 25 ºC, Sigma Aldrich, Sydney, Australia) and transported to the laboratory packed on ice. The faecal slurries were mixed thoroughly with a wooden applicator stick. A small amount of the slurry was smeared onto a glass microscope slide and allowed to air dry. The smear was then heat fixed and stained with EasyStainTM (BFT Decisive Microbiology, North Ryde, NSW, Australia) for 15 minutes before being rinsed from the slide with reagent water (reverse osmosis water with ion exchange, removal of organics, and 0.22 μm filtration: MilliQ (MQ) filtration system, Millipore, North Ryde, Sydney). The slide was then blotted dry and viewed at a magnification of 200x using epifluorescence microscopy (excitation at 450 to 490 nm; long pass emission at 520 nm; Nikon Optiphot-2 with EFD-3 fluorescence attachment). The number of fluorescing Cryptosporidium oocysts were ranked according to the density observed on the slide. Calf faeces from which high numbers of Cryptosporidium oocysts were observed on the slides were sieved through wire mesh with a pore size of approximately 1.5 mm and stored at 4 °C until further extraction was required.

3.2.2.2 Removal of faecal lipids Five millilitre volumes of the sieved faecal slurry were added to 50 mL centrifuge tubes and diluted to 40 mL with reagent water. Samples were defatted by adding 10 mL of diethyl-ether to the diluted faeces and mixing for two minutes by slow inversion. The slurry was then centrifuged at 2500 x g for 30 minutes at 4 °C and the supernatants, including the lipid layer and diethyl-ether, were aspirated from the pellet. This procedure was based on the method of Upton (1997), with the exception that a greater centrifugation force was used. The greater force of 2500 x g gave a more compact

49 faecal lipid layer that was easier to remove than the one obtained at 1000 x g, and the recovery of oocysts was increased (data not shown).

The defatting step was repeated if lipids were still visible in the oocyst suspension after the first defatting step. To achieve this, the pellet was resuspended in 40 mL of reagent water and diethylether added and centrifuged as described above. The pellets from the final defatting step were washed by resuspension in reagent water and centrifugation at 2500 x g for 20 minutes at 4 °C. The oocysts were then separated from other debris by either sucrose or salt flotation.

3.2.2.3 Sucrose flotation The washed pellets resulting from the diethyl-ether defatting step were resuspended in a 55% (w/v) sucrose solution (specific gravity 1.2) (Upton, 1997) and 6 mL volumes were distributed into 15 mL centrifuge tubes. Two millilitre volumes of reagent water were carefully overlaid on top of the sucrose-faeces suspension and centrifuged at 1000 x g for 30 minutes at 4 °C. The water layer and interface between the two layers were carefully collected, with any material below the oocyst layer near the interface being avoided. The oocysts were washed at least three times in reagent water with centrifugation at 2500 x g for 20 minutes. The final oocyst suspension was stored at 4 °C in either regent water or PBS.

3.2.2.4 Salt flotation The defatted and washed pellets were resuspended in a volume of reagent water which was just enough to resuspend the pellet. This suspension was mixed at a ratio of approximately 1:5 with a saturated NaCl solution (specific gravity 1.2) (Weber et al., 1992). Six millilitre volumes were distributed into 15 mL centrifuge tubes and 2 mL volumes of reagent water were carefully overlaid the NaCl solution. The tubes were then centrifuged at 1000 x g for 30 minutes at 4 °C. The water layer and interface between the two layers were collected, and the oocysts were washed as described above for sucrose flotation.

50 3.2.2.5 SephadexTM G-50 gel filtration To obtain Cryptosporidium oocysts which had not been subjected to chemicals or high ionic strength solutions, sieved faecal slurries were passed through a SephadexTM G-50 (SephadexTM G-50 Fine, Amersham Biosciences Pty. Ltd., Baulkham Hills, NSW, Australia) gel filtration medium. The SephadexTM G-50 gel was swelled by soaking the dry gel in an excess of PBS for 4 hours. The gel was then poured into a column made with disposable 25 mL pipettes that had had the tops removed and were plugged at the base with glass wool (Thiriat et al., 1998). Attached to the bottom of the pipette was a piece of tubing with a clamp to regulate the flow through the column. Starting the gravity flow through the column initiated gel packing, and as the gel settled, more SephadexTM G-50 was added until the gel column had a volume of 25 mL. When packed to the final height of 25 mL, a layer of PBS was kept on top of the column until use to minimise the risk of the column drying out.

Before commencing filtration of the oocysts, the PBS was allowed to pass through the column until the meniscus was equal with the top of the SephadexTM G-50 gel. Approximately 0.5 mL of a sieved faecal slurry was added to the top and allowed to filter into the gel. When the meniscus of the faecal slurry was equal with the top of the column, PBS was continuously added to the top of the column so that the column did not run dry. The smaller lipids with which the oocysts were associated in the column could be seen as a pale yellow band of material moving through the column and were collected in a 50 mL centrifuge tube. The oocysts were centrifuged, resuspended in PBS and filtered through 11-μm pore size nylon netting (Millipore, North Ryde, Sydney).

3.2.2.6 Octyl-SepharoseTM Some oocyst suspensions resulting from the SephadexTM G50 columns were further clarified by mixing oocyst suspensions with octyl-SepharoseTM beads (Octyl SepharoseTM 4 Fast Flow, Amersham Biosciences Pty. Ltd., Baulkham Hills, NSW, Australia) to remove some of the remaining lipids. The octyl-SepharoseTM was used to measure the hydrophobicity of Cryptosporidium oocysts, and results of the assay showed no attachment of oocysts to the beads (Chapter four). They were, however,

51 efficient at removing some of the lipids in the ‘untreated’ oocysts, as observed by the colour of the beads changing from white to pale yellow after incubation with the oocyst suspension. The oocyst suspensions prepared using the SephadexTM G-50 method, with or without lipid removal with octyl-SepharoseTM beads, are referred to in this dissertation as untreated oocysts, since no organic solvent or high ionic strength solution was used in their preparation.

3.2.2.7 Flow cytometry sorting Flow cytometry was undertaken with a FACSCaliburTM (Becton Dickinson Australia/New Zealand, North Ryde, NSW, Australia) operating with a 15-W air-cooled argon-ion laser emitting at 488 nm with standard filters supplied. The sheath fluid used was Isoton IITM (Balanced Electrolyte Solution, Beckman Coulter, Australia), and a polypropylene tube (12 x 75 mm) was used as the flow tube. Detection was by way of forward scatter (FSC) and side scatter (SSC) only. The voltage applied to each detector was set at E-1 for FSC and 280 for SSC. Cell sorting was performed to select possible oocysts without any fluorochrome attached. The sort settings are described in Figures 3.1 to 3.4, each with the same parameters (FSC versus SSC). R1 was gated to sort possible Cryptosporidium oocysts.

Oocysts suspensions sorted through the flow cytometer consisted of dilute calf faecal material, oocysts extracted through the SephadexTM G-50 column only, as well as those that had also been further treated with the octyl-SepharoseTM beads, and oocysts that had been extracted using diethyl-ether defatting and sucrose flotation. All oocyst suspensions used with flow cytometry were prepared on the same day from one calf faecal slurry.

To verify whether oocysts were being sorted by the flow cytometer, a total of 100 sorted particles were filtered onto a 13-mm-diameter polycarbonate membrane with pore size 0.8-m (Millipore, Australia) and stained with EasyStainTM. The oocysts were counted using epifluorescence microscopy as described previously.

52 3.3 Results

3.3.1 Diethylether defatting and sucrose flotation

Particle size profiles using the MastersizerTM showed that in some cases after one round of diethyl-ether and sucrose extraction, a wide variety of non-Cryptosporidium oocyst sized particles were in the suspension (Figure 3.5). A distinct peak was detected around the 4-μm size range which corresponds to the expected size of Cryptosporidium oocysts (US-EPA, 2005), however these only made up 5.5% of the particles by volume. A second round of diethyl ether defatting and sucrose flotation removed many of these unwanted particles (Figure 3.5), which suggests that they may have included residual lipids from the faecal slurry. The particle size profile of the oocysts suspension that underwent two diethyl-ether/sucrose processes had a peak in the Cryptosporidium oocysts size range of only 12.5% of the total volume of solids. This was affected by the large peak around 300 μm which is an artefact commonly seen in solutions with low obscuration, even when in the acceptable range, and is attributed to small bubbles in the sample.

Some large particles around 100-μm in size were observed in the size distribution profile after the second round of defatting and sucrose flotation, but these were removed when the sample was filtered through the 11-μm pore size nylon netting (data not shown). Some faecal slurries did not require the second round of defatting and sucrose flotation to obtain a clean suspension, reflecting the variability in the consistency and composition of various faecal suspensions collected. Large numbers of oocysts (106 to 1010) were commonly obtained per extraction.

3.3.2 SephadexTM G-50 column

Many of the lipids and other particulates in the calf faecal slurry were removed in the SephadexTM G-50 column, as was evident from the vast amount of material from the slurry being retained in the column. The oocysts moved through the column in a band that could be seen by the pale yellow colouration of the smaller lipids that moved with the oocysts. Some of these lipids could be removed by their binding to octyl- SepharoseTM beads, which were hydrophobic and did not bind to the oocysts (as

53 described in Chapter four). Oocysts suspensions of around 104 to 105 mL-1 were commonly obtained using this method.

3.3.3 Flow cytometry

Flow cytometry with sorting was used in an attempt to obtain a clean suspension of Cryptosporidium oocysts that had not been subjected to any treatments such as diethyl- ether or high ionic strength solutions. The more highly purified Cryptosporidium oocysts were sorted more efficiently than the less treated oocysts as shown in Table 3.1. Oocysts that had been extracted using diethyl-ether and sucrose were sorted at an efficiency of 88%, and those that had passed through the SephadexTM G-50 column with a further removal of lipids using octyl-SepharoseTM showed a sort efficiency of 77%. However, oocysts that had undergone less purification were sorted at a much lower efficiency due to more miscellaneous faecal particles being present. As the amount of faecal debris increased, the percentage of oocysts being sorted decreased. When the oocyst-containing faecal slurry was passed through the flow cytometer, none of the 100 sorted particles were Cryptosporidium oocysts.

The improved sorting with the more purified oocyst suspensions can be further explained with the aid of the sort charts shown in Figures 3.1 to 3.4. A large number of particles were detected by the flow cytometer in the dilute calf faeces, but the sort area did not contain an obvious discrete population of particles that could potentially be oocysts (Figure 3.1). The sort charts for both the SephadexTM G-50 extracted oocysts and those that had undergone further purification with octyl-SepharoseTM were similar (Figures 3.2 and 3.3), with a distinct group of particles in the sort region being distinguished from the other particles in the oocysts suspensions. However, the suspension that had undergone the octyl-SepharoseTM treatment had removed many of the miscellaneous particles that had moved through the SephadexTM G-50 column at a similar rate to the oocysts, and therefore more of the particles sorted were oocysts. The most highly purified oocyst suspension that had undergone diethyl-ether defatting and sucrose flotation had the least number of miscellaneous particles (Figure 3.4), and the sort region had an obvious concentration of particles of which 88% were shown to be oocysts.

54

Figure 3.1 Flow cytometry and sort gate for diluted calf faecal slurry. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts

55

Figure 3.2 Flow cytometry and sort gate for oocyst suspension prepared using SephadexTM G-50 column. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts

56

Figure 3.3 Flow cytometry and sort gate for oocyst suspension prepared using SephadexTM G-50 column followed by removal of lipids using octyl-SepharoseTM. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts

57

Figure 3.4 Flow cytometry and sort gate for oocysts extracted using diethylether and sucrose. The horizontal axis represents the side scatter (SSC), the vertical axis the forward scatter (FSC) and R1 indicates the sort area selected for oocysts

58 16 Ether Sucrose x1 14 Ether Sucrose x2

12

10

8

6

4 Total solids by volume (%) 2

0 12 14 17 21 25 31 38 46 56 68 83 0.5 1.3 1.6 2.0 2.4 2.9 3.5 4.3 5.2 6.4 7.8 9.5 101 123 150 183 223 272 332 404 492

Particle size (μm)

Figure 3.5 Particle size distributions of oocyst suspensions prepared using one and two rounds of diethylether defatting and sucrose flotation

Table 3.1 Number of oocysts in 100 sorted particles from various oocyst preparations

Preparation Number of oocysts Dilute calf faeces 0 SephadexTM G-50 only 1 SephadexTM G-50 and octyl-SepharoseTM 77 Diethyl-ether sucrose 88

In the more purified oocyst suspensions the flow cytometer was able to sort a high percentage of oocysts from the remaining faecal debris, so providing a useful additional purification step for the oocysts. However, the dilution effect during sorting was such that only around 17 000 counts were being collected per 50 mL of sheath fluid, making the recovered oocyst concentration approximately 15 000 per 50 mL tube. This was too low for use in studies that required high concentrations of pure oocysts such as some of

59 the surface property and aggregation methods discussed in Chapters 4 and 5 respectively, even if centrifugation was used to concentrate oocysts after flow cytometry.

3.4 Discussion

Preparations of Cryptosporidium suspensions containing a high concentration of oocysts were required for many of the tasks in this thesis. Assays used to study how particles act as a whole and where individual populations of particles could not be distinguished (e.g. aggregation studies using particle sizing for detection of aggregates, Chapter five), required the prepared oocyst suspensions to be free of contaminating particles. Losses of oocysts from the suspensions could occur at every stage of purification process. Therefore, if possible it was preferable to keep the number of processing steps to a minimum in order to have the oocysts concentrations as high as possible.

Thiriat et al. (1998) extracted Giardia cysts from faecal material using a series of filtration steps through filters of progressively smaller pore size followed by a series of centrifugation steps and passage of the cysts through a SephadexTM G-50 column. Their method was simplified in the current work so that the diluted calf faeces were passed directly through the column. However, this resulted in only small amounts of the faeces being able to be treated since larger volumes clogged the column.

It was not possible to obtain a large number (>106) of oocysts free from other particles without using defatting or treatments with high ionic strength solutions. Cryptosporidium oocysts moved rapidly through the SephadexTM G-50 column and were separated from the majority of the lipids and other debris in the faecal slurry, yet there were residual lipids of a similar size and density to oocysts. Treatment with octyl- SepharoseTM beads successfully removed some of the remaining lipids. The concentration of oocysts in the final suspensions, however, was lower than that which could be obtained using the diethyl-ether and sucrose. Oocysts extracted using the SephadexTM G-50 method could be used in assays that did not require high oocyst concentrations or highly pure oocyst suspensions.

60 Using the methods of particle sizing and flow cytometry, the purity of the diethylether/sucrose treated oocysts was confirmed. Flow cytometry also confirmed that while the SephadexTM G-50 method was able to separate oocysts from some of the faecal particles, there were still a substantial number of particles around the size of Cryptosporidium oocysts remaining in the suspension. This was true even after the octyl-SepharoseTM treatment, although the resulting oocyst suspension did display a greater concentration of oocysts compared to other similar sized particles.

Cell sorting using flow cytometry is usually performed after staining the target organisms with a fluorochrome such as fluorescein isothiocyanate. This is to improve the selection and sorting of the target organisms (Deere et al., 2002). However, in this case the binding of the stain via surface antigens may change the oocysts surface properties (Chung et al., 2004), so an attempt to sort the oocysts without staining was therefore made. Flow cytometry was shown to be unable to separate unstained Cryptosporidium oocysts from some other similarly-sized particles in the suspension. However, a suspension of purified oocysts at a high concentration could potentially be further purified using flow cytometry and cell sorting, if a more purified suspension of oocysts at a lower concentration was acceptable.

3.5 Conclusions

Both particle sizing and flow cytometry were shown to be able to provide information on the purity of Cryptosporidium oocyst suspensions. The use of a defatting agent was required to obtain an oocyst suspension with high purity and a high concentration. After the first round of diethyl-ether and sucrose treatment some residual lipids remained, though these could be largely removed by employing a second round of treatment.

Oocysts were unable to be distinguished sufficiently from the surrounding debris in unpurified suspensions by flow cytometry. The dilution of the more purified oocysts with sheath fluid also made further clarification with the use this method unsatisfactory if a suspension of oocysts at a high concentration was required. The method using a SephadexTM G-50 column followed by octyl-SepharoseTM treatment was able to produce

61 the least chemically-challenged oocyst suspension of satisfactory purification and concentration for methods that do not require a concentration of more than 104 oocysts per millilitre and can tolerate some non-oocyst particles in the suspension.

62 CHAPTER FOUR

CRYPTOSPORIDIUM OOCYST SURFACE PROPERTIES

4.1 Introduction

The surface properties of microbial cells largely dictate their interactions with environmental surfaces as reviewed in Chapter 2. The important intrinsic properties of pathogens include surface charge, size and hydrophobicity which, in turn, are influenced by environmental factors such as the ionic strength, pH and polyvalent cation concentration of the surrounding medium (Gerba, 1984).

Many different methods have been used for testing the surface properties of microbes. The methods used in this study were hydrophobic interaction chromatography (HIC) columns and microbial adhesion to hydrocarbons (MATH). HIC stationary phases are manufactured by attaching hydrophobic functional groups to an agarose backbone, which in this case is SepharoseTM (Queiroz et al., 2001). The stationary phases of the HIC columns were also used in suspension in order to test for retention of oocysts in the columns due to straining. In addition to hydrophobic ligands, a weak ion exchange ligand was used for determining charge characteristics of the oocysts (Pedersen, 1980). These methods were chosen because they were uncomplicated and inexpensive to perform. The MATH assay, in particular, has been widely used to measure microbial cell hydrophobicity (Geertsema-Doornbusch et al., 1993).

The aim of this chapter was to study the hydrophobicity and electrostatic properties of Cryptosporidium oocysts, including those prepared by various methods described in Chapter three. Knowledge gained on these surface properties should help in an understanding of how Cryptosporidium interact with environmental particles and surfaces and thus how they may be transported in surface runoff.

63 4.2 Materials and methods

4.2.1 Cryptosporidium oocysts

Most of the Cryptosporidium oocysts used in this study were extracted in-house from the faeces of 7-21 day old calves using a range of techniques. All calves were located on the one property, however, isolates from each individual calf were identified separately (Calf 1, Calf 2 etc.). Oocysts extracted using diethyl-ether defatting followed by either salt flotation or sucrose flotations, as described in Chapter three, were used for all surface property techniques described in this chapter. Oocysts extracted using SephadexTM G-50, as described in Chapter 3, were used with the SepharoseTM bead in suspension method to test whether oocysts extracted from faeces using two different methods produced different results. In addition, the Iowa isolate of C. parvum, provided by BTF Decisive Microbiology (Sydney), was used with the in-house HIC columns (as described below) without further treatments.

4.2.2 Hydrophobic interaction chromatography (HIC) columns

4.2.2.1 Pre-cast HIC columns Pre-cast HIC columns (Figure 4.1a) containing SepharoseTM substituted with phenyl-, butyl- or octyl- side chains (HiTrap, Amersham Biosciences Pty. Ltd., Baulkham Hills, NSW, Australia) were used in this study. According to the manufacturer, the hydrophobic ligands were coupled to the SepharoseTM in such a way that the resulting bonds were both stable and uncharged, and therefore should show virtually no non- specific adsorption.

The columns were equilibrated with 20 bed volumes of 4 M NaCl. Approximately 105 Cryptosporidium oocysts extracted from calf faeces by diethyl ether defatting followed by salt flotation were added to 1 mL of 4 M NaCl and forced through each equilibrated column at approximately 2 mL.min-1 (Figure 4.1a) using a syringe. The columns were rinsed with 9 mL of 4 M NaCl and the eluate collected. Using a syringe, 10-mL volumes of decreasing ionic strength solutions were forced through the columns at an approximate rate of 2 mL.min-1. The ionic strengths of NaCl solutions used for elution

64 of the oocysts were 2 M, 1 M, 0.5 M, and 0.1 M, and all eluates were collected separately. The final rinse steps were performed using reagent water. In experiments using the phenyl- and octyl-SepharoseTM columns, only one 10-mL volume of reagent water was forced through the gel, while the butyl-SepharoseTM column was rinsed with three consecutive 10-mL volumes of reagent water to further elucidate the results observed.

Oocysts in each sub-sample were enumerated by fluorescence microscopy (described in Chapter three) following filtration onto 13 mm diameter, 0.8 m pore size polycarbonate filters (Millipore, Sydney, Australia) and stained with EasyStainTM (BTF Decisive Microbiology, Sydney, Australia) according to the manufacturer’s instructions. The spike concentration was determined by dilution, filtration, staining and enumeration of oocysts as described above. To calculate the number of oocysts that remained in the column after each wash step, the cumulative number of oocysts that had washed through the column after each step was subtracted from the concentration of oocysts in the spike solution.

4.2.2.2 In-house HIC columns Octyl-SepharoseTM and SepharoseTM columns were also made in-house. Instead of using the salting-out technique to cause the oocysts to attach to the hydrophobic side chains, a single elution buffer was used. Jones et al. (1996) recommended that buffer systems remain constant to achieve consistency of results between methods for hydrophobicity and charge determinations. Brush et al. (1998) reported that 0.01 M

KNO3 had a conductivity sufficiently low such that it does not interfere with the + - measurement of electrophoretic mobilities, and that the K and NO3 ions are less likely + to react with inorganic compounds on the oocyst surface than other ions such as Na and - Cl . Therefore, 0.01 M KNO3 was chosen as the buffer of choice for use with the in- house HIC columns and also for subsequent hydrophobicity assays used. In addition, reagent water was also used as the eluent in some experiments.

Each column was packed to a bead height of 2.5 cm into Pasteur pipettes that were plugged at the base with glass wool (Figure 4.1b). SepharoseTM columns were used as

65 negative controls since this matrix had no side chain (hydrophobic) substitutions. The columns were washed with 20 bed volumes of the eluent. Approximately 104 oocysts were added to the top of each column in a 100 μL volume of eluent and allowed to pass into the column by gravity. When the meniscus of the spike suspension was level with the top of the matrix in the column, an additional 2 mL volume of elution buffer was carefully added to the top and also allowed to filter through the column under gravitational force.

The number of oocysts that passed through the column was enumerated by filtration of a subsample through a 13 mm diameter, 0.8 m pore size filter, stained, viewed and counted by epifluorescence microscopy as described previously. Results were converted to a percentage of oocysts passing through the octyl-SepharoseTM columns compared to the number passing through the SepharoseTM (control) columns.

Two of the Cryptosporidium isolates used were in-house extractions from two different calves, and a third was the Iowa isolate. The elutent used for all isolates was 0.01 M

KNO3, and reagent water was used as an additional eluent for the Calf 1 isolate only. This gave a comparson for a single oocyst isolate between reagent water, a low ionic strength eluent that would not encourage hydrophobic interactions as according to the electrical double layer interaction theory, and the 0.01 M KNO3 eluent recommended by Brush et al. (1998).

66

Figure 4.1 Hydrophobic interaction chromatography columns: a) Sample being forced through pre-cast HIC octyl-SepharoseTM column; b) samples filtering by gravity through octyl-SepharoseTM columns made in-house with Pasteur pipettes

4.2.3 Interactions with SepharoseTM beads in suspension

SepharoseTM beads, including those with sidechains (octyl-, butyl-, diethylaminoethyl- [DEAE]) (Amersham Biosciences Pty. Ltd., Baulkham Hills, NSW, Australia), and SepharoseTM without sidechains were used in suspension (Figure 4.2) as described previously by Capizzi-Banas et al. (2002), though with modifications to their procedure for determining adhesion. The beads were washed 10 times in either 0.01 M KNO3 or 1

M NaCl in 0.01 M KNO3. After the final wash the beads were resuspended to approximately 50% by volume in the suspension solution. One millilitre volumes of the mixed bead suspension were added to tubes containing approximately 104 oocysts in the same suspension solution. The oocysts used were isolated from faeces using two different extraction methods. One method used diethylether and sucrose while the other used SephadexTM G-50 (as described in Chapter three). Control tubes not containing beads were adjusted to the same volume by the addition of 1 mL of the suspension solution to the oocysts. All tubes were mixed gently for five minutes to allow adsorption of the oocysts to the beads. After allowing the beads to settle undisturbed for ten minutes, 100-L volumes were removed from above the settled beads and filtered through 13 mm diameter, 0.8-m pore size filters, stained, viewed by epifluorescence microscopy and counted as described previously. Results were converted to a

67 percentage of oocysts remaining in suspension compared to the control tube that did not contain beads.

Figure 4.2 SepharoseTM beads after settling: a) SepharoseTM without side-chains, b) DEAE-SepharoseTM and c) octyl-SepharoseTM

4.2.4 Microbial adhesion to hydrocarbons (MATH)

The MATH assay, as described by Rosenberg et al. (1980), was used with suspensions of Cryptosporidium oocysts in four solutions of differing ionic strength. A 1 mL volume of octane was added to a 2 mL suspension of oocysts suspended in either 4 M

NaCl, 0.01 M KNO3, 0.01 M PBS or reagent water, and vortexed vigorously for two minutes. The vigorous mixing created micro-droplets of hydrocarbon which then served as a hydrophobic sub-stratum for microbial adhesion (Figure 4.3). The octane was then allowed to partition from the aqueous layer by leaving the tests undisturbed at room temperature for 15 minutes. The aqueous layer was sampled by passing a Pasteur pipette through the upper octane layer into the aqueous layer and ejecting a small bubble to ensure no octane was in the pipette before withdrawing the sample. Oocysts were filtered onto a 13-mm diameter membrane with pore size of 0.8 m, stained, viewed using epifluorescence microscopy and counted as described previously. Counts obtained in the aqueous layer where octane was added were compared to a control tube in which no octane was added (Figure 4.3).

68

Figure 4.3 MATH assay flow diagram

4.2.5 Statistical analysis

Concentrations of Cryptosporidium oocysts were analysed using analysis of variance (ANOVA) to test for significant differences between means at the = 0.05 level. These analyses were undertaken using the statistical package in Excel (Microsoft® Office Excel 2003, Microsoft, North Ryde, Australia). When means were shown to be significantly different, ranked means were analysed using Student-Newman-Keuls (SNK) pairwise comparisons to determine which pairs of means were significantly different (Zar, 1999). Coefficients of variation were calculated by dividing the standard deviation by the mean, and were converted to a percentage by multiplying by 100.

4.3 Results

4.3.1 Hydrophobic interaction chromatography (HIC) columns

4.3.1.1 Pre-cast HIC columns Cryptosporidium oocysts used were less than four weeks old and were extracted using diethyl-ether defatting and salt flotation. Oocysts that were added to the top of the pre- cast HIC columns (Figure 4.1a) were eluted from the columns with the 4 M NaCl, and also with each volume of eluent that was passed through the column. The cumulative percentage of oocysts removed by the eluent solutions was plotted for each of the phenyl-, butyl- and octyl-SepharoseTM columns (Figure 4.4). Of the spiked oocysts, 61.4% passed directly through the octyl-SepharoseTM HIC column and were eluted in

69 the 4 M NaCl eluent, 30.3% passed directly through the butyl-SepharoseTM HIC column and 27.3% passed directly through the phenyl-SepharoseTM HIC column in the 4 M NaCl eluent.

The results were also calculated as the percentage of Cryptosporidium oocysts being eluted from those remaining in the column after previous elutions. Figure 4.5 shows that the percentage of oocysts being eluted from the columns was highest after the initial spike in 4 M NaCl. Generally the percentage of oocysts released into the eluate decreased as the ionic strengths decreased, or were relatively stable down to 0.1 M NaCl. Passing reagent water through the columns appeared to release a higher percentage of oocysts than the 0.1 M NaCl for both the octyl- and phenyl-SepharoseTM columns. In contrast, the butyl-SepharoseTM column continued to release a similar number of oocysts in the reagent water washes, and the pattern did not change even when two additional reagent water washes were undertaken.

Phenyl 4 M 2 M 1 M 0.5 M Butyl 0.1 M MQ1 ligands MQ2 MQ3 Octyl HIC column hydrophobic HIC column hydrophobic

0 20406080100 Spiked Cryptosporidium oocysts eluted (%)

Figure 4.4 Percentage of spiked oocysts removed from pre-cast HIC columns using eluents of decreasing ionic strengths of NaCl solutions (4 M, 2 M, 1 M, 0.5 M and 0.1 M) followed by up to 3 reagent water washes (MQ1-MQ3)

70 80

60

Octyl 40 Butyl (%) Phenyl

20

0 Removal of remaining oocysts 4 M 2 M 1 M 0.5 M 0.1 M MQ1 MQ2 MQ3 Ionic strength

Figure 4.5 Percentage removal of oocysts remaining in the HIC column prior to each elute passed through the columns

4.3.1.2 In-House HIC columns The Cryptosporidium oocysts used in this assay were less than one month old and were purified from calf faeces using diethyl-ether defatting and sucrose flotation. There was no significant (P > 0.05) difference in the percentage of spiked oocysts passing directly through the columns when using octyl-SepharoseTM beads compared to the control test which contained SepharoseTM without any hydrophobic sidechains (Table 4.1). The coefficient of variation for the Iowa isolate was high at 34%, showing the highest variability of all the isolates, even though there was less than 5% retention of oocysts in the columns compared to the SepharoseTM control columns. In addition, there was no difference in the mean percentage of oocysts passing through the column using the Calf

1 isolate between reagent water and 0.01 M KNO3.

