Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Viral ecology of lakes : a descriptive and ecological study of viruses that infect phytoplankton Clasen, Jessica Liz 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_spring_clasen_jessica_liz.pdf [ 7.16MB ]
Metadata
JSON: 24-1.0053315.json
JSON-LD: 24-1.0053315-ld.json
RDF/XML (Pretty): 24-1.0053315-rdf.xml
RDF/JSON: 24-1.0053315-rdf.json
Turtle: 24-1.0053315-turtle.txt
N-Triples: 24-1.0053315-rdf-ntriples.txt
Original Record: 24-1.0053315-source.json
Full Text
24-1.0053315-fulltext.txt
Citation
24-1.0053315.ris

Full Text

VIRAL ECOLOGY OF LAKES: A DESCRIPTIVE AND ECOLOGICAL STUDY OF VIRUSES THAT INFECT PHYTOPLANKTON by JESSICA LIZ CLASEN B.S., University of North Carolina at Wilmington, 1997 M.S., Arizona State University, 2001  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Oceanography)  THE UNIVERSITY OF BRITISH COLUMBIA February 2008 © Jessica Liz Clasen, 2008  Abstract  Since the 'discovery' of the high abundance of viruses in aquatic environments, it has been generally assumed that viruses in lakes are similar to those in oceans. I directly compared these two systems using a large, robust data set. Viral abundance was significantly different among the surveyed environments. The relationship between viral and bacterial abundance indicated a fundamental difference between lakes and oceans, and suggested that viruses infecting phytoplankton may be more important in lakes. Molecular techniques (PCR & DGGE) were used to document spatial and temporal variations in the richness of viruses that infect eukaryotic phytoplankton (Phycodnaviridae) in lakes at the Experimental Lakes Area (ELA). Phycodnavirus richness was highest in the eutrophic lake, and during the spring/early summer in all the lakes. Viral richness was closely associated with phytoplankton abundance and composition. As a result, richness was influenced by trophic status, while patterns of richness were affected by regional climatic conditions. Phylogenetic analysis of environmental Phycodnavirus DNA polymerase (pol) sequences indicated that freshwater Phycodnaviruses are genetically different from cultured isolates and marine environmental sequences. A genetic distance analysis indicated that pol sequences > 7 % different infected different host species. Therefore, the 20 different freshwater sequences likely infected nine different hosts. Multivariate statistics identified seven possible phytoplankton hosts, including chlorophytes, chrysophytes, diatoms and dinoflagellates.  ii  Finally, the modified dilution experiment was evaluated as an approach for estimating viral-mediated phytoplankton mortality in two lakes at the ELA. Experiments resulted in non-significant apparent growth rate regressions. While a model analysis, indicated that the method was sensitive to poorly constrained parameters such as burst size and length of the lytic cycle, making it unsuitable for estimating mortality rates in these lakes. These studies indicate that Phycodnaviridae are a genetically rich and dynamic component of lakes. Their richness is influenced by both the chemical and physical components of their environment. Although the presence of these viruses indicates that they are a source of phytoplankton mortality, the magnitude of their impact on structuring phytoplankton communities awaits methodological advances. Nonetheless, these findings support the view that viruses infecting phytoplankton are ecologically important components of lake ecosystems.  iii  Table of Contents  Abstract  ^  ii  Table of Contents ^  iv  List of Tables  ^viii  List of Figures  ^ix  Acknowledgements  ^xi  Dedication  ^xiv  Co- Authorship Statement ^  xv  Chapter one: Introduction - General viral biology and viruses in aquatic ecosystems  ^.16  1.1 Introduction ^  17  1.2 General viral biology ^  17  1.2.1 Viral composition ^ 1.2.2 Viral replication cycles  17 ^ 18  1.2.3 Viral host range ^ 1.3 Viruses in the oceans  ^  1.4 Viruses of cyanobacteria ^ 1.5 Viruses of eukaryotic phytoplankton 1.5.1 Abundance  19 19 .20 ^ 21  ^  21  1.5.2 Richness ^  ..22  1.5.3 Sequence analysis ^  ..23  1.5.4 Ecological influence of phytoplankton viruses ^ 24  iv  1.6 Viruses of freshwater eukaryotic phytoplankton.... ^  ...25  1.7 Thesis objectives ^  .26  1.8 References ^  27  Chapter two: Evidence that viral abundance across oceans and lakes is driven by different biological factors ^ 2.1 Summary ^  35 ...36  2.2 Introduction ^  37  2.3 Methods ^  38  2.4 Results ^  45  2.5 Discussion ^  54  2.6 Acknowledgments ^  64  2.7 References ^  .65  Chapter three: Spatial and temporal viral dynamics in lakes are influenced by trophic status and regional climatic conditions^ 3.1 Summary ^  72 ..73  3.2 Introduction ^  74  3.3 Methods ^  76  3.4 Results ^  ...84  3.5 Discussion ^ 3.6 Acknowledgments ^ 3.7 References ^  .95 ...105 106  Chapter four: Identifying freshwater Phycodnaviridae and their potential hosts using DNA pol sequence fragments and a genetic distance analysis ^ ...113 4.1 Summary ^  114 v  4.2 Introduction ^  .115  4.3 Methods ^  116  4.4 Results and discussion ^  125  4.5 Acknowledgments 4.6 References  ^ 139 140  ^  Chapter five: An evaluation of the utility of the modified dilution experiment to estimate viral-mediated phytoplankton mortality  ^ 145  5.1 Summary^  146  5.2 Introduction ^  .147  5.3 Methods ^  150  5.4 Results ^  .157  5.5 Discussion ^ 5.6 Acknowledgments ^  161 ...171  5.7 References ^  173  Chapter six: Conclusions - Summary, suggestions for future research directions and final thoughts ^  ..178  6.1 Summary ^  .179  6.2 Future research directions ^  181  6.2.1 Isolation of new phytoplankton viruses ^  181  6.2.2 Development of a robust and reliable way to estimate viral-mediated phytoplankton mortality ^  .181  6.2.3 Impact of climate change on viral richness and mortality ^ 182 6.3 Final thoughts ^ 6.4 References ^  .183 184 vi  Appendices ^  ..186  1^Binary matrix of AVS bands in Lake 227  187  2^Binary matrix of AVS bands in Lake 239  189  3^Presence or absence of AVS Bands in Lake 240 ^  .190  4^Temporal variation in bacterial abundance ^  192  5^Binary matrix of AVS Bands in Lake 224  .193  vii  List of Tables  Table 2.1 Biological parameters of environments ^  .46  Table 2.2 Discriminant factor analysis  52  Table 2.3 Backward step-wise multiple regression analysis of viral abundance ^  .53  Table 3.1 Physical, chemical and biological parameters of three lakes at ELA  78  Table 4.1 Phycodnaviridae isolates ^  .120  Table 4.2 Environmental samples used in phylogenic analysis ^  .122  Table 5.1 Terms used in the modified dilution experiment  .152  Table 5.2 Model parameters ^  156  Table 5.3 Regression coefficients from modified dilution experiments ^  .159  Table 5.4 Phytoplankton turnover due to grazing and viral lysis ^  .160  viii  ^  List of Figures  Figure 2.1 Map of sampling environments ^  39  Figure 2.2 A comparison of the biological variables determined within each of the three aquatic environments  47  Figure 2.3 Viral abundance as a function of other biological parameters ^49 Figure 2.4 Comparison of viral abundance and bacterial abundance with other published datasets ^  58  Figure 2.5 Comparison of slope coefficients and y-intercepts ^  60  Figure 3.1 Map of lakes sampled ^  .77  Figure 3.2 Spatial variation in viral abundance and richness ^  .85  Figure 3.3 Rarefaction curves  87  ^  Figure 3.4 Temporal variation in viral abundance ^  88  Figure 3.5 Temporal variation in viral richness ^  91  Figure 3.6 Hierarchical classification clustering analysis of viral richness patterns ^ 92 Figure 3.7 Multiple dimensional scaling plot of the eukaryotic phytoplankton community in each lake ^  .94  Figure 3.8 The relationship between AVS richness and phytoplankton abundance ...... .....99 Figure 3.9 Regional climatic conditions during the temporary change in viral richness....103 Figure 3.10 Seasonality of viral abundance and richness ^  104  Figure 4.1 Maximum likelihood tree of DNA polymerase sequences from freshwaters ^ 126 Figure 4.2 Neighbour joining tree of DNA polymerase sequences from freshwaters and other environmental samples  ^ 129  ix  Figure 4.3 Testing the representative amplification of Phycodnaviruses by AVS ^  .132  Figure 4.4 Genetic distance 'within' and 'between' Phycodnavirus groups ^  .135  Figure 4.5 Number of potential hosts ^  .136  Figure 5.1 Stylized modified dilution experiment ^  ..149  Figure 5.2 Modified dilution experiments  ..158  Figure 5.3 Model outputs  .162  Figure 5.4 Model analysis of the modified dilution experiment  164  x  Acknowledgements  No thesis is an individual effort and I would like to take this opportunity to thank some of the people who contributed to this thesis, in one way or another. First, I would like to thank my supervisor Dr. Curtis A. Suttle for his guidance, years of patience, insight and editorial skills. I have truly enjoyed my time in the Suttle lab and I feel like I have learned a huge amount. I extend a huge thanks to my committee members, Drs. Philippe Tortell and John Stockner. Along with their guidance, they offered encouragement, which meant the world to me. My research would not have been possible without the help of the students and staff at the Department of Fisheries and Oceans' Experimental Lakes Area (ELA) in Northwestern Ontario. I am deeply honoured to have been a part of ELA and I would like to extend my thanks to Mark Lyng, Dave Findlay, Dr. Mike Paterson, Stephen Page, Mamie Potter, Les Hummerston, Ian Delorme, Corben Bristow, Shelley Brule, Ken Sandilands, Justin Shead and all the others who were part of the 2003 and 2004 crews. I gratefully acknowledgement my current and past lab mates, including Drs. Andre Comeau, Alex Culley, Alice Ortmann, Emma Hambly, Andrew Lang, Janice Lawrence, Pascale Loret, Steven Short and Christian Winter and soon to be doctors, Emma Brownlee, Caroline Chenard, Matthias Fischer, Julia Gustaysen, Jessica Labonte and JerOme Payet. Thanks to Sean Bridgen, Amy Chan, Karen Reid, Cindy Short and Tanya St. John. Finally, I have had the pleasure of supervising several wonderful students during my Ph.D. including Craig Kalnin, Vesna Posarac, Johan Vande Voorde and Leanne Yeung; these students taught me as much, if not more, then I tried to teach them.  xi  Thanks to others at UBC; who as my colleagues and moral supporters, made my studies more enjoyable. These people include Dr. Claudio DiBacco, Monica Bravo, Dr. Julie Granger, Dr. Brian Hunt, Leigh Gurney, Dr. Catherine Johnson, Allyson Longmuir, Tina Lum, Dr. Jean Marcus, Dr. Laurie Marczak, Russ Markel, Dr. Jon Shurin, Mandy Toperoff and Katsky Venter. I got my first taste of research as a NSF-REU summer student in Dr. Robert Sterner's lab at the University of Minnesota in 1995. That summer fueled my love of science and I would like to thank Bob and Dr. Jotaro Urabe (Kyoto) for the opportunity. Additionally, I would like to thank the others in my academic lineage, including Dr. Joseph Pawlik (UNCW), my Master's advisor Dr. James Elser (ASU), Prof. Winfried Lampert (MPIL) and my science and math teachers at Gordon Bell high school (Winnipeg, Manitoba). These overworked and grossly underpaid inner-city teachers did their best to keep a silly girl occupied and interested; thanks to Mr. Coulter, Mrs. Clark, Mr. Koes, Mr. Lawler and Mr. Wolynec. Thanks to the Lake of the Woods Distinct Properties Owners Association who provided financial support in the form of an Environmental Research Fellowship. I am honoured to have been selected as their first recipient (2003). I am indebted to Peter Haugen of Pipestone Resort, Kenora, Ontario who kindly provided 'research' space in his office and took good care of me while I lived on the lake. My family, both immediate and extended, have always been my biggest cheerleaders and I am forever grateful for all their years of support; especially deserving of thanks are John, Joan, Lauren and Keefe Clasen. I would also like to thank Bones Clasen, my Sweet Pea, who was a constant companion and a wonderful field assistant! Additionally, I would like to acknowledge my Uncle Tom (Dr. Thomas Clasen) who helped me buy my first laptop xii  computer, and my Aunt Judy (Dr. Judith Wynnemer) who has always supported me, which included helping me with my taxes and applications, feeding me numerous Sunday dinners and making me feel less homesick while I lived in North Carolina. Finally, I would like to thank Chris and Ernie Payne. They have had to bare all of my thesis-writing stress. Thanks for being so patience and supportive, it made this marathon bearable to have my support team nearby. I hope the next one we can run together! During my degree, I received funding from the Natural Science and Engineering Research Council of Canada through a postgraduate scholarship and travel awards from the Earth and Ocean Sciences department.  Dedication  To my family, including my grandparents (Ery and Ella Wynnemer & Kay and Millard Clasen), my parents (John and Joan Clasen), my siblings (Lauren and Keefe Clasen & Lynn Niskanen) and my pets (Ginger, Bones, Tommy and Fred). This thesis is as much a labour of their love as it is mine.  xiv  Co-authorship Statement Chapter Two is in press at Freshwater Biology as: Clasen, J.L., S.M. Brigden, J.P. Payet and C.A. Suttle. Evidence that viral abundance across oceans and lakes is driven by different biological factors. Sean M. Brigden counted many of the Pacific samples used in this analysis as part of his honours project. JerOme P. Payet provided viral abundance data from the Arctic which were collected during the Canadian Arctic Shelf Exchange Study (CASES). Since neither the Pacific nor Arctic data has been published yet, S.M. Brigden and J.P. Payet were granted authorship. When chapter three is submitted to Limnology and Oceanography, it will be as: Clasen, J.L., D.L. Findlay and C.A. Suttle. Spatial and temporal viral dynamics in lakes are influenced by trophic status and regional climatic conditions. Dave L. Findlay was a staff member at the Department of Fisheries and Oceans' Freshwater Institute in Winnipeg, Manitoba, which operates the Experimental Lakes Area (ELA). As part of a long term ecological research program, D.L. Findlay determined the composition of the phytoplankton community in several lake samples. Some of these data were used in this study and since it is unpublished, D.L. Findlay was included as an author. Beyond the contributions described above, I collected, processed and analyzed the remaining data, as well as wrote all of the chapters (and associated manuscripts). As a result, I am primary author on all the chapters. Curtis A. Suttle is also an author on all four chapters/manuscripts. As my thesis supervisor, he provided not only guidance and insight but also (much needed) editorial assistance.  xv  Chapter One: Introduction - general viral biology and viruses in aquatic ecosystems  16  1.1 Introduction Viruses are small (60 to 200 nm in diameter) obligate intracellular parasites that are not capable of independent replication, but rather depend on the machinery of their host for transcription and translation of viral proteins. Despite this dependence, viruses continually influence both the ecology and the evolution of the biosphere. The paragraphs below are intended to introduce general viral biology, viruses in aquatic environments, as well as viruses that infect eukaryotic phytoplankton. In order to lay the framework for understanding the objectives of this thesis, this introduction is restricted to describing viruses that infect unicellular organisms.  1.2 General viral biology 1.2.1 Viral composition In general, viruses are composed of a genome surrounded by a protein structure called a capsid. The genome can either be double or single stranded deoxyribonucleic acid (DNA) or ribonucleic acid (RNA), and can range in size from 3.2 kb to 1.2 Mb (Levy et al., 1994; La Scola et al., 2003). Although nucleic acid composition is often used to describe different viruses, morphology can also be used. Capsid structures vary in shape but can be helical, icosahedral or polyhedral; they can be composed of multiple copies of a single protein or several different proteins woven together into a symmetrical structure. Additional structures associated with some viruses include envelopes and tails. For example, many bacteriophages (viruses that infect bacteria) have tails. Finally, a few viruses carry associated enzymes  17  and/or lipids molecules within their capsid that are necessary for successful infections (Levine, 1992).  1.2.2 Viral replication cycles Common viral replication cycles include lytic and lysogenic. Lytic viruses infect a host cell and then immediately commence the replication process that ultimately destroys the cell. On the other hand, lysogenic viruses integrate their genome into the genome of the host after entry into the cell. As the host cell replicates its genome in preparation for cell division, the integrated viral genome (called the prophage) is also copied and passed on to all the daughter cells. Eventually, an environmental or biochemical cue triggers the exclusion of the prophage and the virus enters a lytic replication cycle. The replication cycle of a lytic virus begins with the attachment of the virus to the cell wall or membrane of a host. Attachment can occur via hair-like fibers or cell surface protein recognition. Following attachment, the viral genome and any associated molecules are transferred into the host cell and immediately inhibit host cellular processes (Levy et al., 1994). The transfer mechanisms vary but can include injection, cell wall digestion and membrane fusion. Upon entry, the virus hijacks the machinery of the host to transcribe early viral mRNAs which are then translated into proteins necessary for replicating the viral genome. Replication is followed by the translation of late mRNAs into viral capsid proteins (Hurst & Lindquist, 2000; Nayak, 2000). The new viral genomes and capsids are assembled at the viral assembly area which can be located in either the host's nucleus or cytoplasm. The newly assembled viruses accumulate until the host cell lyses, which releases the particles but leaves behind dissolved organic matter (DOM) and cellular debris (Nayak, 2000). The length of the lytic cycle, which is measured from the time of attachment to cell lysis, varies 18  from minutes to days, depending upon the type of virus, as well as, the size, growth rate and internal nutrient content of the host (Nayak, 2000)  1.2.3 Viral host range Viruses are usually restricted to a small number of potential hosts; generally they infect only one species or a few very closely related species. This specificity often results in a complex evolutionary arms race between the virus and host. However, some viruses are promiscuous, jumping between host species. For example, some cyanophages (viruses that infect cyanobacteria) can infect both Prochlorococcus and Synechococcus (Sullivan et al., 2003). These promiscuous viruses are evolutionarily interesting because they allow genetic material to be exchange between the different host species.  