71 Table 4.1 Mean percentage of three Cryptosporidium oocyst isolates passing directly through the octyl-SepharoseTM HIC columns compared to control SepharoseTM HIC columns Eluent Calf 1+ Calf 2+ Iowa Mean SD N Mean SD N Mean SD N (%) (%) (%) (%) (%) (%) MQ 100 13 5 nd* nd nd nd nd nd

KNO3 100 15 2 97 18 4 95 32 2 * nd = not done + Oocyst isolates Calf 1 and Calf 2 were from different calves located at the same farm

4.3.2 Interactions with SepharoseTM beads in suspension

The Cryptosporidium oocysts used in this assay were less than one month old and were extracted using diethyl-ether defatting and sucrose flotation. The percentages of oocysts bound to the beads were determined by calculating the number of oocysts remaining in suspension after bead settling compared to the control to which no beads were added. On average between 87 and 115% of oocysts were recovered from the tubes containing the DEAE-, octyl-SepharoseTM and SepharoseTM beads (Figure 4.6) when compared to the spike dose. Using one-way ANOVA ( = 0.05), there was no significant difference observed between the percentage recoveries obtained for any of the SepharoseTM beads.

72 140.0

120.0

100.0 Sepharose G-50 oocysts 80.0 Ether/Sucrose oocysts 60.0

40.0 Unbound oocysts (%) 20.0

0.0 DEAE beads Octyl beads Sepharose beads

Figure 4.6 Comparison of SephadexTM G-50 and diethyl-ether/sucrose extracted oocysts in suspension with DEAE- and octyl-SepharoseTM and SepharoseTM beads in 0.01 M KNO3, pH 5.8, error bars are +1 SD for triplicate oocyst suspensions

In an attempt to make the oocysts bind more strongly to the beads, the ionic strength of the solution was increased with the addition of NaCl to a final concentration of 1 M in the 0.1 M KNO3. The percentage recovery of oocysts compared to a control tube not containing Separose beads ranged between 81 and 115% (Figure 4.7). One-way ANOVA analysis ( = 0.05) indicated that there was a significant difference between the groups (p = 0.005). SNK rankings showed differences between the ether/sucrose extracted oocysts in SepharoseTM and both oocyst suspensions in DEAE-SepharoseTM, and between SephadexTM G-50-extracted oocysts in SepharoseTM and ether/sucrose- extracted oocysts in DEAE-SepharoseTM.

73 140.0

120.0

100.0

Sepharose G-50 80.0 oocysts

Ether/Sucrose 60.0 oocysts

40.0 Unbound oocysts (%)

20.0

0.0 DEAE beads Octyl beads Sepharose beads

Figure 4.7: Comparison of SephadexTM G-50 and diethylether/sucrose extracted oocysts in suspension with DEAE and octyl-SepharoseTM and SepharoseTM beads in 0.01 M KNO3 and 1 M NaCl, pH 5.8, error bars + 1 SD for triplicate oocyst suspensions

4.3.3 Microbial adhesion to hydrocarbons

Five isolates of Cryptosporidium oocysts (Calf isolates 1, 3, 4, 5 and the Iowa isolate) were tested using the MATH procedure, with results shown in Figure 4.8. The calf isolates numbered 1, 3, and 4 were all extracted using diethyl-ether and salt flotation, calf isolate number 5 was extracted using diethyl-ether and sucrose flotation, and the Iowa isolate was used as supplied. The Calf 1 isolate was twelve months old, the Calf 3 isolate approximately one month old, the Calf 4 isolate approximately two months old, the Calf 5 isolate was approximately four months old, and the Iowa isolate was used fifteen months after extraction.

All Cryptosporidium oocyst isolates tested in the presence of 4 M NaCl showed a high level of removal in the aqueous layer, with a lesser amount of removal from the aqueous layer when lower ionic strength solutions were used. In the presence of 0.01 M KNO3 the results showed some variability between isolates.

74

The Calf 4 and Calf 5 isolates displayed moderate hydrophobic properties of 47% and

53% removal respectively from the aqueous KNO3 layer. The Calf 1 isolate displayed high hydrophobic properties at 71% removal from the aqueous layer containing 0.01 M

KNO3.

With the exception of the Calf 3 isolate, the oocyst removal from the aqueous layer was highest when suspended in 4 M NaCl and least when suspended in reagent water. The assay with 4 M NaCl was not undertaken with calf isolate 4, but with all other isolates tested the removal percentage from the aqueous layer was 89% or higher. In all instances the percentage removal of oocysts from the aqueous layer when suspended in

0.01 M PBS and 0.01 M KNO3 were similar, although generally the PBS showed a slightly higher removal from the aqueous layer than the KNO3.

The Calf 3 oocyst removal rates from the aqueous layer for all four suspension solutions remained unchanged at between 99 and 100% when the isolate was extracted using diethyl-ether and sucrose flotation (results not shown). In addition the isolate was re- tested when the oocysts were eight weeks old, and again removal from all four suspension solutions was between 99 and 100% (results not shown).

75 120

100

80 4 M NaCl 0.01 M PBS

oocyst removal 60 0.01 M KNO3

(%) MilliQ water 40

20

0 Cryptosporidium Calf 1 Calf 3 Calf 4 Calf 5 Iowa Cryptosporidium isolate

Figure 4.8 Removal efficiency of five different isolates of Cryptosporidium oocysts from the aqueous layer of the MATH assay. Calf isolates 1, 3 and 4 were extracted using diethylether treatment followed by salt flotation, isolate Calf 5 was extracted using diethylether treatment and sucrose flotation, and the Iowa was a commercial suspension

The Calf 3 isolate showed considerably more hydrophobicity with the MATH assay than any of the other isolates (Figure 4.8). Practically all of the oocysts were removed from the aqueous layer regardless of the aqueous solution they were suspended in. Removal rates were 100% when suspended in 4 M NaCl, and 99.9% using any of 0.01

M PBS, 0.01 M KNO3, or reagent water. Because the assay measures the removal of oocysts from the aqueous solution and not the number of oocysts in the octane layer, an attempt was made to check that the removed oocysts were partitioning into the octane layer and not simply lysed by the octane or the methodology in general. All efforts to filter an aliquot of the octane through a 13-mm-diameter membrane and subsequently stain it with a fluorescently labeled anti-Cryptosporidium antibody were unsuccessful (results not shown). There were no objects recognisable as oocysts on the membrane after staining. It is likely, however, that the oocysts could not be stained since the octane would be expected to adhere to the oocyst surface and block the antibody-linked stain from binding to its surface antigen. Some of the octane from a 4 M NaCl test was also added to reagent water and the MATH test assay was repeated, with the hypothesis

76 that in the lower ionic strength some of the oocysts bound to the octane layer may dissociate into the aqueous layer. However no oocyst was observed in the aqueous layer after the assay (results not shown).

A three-way comparison of methods comparing the results of the MATH assay with octane to octyl-SepharoseTM beads used in both the in-house columns and in suspension, with results given in Table 4.2. The eluent and suspension buffer in each case was 0.01 M KNO3 so that results could be directly compared. The oocyst preparations used for this were isolated from two different calves. Both preparations were extracted from faeces with diethyl ether defatting but with different flotation methods. The Calf 1 isolate was approximately twelve months old while the Calf 6 isolate was less than two months old. Although 100% of oocysts remained unbound to the octyl-SepharoseTM beads in suspension and a high proportion of oocysts passed directly through the octyl- SepharoseTM immobilised in a column, more than 50% of the oocysts were removed from the aqueous layer in the MATH assay.

Table 4.2 Percentage of oocysts remaining unbound to the hydrophobic component of the assay, comparison of results from three hydrophobicity assays and two Cryptosporidium isolates

Assay Calf 1+ Calf 6+ (ether/sucrose)* (ether/salt)* Octyl-SepharoseTM (suspension) 100 100 Octyl-SepharoseTM (in-house column) 99 87 MATH (Octane) 34 49 + Oocyst isolates Calf 1 and Calf 6 were from different calves located at the same farm * n=3

77 4.4 Discussion

4.4.1 Hydrophobic interaction chromatography (HIC) columns

Shields and Farrah (2002) reported that 100% of the viruses in their study adhered to the octyl-SepharoseTM in the presence of 4 M NaCl and reported the relative hydrophobicity of their viruses based on how easily they were removed from the column. Since 100% of the Cryptosporidium oocysts did not adhere to the columns, and those that did remain in the column appeared to be removed at a slow but constant rate, it seemed possible that there was another factor affecting the movement of Cryptosporidium through the columns. It was thought that perhaps the oocysts were being physically strained by the SepharoseTM matrix, even though at 90 μm in diameter the beads were unlikely to significantly hinder the passage of the oocysts which are around 4 μm in diameter. There had also been some concern that 4 M NaCl would cause the oocysts to clump together and that could have enabled some of the straining effect, but if this were the case it would be likely that the oocysts would have also been more efficiently retained by the column matrices.

Due to the limited number of pre-cast columns available for the study, the remainder of the HIC column study was undertaken with columns prepared in-house with octyl- SepharoseTM beads. The octyl-SepharoseTM was chosen because the octyl ligand has a straight alkyl chain and is reported to show a more “pure” hydrophobic character than phenyl, which is reported to also be influenced by charge characteristics (Busscher et al., 1995). While the pre-cast butyl-SepharoseTM showed potentially more oocysts adhering to the matrix with the 4 M NaCl, only low numbers of oocysts were removed with subsequent low ionic strength washes and even the addition of reagent water had no effect on the constant, but low, elution rate. Using reagent water as the eluent appeared to have a slight impact on the elution of oocysts from the octyl-SepharoseTM matrix since it increased elution from 10% of remaining oocysts being eluted with the 0.1 M NaCl solution to 22% being eluted with the reagent water. This apparent effect of the reagent water increasing the elution rate was even more obvious with the phenyl- SepharoseTM which had 9% of the remaining oocysts being eluted in the 0.1 M NaCl and 30% of remaining oocysts being eluted with the reagent water. In addition, octane, in contrast to butane, is a liquid at atmospheric pressure and temperature and could be

78 used in the MATH assay and readily compared to results obtained with the octyl- SepharoseTM matrix.

The HIC column methodology used for the columns made in-house was varied from the method used on the pre-cast columns because the salting-out technique did not result in the majority of oocysts adhering to the columns, and therefore the rate of elution using lower ionic strength solutions was meaningless. Determination of hydrophobicity using the in-house columns was to be by measuring the percentage of oocysts that adhered to the columns when using a single ionic strength eluent.

Following the recommendations of Jones et al. (1996) and Brush et al. (1998), subsequent surface property assays were undertaken with 0.01 M KNO3 and also a combination of reagent water, 4 M NaCl or 1 M NaCl since these solutions were all used in the pre-cast HIC columns. Additionally, PBS was trialed as a buffer system that was routinely used in the laboratory.

The results of the octyl-SepharoseTM columns made in-house showed that there was no difference (P > 0.05) between results obtained from a single Cryptosporidium isolate

(Calf 1) when either reagent water or 0.01 M KNO3 was used as the eluent solution, with 100% passing through the column in both cases. Similarly, there was no apparent attachment of the other two Cryptosporidium isolates (Calf 2 and Iowa) to the octyl- TM Sepharose using 0.01 M KNO3 as the eluent.

4.4.2 Interactions with SepharoseTM beads in suspension

Capizzi-Banas et al. (2002) used the DEAE-SepharoseTM and octyl-SepharoseTM beads in suspension for testing the hydrophobicity of Ascaris eggs, because the size of the eggs (55 x 40 m) would prevent them from passing through a column containing microspheres of an average size of 90 m. The same principle was used for this study since the pre-cast HIC results indicated that physical straining was a possibility which could be the dominating influence on the passage of oocysts through the columns.

79 In the suspension tests, oocysts from the same faecal sample but purified by two different extraction techniques were tested. It was hypothesised (discussed in Chapters one and three) that the diethylether/sucrose purification method could affect oocyst surfaces due to the strong chemicals used in the process. Diethylether is a ‘fat solvent’ and could affect any surface lipids on the oocysts. Furthermore, the sucrose and salt flotation involves the use of a hyper-osmotic and dehydrating solution that could also have an effect on oocysts surfaces (Ryley et al., 1976). In contrast, the separation of oocysts from faeces using a SephadexTM G-50 column and a series of filtrations was unlikely to affect any oocysts surface properties.

There was no significant difference (P > 0.05) between the results of oocysts from the two different extraction methods with the same type of beads, and there was no significant difference between the results of the octyl-SepharoseTM and the SepharoseTM suspensions. However, there was a significant difference between the DEAE- SepharoseTM and SepharoseTM suspensions when suspended in a 1 M ionic strength solution. This implies that the hydrophobicity of both oocyst preparations was low, and that the charge of the oocysts may have an effect on attachment when the ionic strength of the solution is relatively high. However, according to the double layer interaction, hydrophobicity, and not the surface charge, that should be enhanced by increasing the ionic strength (discussed in Chapter two).

4.4.3 Microbial adhesion to hydrocarbons (MATH)

The MATH assay has been a commonly used method to measure the hydrophobicity of microbes (Rosenberg et al., 1980; Lichtenberg et al., 1985; Pembrey et al., 1999). In this test, a microbial cell suspension is mixed with a hydrocarbon for a period of time to allow optimal interaction of the microbes with the hydrocarbon phase. After separation of the organic and aqueous phases and depending on their hydrophobicity, cells may remain in the liquid phase or partition either into the liquid-hydrocarbon interface or into the hydrocarbon phase (Rosenberg et al., 1980). The MATH assay, using octane as the hydrocarbon in the system, was also used in this study to measure the hydrophobicity of Cryptosporidium oocysts.

80 Aliphatic hydrocarbons are used in most MATH assays because aromatic hydrocarbons cause lysis of some bacterial species (Pembrey et al., 1999). Therefore, aliphatic octane was chosen over aromatic phenol for the MATH assay to further explore the hydrophobic interactions of oocysts.

The higher the ionic strength of the solution used to suspend the oocysts in the MATH assay, the more oocysts that apparently partitioned into the octane. These results were consistent with the salting-out phenomenon. Martin et al. (1989) and Schneider and Riley (1991) described a cell surface classification scheme that defines isolates with a hydrophobicity greater than 70% as highly hydrophobic, and those with a hydrophobicity less than 30% were classified as highly hydrophilic. Using this classification scheme, the isolate called Calf 3 showed strong hydrophobic properties in all solutions tested including reagent water, although these results could not be confirmed using a mass balance approach. Since aliphatic hydrocarbons are not renowned for lysing bacterial cells, and other MATH tests showed differing results with as low as 5% removal of the oocysts in the presence of octane, it could only be assumed that the removal of oocysts from the aqueous layer meant that they were relocated to the hydrophobic hydrocarbon layer. Neither pre-staining the oocysts with the fluorescently labeled anti-Cryptosporidium antibody or a FISH probing were considered since both methods would impact the oocyst surface, one by coating the oocyst surface with the antibody stain, and the other by permeabilising the oocysts by heating. The use of heat to permeabilise the oocysts has the effect of disrupting the surface proteins which increases the hydrophobicity of the oocyst (Kuznar and Elimelech, 2005). However, it may be useful to use the FISH probe to confirm relocation of oocysts (intact or otherwise) to the octane layer if they already show such marked hydrophobicity.

The age of the oocysts did not appear to have any systematic impact on the MATH assay. The Calf 3 isolate was highly hydrophobic when it was three days old and also two months old, and changing the flotation solution from saturated salt to sucrose also did not affect the results. The oldest isolate (15 month old Iowa strain) showed hydrophobicity results for 4 M NaCl and reagent water that were similar to those obtained with the Calf 1 isolate which was approximately twelve months old, with both isolates less hydrophobic than the three day old Calf 3 isolate. The four month old Calf

81 5 isolate, and the two month old Calf 4 isolate both showed slightly lower hydrophobicity measurements, with the hydrophobicity in reagent water being very low, and would be classed as hydrophilic using the scale of Martin et al. (1989) and Schneider and Riley (1991), contrasting with the highly hydrophobic Calf 3 isolate at three days and two months old.

Drozd and Schwartzbrod (1996) reported low to moderate hydrophobicity of Cryptosporidium oocysts with the MATH assay. They used oocysts extracted from calf faeces using a protocol of ethyl acetate and formaldehyde defatting followed by flotation over NaCl, then sucrose gradient centrifugation followed by resuspension in deionised water and storage in potassium dichromate for an unspecified amount of time. According to the hydrophobicity classification scheme, the oocysts of Drozd and Schwartzbrod (1996) were of low hydrophobicity, reaching a maximum of 40% removal when using 0.05 M MgCl2. Their results were similar to those obtained in the work reported in this Chapter, with low to medium hydrophobicity when oocysts were suspended in low ionic strength solutions. These results contrast with those of Hsu and Huang (2002) who also used the MATH assay with octane to determine the hydrophobicity of Cryptosporidium oocysts. Hsu and Huang (2002) used a 0.01 M -1 NaClO4 buffer system and they calculated the removal in terms of rates (minute ), making comparison between our studies difficult. They did claim that their oocysts demonstrated marked hydrophobicity, which also correlates with the high hydrophobicity results reported in the current study with the Calf 3 isolate. Although the results of these two groups of researchers are difficult to compare due to different suspension solutions and methods being used, both groups report that removal rates from the aqueous layer increased with increasing ionic strength, which was also the case for the current study.

The Calf 3 isolate was extracted using two slightly different methods. Both extraction methods used diethyl-ether defatting, but one was followed by salt flotation while the other sucrose flotation. Neither of these latter treatments appeared to have any effect on the high removal of oocysts from the aqueous layer. Drozd and Schwartzbrod (1996) attempted to perform the MATH assay on oocysts extracted using different methods, however they were unable to purify the oocysts enough for use in the assay without the

82 defatting step. They were able to show, however, that there was no difference in the MATH results between oocysts that had undergone both defatting and sucrose flotation compared to those that had undergone defatting only (Drozd and Schwartzbrod, 1996). Both of these results contradict the results of Brush et al. (1998) who showed that the extraction method employed can affect the hydrophobicity of oocysts.

Neither Drozd and Schwartzbrod (1996) nor Hsu and Huang (2002) reported MATH assay results as variable (ranging from high to low hydrophobicity) as those presented in the current study. Busscher et al. (1995) has reported that although the MATH assay may be the most commonly-employed method to characterise the cell surfaces of microbes, various hydrocarbons used for the assay, including octane, are highly negatively charged in some solutions in which the MATH is carried out. They therefore suggested that the type of hydrocarbons used in the HIC assay may also lead to complicating effects through electrostatic interactions. Variations in the hydrophobicity between isolates in the current study could be due to factors as yet unknown.

4.4.4 Correlation of results from different methods

Cryptosporidium oocysts showed little tendency to attach to the octyl-SepharoseTM beads. In 0.01 M KNO3 there was no attachment of oocysts detected compared to the spike when the beads were in suspension, and there was no adherence of oocysts to the beads when using the in-house HIC columns.

The MATH assay indicated that in the 4 M NaCl solution the oocysts showed marked hydrophobic characteristics. This implies that in the same high ionic strength solution the oocysts should have adhered efficiently to the hydrophobic ligands of the HIC columns. Yet a high proportion of the oocysts passed directly through the columns while suspended in 4 M NaCl. In addition, when 1 M NaCl was added to the 0.01 M

KNO3 electrolyte when using the hydrophobic beads in suspension, there was no difference in attachment to the octyl-SepharoseTM beads when compared to the SepharoseTM beads without the hydrophobic ligands.

83 The MATH assay has been criticised as being influenced by factors other than hydrophobicity (Geertsema-Doornbusch et al., 1993; Busscher et al., 1995; van der Mei et al., 1995). They concede that the method may be useful if performed at an ionic strength, or pH, in which the microbes and the octane are uncharged. Due to our inability to measure the surface charge of the oocysts, and the need to keep suspension buffers the same between methods, adjustment of the test conditions to account for any non-hydrophobic effects was not possible.

4.5 Conclusions

The degree of oocyst hydrophobicity measured appears to depend on the method used, yet the oocyst purification procedure had no impact on the hydrophobicity or electrostatic adhesion to SepharoseTM beads containing hydrophobic or charged sidechains. The oocyst isolate appeared to influence the MATH assay rather than the age of the oocysts. Overall, the results suggested that attachment of oocysts to solid surfaces may be influenced more by electrostatic attraction than hydrophobicity.

84 CHAPTER FIVE

CRYPTOSPORIDIUM OOCYST AGGREGATION

5.1 Introduction

Evidence exists indicating that Cryptosporidium oocysts are able to clump together and form aggregates of variable sizes. Bukhari and Smith (1995) observed oocysts clumping together when they viewed calf faeces in a wet preparation. On further investigation they observed that the majority of oocysts in the aggregates were not viable. Attempts to deliberately cause oocysts to aggregate, however, have been unsuccessful (Butkus et al., 2003). If the formation of aggregates of viable oocysts in the environment was found to be possible, this would have an impact on their environmental behaviour and should be considered in modelling of Cryptosporidium oocyst transport (Ferguson et al., 2005).

The aim of this chapter was to investigate the conditions of ionic strength and pH required for oocysts to aggregate to determine whether oocysts are likely to aggregate under field or aquatic conditions.

5.2 Materials and Methods

5.2.1 Cryptosporidium oocysts

The Cryptosporidium oocysts used in this study were extracted using diethyl-ether defatting and sucrose flotation as described in Chapter 3, and were less than two months old. In addition, on some occasions the purified oocyst suspensions were filtered through 11-μm pore size nylon netting (Millipore, Australia).

5.2.2 Particle size distributions

Particle sizing of Cryptosporidium oocyst suspensions were determined using the Malvern MastersizerTM E (Malvern Instruments Ltd., Malvern, Worcestershire, UK.), essentially as described in Chapter three. However, the control experiments used a lens

85 with a shorter focal length of 100 mm. As a result, the 31 bands separated the particles into a size range of 0.2 to 180 μm. Since each of the 31 bands covered a smaller range of particle sizes, it was thought this would enable any subtle changes in the particle size profiles to be more easily visualised.

After thorough mixing of the turbid oocyst suspension (~5 x 105 oocysts.mL-1) in the MastersizerTM cell, the obscuration was checked to ensure that it was within the acceptable range (10 to 30%) and that it was preferably over 25% and the reading was stable. Adjustments were made to the Masetersizer cell either by the addition of the suspending medium if the obscuration was too high, or by the addition of more oocysts if too low.

5.2.3 pH experiments

Four tubes containing 15 mL of a low ionic strength solution (0.01 M KNO3) were adjusted with either 4 M KOH or 4 M KCl to obtain pH values between 3 and 9. Cryptosporidium oocysts that had undergone diethyl-ether defatting and sucrose flotation (Chapter three) were then added to each tube to a concentration of ~5 x 105 oocysts.mL-1. The pH was checked after the addition of the oocysts, and again after one hour of incubation at room temperature without agitation. The particle size profiles of each oocyst suspension were measured using the MastersizerTM and the 300 mm focal length lens to measure particles sized between 1.2 and 600 μm as described in Chapter three.

5.2.4 Ionic strength experiments

Filtered Cryptosporidium oocysts were added to the stirred MastersizerTM cell containing 15 mL of either N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES) buffer (Sigma, Sydney, Australia) at pH 6.8 or acidified reagent water at pH 3.5 at a concentration of ~5 x 105 oocysts.mL-1. The particle size profile was measured and then the ionic strength was adjusted by the addition of small quantities of a salt solution (either 4 M NaCl or 4 M MgCl2). After the addition of each quantity of salt, the MastersizerTM cell was stirred for ten minutes before the particle size profile was

86 read. If the obscuration had not become stable ten minutes after the addition of the salt solution, a further five minutes was allowed before measuring the particle size profile. Some formations of aggregates were confirmed by epifluorescence microscopy as described in Chapter three.

5.2.5 Effect of stirring on aggregation

Filtered Cryptosporidium oocysts were added to the stirred MastersizerTM cell containing 15 mL of acidified reagent water at a concentration of ~5 x 105 oocysts.mL-1. Particle size profiles of the oocysts suspension were read every five minutes over a period of 15 minutes, and then at 15 minute intervals for a further 1.75 hours. After two hours the magnetic stirring bar was stopped and the particle size readings were continued at intervals of 15 minutes for a further two hours.

The stirring bar was started again and 0.8 mL of 4 M MgCl2 was added to the MastersizerTM cell. This increased the ionic strength of the suspension in the MastersizerTM cell to 0.59 M. The particle size profiles were then measured at five minute intervals for a 15 minute period and then at 15 minute intervals for a further 1.75 hours. Then the stirring bar was stopped and the particle size profiles measured every

15 minutes for a further two hours. An additional 1.6 mL of 4 M MgCl2 was then added to the MastersizerTM cell, bringing the ionic strength of the suspending solution to 1.6 M. The stirring bar was turned on and again particle size profiles were measured every five minutes for a 15 minute period, and every 15 minutes for a further 30 minutes. The stirring bar was then stopped and the particle size profiles measured every 15 minutes for a 45 minute period.

5.3 Results

5.3.1 Effect of pH on aggregation

The pH values of the prepared oocyst suspensions were read both immediately after the addition of oocysts and again an hour later prior to particle sizing. There was little drift in the pH values and so the final value measured just prior to particle sizing was

87 recorded and used. The particle size profiles of the Cryptosporidium oocyst suspensions in 0.01 M KNO3 (ionic strength 0.01 M) between pH 3.3 and 9.0 are shown in Figure 5.1 (full data set in Appendix A). A peak was obtained in the particle size range 3.5 to 4.3 m for each pH tested, corresponding to the size of single oocysts, and all particle size profiles were similar with the exception of some noise at the higher end of the profile. The Cryptosporidium oocyst suspension used had not been filtered through the 11 m pore size nylon netting, and therefore larger sized particles were seen in the size profile.

18 pH = 3.3 16 pH = 5.2 pH = 7.4 14 pH = 9.0 12

10

8

6

4 Total solids by volume (%) 2

0 12 14 17 21 25 31 38 46 56 68 83 0.5 1.3 1.6 2.0 2.4 2.9 3.5 4.3 5.2 6.4 7.8 9.5 101 124 151 183 224 272 332 404 492 Particle size (μm)

Figure 5.1 Particle size distribution profiles of a Cryptosporidium oocyst suspension in 0.01 M KNO3 at pH values between 3.3 and 9.0

5.3.2 Aggregation with changing ionic strength

No aggregation of oocysts was observed at ionic strengths ranging from 0.025 to 0.46 M at a constant pH of 6.8 in the biological buffer HEPES (Figure 5.2, full data set in Appendix B). A peak at 3.5 to 4.3 m and a similar particle size profile was obtained for each of the different ionic strengths tested. Yet the profile for the sample which did not have any NaCl added (0.025 M) had a lower peak for the Cryptosporidium-sized

88 particles compared to the other profiles. This was due to a peak representing 14% of the total solids at in the 332 - 400 μm size range, attributed to noise, perhaps due to not leaving the sample for long enough before measuring the first profile. While all the other profiles show some noise around the same size it contributed to less than 5% of the total volume of solids.

30 0.025 M 0.035 M 25 0.045 M 0.064 M 0.084 M 20 0.10 M 0.12 M 0.17 M 15 0.21 M 0.26 M 0.30 M 10 0.34 M 0.38 M

Total solids by volume (%) 0.42 M 5 0.46 M

0 12 14 17 21 25 31 38 46 56 68 83 0.5 1.3 1.6 2.0 2.4 2.9 3.5 4.3 5.2 6.4 7.8 9.5 101 123 150 183 223 272 332 404 492 Particle size (μm)

Figure 5.2 Particle size distributions of a Cryptosporidium oocyst suspension at pH 6.8 with ionic strengths varying from 0.025 to 0.46 M with NaCl

Since the oocysts did not aggregate at ionic strengths up to 0.46 M, an experiment was performed with the oocysts in a low pH solution and using MgCl2 to adjust the ionic strength. The low pH was used so that the oocysts were closer to their reported isoelectric point to aid the salting-out effect, and the MgCl2 was used since divalent ions have a greater effect on ionic strength than the monovalent ions of NaCl. Therefore, using the MgCl2 to adjust the ionic strength, values of up to 2.3 M were possible within the volume capacity of the MastersizerTM cell.

Using the divalent salt approach it was seen that again there was little evidence of oocyst aggregation up to an ionic strength of 0.45 M (Figure 5.3, full data set in

89 Appendix C). However, the particle size profile for oocysts in an ionic strength of 0.88 M showed that the spread of the curve for the single oocysts peak had become broader and the peak at 3.5 to 4.3 m had reduced in height without any noise around 400 to 600 m to complicate the profile. Subsequent profiles, as the ionic strength increased, showed the peak at 3.5 to 4.3 m became a smaller percentage of the total solids as peaks representing larger particles began to appear in the profiles. Few oocysts remained as single entities at an ionic strength of greater than 2.4 M (Figure 5.3).

Aggregation of the oocysts was confirmed by epifluorescence microscopy. Groups of oocysts were seen on the membranes after staining and microscopy. The staining step involved a series of washing steps that may have affected the aggregates, and the oocysts that had aggregated into large groups were extremely difficult to count, since the top oocysts mask the oocysts beneath them. Therefore there was no attempt to quantitatively confirm the MastersizerTM results by microscopy. Instead, microscopy was only used to confirm that relative differences in groups of oocysts could be seen after staining to prove that aggregation was occurring.

40 0.0002 M 0.093M 35 0.19 M 0.28 M 30 0.37 M 0.45 M 25 0.88 M 1.3 M 20 1.7 M 2.0 M 2.4 M 15 2.7 M 3.0 M Solids by volume (%) 10 3.3 M 3.6 M 5

0 2 12 14 17 21 25 31 38 46 56 68 83 0.5 1.3 1.6 2.4 2.9 3.5 4.3 5.2 6.4 7.8 9.5 101 123 150 183 223 272 332 404 492 Particle size (um)

Figure 5.3 Particle size distributions of a Cryptosporidium oocyst suspension at pH

3.4 with varying the ionic strengths from 0.002 to 3. 6 M using MgCl2

90

5.3.3 Effect of stirring on the particle size profiles

It was possible that shear forces produced from continuous stirring of oocyst suspensions in the MastersizerTM cell could affect oocyst aggregation, by either not allowing aggregation to occur, or by breaking up weak aggregates while they were being formed. A series of experiments was performed to check whether continuous stirring of the Cryptosporidium oocysts had an effect on the particle size profiles. To more efficiently detect any changes in the profiles at around the 4 μm particle size, the 100 mm focal length lens was used to obtain a profile for particles sized 0.5 to 180 μm.

Figure 5.4 (full data set in Appendix D) illustrates the particle size profiles of stirred oocysts in acidified reagent water over a two hour period, and Figure 5.5 (full data set in Appendix D) shows the corresponding particle size profiles obtained for an additional two hour period after stirring had ceased. In both figures a sharp peak was detected in the size range 3.3 – 4.0 μm, and there was no drift in the height of the peak. The profile was also stable after stirring had ceased, indicating minimal effect of stirring on oocyst aggregation under the conditions used.

The results of the previous experiment in which oocyst aggregation was examined indicated that while oocyst aggregation was not occurring at an ionic strength of 0.46 M, it was starting to occur at an ionic strength of 0.88 M. If stirring was affecting the aggregation, then some difference between stirring and not stirring may be detectable at an ionic strength between 0.46 and 0.88 M.

91 40 0 minutes 35 5 minutes 10 minutes 30 15 minutes 25 30 minutes 45 minutes 20 60 minutes 75 minutes 15 90 minutes 105 minutes

Solids by volume (%) 10 120 minutes 5

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.4 Particle size profiles of a stirred Cryptosporidium oocyst suspension in acidified reagent water (pH 2.7) measured at regular intervals over a two hour period

35

30 0 mins 15 minutes 25 30 minutes 45 minutes 60 minutes 20 75 minutes 90 minutes 15 105 minutes 120 minutes 10 Solids by volume (%)

5

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.5 Particle size profiles of an unstirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 measured at 15 minute intervals over a two hour period

92

Figure 5.6 (full data set in Appendix D) shows the particle size profiles obtained from oocysts suspended in acidified reagent water at an ionic strength of 0.59 M with stirring over a two hour period, and Figure 5.7 (full data set in Appendix D) shows the corresponding particle size profiles for a further two hours after the stirring had ceased. In both figures, a sharp peak occurred in the size range 3.3 to 4.0 μm. The first two consecutive readings taken at five minute intervals after the addition of the salt solution to the MastersizerTM cell had considerable noise at the high end of the profile which corresponded to a drop in the height of the peak in the 3.3 to 4.0 μm size range. In the previous experiments, when the salt solution was added, the particle size profiles were measured ten minutes after the addition of the salt solution, and if the obscuration was still unstable after that time the reading was delayed for a further five minutes. This gave the sample enough time to equilibrate within the MastersizerTM cell. Figure 5.8 shows the same data as Figure 5.6 but without the initial 5 and 10 minute profiles which complicate the chart due to noise obtained from the mixing of the salt solution into the oocyst suspension. The profile obtained at 15 minutes was similar to the subsequent profiles indicating that the sample had stabilised. There was little variation in the particle size profiles over the following 105 minutes of stirring.

93 35 5 minutes 30 10 minutes 15 minutes 25 30 minutes 45 minutes 20 60 minutes 75 minutes 15 90 minutes 105 minutes 10

Solids by volume (%) volume by Solids 120 minutes

5

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.6 Particle size profiles of a stirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 and an ionic strength of 0.59 M measured at regular intervals over a two hour period

25 0 mins 15 minutes 20

) 30 minutes 45 minutes 15 60 minutes 75 minutes 90 minutes 10 105 minutes

Solids by volume (% 120 minutes 5 135 mins

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.7 Particle size profiles of an unstirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 and an ionic strength of 0.59 M measured at 15 minute intervals over a two hour period

94 30

15 minutes 25 30 minutes 45 minutes 20 60 minutes 75 minutes 90 minutes 15 105 minutes 120 minutes 10 Solids by volume (%) 5

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.8 Particle size profiles of a stirred Cryptosporidium oocyst suspension in acidified reagent water at pH 2.7 and an ionic strength of 0.59 M measured at 15 minute intervals over a two hour period, with the 0, 5 and 10 minute profiles removed

The ionic strength was adjusted further by the addition of MgCl2 to 1.6 M. Again there was noise in the particle size profiles for 5, 10 and 15 minute measurements (data not shown) due to the sample taking time to equilibrate after the addition of the salt, and so those profiles were not plotted. Figure 5.9 (full data set in Appendix D) illustrates the results of the particle size profiles using stirring for 30 and 45 minutes, and subsequent 15, 30 and 45 minute measurements after the stirring ceased. The time over which the measurements were taken was reduced to 45 minute since there was no apparent change in the profiles after the sample had equilibrated after the addition of the salt in previous experiments (Figures 5.4 to 5.7). Furthermore, it was thought 45 minutes would be long enough to determine whether or not the oocysts were aggregating. All the particle size profiles had a peak in the 3.3 to 4.0 μm size range, although a second peak was also visible at 15 to 18 μm.

95 30 30 minutes stirring 25 45 minutes stirring 15 minutes no stirring

20 30 minutes no stirring 45 minutes no stirring

15

10 Solids by volume (%) volume by Solids 5

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.9 Stirred and unstirred Cryptosporidium oocyst suspension at an ionic strength of 1.6 M and pH 2.7

Figure 5.10 shows a selection of the profiles from the stirred and not-stirred suspensions of oocysts. Included were the profiles of 0 and 120 minutes of stirring and 120 minutes of not stirring for an ionic strength of 0.002 M (simplified to 0 M on the chart), 120 minutes with and without stirring for an ionic strength of 0.59 M, and with and without stirring for an ionic strength of 1.6 M at 45 minutes. In all profiles the main peak is in the size range 3.3 to 4.0 μm. However, with increasing the ionic strength the height of the peaks diminished. Evidence of oocyst aggregation was detected in an ionic strength of 1.6 M due to a peak at 14 to 18 μm that was not present in the lower ionic strength profiles. However, there was no evidence that stirring was affecting the particle size profiles.

96 35

0 M, t=0 30 0 M, t=120, stirring 25 0 M, t=120, no stirring 0.59 M, t=120, stirring 20 0.59 M, t=120, no stirring 1.6 M, t=45, stirring 15 1.6 M, t=45, no stirring 10 Solids by volume (%) volume by Solids 5

0 10 12 15 18 22 27 32 39 47 57 69 83 0.2 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.7 3.3 4.0 4.8 5.8 7.0 8.5 102 123 149 Particle size (μm)

Figure 5.10 Particle size profiles of a Cryptosporidium oocyst suspension at pH 2.7 in different ionic strengths after two hours with and without stirring

5.4 Discussion

A large number of Cryptosporidium oocysts (~107) was required in the MastersizerTM cell to obtain readings within the acceptable range of obscuration on the instrument. It was difficult to obtain sufficiently high concentrations of oocysts to use a fresh suspension with each ionic strength solution tested. Therefore a single oocyst suspension was used, and the ionic strength adjusted by the addition of salts directly to the MastersizerTM cell. The mixing of the salt solution into the oocyst suspension initially caused unreliable and variable measurements of the size profiles, and so a period of 10 minutes was provided between the addition of salts and reading the profiles. This amount of time was sufficient when small amounts of salts were added, but more than 15 minutes was required for the particle size profiles to stabilise when a large amount of salt was added.

The suspension in the MastersizerTM cell was continually stirred to keep any particles or aggregates that might be formed in suspension. There was a possibility that shear forces produced by the stirring could result in retardation, or complete cessation of the aggregation process, or it could have a destructive effect on formed aggregates.

97 However, there was no evidence of the stirring having any such effect when compared to the same suspension that was undisturbed for the same period of time.

For particles to aggregate there must be a high enough concentration in the suspension such that each particle is allowed ample opportunity to collide with other particles. Reducing the pH to levels close to the isoelectric point allows particles to aggregate more easily, since the electrostatic charge is also reduced and they are less likely to repel (Gerba, 1984). Because there was no evidence of aggregation of oocysts suspended in a solution at pH of 6.8 and with ionic strengths up to 0.46 M, the pH for subsequent experiments was lowered by suspending the oocysts in acidified reagent water. Karaman et al. (1999), Drozd and Schwartzbrod (1996), and Hsu and Huang (2002) reported that the isoelectric point (PI) of Cryptosporidium oocysts is approximately 2, 2.5 and 3.3 respectively, although Brush et al. (1998) reported no net charge over a pH range 2 to 10. Since the extraction method used in this study was similar to that used by Karaman et al. (1999), Drozd and Schwartzbrod (1996), and Hsu and Huang (2002), at pH 2.7 the oocysts are likely to have been approaching their isoelectric point. A divalent salt was also used as it had a greater effect on ionic strength and therefore higher ionic strengths could be obtained from the same starting point within the volume capacity of the MastersizerTM cell.

Changing the pH of the suspending medium (pH range 3.3 to 9.0) did not promote aggregation of Cryptosporidium oocysts. It was only after the ionic strength of the suspending medium was increased to 0.59 M or greater at a low pH that evidence of aggregation was observed. These results correlate with those of Butkus et al. (2003) who showed that oocysts that had been extracted using defatting and flotation did not aggregate in a 0.5 M NaCl solution.

Bukhari and Smith (1995) occasionally observed clumps of varying number of Cryptosporidium oocysts when they viewed wet preparations of calf faeces through a microscope. They suggested that these clumps consisted mostly of non-viable oocysts and deduced that the non-viable (presumably more aged) oocyst walls are more adhesive than those of viable oocysts. Anguish and Ghiorse (1997) came to a similar conclusion about the properties of intact and degraded oocysts when they observed that

98 oocysts seeded into a soil matrix were not closely associated with soil particles, however, empty and degraded oocyst fragments frequently appeared to be embedded in the soil matrix. Kuznar and Elimelech (2006) also showed that the surface macromolecules on Cryptosporidium oocysts impeded their attachment to a quartz surface. When the outer glycoprotein layer was removed from the oocysts, they displayed a significantly higher attachment efficiency compared to oocysts from which the glycoproteins had not been removed. This degradation of the surface macromolecules would occur in non-viable oocysts and may account for the observations of Bukhari and Smith (1995) and Anguish and Ghiorse (1997). The viability of the oocysts used in the current study were not assessed. However, the calf faecal samples collected for oocyst purification were selected according to their visible freshness, the AwwaRF-CRCWQT project used oocysts collected from the same farm and extracted in the same manner, and using excystation they were shown to be generally greater than 80% viable (Davies et al., 2005a).

Haynes and Williams (1992) demonstrated that the ionic strength of soils initially increased from 4 - 6 mM to 24 - 41 mM at the site of bovine urination, and nitrification was the cause of a further increase after 30 days to 130 - 140 mM. There was also an associated initial increase in pH of more than one unit, followed by a decline in pH due to nitrification over the next 30 days. These conditions of ionic strength and pH fluctuations in urine-affected areas could be considered typical in catchments with agricultural grazing. Thus, with oocysts showing a tendency to aggregate only at extreme ionic strengths (> 0.59 M), and showing no tendency to aggregate at pH values between 3.3 and 9.0 and at ionic strengths less than 0.46 M, it would seem that oocysts are not likely to aggregate in conditions generally found in the field environment, even in areas greatly affected by animal urine. In fact, only highly saline areas such as sea- water, which has an ionic strength of 0.7 M (Andersen et al., 2000), or salt-affected lands would be likely to cause significant aggregation of viable Cryptosporidium oocysts.

99 5.5 Conclusions

Cryptosporidium oocysts can aggregate when present in ionic strength solutions greater than 0.59 M and at a low pH. However, it is unlikely that oocysts occurring naturally in the environment would encounter conditions of ionic strength high enough to cause aggregation, particularly in drinking water catchment areas.

100 CHAPTER SIX

CRYPTOSPORIDIUM ATTACHMENT TO SOIL

6.1 Introduction

Bacteria and viruses have been reported to attach to solid surfaces (Loveland et al., 1996; Fang et al., 2000), with their surface chemistry being thought to play a major role in this process. Similarly, researchers have reported the attachment of Cryptosporidium oocysts to environmental particles (Medema et al., 1998; Kuczynska and Shelton, 1999; Kuczynska et al., 2005; Searcy et al., 2005). However, there is also a body of work suggesting that oocysts do not readily attach to soil particles (Anguish and Ghiorse, 1997; Butkus et al., 2003; Dai and Boll, 2003).

The aim of this chapter was to determine whether Cryptosporidium oocysts would attach to soil particles from the field site where the AwwaRF-CRCWQT transport studies were carried out in Sydney’s drinking water catchment (Davies et al., 2005a; Ferguson et al., 2006).

6.2 Materials and Methods

6.2.1 Soil characteristics

The soil used in this study was collected from Arthursleigh Farm, Marulan, approximately 200 km southwest of Sydney, 34°33’06.2’’ north, 150°03’22.6’’ east. This was the field site where overland transport of Cryptosporidium was studied in the AwwaRF-CRCWQT project, “Fate and Transport of Surface Water Pathogens in Watersheds” (Davies et al., 2005a; Ferguson et al., 2006). The loam soil was collected from a depth of 5 – 20 cm below the soil surface. Water decanted after shaking 20 grams of soil in 50 mL of ultra-pure water had a pH of 5.7 and had an electrical -1 conductivity of 105 μS.cm , the median grain size (d50) was 0.07 mm, the uniformity coefficient (d60/d10) was 348, and particles less than 2 μm in size made up 23.8% of the soil (dry weight) (Davies et al., 2005a). The dry weight of the soil at the time of this

101 study was tested in triplicate by standard methods (Palmer and Troeh, 1995; APHA, 1998).

6.2.2 Settling columns

The pipette method (Irani and Callis, 1963; Palmer and Troeh, 1995) was used to measure aqueous oocyst settling under gravity. Soil slurries were prepared in settling cylinders 6 x 43 cm by adding artificial rainwater [4.07 g NaNO3, 3.24 g NaCl, 0.35 g

KCl, 1.65 g CaCl2.2H2O, 2.98 g MgSO4.7H2O, and 3.41 g (NH4)2SO4 per litre of reagent water as per Laegdsmand et al. (1999)] to 20 g of sieved soil. Cryptosporidium oocysts (~105) were mixed into the soil slurry and left to stand for one hour in a constant temperature room set to 20 ºC to allow oocyst attachment to the loam soil particles.

The oocysts used were purified using diethyl-ether defatting and sucrose flotation (Chapter three), and had been stored in PBS for five months at 4 ºC. Additional artificial rainwater was added to the column to obtain a final volume of one litre, the suspensions were mixed thoroughly by inversion, and a sample was immediately collected from the column (0 minutes). The settling columns were then left in a constant temperature room set to 20 ºC for the duration of the settling. Fractions of the suspension were collected from 10 cm below the surface at regular intervals (26, 88, 300, and 900 minutes) over a 15 hour period. These sampling times were chosen based on the calculated settling time for particles sized around 4 μm, which should have settled greater than 10 cm after 15 hours as according to Stoke’s Law (Chapter two). Figure 6.1 shows three columns after approximately 100 minutes of settling. Having been sampled twice in this time period, the volume had dropped from the initial one litre volume in the columns.

The particle size distribution of each collected fraction was measured as described in Chapter 3, and the oocysts in the fractions were quantified. Cryptosporidium oocysts were separated from the soil debris by IMS and enumerated using fluorescence microscopy as described in Chapter three.

102

Figure 6.1 Settling columns after settling for 100 minutes. Each column originally contained 20 g soil (wet weight), 1 L artificial rain water and ~105 Cryptosporidium oocysts

6.2.3 Particle size distribution profiles

Using the Malvern MastersizerTM E, a sub-sample from each fraction collected was subjected to particle sizing as described in Chapter 3. The 300-mm focal length lens was used so that particle sizes from 0.5 to 600 μm were measured. Due to the different turbidities of each fraction, it was necessary to add different volumes to the MastersizerTM cell in order to ensure the obscuration was within acceptable levels. The sub-sample volumes varied between 0.5 and 3 mL, with the volume of reagent water in the MastersizerTM cell ranging from 15 mL to 12 mL so that the cell did not overflow when the larger sub-sample volumes were added.

103 6.2.4 Quantification of Cryptosporidium oocysts

The oocysts in a 10-mL sub-sample of each fraction were separated from the debris using IMS (Dynabeads® anti-Cryptosporidium kit, Dynal, Melbourne, Australia). The methodology used was based on the IMS protocol in the US-EPA method 1622 (US- EPA, 2001). The 1 mL sub-sample was added to a tube containing 1 mL of 10x SL- Buffer-A and 1 mL of 10x SL-Buffer-B. A 100 μL volume of the anti-Cryptosporidium beads were added to the tube which was then rotated at 18 rpm (Sample Mixer, Dynal, Melbourne, Australia) for one hour at room temperature. The samples were then placed in the presence of a magnet (MPC®-1, Dynal, Melbourne, Australia) and rocked for two minutes to collect the beads on the wall of the tube. The debris and buffers were removed using a Pasteur pipette. The tube was removed from the magnet and a 1 mL volume of 1x SL-Buffer-A was added. The beads were gently resuspended from the wall of the tube, transferred to a 1.5 mL centrifuge tube, again placed in the presence of a magnet (MPC®-M, Dynal, Melbourne, Australia) and rocked to collect the beads on the wall of the tube. The supernatant was removed, the tube was removed from the magnet, and 50 μL of 0.1 M HCl was added to the tube and agitated vigorously by vortexing for 30 seconds. The tube was incubated at room temperature for 10 minutes, vortexed vigorously again for 15 seconds, and placed in the presence of a magnetic for the beads to collect again on the wall of the tube. The supernatant was removed, filtered through a membrane, washed with reagent water, stained and viewed using fluorescence microscopy as described in Chapter three.

6.2.5 Collection of fractions method trial

To determine whether a single column could be used for data collection for all time points, a sampling regime that involved the collection of multiple samples from a single column was tested against a series of sacrificial columns. A series of five settling columns, each containing 20 g of soil (wet weight) and one litre of artificial rainwater, was placed in a constant temperature room set to 20 ºC. Column A was sampled at each of the time points t = 0, 26, 88, 300 and 900 minutes. Columns B, C, D, and E were each sampled once only, after 26, 88, 300 and 900 minutes of settling respectively. The particle size profiles of the fractions collected from 10 cm below the surface were measured.

104

6.2.6 Temperature monitoring

An iButton® (a computer chip enclosed in a 16 mm diameter stainless steel can) (Maxim/Dallas Semiconductor Corp, Texas, US) was used to measure and store temperature values. The resolution of the iButton® was 0.5 ºC. In order to measure the temperature inside the columns, an iButton® was placed inside a dummy settling column (a column that didn’t have samples taken from it) at 10 cm from the surface, by hanging it on a wire hook from the top of the column. A second iButton® was used to measure the room temperature for the duration of the experiments. Temperatures were measured at 5 minute intervals. The data was downloaded onto a computer at the conclusion of the experiment.

6.2.7 Statistical analysis

The numbers of Cryptosporidium oocysts in the collected fractions were analysed using ANOVA and Student-Newman-Keuls (SNK) rankings to test for significant differences between means at the = 0.05 level as described in Chapter four.

6.3 Results

6.3.1 Sample collection protocol development

Control experiments were performed to determine whether the collection of multiple samples from a single settling column would significantly disturb the settling, or if sacrificial columns were required for this study. The particle size profiles from the fractions sampled 10 cm below the surface on each of the five sampling occasions were compared to the profiles obtained from the sacrificial columns (Figure 6.2). Larger particles settle more quickly than the smaller particles so that at each sampling time curve peaks are at a lower particle size and the range of the detected particles also reduced. Since the curves are based on the percentage by volume of the total solids in the sample, as the larger particles settle, the peaks of the remaining particles became

105 higher. This is caused by the remaining particles now representing a larger proportion of the remaining total volume of solids in the fractions.

20 A0 A26 B26 15 A88 C88 A300 10 D300 A900 E900

Solids by volume (%) volume by Solids 5

0 12 14 17 21 25 31 37 46 56 68 83 0.5 1.3 1.6 2.0 2.4 2.9 3.5 4.3 5.2 6.4 7.8 9.5 101 123 150 183 223 272 331 404 492 Particle size (m)

Figure 6.2 Particle size profiles for samples collected from five settling columns containing soil. Column A was sampled at each of the five time points, columns B - E were sacrificial columns and were sampled at one time point only

There was very little difference between the particle size profiles regardless of whether the column was sampled once or many times, as shown by the very close particle size profile curves for the different sampling times in Figure 6.2 (full data set in Appendix E). The exception was for the 900 minute fraction from the column that was sampled on multiple occasions (A900) and the 900 minute fraction from the sacrificial column (E 900). The peaks for both fractions were around 2-m in size, however, the peak for the 900 minute fraction of the sacrificial column showed that the particles in the 2.0 - 2.4 m size range consisted of 19% of the total volume of solids, and the equivalent volume from the column sampled on multiple occasions was only 3.1%. There was considerable noise observed in the particle size range 300 m or higher for the A900 fraction. This noise artefact was often observed when the sample did not contain many particles and the obscuration was low. Since the profiles are shown as a percentage of the total volume of solids, noise at around 300 μm significantly reduced any other

106 peaks. Given that both columns gave particle size peaks at the same point, and that the curves mimic each other even though they were some distance apart on the vertical axis, it was considered that they were equivalent.

6.3.2 Gravity settling conditions

Although the constant temperature room where this work was undertaken was set to 20 ºC, the iButtons® placed both in the room and inside a ‘dummy’ column both recorded the temperature as being a constant 22 ºC, with the iButton® in the column fluctuating between 22 and 22.5 ºC. Therefore 22 ºC was used as the experiment temperature for any temperature-dependent parameter values (such as water density) used in Stoke’s equation.

The dry weight of the soil was 88.9% with a standard deviation of 0.04% (n=3).

6.3.3 Soil settling

As observed in the protocol development experiment, the larger sized particles were removed from the fractions over time, and there was a corresponding increase in the peak height of the particle size curves. The density of soil used in this study was not measured. Densities ranging from 1442 - 1602 kg.m-3, however, were chosen as being representative for the loam soil (http://www.simetric.co.uk/si_materials.htm). Other values used in the equation were 1.002 x 10-3 Ns.m-2 for the dynamic viscosity of water at 22 ºC, 9.8 m.s-2 for the acceleration due to gravity, and 997.5 kg.m-3 for the density of water at 22 ºC. Using these densities in Stoke’s equation, a range of particles sizes expected to be excluded at each sampling occasion was obtained (Table 6.1). The actual particle sizes that were present in the main peak as well as the entire range of the main peak are also described in Table 6.1. The entire range of the peak was chosen as those sized particles that contributed to at least 1% of the total solids (by volume) in the sample. While the peak of each curve consisted of smaller particles than the calculated cut-off size for each sampling occasion (using Stoke’s equation), there were still between 12 and 18% of particles in the size profiles that would not be expected if the particles settled according to Stoke’s Law. It is possible that, due to the volume of

107 sample which was removed for analysis at each time point, some particulates lower than the 10 cm depth may have been disturbed and collected during the sampling process.

6.3.4 Cryptosporidium oocyst settling

The Cryptosporidium oocysts counted in the 10 mL volumes of sampled fractions were compared directly without adjustment for recovery efficiency of the method. Errors associated with the methodology were assumed to be constant between the different fractions since the amount of solids in each IMS tube was within the limits specified by the manufacturer. A similar experiment in our laboratory where particles in runoff from a bovine faecal pat were settled and showed that the oocyst recovery was similar between fractions sampled at different time points (Brookes et al., 2006). Furthermore, another study evaluating different Cryptosporidium and Giardia concentration techniques from water showed that the IMS step was highly reproducible even with highly turbid sample concentrates from different water sources (Ferguson et al., 2004).

Particle size profiles for each of the collected fractions is shown in Figure 6.3 (full data set in Appendix F), and the Cryptosporidium numbers detected in each of the fractions are shown in Figure 6.4. One-way ANOVA showed a significant difference between the oocyst concentrations (P < 0.05) from the different fractions. SNK analysis ranked the Cryptosporidium results into three groups, with groups A and B overlapping (Table 6.2). There was a significant reduction in the number of oocysts in the 900 minute fraction, with only 27% of oocysts remaining. At the same time, particle sizing showed that the majority of particles present were around 2 μm in diameter, with only 7% of the particles by volume in the fraction being around 3.5 - 5.2 μm in size.

Similarly to the density of loam soil particles, the densities of Cryptosporidium oocysts reported in the literature vary. The highest (1108 kg.m-3) and lowest (1005 kg.m-3) reported values for oocyst densities (Young and Komisar, 2005b) were used in Stoke’s equation, giving settling rates for oocysts of an average 5 μm in size (range 4 – 6 μm) of 1.5 x 10-6 m.s-1 and 1.0 x 10-7 m.s-1. These rates were used to determine the theoretical distance that single oocysts would have expected to settle through the column on sampling occasions (Table 6.2). Even at 900 minutes, single oocysts would not have

108 settled further than 10 cm from the surface of the columns. Therefore, a longer settling time would have been required to study the migration of single oocysts further than 10 cm. Settling of the oocysts over time was evident, however, from the general decrease in the numbers detected in each fraction, the overlapping SNK rankings for the 0, 26, 88 and 300 minute fractions, and the significant difference in oocyst numbers in those fractions compared to the 900 minute fraction. All these factors indicated an association of the oocysts with the more dense soil particles. The percentage of oocysts remaining in the collected fractions at 26, 88, 300 and 900 minutes were 82, 73, 60 and 27% respectively. Since single oocysts would not have settled further than 10 cm according to Stoke’s Law, the decreasing oocyst counts and overlapping SNK rankings indicate that oocyst settling was most likely associated with soil particles.

The overlapping SNK values make it difficult to specify which particles the oocysts were associating with during settling. To quantify which particle sizes the oocysts are associated with, the size of the settling particles has to be determined. For this study, the particles that were deemed to have settled a significant distance were those from the point at which two consecutive particle size profiles crossed (Figure 6.3) to the point at which the particles in the fraction sampled from the lesser time period contributed less than 1% of the total solids. This gave overlapping size ranges for the settled particles from each fraction, but using this approach it could be argued that approximately 18% of the oocysts settled with particles sized 14 – 124 μm, 9% settled with particles sized 6.4 – 25 μm, 12% settled with particles sized 3.5 - 17 μm, and 33% settled with particles sized 2.4 – 12 μm. From these results it could not be determined whether the remaining 27% of oocysts were attached to small (colloidal) particles or if they remained unassociated with the remaining suspended soil particles.

109 25

t=0 min t=26 mins 20 t=88 mins t=300 mins t=900 mins 15

10 Solids by volume (%) 5

0 12 14 17 21 25 31 37 46 56 68 83 0.5 1.3 1.6 2.0 2.4 2.9 3.5 4.3 5.2 6.4 7.8 9.5 101 123 150 183 223 272 331 404 492 Particle size (μm)

Figure 6.3 Particle size distributions of fractions collected from settling columns at 10 cm below the surface of triplicate settling columns at various times between 0 and 900 minutes, error bars ± 1 SD between columns

700

600

500

400

300

200

Oocysts counted per 10 mL Oocysts counted per 100

0 0 26 88 300 900 Time of sampling (minutes)

Figure 6.4 Oocyst counts from each sampled fraction of triplicate settling columns, error bars are ±1 SD

110

Table 6.1 Estimated and measured particle size fractions in three settling columns associated with each size fraction Settling Estimated particle size in Particle size range of fraction measured time fraction (μm) (μm) (mins) Peak range Entire range 0 All particles 31.0 - 37.8 0.5 - 123.6 26 <16 - <14 7.8 - 9.5 0.5 - 68.3 88 <8.8 – 7.6 4.3 - 5.2 0.5 - 17.1 300 <4.8 - <4.1 2.0 - 2.4 0.5 - 11.6 900 <2.8 - <2.4 2.0 - 2.4 0.5 - 6.3

Table 6.2 Estimated distance of single oocyst settlement using high and low reported oocyst density in Stoke’s equation, compared to the number of oocysts counted in each fraction and their statistical SNK ranking Settling time Predicted distance of Oocyst count per SNK ranking* single oocyst settling 10 mL (min) (cm) Mean (SD) 0 n/a 604 (35) A 26 0.02 – 0.002 497 (44) A B 88 0.08 – 0.008 443 (64) A B 300 0.27 – 0.02 368 (130) B 900 0.81 – 0.06 168 (43) C * Observations with the same letter indicate no significant differences between sample means at  = 0.05.

6.4 Discussion

Differential settling occurs when larger/denser particles settle more rapidly than smaller particles. The terminal velocity of settling particles and aggregates are often assumed to follow Stoke’s law which describes the relationship of settling time to particle density- diameter. For a sedimentation experiment, the concentration should be low enough that

111 every particle has sufficient space to settle independently, but high enough that it can be accurately detected.