1.3 Viruses in the oceans Early last century, seawater was thought to be devoid of viruses and bacteria because of the high concentrations of salt; in fact seawater was used as an antiseptic. When viruses were discovered in seawater, their abundance was thought to be too low to be ecologically important. However, the use of transmission electron microscopy in the late 1980s shattered this dogma; Bergh et al. (1989) and Proctor & Fuhrman (1990) discovered high abundances of viruses (> 10 6 ) in seawater. In the pelagic environment, viruses are generally an order of magnitude more abundant than bacteria. Their average abundance is 10 7 to 10 8 viruses mL-1 , but can range from < 10 4 to > 10 10 viruses mL -1 (see Fuhrman & Suttle, 1993; Suttle, 2000; Wommack & Colwell, 2000; Suttle, 2007). Studies indicate substantial temporal and spatial  19  variation in viral abundances, including by day, month, year, location and depth (reviewed in Wommack & Colwell, 2000). Viral infections have been observed in many aquatic organisms, including bacteria, phytoplankton, fish and marine mammals (Osterhaus et al., 1989; LaPatra, 1998; Paul & Kellogg, 2000; Suttle, 2000; Brussaard, 2004). However, as bacteria, cyanobacteria and eukaryotic phytoplankton are the most abundant organisms in aquatic ecosystems, they are also the greatest sources of viruses (reviewed in Paul & Kellogg, 2000; Suttle, 2000). The often observed strong association between total viral and bacterial abundances and/or chlorophyll a concentrations further supports the claim that most viruses in the ocean infect bacteria and phytoplankton (reviewed in Wommack & Colwell, 2000; Suttle, 2005; Suttle, 2007). As primary producers, phytoplankton occupy a unique position in aquatic ecosystems. As a result, viruses that infect phytoplankton are likely ecologically significant components of aquatic environments.  1.4 Viruses of cyanobacteria Cyanophages are dsDNA viruses that infect cyanobacteria and are members of the Myoviridae, Siphoviridae and Podoviridae families (Mann, 2003). There are abundant in aquatic environments (10 3 to 10 5 cyanophage mL -1 ), infecting the ecologically important cyanobacteria genera Prochlorococcus and Synechococcus (Fuhrman & Suttle, 1993; Suttle & Chan, 1993; Suttle & Chan, 1994; Sullivan et al., 2003). Phylogenetic analyses with genes that encode either structural proteins (g20, g23) or core photosynthetic proteins (psbA, psbD) indicate that these groups of viruses are genetically rich (see Mann et al., 2003; Filee et al., 2005; Short & Suttle, 2005). Experimental evidence suggests that 3 to 10 % of  20  Synechococcus cells are destroyed daily by viral lysis (Proctor & Fuhrman, 1990; Suttle,  2004), indicating that these viruses may be an important in nutrient cycling in aquatic ecosystems (Wilhelm & Suttle, 1999) and latent gene transfer (Mann, 2003). Despite this, further discussion will be limited to viruses that infected eukaryotic phytoplankton as they were the primary focus of the thesis research.  1.5 Viruses of eukaryotic phytoplankton Mayer & Taylor (1979) were among the first to describe a virus that infects a marine eukaryotic phytoplankton (Micromonas pusilla virus); later, Van Etten and colleagues (1983) discovered a group of viruses that are known to infect a freshwater Chlorella-like alga. Currently, viruses have been isolated that infect a number of eukaryotic algae, including ones that infect Chlorella-like algae, Micromonas pusilla, Chrysochromulia spp., Aureococcus anaophagefferns, Emiliania huxleyi, Phaeocystis pouchetii, Heterocapsa spp., Heterosigma akashiwo, Pyramimonas orientalis, Chaetocerus salsugineum and Rhizosolenia setigera (see  Brussaard, 2004; Shirai et al., 2006). Most of the viruses that infect these hosts are members of the viral family Phycodnaviridae, which are large (> 100 nm in diameter), icosahedral viruses with big (160 — 560 kb) dsDNA genomes (Wilson et al., 2005b). Research conducted using these virus-host systems, as well as recently developed molecular techniques have provided insight into the abundance, richness and ecological influence of viruses that infect eukaryotic phytoplankton.  1.5.1 Abundance of eukaryotic phytoplankton viruses It is difficult to determine the abundance of phytoplankton-specific viruses because viruses can not be distinguished by morphology alone. Traditionally, culture-dependent  21  techniques, such as most probable number (MPN) and plaque-forming unit (PFU) assays, have been used to determine the abundance of specific phytoplankton viruses. Despite limitations, these assays have generated interesting distribution and abundance data. For example, Cottrell & Suttle (1991) reported that Micromonas pusilla viruses were more abundant during the winter months. Additionally, a survey by Van Etten et al. (1985) found that the abundance of viruses infecting a Chlorella-like alga were highest in nutrient rich drainage ditches in the spring, and suggested that there was a correlation between viral abundance and ecosystem productivity.  1.5.2 Richness of eukaryotic phytoplankton viruses The plethora of viruses that infect the same species of phytoplankton indicate that viral richness is high. For example, there are 52 viruses that infect Chlorella-like algae currently isolated, as well as, several different Micromonas pusilla and Emiliania huxleyi viruses. However, since only a few phytoplankton viruses have been isolated, culture independent techniques are necessary to determine the true richness of Phycodnaviridae. Sequence analysis of three Phycodnaviruses, including two that infect a Chlorellalike alga (PBCV-1 and NY-2A) and one that infects Micromonas pusilla (MpV-SP1), revealed the presence of relatively conserved motifs in the DNA polymerase that flank the universally conserved YGDTDS (Asp-Asp) motif (Chen & Suttle, 1995). DNA polymerase (hereafter DNA pol) is an enzyme necessary for DNA replication and is coded by many viruses including all the nucleocytoplasmic large DNA viruses (NCLCV), which is a group that encompasses the Phycodnaviridae (Dunigan et al., 2006). Chen & Suttle (1995) and Chen et al. (1996) designed algal virus specific primers (hereafter, AVS-1 and 2) that  22  specifically target Phycodnaviridae. The primers allow Phycodnaviridae to be detected in environmental samples without the need for cultured hosts or further virus isolation. Short & Suttle, (1999; 2000) coupled PCR amplification with denaturing gradient gel electrophoresis (DGGE) to create a Phycodnaviridae community 'fingerprint'. DGGE separates dsDNA fragments based upon their individual nucleotide composition by passing them through a polyacrylamide gel that has an increasing gradient of denaturants (urea and formamide). The individual sequences separate because they denature at different concentrations, this changes their secondary structure and stops their migration through the gel creating a fingerprint. DGGE is an extremely sensitive tool that has the potential to separate sequences that differ by only one base pair (Myers et al., 1987). Therefore, DGGE `fingerprints' give an indication of both the viral richness and community composition of a sample. DGGE 'fingerprints' generated from marine environmental samples demonstrated considerable temporal and spatial variation in Phycodnavirus richness (Short & Suttle, 2002; 2003). For example, a year-long investigation of viruses at Jericho Pier, Vancouver, BC, determined that the variation in richness was partly associated with chlorophyll a concentrations and tidal height, suggesting that both the physical and biological components of the environment influenced the Phycodnavirus community (Short & Suttle, 2003).  1.5.3 Sequence analysis of eukaryotic phytoplankton viruses Phylogenic analyses of AVS amplified DNA pol sequences from environmental samples confirm that the fragments are from Phycodnaviruses (Chen et al., 1996; Short & Suttle, 2003). Additionally, sequences from the Gulf of Mexico, as well as, the Pacific and Southern Oceans were similar, suggesting that closely related Phycodnaviruses are cosmopolitan in distribution (Chen et al., 1996; Short & Suttle, 2002). 23  1.5.4 Ecological influence of phytoplankton viruses Understanding the ecological significance of eukaryotic phytoplankton viruses has been limited by methodological issues; however, several researchers have used experiments to estimate viral-mediated phytoplankton mortality, as it is the most direct affect viruses have on phytoplankton communities. For example, viruses were estimated to cause 25-100 % of Emiliania. huxleyi mortality in a bloom (Bratbak et al., 1993); whereas, 4.4 to 15 % of Micromonas spp. mortality has been attributed to viruses (Evans et al., 2003;.Cottrell & Suttle, 1995, respectively). In one of the first community level experiments conducted, Suttle et al. (1990) reported a 78 % reduction in photosynthetic rates and fluorescence with a modest addition of a concentrated viral community. By directly killing phytoplankton, viruses can affect community composition. A viral-mediated demise of a phytoplankton bloom is perhaps the most obvious example; however, viruses appear to affect composition in non-bloom conditions as well. Suttle, (1992) suggested that the decrease in  14 C  incorporation with the addition of a viral  concentrate to a natural community was the result of species-specific infections, because the response was not linearly related to the amount added. Additionally, there was an increase in chlorophyll a concentrations after 48 hours, suggesting that the non-infected phytoplankton cells grew. The kill the winner hypothesis (Thingstad, 2000) suggests that viruses promote diversity by killing the most abundant host, allowing other phytoplankton to flourish. Phytoplankton viral infections can affect nutrient cycling because, as previously mentioned nutrients are released into the dissolved or particulate pool during lytic events. Gobler et al. (1997) reported the release of carbon, nitrogen, phosphorous, selenium and iron during the lysis of infected Aureococcus. anophagefferens cells; while experiments conducted with a Chlorella-like alga, estimated that > 90 % of host cellular phosphorous was 24  released during cell lysis (Clasen, 2001). These results suggest that phytoplankton viruses are ecologically important, affecting mortality rates, community composition and nutrient cycling. A majority of these results stem from investigations conducted in marine environments. In contrast, relatively little is known about phytoplankton viruses in lakes.  1.6 Viruses of freshwater eukaryotic phytoplankton The closest examples of viruses that infect freshwater eukaryotic phytoplankton are the viruses that infect an endosymbiotic Chlorella-like alga. This green alga is normally found in a symbiotic relationship with Paramecium spp. or Hydra spp. (Van Etten & Meints, 1999). Several (52) of theses viruses have been isolated and a few genomes have been completely sequenced (see Dunigan et al., 2006; Fitzgerald et al., 2007). Although there is much known about the biology of these viruses (see Van Etten et al., 2002; Van Etten, 2003; Kang et al., 2005) very little is known about their ecology. In fact the only ecological study is one in which Chlorella-like viruses were found to exist in lakes, rivers and ditches in a few states (Van Etten et al., 1985). However, it is difficult to extract any potential ecological significance from viruses that infect an endosymbiotic alga. Maranger & Bird (1995) investigated the relationship between viral abundance and other biological variables in 22 Quebec lakes. Unlike most oceanic environments, these authors found a significant relationship between viral abundance and chlorophyll a concentrations, which they suggested was evidence that viruses infecting phytoplankton are more important in freshwater than marine environments. Given this suggestion and the lack of information about freshwater phytoplankton viruses, this thesis sought to investigate  25  viruses that infect freshwater eukaryotic phytoplankton in order to understand the role these viruses occupy in lake ecosystem dynamics.  1.7 Thesis objectives Overall, I am interested in the influence viruses have on structuring phytoplankton communities in lakes. Given the relative infancy of the field and the paucity of data, my thesis was designed to lay some of the initial framework necessary to address this issue. The objectives of my thesis were: 1.  To compare viral abundance in freshwater and marine environments using a large and robust data set. The aim was to explore the relationship between viral abundances and other biological parameters in both environments and determine if they were the same.  2.  To document the spatial and temporal variations in Phycodnaviridae richness in three different lakes and identify the variables that influence this variation.  3.  To document the genetic identity of some of the Phycodnaviruses present in lakes and determine their potential hosts using genetic distance and a multivariate statistical analysis.  4.  To theoretically and empirically evaluate an experimental technique used to estimate viral-mediated phytoplankton mortality rates.  26  1.8 References Bergh, 0. , Borsheim, K. Y. , Bratbak, G. & Heldal, M. (1989) High abundance of viruses found in aquatic environments. Nature 340, 467-468.  Bratbak, G. , Egge, J. K. & Heldal, M. (1993) Viral mortality of the marine alga Emiliania huxleyi (Haptophyceae) and termination of alga blooms. Marine Ecology Progress Series 93,  39-48.  Brussaard, C. P. D. (2004) Viral control of phytoplankton populations-a review. The Journal of Eukaryotic Microbiology 51, 125-138.  Chen, F. & Suttle, C. A. (1995) Amplification of DNA polymerase gene fragments from viruses infecting microalgae. Applied and Environmental Microbiology 61, 1274-1278.  Chen, F. , Suttle, C. A. & Short, S. M. (1996) Genetic diversity in marine algal virus communities as revealed by sequence analysis of DNA polymerase genes. Applied and Environmental Microbiology 62, 2869-2874.  Clasen, J. L. (2001). Aquatic Viral Ecology. Department of Biology. Tempe, Arizona State University. Master's: 75.  27  Cottrell, M. T. & Suttle, C. A. (1991) Wide-spread occurrence and clonal variation in viruses which cause lysis of a cosmopolitan eukaryotic marine phytoplankter, Micromonas pusilla. Marine Ecology Progress Series 78, 1-9.  Cottrell, M. T. & Suttle, C. A. (1995) Dynamics of a lytic virus infecting the photosynthetic marine picoflagellate Micromonas pusilla. Limnology and Oceanography 40, 730-739.  Dunigan, D. D. , Fitzgerald, L. A. & Van Etten, J. L. (2006) Phycodnaviruses: A peek at genetic diversity. Virus Research 117, 119-132.  Evans, C. , Archer, S. D. , Jacquet, S. & Wilson, W. H. (2003) Direct estimates of the contribution of viral lysis and microzooplankton grazing to the decline of a Micromonas spp. population. Aquatic Microbial Ecology 30, 207-219.  Filee, J. , Francoise, T. , Suttle, C. A. & Krisch, H. M. (2005) Marine T4-type bacteriophages, a ubiquitous component of the dark matter of the biosphere. Proceedings of the National Academy of Sciences 102, 12471-12476.  Fitzgerald, L. A. , Graves, M. V. , Li, X. , Feldblyum, T. , Hartigan, J. & Van Etten, J. L. (2007) Sequence and annotation of the 314-Kb Mt325 and the 321-Kb Fr483 viruses that infect Chlorella Pbi. Virology 358, 459-471.  Frederickson, C. F. , Short, S. M. & Suttle, C. A. (2003) The physical environment affects cyanophage communities in British Columbia inlets. Microbial Ecology 46, 348-357. 28  Fuhrman, J. A. & Suttle, C. A. (1993) Viruses in marine planktonic systems. Oceanography  6, 51-63.  Gobler, C. J. , Hutchins, D. A. , Fisher, N. S. , Cosper, E. M. & Sanudo-Wilhelmy, S. A. (1997) Release and bioavailability of C, N, P, Se and Fe following viral lysis of a marine chrysophyte. Limnology and Oceanography 42, 1492-1504.  Hurst, C. J. & Lindquist, H. D. A. (2000). Defining the ecology of viruses. Viral ecology. C. J. Hurst. London, Academic Press: 1-40.  Kang, M. , Dunigan, D. D. & Van Etten, J. L. (2005) Chlorovirus: a genus of Phycodnaviridae that infects certain Chlorella-like green algae. Molecular Plant Pathology  6, 213-224.  La Scola, B. , Audic, S. , Robert, C. , Jungang, L. , de Lamballerie, X. , Drancourt, M. , Birtles, R. , Claverie, J. M. & Raoult, D. (2003) A giant virus in amoebae. Science 299, 20332033.  LaPatra, S. E. (1998) Factors affecting pathogenicity of infectoius hematopoietic necrosis (IHNV) for salmonid fish. Journal of Aquatic Animal Health 10, 121-131.  Levine, A. J. (1992) Viruses. Scientific American Library, New York.  29  Levy, J. A. , Fraenkel-Conrat, H. & Owens, R. A. (1994) Virology. Prentice Hall, Englewood Cliffs, NJ.  Mann, N. H. (2003) Phages of the marine cyanobacterial picophytoplankton. FEMS Microbiology Reviews 27, 17-34.  Mann, N. H. , Cook, A. , Millard, A. , Bailey, S. & Clokie, M. (2003) Marine ecosystems: bacterial photosynthesis genes in a virus. Nature 424, 741.  Maranger, R. & Bird, D. F. (1995) Viral abundance in aquatic systems: a comparison between marine and fresh waters. Marine Ecology Progress Series 121, 217-226.  Mayer, J. A. & Taylor, F. J. R. (1979) A virus which lyses the marine nanoflagellate Micromonas pusilla. Nature 281, 299-301.  Myers, R. M. , Maniatis, T. & Lerman, L. S. (1987) Detecting and localization of single base changes by denaturing gradient gel electrophoresis. Methods in Enzymology. 155, 501-527.  Nayak, D. B. (2000). Virus morphology, replication and assembly. Viral Ecology. C. J. Hurst. London, Academic Press: 63-124.  Osterhaus, A. D. M. E. , Groen, J. , Uytdehaag, F. G. C. M. , Vanderbildt, M. W. G. , Visser, I. K. G. , Bergman, A. & Klingeburn, B. (1989) Distemper virus in Baikal seals. Nature 338, 209-210. 30  Paul, J. H. & Kellogg, C. A. (2000). Ecology of bacteriophages in nature. Viral Ecology. C. J. Hurst. London, Academic Press: 211-246.  Proctor, L. M. & Fuhrman, J. A. (1990) Viral mortality of marine bacteria and cyanobacteria.  Nature (London) 343, 60-62.  Shirai, Y. , Takao, Y. , Mizumoto, H. , Tomaru, Y. , Honda, D. & Magasaki, K. (2006) Genomic and phylogenetic analysis of a single-stranded RNA virus infecting Rhizosolenia  setigera (Stramenopiles : Bacillariophyceae). Journal of the Marine Biological Association of the United Kingdom 86, 475-483.  Short, C. M. & Suttle, C. A. (2005) Nearly identical bacteriophage structural gene sequences are widely distributed in both marine and freshwater environments. Applied and  Environmental Microbiology 71, 480-486.  Short, S. M. & Suttle, C. A. (1999) Use of the polymerase chain reaction and denaturing gradient gel electrophoresis to study diversity in natural virus communities. Hydrobiologia  401, 19-32.  Short, S. M. & Suttle, C. A. (2000) Denaturing gradient gel electrophoresis resolves virus sequences amplified with degenerate primers. BioTechniques 28, 20-26.  31  Short, S. M. & Suttle, C. A. (2002) Sequence analysis of marine virus communities reveals that groups of related algal viruses are widely distributed in nature. Applied and Environmental Microbiology 68, 1290-1296.  Short, S. M. & Suttle, C. A. (2003) Temporal dynamics of natural communities of marine algal viruses and eukaryotes. Aquatic Microbial Ecology 32, 107-119.  Sullivan, M. B. , Waterbury, J. B. & Chisholm, S. W. (2003) Cyanophages infecting the oceanic cyanobacterium Prochlorococcus. Nature 424, 1047-1051.  Suttle, C. A. (1992) Inhibition of photosynthesis in phytoplankton by the submicron size fraction concentrated from seawater. Marine Ecology Progress Series 87, 105-112.  Suttle, C. A. (1994) The significance of viruses to mortality in aquatic microbial communities. Microbial Ecology 28, 237-243.  Suttle, C. A. (2000). Ecological, evolutionary, and geochemical consequences of viral infection of cyanobacteria and eukaryotic algae. Viral Ecology. C. J. Hurst. London, Academic Press: 247-296.  Suttle, C. A. (2005) Viruses in the sea. Nature 437, 356-361.  Suttle, C. A. (2007) Marine viruses-major players in the global ecosystem. Nature Review 5, 801-812. 32  Suttle, C. A. & Chan, A. M. (1993) Marine cyanophages infecting oceanic and coastal strains of Synechococcus: abundance, morphology, cross-infectivity and growth characteristics. Marine Ecology Progress Series 92, 99-109.  Suttle, C. A. & Chan, A. M. (1994) Dynamics and distribution of cyanophages and their effect on marine Synechococcus spp. Applied and Environmental Microbiology 69, 31673174.  Suttle, C. A. , Chan, A. M. & Cottrell, M. T. (1990) Infections of phytoplankton by viruses and reduction of primary production. Nature 347, 467-469.  