A soil suspension of 0.5% (w/v) has been reported to allow individual particles to settle without being influenced by those around them, however the sampled fraction must also contain enough solids for analysis (Irani and Callis, 1963). Often these two restrictions cannot be met simultaneously, and a compromise must be made (Irani and Callis, 1963). Dai and Boll (2003) used a soil concentration of 2 mg.L-1 (0.0002%) since that was the maximum soil concentration for reliable detection of oocysts using flow cytometry. In the current study, a soil concentration of 20 g.L-1 was used. Solids comprised 88.9% of the wet weight, so the dry weight soil concentration in the column was 1.8%, which was high compared to the recommendations of Irani and Callis (1963). In this experiment, however, for the soil particles to be at a high enough concentration to be able to use a small sub-sample of the fraction for an acceptable obscuration in the Malvern MastersizerTM a high soil concentration was required. A lower concentration of soil particles would have required a larger amount of sample to be taken at each time point and this would have had the effect of disturbing the particle settling. Detection of Cryptosporidium oocysts was possible in samples containing 1.5% soil particles because IMS was used to clarify samples prior to microscopy.

Other considerations when using Stoke’s Law are that the diameter of the cylinder should be greater than 4 cm to minimise wall effects, which cause particles to settle at a slower rate due to frictional increases near the wall, and which increase with the size of particles. Also, a constant temperature must also be maintained so that thermal convection currents do not disturb settling or redistribute the particles during settling (Palmer and Troeh, 1995). These recommendations were fulfilled in this study.

The soil settling in the current study was in agreement with the predicted values calculated using Stoke’s equation. While there were particles in the fractions that were larger in size than those predicted by Stoke’s Law, it is likely that during the collection of the fractions some of the larger particles that had already settled past the 10 cm mark may have been drawn into the pipette. Young and Komisar (2005b) used settling

112 columns with ports located in the side of the columns for possibly more accurate sample collection.

In contrast to the soil particles, most of the Cryptosporidium oocysts did not behave as single oocysts according to Stoke’s Law. Attachment of the oocysts to the soil particles would explain the significantly faster than expected settling rates, since even when using the most extreme value for oocyst densities found in the literature (Young and Komisar, 2005b) single oocysts would not have been expected to settle further than 0.8 cm in 900 minutes. The results of this study, however, indicate that at least 72% of the oocysts used settled further than 10 cm in 900 minutes, which implies an association with the soil particles. Due to exceeding the recommended concentration of solids which allows for particles to settle without being influenced by those around them (Irani and Callis, 1963), the oocyst settling behaviour may have been enhanced by being ‘dragged’ by some of the faster settling particles.

Previous studies reported oocyst attachment to soils of 72% with sandy loam, and 93.1% with clay loam soil (Kuczynska et al., 2005), with oocyst recovery from soils being inversely correlated with clay content (Kuczynska and Shelton, 1999). Medema et al. (1998) also reported a significant proportion of oocysts readily attached to particles in wastewater effluent, with 30% of oocysts attaching to particles during the initial few minutes of mixing, and up to 74% attachment after 24 hours of incubation. In contrast, Dai and Boll (2003) report no attachment of oocysts to natural soil particles, and using computer-assisted laser scanning and video microscopy, Anguish and Ghiorse (1997) observed that viable oocysts were not closely associated with soil particles.

Other researchers have shown that attachment of oocysts to surfaces is enhanced when the oocysts are non-viable or not infectious (Bukhari and Smith, 1995; Anguish and Ghiorse, 1997; Kuznar and Elimelech, 2006). While the age of the oocysts used by Kuczynska et al. (2005) is unclear, Medema et al. (1998) used purified oocysts between two and eight months of age that had been stored in potassium dichromate. It is possible that the attachment observed by Medema et al. (1998), and also the association of the five month old oocysts with soil particles observed in the current study, could largely be due to the attachment of degraded oocysts to the soil particles. Searcy et al.

113 (2005), however, observed faster settling of their oocysts that were less than two months old in columns containing suspended sediments (0.2%) compared to columns that did not contain the suspended particles. Attachment of oocysts to the sediment particles was confirmed by fluorescence microscopy (Searcy et al., 2005).

6.5 Conclusions

During settling studies, Cryptosporidium oocysts were associated with the soil particles from the field site used for transport studies described in Davies et al. (2005a) and Ferguson et al. (2006). However, it was not possible to unequivocally confirm attachment, with at least some ‘free’ oocyst settling possibly influenced by faster settling soil particles. It is likely that attachment does occur, with attachment possibly being influenced by factors associated with oocyst age.

114 CHAPTER SEVEN

DISCUSSION

In 1998 Cryptosporidium oocysts and Giardia cysts were detected in the drinking water distribution system of the city of Sydney, Australia. The most likely source of contamination was determined to be from animals in the water catchment area (McClellan, 1998). Genotyping techniques were largely unavailable at the time, and so the oocyst source(s) could only be inferred from knowledge of catchment activities. Heavy rainfall during and prior to the events, and the formation of a thermocline in the main reservoir, were identified as some of the events that led to high numbers of oocysts and cysts being detected in the finished drinking water (McClellan, 1998). This led to the initiation of a project researching the fate and transport of pathogens in catchments, focussing on the microorganisms Escherichia coli, PRD1 bacteriophage and Cryptosporidium. Using simulated rainfall, the movement of these organisms was measured down slopes with different degrees of fall, different amounts of groundcover, and using different intensities and duration of rainfall (Davies et al., 2005a).

Ultimately the aim of the project was to advance the state of knowledge on sources, fate and transport of pathogens in catchments (Ferguson et al., 2003) to enable the movement of these organisms to be modeled. Little information was available with regard to the attachment of Cryptosporidium oocysts to environmental particles. Without this information it was not known whether the oocysts would move as single entities and should be modeled as particles in the 4 – 6 μm size range, or whether attachment to particles that could retard their movement was likely. In addition, the oocysts used in the transport studies were purified using chemical defatting followed by density gradient purification. There was some concern that, due to reports of oocyst treatment changing oocyst behaviour (Ongerth and Pecoraro, 1996; Brush et al., 1998), the chemical treatment of the oocysts would cause them to display a different behaviour in the transport studies compared to the behaviour of native oocysts.

Most oocysts used in research are defatted using a chemical agent such as diethyl-ether followed by flotation over a high density solution such as salt, sucrose, Percoll® or a combination of these (Waldman et al., 1986; Weber et al., 1992; Clavel et al., 1996;

115 Upton, 1997; Kuczynska and Shelton, 1999). Brush et al. (1998) and Dai and Boll (2003) used oocysts that had been extracted by continuous flow centrifugation on a bed of sucrose or salt. While Brush et al. (1998) reported that the electrophoretic mobility of oocysts varied with the method of purification, and that hydrophobic interactions changed with oocyst age, Dai and Boll (2003) reported that the oocyst purification method had no impact on their studies of oocyst attachment to soil particles, however. This reflects the conflicting nature of much of the available information.

The concern addressed in the current study was that even the use of a high ionic strength solution, without defatting, purification methods may have an impact on the oocyst surfaces. Both the NaCl and the sucrose solutions are hyper-osmotic, and extended contact can cause dehydration effects in the closely-related Eimeria oocysts (Ryley et al., 1976). This group of researchers also found that the dehydration effects of NaCl could be reversed without loss of viability, but the effects of extended contact with zinc sulphate, sucrose or saturated salt on Eimeria oocysts were not reversible. To the author’s knowledge, this type of study has not been attempted on Cryptosporidium oocysts, although oocysts suspended in an ionic strength similar to seawater for 40 days show dehydration effects (Freire-Santos et al., 1999). The length of oocyst contact with the flotation solutions, however, is generally significantly less than the contact time used to show dehydration effects in Eimeria oocysts.

Since commonly used flotation solutions could potentially have an effect on the Cryptosporidium oocyst surface, attempts were made in the current study to obtain a purified oocyst suspension without using chemical defatting or the use of dehydrating or high ionic strength solutions. If a highly purified suspension or high concentration of oocysts was not required, then it was possible to obtain a suspension of oocysts containing ~105 oocysts that had not gone through defatting or flotation, but still had many of the faecal solids and lipids removed. This was done by passing slurry of faecal material containing a high concentration of Cryptosporidium oocysts through a column containing SephadexTM G-50 gel, and subsequently removing further lipids by treatment with hydrophobic octyl-SepharoseTM beads. Since a large number of oocysts were required for most of the assays subsequently performed, however, the method was of limited usefulness for oocyst purification. Flow cytometry with cell sorting was able to

116 remove some of the remaining debris after extraction of oocysts using either the diethyl- ether and sucrose method or SephadexTM G-50 and octyl-SepharoseTM beads method.

In order to study attachment of oocysts to particles, the first step was to determine the general surface properties of the oocysts, given it is their surfaces that interact with the environment and give them their characteristic properties. Researchers who have studied the surface properties of Cryptosporidium oocysts are not always in agreement, in part due to the assays used for hydrophobicity measurements being influenced by surface charge (Chapter two).

The current study reinforced the issues raised by Pembrey et al. (1999) and Jones et al. (1996) who stressed the difficulties of comparing results from different methods. The hydrophobicity results obtained in the current study were method dependent. Using the MATH assay the oocysts ranged from low to high hydrophobicity in reagent water, medium to high hydrophobicity in low ionic strength solutions (0.01 N KNO3 and 0.01 M PBS), and high hydrophobicity in a high ionic strength solution (4 M NaCl). The increase in apparent hydrophobicity was consistent with the well-recognised phenomenon of salting-out (Lindahl et al., 1981). The increased ionic strength has the effect of compressing the hair-like structures on the oocyst surface (Considine et al., 2000), and also compressing the double layer of diffuse ions surrounding the oocyst, allowing oocysts to approach other oocysts or surfaces more closely and overcome the repulsion barrier in order for the van der Waals attraction energies to dominate.

In contrast to the MATH assay, the results of hydrophobicity testing using hydrophobic ligands attached to SepharoseTM beads indicated that the purified oocysts showed no hydrophobicity. Some straining effect may have been observed when the SepharoseTM matrix was immobilised in a column. Straining generally becomes an important removal mechanism when the diameters of the suspended particles are larger than 0.2 times the diameter of the particles comprising the porous media (Bouwer, 1984; cited in Stevik et al., 2004). Thus, the 4 – 6 μm sized Cryptosporidium oocysts would not be likely to be strained through a matrix consisting of beads 90-μm in size if they were single entities. Aggregates, however, only have to be 18 μm in size or greater for the effects of straining to occur. Work reported in Chapter 5 described the formation of

117 oocyst aggregates at a low pH and in an ionic strength of 0.59 M that had an effective diameter of 18 μm, and at 3.6 M there were no single Cryptosporidium oocysts remaining in the suspension. Once an aggregate has formed, it is not known how easily the aggregate can disassociate. In the light of the aggregation results from Chapter five, however, it is clear that straining may have been an influencing factor in the retention of oocysts in the HIC columns reported in Chapter four, particularly when oocysts were suspended in 4 M NaCl. In future experiments it would be preferable to monitor both the ionic strength and pH of solutions used. In addition, it is recommended to systematically test for oocyst aggregation with varying combinations of ionic strength and pH in the range relevant to the experiments undertaken.

Using the octyl-SepharoseTM beads in suspension, it was clear that Cryptosporidium did not associate with the matrix. There was no difference detected between attachment results of purified oocysts with octyl-SepharoseTM beads and SepharoseTM without any attached ligands. In fact, the octyl-SepharoseTM beads were found to be more useful in removing calf faecal lipids from oocyst suspensions. The octyl-SepharoseTM beads were thus employed to remove faecal lipids in oocyst suspensions that had passed through a SephadexTM G-50 column. Furthermore, no difference was detected between purified and ‘unpurified’ oocysts with the octyl-SepharoseTM or DEAE-SepharoseTM beads in suspension, showing that different extraction methods did not affect surface properties as determined by the methods used in the current study.

The DEAE-SepharoseTM beads were used to study possible electrostatic effects of attachment. A significant but low level of attachment of oocysts to the DEAE- SepharoseTM beads was observed, suggesting that electrostatic attraction may influence oocyst attachment to particles. Further investigation into the surface charges on Cryptosporidium oocysts was impeded by occupational health and safety issues. The required instrumentation was not located in a designated physical containment level two (PC2) laboratory, and so it was not possible to use viable or potentially viable oocysts in the laboratories where they were located. Although gamma irradiation of oocyst suspensions was an option, it was thought that this treatment may have an impact on the oocyst surface due to disruption of surface proteins, and ultimately the results would not be useful due to these complicating factors.

118

The purified oocysts in the current study showed little tendency to aggregate until extreme ionic strengths and pH were obtained (> 0.59 M, pH 3.4), and therefore it would be unlikely for them to aggregate under normal environmental conditions. These results were similar to those of Butkus et al. (2003) who did not observe aggregation in a 0.5 M ionic strength solution with oocysts that were purified by sucrose flotation only. Those who have observed oocysts aggregating naturally in calf faecal material also noted that the oocysts were predominantly non-viable (Bukhari and Smith, 1995). Given that the results of Kuznar and Elimelech (2006) indicated that in the absence of surface macromolecules on Cryptosporidium oocysts the oocysts had significantly higher attachment efficiency to a quartz surface, it was hypothesised that the lack of aggregation observed in the current study was due to the viability of the oocyst suspension. Although the oocyst viability was not tested, excystation results from related studies using oocysts from the same farm and extracted using the same methods showed that the oocysts were predominantly greater than 80% viable. This implies that the oocysts used in the current study were likely to be predominantly viable. It is recommended, however, that for future experiments the viability of the Cryptosporidium oocysts suspensions are determined at the time of the experimentation.

Results from the settling columns suggested that oocysts were associated with soil particles and therefore settled faster than Stoke’s Law predicts for single oocysts. Since the oocysts did not aggregate at ionic strengths and pH values that were typical of those obtained in the environment, and that environmental particles are predominantly negatively charged (Loder and Liss, 1985; Loveland et al., 1996), the apparent association of the oocysts with the soil particles was confusing.

There are also certain minerals in soils that are positively charged in the pH range of natural waters, and while they are low in abundance they have large surface areas and can therefore play a significant role in transport (Loveland et al., 1996). However, the electrophoretic mobility and hydrophobicity were not determined for the soil particles used in the current study.

119 The Cryptosporidium oocysts in the settling experiment had aged to five months, which is similar to the age of oocysts used by Medema et al. (1998) who reported 34% of their oocysts immediately attaching to environmental wastewater particles and up to 75% of particles showing attachment after 24 hours. The oocysts used by Dai and Boll (2003) were reported to be less than six months of age, and therefore it is difficult to determine whether the lack of association between their oocysts and soil particles was age related or due to other factors. While the authors attributed the lack of attachment to soil particles to a negative charge of the oocysts, they also acknowledged that the small number of particles in their settling columns (2 mg/L) may have contributed. Infact, it is likely that with this concentration of soil particles that there were insufficient opportunities for oocyst collision with the soil particles to evaluate whether attachment would occur, even after mixing the sample for 24 hours prior to settling. The opposite effect may have been an issue in the current study. That is, that there were more solids in the settling columns than was recommended when applying Stoke’s Law, and so the particles may have been influenced by those settling around them. In other words, it was hypothesised that Cryptosporidium oocysts may have been ‘dragged’ along with some of the faster settling soil particles without being physically attached to them. It is therefore recommended that for any future settling column experiments that there be a period allowed whereby the oocysts and soil particles are able to be in close contact in order for any attachment to occur. This would be prior to diluting the soil to the recommended concentration of 2 mg/L. This should be sufficient to avoid the effect of ‘dragging’ in the future.

Cryptosporidium oocysts may be excreted in the faeces of humans and animals, and in order for them to move into waterways they must first be released from the faecal pat/septic zone. Researchers who have studied the release of Cryptosporidium oocysts from faecal pats have reported that the mechanical forces associated with droplet impact result in a higher release rate of oocysts than when using mist (Bradford and Schijven, 2002; Schijven et al., 2004) although some of this droplet impact is reduced with pooling water (Dai, 2003). Muirhead et al. (2005) reported that only 8% of the Escherichia coli cells released from faecal cow pats with rainfall were attached to particles, and concluded that most E. coli eroded from cow pats would occur as readily- transportable single cells. While Schijven et al. (2004) studied the release of

120 Cryptosporidium from bovine faecal material due to rainfall, and Bradford and Schijven (2002) the change in release due to different ionic strengths, the issue of whether oocysts are released from faecal material as single entities or attached to faecal particles has not been rigourously studied. Brookes et al. (2006), however, showed that oocysts in run-off collected from artificial rainfall on cow pats that were spiked with purified oocysts settled at a similar rate to that predicted by Stoke’s Law, suggesting that purified oocysts were released from cow pats as single oocysts and unassociated with large faecal particles.

Bradford and Schijven (2002) observed that rainfall only depleted the finer manure particles on the surface of the faecal pat, and therefore the faecal pat will act as a long- term source of contamination. Similarly, in the AwwaRF-CRCWQT project, the faecal pats were shown to provide a low level source of contamination, with only a small percentage of the oocysts being removed from the pat per rain event (Davies et al., 2005a). Furthermore, one week old faecal pats released significantly fewer oocysts than fresh cowpats. The aging of the pats had the effect of forming a harder ‘crust’ on the outside edge of the pat making them less permeable to water. Biotic activity and desiccation have both been shown to affect oocyst viability (Fayer et al., 2000b), and with a decrease in viability and degradation of the surface proteins, there would have been an expected increase in the hydrophobicity of the oocysts (Kuznar and Elimelech, 2005). Therefore, the oocysts may have become more hydrophobic with aging, and possibly result in attachment to faecal solids, making them less likely to be transported from the cowpat.

The AwwaRF-CRCWQT project reported that oocysts moved further overland, and more rapidly, when there was no vegetation compared to grassed soil (Davies et al., 2005a). This was largely a function of the volume of run off generated, with significantly less runoff being generated when vegetation was present and a corresponding increase in soil infiltration of the rain (Davies et al., 2004). Some oocysts were transported rapidly, with oocysts being detected ten metres from the faecal pat within five minutes of rainfall and, based on the results in the current study, possibly due to the movement of free individual oocysts.

121 Although the oocysts used in the AwwaRF-CRCWQT transport studies were extracted using diethyl-ether defatting and sucrose flotation, there is little evidence that the results would have been affected by the method of oocyst extraction. The purified oocysts in the current study were also shown to be stable in low ionic strength solutions, with no attachment to hydrophobic ligands on SepharoseTM beads and low to moderate removal by the MATH assay, which is well known to be influenced by factors other than hydrophobicity. Furthermore, partitioning of the oocysts to the hydrophobic octane layer could not be confirmed. Although there was some attachment of oocysts to DEAE-SepharoseTM, it was only a small proportion, indicating that in terms of hydrophobicity and surface charge the oocysts demonstrated stability in low ionic strength solutions.

A significant conclusion for the current study is that rainfall-released cow pat oocysts are unlikely to aggregate in most environments. The exception would be if the runoff was produced in highly saline areas, since an ionic strength of around 0.59 M was required for aggregation to occur. In addition, the attachment of oocysts to particles in the soil used for the transport studies was inconclusive. Although attachment could not be confirmed, however, there was an association of the oocysts with the soil particles as they settled. The run-off samples collected from the bare soil plots in the AwwaRF- CRCWQT transport studies were highly turbid (Davies et al., 2005a), indicating that some of the soil particles were highly mobile. It is possible that if the oocyst settling presented in Chapter 6 was due to the faster settling particles influencing the smaller ones (i.e. dragging), then the same mechanism could be involved in the transport of oocysts ten metres in less than five minutes of rainfall.

In conclusion, the oocysts used in the current study displayed no hydrophobicity characteristics with octyl-SepharoseTM beads, but variable attachment to octane in the MATH assay dependent upon the isolate and not age factors. There were a small number of oocysts that displayed the ability to attach to DEAE ligands, implying that some of the oocysts were charged. The oocyst extraction method used did not appear to affect the surface properties of the oocysts with respect to attachment to DEAE and octyl ligands. Purified oocysts did not aggregate in solutions of ionic strength and pH values that would generally be expected in environmental aquatic conditions.

122 Attachment of oocysts to soil particles was not conclusively shown. Although as association of oocysts with settling soil particles occurred, the amount of soil in the settling columns exceeded the recommended 0.2 – 0.5% (Irani and Callis, 1963) and the oocyst settling may have been influenced by the faster settling soil particles. Approximately 27% of the oocysts spiked into the settling columns remained unattached or were associated with soil particles less than 2.4 m in size. This implies that the results of the overland transport studies from the AwwaRF-CRCWQT project were valid, since there was no evidence that the oocyst extraction method affected the oocysts. Further research, however, is required before determining whether oocysts should be modeled as single entities or associated with soil particles.

123 CHAPTER EIGHT

FUTURE RESEARCH

During the course of the research presented in this thesis, it became clear that obtaining a pure suspension of Cryptosporidium oocysts at a high concentration was difficult to achieve without the use of chemicals such as diethylether or high ionic strength solutions that could possibly impact the surfaces of oocysts. This inability to obtain a highly pure suspension of oocysts that had not gone through a harsh purification procedure meant that particle sizing was an inappropriate method for aggregation studies of these oocysts. Development of a more suitable process for studying the aggregation of such oocysts is desirable.

To further the understanding of whether oocysts would attach to the soil particles, the use of partially purified using physical means instead of chemical, could be used in settling columns. Some modifications of the method used in this study would be required, and a system similar to that used by Young and Komisar (2005b) would give more controlled conditions than was used in this study. The work of Young and Komisar (2005b) also indicated that oocyst suspensions were comprised of populations of both viable and non-viable oocysts, and oocysts of varying densities, and so this possibility must be taken into account for any future settling column studies.

In addition to the study of oocyst attachment to soil, settling columns could be used to study whether or not oocysts in faecal material are found in aggregates or attached to faecal particles. There is little information available as to whether oocysts are eluted from faecal pats as single entities or not, yet this is important information to include in any model on oocyst movement. While there has been some research into quantifying oocyst elution using the force of rain drops, little is known about the nature of the oocysts once eluted. The assumption is made that oocysts are eluted from faecal pats as single entities and are not attached or aggregated at that stage.

The research presented in this study showed that oocysts are not likely to aggregate in rainwater systems, it is possible that they would aggregate in highly saline areas. It

124 would be worthwhile to study the strength of aggregates, and to determine under what conditions the aggregate would break into smaller aggregates or single entities.

It is important to note that future research involving the study of surface properties of Cryptosporidium oocysts should note not only the age of the oocysts, storage conditions and extraction procedures used, but also the viability of oocyst suspensions. This information may have been able to explain discrepencies in some of the results obtained for this research.

125 REFERENCES

Aboytes, R., G. D. di Giovanni, F. A. Abrams, C. Rheinecker, W. McElroy, N. Shaw and M. W. LeChevallier. 2004. Detection of infectious Cryptosporidium in filtered drinking water. Journal of the American Water Works Association. 96:88-98. Abu-Zreig, M., R. P. Rudra and H. R. Whiteley. 2001. Validation of a vegetated filter strip model (VFSMOD). Hydrological Processes. 15:729-742. Ahimou, F., M. Paquot, P. Jacques, P. Thonart and P. G. Rouxhet. 2001. Influence of electrical properties on the evaluation of the surface hydrophobicity of Bacillus subtilis. Journal of Microbiological Methods. 45:119-126. Andersen, R. J., G. R. Bigg, D. R. Corbett, S. J. de Mora, B. A. McKee, W. L. Miller, J. M. Smoak, P. W. Swarzenski and D. E. Williams. 2000. Chemistry in the Marine Environment. The Royal Society of Chemistry. Anguish, L. J. and W. C. Ghiorse. 1997. Computer-assisted laser scanning and video microscopy for analysis of Cryptosporidium parvum oocysts in soil, sediment, and feces. Applied and Environmental Microbiology. 63:724-733. APHA. 1998. Standard methods for the examination of water and wastewater. 20th ed. American Public Health Association, Washington D.C. Arakawa, T. and S. N. Timasheff. 1982. Preferential interactions of proteins with salts in concentrated solutions. Biochemistry. 21:6545-6552. Arakawa, T. and S. N. Timasheff. 1984. Mechanism of protein salting in and salting out by divalent cation salts: balance between hydration and salt binding. Biochemistry. 23:5912-5923. Ashbolt, N. J. and D. J. Roser. 2003. Interpretation and management implications of event and baseflow pathogen data. In: “Watershed management for water supply systems.” Pfeiffer, M. J., D. J. V. Abs and K. N. Brooks (eds.), pp. CD-ROM American Water Resources Association, New York City, New York June 30 – July 2, 2003. Atherholt, T. B., M. W. LeChevallier, W. D. Norton and J. S. Rosen. 1998. Effect of rainfall on Giardia and crypto. Journal of the American Water Works Association. 90:66-80. Atwill, E. R., L. Hou, B. M. Karle, T. Harter, K. W. Tate and R. A. Dahlgren. 2002. Transport of Cryptosporidium parvum oocysts through vegetated buffer strips and estimated filtration efficiency. Applied and Environmental Microbiology. 68:5517- 5527. Atwill, E. R., B. Hoar, M. das Graças Cabral Pereira, K. W. Tate, F. Rulofson and G. Nader. 2003. Improved quantitative estimates of low environmental loading and sporadic periparturient shedding of Cryptosporidium parvum in adult beef cattle. Applied and Environmental Microbiology. 69:4604-4610. Aydin, Y. and I. A. Özkul. 1996. Infectivity of Cryptosporidium muris directly isolated from the murine stomach for various laboratory animals. Veterinary Parasitology. 66:257-262.

126 Barker, I. K. and P. L. Carbonell. 1974. Cryptosporidium agni sp.n. from lambs, and Cryptosporidium bovis sp.n. from a calf, with observations on the oocyst. Zeitschrift für Parasitenkunde. 44:289-298. Barnes, D. A., A. Bonnin, J.-X. Huang, L. Gousset, J. Wu, J. Gut, P. Doyle, J.-F. Dubremetz, H. Ward and C. Petersen. 1998. A novel multi-domain mucin-like glycoprotein of Cryptosporidium parvum mediates invasion. Molecular and Biochemical Parasitology. 96:93-110. Barta, J. R. and R. C. A. Thompson. 2006. What is Cryptosporidium? Reappraising its biology and phylogenetic affinities. TRENDS in Parasitology. 22:463-468. Barwick, R. S., H. O. Mohammed, M. E. White and R. B. Bryant. 2003. Factors Associated with the Likelihood of Giardia spp. and Cryptosporidium spp. in Soil from Dairy Farms. Journal of Dairy Science. 86:784-791. Becher, K. A., I. D. Robertson, D. M. Fraser, D. G. Palmer and R. C. A. Thompson. 2004. Molecular epidemiology of Giardia and Cryptosporidium infections in dairy calves originating from three sources in Western Australia. Veterinary Parasitology. 123:1-9. Belosevic, M., R. A. Guy, R. Taghi-Kilani, N. F. Neumann, L. L. Gyürék, L. R. J. Liyanage, P. J. Millard and G. R. Finch. 1997. Nucleic acid stains as indicators of Cryptosporidium parvum oocyst viability. International Journal for Parasitology. 27:787-798. Bergendahl, J. and D. Grasso. 1999. Prediction of colloid detachment in a model porous media: thermodynamics. American Institute of Chemical Engineers. AIChE Journal. 45:475-484. Bergendahl, J. and D. Grasso. 2000. Prediction of colloid detachment in a model porous media: hydrodynamics. Chemical Engineering Science. 55:1523-1532. Birdi, K. S. 1997. Handbook of surface and colloid chemistry CRC Press LLC, Boca Raton. Blasdall, S. A., J. E. Ongerth and N. J. Ashbolt. 2003. Differentiation of Cryptosporidium parvum subtypes by a novel microsatelite-telomere PCR with PAGE. In: “Cryptosporidium: from molecules to disease.” Thompson, R. C. A., A. Armson and U. M. Ryan (eds.), Elsevier B. V., Amsterdam. Bodley-Tickell, A. T., S. E. Kitchen and A. P. Sturdee. 2002. Occurrence of Cryptosporidium in agricultural surface waters during an annual farming cycle in lowland UK. Water Research. 36:1880-1886. Bradford, S. A. and J. Schijven. 2002. Release of Cryptosporidium and Giardia from dairy calf manure: impact of solution salinity. Environmental Science and Technology. 36:3916-3923. Brookes, J. D., C. M. Davies, M. R. Hipsey and J. P. Antenucci. 2006. Association of Cryptosporidium with bovine faecal particles and implications for risk reduction by settling within water supply reservoirs. Journal of Water and Health. 4:87-98. Brush, C. F., M. F. Walter, L. J. Anguish and W. C. Ghiorse. 1998. Influence of pretreatment and experimental conditions on electrophoretic mobility and hydrophobicity of Cryptosporidium parvum oocysts. Applied and Environmental Microbiology. 64:4439-4445.