Thingstad, T. F. (2000) Elements of a theory for the mechanisms controlling abundance, diversity and biogeochemical role of lytic bacterial viruses in aquatic systems. Limnology and Oceanography 45, 1320-1328.  Van Etten, J. L. (2003) Unusual life style of giant chlorella viruses. Annual Review of Genetics 37, 153-195.  Van Etten, J. L. , Burbank, D. W. , Kuczmarski, D. & Meints, R. H. (1983) Virus infection of culturable Chlorella-like algae and development of a plaque assay. Science 219, 994-996.  Van Etten, J. L. , Graves, M. V. , Muller, D. G. , Boland, W. & Delaroque, N. (2002) Phycodnaviridae-large DNA algal viruses. Archives of Virology 147, 1479-1516.  33  Van Etten, J. L. & Meints, R. H. (1999) Giant viruses infecting algae. Annual Reviews in Microbiology 53, 447-494.  Van Etten, J. L. , Van Etten, C. H. , Johnson, J. K. & Burbank, D. E. (1985) A survey for viruses from freshwater that infect a eukaryotic Chlorella-like green algal. Applied and Environmental Microbiology 49, 1326-1328.  Wilhelm, S. W. & Suttle, C. A. (1999) Viruses and nutrient cycles in the sea. Bioscience 49, 781-788.  Wilson, W. H. , Van Etten, J. L. , Schroeder, D. S. , Nagasaki, K. , Brussaard, C. P. D. , Delaroque, N. , Bratbak, G. & Suttle, C. A. (2005b). Phycodnaviridae Virus Taxonomy: Classification and Nomenclature of Viruses, Eighth Report of the International Committee on the Taxonomy of Viruses. C. M. Fauquet, M. A. Mayo, J. Maniloff, U. Desselberger and L. A. Ball. San Diego, Elsevier Academic Press: 163-175.  Wommack, K. E. & Colwell, R. R. (2000) Virioplankton: Viruses in aquatic ecosystems. Microbiology and Molecular Biology Reviews 64, 69-114.  34  Chapter two: Viral abundance across marine and freshwater systems is driven by different biological factors  A version of this chapter has been accepted for publication. Clasen, J.L., Brigden S.M., Payet, J.P. & Suttle, C.A. (In Press) Evidence that viral abundance across oceans and lakes is driven by different biological factors. Freshwater  Biology.  35  2.1. Summary Samples from 16 lakes in central (n = 145) and western (n = 12) North America; the coastal northeast Pacific (n = 302) and the western Canadian Arctic Oceans (n = 142) were collected and processed for viral, bacterial and cyanobacterial abundances, and chlorophyll a concentrations. Viral abundance was significantly different among these environments (ANOVA, F = 432, p < 0.0001); it was highest in the coastal Pacific Ocean, and lowest in the coastal Arctic Ocean. The abundance of bacteria (ANOVA, F = 189, p < 0.0001) and cyanobacteria (ANOVA, F = 45, p < 0.0001), as well as, chlorophyll a concentrations (ANOVA, F = 322, p < 0.0001) also differed among environments, with bacterial abundance and chlorophyll a concentrations highest in lakes, while cyanobacteria abundance were highest in the coastal Pacific Ocean. Consequently, the relationships of these variables with viral abundance varied among environments. A discriminant analysis indicated that the environments were predictably different. Multiple regression analyses included bacterial and cyanobacterial abundances, and chlorophyll a concentrations as significant variables in explaining viral abundance in lakes (r 2 = 0.387, p < 0.0001). The significant variables in the models for the coastal oceans were different than the lake model. Differences in viral production and loss rates are likely responsible for the variation in viral abundance among environments. For example, fresh waters have the fewer viruses compared to bacteria, despite previously documented higher burst sizes and frequency of infected cells in freshwater environments, suggesting loss rates may be more important in lakes. Viral abundance and the biological variables related to viral abundance differ between freshwater and marine environments implying different drivers of viral abundance in lakes and oceans.  36  2.2 Introduction Viruses are dynamic and ecologically important members of aquatic ecosystems that can influence nutrient cycles, community composition and horizontal gene transfer (see Fuhrman, 1999; Paul, 1999; Wilhelm & Suttle, 1999; Wommack & Colwell, 2000; Suttle, 2005). The factors influencing viral abundance and dynamics in aquatic environments are complex; however, correlation analysis can be used to identify physical, chemical and biological variables that are associated with changes in viral abundance. Such analyses have already provided insights into the factors regulating viral abundance in aquatic environments (reviewed in Suttle, 2000; Wommack & Colwell, 2000). For example, correlation analysis indicated that viral abundance is typically associated with bacterial abundance and/or chlorophyll a concentrations (Cochlan et al., 1993; Paul et al., 1993; Jiang & Paul, 1994; Steward et al., 1996; Weinbauer & Suttle, 1997; Kepner et al., 1998). The majority of these studies focused on marine rather than freshwater environments. However, Maranger & Bird (1995) used correlation and multiple regression analyses to compare viral abundance from lakes and marine systems with bacterial abundance and production, as well as with chlorophyll a and total dissolved phosphorous concentrations. Interestingly, they found a strong correlation between viral abundance and chlorophyll a concentrations, but not with bacterial abundance in lakes, whereas, in marine systems viral abundance was correlated strongly with bacterial abundance. Based upon these results, the authors suggested that the factors controlling the abundance of viruses differ between lakes and oceans. However, as viral abundance has frequently been underestimated due to a number of methodological problems (see Hennes & Suttle, 1995; Weinbauer & Suttle, 1997; Bettarel et  37  al., 2000; Wen et al., 2004; Suttle, 2005), we re-visited the relationships between the  abundance of viruses and potential hosts as inferred from cell counts of heterotrophic bacteria and unicellular cyanobacteria, as well as, chlorophyll a concentrations (i.e. biological variables). Ultimately, we wish to understand the influence of biological variables on viral abundance in both freshwater and marine systems. Given the uncertain accuracy of many estimates of viral abundance due to inherent errors in transmission electron microscopy and fixation problems in many studies using epifluorescence microscopy, we used several relatively large data sets from our own laboratory in which we had confidence in the accuracy of abundance estimates. We compared viral abundance with several biological variables for freshwater and marine environments to determine which biological factors drive changes in viral abundance.  2.3 Methods Sampling Lakes Samples were collected from lakes in Wisconsin (WIS 1999), British Columbia (BCC 2002), Lake of the Woods, Ontario (LOW 2003) and the Experimental Lakes Areas, Ontario (ELA 2003 and 2004) during the ice-free season (June to October) (Figure 2.1). The 157 freshwater samples came from lakes that varied in size, depth and trophic status (Juday & Birge, 1941; Cleugh & Hauser, 1971). Samples were collected by either a Van Dorn bottle, a pump or by carefully submerging and filling a carboy with subsurface water. Samples were taken from depths less then 2 m, except for samples from Wisconsin (WIS 1999) which were  38  Arctic BFS (n = 142)  Pacific SOG (n = 302)  Figure 2. 1. Map of the sampling environments used in this study. Lake samples were collected from Wisconsin (WIS), British Columbia (BCC) and Ontario (Lake of the Woods (LOW) and the Experimental Lakes Area (ELA)). All the coastal Pacific Ocean samples were collected in/near the Straight of Georgia, British Columbia (SOG), while coastal Arctic Ocean samples were collected in the Southwestern part of the Beaufort Sea (BFS).  39  collected at depths from 0.5 to 32 m. Water samples were processed immediately for viral, bacterial and cyanobacterial abundances, as well as chlorophyll a concentrations using the methods described below.  Coastal Pacific Ocean  Coastal Pacific Ocean samples were collected during four research cruises in and near the Strait of Georgia, British Columbia (Figure 2.1). These cruises occurred during August 1996, June 1997, July 2000 and July 2002 (hereafter, SOG 1996, 1997, 2000 and 2002, respectively). The Strait of Georgia lies between continental British Columbia and Vancouver Island. It is ca. 28 km wide and 240 km long and is characterized by areas of high primary productivity and flushing rates (Harrison et al., 1983). Water samples were collected from 0.5 to 100 m using Go-Flo bottles mounted on a rosette equipped with a CTD and fluorometer. Upon reaching deck, samples for viruses, bacteria, cyanobacteria and chlorophyll a were collected and processed immediately.  Coastal Arctic Ocean  In the Canadian Arctic Ocean, samples were collected in the Southeastern Beaufort Sea in October 2003 and from May to August 2004 (BFS 2003 & 2004) during the Canadian Arctic Shelf Exchange Study (CASES) (Figure 2.1). Water was collected at several stations, ranging from stations heavily influenced by the Mackenzie River to more oceanic stations in the Amundsen Gulf (Garneau et al., 2006; Garneau et al., In press). All stations were located on the Arctic shelf which is an important area of primary production in the Arctic Ocean (Carmack et al., 2004). Water from depths ranging from 1.0 to 530 m was obtained with  40  Niskin or Go-Flo bottles mounted on a rosette. Samples were immediately processed for viruses, bacteria and chlorophyll a once onboard.  Enumeration Viruses  Total viral abundance was determined by epifluorescence microscopy (EFM) of samples stained with either YO-PRO-1  TM  (WIS 1999, SOG 1996 and BFS 2003 & 2004) or  SYBR Green I'm (BCC 2002, LOW 2003, ELA 2003 & 2004 and SOG 1997 & 2000). However, there is no difference in the accuracy of YO-PRO-1  TM  and SYBR Green I TM to  determine viral abundance when the samples are processed correctly (see methodological consideration in the discussion). Yo-Pro-1 TM. For samples stained with YO-PRO-1 TM (4-13-methy-2,3-dihydro-  (benzo-1,3-oxazole1-2-methylmethyledene1-1-(3' -trimethyl-amonium-propyl)-quinolinium diiodide; Invitrogen, Burlington, ON, Canada), 0.8 to 1.6 mL of the sample was gently ( < 250 mm Hg) filtered onto a 0.02 pm pore-sized Anodisc filter (Whatman, Florham Park, NJ, USA) and placed sample side up on an 80 pL drop of YO-PRO-1  TM  (final concentration of  50 pM) (Hennes & Suttle, 1995). The slides were incubated in the dark for 48 h, rinsed with filtered-deionized water and then mounted on a slide with 100 % glycerol. Slides were frozen until enumerated on an epifluorescence microscope with a wide-blue filter set (450480 nm). Total viral abundance (viruses mL -1 ) was determined by counting a minimum of 300 viruses over 20 random fields on each slide. The incubation period in the samples from the Wisconsin lakes (WIS 1999) was reduced to 4 min by microwave radiation (Xenopoulos & Bird, 1997).  41  TM (Invitrogen) was used (Noble & Fuhrman, SYBR Green ITA1 . When SYBR Green I  1998), the sample was fixed with glutaraldehyde to a final concentration of 2 % and processed within 1 h of fixation (Wen et al., 2004). A volume of the fixed sample (0.8 to 1.6 mL) was filtered onto a 0.02 [tm pore-sized Anodisc filters (Whatman) and placed sample side up on an 80 pL drop of SYBR Green I TM (diluted 400 x) and incubated in the dark for 15 min. After incubation, the filter was carefully blotted dry with a Kimwipe TM and mounted on a slide using a mixture of 50 % glycerol, 50 % phosphate buffered saline (PBS) and 0.1 % p-phenylenediamine. Slides were frozen until enumerated as described above.  Bacteria Bacterial abundance was determined by EFM of cells stained with YO-PRO-1 TM (WIS 1999, BFS 2003 & 2004), SYBR Green I TM (SOG 2000), Acridine Orange (SOG 1996 & 1997) or 4',6-diamidino-2-phenylindole (BCC 2002, LOW 2003, ELA 2003 & 2004). However, any difference observed between environments can not be explained by the use of these different stains (see methodological considerations in the discussion). Bacterial abundance (bacteria mL -1 ) was estimated by counting more than 200 bacteria over 20 random fields on each slide. Bacteria enumerated with YO-PRO-1 TM and SYBR Green I  Tm  were processed as  described for viruses. When samples were stained with Acridine Orange (AO; Hobbie et al., 1977) or 4',6-diamidino-2-phenylindole (DAPI; Porter & Feig, 1980), 0.8 to 5 mL of the sample was incubated for 3 min with AO (2 mL of 0.1 mg mL -1 ) or 10 min with DAPI (250 uL of 0.1mg mL -1 ) and then filtered onto a black 0.2 lam pore-sized polycarbonate filter. The filter was mounted on a slide with low fluorescence immersion oil and frozen until enumerated by EFM using either wide-blue (AO) or UV (DAPI) filter sets. 42  Chlorophyll a The concentration of chlorophyll a (rig L -1 ) was determined by gently (< 250 mm Hg) filtering 15 to 100 mL of freshly collected water through a GF/F filter (Whatman) under low light (WIS 1999, BCC 2002, LOW 2003, ELA 2003 & 2004, SOG 1997 and BFS 2003 & 2004). The filters were frozen until extracted overnight in 90 % acetone and the concentration of chlorophyll a was determined by fluorometry (Wetzel & Likens, 1991).  Cyanobacteria For SOG 1996, cyanobacteria abundances were determined from YO-PRO-1  TM  virus  slides (see above) by counting autofluorescent cyanobacteria cells under wide-green excitation (510-550 nm). For the other samples (LOW 2003, ELA 2003 & 2004 and SOG 1997 & 2000) cyanobacterial abundance was determined by filtering 5-15 mL of water onto a black 0.2 iim pore-sized polycarbonate filter. The filter was filtered to dryness, mounted onto a glass slide with 100 % glycerol and frozen until enumerated. For all the stations (including SOG 1996), abundance (cyanobacteria cell mL -I ) was determined by counting > 200 autofluorescent cyanobacteria under wide-green excitation (510-550 nm).  Data Analysis ANOVA analysis Log (base 10) transformed biological variables, including viral abundance (viruses mL -I ), bacterial abundance (cells mL -1 ), chlorophyll a concentrations (lig L -1 ), cyanobacterial abundance (cells mL-I ) and the virus-to-bacteria ratio, were compared among environments  43  by ANOVA with a Tukey post-hoc test (SYSTAT v. 11). Viral abundance was also plotted against the other biological variables to determine the relationship among the variables.  Discriminant analysis Discriminant analyses were used to statistically describe differences among environments (SYSTAT v. 11). Viral, bacterial and cyanobacterial abundances, and chlorophyll a concentrations were used as predictors in the analyses. A discriminant analysis predicts the likelihood that an 'unknown' sample belongs to a particular group (Manly, 2005). All data were log (base 10) transformed to satisfy the assumptions of this analysis, which include homoscedaticity and normal distribution.  Multiple regression analysis To determine the variables that explained the most variation in viral abundance within each environment, the biological variables were regressed against viral abundance in a backwards step-wise multiple regression analysis using SYSTAT v. 11. This analysis systemically removes independent variables from the model that contribute little to explaining the dependent variable. At each removal step, the data are scanned for variables that should be re-entered into the model because their contribution has increased due to interactions with other variables (Sokal & Rohlf, 1995). Independent variables in the regression models included bacterial abundance (cell m1: 1 ), cyanobacterial abundance (cyanobacteria m1: 1 ) and chlorophyll a concentrations (jig 1: 1 ); however, because cyanobacteria data were not determined from the Arctic Ocean samples, they were not  44  included as an independent variable in the Arctic Ocean regression model. All the data were log (base 10) transformed prior to analysis to satisfy the assumptions of regression analyses.  2.4 Results The abundance of viruses was determined in 601 samples, representing three distinct environments, and ranged from 2.2 x 10 5 to a 3.9 x 10 8 viruses mL -1 (Table 2.1). There were significant differences in viral abundance (Figure 2.2a. ANOVA, n = 601, F2,598 = 432.6, p < 0.0001), and bacterial abundance (Figure 2.2a. ANOVA, n = 601,  F2,598 =  189.3, p = <  0.0001) among environments. Lakes had significantly higher amounts of chlorophyll a than either the coastal Pacific or Arctic Oceans (Figure 2.2b. ANOVA, n = 397,  F2,394 = 322.8,  p<  0.0001); while the coastal Pacific Ocean had higher abundances of unicellular cyanobacteria (Figure 2.2b. ANOVA, n = 201 ,  F1,289 = 45.0,  p < 0.0001). Finally, ratios of viruses to  bacteria were significantly different among environments (Figure 2.2c. ANOVA, n = 601, F2,598 = 435.5,  p < 0.0001).  Overall, viral abundance was related to bacterial abundance, cyanobacterial abundance and chlorophyll a concentrations (bacteria: r 2 = 0.22, p < 0.0001, n =565; chlorophyll a: r2 = 0.022, p = 0.005, n =361; cyanobacteria: r 2 = 0.02, p = 0.005, n =391); however, when the data were analyzed separately by environment, striking patterns emerged (Figure 2.3). For example, there was little overlap among the clusters of viral and bacterial abundances generated from each environment (Figure 2.3a). Similarly, clusters of viral abundance and chlorophyll a concentrations or cyanobacterial abundance were distinctly  45  ^ ^  ^.  Ri e • .—• z 6 CU 'Lk 0 z^irt^• ...I :CI t4  J.4^ ; • IM 6  ...,  (:j Cl) ......,  F,2...,  ---,  "--..., <3 ---, c.,1 V') 0 cr.) •—■ ••••,^kr) Q\, 0 N • I N 71. 71 ^I '-1 °° _. CD C .......  CD ,--4  171 ^ m.^„, ,—., ,C ,, c..) 7' v__, Le)^0 y  A■w^: 14 0  "C) al V ^E '—° ej^CI X ch '-I^k X  ,,,.,...; .....„.  ,--,  .....,  ,—,  cs4  ....,  WI C  p^,4 1■1 •-■I^c \I, (:::! Ch = 6 're :j.^cr■ —4^1  ct^el ^ca ...,^,.... C...) vl t z © ,.., " 0--. =1 4^  6 .....  ,..., ,r) CD ,  r".., 'C'CD  k  -  6  X  ,..., c'' %.•  ^  In• D\1 d  ,0^  —' r'":  o ,--I  ./.1  GC - ,- t....^cr) 71-^71 . ■c) 7f. Cr 6 .4 • I CT ^r) Cs1 Cr) bt (",21 M C; 6 d 6  *6'  'V ,.'‘  X In  x ,—  ..4 ,0 —1 O  c^ad^z.. c,^....., 7z" : U S  1 —  7. ct 5 <14 V  vp  -4  CU  ,  X C:J■  C7'■ ,•  'Q ""  c; .....  71" Cr;^ci  c:5 ......  MI  4—'  0^CI 0^,..,4 U al l U  m^c.) cl^;..),  '.'"^4 .  7t^(.)  0; ,I, ,c;  cA ,—.' In 0 N Cr) ,-,  X  e...., \0 ,-.... r'-..., 0 ‘....) 1...-1 C) .--.^,--I C D^,-- I 0 C >< .—I ,--1 ,-0  X  6  r— cd . 1-..1 .•-•^, et Z = pp XI^a.) ©^1 4 cg ‘.-1.^,cf-, 04  ^etc^,?..,  E  ,...... 1 '"  Z Z  4W.)i y  ....•  E  CI? ^WI .ir  E) a)  E 0  •6 p. ›' 0 W  46  Figure 2.2. A comparison of the biological variables determined within each of the three aquatic environments. a) Viral abundance (viruses mL -1 ) and bacterial abundance (cells mL ), b) chlorophyll a concentrations (pg L -1 ) and cyanobacterial abundance (cells mL -1 ) and c) -1 virus-to-bacteria ratio. Error bars indicate + 1 standard error, when not visible the error was less then the width of the bar. The abundance of cyanobacteria was not determined (nd) from the coastal Arctic Ocean samples. For each variable, the means are significantly different (ANOVA) from each other when marked with different letters.  47  a) c ,71 e+8 a as --I -0 E  ir)  t(-) 1 e+7  ow Viruses Bacteria a  --I  o  E w1e+6 ))  T2 . 2 e+5  Lake Pacific Arctic  0 -J  E )1e+6  °  0 a) c  T. ca  b^Chlorophyll a Cyanobacten  1 e+4  a) c c -91e+2 o 0 Tti (1) 1e+0 CL  2 c' 0 20.01 00  Lake Pacific Arctic  60 0 Ca  1.4 Lake  Pacific Arctic  48  Figure 2.3. Viral abundance (viruses mL -1 ) as a function of either a) bacterial abundance (cells mL -1 ), b) chlorophyll a concentrations (pg L -1 ) or c) cyanobacterial abundance (cells mL -1 ) within each of the environments. In all cases, lake samples are closed circles, coastal Pacific Ocean samples are open squares and coastal Arctic Ocean samples are crosses. The ovals represent ± 1 standard deviation around the means.  49  J  E  S 100.0 (/)  a) tn  2  10.0  0 V  • Lake ^ Pacific x Arctic  1.0  0.1^1.0^10.0^100.0 Bacterial Abundance (cells x 10 5 mL - ) /1111^1^1 111111^1^1.1111„ ^1^111119^1^,,,11„1^1^1 I  _b  x X - x^ X , 111111i^1^1 1111111Y 1  ^1111111  1^1 1111111^1^1 1111111  1^1 1  • Lake ^ Pacific x Arctic  0.01^0.1 1.0^10.0 100.0 Chlorophyll a Concentrations (Ng L -1 )  10.0  • Lake Pacific 111111^1^1 1111111^1^1 1111111^1^1111111^1^1 1111111  ^.01^1.0^10.0 100.0 1000.0 Cyanobacterial Abundance (cell x 10 3 mL -1 )  50  different between the lakes and the marine regions, including the Pacific and Arctic coastal environments (Figure 2.3b and c, respectively). To examine the extent to which the data could be separated by environment, a series of discriminant function analyses were performed. With just viral abundance in the model, the data was placed into the correct environment — 67 % of the time (Table 2.2). The accuracy of the models increased when bacterial abundance was added. Data from the lakes, coastal Pacific and Arctic oceans were correctly identified 90 %, 92 % and 88 % of the time, respectively (overall average = 90 %). The addition of chlorophyll a concentrations and cyanobacterial abundance increased the overall accuracy to 94 % and 99 % (Table 2.2). Backward step-wise multiple regression analyses were used to determine which biological variables explained the most variation in viral abundance within each environment. The models for lakes and the coastal Pacific Ocean regressed viral abundance (viruses mL -1 ) against bacterial abundance (cells mL -I ), chlorophyll a concentrations (pg L -1 ) and cyanobacterial abundance (cells mL -1 ). The Arctic Ocean model regressed viral abundance (viruses mL-I ) against bacterial abundance (cells mL -I ) and chlorophyll a concentrations (pg L-I ). The final models explained 39 %, 52 % and 26 % of the variation in lakes, and the coastal Pacific and Arctic Oceans, respectively (Table 2.3). Bacterial abundance remained in the final models for all three systems, explaining some of the variation. However, the models differed in their remaining variables. Part of the variation in viral abundance in the coastal Pacific Ocean was explained by the abundance of cyanobacteria; chlorophyll a concentrations explained some of the variation in the coastal Arctic ocean, whereas, both cyanobacterial abundance and chlorophyll a concentrations contributed to the final model in lakes (Table 2.3).  51  Table 2.2. Discriminant Factor Analysis. The percent of 'unknown' samples that are placed in the correct environments based on several discriminant analyses with the predictors listed in the top row. Environment  Viruses  Viruses, bacteria  Viruses, bacteria, chlorophyll a  Lakes  49 %  90 %  93 %  Viruses, bacteria, chlorophyll a, cyanobacteria 98 %  Coastal Pacific  79 %  92 %  100%  100%  Coastal Arctic  63 %  88 %  90%  nd  Total  67 %  90 %  94 %  99%  nd: Not determined; cyanobacterial abundances were not determined from the coastal Arctic Ocean samples.  52  Table 2.3. Backward step-wise multiple regression analyses of viral abundance. Data included as independent variables in the lake and coastal Pacific Ocean models included the log of bacterial abundance, chlorophyll a concentrations and cyanobacterial abundance. The coastal Arctic Ocean model contained only the log of bacterial abundance and chlorophyll a concentrations. VA = viral abundance, BA = bacterial abundance, Chla = chlorophyll a concentrations, Cyanos = cyanobacterial abundance. Environment  Backwards step-wise regression models  Lake  y = 4.831 + 0.486 Log BA — 0.291 Log Cyanos + 0.197 Log Chla (r2 = 0.387, p < 0.0001 n = 95) y = 4.696 + 0.370 Log BA + 0.146 Log Cyanos (r2 = 0.517, p < 0.0001 n = 295) y = 4.559 + 0.364 Log BA + 0.0.95 Log Chla (r2 = 0.263, p < 0.0001 n = 138)  Coastal Pacific Coastal Arctic  53  2.5 Discussion There were several major findings that stemmed from our analyses. The first is that there were clear differences in viral abundance among environments. Additionally, the relationships between viral abundance and the other biological variables were different in all three environments. Secondly, different biological variables were significant predictors of the observed variation in viral abundance among environments. All together, our results suggest that there is a significant difference between lakes and oceans in terms of viral abundance and associated biological variables. We suggest that differences in viral production and loss rates in the two systems may contribute to these observations. These results are discussed below.  Methodological considerations Although both chemistry and physics have profound affects on the biota of aquatic systems, these data were not available for all environments. Moreover, we were most interested in comparing our results with those from other studies, including Maranger & Bird (1995). For these reasons, our analyses only examined the relationships between viruses and other biological variables. We did not include estimates of viral abundance from many other studies because most were based on transmission electron microscopy (TEM) or on epifluorescence microscopy (EFM) using samples that were fixed and stored for > 1 hour before processing, and these methods can lead to significant underestimates of viral abundance (see Hennes & Suttle, 1995; Weinbauer & Suttle, 1997; Wen et al., 2004). In this study, a number of different protocols and dyes were used for estimating the abundances of viruses and bacteria because the data sets span several years, researchers and  54  projects. However, for the reasons outlined below, the differences observed among environments are real and not the result of biases associated with the different methodologies. Both Yo-Pro-1 TM and SYBR Green I TM were used to estimate viral abundance in our study. However, with the exception of samples from the Wisconsin lakes (WIS 1999), which were processed using a modified Yo-Pro-1  TM  method of Xenopoulos & Bird (1997), the  samples were handled as outlined in Wen et al. (2004). Bettarel et al. (2000) compared different methods for estimating viral abundance in aquatic samples, including the original (Hennes and Suttle 1995) and modified Yo-Pro-1 TM (Xenopoulos & Bird 1997) and the SYBR Green I TM (Noble & Fuhrman, 1998) methods. They did not find a significant difference between estimates of viral abundance made using the original and modified YoPro-1 TM methods, but reported lower estimates when SYBR Green I  TM  was used, likely  because the samples were not processed immediately after fixation. However, when water samples are fixed and processed immediately, as was done in our study, there is no difference in the accuracy of Yo-Pro-1 TM and SYBR Green ITM (Wen et al., 2004). Within our own dataset, there was no significant difference between viral abundances in the same environment determined using the different stains (Pacific, t144 = -1.458, p = 0.147). Estimates of bacterial abundance in this study were made using four different fluorochromes (AO, DAPI, Yo-Pro-1  TM  and SYBR Green I Tm ). Posch et al. (2001) reported  that estimates of bacterial abundance made with DAPI were about 10 % lower than those made with AO, while Suzuki et al. (1993) reported that DAPI counts in coastal seawater averaged only about 70 % of those made with AO. In our study, DAPI was only used for freshwater samples, with the exception of the Wisconsin (WIS 1999) samples which were stained using the modified Yo-Pro-1  TM  method. As estimates of bacterial abundance was 55  significantly higher in lakes than in either the Pacific or Arctic Ocean samples, if bacterial abundance was underestimated in the lake samples by using DAPI it would only exaggerate the differences reported in this study. Although there are no direct comparisons of bacterial abundance estimates made with SYBR Green I  TM  and Yo-Pro-1 TM relative to DAPI or AO,  one would anticipate that if there was a bias, estimates made with Yo-Pro-1  TM  and SYBR  Green I TM would be higher because of their higher fluorescence yield. However, there is no evidence for such a bias; our bacterial abundances are similar to those reported previously for each of the environments (Chrzanowski et al., 1996; Wilhelm et al., 2002; Kirchman et al., 2007). Additionally, if a bias did exist, it would have to be very large to alter the patterns observed in this study, it would mean that nearly all the estimates of bacterial abundance in the literature are incorrect. Additionally although, direct comparisons within our data set are difficult, bacterial abundance determined within lakes using Yo-Pro-1  TM  and DAPI were  similar (Lakes, t45 = 0.247, p = 0.806). All estimates of cyanobacterial abundance were made by counting autofluorescent cells under wide-green excitation, and would not have been affected by the addition of YoPro-1 1'm or the use of a different filter type for the SOG 1996 samples. As is frequently done in other studies, we used chlorophyll a as a proxy for phytoplankton biomass, even though the ratio of Carbon to chlorophyll a can vary several fold as a function of both species composition, light and nutrient history. Nonetheless, it provides a convenient approximation of phytoplankton biomass and allows the comparison of our results to those of others, including Maranger & Bird, (1995). Finally, as previously mentioned several researchers enumerated viruses and the other biological variables. Multiple researches could introduce a bias, especially if the researcher only worked in one environment. However in our study, a few researchers determined 56  abundances in different environments, reducing this potential bias. For example, one researcher processed and counted samples from both the coastal Pacific and Lake environments. Moreover, all the researches were trained by the same individual, further reducing any potential enumeration bias associated with the use of multiple researchers. Overall, we believe that the results in this study reflect innate differences among environments and are not rooted in methodological artifacts.  Differences across environments Viral abundance was significantly different among environments (Figure 2.2 and Table 2.1). Additionally, there were clear differences between all three environments in terms of the relationships of viral abundance and other biological variables (Figure 2.3). Discriminant analyses confirmed that the environments were predictably different when biological variables were included in the analysis (Table 2.2). Multiple regression analysis indicated that viral abundance in lakes was a least partly (39 %) explained by bacterial and cyanobacterial abundances, as well as, chlorophyll a concentrations. In contrast, in the Pacific and Arctic Oceans bacteria and either cyanobacterial abundance (coastal Pacific) or chlorophyll a concentrations (coastal Arctic) explained a proportion of the variation (52 and 26 %, respectively; see Table 2.3). These results are consistent with different biological factors influencing viral abundance in the different aquatic systems. The data of Culley & Welschmeyer (2002) provide further evidence that the relationship between viral and bacterial abundances differs between lakes in North America and the eastern Pacific (Figure 2.4a). Culley & Welschmeyer (2002) used the original YoPro-1' method to estimate viral abundance and produced data that are directly comparable to those in the present study. Their relationship between viral and bacterial abundances 57  —0—Pacific regression 0 (2)^et al. 2002 00^regression  0.0^1.0^2.0 Bacterial Abundance (cells x 10 7 mL -1 )  .7" 100.0  E (tS r  0)  '5  10.0  a)  1.0  •  Lake  ^ Pacific X Arctic 0.1 0.1  •  Deep  10.0^100.0  Bacterial Abundance (cells x 10 5 mL -1 )  Figure 2.4. a) The regression from Culley et al. (2002) transect from the coastal to open Pacific Ocean compared to the regression generated from coastal Pacific samples used in this study. b) A comparison of viral abundance (viruses mL -1 ) and bacterial abundance (cells ml.. -1 ) from this study to another published data set; the deep Pacific Ocean (Ortmann, 2005). Lake samples are closed circles, coastal Pacific Ocean samples are open squares, coastal Arctic Ocean samples are crosses and deep Pacific Ocean samples are open diamonds. The ovals represent ± 1 standard deviations around the means and the line is the regression line.  58  collected on a transect from southern California to Hawaii is very similar to that for the coastal Pacific samples (Figure 2.4a), providing evidence that the relationship between viral and bacterial abundances holds for water from both productive coastal and oligotrophic oceanic waters. Taken into consideration with the data from the Arctic Ocean, and from within and outside neutrally buoyant hydrothermal vent plumes in the deep Pacific ocean (Ortmann & Suttle, 2005), it is evident that there are differences among environments with respect to the relationships between the abundances of viruses and bacteria (Figure 2.4b). Additionally, recent data from Parada et al. (2007) showing very high abundances of viruses relative to bacteria in the deep waters of the North Atlantic support the claim that there are different controls on viral abundance among systems. Interestingly, although the different environments cluster into separate groups (Figure 2.4b), the regression slope coefficients from the diverse marine environments (coastal Pacific, coastal Arctic, deep Pacific) were not significantly different from each other, suggesting similar mechanisms are driving this relationship within oceans. However, the marine slope coefficients were significantly different from the slope coefficient from lakes (Figure 2.5, ANOVA, p < 0.0001); clearly demonstrating that the relationships between viral and bacterial abundances differ between lakes and oceans. Maranger & Bird (1995) directly compared viral abundance and other biological variables for 22 Quebec lakes with marine data from the literature. They also found that marine and fresh waters differed significantly, but the relationships were not the same as those in our study. The authors found a significant relationship between viral abundance and chlorophyll a concentrations, but unlike our results, did not find a strong correlation between viral and bacterial abundances in lakes (Figure 2.2a). Maranger & Bird (1995) fixed and stored their samples at 4 degrees for up to a month before determining viral abundance by 59  Lake Pacific Arctic Deep  b  0 E 1,5 2 2.),  T  1  ‘ )  ce 8 a) c C a)  a  1 .  a  C  Lake^Pacific Arctic^Deep  Figure 2.5. a) A comparison of the slope coefficients and b) yintercepts generated from Figure 2.4 (viral abundance regressed against bacterial abundance in lakes and the coastal Pacific and Arctic Oceans and the deep Pacific Ocean). Error bars indicate variation about the regression slope or y-intercept. Means indicated by different letters are significantly different (ANOVA) from each other.  60  TEM; both their fixation and enumeration methods would likely have led to inaccurate estimates of viral abundance (Hennes & Suttle, 1995; Bettarel et al., 2000; Wen et al., 2004). Unfortunately, there is no reliable way to correct for changes in viral abundance due to fixation artifacts or TEM enumeration. Any data obtained from samples that were fixed and stored at 4 degrees for more then an hour can not be directly compared because fixation affects (including viral decay rates) vary substantially within and among environments. As a result, the higher virus-to-bacteria ratios (VBR) in lakes compared to oceans reported by Maranger and Bird (1995) should be interpreted carefully. In our study where direct comparisons are possible, the ocean VBR was higher than those found in lakes (Figure 2.2, Table 2.1). Other studies in freshwater or marine environments where the samples were immediately processed have VBR similar to the ones reported in this study (Culley & Welschmeyer, 2002; Seymour et al., 2006; Colombet et al., 2006); however, it should be noted that VBR range widely in environmental samples (see Wommack & Colwell, 2000). Yet, what is most interesting is not the absolute VBR in each environment but rather the relationship between viral and bacterial abundances. For example, while the Pacific and Arctic VBR are significantly different from each other (Figure 2.2, Table 2.1) their regression coefficients and y-intercepts are not (Figure 2.3), suggesting that similar processes are driving the relationship between these two variables despite differences in their absolute abundances.  Causes of environmental differences The slope coefficients from plots of viral versus bacterial abundances were significantly higher in the lake environment than those in the marine environments (Figure 2.5), suggesting that the relationship between viral and bacterial abundances is different in 61  the two systems. Viral abundance is affected by viral production and loss processes, such as contact rates, burst size, percent of infected cells, host diversity, UV radiation and viral decay rates. Variations in any of these factors could explain the observed differences in the virus and bacteria relationships in freshwater and marine samples. Lakes and the coastal Pacific Ocean had comparable bacterial abundances (Figure 2.2, Table 2.1) and consequently would have similar contact rates (assuming every encounter leads to an infection). However, viral abundances in the coastal Pacific Ocean were almost an order of magnitude higher than in lakes (Figure 2.2 and Table 2.1), indicating that contact rates between viruses and bacteria cannot explain the differences between marine and freshwater systems. The observed variation in the virus and bacteria relationship could be caused by differences in the burst sizes and/or the percent of bacteria infected in lakes and oceans. A recent review by Parada et al. (2006) reported higher burst sizes and frequencies of visibly infected cells in freshwaters, suggesting that viral production rates are higher in freshwater systems. Therefore, if viral loss rates are similar in marine and fresh waters, one would expect this to result in more viruses compared to bacteria in lakes. This is the opposite of what we observed (Figures 2.2, 2.3 and 2.4b), suggesting that mechanisms other than burst size and infection rates are controlling the relative abundance of viruses and bacteria in lakes. Perhaps assuming equal viral loss rates in the two environments is unrealistic, as unequal loss rates could explain the difference. Processes that could influence loss rates include viral adsorption to particles, degradation by virucidal compounds or UV radiation (Berry & Noton, 1976; Kapuscinski & Mitchell, 1980; Suttle & Chen, 1992). The magnitude of viral loss is largely unknown and since it could explain our observed difference between freshwater and marine environments, further research is warranted.  62  Alternatively, the decoupling of viruses and bacteria in lakes compared to the marine environments could be due to an increased importance of viruses infecting other organisms in lakes. Lakes had among the highest photosynthetic biomass (as measured by chlorophyll a concentrations and cyanobacterial abundance; Figure 2.2, Table 2.1) and both cyanobacteria and chlorophyll a were significant predictors of viral abundance in the lake regression model (Table 2.3), whereas this was not the case for the marine data. This suggests that the higher slope coefficient for the lake data could be related to the higher photosynthetic biomass; lakes could have a higher contribution of viruses from photosynthetic hosts. Interestingly, this idea was originally suggested by Maranger & Bird (1995) based upon their lakes survey in Quebec.  