127 Brush, C. F., W. C. Ghiorse, L. J. Anguish, J.-Y. Parlange and H. G. Grimes. 1999. Transport of Cryptosporidium parvum Oocysts through Saturated Columns. Journal of Environmental Quality. 28:809-815. Buckley, R. and W. Warnken. 2003. Giardia and Cryptosporidium in pristine protected catchments in central Eastern Australia. Ambio. 32:84-86. Bukhari, Z. and H. V. Smith. 1995. Effect of three concentration techniques on viability of Cryptosporidium parvum oocysts recovered from bovine feces. Journal of Clinical Microbiology. 33:2592-2595. Bukhari, Z., W. G. Glen and J. L. Clancy. 1999. Effects of pH on a fluorogenic vital dye assay (4',6'-diamidino-2-phenyl-indole and propodium iodide) for Cryptosporidium sp. oocysts. Water Research. 33:3037-3039. Bukhari, Z., M. M. Marshall, D. G. Korich, C. R. Fricker, H. V. Smith, J. Rosen and J. L. Clancy. 2000. Comparison of Cryptosporidium parvum viability and infectivity assays following ozone treatment of oocysts. Applied and Environmental Microbiology. 66:2972-2980. Bunt, C. R., D. S. Jones and I. G. Tucker. 1993. The effects of pH, ionic strength and organic phase on the bacterial adhesion to hydrocarbons (BATH) test. International Journal of Pharmaceutics. 99:93-98. Bunt, C. R., D. S. Jones and I. G. Tucker. 1995. The effects of pH, ionic strength and polyvalent ions on the cell surface hydrophobicity of Escherichia coli evaluated by the BATH and HIC methods. International Journal of Pharmaceutics. 113:257-261. Busscher, H. J., A. H. Weerkamp, H. C. van der Mei, A. W. J. van Pelt, H. P. de Jong and J. Arends. 1984. Measurement of the surface free energy of bacterial cell surfaces and its relevance for adhesion. Applied and Environmental Microbiology. 48:980-983. Busscher, H. J., J. Sjollema and H. C. Van der Mei. 1990. Relative importance of surface free energy as a measure of hydrophobicity in bacterial adhesion to solid surfaces. In “Microbial cell surface hydrophobicity.” Doyle, R. J. and M. Rosenberg (eds.), American Society for Microbiology, Washington DC, pp. 335- 361. Busscher, H. J., B. van de Belt-Gritter and H. C. van der Mei. 1995. Implications of microbial adhesion to hydrocarbons for evaluating cell surface hydrophobicity 1. Zeta potentials of hydrocarbon droplets. Colloids and Surfaces B: Biointerfaces. 5:111-116. Busscher, H. J. and W. Norde. 2000. Limiting values for bacterial zeta potentials. Journal of Biomedical Materials Research. 50:463-464. Bustamante, H. A., S. R. Shanker, R. M. Pashley and M. E. Karaman. 2001. Interaction between Cryptosporidium oocysts and water treatment coagulants. Water Research. 35:3179-3189. Butkus, M. A. and D. Grasso. 1998. Impact of aqueous electrolytes on interfacial energy. Journal of Colloid and Interface Science. 200:172-181. Butkus, M. A., J. T. Bays and M. P. Labare. 2003. Influence of surface characteristics on the stability of Cryptosporidium parvum oocysts. Applied and Environmental Microbiology. 69:3819-3825.

128 Butt, H.-J. 1991. Measuring electrostatic, van der Waals, and hydration forces in electrolyte solutions with an atomic force microscope. Biophysical Journal. 60:1438-1444. Camesano, T. A. and N. I. Abu-Lail. 2002. Heterogeneity in bacterial surface polysaccharides, probed on a single-molecule basis. Biomacromolecules. 3:661-667. Campbell, A. T., L. J. Robertson and H. V. Smith. 1992. Viability of Cryptosporidium parvum oocysts: correlation of in vitro excystation with inclusion or exclusion of fluorogenic vital dyes. Applied and Environmental Microbiology. 58:3488-3493. Campbell, A. T., L. J. Robertson and H. V. Smith. 1993. Effects of preservatives on viability of Cryptosporidium parvum oocysts. Applied and Environmental Microbiology. 59:4361-4362. Capizzi-Banas, S., M. Maux and J. Schwartzbrod. 2002. Surface hydrophobicity of Ascaris eggs and Giardia cysts. Helminthologia. 39:197-204. Capowiez, Y. and A. Bérard. 2006. Assessment of the effects of imidacloprid on the behavior of two earthworm species (Aporrectodea nocturna and Allolobophora icterica) using 2D terraria. Ecotoxicology and Environmental Safety. 64:198-206. Casemore, D. P., S. E. Wright and R. L. Coop. 1997. Cryptosporidiosis – human and animal epidemiology. In “Cryptosporidium and cryptosporidiosis.” Fayer, R. (ed.), CRC Press LLC, Boca Raton, pp. 65-92. Chalmers, R. M., K. Elwin, W. J. Reilly, H. Irvine, A. L. Thomas and P. R. Hunter. 2002. Cryptosporidium in farmed animals: the detection of a novel isolate in sheep. International Journal for Parasitology. 32:21-26. Chattopadhyay, S. and R. W. Puls. 1999. Adsorption of bacteriophages on clay minerals. Environmental Science and Technology. 33:3609-3614. Chattopadhyay, S. and R. W. Puls. 2000. Forces dictating colloidal interactions between viruses and soil. Chemosphere. 41:1279-1286. Chen, G. and K. A. Strevett. 2001. Impact of surface thermodynamics on bacterial transport. Environmental Microbiology. 3:237-245. Chung, J., W. Hijnen and N. J. Ashbolt. 2004. Potential Cryptosporidium oocyst surrogates for sand filtration and the importance of their surface properties. Programme and abstracts of the International Giardia and Cryptosporidium conference, Amsterdam, The Netherlands. September 20 – 24. Clavel, A., A. Arnal, E. Sánchez, M. Varea, J. Quílez, I. Ramírez and F. J. Castillo. 1996. Comparison of 2 centrifugation procedures in the formalin-ethyl acetate stool concentration technique for the detection of Cryptosporidium oocysts. International Journal for Parasitology. 26:671-672. Collins, R. and K. Rutherford. 2004. Modelling bacterial water quality in streams draining pastoral land. Water Research. 38:700-712. Considine, R. F., D. R. Dixon and C. J. Drummond. 2000. Laterally-resolved force microscopy of biological microspheres-oocysts of Cryptosporidium parvum. Langmuir. 16:1323-1330.

129 Considine, R. F., D. R. Dixon and C. J. Drummond. 2002. Oocysts of Cryptosporidium parvum and model sand surfaces in aqueous solutions: an atomic force microscope (AFM) study. Water Research. 36:3421-3428. Cox, P., M. Griffith, M. Angles, D. Deere and C. Ferguson. 2005. Concentrations of pathogens and indicators in animal feces in the Sydney watershed. Applied and Environmental Microbiology. 71:5929-5934. Coyne, M. S., R. A. Gilfillen, R. W. Rhodes and R. L. Blevins. 1995. Soil and fecal coliform trapping by grass filter strips during simulated rain. Journal of Soil and Water Conservation. 50:405-408. Craun, G. F., R. L. Calderon and N. Nwachuku. 2003. Causes of waterborne outbreaks reported in the United States, 1991-1998. In: “Drinking water and infectious disease, establishing the links.” Hunter, P. R., M. Waite and E. Ronchi (eds.), CRC Press LLC, London. Current, W. L., N. C. Reese, J. V. Ernst, W. S. Bailey, M. B. Heyman and W. M. Weinstein. 1983. Human cryptosporidiosis in immunocompetent and immunodeficient persons. Studies of an outbreak and experimental transmission. The New England Journal of Medicine. 308:1252-1257. Curriero, F. C., J. A. Patz, J. B. Rose and S. Lele. 2001. The association between extreme precipitation and waterborne disease outbreaks in the United States, 1948- 1994. American Journal of Public Health. 91:1194-1199. D'Antonio, R. G., R. E. Winn, J. P. Taylor, T. L. Gustafson, W. L. Current, M. M. Rhodes, G. W. Gary, Jr. and R. A. Zajac. 1985. A waterborne outbreak of cryptosporidiosis in normal hosts. Annals of Internal Medicine. 103:886-888. Dahlbäck, B., M. Hermansson, S. Kjelleberg and B. Norkrans. 1981. The Hydrophobicity of bacteria - an important factor in their initial adhesion at the air- water interface. Archives of Microbiology. 128:267-270. Dai, X. and R. M. Hozalski. 2002. Effect of NOM and biofilm on the removal of Cryptosporidium parvum oocysts in rapid filters. Water Research. 36:3523-3532. Dai, X. 2003. Transport kinetics of Cryptosporidium parvum and Giardia lamblia in overland flow. PhD thesis, Biological and Agricultural Engineering, University of Idaho. Dai, X. and J. Boll. 2003. Evaluation of attachment of Cryptosporidium parvum and Giardia lamblia to soil particles. Journal of Environmental Quality. 32:296-304. Dai, X., J. Boll, M. E. Hayes and D. E. Aston. 2004. Adhesion of Cryptosporidium parvum and Giardia lamblia to solid surfaces: the role of surface charge and hydrophobicity. Colloids and Surfaces B: Biointerfaces. 34:259-263. Dai, X. and J. Boll. 2006. Settling velocity of Cryptosporidium parvum and Giardia lamblia. Water Research. 40:1321-1325. Darnault, C. J. G., P. Garnier, Y.-J. Kim, K. L. Oveson, T. S. Steenhuis, J.-Y. Parlange, M. Jenkins, W. C. Ghiorse and P. Baveye. 2003. Preferential transport of Cryptosporidium parvum oocysts in variably saturated subsurface environments. Water Environment Research. 75:113-120.

130 Davies, C. M., C. Kaucner, D. Deere and N. J. Ashbolt. 2003. Recovery and enumeration of Cryptosporidium parvum from animal fecal matrices. Applied and Environmental Microbiology. 69:2842-2847. Davies, C. M., C. M. Ferguson, C. Kaucner, M. Krogh, N. Altavilla, D. A. Deere and N. J. Ashbolt. 2004. Dispersion and transport of Cryptosporidium Oocysts from fecal pats under simulated rainfall events. Applied and Environmental Microbiology. 70:1151-1159. Davies, C., C. Kaucner, N. Altavilla, N. Ashbolt, C. Ferguson, M. Krogh, W. Hijnen, G. Medema and D. Deere. 2005a. Fate and transport of surface water pathogens in watersheds. American Water Works Research Foundation Report 1P-4.75C- 91078F-10/05-CM, 261 pp. Denver, Colorado, USA. Davies, C. M., N. Altavilla, M. Krogh, C. M. Ferguson, D. A. Deere and N. J. Ashbolt. 2005b. Environmental inactivation of Cryptosporidium oocysts in catchment soils. Journal of Applied Microbiology. 98:308-317. Deere, D. D., G. Vesey, N. Ashbolt and M. Gauci. 2002. Flow cytometry and cell sorting for monitoring microbial cells. In: “Encyclopedia of environmental microbiology.” Bitton, G. (ed.), Wiley and Sons, New York. Vol. 4:1944-1958. Deng, M. Q., D. O. Cliver and T. W. Mariam. 1997. Immunomagnetic capture PCR to detect viable Cryptosporidium parvum oocysts from environmental samples. Applied and Environmental Microbiology. 63:3134-3138. Dong, H. 2002. Significance of electrophoretic mobility distribution to bacterial transport in granular porous media. Journal of Microbiological Methods. 51:83-93. Drozd, C. and J. Schwartzbrod. 1996. Hydrophobic and electrostatic cell surface properties of Cryptosporidium parvum. Applied and Environmental Microbiology. 62:1227-1232. Ducker, W. A., T. J. Senden and R. M. Pashley. 1991. Direct measurement of colloidal forces using an atomic force microscope. Nature. 353:239-241. Dufour, A. P., M. Feige, M. Messner, S. Regli, C. Rodgers, F. W. Schaefer III, S. Schaub, J. Sinclair and L. J. Wymer. 1999. Criteria for evaluation of proposed protozoan detection methods. Lindquist, H. D. A. (ed.), U.S. Environmental Protection Agency, Washington, D. C. Enemark, H. L., V. Bille-Hansen, P. Lind, P. M. H. Heegaard, H. Vigre, P. Ahrens and S. M. Thamsborg. 2003. Pathogenicity of Cryptosporidium parvum-evaluation of an animal infection model. Veterinary Parasitology. 113:35-57. Entrala, E., J.-M. Molina-Molina, M.-J. Rosales-Lombardo, M. Sánchez-Moreno and C. Mascaró-Lazcano. 2000. Cryptosporidium parvum: oocysts purification using potassium bromide discontinuous gradient. Veterinary Parasitology. 92:223-226. Entry, J. A., R. K. Hubbard, J. E. Thies and J. J. Fuhrmann. 2000. The influence of vegetation in riparian filterstrips on coliform bacteria: I. Movement and survival in water. Journal of Environmental Quality. 29:1206-1214. Esquibel-King, M. A., A. C. Dias-Cabral, J. A. Queiroz and N. G. Pinto. 1999. Study of hydrophobic interaction adsorption of bovine serum albumin under overloaded conditions using flow microcalorimetry. Journal of Chromatography A. 865:111- 122.

131 Fall, A., R. C. A. Thompson, R. P. Hobbs and U. Morgan-Ryan. 2003. Morphology is not a reliable tool for delineating species within Cryptosporidium. The Journal of Parasitology. 89:399-402. Fang, H. H. P., K.-Y. Chan and L.-C. Xu. 2000. Quantification of bacterial adhesion forces using atomic force microscopy (AFM). Journal of Microbiological Methods. 40:89-97. Fayer, R., L. Gasbarre, P. Pasquali, A. Canals, S. Almeria and D. Zarlenga. 1998a. Cryptosporidium parvum infection in bovine neonates: dynamic clinical, parasitic and immunologic patterns. International Journal for Parasitology. 28:49-56. Fayer, R., J. M. Trout and M. C. Jenkins. 1998b. Infectivity of Cryptosporidium parvum oocysts stored in water at environmental temperatures. The Journal of Parasitology. 84:1165-1169. Fayer, R., U. Morgan and S. J. Upton. 2000a. Epidemiology of Cryptosporidium: transmission, detection and identification. International Journal for Parasitology. 30:1305-1322. Fayer, R., J. M. Trout, E. Walsh and R. Cole. 2000b. Rotifers ingest oocysts of Cryptosporidium parvum. The Journal of Eukaryotic Microbiology. 47:161-163. Fayer, R. 2004. Cryptosporidium: a water-borne zoonotic parasite. Veterinary Parasitology. 126:37-56. Fayer, R., M. Santín, J. M. Trout and E. Greiner. 2006. Prevalence of species and genotypes of Cryptosporidium found in 1-2-year-old dairy cattle in the eastern United States. Veterinary Parasitology. 135:105-112. Feick, J. D. and D. Velegol. 2002. Measurements of charge nonuniformity on polystyrene latex particles. Langmuir. 18:3454-3458. Ferguson, C., A. M. d. R. Husman, N. Altavilla, D. Deere and N. Ashbolt. 2003. Fate and transport of surface water pathogens in watersheds. Critical Reviews in Environmental Science and Technology. 33:299-361. Ferguson, C., C. Kaucner, M. Krogh, D. Deere and M. Warnecke. 2004. Comparison of methods for the concentration of Cryptosporidium oocysts and Giardia cysts from raw waters. Canadian Journal of Microbiology. 50:675-682. Ferguson, C. M., B. Croke, N. J. Ashbolt and D. A. Deere. 2005. A deterministic model to quantify pathogen loads in drinking water catchments: pathogen budget for the Wingecarribee. Water Science and Technology. 52:191-197. Ferguson, C. M., C. M. Davies, C. Kaucner, N. J. Ashbolt, M. Krogh, J. Rodehutskors and D. A. Deere. 2006. Field scale quantification of microbial transport from bovine faeces under simulated rainfall events. Journal of Water and Health. 5:83-95. Freire-Santos, F., A. M. Oteiza-López, C. A. Vergara-Castiblanco and M. E. Ares- Mazás. 1999. Effect of salinity, temperature and storage time on mouse experimental infection by Cryptosporidium parvum. Veterinary Parasitology. 87:1- 7. Gannon, J., Y. Tan, P. Baveye and M. Alexander. 1991a. Effect of sodium chloride on transport of bacteria in a saturated aquifer material. Applied and Environmental Microbiology. 57:2497-2501.

132 Gannon, J. T., V. B. Manilal and M. Alexander. 1991b. Relationship between cell surface properties and transport of bacteria through soil. Applied and Environmental Microbiology. 57:190-193. Gantzer, C., L. Gillerman, M. Kuznetsov and G. Oron. 2001. Adsorption and survival of faecal coliforms, somatic coliphages and F-specific RNA phages in soil irrigated with wastewater. Water Science and Technology. 43:117-124. Gardner, K. H. and T. L. Theis. 1996. A unified kinetic model for particle aggregation. Journal of Colloid and Interface Science. 180:162-173. Gardner, K. H., T. L. Theis and T. C. Young. 1998a. The significance of shear stress in the agglomeration kinetics of fractal aggregates. Water Research. 32:2660-2668. Gardner, K. H., T. L. Theis and T. C. Young. 1998b. Colloid aggregation: numerical solution and measurements. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 141:237-252. Gary, H. L., S. R. Johnson and S. L. Ponce. 1983. Cattle grazing impact on surface water quality in a Colorado Front Range stream. Journal of Soil and Water Conservation. 38:124-128. Geertsema-Doornbusch, G. I., H. C. van der Mei and H. J. Busscher. 1993. Microbial cell surface hydrophobicity. Journal of Microbiological Methods. 18:61-68. Gennaccaro, A. L., M. R. McLaughlin, W. Quintero-Betancourt, D. E. Huffman and J. B. Rose. 2003. Infectious Cryptosporidium parvum oocysts in final reclaimed effluent. Applied and Environmental Microbiology. 69:4983-4984. Gerba, C. P. 1984. Applied and theoretical aspects of virus adsorption to surfaces. Advances in Applied Microbiology. 30:133-168. Ghodrati, M., M. Chendorain and Y. J. Chang. 1999. Characterization of macropore flow mechanisms in soil by means of a split macropore column. Soil Science Society of America Journal. 63:1093-1101. Gilbert, P., D. J. Evans, E. Evans, I. G. Duguid and M. R. W. Brown. 1991. Surface characteristics and adhesion of Escherichia coli and Staphylococcus epidermidis. The Journal of Applied Bacteriology. 71:72-77. Gitis, V., R. C. Haught, R. M. Clark and G. Rothenberg. 2002. Assessing the removal of inorganic colloids and Cryptosporidium parvum from drinking water. Journal of Environmental Monitoring. 4:244-248. Glynn, J. R., Jr., B. M. Belongia, R. G. Arnold, K. L. Ogden and J. C. Baygents. 1998. Capillary electrophoresis measurements of electrophoretic mobility for colloidal particles of biological interest. Applied and Environmental Microbiology. 64:2572- 2577. Gmachowski, L. 1996. Hydrodynamics of aggregated media. Journal of Colloid and Interface Science. 178:80-86. Grasso, D., B. F. Smets, K. A. Strevett, B. D. Machinist, C. J. Van Oss, R. F. Giese and W. Wu. 1996. Impact of physiological state on surface thermodynamics and adhesion of Pseudomonas aeruginosa. Environmental Science and Technology. 30:3604-3608.

133 Gregory, J. 1994. Cryptosporidium in water: treatment and monitoring methods. Filtration and Separation. 31:283-289. Guardani, R., C. A. O. Nascimento and R. S. Onimaru. 2002. Use of neural networks in the analysis of particle size distribution by laser diffraction: tests with different particle systems. Powder Technology. 126:42-50. Guselle, N. J., A. J. Appelbee and M. E. Olson. 2003. Biology of Cryptosporidium parvum in pigs: from weaning to market. Veterinary Parasitology. 113:7-18. Hancock, C. M., J. B. Rose and M. Callahan. 1998. Crypto and Giardia in US groundwater. Journal of the American Water Works Association. 90:58-61. Harris, J. R. and F. Petry. 1999. Cryptosporidium parvum: structural components of the oocyst wall. The Journal of Parasitology. 85:839-840. Harter, T., S. Wagner and E. R. Atwill. 2000. Colloid transport and filtration of Cryptosporidium parvum in sandy soils and aquifer sediments. Environmental Science and Technology. 34:62-70. Harvey, R. W. and S. P. Garabedian. 1991. Use of colloid filtration theory in modeling movement of bacteria through a contaminated sandy aquifer. Environmental Science and Technology. 25:178-185. Harvey, R. W., D. W. Metge, N. Kinner and N. Mayberry. 1997. Physiological considerations in applying laboratory-determined buoyant densities to predictions of bacterial and protozoan transport in groundwater. Results of in-situ and laboratory tests. Environmental Science and Technology. 31:289-295. Hawkins, P., P. Cox, I. Fisher, G. Kastl, V. Jegatheesan, M. Warnecke, M. Angles, H. Bustamante and A. Chiffings. 2001. The 1998 Sydney water crisis - an alternative point of view. Water Journal of the Australian Water and Wastewater Association. 28:37-40. Hayes, E. B., T. D. Matte, T. R. O'Brien, T. W. McKinley, G. S. Logsdon, J. B. Rose, B. L. P. Ungar, D. M. Word, P. F. Pinsky, M. L. Cummings, M. A. Wilson, E. G. Long, E. S. Hurwitz and D. D. Juranek. 1989. Large community outbreak of cryptosporidiosis due to contamination of a filtered public water supply. The New England Journal of Medicine. 320:1372-1376. Haynes, R. J. and P. H. Williams. 1992. Changes in soil solution composition and pH in urine-affected areas of pasture. Journal of Soil Science. 43: 323-334. Hazen, K. C. and B. W. Hazen. 1987. A polystyrene microsphere assay for detecting surface hydrophobicity variations within Candida albicans populations. Journal of Microbiological Methods. 6:289-299. Heisz, M. 1997. Studies on the survival and viability testing of Cryptosporidium parvum oocysts in the water environment. MSc. Thesis, Microbiology and Immunology, University of Ottawa Hermansson, M., S. Kjelleberg, T. K. Korhonen and T.-A. Stenström. 1982. Hydrophobic and electrostatic characterization of surface structures of bacteria and its relationship to adhesion to an air-water interface. Archives of Microbiology. 131:308-312.

134 Hijjawi, N. S., B. P. Meloni, U. M. Ryan, M. E. Olson and R. C. A. Thompson. 2002. Successful in vitro cultivation of Cryptosporidium andersoni: evidence for the existence of novel extracellular stages in the life cycle and implications for the classification of Cryptosporidium. International Journal for Parasitology. 32:1719- 1726. Hijnen, W. A., A. J. Brouwer-Hanzens, K. J. Charles and G. J. Medema. 2005. Transport of MS2 phage, Escherichia coli, Clostridium perfringens, Cryptosporidium parvum, and Giardia intestinalis in a gravel and a sandy soil. Environmental Science and Technology. 39:7860-7868. Hörman, A., R. Rimhanen-Finne, L. Maunula, C.-H. von Bonsdorff, N. Torvela, A. Heikinheimo and M.-L. Hänninen. 2004. Campylobacter spp., Giardia spp., Cryptosporidium spp., noroviruses, and indicator organisms in surface water in southwestern Finland, 2000-2001. Applied and Environmental Microbiology. 70:87- 95. Hoxie, N. J., J. P. Davis, J. M. Vergeront, R. D. Nashold and K. A. Blair. 1997. Cryptosporidiosis-associated mortality following a massive waterborne outbreak in Milwaukee, Wisconsin. American Journal of Public Health. 87:2032-2035. Hsu, B.-M., C. Huang and J. R. Pan. 2001. Filtration behaviors of Giardia and Cryptosporidium - ionic strength and pH effects. Water Research. 35:3777-3782. Hsu, B.-M. and C. Huang. 2002. Influence of ionic strength and pH on hydrophobicity and zeta potential of Giardia and Cryptosporidium. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 201:201-206. Hunter, P. R. and Q. Syed. 2001. Community surveys of self-reported diarrhoea can dramatically overestimate the size of outbreaks of waterborne cryptosporidiosis. Water Science and Technology. 43:27-30. Inoue, M., S. Uga, T. Oda, S. K. Rai, G. Vesey and H. Hotta. 2006. Changes of physical and biochemical properties of Cryptosporidium oocysts with various storage conditions. Water Research. 40:881-886. Ionas, G., J. J. Learmonth, E. A. Keys and T. J. Brown. 1998. Distribution of Giardia and Cryptosporidium in natural water systems in New Zealand - a nationwide survey. Water Science and Technology. 38:57-60. Irani, R. R. and C. F. Callis. 1963. (eds.). Particle Size: Measurement, Interpretation, and Application. John Wiley and Sons, New York. Jacobsen, O. H., P. Moldrup, C. Larsen, L. Konnerup and L. W. Petersen. 1997. Particle transport in macropores of undisturbed soil columns. Journal of Hydrology. 196:185-203. Jenkins, M. B., L. J. Anguish, D. D. Bowman, M. J. Walker and W. C. Ghiorse. 1997. Assessment of a dye permeability assay for determination of inactivation rates of Cryptosporidium parvum oocysts. Applied and Environmental Microbiology. 63:3844-3850. Jenkins, M. B., M. J. Walker, D. D. Bowman, L. C. Anthony and W. C. Ghiorse. 1999. Use of a sentinel system for field measurements of Cryptosporidium parvum oocyst inactivation in soil and animal waste. Applied and Environmental Microbiology. 65:1998-2005.

135 Jenkins, M., J. M. Trout, J. Higgins, M. Dorsch, D. Veal and R. Fayer. 2003. Comparison of tests for viable and infectious Cryptosporidium parvum oocysts. Parasitology Research. 89:1-5. Jervis, H. R., T. G. Merrill and H. Sprinz. 1966. Coccidiosis in the guinea pig small intestine due to a Cryptosporidium. American Journal of Veterinary Research. 27:408-414. Jewett, D. G., T. A. Hilbert, B. E. Logan, R. G. Arnold and R. C. Bales. 1995. Bacterial transport in laboratory columns and filters: Influence of ionic strength and pH on collision efficiency. Water Research. 29:1673-1680. Joachim, A. 2004. Human cryptosporidiosis: an update with special emphasis on the situation in Europe. Journal of Veterinary Medicine. B, Infectious Diseases and Veterinary Public Health. 51:251-259. Jones, D. S., C. G. Adair, W. M. Mawhinney and S. P. Gorman. 1996. Standardisation and comparison of methods employed for microbial cell surface hydrophobicity and charge determination. International Journal of Pharmaceutics. 131:83-89. Jones, J. F., J. D. Feick, D. Imoudu, N. Chukwumah, M. Vigeant and D. Velegol. 2003. Oriented adhesion of Escherichia coli to polystyrene particles. Applied and Environmental Microbiology. 69:6515-6519. Karaman, M. E., R. M. Pashley, H. Bustamante and S. R. Shanker. 1999. Microelectrophoresis of Cryptosporidium parvum oocysts in aqueous solutions of inorganic and surfactant cations. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 146:217-225. Karim, M. R., F. D. Manshadi, M. M. Karpiscak and C. P. Gerba. 2004. The persistence and removal of enteric pathogens in constructed wetlands. Water Research. 38:1831-1837. Kaucner, C. and T. Stinear. 1998. Sensitive and rapid detection of viable Giardia cysts and Cryptosporidium parvum oocysts in large-volume water samples with wound fiberglass cartridge filters and reverse transcription-PCR. Applied and Environmental Microbiology. 64:1743-1749. Kistemann, T., T. Claßen, C. Koch, F. Dangendorf, R. Fischeder, J. Gebel, V. Vacata and M. Exner. 2002. Microbial load of drinking water reservoir tributaries during extreme rainfall and runoff. Applied and Environmental Microbiology. 68:2188- 2197. Korich, D. G., J. R. Mead, M. S. Madore, N. A. Sinclair and C. R. Sterling. 1990. Effects of ozone, chlorine dioxide, chlorine, and monochloramine on Cryptosporidium parvum oocyst viability. Applied and Environmental Microbiology. 56:1423-1428. Korich, D. G., M. M. Marshall, H. V. Smith, J. O'Grady, Z. Bukhari, C. R. Fricker, J. P. Rosen and J. L. Clancy. 2000. Inter-laboratory comparison of the CD-1 neonatal mouse logistic dose-response model for Cryptosporidium parvum oocysts. The Journal of Eukaryotic Microbiology. 47:294-298. Kuczynska, E. and D. R. Shelton. 1999. Method for detection and enumeration of Cryptosporidium parvum oocysts in feces, manures, and soils. Applied and Environmental Microbiology. 65:2820-2826.

136 Kuczynska, E., D. R. Shelton and Y. Pachepsky. 2005. Effect of bovine manure on Cryptosporidium parvum oocyst attachment to soil. Applied and Environmental Microbiology. 71:6394-6397. Kulkarni, P., G. Dutari, P. Biswas and R. Haught. 2004. Gravity settling characteristics of Cryptosporidium parvum oocysts in aqueous suspension using in situ static light scattering. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 233:1-10. Kuznar, Z. A. and M. Elimelech. 2005. Role of surface proteins in the deposition kinetics of Cryptosporidium parvum oocysts. Langmuir. 21:710-716. Kuznar, Z. A. and M. Elimelech. 2006. Cryptosporidium oocyst surface macromolecules significantly hinder oocyst attachment. Environmental Science and Technology. 40:1837-1842. Laegdsmand, M., K. G. Villholth, M. Ullum and K. H. Jensen. 1999. Processes of colloid mobilization and transport in macroporous soil monoliths. Geoderma. 93:33- 59. Langmuir, I. 1938. The role of attractive and repulsive forces in the formation of tactoids, thixotropic gels, protein crystals and coacervates. Journal of Chemical Physics. 6:873-896. Le Berre, F., G. Chauveteau and E. Pefferkorn. 1998. Shear induced aggregation/fragmentation of hydrated colloids. Journal of Colloid and Interface Science. 199:13-21. LeChevallier, M. W., M. Abbaszadegan and G. D. Di Giovanni. 2000. Detection of infectious oocysts in environmental water samples using an integrated cell culture- PCR (CC-PCR) system. Water, Air and Soil Pollution. 123:53-65. Lichtenberg, D., M. Rosenberg, N. Sharfman and I. Ofek. 1985. A kinetic approach to bacterial adherence to hydrocarbon. Journal of Microbiological Methods. 4:141- 146. Lindahl, M., A. Faris, T. Wadström and S. Hjertén. 1981. A new test based on 'salting out' to measure relative surface hydrophobicity of bacterial cells. Biochimica et Biophysica Acta. 677:471-476. Lindsay, D. S., S. J. Upton, D. S. Owens, U. M. Morgan, J. R. Mead and B. L. Blagburn. 2000. Cryptosporidium andersoni n. sp. (Apicomplexa: Cryptosporiidae) from cattle, Bos taurus. Journal of Eukaryotic Microbiology. 47:91-95. Llovo, J., A. Lopez, J. Fabregas and A. Muñoz. 1993. Interaction of lectins with Cryptosporidium parvum. The Journal of Infectious Diseases. 167:1477-1480. Lloyd, S. and J. Smith. 1997. Pattern of Cryptosporidium parvum oocyst excretion by experimentally infected dogs. International Journal for Parasitology. 27:799-801. Loder, T. C. and P. S. Liss. 1985. Control by organic coatings of the surface charge of estuarine suspended particles. Limnology and Oceanography. 30:418-421. Loveland, J. P., J. N. Ryan, G. L. Amy and R. W. Harvey. 1996. The reversibility of virus attachment to mineral surfaces. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 107:205-221.