Potential ecological implications The ecological implications of the different relationships between viruses and host cells in the environments are potentially important. For example, the observed lower abundance of viruses relative to bacteria in lakes suggests that viral-mediated mortality of bacteria may be less important in lakes than oceans. Of course more extensive sampling and research is necessary to support this statement. Additionally, chemical and physical variables should be included in future regression analyses in order to explain more of the observed variations in viral abundance; which could further clarify the differences between the environments. However, if further analysis indicates that the difference between environments is truly driven by biological variables, the role of viruses in biogeochemical cycling could be affected. For example, nutrients moving through the viral shunt (Thingstad  et al., 1993; Suttle, 2005) could be suppressed in lakes if viral-mediated bacterial mortality is  63  less important in these environments. This could have ecosystem-wide implications, including affects on ecosystem productivity and should be explored further.  2.6 Acknowledgements The authors gratefully acknowledge the assistance of present and past members of the Suttle lab who collected many of the samples, as well as, the crews of the CCGS Vector and the CCGS Amundsen. A.C. Ortmann kindly provided her original deep Pacific abundance data. M.-E. Garneau and W.F. Vincent from the Canadian Arctic Shelf Exchange Study (CASES) provided the Arctic Ocean chlorophyll a data. Collection of some of the lake samples was made possible by the support of the staff and students at the University of Wisconsin's Trout lake station and J.J. Elser (IRCEB NSF grant DEB 9977047). The manuscript was improved by comments from C. Chenard, J. Grainger and A.C. Ortmann and two anonymous reviewers. This research was supported by Discovery and Network (CASES) grants from the Natural Science and Engineering Research Council of Canada (C.A.S.) and a graduate student fellowship (J.L.C.), as well as an Environmental Research Fellowship from the Lake of the Woods Distinct Properties Owners Association (J.L.C.).  64  2.7 References Berry, S. A. & Noton, B. G. (1976) Survival of bacteriophages in seawater. Water Research 10, 323-327.  Bettarel, Y. , Simo-Ngando, T. , Amblard, C. & Laveran, H. (2000) A comparison of methods for counting viruses in aquatic systems. Applied and Environmental Microbiology 66, 2283-2289.  Carmack, E. C. , MacDonald, R. W. & Jasper, S. (2004) Phytoplankton productivity on the Canadian Shelf of the Beaufort Sea. Marine Ecology Progress Series 277, 37-50.  Chrzanowski, T. H. , Kyle, M. , Elser, J. J. & Sterner, R. W. (1996) Element ratios and growth dynamics of bacteria in an oligotrophic Canadian shield lake. Aquatic Microbial Ecology 11, 119-125.  Cleugh, T. R. & Hauser, B. W. (1971) Results of initial survey of experimental lakes area, Northwestern Ontario. Journal of the Fisheries Research Board of Canada 28, 129-137.  Cochlan, W. P. , Wilmer, J. , Steward, G. F. , Smith, D. C. & Azam, F. (1993) Spatial distribution of viruses, bacteria and chlorophyll a in neritic, oceanic and estuarine environments. Marine Ecology Progress Series 92, 77-87.  65  Colombet, J. , Sime-Ngando, T. , Cauchie, H. M. , Fonty, G. , Hoffmann, L. & Demeure, G. (2006) Depth-related gradients of viral activity in Lake Pavin. Applied and Environmental Microbiology 72, 4440-4445.  Culley, A. I. & Welschmeyer, N. A. (2002) The abundance, distribution, and correlation of viruses, phytoplankton, and prokaryotes along a Pacific transect. Limnology and Oceanography 47, 1508-1513.  Fuhrman, J. A. (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399, 541-548.  Garneau, M.-E. , Roy, S. , Lovejoy, C. , Gratton, Y. & Vincent, W. F. (In press) Seasonal dynamics of bacterial biomass and production on the Arctic shelf: Franklin Bay, western Canadian Arctic. Journal of Geophysical Research.  Garneau, M.-E. , Vincent, W. F. , Alonso-Stez, L. , Gratton, Y. & Lovejoy, C. (2006) Prokaryotic community structure and heterotrophic production in a river-influenced coastal arctic ecosystem. Aquatic Microbial Ecology 42, 27-40.  Harrison, P. J. , Fulton, J. D. , Taylor, F. J. R. & Parsons, T. R. (1983) Review of the biological oceanography of the Straight of Georgia: pelagic environment. Canadian Journal of Fisheries and Aquatic Sciences 40, 1064-1094.  66  Hennes, K. P. & Suttle, C. A. (1995) Direct counts of viruses in natural waters and laboratory cultures by epifluorescence microscopy. Limnology and Oceanography 40, 1050-1055.  Hobbie, J. E. , Daley, R. J. & Jasper, S. (1977) Use of Nuclepore filters for counting bacteria by fluorescence microscopy. Applied and Environmental Microbiology 33, 1225-1228.  Jiang, S. C. & Paul, J. H. (1994) Seasonal and deil abundance of viruses and occurrence of lysogency/bacteriocinogeny in the marine environment. Marine Ecology Progress Series 104, 163-172.  Juday, C. & Birge, E. A. (1941) Hydrography and morphometry of some Northeastern Wisconsin lakes. Transactions of the Wisconsin Academy of Science 33, 21-72.  Kapuscinski, R. B. & Mitchell, R. (1980) Processes controlling viruses inactivation in coastal waters. Water Research 14, 363-371.  Kepner, R. L. , Wharton, R. A. & Suttle, C. A. (1998) Viruses in Antarctic lakes. Limnology and Oceanography 43, 1754-1761.  Kirchman, D. L. , Elifantz, H. , Dittel, A. I. , Malmstrom, R. R. & Cottrell, M. T. (2007) Standing stocks and activity of Archaea and Bacteria in the western Arctic Ocean. Limnology and Oceanography 52, 495-507.  67  Manly, B. F. J. (2005) Multivariate Statistical Methods: A primer. Chapman and Hall/CRC, Boca Raton.  Maranger, R. & Bird, D. F. (1995) Viral abundance in aquatic systems: a comparison between marine and fresh waters. Marine Ecology Progress Series 121, 217-226.  Noble, R. T. & Fuhrman, J. A. (1998) Use of SYBR Green I for rapid epifluorescence counts of marine viruses and bacteria. Aquatic Microbial Ecology 14, 113-118.  Ortmann, A. C. & Suttle, C. A. (2005) High abundance of viruses in a deep-sea hydrothermal vent system indicates viral mediated microbial mortality. Deep-Sea Research 1 52, 15151527.  Parada, V. , Herndl, G. J. & Weinbauer, M. G. (2006) Viral burst size of heterotrophic prokaryotes in aquatic systems. Journal of the Marine Biological Association of the United Kingdom 86, 613-621.  Parada, V. , Sintes, E. , van Aken, H. M. , Weinbauer, M. G. & Herndl, G. J. (2007) Viral abundance, decay and diversity in the meso- and bathypelagic waters on the North Atlantic. Applied and Environmental Microbiology 73, 4429-4438.  Paul, J. H. (1999) Microbial gene transfer: an ecological perspective Journal of Molecular Microbiology and Biotechnology 1, 11-26.  68  Paul, J. H. , Rose, J. B. , Jiang, S. C. , Kellogg, C. A. & Dickson, L. (1993) Distribution of viral abundance in the reef environment of Key Largo, Florida. Applied and Environmental Microbiology 59, 718-724.  Porter, K. G. & Feig, Y. S. (1980) The use of DAPI for identifying and counting aquatic microflora. Limnology and Oceanography 25, 943-948.  Posch, T. , Loferer-Krossbacher, M. , Gao, G. , Alfreider, A. , Pernthaler, J. & Psermer, R. (2001) Precision of bacterioplankton biomass determination: a comparison of two fluorescent dyes, and of allometric and linear volume-to-carbon conversion factors. Aquatic Microbial Ecology 25, 55-63.  Seymour, J. R. , Seuront, L. , Doubell, M. , Waters, R. L. & Mitchell, J. G. (2006) Microscale patchiness of virioplankton. Journal of the Marine Biological Association of the United Kingdom 86, 551-561.  Sokal, R. R. & Rohlf, F. J. (1995) Biometry. W.H. Freeman and Company, New York.  Steward, G. F. , Smith, D. C. & Azam, F. (1996) Abundance and production of bacteria and viruses in the Bering and Chukchi Seas. Marine Ecology Progress Series 131, 287-300.  Suttle, C. A. (2000). Ecological, evolutionary, and geochemical consequences of viral infection of cyanobacteria and eukaryotic algae. Viral Ecology. C. J. Hurst. London, Academic Press: 247-296. 69  Suttle, C. A. (2005) Viruses in the sea. Nature 437, 356-361.  Suttle, C. A. & Chen, F. (1992) Mechanisms and rates of decay of marine viruses in seawater. Applied and Environmental Microbiology 58, 3721-3729.  Suzuki, M. T. , Sherr, E. B. & Sherr, B. F. (1993) DAPI direct counting underestimates bacterial abundances and average cell size compared to AO direct counting. Limnology and Oceanography 38, 1566-1570.  Thingstad, T. F. , Heldal, M. , Bratbak, G. & Dundas, I. (1993) Are viruses important partners in pelagic food webs. Trends in Ecology & Evolution 8, 209-212.  Weinbauer, M. G. & Suttle, C. A. (1997) Comparison of epifluorescence and transmission electron microscopy for counting viruses in natural marine waters. Aquatic Microbial Ecology 13, 225-232.  Wen, K. , Ortmann, A. C. & Suttle, C. A. (2004) Accurate estimation of viral abundance by epifluorescence microscopy. Applied and Environmental Microbiology 70, 3862-3867.  Wetzel, R. G. & Likens, G. E. (1991) Limnological Analysis. Springer-verlag, New York.  Wilhelm, S. W. , Brigden, S. M. & Suttle, C. A. (2002) A dilution technique for the direct measurement of viral production : A comparison in stratified and tidally mixed coastal waters. Microbial Ecology 43, 168-173. 70  Wilhelm, S. W. & Suttle, C. A. (1999) Viruses and nutrient cycles in the sea. Bioscience 49, 781-788.  Wommack, K. E. & Colwell, R. R. (2000) Virioplankton: Viruses in aquatic ecosystems.  Microbiology and Molecular Biology Reviews 64, 69-114.  Xenopoulos, M. A. & Bird, D. F. (1997) Virus a la sauce Yo-Pro: Microwave-enhanced staining for counting viruses by epifluorescence microscopy. Limnology and Oceanography  42, 1648-1650.  71  Chapter Three: Spatial and temporal viral dynamics in lakes are influenced by trophic status and regional climatic conditions  A version of this chapter will be submitted for publication. Clasen, J.L., Findlay, D.L. & Suttle, C.A. Spatial and temporal viral dynamics in lakes are influenced by trophic status and regional climatic conditions.  72  3.1 Summary To determine the importance of viruses that infect eukaryotic phytoplankton (Phycodnaviridae) in structuring phytoplankton communities, the temporal and spatial  variations of these viruses in three lakes in the Experimental Lakes Area, Canada were investigated over the ice-free season of 2004. Viral abundance and richness were determined in Lake 227, Lake 239 and Lake 240. Richness of the Phycodnavirus community was assessed using denaturing gradient gel electrophoresis (DGGE) of algal virus specific (AVS1 and 2) amplified PCR products. Each lake was distinct in both viral abundance and richness. The eutrophic Lake 227 had substantially higher abundances of viruses (p < 0.0001) and overall viral richness than the other two lakes. Among the lakes, there were synchronized temporal patterns in both viral abundance and richness, including an association between viral abundance and chlorophyll a concentrations in the spring and early summer. Analyses indicated that the Phycodnavirus communities in the spring and early summer months were richer than those present during the fall months, further suggesting that eukaryotic phytoplankton infections are ecologically important in the spring and early summer months. Our data imply that the overall Phycodnavirus community richness is influenced by trophic status; whereas patterns of richness are affected by regional climatic conditions. Since richness can directly affect host-virus interactions, these results have potentially important implications for understanding the influence viruses have on phytoplankton communities.  73  3.2 Introduction Spurred by the discovery of high abundances of viruses in the oceans (Bergh et al., 1989; Proctor & Fuhrman, 1990), researchers have found that viruses exist in all aquatic environments, including coastal and deep oceans, lakes, rivers, estuaries, sea ice, sediments, hydrothermal vents and rain (see Suttle, 2000; Wommack & Colwell, 2000; Weinbauer, 2004). Despite their ubiquitous distribution, viral abundance and dynamics vary among environments. Clasen et al. (In Press; Chapter two) found a significant difference between viral abundances in different aquatic systems, with more viruses present in the coastal Pacific Ocean than in lakes. Additionally, the relationship between the abundances of viruses and bacteria was substantially different between lakes and oceans. In lakes there was a greater change in viral abundance with bacterial abundance than was found in marine environments. This decoupling between the abundances of viruses and bacteria may be explained by an increased contribution of viruses that infect phytoplankton to the viral pool in lakes. This hypothesis is supported by the inclusion of both chlorophyll a concentrations and cyanobacterial abundance as significant predictors of viral abundance in a backward-stepwise regression analysis (Clasen et al., In Press; Chapter two). Chlorophyll a concentrations were also found to be significant predictors of viral abundance in another lake study (Maranger & Bird, 1995), adding credence to the idea that viruses of phytoplankton may be an important component of the virioplankton in lakes. In general, aquatic viruses are ecologically important, affecting community composition, nutrient cycling and genetic diversity via lateral gene transfer and selective mortality (see Fuhrman, 1999; Paul & Kellogg, 2000; Wommack & Colwell, 2000; Suttle, 2005). For example, Suttle (2005) estimated that 150 Gt of carbon a year is lost through  74  viral-mediated cell lysis, mostly from the 20-40 % of bacteria cells that are lysed daily. Viruses that infect phytoplankton may be potentially important, because as primary producers, eukaryotic phytoplankton and cyanobacteria convert inorganic carbon into organic carbon, fueling the ecosystem. Therefore, viral infections of phytoplankton could shuffle a previously underappreciated amount of carbon (and other nutrients) away from higher trophic levels to the microbial food web via the viral shunt (Thingstad et al., 1993; Wilhelm & Suttle, 1999; Suttle, 2005) To date, culture dependent and independent techniques suggest that viruses that infect phytoplankton are both abundant and genetically rich (Suttle, 2000; Brussaard, 2004; Dunigan et al., 2006). So far, — 15 groups of viruses that infect eukaryotic phytoplankton have been isolated and cultured. Most of these viruses are members of the family Phycodnaviridae, which are large, icosahedral, dsDNA viruses (Brussaard, 2004; Dunigan et al., 2006). Viruses in this family infect a diverse range of hosts, including species in the  genera Chlorella, Phaeocystis, Chrysochromulina, Micromonas, Heterosigma and Emiliania (reviewed in Brussaard, 2004). Chen and Suttle (1995) developed algal virus specific degenerate primers (hereafter AVS-1 and 2) which amplify the highly conserved 0-family (alike) DNA polymerase genes of many Phycodnaviridae, as a culture independent mechanism to explore natural viral richness. Short and Suttle (2003) paired AVS-PCR and denaturing gradient gel electrophoresis (DGGE) to separate the PCR amplified products based upon their nucleotide composition (Short & Suttle, 1999), in order to investigate Phycodnaviruses in marine environmental samples. Their year-long study revealed high genetic richness and complex interactions between both the biological and physical environments and the viral community (Short & Suttle, 2003).  75  Despite their potential importance, little is known about Phycodnaviruses in freshwaters. In fact, only one group of Phycodnaviridae (Chlorovirus) is known to infect freshwater hosts, an endosymbiotic Chlorella-like algae (Meints et al., 1981; Van Etten & Meints, 1999). Although a great deal is known about the biology of Chlorella viruses (see Van Etten & Meints, 1999; Dunigan et al., 2006), with the exception of one environmental survey (Van Etten et al., 1985), the ecology of Phycodnaviruses in lakes remains largely unexplored. Given the suggestion that phytoplankton viruses may be an ecologically important component of lake ecosystems and the fact that they remain largely unstudied in freshwater habitats, a study of viruses that infect eukaryotic phytoplankton was conducted in several lakes in the Experimental Lakes Area in Northwestern Ontario, Canada.  3.3 Methods Sampling Three lakes (Lakes 227, 239 and 240) located in the Experimental Lakes Area (ELA) in Northwestern Ontario were used in this investigation (Figure 3.1). The lakes were physically, chemically and biologically different (see Table 3.1 and Cleugh & Hauser, 1971). These differences can be partly explained by the artificial fertilization of Lake 227 with either nitrogen and/or phosphorus, which has occurred since 1969 (Findlay et al., 1994). All three lakes were sampled approximately every two weeks during the ice free season of 2004 (mid-May to mid-October) in conjunction with the long-term sampling program established  76  Figure 3.1 Map of the lakes sampled.  This figure has been removed for copyright reasons. The original image is available at www.umanitoba.ca/institutes/fisheries  77  ^  a) taEl  ;2 czt 6 -4 -o a) 6,  g  0 tu) (i)  a) -0 "c1 0 0 ••-• 0  vp  N C4  CL)  1.4 ct  "cl  ^"I°  0  o c.) 1-) 4-44  N cc bi) 0  (a u Li cs Z  4.-7 4.: cuo cl al Z Z 023 03 = ,..o WO t5LI  71-  c5 an  c)  ,_, 0, on  r-  cA  t) s u s u s  --, 71:  ,t  --, ,c5, kr,  --, 0,-; --,  .0 0\UcNUo  WI  -o  ct 03  0  .1. eq  0,, -% M o.4  M ei cu  ,..  0 •4  el  h N  4) C.) al t  ,.5  E  0., 0 -a.) 0  E —  ", •,.., ° 6..., z  .:' `,LA (I Z 03 -4 CIO -o ,.4 c4 H  VD  >, 7) o VD  >-, 7, o -g,  II  . __,  -g,  II  &o -  II  ....... 0.  < .1 44 (ct -o o .-  < ..1 W (c1 -75 = > ct  44  - >-, o 44  V) "17  > cc: c„ ,,;-^a)  a)^+-' II v.., vict  --  "C)  .,  .4 H  II  .... 0. __,  ,.....,  (7)  „  O,  c.i  ---, N II  2  00  c:>  (I) a.) Z at "t7  az as cz a) c.) at ...0  N II  CI  , ,' _ bp = ■-■  0 •", t.j s.zi V, c)  2 7,,, , 1=', ct 0 c...)^0 .1 ',,, 0 0 V0 •.,^li ..^Il  N II  .0-, -4  L::,  ;---1  E  0  (,)^  -  0  v, ti 6,^( 10 4 ''''' .5 U , .-^v, , ct^  .-4 . ,_,^c4^a.) 64^CI)^›, II  g^o^ --,-, ^>1 az 2 ,,v..1.^I. tr) muUU  ca 0 bp  2 1 —,—, 0  V., ••  a., ._,  ,.  ms^cl-) — o,..0  .,  8 ..-1 714-+^c4 0, 1 .4 ali ;.. 0 0 .0 ,..  a,  ,..0  o  4  ,-,=3  a) c)  0 ° 0 ,.^¢. ■0 ° 00 CL, 0 N C5 c:, ,c)^0^A-■ c-,1 N —1 r--- 0^;--1^,.., 0 o  .. -4  .5  a cu^ vC^ii13 ) ,-4 .4c7;; II V.., cZ ,—,  r-i r^)^• --,  ON  00 C- a 0 —1 ° en 0 e.1^0 (-NI VD. Ci Cr■^r--: C4^(---: 71- Ci CA^',,.:j: t,.;.^0^cb,p 4cA en E-- --1 ('01 c^ —0 en Lri C:3.^.7,n `c 1)'^ • , ;..,^0 N  0  ,....,  (7)  cn  0  „  a  II  a N  0 N II  6 N CA II  II  ... =  N II  vD = oo Z cn --,^0 7,,--, --4^(-4 7t.^t---: 00 4 ,--( 14 --1 6 cr; —' ,--4 N 00 —' rfi 4 tri '-' o o o o ....• 2 ,._, co N c). C:7 --.■ cNi -, d ■6  ‘..0  ,-12 Uc:"  E  cA  —'-' (L)  5 z ms ,-. ct^,----..  8 P. 11)  ..t: "; 0 8  6, -o q-, o  ,. st<, c..)  0  61 -2 0 '-'^ ,_, =I :<., < .1 clf.  ;..,^0 a)^0 -c) i. -g 0 --, a)^z  5 07. ,u-, 1-) E 5 a)^E a -7. ceZID .- t ct^cr3 g 0t .,..,as ...- — _,-,,..,^.. 1 ,.0^ , $-,^c4 7.',  ci-, cl'zi ° Z 'ci; ,  ct  a) ,--  cz a) 1.4 4-, 6.,  cA•ON"  v)  ,—, E  ..z) 0 '71C^" CJ.^" ^■LD^" tr) 44 ,—i N , __, ,__, • ^. cn ^--. M N 71- — ON N ..W o o 7:3 -/"" CI  CV ,-  /4  ,-4  a 6, as  al  4  ci) t:k0  c a)  Z cl  --,  ••-•  cl  `) "C:1  cll  -o 5  a) C.) .144^C14  ..4  a.)  