137 MacKenzie, W. R., N. J. Hoxie, M. E. Proctor, M. S. Gradus, K. A. Blair, D. E. Peterson, J. J. Kazmierczak, D. G. Addiss, 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. New England Journal of Medicine. 331:161-167. Martin, M. A., M. A. Pfaller, R. M. Massanari and R. P. Wenzel. 1989. Use of cellular hydrophobicity, slime production, and species identification markers for the clinical significance of coagulase-negative staphylococcal isolates. American Journal of Infection Control. 17:130-135. Mawdsley, J. L., R. D. Bardgett, R. J. Merry, B. F. Pain and M. K. Theodorou. 1995. Pathogens in livestock waste, their potential for movement through soil and environmental pollution. Applied Soil Ecology. 2:1-15. McBride, M. B. and P. Baveye. 2002. Diffuse double-layer models, long-range forces, and ordering in clay colloids. Soil Science Society of America Journal. 66:1207- 1217. McClellan, P. 1998. Sydney Water Inquiry. Final Report. Premier's Department, NSW Government, Sydney, Australia. McDowell-Boyer, L. M., J. R. Hunt and N. Sitar. 1986. Particle transport through porous media. Water Resources Research. 22:1901-1921. McGechan, M. B. and D. R. Lewis. 2002. SW--Soil and Water: Transport of Particulate and Colloid-sorbed Contaminants through Soil, Part 1: General Principles. Biosystems Engineering. 83:255-273. McGlade, T. R., I. D. Robertson, A. D. Elliot, C. Read and R. C. A. Thompson. 2003. Gastrointestinal parasites of domestic cats in Perth, Western Australia. Veterinary Parasitology. 117:251-262. McNeil-Watson, F., W. Tscharnuter and J. Miller. 1998. A new instrument for the measurement of very small electrophoretic mobilities using phase analysis light scattering (PALS). Colloids and Surfaces A: Physicochemical and Engineering Aspects. 140:53-57. Medema, G. J., F. M. Schets, P. F. M. Teunis and A. H. Havelaar. 1998. Sedimentation of free and attached Cryptosporidium oocysts and Giardia cysts in water. Applied and Environmental Microbiology. 64:4460-4466. Meehan, W. R. and W. S. Platts. 1978. Livestock grazing and the aquatic environment. Journal of Soil and Water Conservation. 33:274-278. Meisel, J. L., D. R. Perera, C. Meligro and C. E. Rubin. 1976. Overwhelming watery diarrhea associated with a Cryptosporidium in an immunosuppressed patient. Gastroenterology. 70:1156-1160. Missana, T. and A. Adell. 2000. On the applicability of DLVO theory to the prediction of clay colloids stability. Journal of Colloid and Interface Science. 230:150-156. Monis, P. T. and R. C. A. Thompson. 2003. Cryptosporidium and Giardia-zoonoses: fact or fiction? Infection, Genetics and Evolution. 3:233-244. Morgan, U. M., C. C. Constantine, D. A. Forbes and R. C. A. Thompson. 1997. Differentiation between human and animal isolates of Cryptosporidium parvum

138 using rDNA sequencing and direct PCR analysis. Journal of Parasitology. 83:825- 830. Morgan, U. M., K. D. Sargent, P. Peplazes, D. A. Forbes, F. Spano, H. Hertzberg, A. Elliot and R. C. A. Thompson. 1998. Molecular characterization of Cryptosporidium from various hosts. Parasitology. 117:31-37. Morgan, U. M., J. R. Buddle, A. Armson, A. Elliot and R. C. A. Thompson. 1999a. Molecular and biological characterisation of Cryptosporidium in pigs. Australian Veterinary Journal. 77:44-47. Morgan, U. M., L. Xiao, R. Fayer, A. A. Lal and R. C. A. Thompson. 1999b. Variation in Cryptosporidium: towards a taxonomic revision of the genus. International Journal for Parasitology. 29:1733-1751. Morgan-Ryan, U. M., A. Fall, L. A. Ward, N. Hijjawi, I. Sulaiman, R. Fayer, R. C. A. Thompson, M. Olson, A. Lal and L. Xiao. 2002. Cryptosporidium hominis n. sp. (Apicomplexa: Cryptosporidiidae) from Homo sapiens. The Journal of Eukaryotic Microbiology. 49:433-440. Mozes, N. and P. G. Rouxhet. 1987. Methods for measuring hydrophobicity of microorganisms. Journal of Microbiological Methods. 6:99-112. Muirhead, R. W., R. P. Collins and P. J. Bremer. 2005. Erosion and subsequent transport state of Escherichia coli from cowpats. Applied and Environmental Microbiology. 71:2875-2879. Musial, C. E., M. J. Arrowood, C. R. Sterling and C. P. Gerba. 1987. Detection of Cryptosporidium in water by using polypropylene cartridge filters. Applied and Environmental Microbiology. 53:687-692. Nanduri, J., S. Williams, T. Aji and T. P. Flanigan. 1999. Characterization of an immunogenic glycocalyx on the surfaces of Cryptosporidium parvum oocysts and sporozoites. Infection and Immunity. 67:2022-2024. Nielsen, J. L., L. H. Mikkelsen and P. H. Nielsen. 2001. In situ detection of cell surface hydrophobicity of probe-defined bacteria in activated sludge. Water Science and Technology. 43:97-103. Nime, F. A., J. D. Burek, D. L. Page, M. A. Holscher and J. H. Yardley. 1976. Acute enterocolitis in a human being infected with the protozoan Cryptosporidium. Gastroenterology. 70:592-598. O'Donoghue, P. J. 1995. Cryptosporidium and cryptosporidiosis in man and animals. International Journal for Parasitology. 25:139-195. Olson, M. E., C. L. Thorlakson, L. Deselliers, D. W. Morck and T. A. McAllister. 1997. Giardia and Cryptosporidium in Canadian farm animals. Veterinary Parasitology. 68:375-381. Olson, M. E., J. Goh, M. Phillips, N. Guselle and T. A. McAllister. 1999. Giardia cyst and Cryptosporidium oocyst survival in water, soil, and cattle feces. Journal of Environmental Quality. 28:1991-1996. Ongerth, J. E. and J. P. Pecoraro. 1996. Electrophoretic mobility of Cryptosporidium oocysts and Giardia cysts. Journal of Environmental Engineering. 122:228-231.

139 Palmer, R. G. and F. R. Troeh. 1995. Introductory Soil Science: Laboratory Manual. Oxford University Press, New York. Panciera, R. J., R. W. Thomassen and F. M. Garner. 1971. Cryptosporidial infection in a calf. Veterinary Pathology. 8:479-484. Park, C. H. and P. M. Huck. 2003. A conceptual model for Cryptosporidium transport in watersheds. Water Quality Research Journal of Canada. 38:77-113. Pedersen, K. 1980. Electrostatic interaction chromatography, a method for assaying the relative surface charges of bacteria. FEMS Microbiology Letters. 12:365-367. Pedrós-Alió, C., J. Mas, J. M. Gasol and R. Guerrero. 1989. Sinking speeds of free- living photogrophic bacteria determined with covered and uncovered traps. Journal of Plankton Research. 11:887-905. Pembrey, R. S., K. C. Marshall and R. P. Schneider. 1999. Cell surface analysis techniques: what do cell preparation protocols do to cell surface properties? Applied and Environmental Microbiology. 65:2877-2894. Peng, M. M., L. Xiao, A. R. Freeman, M. J. Arrowood, A. A. Escalante, A. C. Weltman, C. S. L. Ong, W. R. MacKenzie, A. A. Lal and C. B. Beard. 1997. Genetic polymorphism among Cryptosporidium parvum isolates: evidence of two distinct human transmission cycles. Emerging Infectious Diseases. 3:567-573. Perz, J. F. and S. M. Le Blancq. 2001. Cryptosporidium parvum infection involving novel genotypes in wildlife from lower New York state. Applied and Environmental Microbiology. 67:1154-1162. Pokorny, N. J., S. C. Weir, R. A. Carreno, J. T. Trevors and H. Lee. 2002. Influence of temperature on Cryptosporidium parvum oocyst infectivity in river water samples as detected by tissue culture assay. The Journal of Parasitology. 88:641-643. Power, M. L., M. B. Slade, N. C. Sangster and D. A. Veal. 2004. Genetic characterisation of Cryptosporidium from a wild population of eastern grey kangaroos Macropus giganteus inhabiting a water catchment. Infection, Genetics and Evolution. 4:59-67. Power, M. L., N. C. Sangster, M. B. Slade and D. A. Veal. 2005. Patterns of Cryptosporidium oocyst shedding by eastern grey kangaroos inhabiting an Australian watershed. Applied and Environmental Microbiology. 71:6159-6164. Proctor, S. J. and R. L. Kemp. 1974. Cryptosporidium anserinum sp. n. (Sporozoa) in a domestic goose Anser anser L., from Iowa. Journal of Protozoology. 21:664-666. Queiroz, J. A., C. T. Tomaz and J. M. S. Cabral. 2001. Hydrophobic interaction chromatography of proteins. Journal of Biotechnology. 87:143-159. Quinn, C. M., G. P. Archer, W. B. Betts and J. G. O'Neill. 1996. Dose-dependent dielectrophoretic response of Cryptosporidium oocysts treated with ozone. Letters in Applied Microbiology. 22:224-228. Rahe, T. M., C. Hagedorn, E. L. McCoy and G. F. Kling. 1978. Transport of antibiotic- resistant Escherichia coli through western Oregon hillslope soils under conditions of saturated flow. Journal of Environmental Quality. 7:487-494.

140 Ramirez, N. E., L. A. Ward and S. Sreevatsan. 2004. A review of the biology and epidemiology of cryptosporidiosis in humans and animals. Microbes and Infection. 6:773-785. Ranucci, L., H.-M. Müller, G. La Rosa, I. Reckmann, M. A. G. Morales, F. Spano, E. Pozio and A. Crisanti. 1993. Characterization and immunolocalization of a Cryptosporidium protein containing repeated amino acid motifs. Infection and Immunity. 61:2347-2356. Razatos, A., Y.-L. Ong, M. M. Sharma and G. Georgiou. 1998. Molecular determinants of bacterial adhesion monitored by atomic force microscopy. Proceedings of the National Academy of Sciences of the United States of America. 95:11059-11064. Ren, J. and A. I. Packman. 2002. Effects of background water composition on stream- subsurface exchange of submicron colloids. Journal of Environmental Engineering. 128:624-634. Robertson, L. J., A. T. Campbell and H. V. Smith. 1992. Survival of Cryptosporidium parvum oocysts under various environmental pressures. Applied and Environmental Microbiology. 58:3494-3500. Robertson, L. J., A. T. Campbell and H. V. Smith. 1993. In vitro excystation of Cryptosporidium parvum. Parasitology. 106:13-19. Robertson, B., M. I. Sinclair, A. B. Forbes, M. Veitch, M. Kirk, D. Cunliffe, J. Willis and C. K. Fairley. 2002. Case-control studies of sporadic cryptosporidiosis in Melbourne and Adelaide, Australia. Epidemiology and Infection. 128:419-431. Rochelle, P. A., R. De Leon, M. H. Stewart and R. L. Wolfe. 1997. Comparison of primers and optimization of PCR conditions for detection of Cryptosporidium parvum and Giardia lamblia in water. Applied and Environmental Microbiology. 63:106-114. Rochelle, P. A., M. M. Marshall, J. R. Mead, A. M. Johnson, D. G. Korich, J. S. Rosen and R. De Leon. 2002. Comparison of in vitro cell culture and a mouse assay for measuring infectivity of Cryptosporidium parvum. Applied and Environmental Microbiology. 68:3809-3817. Rochelle, P. A., S. J. Upton, B. A. Montelone and K. Woods. 2005. The response of Cryptosporidium parvum to UV light. TRENDS in Parasitology. 21:81-87. Rose, J. B., A. Cifrino, M. S. Madore, C. P. Gerba, C. R. Sterling and M. J. Arrowood. 1986. Detection of Cryptosporidium from wastewater and freshwater environments. Water Science and Technology. 18:233-239. Rose, J. B., H. Darbin and C. P. Gerba. 1988. Correlations of the protozoa, Cryptosporidium and Giardia, with water quality variables in a watershed. Water Science and Technology. 20:271-276. Rose, J. B. 1997. Environmental ecology of Cryptosporidium and public health implications. Annual Review of Public Health. 18:135-161. Rose, J. B., D. E. Huffman and A. Gennaccaro. 2002. Risk and control of waterborne cryptosporidiosis. FEMS Microbiology Reviews. 26:113-123.

141 Rosenberg, M., D. Gutnick and E. Rosenberg. 1980. Adherence of bacteria to hydrocarbons: a simple method for measuring cell-surface hydrophobicity. FEMS Microbiology Letters. 9:29-33. Rosenberg, M. and R. J. Doyle. 1990. Microbial cell surface hydrophobicity: history, measurement, and significance. In Microbial cell surface hydrophobicity Doyle, R. J. and M. Rosenberg (eds.), pp. 1-37 American Society for Microbiology, Washington, DC. Roser, D. J., N. J. Ashbolt, R. Leeming, R. Kagi and T. D. Waite. 2003. In: “26th International Hydrology and Water Resources Symposium.” M. J. Boyd (ed.). Barton A.C.T. Institution of Engineers, Novotel, Wollongong, Australia. 10 – 14 November. Ruecker, N. J. and N. F. Neumann. 2006. Re-assessing the public health risk associated with the waterborne transmission of Cryptosporidium spp. Reviews in Medical Microbiology. 17:1-9. Ryan, U., C. Read, P. Hawkins, M. Warnecke, P. Swanson, M. Griffith, D. Deere, M. Cunningham and P. Cox. 2005a. Genotypes of Cryptosporidium from Sydney water catchment areas. Journal of Applied Microbiology. 98:1221-1229. Ryan, U. M., C. Bath, I. Robertson, C. Read, A. Elliot, L. McInnes, R. Traub and B. Besier. 2005b. Sheep may not be an important zoonotic reservoir for Cryptosporidium and Giardia parasites. Applied and Environmental Microbiology. 71:4992-4997. Ryley, J. F., R. Meade, J. Hazelhurst and T. E. Robinson. 1976. Methods in coccidiosis research: separation of oocysts from faeces. Parasitology. 73:311-326. Santín, M., J. M. Trout, L. Xiao, L. Zhou, E. Greiner and R. Fayer. 2004. Prevalence and age-related variation of Cryptosporidium species and genotypes in dairy calves. Veterinary Parasitology. 122:103-117. Schets, F. M., G. B. Engels, M. During and A. M. de Roda Husman. 2005. Detection of infectious Cryptosporidium oocysts by cell culture immunofluorescence assay: applicability to environmental samples. Applied and Environmental Microbiology. 71:6793-6798. Schiemann, D. A., M. R. Crane and P. J. Swanz. 1987. Surface properties of Yersinia species and epithelial cell interactions in vitro by a method measuring total associated, attached and intracellular bacteria. Journal of Medical Microbiology. 24:205-218. Schijven, J. F. and S. M. Hassanizadeh. 2000. Removal of viruses by soil passage: overview of modeling, processes, and parameters. Critical Reviews in Environmental Science and Technology. 30:49-127. Schijven, J. F., S. A. Bradford and S. Yang. 2004. Release of Cryptosporidium and Giardia from dairy cattle manure: physical factors. Journal of Environmental Quality. 33:1499-1508. Schneider, P. F. and T. V. Riley. 1991. Cell-surface hydrophobicity of Staphylococcus saprophyticus. Epidemiology and Infection. 106:71-75.

142 Searcy, K. E., A. I. Packman, E. R. Atwill and T. Harter. 2005. Association of Cryptosporidium parvum with suspended particles: impact on oocyst sedimentation. Applied and Environmental Microbiology. 71:1072-1078. Searcy, K. E., A. I. Packman, E. R. Atwill and T. Harter. 2006. Deposition of Cryptosporidium oocysts in streambeds. Applied and Environmental Microbiology. 72:1810-1816. Sherer, B. M., J. R. Miner, J. A. Moore and J. C. Buckhouse. 1992. Indicator bacterial survival in stream sediments. Journal of Environmental Quality. 21:591-595. Shields, P. A. and S. R. Farrah. 2002. Characterization of virus adsorption by using DEAE-sepharose and octyl-sepharose. Applied and Environmental Microbiology. 68:3965-3968. Sischo, W. M., E. R. Atwill, L. E. Lanyon and J. George. 2000. Cryptosporidia on dairy farms and the role these farms may have in contaminating surface water supplies in the northeastern United States. Preventive Veterinary Medicine. 43:253-267. Skerrett, H. E. and C. V. Holland. 2000. The occurrence of Cryptosporidium in environmental waters in the greater Dublin area. Water Research. 34:3755-3760. Šlapeta, J. 2006a. Cryptosporidium species found in cattle: a proposal for a new species. TRENDS in Parasitology. 22:469-474. Šlapeta, J. 2006b. Response to Xiao et al.: Further debate on the description of Cryptosporidium pestis. TRENDS in Parasitology: in press. Slavin, D. 1955. Cryptosporidium meleagridis (sp. nov.). Journal of Comparative Pathology. 65:262-270. Slifko, T. R., A. Coulliette, D. E. Huffman and J. B. Rose. 2000. Impact of purification procedures on the viability and infectivity of Cryptosporidium oocysts. Water Science and Technology. 41:23-29. Smets, B. F., D. Grasso, M. A. Engwall and B. J. Machinist. 1999. Surface physicochemical properties of Pseudomonas fluorescens and impact on adhesion and transport through porous media. Colloids and Surfaces B: Biointerfaces. 14:121-139. Smith, H. V. and J. B. Rose. 1998. Waterborne cryptosporidiosis: current status. Parasitology Today. 14:14-22. Smith, J. J., T. S. Gunasekera, C. R. M. Barardi, D. Veal and G. Vesey. 2004. Determination of Cryptosporidium parvum oocyst viability by fluorescence in situ hybridization using a ribosomal RNA-directed probe. Journal of Applied Microbiology. 96:409-417. Smith, H. V., R. A. Nichols, M. Mallon, A. Macleod, A. Tait, W. J. Reilly, L. M. Browning, D. Gray, S. W. Reid and J. M. Wastling. 2005. Natural Cryptosporidium hominis infections in Scottish cattle. The Veterinary Record. 156:710-711. Sréter, T. and Z. Széll. 1998. The low sedimentation speed of Cryptosporidium oocysts: a further explanation for waterborne outbreaks. The Journal of Protozoology Research. 8:58-63.

143 Starkey, S. R., S. E. Wade, S. Schaaf and H. O. Mohammed. 2005. Incidence of Cryptosporidium parvum in the dairy cattle population in a New York City watershed. Veterminary Parasitology. 131:197-205. Stenström, T. A. 1989. Bacterial hydrophobicity, an overall parameter for the measurement of adhesion potential to soil particles. Applied and Environmental Microbiology. 55:142-147. Stevik, T. K., K. Aa, G. Ausland and J. F. Hanssen. 2004. Retention and removal of pathogenic bacteria in wastewater percolating through porous media: a review. Water Research. 38:1355-1367. Stinear, T., A. Matusan, K. Hines and M. Sandery. 1996. Detection of a single viable Cryptosporidium parvum oocyst in environmental water concentrates by reverse transcription-PCR. Applied and Environmental Microbiology. 62:3385-3390. Stott, R., E. May, E. Matsushita and A. Warren. 2001. Protozoan predation as a mechanism for the removal of Cryptosporidium oocysts from wastewaters in constructed wetlands. Water Science and Technology. 44:191-198. Suresh, P. and J. E. Rehg. 1996. Comparative evaluation of several techniques for purification of Cryptosporidium parvum oocysts from rat feces. Journal of Clinical Microbiology. 34:38-40. Swanton, S. W. 1995. Modelling colloid transport in groundwater; the prediction of colloid stability and retention behaviour. Advances in Colloid and Interface Science. 54:129-208. Tanford, C. 1972. Hydrophobic free energy, micelle formation and the association of proteins with amphiphiles. Journal of Molecular Biology. 67:59-74. Thill, A., S. Moustier, J. Aziz, M. R. Wiesner and J. Y. Bottero. 2001. Flocs restructuring during aggregation: experimental evidence and numerical simulation. Journal of Colloid and Interface Science. 243:171-182. Thiriat, L., F. Sidaner and J. Schwartzbrod. 1998. Determination of Giardia cyst viability in environmental and faecal samples by immunofluorescence, fluorogenic dye staining and differential interference contrast microscopy. Letters in Applied Microbiology. 26:237-242. Thomas, F., E. Bard, M. C. Rouillier, B. Prélot and L. Mathieu. 2001. Filtration-elution of Cryptosporidium oocysts assisted by electrostatic interactions. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 195:135-142. Triffit, M. J. 1925. Observations on two new species of Coccidia parasitic in snakes. Protozoology. 1:19-26. Trimble, S. W. and A. C. Mendel. 1995. The cow as a geomorphic agent - a critical review. Geomorphology. 13:233-253. Tyrrel, S. F. and J. N. Quinton. 2003. Overland flow transport of pathogens from agricultural land receiving faecal wastes. Journal of Applied Microbiology. 94:87S- 93S. Tyzzer, E. E. 1907. A sporozoan found in the peptic glands of the common mouse. Proceedings of the Society for Experimental Biology and Medicine. 5:12-13.

144 Tyzzer, E. E. 1910. An extracellular coccidium, Cryptosporidium muris (gen. et sp. nov.), of the gastric glands of the common mouse. Journal of Medical Research. 23:487-511. Tyzzer, E. E. 1912. Cryptosporidium parvum (sp. nov.), a coccidium found in the small intestine of the common mouse. Archiv fuer Protistenkunde. 26:394-412. Uga, S., J. Matsuo, E. Kono, K. Kimura, M. Inoue, S. K. Rai and K. Ono. 2000. Prevalence of Cryptosporidium parvum infection and pattern of oocyst shedding in calves in Japan. Veterinary Parasitology. 94:27-32. Uguen, C., P. Brasseur, A. Moreno-Sabater, L. Favennec and J. J. Ballet. 1997. Evaluation of viability and infectivity of waterborne Cryptosporidium parvum oocysts. The Journal of Eukaryotic Microbiology. 44:73S. Upton, S. J. 1997. In vitro cultivations. In: “Cryptosporidium and Cryptosporidiosis.” Fayer, R. (ed.), CRC Press LLC, Boca Raton, pp. 181-207. US-EPA. 2001. Method 1623: Cryptosporidium in water by filtration/IMS/IFA. pp. http://www.epa.gov/waterscience/methods/1622.pdf, Office of Water, United Stated Environment Protection Agency, Washington D. C. US-EPA. 2005. Method 1623: Cryptosporidium and Giardia in water by filtration/IMS/IFA. pp. http://www.epa.gov/microbes/1623de1605.pdf, Office of Water, United States Environment Protection Agency, Washington D.C. van der Mei, H. C., B. van de Belt-Gritter and H. J. Busscher. 1995. Implications of microbial adhesion to hydrocarbons for evaluating cell surface hydrophobicity 2. Adhesion mechanisms. Colloids and Surfaces B: Biointerfaces. 5:117-126. van Loosdrecht, M. C. M., J. Lyklema, W. Norde, G. Schraa and A. J. Zehnder. 1987a. Electrophoretic mobility and hydrophobicity as a measure to predict the initial steps of bacterial adhesion. Applied and Environmental Microbiology. 53:1898-1901. van Loosdrecht, M. C. M., J. Lyklema, W. Norde, G. Schraa and A. J. Zehnder. 1987b. The role of bacterial cell wall hydrophobicity in adhesion. Applied and Environmental Microbiology. 53:1893-1897. van Voorthuizen, E. M., N. J. Ashbolt and A. I. Schäfer. 2001. Role of hydrophobic and electrostatic interactions for initial enteric virus retention by MF membranes. Journal of Membrane Science. 194:69-79. Vanni, M. and G. Baldi. 2002. Coagulation efficiency of colloidal particles in shear flow. Advances in Colloid and Interface Science. 97:151-175. Vergara-Castiblanco, C. A., F. Freire-Santos, A. M. Oteiza-Lopez and M. E. Ares- Mazas. 2000. Viability and infectivity of two Cryptosporidium parvum bovine isolates from different geographical location. Veterinary Parasitology. 89:261-267. Vesey, G., K. R. Griffiths, M. R. Gauci, D. Deere, K. L. Williams and D. A. Veal. 1997. Simple and rapid measurement of Cryptosporidium excystation using flow cytometry. International Journal for Parasitology. 27:1353-1359. Wade, S. E., H. O. Mohammed and S. L. Schaaf. 2000. Prevalence of Giardia sp. Cryptosporidium parvum and Cryptosporidium andersoni (syn. C. muris) [correction of Cryptosporidium parvum and Cryptosporidium muris (C. andersoni)]

145 in 109 dairy herds in five counties of southeastern New York. Veterinary Parasitology. 93:1-11. Waldman, E., S. Tzipori and J. R. Forsyth. 1986. Separation of Cryptosporidium species oocysts from feces by using a Percoll discontinuous density gradient. Journal of Clinical Microbiology. 23:199-200. Walker, S. E., S. Mostaghimi, T. A. Dillaha and F. E. Woeste. 1990. Modeling animal waste management practices: impacts on bacteria levels in runoff from agricultural lands. Transactions of the American Society of Agricultural Engineers. 33:807-817. Walker, M. J. and C. D. Montemagno. 1999. Sorption of Cryptosporidium parvum oocysts in aqueous solution to metal oxide and hydrophobic substrates. Environmental Science and Technology. 33:3134-3139. Walter, M. T., J.-Y. Parlange, M. F. Walter, X. Xin and C. A. Scott. 2001. Modeling pollutant release from a surface source during rainfall runoff. Journal of Environmental Quality. 30:151-159. Warren, K. S., R. A. Swan, U. M. Morgan-Ryan, J. A. Friend and A. Elliot. 2003. Cryptosporidium muris infection in bilbies (Macrotis lagotis). Australian Veterinary Journal. 81:739-741. Weber, R., R. T. Bryan and D. D. Juranek. 1992. Improved stool concentration procedure for detection of Cryptosporidium oocysts in fecal specimens. Journal of Clinical Microbiology. 30:2869-2873. Wynn, T. and S. Mostaghimi. 2006. The effects of vegetation and soil type on streambank erosion, southwestern Virginia, USA. Journal of the American Water Resources Association. 42:69-82. Xia, F., D. Nagrath, S. Garde and S. M. Cramer. 2004. Evaluation of selectivity changes in HIC systems using a preferential interaction based analysis. Biotechnology and Bioengineering. 87:354-363. Xiao, L., K. Alderisio, J. Limor, M. Royer and A. A. Lal. 2000. Identification of species and sources of Cryptosporidium oocysts in storm waters with a small-subunit rRNA-based diagnostic and genotyping tool. Applied and Environmental Microbiology. 66:5492-5498. Xiao, L., A. Singh, J. Limor, T. K. Graczyk, S. Gradus and A. Lal. 2001. Molecular characterization of Cryptosporidium oocysts in samples of raw surface water and wastewater. Applied and Environmental Microbiology. 67:1097-1101. Xiao, L., R. Fayer, U. Ryan and S. J. Upton. 2004. Cryptosporidium taxonomy: recent advances and implications for public health. Clinical Microbiology Reviews. 17:72- 97. Xiao, L., R. Fayer, U. Ryan and S. J. Upton. 2006a. Response to the newly proposed species Cryptosporidium pestis. TRENDS in Parasitology: in press. Xiao, L., L. Zhou, M. Santín, W. Yang and R. Fayer. 2006b. Distribution of Cryptosporidium parvum subtypes in calves in eastern United States. Parasitology Research: in press.

146 Yang, S., M. C. Healey and C. Du. 2006. Infectivity of preserved Cryptosporidium parvum oocysts for immunosuppressed adult mice. FEMS Immunology and Medical Microbiology. 13:141-145. Young, P. L. and S. J. Komisar. 2005a. Impacts of viability and purification on the specific gravity of Cryptosporidium oocysts. Water Research. 39:3349-3359. Young, P. L. and S. J. Komisar. 2005b. Settling behavior of unpurified Cryptosporidium oocysts in laboratory settling columns. Environmental Science and Technology. 39:2636-2644. Zar, J. H. 1999. Biostatistical analysis. 4th ed. Ryu, R., S. L. Snavely, N. Bauer and L Tarabokjia (eds.), Pearson Education, Upper Saddle River, New Jersey. Zaimes, G. N., R. Schultz, C. and T. M. Isenhart. 2006. Riparian land uses and precipitation influences on stream bank erosion in central Iowa. Journal of the American Water Resources Association. 42:83-97. Zarlenga, D. S. and J. M. Trout. 2004. Concentrating, purifying and detecting waterborne parasites. Veterinary Parasitology. 126:195-217. Zhou, L., H. Kassa, M. L. Tischler and L. Xiao. 2004. Host-adapted Cryptosporidium spp. in Canada geese (Branta canadensis). Applied and Environmental Microbiology. 70:4211-4215. Zita, A. and M. Hermansson. 1997. Determination of bacterial cell surface hydrophobicity of single cells in cultures and in wastewater in situ. FEMS Microbiology Letters. 152:299-306.