O z1  --,  a) '"  E  o E 4-1  (1-) 1)^;-■  (  715 CLI  x 6 W.2  78  at the ELA. An integrated water sampler (Shearer, 1978) was used to collected water from the euphotic zone (0-16 m) to determine phytoplankton abundance and composition. Additionally, a 20-L polyethylene carboy rinsed with 10 % HC1 and then lake water was carefully submerged and filled with sub-surface water (0.5 m). A portion of this water was used to determine viral abundance and chlorophyll a concentrations, while the viral community in the remaining volume (- 18 L) of lake water was concentrated using ultrafiltration (see the concentration of natural viral communities section below). Daily mean air temperatures ( ° C) and speed of maximum wind gusts (km h . ') were obtained from Environment Canada for Kenora, Ontario, Canada which is approximate 50 km West of ELA. Additional climate data was obtained from National Oceanic and Atmospheric Administration (NOAA) and include clear sky UV index for Minneapolis, Minnesota, USA (744 km South of ELA); Sioux Falls, SD, USA; Portland, ME, USA; Detroit, MI, USA; Milwaukee, WI, USA and Chicago, IL, USA.  Abundance Viral abundance and chlorophyll a concentrations were determined from each water sample using the methods of Noble & Fuhrman (1998) and Wetzel & Likens (1991), respectively. Briefly, viral abundance was determined from a fixed (2 % final concentration of glutaraldehyde) water sample using the nucleic-acid stain SYBR Green I' (Invitrogen, Burlington, ON, Canada). A portion (250 [IL) of the fixed lake sample was filtered onto a 0.02 inn pore-sized Anodisc filter (Whatman, Florham Park, New Jersey, USA) within an hour of fixation as suggested by Wen et al. (2004). The filter was placed sample side up on a drop of diluted (400 x) SYBR Green I' and incubated in the dark for - 15 min. Upon completion, the filter was blotted dry and mounted on a clean glass side with 40 [iL of 79  mounting medium (50 % glycerol, 50 % phosphate buffered saline and 0.1% pphenylenediamine). Duplicate slides were made from each lake sample. Slides were frozen until enumerated using epifluorescence microscopy with a wide-blue filter set (450-480 nm). Viral abundance was determined by counting > 300 viruses over at least 20 random fields of view on each slide (Suttle, 1993), using the 100 x objective. The duplicate slides were then averaged to determine viral abundance in the sampled lake water (viruses mL -1 ). Chlorophyll a concentrations (pg L -1 ) were determined from duplicate filters. In each case, fresh water samples of 50 to 100 mL were gently (< 250 mmHg) filtered onto glassfiber GF/F filters (Whatman). Filters were frozen until processed, which occurred within 4 months. Chlorophyll a pigments were extracted overnight in 90 % acetone and concentrations (pg L i ) were determined by fluorometry (Wetzel & Likens, 1991). Eukaryotic phytoplankton abundance (cells L -1 ) and composition (to species level) were determined from the integrated water samples. In each case, the UtermOhl technique (1958) as modified by Nauwerck (1963) was used to determine phytoplankton abundance. Briefly, a sub-sample (10 mL) of a 125 mL Lugol's fixed integrated water sample was settled for 24 hours in UtermOhl settling chambers. Cells were counted on an invert microscope with a phase-contrast filter, using the 10 x and 40 x objectives. All phytoplankton samples were counted by D.L. Findlay of the Freshwater Institute in Winnipeg. Manitoba.  Concentration of natural viral communities The virioplankton community in each of the lake samples was concentrated using ultrafiltration (Suttle et al., 1991). In each case, — 18 L of lake water was pressure-filtered through several 0.45 pm pore-sized 142 mm diameter filters (Durapore, Millipore, Billerica, MA, USA) to remove large particles and organisms. The filtrate was then concentrated to a 80  final volume of — 200 mL using a 10 kDa molecular weight cut-off tangential flow ultrafiltration cartridge (S1Y10; Millipore). These viral concentrates (VC) were stored at 4 ° C in the dark until further processed (2-7 months).  DNA Extraction and PCR Viral DNA was extracted from each viral concentrate (VC) using the MoBio Ultra Clean Soil DNA extraction kit (MoBio, Carlsbad, CA, USA) which included a solution to remove inhibitors. The volume of VC extracted was standardized to the same initial volume of lake water (50 mL of lake water) using the VC concentration factor, extracted DNA was frozen until used as template in polymerase chain reactions (PCR). Algal virus specific primers (AVS-1 and 2) were used to amplify the DNA polymerases of viruses that specifically infect eukaryotic phytoplankton in the family Phycodnaviridae (Chen & Suttle, 1995; Chen et al., 1996). In each case, two reaction rounds  were carried out to increase the yield of amplicons. In the first round, 5 pi, of DNA template was added to a 45 pt PCR mix, containing 5.0 pL of 10 x PCR buffer, 1.5 pL of 50 mM MgC12, 1.0 pL of each of the 2.0 mM dNTPs, 1.0 pL of 10 nM of the algal virus specific primer-1 (AVS-1) and 3.0 gL of 10 nM of AVS-2, 0.625 U of PLATINUM ® Taq DNA polymerase (Invitrogen) and water. The negative control was prepared as above but contained no DNA template and Micromonas pusilla virus (MpV) was used as a positive control. The PCR was performed on a PCR Express Hybaid Tm Thermal cycler (Middlesex, UK) with the following reaction conditions: initial denaturation at 95 ° C for 90 s, followed by 40 cycles of denaturation at 95 ° C for 45 s, annealing at 45 ° C for 45 s and extension at 72 ° C for 45 s; and a final extension at 72 ° C for 7 min. To confirm amplification, 10 pL of PCR products and 2 pL of 6 x loading buffer were loaded onto a 1.5 % agarose gel flanked 81  with a 100 by ladder (Invitrogen). The gel was run at 90 V for 60 min in TBE (0.5 x) buffer, stained with Ethidium Bromide (EtBr) and then viewed on a UV transilluminator. Bands (- 800 bp) were plugged with a clean Pasteur pipette and placed in a sterile 0.5 mL micro-centrifuge tube. DNA was eluted by adding 100 pL of 1 x TAE and heating this solution to 65 ° C for 60 min. Two ',IL of the eluted DNA was used in a second round PCR. The 2 nd round reaction was prepared as described above, except the number of cycles was reduced to 26. Positive and negative controls were prepared as before; however, the negative control from the first round reaction was also plugged and included as a control in the second round reaction. As before, electrophoresis (1.5 % agarose gel in 0.5 x TBE, 90 V for 60 min) was done to check for amplification. The remaining second round PCR products were stored at -20 ° C.  Denaturing gradient gel electrophoresis (DGGE)  The richness and temporal variation in the AVS amplified viral community was assessed by denaturing gradient gel electrophoresis (DGGE), a technique that separates AVS amplified dsDNA fragments according to their nucleotide sequences as they pass through a polyacrylamide gel that has an increasing concentration of denaturant (Short & Suttle, 1999). Second round PCR products (40 pL) and 6 x loading buffer (101.1L) were loaded onto a 6 to 7 % polyacrylamide gel, which had a 20 to 40 % gradient of denaturant (100 % denaturant is defined as 7 M urea and 40 % deionized formamide). Samples were run at 60 V for 15 h in 60 ° C 1 x TAE buffer, using a D-code Tm electrophoresis system (Bio-Rad, Hercules, CA, USA). Upon completion, the gel was stained in a 1 x SYBR Green I Tm (Invitrogen) solution for > 3 h. Gels were visualized and photographed with the AlphaImager Tm system (Alpha Innotech, San Leandro, CA, USA). All gels were flanked with control markers, which were 82  used to correct any 'smiling' effects on the gels. The marker was a composite of several environmental DGGE bands that had been plugged, eluted and re-amplified via PCR. Gel images were imported into GelCompar IITM (Applied Maths). Individual bands were identified by peaks on a background corrected densitometric curve generated by GelComparII TM . Each band was assumed to be unique viruses and, therefore, the viral richness of the Phycodnavirus community was determined by either counting the number of unique bands in each lake (i.e. on each gel) or the number of bands present on each sampling date (i.e. in each lane).  Statistics Analysis of viral abundance and richness  Analyses of variance (SYSTAT v. 11) were used to ascertain spatial and temporal differences between lakes and dates, respectively. Viral abundance data (viruses mL -1 ) were log (base 10) transformed to satisfy ANOVA assumptions, including homoscedasity and normal distribution. The viral richness data did not need to be transformed. If needed, a Tukey post-hoc test was used to further examine significant differences between lakes or months. Rarefaction curves (which plot richness as a function of occurrence) were constructed (EcoSim700, v.7.72) using the viral richness data from each lake to determine if sampling was complete (Raup, 1975).  Cluster analysis of viral community richness  Within each lake, the relatedness of the AVS amplified viral communities was compared using the DGGE banding pattern on each gel and a hierarchical classification clustering analysis (SYSTAT v.11). The presence or absence of each band (identified by the 83  location of peaks on a densitometric curve) on the sampling dates was converted into a binary matrix. The clustering analysis then grouped similar communities based upon this binary matrix; however, intensity of the bands were not included as it can be an unreliable measure of abundance (Gelsomino et al., 1999). This cluster analysis used complete linkage as the joining algorithm and generated a diagram showing the percent difference between the sampled communities.  Multiple dimensional scaling (MDS) of phytoplankton abundance and composition  MDS was used to reduce the complex eukaryotic phytoplankton community present at each sampling date to two dimensions, allowing for easy comparison (SYSTAT v.11). In this case, the presence and absence of each phytoplankton species, as well as the abundance of that species were included in the analysis (determined by D.L. Findlay). All data were log (base 10) transformed to satisfy the assumptions of the analysis (log [abundance +1]). For each lake, a Kruskal model of monotonic MDS with phytoplankton species in the chlorophyte, cryptophyte, chrysophyte, diatom and dinoflagellate groups were used in the analysis. Stress values were reported for each analysis, which is an indication of the goodness of fit.  3.4 Results Spatial variation in viral abundance and richness  Lake 227 had a higher average viral abundance than the other two lakes (ANOVA, F 2,56 =  21 . 559 , p < 0.0001; Table 3.1, Figure 3.2a), while Lake 239 did not significantly differ  84  1e+8  ANOVA, p < 0.0001 a  E a) 1e +7 -  2  b  Cs  .0 1e +6 -  1e+5 227^239^240 14 - b  100 .  Average bands per date (+SE)  = Numbers of bands in each lake  n. E  ANOVA,}p = 0.275  12 -  - 80  10 -  43eL "C3 CO CO^  - 60 CT  8  6  - 40  rn  Ei  E  z 20  0 227^239^240  Figure 3.2. Spatial variation in viral abundance and richness. a) Plots of viral abundances (viruses mL -1 ) in the lakes. The error bars represent + 1 S.E. Means indicated by different letters are significantly different based on an ANOVA. b) The squares are the average number of bands on the DGGE per sample date. Error bars indicate + 1 S.E. and ANOVA results are indicated. The bars are the total number of unique bands per lake and represent overall Phycodnavirus richness in each of the lakes.  85  from Lake 240 (ANOVA, F 1,36 1.339, p = 0.246). Based upon changes in viral abundance through the ultrafiltration steps, the efficiency of concentration was only — 19 % (range 5 to 40 %; highest in Lake 240). Despite this, Lake 227 had a higher overall viral richness with more unique bands on the DGGE than the other two lakes (Lake 227 = 55, Lake 239 = 38 and Lake 240 = 51; Figure 3.2b). Although the average number of bands per sample date varied between the lakes, the lakes were not significantly different from each other in this parameter (ANOVA, F  2,30 =  1.346, p = 0.276, Figure 3.2b). In all three lakes, the rarefaction  curves were beginning to flatten, particularly in Lakes 239 and 240 (Figure 3.3).  Temporal variation in viral abundance and potential phytoplankton hosts There was a peak in viral abundance in all three lakes in the spring and early summer, followed by a sharp decline in abundance (Figure 3.4a and insert). There was, however, no significant difference in viral abundance between individual months (ANOVA, F 5, 53 = 2.285, p = 0.059, Figure 3.4a insert). The peak in viral abundance in the spring was associated with chlorophyll a concentrations in all three lakes during May, June and July (hereafter called spring/summer) (regression = F  1, 16 =  6.659, p< 0.0001, r 2 = 0.29), but not  during late July, August, September and October (hereafter called summer/fall) (Figure 3.4b, c, d). However, the two variables were slightly out of phase in the spring in Lake 227 (Figure 3.b). In all cases, a similar pattern was observed when phytoplankton abundance (cells L i ) was used instead of chlorophyll a concentrations (Figure 3.4b, c, d).  Temporal variation in viral richness and potential phytoplankton hosts Although viral richness in each of the lakes varied substantially throughout the  86  60  50 U) u)  a) c ...c  40 -  0  E us  30 -  —III— L239 —0— L227 —.— L240  0-  0^20^40^60^80^100^120^140  Occurrence Figure 3.3. Rarefaction curves. Curves for Lakes 227, 239 and 240 are shown. The curves are based upon the occurrence of individual DGGE bands found in all of the samples collected from a lake.  87  Figure 3.4. Temporal variation in viral abundance. a) Total viral abundances (viruses mL -1 ) in each of the three lakes determined from samples collected from mid-May to mid-October. The inserted panel is the average monthly viral abundance in all of the lakes + 1 S.E. b) Viral abundance (viruses mL -1 ), chlorophyll a concentrations (pg L -1 ) and phytoplankton abundance (cell L -1 ) in Lake 227 over the sampled period. c) Lake 239 and d) Lake 240. In all cases, the solid lines are viral abundances, the dashed lines are chlorophyll a concentrations and the bars are phytoplankton abundance. Note the different scales used in each of the panels.  88  3e+7  1.4e+7  a E  NOVA, p = 0.059  1.2e+7 a s  3e+7  1.0e+7  2e+7  .5  8.0e+6  +4  28+7  8  6.0e+6 et +7  5e+6  4.0e+6  N^/ 4 I  Lake 227 Lake 230 Lake 240  2.0e+6 5-  0  0.0 May  Jun  Jul  Aug  Sep  Oct  Nov  1.3 , 9  8e+8 a 6e+8  2  A 40+8  20+8  1. 4e+7 1.26+7 E g  26+7 al... Viral Abundance Chforophl a  C  Phytoplankton Comity  2e+7  1.06+7  3  e•-•  8.0e+6 -  10.7  C.  Y  6.01+8 4.00+6  5e.6 2 2.019+6 -  0  0.0 Jun  J  Aug^Sep^0 t  3e+7  30+7  E 2e+7  3 8  6  2e+7  0  3  le+7  54,6  0 Jun  Jul  Aug  Sep  Oct  89  season, careful examination suggested there was evidence of synchronous patterns among the lakes. First, all three lakes were richer in at the beginning of the season (Figure 3.5a,b,c) compared with the summer/fall months, although this trend was not statistically supported when analyzed monthly (Figure 3.5d; ANOVA, F5, 27 = 1.400, p = 0.256). Second, there were synchronized changes in the viral richness banding patterns in all the lakes. For example, hierarchical clustering analyses based upon the binary matrices generated from the gels in Figure 3.5 (see Appendices 1 to 3) indicated that the viral communities present in spring/summer clustered separately from those present in the summer/fall (Figure 3.6), except for the earliest spring samples (May 20, May 26 and June 1) which clustered with the late fall samples (Oct 4, 6 and 13). Additionally, the cluster analysis indicated that from June 23 to July 1 there was a temporary change in the richness banding patterns in all three lakes that reverted back to the previous banding pattern on the next sampling date (see June 23, July 1 and June 29 on Figures 3.5a,b and c, respectively, as well as the underlined dates in Figure 3.6). For example, in Lake 239 the difference among the viral community present on July 1 and those present either before (June 16) or after (July 15) this date was greater than 60 % in both cases; however, the difference among communities flanking the change observed on July 1 (i.e. between the communities present on June 16 and July 15) was only 15 % different (Figure 3.6b). Similar to the viral communities, there were also substantial changes in the eukaryotic phytoplankton communities between spring/summer and summer/fall, as indicated by the relatively small distances between most of the consecutively collected spring and early summer samples on the MDS analyses of phytoplankton abundance and composition (Figure 3.7; data available from the Freshwater Institute). However, phytoplankton samples were collected with less frequency in Lake 240, making it difficult to determine trends in this lake. 90  a: 227  b: 239  CD^0 CO^1—^CO^LO^‘— N N N 1`.■ N CO -N-NCO (0 T■ t■ CO CO 0) (T)^v-  a  c: 240  0^LO^CT) C^N  Nvi-  A—  NN  ,-  NCON  so  N- N- CO 5 co  L0^LO CO 0 *Zr NGON6 C":5 N- CO 5 co  zan  C  25  d ANOVA, p = 0.256 20  15  I  10  5  0 May^Jun^Jul  ^  Aug^Sep^Oct  Figure 3.5. Temporal variation in viral richness. a) DGGE gel of Phycodnavirus richness as amplified by the AVS primers in Lake 227, b) Lake 239 and c) Lake 240. In all cases, sample dates (month/day) are listed above each gel lane. d) The average monthly viral richness in all of the lakes + 1 S.E.  91  Figure 3.6. Hierarchical classification clustering analysis of viral richness patterns. a) A cluster analysis of the relatedness (plotted as percent difference) of the Phycodnavirus communities present on each of the sampling dates in Lake 227, b) Lake 239 and c) Lake 240. In all cases, distinct spring and summer clusters were observed and are indicated by the solid vertical lines. The earliest spring samples collected are italicized, while samples corresponding to the temporary change in the richness banding patterns are underlined (see text and Figure 3.5a, b and c).  92  a^SEPT2 ^ SEPT15 ^ MAY26 ^ OCT13 ^ AUG8 —I SEPT29 ^ AUG18^ JUNE20 JULY7 ^ J I JUNE23 JULY21 ^  I^I^I  0.0 0.1 0.2 0.3 0.4 0.5 0.6 Distances  b^JUNE4 JUNE16 JULY15 JULY26 JULY1 JULY29 1 AUG12 AUG26 ^ OCT4 ^ SEPT22I ^ MAY20 SEPT9 ^ t^  III!  0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Distances  C^JULY13^ IJUNE15^ AUG24^ SEPT8 ^ AUG10 ^ SEPT21 ^ JULY28 ^ JUNE29 ^ JUNE1 ^ OCT6 ^ I^I  ir  0.0 0.1 0.2 0.3 0.4 0.5 0.6 Distances  93  Stress = 0.132 ° AUG1 0  SEPT15 ° AUG18  -1^0^1 Dimension-1  -1^0^1 Dimension-1  -1^0^1 Dimension-1  ^  ^  ^  2  2  2  Figure 3.7. Multiple dimensional scaling plots of the eukaryotic phytoplankton communities in each lake. a) A MDS plot of the phytoplankton communities present at each sampling date in Lake 227, b) Lake 239 and c) Lake 240. The distance between any two data points is proportion to how similar phytoplankton communities are; small distances between communities indicate that are similar in both phytoplankton abundance and composition. The ovals in each panel encompass most of the spring/summer samples but does not represent a statistical test. Dates underlined corresponding to phytoplankton communities present when the temporary change in viral richness occurred.  94  3.5 Discussion The spatial and temporal study of viruses in three lakes at the Experimental Lakes Area in Ontario, Canada, yielded several interesting findings. First, to our knowledge this is the first report of Phycodnavirus richness in freshwaters. Second, overall viral abundance and richness were higher in the more productive environments. Third, the association between total viral abundance and chlorophyll a concentrations (and phytoplankton abundance) and the fact that viruses infecting eukaryotic phytoplankton are more genetically rich in the spring/summer suggest that these viruses may be more ecologically important during this time. As a result, overall Phycodnavirus communities appear to be influenced by trophic status; whereas patterns of richness are affected by regional climatic conditions. These results have potentially important ecological implications and emphasize the need to include viruses in future studies of freshwater systems.  Phycodnaviridae in lakes Although sequences from two viruses that infect a freshwater Chlorella-like alga were included in designing the algal virus specific primers (AVS-1 and 2; Chen & Suttle, 1995), the only published environmental data obtained using these primers are from marine samples (Short & Suttle, 1999; 2002; 2003). To our knowledge, this study is the first to amplify the 0-family (a-like) DNA polymerase gene fragments of Phycodnaviridae from environmental freshwater samples. Sequence analysis of amplified PCR products indicates that the AVS primers have a high fidelity for the DNA polymerase of Phycodnaviruses. All 62 currently available sequences cluster in strongly supported clades with known Phycodnaviridae isolates (Chen et al., 1996; Short & Suttle, 2002; 2003; Clasen, Chapter 4),  95  suggesting that the observed DGGE bands were likely amplified from Phycodnaviruses. The variable and low concentration efficiency could influence richness determination. However, the fact that the rarefaction curves are beginning to flatten indicates that the sampling for the viruses amplified by the AVS primers was adequate; despite the low efficiency, it came close to capturing the true richness of these Phycodnaviruses in all three lakes (Figure 3.3).  Spatial variation Viral abundance and overall richness were higher in Lake 227 than in the other two lakes (Figures 3.2a and 3.2b, respectively). Lake 227 is artificially fertilized as part of a long-term project (Findlay et al., 1994) and is classified as eutrophic. It has an average summer chlorophyll a concentration of 30 jig L -1 and a phytoplankton community composed primarily of species that thrive at higher nutrient concentrations, such as filamentous cyanobacteria and chlorophytes (Table 3.1; Wetzel, 2001). The high nutrient concentrations in Lake 227 support more bacteria, cyanobacteria and eukaryotic phytoplankton than Lakes 239 or 240 (Table 3.1). Intuitively, the higher density of potential host cells results in higher viral abundances. Similar high viral abundances with increased trophic status have been seen in other freshwater (Bettarel et al., 2003; Liu et al., 2006) and marine (Fuhrman, 1999; Wommack & Colwell, 2000) environments. Additionally, the eutrophication of Lake 227 is likely responsible for the observed high overall viral richness. Compared to the other lakes, Lake 227 had more unique DGGE bands, which represent different viral sequences (Figure 3.2b). Additionally, Lake 240 has slightly a higher trophic classification than Lake 239 based upon average chlorophyll a (Table 3.1) and nutrient concentrations (Cleugh & Hauser, 1971), and it also has slightly higher viral abundance and richness than Lake 239 (Figure 3.2). Therefore, the differences in 96  viral abundance and viral richness among the lakes can be attributed to differences in trophic status. A similar relationship between viral richness and ecosystem productivity was demonstrated (although not discussed) before using different primers sets. The richness of cyanophage determined from a conserved structural gene (g20) in two British Columbia lakes was higher in the oligo-mesotrophic Cultus Lake than the ultra-oligotrophic Chilliwack Lake (Short & Suttle, 2005), further supporting the idea that trophic status influences viral communities.  Temporal variation Although there were differences in total viral abundance and richness among the lakes, temporal changes in viral abundance and richness occurred at similar times. For example, in all three lakes viral abundances were highest during May and June and declined in late July (Figure 3.4a), suggesting a strong seasonality. It appears that the virioplankton community is influenced by eukaryotic phytoplankton in the spring since total viral abundances in the lakes were associated with both chlorophyll a concentration and phytoplankton abundance during this time, despite a slight mismatch in Lake 227 (Figure 3.4b,c,d), but not bacterial abundances (see Appendix.5). The temporal variation in richness provides further evidence of seasonality within the viral communities. In the three lakes viral richness was highest in the June, July and early August (Figure 3.5). In each of the lakes, the virioplankton communities cluster by season, with a few clusters containing most of the spring/summer samples, while the others contained the late summer/fall samples (Figure 3.6). The earliest spring samples (May 20, 26 and June 1) not only had low richness but also clustered closely with the samples collected on October 4, 6 and 13, likely due to the deep mixing that occurred during the early spring and late fall. 97  Other researchers have reported similarly strong seasonal trends in viral abundance in lakes in China and France (Bettarel et al., 2004; Liu et al., 2006); as well as a clear seasonality in cyanophage richness in Lake Bourget, France, using DGGE and sequence analyses (Dorigo et al., 2004). Seasonal changes in viral abundance and richness have also been found in marine viral communities (Marston & Sallee, 2003; Short & Suttle, 2003; Wang & Chen, 2004; Auguet et al., 2005). The strong seasonality in both viral abundance and richness observed in our data is associated with changes in the phytoplankton community during the clear-water-phase, which is characterized by a decline in the abundance of eukaryotic phytoplankton and a shift in community composition. In Lakes 239 and 227, the decrease in viral richness (see Figures 3.5 and 3.6) coincided with a decline in both chlorophyll a concentrations and phytoplankton abundances (Figure 3.8); this pattern doesn't seem to occur in Lake 240, but this may be the result of less phytoplankton abundance data available for this lake. In addition to the decrease in phytoplankton abundance, the composition of the phytoplankton community also changed between the spring/summer and the summer/fall, as indicated by the distances between consecutive dates on the MDS analysis of the phytoplankton communities (Figure 3.7). For example, there are substantial shifts in the diatom, chlorophytes and chrysophytes groups during this time (data available from the Freshwater Institute). A change in the phytoplankton community also represents a change in the potential hosts available, which intuitively would directly affect the composition of the Phycodnavirus communities As the water column warms after spring turnover, it stratifies and nutrients become limiting in the epilimnion (Lampert & Sommer, 1997). The lake enters the clear-water phase during which there is decline in the abundance of eukaryotic phytoplankton and a shift in  98  Figure 3.8. The relationship between Phycodnavirus richness and phytoplankton abundance. a) Overall viral richness present at each sampling dates plotted with total eukaryotic phytoplankton abundance (cells L -1 ) in Lake 227, b) Lake 239 and c) Lake 240. (Note the less frequent phytoplankton density data). In all the panels, the dashed vertical lines represent the decline in viral richness that corresponded to a decrease in phytoplankton abundance.  99  •  a  25  1e+9  • ▪ Phycodnavirus Richness rZZI Phytoplankton Abundance  20 -  8e+8 T.,  H  m  6e+8  15 2  4e , 8 0  C  C  0  2e+8  0 Sep  Oct  Nov  2e+7 Phycodnavirus Richness f77,1 Phytoplankton Abundance  -  2e+7  2) 10 -  1e+7  >, a-^5-  5e+6 go  15  -o  :g o.  Sep  25  Oct^Nov  ^ 2.0e+7  C  ♦•• Phycodnavirus Richness Phytoplankton Abundance  20 - 1.5e+7 15  tr  1.0e+7  2  ,  10  0  C C  5.0e+6 5o_ 0 ^ May^Jun  OD Jul  Aug  Sep  Oct  Nov  1 00  composition from predominately diatoms and chlorophytes to small flagellates (Wetzel, 2001). Therefore, the change from higher viral abundance and richness in the spring/summer to the pattern observed in the fall can be attributed to the onset of the clear-water-phase. In Lakes 240 and 227, Phycodnavirus richness increased again in the fall before dropping to low values. In Lake 240, this peak in richness is associated with a chlorophyll a peak; however this is not the case in Lake 227. Further investigation is needed but the additional peak in richness is likely related to the particular phytoplankton group present. However, in all three lakes, the variation in the richness of algal viruses as well as the association of total viral abundance with chlorophyll a concentrations and phytoplankton abundance are consistent with phytoplankton viruses being important in the spring/summer months in the lakes. The clear-water phase is driven by climatic factors including temperature, wind speed, and insolation (Lampert & Sommer, 1997), which indirectly affects viruses through changes in the phytoplankton community. There is other evidence that climate affects the virioplankton community through hosts. Ram et al. (2005) found that viral abundances collected during their year-long study of a reservoir in France were tightly coupled with water temperature. Furthermore, Bettarel et al. (2003) suggested that similar patterns in the abundances of viruses in two lakes in the Massif Central region of France were influenced by the light conditions. Beyond freshwater environments, researchers have found that viral abundance and genetic richness can be strongly influenced by physical parameters in both estuaries and oceanic environments; parameters that would in turn affect host abundance and diversity (Frederickson et al., 2003; Short & Suttle, 2003; Wang & Chen, 2004; Auguet et al., 2005). However, climate may also directly affect viral communities. For example during June 23, July 29 and July 1 a temporary change in viral richness occurred in all three lakes 101  (Figures 3.5, 3.8), even though the phytoplankton communities were similar during this time, as indicated by the close proximity on the MDS analysis (Figure 3.7). During this time, the daily mean air temperature (Kenora, ON, Canada) increased from 8 °C to 25 °C from June 23 to July 1 (Figure 3.9a), while the amount of UV radiation decreased (Figure 3.9b) and wind speeds diminished during this time (> 43 to < 30 km h-1 , data not shown). This combination of increasing temperature and low wind speeds would have likely lead to substantial daily fluctuations in surface water temperatures in the lakes. This change in temperature coupled with an decrease in UV irradiance may have contributed to the temporary change in viral richness by inducing latent infections or by decreasing the damage to viruses by UV radiation (Suttle & Chen, 1992). Although the mechanisms are unclear, the richness data are consistent with climatic conditions directly influence virioplankton community composition, however further research is necessary to specifically test this idea.  Potential ecological consequences All three lakes had higher viral abundances and richness in the spring/summer than during the summer/fall (ANOVA 1,31 =  abundance, F 1,57 =  4.433, p = 0.040, ANOVA  richness,  F  5.030, p = 0.032; Figure 3.10). These data, as well as, the association of viral  abundance with chlorophyll a concentrations and phytoplankton abundance in the spring/summer indicate that viruses of eukaryotic phytoplankton are a more important component of the ecosystem in the these months than the summer/fall months. This suggests that there may be increased viral-mediated mortality, nutrient cycling and potential for genetic exchange during the spring/summer. The importance and magnitude of these effects are likely influenced by both spatial and temporal variations. In this study, spatial differences in viral abundance and richness were driven by trophic status, while changes in 102  25  a)  20  T. a 15  E a.)  F-  •c  10-  a)  - Temperature Lake 227 A Lake 239 Lake 240 O  2 O  Jun 14^Jun 21^Jun 28^Jul 05^Jul 12^Jul 19  11  10  9  8  7  6 Jun14  ^  Jun21  ^  Jun28  ^  Jul05  ^  Jul12  Figure 3.9. Regional climatic conditions during the temporary change in viral richness. a) Daily mean air temperature (°C) data from Kenora, Ontario, Canada (solid line). The symbols correspond to mornings when in situ water temperatures data were collected. b) UV index from Minneapolis, MN, USA (thick black line); UV index data from Sioux Fall, SD, USA; Portland, ME, USA; Detroit, MI, USA; Milwaukee, WI, USA and Chicago, IL, USA are also included. The insert is the UV index from the cities listed above over then entire investigated period (Mid-May to Mid-October, 2004). In both panels, the horizontal bars flank the dates around the temporary change in viral richness.  103  1e+8  a^ ANOVA, p = 0.040  le+5 Spring/summer  25  b  ^  ^  Summer/fall  ANOVA, p = 0.032  20 U)  a) 15 co  "5 co C 0 0  10  0__c 5  0 Spring/summer  ^  Summer/fall  Figure 3.10. Seasonality in viral abundance and richness. a) Plots of viral abundance (viruses mL-1) and b) viral richness in the spring/summer and summer/fall seasons in all three lakes.  104  temporal patterns of viral abundance and richness were influenced by regional climatic conditions, which indirectly affect viruses through phytoplankton abundance and richness . The 2 to 3 ° C increase in temperature that is predicted to occur by 2020 in Northwestern Ontario will likely affect phytoplankton abundance and composition (Mugnuson et al., 1997), which may in turn affect Phycodnavirus community composition. The potential effects of climate change on aquatic viruses and the consequent implications for phytoplankton mortality and the viral shunt, emphasizes the need to consider viruses and viral-mediated processes in future research in lake  3.6 Acknowledgements We gratefully acknowledge the staff and students of the Experimental Lakes Area for their assistance with conducting this research, in particular Mark Lyng, Stephen Page, Justin Shead and Ken Sandilands. The past and present members of the Suttle and Shurin labs provided numerous helpful discussions, as well as technical and statistical assistance. Thanks to Caroline Chenard, Johan Vande Voorde, Matthias Fischer and Margaret Orlowski for reviewing and editing early drafts of this manuscript. Financial support came from the Natural Sciences and Engineering Research Council of Canada in the form of a graduate student fellowship to J.L.C. and a Discovery Grant to C.A.S. ecosystems.  105  3.7 References Auguet, J. C. , Montanie, H. , Delmas, D. , Hartmann, H. J. & Huet, V. (2005) Dynamic of virioplankton abundance and its environmental control in the Charente Estuary (France). Microbial Ecology 50, 337-349.  Bergh, 0. , Borsheim, K. Y. , Bratbak, G. & Heldal, M. (1989) High abundance of viruses found in aquatic environments. Nature 340, 467-468.  Bettarel, Y. , Sime-Ngando, T. , Amblard, C. , Carrias, J.-F. & Portelli, C. (2003) Virioplankton and microbial communities in aquatic systems: a seasonal study of two lakes of differing trophy. Freshwater Biology 48, 810-822.  Bettarel, Y. , Sime-Ngando, T. , Amblard, C. & Dolan, J. (2004) Viral activity in two contrasting lake ecosystems. Applied and Environmental Microbiology 70, 2941-2951.  Brussaard, C. P. D. (2004) Viral control of phytoplankton populations-a review. The Journal of Eukaryotic Microbiology 51, 125-138.  Chen, F. & Suttle, C. A. (1995) Amplification of DNA polymerase gene fragments from viruses infecting microalgae. Applied and Environmental Microbiology 61, 1274-1278.  Chen, F. , Suttle, C. A. & Short, S. M. (1996) Genetic diversity in marine algal virus communities as revealed by sequence analysis of DNA polymerase genes. Applied and Environmental Microbiology 62, 2869-2874.  106  Clasen, J. L. , Brigden, S. M. , Payet, J. P. & Suttle, C. A. (In Press) Evidence that viral abundance across oceans and lakes is driven by different biological factors. Freshwater Biology.  Cleugh, T. R. & Hauser, B. W. (1971) Results of initial survey of experimental lakes area, Northwestern Ontario. Journal of the Fisheries Research Board of Canada 28, 129-137.  Dorigo, U. , Jacquet, S. & Humbert, J.-F. (2004) Cyanophage diversity, inferred from g20 gene analyses, in the largest natural lake in France, Lake Bourget. Applied and Environmental Microbiology 70, 1017-1022.  Dunigan, D. D. , Fitzgerald, L. A. & Van Etten, J. L. (2006) Phycodnaviruses: A peek at genetic diversity. Virus Research 117, 119-132.  Findlay, D. L. , Kasian, S. E. M. , Hendzel, L. L. , Regehr, G. W. , Schindler, E. U. & Shearer, J. A. (1994) Biomanipulation of Lake 227 in the Experimental Lakes Area (ELA): effects of phytoplankton and nutrients. Canadian Journal of Fisheries and Aquatic Sciences 51, 2794-2807.  Frederickson, C. F. , Short, S. M. & Suttle, C. A. (2003) The physical environment affects cyanophage communities in British Columbia inlets. Microbial Ecology 46, 348-357.  Fuhrman, J. A. (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399, 541-548.  107  Gelsomino, A. , Keijzer-Wolters, A. C. , Cacco, G. & Dirk van Elsas, J. (1999) Assessment of bacterial community structure in soil by polymerase chain reaction and denaturing gradient gel electrophoresis. Journal of Microbiology Methods 38, 1-15.  Lampert, W. & Sommer, U. (1997) Limnoecology: The ecology of lakes and streams. Oxford University Press, New York.  Liu, Y.-M. , Zhang, Q.-Y. , Yuan, X.-P. , Li, Z.-Q. & Gui, J.-F. (2006) Seasonal variation of virioplankton in a eutrophic shallow lake. Hydrobiologia 560, 323-334.  Maranger, R. & Bird, D. F. (1995) Viral abundance in aquatic systems: a comparison between marine and fresh waters. Marine Ecology Progress Series 121, 217-226.  Marston, M. F. & Sallee, J. L. (2003) Genetic diversity and temporal variation in the cyanophage community infecting marine Synechococcus species in Rhode Island's coastal waters. Applied and Environmental Microbiology 69, 4639-4647.  Meints, R. H. , Van Etten, J. L. , Kuczmarski, D. , Lee, K. & Ang, A. (1981) Viral infection of the symbiotic Chlorella-like alga present in Hydra viridis. Virology 113, 698-703.  Mugnuson, J. J. , Webster, K. E. , Assel, R. A. , C.J., B. , Dillon, P. J. , Eaton, J. G. , Evans, H. E. , Fee, E. J. , Hall, R. I. , Mortsch, L. R. , Schindler, D. W. & F.H., Q. (1997) Potential effects of climate changes on aquatic systems: Laurentian Great Lakes and precambrian shield region. Hydrological Processes 11, 825-871. 108  Nauwerck, A. (1963) Die beziehungen zwischen zooplankton and phytoplankton in See Erken. Symbolae Botanicae Upsaliensis 17, 1-163.  Noble, R. T. & Fuhrman, J. A. (1998) Use of SYBR Green I for rapid epifluorescence counts of marine viruses and bacteria. Aquatic Microbial Ecology 14, 113-118.  Paul, J. H. & Kellogg, C. A. (2000). Ecology of bacteriophages in nature. Viral Ecology. C. J. Hurst. London, Academic Press: 211-246.  Proctor, L. M. & Fuhrman, J. A. (1990) Viral mortality of marine bacteria and cyanobacteria. Nature (London) 343, 60-62.  Ram, A. S. P. , Boucher, D. , Sime-Ngando, T. , Debroas, D. & Romagoux, J. C. (2005) Phage bacteriolysis, protistan bacterivory potential and bacterial production in a freshwater reservoir: coupling with temperature. Microbial Ecology 50, 64-72.  Raup, D. M. (1975) Taxonomic diversity estimation using rarefaction. Paleobiology 1, 333342.  Shearer, J. A. (1978) Two devices for obtaining water samples integrated over depth. Canadian Fisheries and Marine Service Technical Reports 772, 1-10.  109  Short, C. M. & Suttle, C. A. (2005) Nearly identical bacteriophage structural gene sequences are widely distributed in both marine and freshwater environments. Applied and Environmental Microbiology 71, 480-486.  Short, S. M. & Suttle, C. A. (1999) Use of the polymerase chain reaction and denaturing gradient gel electrophoresis to study diversity in natural virus communities. Hydrobiologia 401, 19-32.  Short, S. M. & Suttle, C. A. (2002) Sequence analysis of marine virus communities reveals that groups of related algal viruses are widely distributed in nature. Applied and Environmental Microbiology 68, 1290-1296.  