147 APPENDIX A

Raw particle size data (range 0.5 – 600 m) for Cryptosporidium oocysts suspended in HEPES buffer at pH values ranging from 3.3 to 9.0

Lower limit of size range (m) pH = 3.3 pH = 5.2 pH = 7.4 pH = 9.0 0.5 0 0 0 0 1.3 0.0109 0.0106 0.0090 0.0113 1.6 1.6738 1.4687 1.3422 1.6054 2.0 4.7249 4.3283 4.2967 5.0007 2.4 8.7207 8.0729 8.2325 9.5682 2.9 12.5661 11.6554 12.0644 14.3002 3.5 14.5865 13.4503 14.0262 16.2215 4.3 13.3642 12.3403 12.8206 14.8315 5.2 10.2408 9.2682 9.4137 10.9689 6.4 7.0085 6.2861 6.0842 7.0886 7.8 4.9930 4.4589 4.0760 4.7424 9.5 3.9887 3.5701 3.1386 3.6077 12 3.6237 3.3237 2.8974 2.9858 14 3.3494 3.036 2.6597 2.4780 17 2.8233 2.5122 2.0562 1.7147 21 2.0877 1.7202 1.2154 0.8334 25 1.2438 0.9297 0.4273 0.1535 31 0.5408 0.2856 0.0057 0.0004 38 0.1275 0.0016 0 0 46 0.0391 0 0 0 56 0.2100 0.0013 0 0.0001 68 0.5562 0.3665 0 0.0133 83 1.0274 0.9884 0 0.0519 101 1.3491 1.6673 0 0.0961 124 1.0404 1.6805 0 0.1342 151 0.1026 1.2124 0 0.1761 183 0.0006 0.8447 0 0.816 224 0 1.0066 0 2.4956 272 0 2.0483 0.0001 0.0353 332 0 3.4487 0.0590 0 404 0 0.0161 0.9451 0 492 0 0 14.2301 0

148 APPENDIX B

Raw particle size data (range 0.5 – 600 m) for a Cryptosporidium oocys suspension at pH 6.8 with ionic strengths varying from 0.025 to 0.46 M

Lower limit of size range (m) 0.025 M 0.035 M 0.045 M 0.064 M 0.084 M 0.10 M 0.12 M 0.17 M 0.5 0 0 0 0 0 0 0 0 1.3 0.0112 0 0 0 0 0 0.0084 0 1.6 1.4917 0.4283 0.2444 0.4870 0.3177 0.9005 0.8651 0.2268 2.0 4.2431 3.9251 3.0889 4.0516 3.5117 4.9157 4.9958 3.2213 2.4 7.9651 11.4686 10.4050 11.1186 10.8093 11.1056 11.0593 11.1157 2.9 11.5525 22.3865 20.3641 19.5439 20.0836 17.9838 17.8284 22.0480 3.5 12.7952 23.7300 24.7685 22.5284 23.8704 21.0799 21.1984 26.7575 4.3 10.8099 16.9120 17.1073 16.6752 17.2081 17.0114 17.6668 18.677 5.2 6.7271 6.9139 6.4888 7.4754 7.0798 9.3176 9.7889 7.0544 6.4 2.8807 1.3196 1.1079 1.7054 1.4094 3.1066 3.4022 1.1777 7.8 0.5825 0.0249 0.0200 0.0342 0.0270 0.0679 0.0748 0.0210 9.5 0.0053 0 0 0 0 0 0 0 12 0 0.0018 0.0004 0.0020 0.0023 0.0023 0.0006 0 14 0 0.0539 0.0122 0.0400 0.0199 0.0530 0.0007 0.0126 17 0 0.7592 0.6381 0.7086 0.6800 0.7187 0.0003 0.6802 21 0 0.9952 0.8557 0.9133 0.8952 0.9027 0 0.8837 25 0 0.8656 0.7640 0.7638 0.7493 0.7231 0.0104 0.6703 31 0 0.6592 0.6624 0.5738 0.5382 0.5162 0.5433 0.4174 38 0.0004 0.6101 0.7733 0.5980 0.5310 0.5003 0.4558 0.3916 46 0.0894 0.7390 1.0769 0.8280 0.7619 0.7236 0.6377 0.5962 56 0.5558 0.8981 1.3472 1.0403 1.0547 0.9381 0.8676 0.8107 68 0.8647 1.0030 1.5031 1.0879 1.2896 1.0023 1.0027 0.8818 83 1.3058 1.0748 1.5039 0.9773 1.3660 0.9545 1.0003 0.8330 101 1.7910 0.9774 1.3651 0.8299 1.2179 0.7911 0.5880 0.6119 123 1.3753 0.5956 1.0064 0.6419 0.7494 0.5376 0.1036 0.2119 150 0.1356 0.1333 0.5582 0.6251 0.2712 0.3692 0.0003 0.0032 183 0.0030 0.0044 0.3377 0.8793 0.1812 0.4666 0.0033 0 223 0.9734 0.3699 0.6100 1.6238 0.6766 1.0752 0.3334 0.0065 272 4.1410 1.1885 1.2548 3.0033 1.6588 2.4872 1.6580 0.5001 332 14.0240 1.5436 1.5460 1.2412 2.1977 1.7431 3.6743 1.3378 404 10.7440 0.4172 0.5859 0.0029 0.8368 0.0062 2.0222 0.7815 492 4.9325 0.0013 0.0038 0 0.0056 0 0.2095 0.0701

149

Lower limit of size range (m) 0.21 M 0.26 M 0.30 M 0.34 M 0.38 M 0.42 M 0.46 M 0.5 0 0 0 0 0 0 0 1.3 0 0 0 0 0.0127 0 0 1.6 0.2540 0.3886 0.8097 0.4745 1.2428 0.4541 0.5605 2.0 3.2843 3.7004 5.0493 4.2154 6.0867 4.1839 4.5793 2.4 11.2531 11.5105 12.5706 11.9601 13.2290 12.7484 13.4982 2.9 22.1524 22.3922 20.9697 21.9845 20.0690 23.4033 24.0718 3.5 26.6924 26.2039 24.2281 25.7895 21.9562 26.1240 26.7528 4.3 18.5856 18.8177 18.0705 18.2209 16.6244 17.4664 17.4251 5.2 6.8908 7.5982 8.4261 7.1741 8.2411 6.2479 6.2217 6.4 1.1366 1.4590 2.0742 1.3342 2.2600 0.8685 0.9210 7.8 0.0202 0.0275 0.0424 0.0249 0.0475 0.0141 0 9.5 0 0 0 0 0 0 0 12 0 0 0 0.0002 0.0005 0 0 14 0.0123 0 0 0.0150 0.0054 0.0111 0 17 0.6652 0 0 0.0005 0.0003 0.5926 0 21 0.8797 0 0 0 0 0.7417 0 25 0.6939 0.0135 0 0.0056 0.0069 0.5102 0 31 0.4034 0.7001 0 0.3022 0.3387 0.1768 0 38 0.2902 0.5546 0.0060 0.3239 0.1923 0.0018 0 46 0.4241 0.6990 0.3398 0.5958 0.2811 0.0496 0.0015 56 0.6697 0.8737 0.8138 0.8044 0.5142 0.3314 0.1872 68 0.8756 0.9132 1.0921 0.8913 0.7523 0.6760 0.7714 83 0.9316 0.8433 1.1213 0.9768 1.0394 0.7598 1.3627 101 0.7149 0.5840 0.7862 1.0367 1.2406 0.5483 1.6410 123 0.2464 0.1849 0.2415 0.9533 1.1381 0.1994 1.3365 150 0.0035 0.0024 0.0031 0.7646 0.7975 0.0033 0.6341 183 0 0 0 0.6455 0.6359 0 0.0985 223 0.0138 0.0113 0.0085 0.7125 1.0143 0.0046 0.0008 272 0.7524 0.6691 0.6737 0.7907 1.8283 0.5424 0 332 1.5197 1.4308 1.8328 0.0030 0.4446 1.8792 0 404 0.6077 0.4285 0.8022 0 0.0001 1.3341 0 492 0.0265 0 0.0384 0 0 0.1271 0

150 APPENDIX C

Raw particle size data (range 0.5 – 600 m) measured in duplicate for Cryptosporidium oocysts suspended at pH 3.4 in ionic strengths varying from 0.002 to 3.57 M

Lower limit of size 0.0002 M 0.093M 0.185 M 0.28 M range (m) 1 2 1 2 1 2 1 2 0.5 0 0 0 0 0 0 0 0 1.3 0.0013 0.0012 0.0014 0.0014 0.0012 0.0013 0.0010 0.0012 1.6 0.0312 0.0336 0.0370 0.0347 0.0439 0.0708 0.0831 0.0596 2.0 1.3565 1.3487 1.5571 1.4776 1.5070 2.1154 2.0110 1.8488 2.4 8.1474 8.0477 8.4352 8.2995 8.2467 9.8313 10.3468 9.1680 2.9 25.4529 25.1453 24.6355 24.7717 23.8352 24.5144 23.8343 24.5184 3.5 38.3057 38.3868 36.3588 36.8169 37.2918 33.8004 33.8845 34.8929 4.3 21.9021 22.072 22.9965 23.1435 22.5810 22.0166 22.2388 22.5119 5.2 4.4336 4.6064 5.4401 4.9547 5.8333 6.7536 6.6897 6.2638 6.4 0.3694 0.3583 0.5384 0.500 0.6598 0.8961 0.9108 0.7354 7.8 0 0 0 0 0 0 0 0 9.5 0 0 0 0 0 0 0 0 12 0 0 0 0 0 0 0 0 14 0 0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 21 0 0 0 0 0 0 0 0 25 0 0 0 0 0 0 0 0 31 0 0 0 0 0 0 0 0 38 0 0 0 0 0 0 0 0 46 0 0 0 0 0 0 0 0 56 0 0 0 0 0 0 0 0 68 0 0 0 0 0 0 0 0 83 0 0 0 0 0 0 0 0 101 0 0 0 0 0 0 0 0 123 0 0 0 0 0 0 0 0 150 0 0 0 0 0 0 0 0 183 0 0 0 0 0 0 0 0 223 0 0 0 0 0 0 0 0 272 0 0 0 0 0 0 0 0 332 0 0 0 0 0 0 0 0 404 0 0 0 0 0 0 0 0 492 0 0 0 0 0 0 0 0

151

Lower limit of size 0.37 M 0.45 M 0.88 M 1.28 M range (m) 1 2 1 2 1 2 1 2 0.5 0 0 0 0 0 0 0 0 1.3 0.0007 0.0006 0.0002 0 0.0063 0.0046 0.0065 0.0072 1.6 0.1074 0.1144 0.3746 0.3660 1.3139 1.2052 1.5907 1.6327 2.0 2.3347 2.3391 3.7586 3.8248 5.6921 5.6958 5.8743 5.8497 2.4 10.0962 10.0568 11.6160 11.7409 12.5976 12.5118 12.0507 12.5055 2.9 24.1942 24.3258 22.9520 22.6836 19.8488 19.9763 17.8094 17.7986 3.5 32.8136 32.8664 28.2052 28.1583 23.3389 23.3172 20.4556 20.5106 4.3 22.1962 22.1927 21.2472 21.3857 19.5810 19.5164 17.5810 17.6424 5.2 7.1148 7.0348 9.3128 9.3192 11.4462 11.4809 11.9142 11.3004 6.4 1.1184 1.0466 2.3331 2.3205 4.7575 4.7962 5.6523 5.7203 7.8 0.0237 0.0228 0.1986 0.1992 1.3037 1.3554 2.6109 2.6545 9.5 0 0 0.0017 0.0018 0.1140 0.1401 1.4527 1.4813 12 0 0 0 0 0.0001 0.0002 1.2326 1.2279 14 0 0 0 0 0 0 0.9872 0.9476 17 0 0 0 0 0 0 0.5493 0.5245 21 0 0 0 0 0 0 0.1933 0.1920 25 0 0 0 0 0 0 0.0358 0.0048 31 0 0 0 0 0 0 0.0036 0 38 0 0 0 0 0 0 0 0.0003 46 0 0 0 0 0 0 0 0 56 0 0 0 0 0 0 0 0 68 0 0 0 0 0 0 0 0 83 0 0 0 0 0 0 0 0 101 0 0 0 0 0 0 0 0 123 0 0 0 0 0 0 0 0 150 0 0 0 0 0 0 0 0 183 0 0 0 0 0 0 0 0 223 0 0 0 0 0 0 0 0 272 0 0 0 0 0 0 0 0 332 0 0 0 0 0 0 0 0 404 0 0 0 0 0 0 0 0 492 0 0 0 0 0 0 0 0

152

Lower limit of size 1.66 M 2.02 M 2.37 M 2.69 M range (m) 1 2 1 2 1 2 1 2 0.5 0 0 0.0036 0.0034 0.0060 0.0059 0.0871 0.0812 1.3 0.0137 0.0147 0.5598 0.5626 0.6106 0.5997 0.8219 0.7373 1.6 1.8676 1.9330 2.0667 2.0755 1.3100 1.2882 1.4956 1.3284 2.0 5.4091 5.3880 4.0100 3.8776 1.9836 1.9403 1.9544 1.7148 2.4 9.9674 9.9531 6.0620 6.0169 2.5218 2.4752 0.0389 0.0341 2.9 14.2299 14.1688 7.6057 7.3077 2.8124 2.7289 0 0 3.5 16.0589 15.9375 7.9547 7.8147 2.7698 2.6804 0.0421 0.0344 4.3 14.1258 14.0197 7.0492 6.8783 2.4719 2.3852 1.5745 1.4052 5.2 9.8850 9.8613 5.6959 5.6359 1.0677 1.0137 0.5817 1.0293 6.4 8.3976 6.4203 5.0605 5.0798 1.0948 1.0407 0.5545 0.9766 7.8 2.8855 4.9151 5.7267 5.7484 1.5211 1.4622 0.7372 1.1435 9.5 4.5649 4.6871 7.2619 7.2641 2.4001 2.3176 1.0076 1.4875 12 4.5720 4.6289 8.9643 8.8981 3.9387 3.8576 1.5184 2.0278 14 3.8765 3.9135 9.7257 9.6986 6.2402 6.1512 2.2852 2.8812 17 2.5171 2.5316 9.0042 9.0848 9.3611 9.2472 3.6469 4.2672 21 1.1623 1.1575 6.8058 7.0214 12.7941 12.5901 5.8324 6.4128 25 0.3770 0.3823 4.0209 4.4050 15.0622 15.3090 9.2516 9.2754 31 0.0775 0.0760 1.7664 1.919 14.3633 14.3903 13.0137 13.3204 38 0.0114 0.0111 0.5420 0.5857 10.2468 10.5697 16.1029 14.8406 46 0.0008 0.0008 0.1056 0.1132 5.4174 5.7742 15.7741 14.9251 56 0 0 0.0085 0.0092 1.6684 1.8050 12.414 11.2825 68 0 0 0 0 0.3048 0.3289 7.2331 6.6384 83 0 0 0 0 0.0313 0.0360 3.0168 3.0081 101 0 0 0 0 0.0019 0.0028 0.8949 0.9539 123 0 0 0 0 0.0004 0 0.1313 0.1807 150 0 0 0 0 0 0 0.0191 0.0137 183 0 0 0 0 0 0 0.0015 0 223 0 0 0 0 0 0 0 0 272 0 0 0 0 0 0 0 0 332 0 0 0 0 0 0 0 0 404 0 0 0 0 0 0 0 0 492 0 0 0 0 0 0 0 0

153

Lower limit of size 3.00 M 3.29 M 3.57 M range (m) 1 2 1 2 1 2 0.5 0.0232 0.0229 0.0204 0.0301 0 0 1.3 0.1428 0.1533 0.1408 0.1952 0 0 1.6 0.2419 0.2740 0.2433 0.3262 0 0 2.0 0.3107 0.3393 0.3135 0.0050 0 0 2.4 0.3381 0.3730 0.3450 0 0 0 2.9 0.3358 0.3741 0.3371 0.0022 0 0 3.5 0.3113 0.3341 0.3014 0.3147 0 0 4.3 0.2897 0.3010 0.2669 0.2229 0 0.0027 5.2 0.2938 0.2975 0.2657 0.2137 0.0020 0.2905 6.4 0.3436 0.3518 0.3260 0.2391 0.2774 0.2461 7.8 0.4437 0.4800 0.4713 0.3194 0.2384 0.2492 9.5 0.5964 0.6801 0.7045 0.4570 0.2857 0.3151 12 0.8250 0.9705 1.0572 0.6536 0.3362 0.3950 14 1.1900 1.3947 1.5867 0.9440 0.4597 0.5544 17 1.7964 2.0152 2.3846 1.3868 0.6582 0.8002 21 2.6271 2.9037 3.4981 2.0241 0.9650 1.1839 25 3.7327 4.0511 4.8938 2.8323 1.3611 1.6698 31 4.9519 5.3432 6.4393 3.7592 1.8044 2.2090 38 5.8828 6.5079 7.8310 4.5969 2.1633 2.7402 46 6.1911 7.0707 8.6052 5.0435 2.2611 2.8375 56 5.7990 7.0791 8.3464 4.8616 2.0287 2.6274 68 5.0747 6.1663 7.3770 4.2706 1.6551 2.2364 83 4.5142 5.4571 6.2203 3.6170 1.4156 1.9175 101 4.0886 4.7989 5.0738 3.0687 1.3911 1.7804 123 3.6241 3.9181 4.0354 2.7152 1.5460 1.8079 150 2.9477 2.9444 3.1302 2.5791 1.8503 1.9944 183 2.1241 2.0499 2.1320 2.6725 2.3105 2.3396 223 1.7946 1.7937 1.2226 2.9982 3.5917 3.3894 272 2.8793 2.6671 1.0430 4.3536 6.2725 5.7564 332 5.7297 5.0520 2.3768 7.3289 11.1367 9.9152 404 10.4696 8.7406 6.3729 13.1734 19.1177 17.4610 492 20.0863 15.0947 12.6377 24.7951 36.8714 35.2807

154 APPENDIX D

Raw particle size data (range 0.2 – 180 m) for Cryptosporidium oocysts at pH 2.7 in various ionic strength solutions with and without stirring of the oocyst suspension

Lower limit of size 0.002 M, stirring range (m) 0 min 5 min 10 min 15 min 30 min 45 min 60 min 75 min 0.20 0 0 0 0 0 0 0 0 0.48 0 0 0 0 0 0 0 0 0.59 0 0 0 0 0 0 0 0 0.71 0 0 0 0 0 0 0 0 0.86 0 0 0 0 0 0 0 0 1.04 0 0 0 0 0 0 0 0 1.26 0.0002 0.0002 0 0.0002 0 0.0002 0 0 1.52 0.0402 0.0285 0.0067 0.0165 0.0053 0.0095 0.0025 0.0058 1.84 1.1513 0.9951 0.7316 0.8929 0.5594 0.7600 0.4864 0.6644 2.23 4.8007 4.5207 4.7836 4.6714 3.5210 4.2921 3.7038 4.1636 2.7 21.9901 21.4237 21.3283 21.1535 21.8721 20.9193 20.5768 20.3762 3.27 32.7886 33.6855 32.1353 32.5619 34.5354 32.8624 33.9607 32.2798 3.95 18.2376 18.8144 19.1738 19.237 18.3226 18.5386 18.3845 18.2750 4.79 4.3899 4.2020 5.6331 4.6454 3.8137 4.4003 3.8854 4.3950 5.79 0.8355 0.8758 0.8518 0.8670 0.6987 0.8459 0.7020 0.9027 7.01 0.0023 0.0053 0.0163 0.0018 0.0022 0.0031 0.0022 0.0074 8.48 0 0 0 0 0 0 0 0 10.27 0 0 0 0 0 0 0 0 12.43 0.0008 0.0008 0.0008 0.0008 0.0009 0.0009 0.0015 0.0014 15.05 0.2897 0.2647 0.2896 0.2703 0.2928 0.3101 0.4589 0.4734 18.21 0.6123 0.5559 0.6377 0.5517 0.5706 0.5956 0.8093 0.8793 22.04 0.6797 0.6252 0.7221 0.6458 0.6827 0.6765 0.8522 0.8833 26.68 0.7135 0.6700 0.7204 0.7380 0.7201 0.6964 0.8188 0.8003 32.29 0.8741 0.7158 0.6785 0.7793 0.7376 0.7720 0.8391 0.7905 39.08 0.8461 0.8305 0.7100 0.9180 0.8730 0.8865 0.8186 0.7698 47.3 0.7804 0.8612 0.6690 0.9247 0.9180 0.9091 0.6789 0.6961 57.25 0.4389 0.5490 0.4225 0.5983 0.5914 0.5956 0.3491 0.3450 69.3 0.0449 0.1169 0.0808 0.1162 0.1032 0.2020 0.0932 0.1323 83.87 0.1574 0.1603 0.1202 0.1233 0.1427 0.2962 0.3566 0.4851 101.52 0.7652 0.6715 0.7775 0.6168 0.6417 0.8529 1.0736 1.2228 122.87 1.8475 1.6725 2.2213 1.5952 1.6812 1.9292 2.2885 2.5089 148.72 7.7132 7.7547 7.2892 8.0739 8.7136 8.6455 8.8577 8.9419

155

Lower limit of size 0.002 M, stirring 0.002 M, no stirring range (m) 90 min 105 min 120 min 15 min 30 min 45 min 60 min 75 min 0.20 0 0 0 0 0 0 0 0 0.48 0 0 0 0 0 0 0 0 0.59 0 0 0 0 0 0 0 0 0.71 0 0 0 0 0 0 0 0 0.86 0 0 0 0 0 0 0 0 1.04 0 0 0 0 0 0 0 0 1.26 0 0 0 0 0 0 0 0 1.52 0 0.0013 0 0 0.0016 0.0009 0 0.0014 1.84 0.3285 0.5572 0.4883 0.3189 0.1731 0.1300 0.2329 0.1535 2.23 3.0990 3.7460 3.5021 3.3144 2.3568 2.0067 2.4542 2.1028 2.7 21.6562 19.6455 18.7249 17.9224 17.1873 16.5305 15.4805 14.1202 3.27 30.8504 31.8850 32.4776 32.0906 32.9003 31.1321 30.2694 32.3089 3.95 19.0587 18.9921 19.5410 19.9061 20.5377 21.4783 22.0194 20.3182 4.79 4.9649 4.6003 4.5983 4.8327 4.6259 5.4154 4.9613 6.3460 5.79 0.8342 0.9024 0.9317 0.9294 0.7634 0.9540 1.0622 1.0338 7.01 0.0172 0.0106 0.0153 0.0121 0.0075 0.0277 0.0462 0.0352 8.48 0 0 0 0 0.0004 0 0 0 10.27 0 0 0 0 0 0 0 0 12.43 0.0012 0.0017 0.0015 0.0015 0.0016 0.0010 0.0013 0.0010 15.05 0.3889 0.5151 0.4808 0.5031 0.5192 0.3484 0.4354 0.3500 18.21 0.7488 0.8574 0.8712 0.9684 0.9387 0.6999 0.8780 0.7110 22.04 0.7808 0.8847 0.8684 0.9327 0.9778 0.8102 0.9634 0.8226 26.68 0.7492 0.8298 0.7837 0.8469 0.8700 0.8202 0.9634 0.8668 32.29 0.7598 0.7625 0.7215 0.7412 0.7591 0.8189 0.9343 0.8426 39.08 0.7720 0.6900 0.6638 0.6446 0.6584 0.8178 0.8462 0.8457 47.3 0.6779 0.5450 0.5187 0.4869 0.5208 0.7344 0.6530 0.7413 57.25 0.4251 0.3283 0.3675 0.2955 0.3379 0.5409 0.3404 0.5005 69.3 0.2514 0.2196 0.5215 0.2186 0.2412 0.4419 0.1783 0.3152 83.87 0.6106 0.7699 0.8693 0.8289 0.8560 0.8669 0.8330 0.7976 101.52 1.5428 1.5627 1.7850 1.6700 1.7779 1.8752 1.8129 1.9769 122.87 3.2013 2.8676 2.8830 3.2083 3.3434 3.6599 3.6413 3.9906 148.72 8.2812 8.8252 8.3848 9.3269 9.6439 9.8888 10.9933 10.8181

156

Lower limit of size 0.002 M, no stirring 0.59 M, stirring range (m) 90 min 105 min 120 min 5 min 10 min 15 min 30 min 45 min 0.20 0 0 0 0 0 0 0 0 0.48 0 0 0 0 0 0 0 0 0.59 0 0 0 0 0 0 0 0 0.71 0 0 0 0 0 0 0 0 0.86 0 0 0 0 0 0 0 0 1.04 0 0 0 0 0 0 0 0 1.26 0 0 0 0 0 0 0 0 1.52 0.0004 0.0012 0.0010 0.0009 0.0005 0.0004 0.0015 0 1.84 0.1012 0.1349 0.1562 0.2251 0.1031 0.1033 0.1690 0.3698 2.23 1.7327 1.8785 1.8562 1.0996 1.7867 1.8595 2.4217 3.4446 2.7 14.3568 13.8371 13.5365 2.9769 14.0583 13.3437 15.4690 14.6411 3.27 31.1330 29.9984 29.9974 4.4134 27.4227 27.2270 24.5556 22.8545 3.95 21.4308 22.3710 22.0561 3.8590 19.3587 18.1845 16.2309 15.9215 4.79 5.7574 5.6090 5.6658 2.1413 5.6418 5.0759 5.1188 5.8970 5.79 0.9715 1.0896 0.9854 0.8751 1.2810 1.1887 1.1703 1.5463 7.01 0.0314 0.0484 0.0412 0.1452 0.0537 0.0402 0.0302 0.0648 8.48 0 0 0 0.0003 0 0 0 0 10.27 0 0 0 0 0 0 0 0.0030 12.43 0.0009 0.0009 0.0008 0 0.0017 0.0016 0.0016 0.0198 15.05 0.3166 0.3123 0.3148 0 0.5070 0.4755 0.4648 0.5028 18.21 0.6892 0.6875 0.6931 0 0.8042 0.8154 0.6635 0.6713 22.04 0.8372 0.8654 0.8745 0 0.8885 0.9777 0.7012 0.7061 26.68 0.9775 0.9445 0.8732 0 0.8389 1.1253 0.8592 0.8752 32.29 0.9201 0.9932 0.9802 0 0.8889 1.3490 1.2315 1.2900 39.08 0.9162 1.0214 1.2594 0 0.9327 1.5628 1.6695 1.7527 47.3 0.8074 0.8687 0.8856 0 0.7333 1.5196 1.7165 1.7196 57.25 0.5394 0.5219 0.5156 0 0.0776 0.9456 1.0429 0.9433 69.3 0.3959 0.3090 0.4785 0.0126 0.0083 0.3673 0.2806 0.0060 83.87 0.8965 0.8267 0.8462 0.0675 0.6202 0.9276 0.7894 0.4484 101.52 2.0583 1.9726 2.0593 3.6542 2.2481 2.1758 2.5359 2.3905 122.87 3.9948 3.8910 3.9975 11.6819 4.8378 4.5329 5.4285 5.7562 148.72 11.1347 11.8167 11.9254 68.8471 16.9069 16.2008 17.4476 18.1756

157

Lower limit of size 0.59 M, stirring 0.59 M, no stirring range (m) 60 min 75 min 90 min 105 min 120 min 15 min 30 min 45 min 0.20 0 0 0 0 0 0 0 0 0.48 0 0 0 0 0 0 0 0 0.59 0 0 0 0 0 0 0 0 0.71 0 0 0 0 0 0 0 0 0.86 0 0 0 0 0 0 0 0 1.04 0 0 0 0 0 0 0 0 1.26 0 0 0 0 0 0 0 0 1.52 0 0 0 0 0 0 0 0 1.84 0.3165 0.3708 0.4237 0.4245 0.4312 0.4278 0.4402 0.3386 2.23 3.3403 3.4603 3.6827 3.6930 3.7240 3.7050 3.6859 3.3552 2.7 15.4091 15.5392 15.6195 15.5411 15.4872 15.5457 15.5511 15.8794 3.27 24.3983 23.5116 23.1185 23.0891 22.9891 22.8378 22.524 23.6063 3.95 16.0371 16.0454 15.6231 15.6252 15.6096 14.8158 14.9685 14.5819 4.79 5.5086 5.5276 5.6477 5.6696 5.5843 5.3669 5.3632 4.8966 5.79 1.3708 1.4005 1.4205 1.4077 1.3811 1.3378 1.3095 1.1710 7.01 0.0480 0.0459 0.0438 0.0426 0.0404 0.0151 0.0344 0.0103 8.48 0 0 0 0 0 0 0 0 10.27 0.0006 0.0020 0.0032 0.0029 0.0034 0.0017 0.0024 0.0010 12.43 0.1662 0.2157 0.2686 0.2599 0.2775 0.1972 0.2419 0.1608 15.05 0.6717 0.7159 0.7975 0.7915 0.8083 0.7234 0.7723 0.6541 18.21 0.7707 0.7933 0.8767 0.8729 0.9081 0.8582 0.8374 0.7823 22.04 0.6263 0.6152 0.6808 0.7167 0.7402 0.7488 0.6590 0.6506 26.68 0.5238 0.5416 0.5579 0.5745 0.5885 0.6456 0.5784 0.5787 32.29 0.7159 0.7550 0.7109 0.7307 0.7405 0.7779 0.7688 0.7529 39.08 1.1792 1.2478 1.1390 1.1189 1.1347 1.1355 1.1955 1.1883 47.3 1.6039 1.6667 1.4515 1.4423 1.5158 1.5232 1.5807 1.6428 57.25 1.4933 1.4437 1.3546 1.3110 1.3798 1.3923 1.4843 1.6442 69.3 0.8780 0.7296 0.9370 0.8554 0.9856 1.0131 1.0225 1.2347 83.87 0.9831 0.8376 1.2662 1.2036 1.3028 1.3799 1.2992 1.4293 101.52 2.0607 2.058 2.4984 2.4878 2.5538 2.6620 2.6379 2.5929 122.87 4.6181 4.9002 5.4306 5.3759 5.4964 5.9081 5.4542 5.1403 148.72 17.2797 17.5765 16.4473 16.7633 16.3175 16.9812 17.5887 17.708

158

Lower limit of size 0.59 M, no stirring 2.70 M stirring range (m) 60 min 75 min 90 min 105 min 120 min 135 min 5 min 10 min 0.20 0 0 0 0 0 0 0 0 0.48 0 0 0 0 0 0 0 0 0.59 0 0 0 0 0 0 0 0 0.71 0 0 0 0 0 0 0 0 0.86 0 0 0 0 0 0 0 0 1.04 0 0 0 0 0 0 0 0 1.26 0 0 0 0 0 0 0 0 1.52 0 0 0 0 0 0 0.0071 0.0073 1.84 0.2978 0.3767 0.3556 0.3608 0.3837 0.3795 0.2188 0.2190 2.23 3.3765 3.4524 3.4008 3.4228 3.5021 3.4825 0.9509 0.9605 2.7 16.1906 15.3504 15.3693 15.3827 15.9812 14.7996 2.6346 2.5834 3.27 23.4943 22.4974 22.3129 22.4616 22.4604 21.3216 3.4292 3.3066 3.95 14.1749 14.2501 14.0773 14.0385 13.2160 13.5681 2.8655 2.7429 4.79 4.2853 4.9133 4.8259 4.8277 4.1751 4.8661 1.7207 1.6073 5.79 0.8900 1.1668 1.1363 1.1407 0.9201 1.1520 0.7919 0.7379 7.01 0.0035 0.0090 0.0080 0.008 0.0047 0.0072 0.2123 0.2097 8.48 0 0 0 0 0 0 0.0023 0.0024 10.27 0.0002 0.0010 0.0004 0.0011 0.0014 0.0004 0 0 12.43 0.1386 0.1657 0.1354 0.1609 0.1767 0.1333 0 0 15.05 0.6618 0.6565 0.6079 0.6332 0.6542 0.5919 0 0 18.21 0.8151 0.7442 0.7359 0.7484 0.7110 0.6936 0 0 22.04 0.7223 0.6314 0.6385 0.6268 0.5790 0.5958 0 0 26.68 0.6636 0.5936 0.6098 0.5676 0.5477 0.5912 0 0 32.29 0.8274 0.8535 0.8303 0.7483 0.8123 0.8528 0 0 39.08 1.2113 1.2932 1.3323 1.1992 1.3915 1.3578 0 0 47.3 1.5554 1.7009 1.7106 1.6655 1.8439 1.8713 0 0 57.25 1.4507 1.5793 1.6674 1.7471 1.8843 1.7370 0 0 69.3 1.0483 0.9300 0.9453 1.4427 1.3051 0.9862 0.0148 0.0001 83.87 1.3925 1.2447 1.2330 1.6691 1.3470 1.2743 0.1704 0.0470 101.52 2.9494 2.6575 2.4462 2.9080 2.4203 2.6599 4.6876 3.9290 122.87 6.1821 5.5073 5.1778 5.5814 5.0252 5.7373 13.1090 11.5176 148.72 17.6684 19.4251 20.4431 18.6579 20.657 21.3403 69.1849 72.1292

159

Lower limit of 2.70 M, no stirring 2.70 M, no stirring size range (m) 15 min 30 min 45 min 15 min 30 min 45 min 0.20 0 0 0 0 0 0 0.48 0 0 0 0 0 0 0.59 0 0 0 0 0 0 0.71 0 0 0 0 0 0 0.86 0 0 0 0 0 0 1.04 0 0 0 0 0 0 1.26 0.0001 0.0002 0.0002 0.0001 0.0001 0.0001 1.52 0.0056 0.0175 0.0230 0.0548 0.0701 0.0701 1.84 0.1090 0.9655 1.0391 1.5045 1.6213 1.6213 2.23 0.4184 6.0954 6.2659 7.2880 7.7101 7.7101 2.7 1.0701 20.9618 20.8887 21.1317 21.5733 21.5733 3.27 1.1669 27.1483 26.5902 25.895 26.1695 26.1695 3.95 0.8281 19.8325 19.6397 19.4624 19.4387 19.4387 4.79 0.3999 9.1250 8.9108 9.5812 9.9617 9.9617 5.79 0.1475 2.7356 2.6822 3.1344 3.2699 3.2699 7.01 0.0370 0.2041 0.1945 0.2665 0.2590 0.2590 8.48 0.0047 0.0018 0.0022 0.0002 0 0 10.27 0 0.3962 0.4095 0.2919 0.1886 0.1886 12.43 0 1.9941 2.0134 1.8690 1.6796 1.6796 15.05 0 2.9928 3.1307 2.8571 2.5917 2.5917 18.21 0 2.8797 3.0016 2.7378 2.4382 2.4382 22.04 0 2.0451 2.2159 1.9154 1.6153 1.6153 26.68 0 1.1816 1.3474 1.0509 0.8343 0.8343 32.29 0 0.6373 0.7598 0.5084 0.3376 0.3376 39.08 0 0.4083 0.4481 0.2579 0.1438 0.1438 47.3 0 0.2458 0.2797 0.1383 0.0714 0.0714 57.25 0 0.1086 0.1282 0.0476 0.0229 0.0229 69.3 0.0974 0.0221 0.0284 0.0065 0.0027 0.0027 83.87 4.8863 0.0007 0.0009 0.0001 0 0 101.52 11.3552 0 0 0 0 0 122.87 19.1786 0 0 0 0 0 148.72 60.2952 0 0 0 0 0

160 APPENDIX E

Raw particle size data (range 0.5 to 600 m) measured in duplicate for fractions collected from five settling columns containing soil. Column A was sampled at each of five time points (0, 26, 88, 300 and 900 minutes) and columns B to E were sacrifical columns sampled at one time point only.

Lower limit of size range (m) A 0 A 0 A 26 A 26 B 26 B 26 A 88 A 88 0.5 0.2246 0.2677 0.8074 0.8570 0.8483 0.8548 0.6952 0.6805 1.3 0.9982 1.1813 2.8095 2.8727 3.0477 3.0514 3.3257 3.3045 1.6 1.5773 1.8450 4.2806 4.3227 4.6189 4.6400 5.4661 5.4598 2.0 1.8789 2.1891 5.0125 5.0031 5.4708 5.4791 6.9424 7.1414 2.4 1.9724 2.2753 5.2543 5.2168 5.9482 5.8160 7.8974 7.8430 2.9 1.9801 2.2586 5.5258 5.4846 6.1815 6.1278 8.7425 8.5466 3.5 2.0489 2.3328 6.2287 6.0691 6.9233 6.8445 9.4067 9.2848 4.3 2.2620 2.5864 7.4291 7.3202 8.2930 8.0876 10.1716 10.0111 5.2 2.6779 3.0685 9.0094 8.9302 9.7791 9.7862 10.6081 10.4479 6.4 3.2330 3.7537 10.3008 10.3396 10.9239 10.6226 10.0591 9.8558 7.8 3.8422 4.4533 10.5533 10.5478 10.8458 10.9491 8.2897 8.1247 9.5 4.6910 5.2606 9.5865 9.5736 9.4851 9.1860 5.8444 5.7697 12 5.2779 6.0266 7.5178 7.5740 6.9975 6.8278 3.4403 3.4034 14 5.9789 6.6972 5.1111 5.1150 4.3523 4.2848 1.6719 1.6630 17 6.6082 7.2672 2.9711 3.0020 2.1902 2.2028 0.7189 0.7208 21 7.0844 7.4604 1.5274 1.5576 0.8661 0.9144 0.4120 0.4141 25 7.2392 7.3373 0.8287 0.8526 0.3628 0.3944 0.3773 0.3849 31 7.0478 6.6636 0.6419 0.6478 0.2886 0.3495 0.3266 0.3184 38 6.4635 5.7346 0.6791 0.6784 0.3934 0.4715 0.2365 0.1905 46 5.5234 4.6463 0.6869 0.6937 0.4479 0.5364 0.1563 0.1133 56 4.4376 3.5257 0.5437 0.5742 0.3199 0.4696 0.0025 0.1473 68 3.3613 2.5321 0.2959 0.3512 0.1342 0.2539 0 0.2584 83 2.5142 1.8548 0.0730 0.1402 0.0027 0.0780 0 0.3586 101 1.9943 1.4951 0.0004 0.0014 0 0.0006 0 0.2747 124 1.5758 1.3855 0.0004 0 0 0 0 0.0169 151 1.2891 1.3344 0.0773 0.0004 0.0010 0 0 0 183 0.8628 0.9503 0.3919 0.0892 0.1887 0.0009 0.0052 0 224 0.2129 0.0172 0.8528 0.2784 0.5037 0.0652 0.7718 0.0048 272 0.0011 0 1.0008 0.4347 0.5835 0.1845 4.4167 0.3196 332 0.0556 0 0.0021 0.5250 0.0017 0.3438 0.0152 1.1041 404 4.0335 0.0788 0.0007 0.5568 0 0.5696 0 2.0609 492 1.0521 3.5208 0 0.3902 0 0.6071 0 1.7764

161

Lower limit of size range (m) C 88 C 88 A 300 A 300 D 300 D 300 A 900 A 900 0.5 0.7222 0.7531 1.5784 1.6025 1.4695 1.3576 0.4814 0.8333 1.3 3.6458 3.5966 6.6178 6.7430 6.7143 6.1468 1.4784 2.5706 1.6 5.7490 5.8677 10.3885 10.5810 10.4215 9.8463 2.1175 3.6696 2.0 7.2684 7.4033 12.3174 12.5394 12.5836 12.1489 2.2530 3.8880 2.4 8.2387 8.3373 12.3581 12.5822 12.8918 12.3838 2.0011 3.3459 2.9 8.9700 9.0356 11.0799 11.3361 11.8322 11.3125 1.4025 2.4142 3.5 9.7234 9.6275 8.9651 9.1711 9.9528 9.5488 0.8529 1.4623 4.3 10.3107 10.2505 6.7944 6.9555 7.7956 7.4991 0.4498 0.7706 5.2 10.5808 10.5695 4.8409 4.9571 5.6768 5.4704 0.2316 0.3975 6.4 9.8904 9.7911 3.1847 3.2358 3.7181 3.5514 0.1436 0.2508 7.8 7.8958 7.8617 1.9302 1.9109 2.1283 2.0129 0.1155 0.2071 9.5 5.3441 5.3547 1.1025 1.0285 1.0196 0.9603 0.1038 0.1877 12 2.9725 2.9205 0.6352 0.5397 0.4050 0.4035 0.0926 0.1654 14 1.2505 1.1808 0.4317 0.3729 0.2004 0.2463 0.0797 0.1403 17 0.3561 0.2984 0.4037 0.4306 0.2717 0.3227 0.0576 0.1081 21 0.1456 0.1023 0.5011 0.6131 0.4456 0.4849 0.0276 0.0647 25 0.2630 0.2366 0.6494 0.7822 0.6063 0.6178 0.0048 0.0238 31 0.4010 0.3722 0.7593 0.8400 0.6639 0.6617 0 0.0019 38 0.4178 0.3745 0.7428 0.7399 0.6036 0.6298 0 0 46 0.3164 0.2674 0.5743 0.5159 0.4375 0.4680 0 0 56 0.1595 0.1214 0.2960 0.2451 0.2378 0.2860 0 0 68 0.0296 0.0064 0.0401 0.0322 0.0717 0.1303 0 0 83 0.0001 0 0 0 0.0011 0.0268 0 0 101 0 0 0 0 0 0.0001 0 0 124 0 0 0.0017 0 0.0007 0 0 0 151 0 0 0.1604 0 0.0125 0 0 0 183 0.0001 0.0001 0.7446 0.0039 0.2379 0.0002 0 0 224 0.0243 0.0392 2.3251 0.3084 1.5337 0.0339 0.0047 0.0049 272 0.2633 0.2795 5.6476 0.9417 8.0331 0.3850 0.0579 0.0539 332 0.8805 0.8646 4.8967 2.1471 0.0333 1.4336 1.6438 1.5443 404 2.0244 2.1441 0.0335 4.6774 0 4.7472 17.5192 15.9297 492 2.1558 2.3433 0 4.1673 0 6.8836 68.881 61.9653

162

Lower limit of size range (m) E 900 E 900 0.5 3.9554 4.0158 1.3 12.1975 12.4470 1.6 17.3460 17.7538 2.0 18.3464 18.8305 2.4 15.6946 16.1515 2.9 11.2564 11.5859 3.5 6.7496 6.8765 4.3 3.4449 3.4025 5.2 1.5930 1.4706 6.4 0.8012 0.7653 7.8 0.5246 0.6882 9.5 0.4338 0.8097 12 0.4239 0.7970 14 0.5138 0.6128 17 0.6511 0.3434 21 0.7044 0.1815 25 0.6161 0.2927 31 0.5205 0.6128 38 0.6006 0.8651 46 0.8211 0.8151 56 0.9917 0.4778 68 0.9419 0.1566 83 0.6149 0.0370 101 0.2176 0.0093 124 0.0371 0.0016 151 0.0019 0.0001 183 0 0 224 0 0 272 0 0 332 0 0 404 0 0 492 0 0

163 APPENDIX F

Raw particle size data (range 0.5 to 600 m) for fractions collected from settling columns 10 cm below the surface. Each of the three columns were sampled on 5 occasions (0, 26, 88, 300 and 900 minutes)

Lower limit of 0 min size range (m) A A B B C C Mean SD 0.5 0.3254 0.3373 0.2625 0.5384 0.2936 0.3065 0.3386 0.0273 1.3 1.2130 1.2460 0.9444 2.0374 1.0937 1.1604 1.2650 0.1118 1.6 1.8956 1.9000 1.4133 3.1325 1.6718 1.7856 1.9406 0.1923 2.0 2.1787 2.2230 1.6307 3.6911 1.9776 2.1000 2.2715 0.2220 2.4 2.2465 2.2765 1.6458 3.8193 2.0073 2.1747 2.3349 0.2427 2.9 2.2261 2.2355 1.5924 3.7758 1.9927 2.1609 2.3063 0.2523 3.5 2.2914 2.2842 1.6041 3.8650 2.0182 2.2130 2.3555 0.2723 4.3 2.5419 2.5224 1.7631 4.2453 2.2274 2.4272 2.5935 0.3058 5.2 2.9989 3.0727 2.0928 4.8925 2.6259 2.8085 3.0428 0.3680 6.4 3.5357 3.5182 2.5007 5.6510 3.1263 3.2271 3.5408 0.4029 7.8 4.0158 4.0496 2.8711 6.2554 3.6205 3.6326 4.0110 0.4515 9.5 4.4931 4.5351 3.2297 6.5998 4.1434 4.1526 4.4723 0.4954 12 4.9811 5.0042 3.5592 6.7136 4.6869 4.6853 4.9022 0.5544 14 5.4809 5.4686 3.9505 6.5126 5.2679 5.3484 5.3396 0.6018 17 5.9471 5.9569 4.4568 6.1982 5.7737 6.2738 5.8400 0.6425 21 6.3909 6.4237 5.0446 5.6126 6.2627 7.2899 6.3306 0.7196 25 6.7204 6.7791 5.8066 4.9589 6.6020 8.1570 6.7401 0.7590 31 6.8085 6.8945 6.6461 4.2854 6.7048 8.5269 6.9133 0.7210 37 6.6710 6.6290 7.4551 3.6492 6.5432 8.3850 6.8167 0.7303 46 6.2941 5.9989 7.9298 3.0873 6.1108 6.9745 6.1957 0.7343 56 5.1755 5.1919 7.9735 2.5724 5.4578 5.5653 5.3573 1.0986 68 4.0873 3.9231 7.1867 2.0752 4.7288 3.7634 4.2182 1.2950 83 3.0761 2.9504 6.1665 1.7014 4.0503 2.4055 3.2508 1.3553 101 2.2308 2.1616 4.8773 1.3649 3.3425 1.4094 2.3994 1.2287 123 1.4823 1.5936 3.4804 1.0312 2.6005 0.8997 1.7124 0.9411 150 1.0839 1.2534 2.2703 0.7135 1.9293 0.6916 1.2333 0.5932 183 0.9934 1.1136 1.1020 0.4800 1.3995 0.6699 0.9183 0.2364 223 1.0705 1.0736 0.4475 0.3381 1.0149 0.5322 0.7155 0.2951 272 1.0159 0.9513 0.0961 0.2004 0.6289 0.2725 0.4910 0.4018 331 0.5266 0.4318 0.0004 0.0015 0.0958 0.0006 0.1510 0.2581 404 0.0017 0.0023 0 0 0.0016 0 0.0008 0.0009 492 0 0 0 0 0 0 0 0

164

Lower limit of 26 min size range (m) A A B B C C Mean SD 0.5 0.8001 0.8406 0.9179 0.9020 0.8278 0.8921 0.8634 0.0471 1.3 2.8901 3.0531 3.2755 3.2657 3.0657 3.2300 3.1300 0.1531 1.6 4.3918 4.6317 4.9561 5.1327 4.6955 4.9067 4.7858 0.2650 2.0 5.1287 5.4388 5.7683 5.7940 5.5205 5.7403 5.5651 0.2580 2.4 5.3002 5.6014 5.9178 5.9609 5.7270 5.9450 5.7421 0.2589 2.9 5.3220 5.6260 5.8417 5.9483 5.7992 5.9671 5.7507 0.2432 3.5 5.6288 5.9382 6.0624 6.1708 6.1162 6.2914 6.0346 0.2306 4.3 6.3631 6.6954 6.8013 6.9070 6.9311 7.1264 6.8041 0.2596 5.2 7.4907 7.8496 8.0348 8.1001 8.1066 8.3800 7.9936 0.2996 6.4 8.5711 8.9377 9.1512 9.2652 9.2904 9.4335 9.1082 0.3108 7.8 8.9230 9.2958 9.6798 9.6833 9.5264 9.7708 9.4799 0.3200 9.5 8.4298 8.7166 9.3710 9.3069 8.9481 9.1281 8.9834 0.3619 12 7.0433 7.2739 7.9923 7.9397 7.4346 7.5033 7.5312 0.3725 14 5.1938 5.3293 6.0109 6.0018 5.4071 5.3778 5.5535 0.3584 17 3.3843 3.4541 3.9199 3.8799 3.3833 3.3064 3.5547 0.2718 21 2.0433 2.0449 2.1725 2.1255 1.7722 1.7102 1.9781 0.1910 25 1.3800 1.3028 1.0111 0.9735 0.8024 0.7629 1.0388 0.2542 31 1.2971 1.0802 0.4433 0.4070 0.3785 0.3585 0.6608 0.4156 37 1.5113 1.1654 0.3172 0.3409 0.3116 0.2999 0.6577 0.5386 46 1.7707 1.2858 0.3827 0.3849 0.3762 0.3662 0.7611 0.6137 56 1.8543 1.2861 0.4392 0.4162 0.4398 0.4308 0.8111 0.6149 68 1.7206 1.0730 0.4519 0.3785 0.5111 0.4958 0.7718 0.5277 83 1.4565 0.8491 0.4148 0.3195 0.6667 0.6161 0.7205 0.4065 101 1.0942 0.6222 0.3550 0.2350 0.8471 0.7137 0.6445 0.3162 123 0.6830 0.3929 0.2344 0.1288 0.8809 0.6647 0.4975 0.2917 150 0.3042 0.1824 0.0778 0.0319 0.6417 0.4269 0.2775 0.2302 183 0.0239 0.0279 0.0017 0.0001 0.1611 0.1490 0.0606 0.0741 223 0.0002 0.0003 0 0 0.0008 0.0016 0.0005 0.0006 272 0 0 0 0 0 0 0 0 331 0 0 0 0 0.0091 0 0.0015 0.0037 404 0 0 0 0 0.4835 0 0.0806 0.1974 492 0 0 0 0 0.9376 0 0.1563 0.3828

165

Lower limit of 88 min size range (m) A A B B C C Mean SD 0.5 0.8481 0.8837 0.9960 0.9809 0.7833 0.7922 0.8807 0.0914 1.3 4.0628 4.1377 4.7751 4.8033 3.7940 3.8098 4.2305 0.4536 1.6 6.5873 6.7329 7.7784 7.9256 6.2094 6.2287 6.9104 0.7584 2.0 8.3184 8.4812 9.7324 9.9135 7.8747 7.8834 8.7006 0.9033 2.4 9.1182 9.3505 10.7010 10.9755 8.8693 8.8682 9.6471 0.9439 2.9 9.5535 9.7116 11.0492 11.4208 9.5130 9.5684 10.1361 0.8619 3.5 9.8093 9.9778 11.4087 11.4994 10.1258 10.1466 10.4946 0.7535 4.3 9.9484 10.1129 10.9766 11.3358 10.6976 10.7868 10.6430 0.5251 5.2 9.7844 9.9787 10.3724 10.5872 10.9941 11.0690 10.4643 0.5231 6.4 8.8630 9.0431 8.9080 8.7704 10.3393 10.3996 9.3872 0.7661 7.8 7.0892 7.2356 6.5605 6.1941 8.5484 8.6309 7.3765 1.0113 9.5 4.9865 5.0479 4.0310 3.5758 6.0654 6.1345 4.9735 1.0381 12 2.9374 2.9939 1.9492 1.5294 3.5456 3.5805 2.7560 0.8424 14 1.5085 1.5036 0.6632 0.4454 1.6356 1.6329 1.2315 0.5322 17 0.7567 0.7198 0.0977 0.0428 0.4683 0.4681 0.4256 0.3013 21 0.5374 0.4904 0.0005 0.0002 0.0075 0.0005 0.1728 0.2647 25 0.5813 0.5372 0 0 0.0002 0 0.1865 0.2891 31 0.5933 0.5848 0 0 0.0144 0 0.1988 0.3024 37 0.5204 0.5097 0 0 0.0884 0 0.1864 0.2569 46 0.3406 0.3461 0 0 0.1130 0 0.1333 0.1685 56 0.2195 0.2081 0 0 0.0930 0 0.0868 0.1048 68 0.2410 0.2071 0 0 0.0748 0 0.0872 0.1104 83 0.3961 0.3403 0 0 0.0772 0 0.1356 0.1835 101 0.5024 0.4303 0 0 0.0573 0 0.1650 0.2356 123 0.3863 0.3225 0 0 0.0104 0 0.1199 0.1828 150 0.1200 0.1114 0 0 0 0 0.0386 0.0598 183 0.0012 0.0012 0 0 0 0 0.0004 0.0006 223 0 0 0 0 0 0 0 0 272 0 0 0 0 0 0 0 0 331 0.0070 0 0 0 0 0 0.0012 0.0029 404 0.4939 0 0 0 0 0 0.0823 0.2016 492 0.8887 0 0 0 0 0 0.1481 0.3628

166

Lower limit of 300 min size range (m) A A B B C C Mean SD 0.5 1.8804 1.9214 1.8792 1.8992 1.7793 1.8004 1.8600 0.0568 1.3 7.6556 7.8055 7.6958 7.7869 7.5044 7.6536 7.6836 0.1091 1.6 11.9031 12.1750 12.0184 12.1283 11.7777 12.0431 12.0076 0.1468 2.0 14.0684 14.2722 14.1239 14.3561 13.9950 14.3157 14.1886 0.1465 2.4 13.8851 14.1543 14.0345 14.2540 14.0477 14.4278 14.1339 0.1900 2.9 12.2462 12.4548 12.5017 12.5700 12.6159 12.9626 12.5585 0.2359 3.5 9.8800 10.0156 9.9555 10.1332 10.3112 10.6718 10.1612 0.2921 4.3 7.4528 7.5162 7.4464 7.6208 7.9048 8.1327 7.6790 0.2800 5.2 5.3300 5.3495 5.2458 5.3578 5.6260 5.8303 5.4566 0.2236 6.4 3.5751 3.5580 3.3770 3.4510 3.6813 3.7889 3.5719 0.1496 7.8 2.2431 2.2297 2.0011 2.0233 2.1732 2.2335 2.1507 0.1102 9.5 1.3539 1.3444 1.0721 1.0949 1.1674 1.1948 1.2046 0.1208 12 0.8230 0.8183 0.6246 0.6036 0.6250 0.6152 0.6850 0.1054 14 0.5416 0.5324 0.4191 0.4106 0.3623 0.3475 0.4356 0.0832 17 0.3594 0.3464 0.3416 0.3267 0.2120 0.2159 0.3003 0.0677 21 0.1904 0.1693 0.2518 0.2377 0.1146 0.1026 0.1777 0.0616 25 0.0307 0.0119 0.1200 0.1078 0.0319 0.0135 0.0526 0.0483 31 0 0 0.0019 0.0016 0.0003 0 0.0006 0.0009 37 0 0 0 0 0 0 0 0 46 0 0 0 0 0 0 0 0 56 0 0 0 0 0 0 0 0 68 0.0012 0.0014 0.0022 0.0022 0 0 0.0012 0.0010 83 0.1063 0.1240 0.1450 0.1592 0.0005 0.0005 0.0893 0.0711 101 0.4166 0.4942 0.4679 0.5361 0.0042 0.0133 0.3221 0.2458 123 0.8061 0.9016 0.8549 0.9818 0.0679 0.1281 0.6234 0.4115 150 1.2803 1.2524 1.2693 1.2248 0.2211 0.3680 0.9360 0.4994 183 1.8711 1.4390 1.6634 1.4266 0.6488 0.8416 1.3151 0.4746 223 1.7808 0.9231 1.7004 1.0608 1.8026 1.3327 1.4334 0.3841 272 0.3192 0.1887 0.7849 0.2449 3.2938 0.9637 0.9659 1.1827 331 0.0000 0.0009 0.0016 0.0002 0.0311 0.0024 0.0060 0.0123 404 0 0 0 0 0 0 0 0 492 0 0 0 0 0 0 0 0

167

Lower limit of size 900 min range (m) A A B B C C Mean SD 0.5 4.1817 4.3331 5.4095 5.4770 4.6330 4.5955 4.7716 0.5468 1.3 12.8452 13.3269 15.2192 15.4144 13.9597 13.7814 14.0911 1.0270 1.6 18.1823 18.8135 20.6796 20.9613 19.1449 19.2215 19.5005 1.0898 2.0 18.9635 19.5795 20.3992 20.7227 19.2650 19.6157 19.7576 0.6735 2.4 15.8578 16.3220 15.7063 15.9659 15.2586 15.5083 15.7698 0.3701 2.9 10.9114 11.1856 9.4605 9.6679 9.6697 9.9809 10.1460 0.7238 3.5 6.2131 6.2385 4.2119 4.3224 4.8220 5.0705 5.1464 0.8937 4.3 2.8937 2.9105 1.2060 1.2368 1.8811 2.0172 2.0242 0.7552 5.2 1.2826 1.3024 0.1604 0.1687 0.6165 0.6903 0.7035 0.5065 6.4 0.7021 0.6975 0 0 0.2350 0.2787 0.3189 0.3169 7.8 0.4866 0.4502 0 0 0.1162 0.1498 0.2005 0.2165 9.5 0.3083 0.2740 0 0 0.0490 0.0698 0.1169 0.1382 12 0.1398 0.1168 0 0 0.0106 0.0175 0.0475 0.0634 14 0.0369 0.0279 0 0 0.0001 0.0002 0.0109 0.0169 17 0.0004 0.0003 0 0 0 0 0.0001 0.0002 21 0 0 0 0 0 0 0 0 25 0 0 0 0 0 0 0 0 31 0 0 0 0 0 0 0 0 37 0 0 0 0 0 0 0 0 46 0 0 0 0 0 0 0 0 56 0 0 0 0 0 0 0 0 68 0 0 0.0001 0.0002 0.0001 0.0001 0.0001 0.0001 83 0 0 0.0170 0.0161 0.0061 0.0185 0.0096 0.0086 101 0 0 0.0815 0.0668 0.0498 0.1538 0.0587 0.0576 123 0 0 0.1565 0.1134 0.1834 0.5433 0.1661 0.2003 150 0 0 0.2572 0.1755 0.6580 1.5553 0.4410 0.5970 183 0 0 0.7184 0.5317 2.1350 3.8371 1.2037 1.5091 223 0 0 1.5702 1.3188 1.9270 2.8656 1.2803 1.1221 272 0.0010 0.0002 1.4758 1.3264 0.0653 0.0293 0.4830 0.7131 331 0.2678 0.0209 1.3325 1.1424 0.0025 0 0.4610 0.6128 404 1.4010 0.6904 0.9628 0.6978 0.2411 0 0.6655 0.5008 492 5.3250 3.7098 0.9753 0.6737 5.0701 0 2.6257 2.3607

168