Short, S. M. & Suttle, C. A. (2003) Temporal dynamics of natural communities of marine algal viruses and eukaryotes. Aquatic Microbial Ecology 32, 107-119.  Suttle, C. A. (1993). Enumeration and isolation of viruses. Handbook of methods in aquatic microbial ecology, Lewis Publishers: 121-134.  Suttle, C. A. (2000). Ecological, evolutionary, and geochemical consequences of viral infection of cyanobacteria and eukaryotic algae. Viral Ecology. C. J. Hurst. London, Academic Press: 247-296.  Suttle, C. A. (2005) Viruses in the sea. Nature 437, 356-361.  110  Suttle, C. A. , Chan, A. M. & Cottrell, M. T. (1991) Use of ultrafiltration to isolate viruses from seawater which are pathogens of marine phytoplankton. Applied and Environmental Microbiology 57, 721-726.  Suttle, C. A. & Chen, F. (1992) Mechanisms and rates of decay of marine viruses in seawater. Applied and Environmental Microbiology 58, 3721-3729.  Thingstad, T. F. , Heldal, M. , Bratbak, G. & Dundas, I. (1993) Are viruses important partners in pelagic food webs. Trends and Ecology & Evolution 8, 209-212.  UtermOhl, H. (1958) Zur Vervollkommnung der quantitativen Phytoplankton-Methodik. Mitteilung linternationale Vereinigung fuer Theoretische unde Angewandle Limnologie 9, 1-  38.  Van Etten, J. L. & Meints, R. H. (1999) Giant viruses infecting algae. Annual Reviews in Microbiology 53, 447-494.  Van Etten, J. L. , Van Etten, C. H. , Johnson, J. K. & Burbank, D. E. (1985) A survey for viruses from freshwater that infect a eukaryotic Chlorella-like green algal. Applied and Environmental Microbiology 49, 1326-1328.  Wang, K. & Chen, F. (2004) Genetic diversity and population dynamics of cyanophage communities in the Chesapeake Bay. Aquatic Microbial Ecology 34, 105-116.  111  Weinbauer, M. G. (2004) Ecology of prokaryotic viruses. FEMS Microbiology Reviews 28, 127-181.  Wen, K. , Ortmann, A. C. & Suttle, C. A. (2004) Accurate estimation of viral abundance by epifluorescence microscopy. Applied and Environmental Microbiology 70, 3862-3867.  Wetzel, R. G. (2001) Limnology: Lake and river ecosystems. Academic Press, San Diego.  Wetzel, R. G. & Likens, G. E. (1991) Limnological Analysis. Springer-verlag, New York.  Wilhelm, S. W. & Suttle, C. A. (1999) Viruses and nutrient cycles in the sea. Bioscience 49, 781-788.  Wommack, K. E. & Colwell, R. R. (2000) Virioplankton: Viruses in aquatic ecosystems. Microbiology and Molecular Biology Reviews 64, 69-114.  112  Chapter four: Identifying freshwater Phycodnaviridae and their potential phytoplankton hosts using DNA pol sequence fragments and a genetic distance analysis  A version of this chapter will be submitted for publication. Clasen, J.L.,. & Suttle, C.A. Identifying freshwater Phycodnaviridae and their potential phytoplankton hosts using DNA pol sequence fragments and a genetic distance analysis.  113  4.1 Summary Viruses that infect phytoplankton are a potentially important component of aquatic ecosystems; yet they remain relatively unstudied, particularly in lakes. In order to study viruses infecting eukaryotic phytoplankton (Phycodnaviridae) in lakes and to estimate the number of potential host species, samples were collected bimonthly from four lakes at the Experimental Lakes Area in Ontario, Canada, during the ice-free period (mid-May to midOctober) of 2004. From each lake, Phycodnaviridae DNA polymerase (pol) gene fragments were amplified using algal virus specific (AVS) primers and separated by denaturing gradient gel electrophoresis; 20 bands were extracted from the gels and sequenced. Phylogenetic analysis indicated that freshwater Phycodnaviruses are genetically different from known isolates and marine environmental sequences and belong in previously undescribed phylogenetic groups. An analysis of the genetic distances 'within' and 'between' phylogenetic groups of Phycodnavirus isolates indicated that DNA pol sequences that differ by more than 7 % at the inferred amino acid level are from viruses that infect different hosts. Application of this threshold value to environmental sequences indicated that the  Phycodnaviridae DNA pol sequences from these lakes came from viruses that infected at least nine different species of phytoplankton. A multivariate statistical analysis suggested that likely freshwater hosts included Mallomonas sp., Monoraphidium sp. and Cyclotella sp. Using this approach it should be possible to start to unravel the relationships between environmental viruses and the hosts they infect.  114  4.2 Introduction Since the 'discovery' of high abundances of viruses in oceans (Bergh et al., 1989), a growing body of research has demonstrated that viruses are dynamic members of aquatic environments (Suttle, 2000; Wommack & Colwell, 2000; Brussaard, 2004; Suttle, 2005). Viruses infecting phytoplankton are of particular ecological importance and have been implicated in bloom termination (Bratbak et al., 1993; Nagasaki et al., 1994), changing community composition (Castberg et al., 2001) and nutrient cycling (Gobler et al., 1997; Wilhelm & Suttle, 1999). Much of our current knowledge about viruses infecting phytoplankton comes from model systems. Mayer & Taylor (1979) and Van Etten et al. (1983) were among the first to isolate viruses infecting eukaryotic unicellular algae that cause lysis of Micromonas pusilla and a Chlorella-like alga, respectively. There are now about a dozen virus-host systems in culture for unicellular algae (reviewed in Brussaard, 2004). Most of these are Phycodnaviridae, a family of large (100-220 nm in diameter), polyhedral, dsDNA viruses  that infect eukaryotic algae (Wilson et al., 2005b). Undoubtedly, these isolates represent a small percentage of the Phycodnaviridae that infect eukaryotic phytoplankton. Moreover, the only cultured freshwater representatives of these viruses are ones that infect Chlorellalike algae that are typically found in symbiotic relationships with Paramecium and Hydra (Van Etten et al., 2002). Consequently, culture-independent techniques are necessary to explore the richness of Phycodnaviridae in environmental samples. The genetic richness of Phycodnaviridae has been surveyed using PCR-based methods and degenerate primers (AVS-1 and 2) that amplify a-like DNA polymerase (pol) gene fragments from Phycodnaviruses (Chen & Suttle, 1995; Chen et al., 1996). Combining  115  this method with denaturing gradient gel electrophoresis (PCR-DGGE), Short and Suttle (1999; 2002; 2003) directly compared the genetic richness of Phycodnaviruses among marine environments and found similar sequences in widely separated locations, suggesting that closely related Phycodnaviruses are cosmopolitan in distribution. This study extends previous work by phylogenetically analyzing freshwater Phycodnaviridae DNA pol sequences and assessing their relatedness to those from  Phycodnavirus isolates and marine environmental sequences. Additionally, this study determines the number and possible identity of freshwater phytoplankton hosts using a genetic distance analysis. DNA pol sequences from Phycodnavirus isolates form monophyletic clusters based on the hosts which they infect (Chen & Suttle, 1995; Brussaard et al., 2004). This suggests that for environmental DNA pol sequences, the number of  different phytoplankton species infected can be inferred from the number of discrete phylogenetic clades (Suttle, 2003). Determining the genetic identity of freshwater Phycodnaviruses and the organisms they infect is an important step in understanding the ecological significance of viruses in lakes.  4.3 Methods Sampling and virioplankton concentration  Four lakes (Lakes 224, 227, 239 and 240) located in the Experimental Lakes Area (ELA), Ontario (see Cleugh & Hauser, 1971) were sampled approximately every two weeks during the ice-free period from mid-May to mid-October, 2004 by carefully submerging (0.5 m) and filling a pre-rinsed (10 % HCl followed by lake water) 20 L polyethylene carboy.  116  Each time, an integrated water sampler (Shearer, 1978) was used to collected water from the euphotic zone (0-16 m) to determine phytoplankton community composition. The virioplankton community in each lake sample was concentrated using ultrafiltration (Suttle et al., 1991). In each case, - 18 L of lake water was pressure-filtered through several 0.45 pm pore-sized, 142 mm diameter polyvinylidine difluoride filters (Durapore, Millipore, Billerica, MA, USA) to remove large particles and organisms. The remaining particulate material in the filtrate was then concentrated - 80-fold to a final volume of - 200 mL, using a 10 kDa molecular weight cut-off tangential flow ultrafiltration cartridge (S1Y10, Millipore). The concentrates were stored at 4 ° C in the dark until processed further, which occurred within 2-7 months.  Polymerase chain reaction and denaturant gradient gel electrophoresis (PCR-DGGE) Viral DNA was extracted from each concentrate using the MoBio Ultra Clean Soil DNA extraction kit (MoBio, Carlsbad, CA, USA), which included a solution to remove inhibitors. The volume of concentrate extracted was standardized to the same initial volume of lake water (50 mL). Extracted DNA was frozen until used as template in polymerase chain reactions (PCR). The a-like DNA polymerases of Phycodnaviridae were amplified using two rounds of PCR with the algal virus specific primers (AVS-1 and 2; Chen & Suttle, 1995; Chen et al., 1996). In the first round, 5 pL of DNA template was added to a 45 pL PCR mix, containing 5.0 pL of 10 x PCR buffer, 1.5 pL of 50 mM MgC12, 1.0 pL of each of the 2.0 mM dNTPs, 1.0 pL of 10 nM AVS-1 and 3.0 pL of 10 nM of AVS-2, 0.625 U of PLATINUM ® Taq DNA polymerase (Invitrogen, Burlington, ON, Canada) and water. The negative control was prepared as above but contained no DNA template, while Micromonas pusilla virus (MpV) 117  was used as a positive control. The PCR was performed on a PCR Express Hybaid  TM  Thermal cycler (Middlesex, UK) using the following reaction conditions: initial denaturation at 95 ° C for 90 s, followed by 40 cycles of denaturation at 95 ° C for 45 s, annealing at 45 ° C for 45 s and extension at 72 ° C for 45 s; and a final extension at 72 ° C for 7 min. To confirm amplification, 10 p.I.., of PCR product and 2 i_iL of 6 x loading buffer were loaded onto a 1.5 % agarose gel flanked by a 100 by ladder (Invitrogen). The gel was run at 90 V for 60 min in TBE buffer (0.5 x), stained with Ethidium bromide (EtBr) and then viewed on a UV transilluminator (AlphaImager TM 3400, Alpha Innotech, San Leandro, CA, USA). Bands (800 bp) were plugged with a clean Pasteur pipette and placed in a sterile 0.5 mL microcentrifuge tube. DNA was eluted by adding 100 !IL of 1 x TAE and heating to 65 ° C for 60 min. Two RI., of the eluted DNA was used in a second round PCR. The 2 nd round reaction was prepared as described above, except the number of cycles was reduced to 26. Positive and negative controls were prepared as before; however, the negative control from the first round reaction was also plugged and included as a control in the second round reaction. The second round PCR products were stored at -20 ° C. Denaturing gradient gel electrophoresis (DGGE) was used to separate the AVS amplified viral dsDNA fragments by differences in their nucleotide composition (Short & Suttle, 1999). Second round PCR products (40 pL) and 6 x loading buffer (10 RL) were loaded onto a 6 to 7 % polyacrylamide gel, which had a 20 to 40 % gradient of denaturant (100 % denaturant is defined as 7 M urea and 40 % deionized formamide). Samples were run at 60 V for 15 h in a 60 ° C 1 x TAE buffer, using a D-code TM electrophoresis system (Bio-Rad, Hercules, CA, USA). Upon completion, the gel was stained in a 1 x SYBR Green Imi solution (Invitrogen) for > 3 h. Gels were visualized and photographed as described above. 118  Cloning and sequencing  To infer the genetic relationship among the Phycodnaviruses, several randomly selected DGGE bands from each lake were sequenced. Bands were excised with a Pasteur pipette and placed in sterile micro-centrifuge tubes. DNA was eluted by adding 100 [IL of 1 x TAE and heating to 95 ° C for 15 min. Two III, of the eluted DNA were used as template in a PCR with AVS primers, using the conditions previously described. Amplified DNA was purified (MinElute PCR purification kit; Qiagen Science, Germantown, MD, USA), ligated and transformed into E.coli (stain TOP10) using a TOPO TA Cloning kit (Invitrogen), following the methods recommended by the manufacturer. After an overnight incubation (37  ° C), E. coli colonies were harvested and the inserted pol fragments were amplified via T3 and T7 primers using the manufacturer's recommendations. Amplified PCR products (— 800 bps) were cleaned (MinElute PCR purification kit; Qiagen Science), diluted and submitted for sequencing. All sequencing was done by the University of British Columbia's Nucleic Acid and Protein Service Facility, using Applied Biosystems BioDye v3.1 Terminator Chemistry (Applied Biosystems, Foster City, CA, USA).  Sequence analysis  Phylogenetic analyses were used to compare the amplified DNA pol fragments to sequences from Phycodnaviridae isolates, as well as environmental samples. Two phylogenies were created. The first included the DNA pol sequences from the lakes (n = 20) and from Phycodnavirus isolates available in GenBank, including Micromonas pusilla viruses (MpV), Paramecium bursaria Chlorella viruses (PBCV), Phaeocystis globosa viruses (PgV), Chrysochromulina breviflum viruses (CbV), Emiliania huxleyi viruses (EhV), Feldmannia sp. virus (FSV) and Ectocarpus siliculousus virus (EsV) (see Table 4.1). The 119  Table 4.1. Phycodnaviridae isolates. Sequences of DNA polymerase (pol) from known Phycodnaviridae isolates were accessed through GenBank. Bold sequences were used in the environmental phylogeny (Figure 4.2).  Virus MpV PBCV  PgV  CbV EhV FSV EsV  Description Micromonas pusilla virus  Strain  PL1 SP1 SP2 Paramecium bursaria Chlorella virus PBCV1 infecting Chlorella-like strain SC1B NC64a AR93.2 NYbl CH57 AL1A CA4B SH6A NYsl AR158 NY2A Paramecium bursaria Chlorella virus CWM1 infecting Chlorella-like strain Pbi CVB1 CVR1 CVA1 Phaeocystis globosa virus 3T 4T 6T 10T 5T 7T Chrysochromulina breviflum virus PW3 PW1 Emiliania huxleyi virus EHV 208 Feldmannia sp. virus FSV Ectocarpus siliculousus virus EsV  GenBank number U32982 U32975 U32976 AJ890364 AF344238 AF344203 AF344234 AF344210 AF344198 AF344209 AF344239 AF344235 AF344202 AF344230 AF344214 AF344212 AF344215 AF344211 AY345136 AY345137 AY345139 AY345142 AY345138 AY345140 U32984 U32983 U42580 AF453867 AF013260 AF204951  120  second phylogeny contained all the sequences from the lake samples (n = 20), some Phycodnaviruses isolates, all the marine environmental AVS sequences in GenBank, as well as, the top 20 matches to each of the 20 lake sequences in the Global Ocean Survey (Rusch et al., 2007) NCBI environmental database (see Tables 4.1 and 4.2).  For each phylogeny, translated sequences were aligned in CLUSTAL X, and edited by eye in BioEditTm to move misplaced amino acids and remove uninformative gaps. The edited alignments were used to construct neighbour-joining (NJ) and maximum likelihood (ML) trees. Bootstrapped NJ trees (reps = 1000) were constructed in PAUP version 4.0b8, while quartet puzzling ML trees were constructed in Tree Puzzle (ver 5.2). Phylogenetic trees were drawn using TreeView version 1.6.6. In both cases, tree topologies are shown with NJ bootstrap and ML support > 50 indicated at the nodes (NJ/ML). African Swine Fever Virus (ASFV) was used as an outgroup in both of the phylogenetic analyses (as in Short & Suttle, 2003).  Determining the representative amplification of Phycodnaviruses by AVS To confirm that the AVS primers amplify gene fragments that are representative of the overall richness of the Phycodnaviridae, DNA pol gene fragments from the complete genomes of the Phycodnaviruses PBCV-1 and EhV (accession numbers U42580 and AJ890364, respectively) were blasted against the metagenomic Global Ocean Survey (hereafter GOS) NCBI database (Rusch et al., 2007). The genomes and GOS sequences were obtained without the use of the AVS primers and are, therefore, free from any potential skewed amplication bias associated with the primers. The top 20 matches from the blast searches conducted with each genome were combined and any GOS sequences lacking the  121  Table 4.2. Environmental samples. Freshwater DNA polymerase (pol) sequences were obtained from samples collected from four lakes at the Experimental Lake Area (L224, L227, L239 and L240). Other environmental DNA pol sequences from the Gulf of Mexico, and the Pacific and Southern Oceans were accessed through GenBank. Metagenomic data from the Global Ocean Survey was accessed through the NCBI environmental database; GOS locations are not identified since sequences often came from several different locations (see http://camera.calit2.net/index.php). Sample prefix  L224 L227 L239 L240 BS ESO2 JPays MI OTU PS SI SO GOS  Location  Reference  Lake 224, ELA, ON, CA Lake 227, ELA, ON, CA Lake 239, ELA, ON, CA Lake 240 , ELA, ON, CA Barkley Sound, BC, CA Marine aerosols, East Sea, Korea Jericho Pier, BC, CA Malaspina Inlet, BC, CA Gulf of Mexico,TX, USA Pendrell Sound, BC, CA Salmon Inlet, BC, CA Southern Ocean Halifax to Galapogos Islands  This study This study This study This study (Short & Suttle, 2002) (Cho et al., unpublished)  * * * * t AY436587 to AY436589  (Short & Suttle, 2003) (Short & Suttle, 2002) (Chen et al., 1996) (Short & Suttle, 2002) (Short & Suttle, 2002) (Short & Suttle, 2002) (Rusch et al., 2007)  AY145089 to AY145098 t U36931 to U36935 t t t  GenBank number  AACY020457048, AACY020716371, AACY020076678, AACY020168926, AACY020008685, AACY023220264, AACY022626042, AACY020325924, AACY021388666, AACY020002798, AACY020040066, AACY020013999, AACY020462121, AACY021701173, AACY021524699, AACY022625378, AACY023197486, AACY023235187, AACY020017444, AACY020013609, AACY020006926, AACY020028759, AACY020011590, AACY020009009, AACY020010354, AACY021532305, AACY020038007, AACY022638056, AACY023984140, AACY020034483, AACY023336016, AACY023234611, AACY020069902, AACY020070781, AACY023362373  *Accession numbers EU408225 through EU408244 tAccession numbers AF405572 through AF405604  122  conserved DNA pol YGDTDS (Asp-Asp) motif were removed. The remaining GOS sequences along with the Phycodnavirus isolates were used to construct a phylogeny using the methods described above. The topology of this GOS tree was then visually compared to the one created using all of the environmental sequences obtained with the AVS primers to determine if the primers are introducing a bias by preferentially amplifying only part of the Phycodnavirus community.  Predicting the number of potential hosts using a genetic distance analysis The genetic distance was compared between cultured Phycodnavirus isolates to determine the distances between those that infect the same species relative to those that infect different species. Using the phylogenetic methods described above, a ML tree of the Phycodnavirus isolates was generated with sequences from isolates of PBCV-1 (including viruses infecting Chlorella stains NC64a and Pbi), EhV, MpV, PgV and CbV (see Table 4.1). The maximum branch lengths within each of the six groups of viruses infecting the same host were determined (hereafter `within'), as were, the minimum branch lengths between all possible combinations of the different hosts groups (hereafter `between'). A discriminant analysis (DA) was conducted with the distance data; DA predicts the likelihood that an `unknown' sample belongs in a particular group and, therefore, measures the robustness of the grouping criteria. Finally, the 'within' and 'between' group means, S.E., S.D. and 95 % upper and lower confidence intervals were calculated and compared (by ANOVA) to determine the maximum genetic distance separating viruses infecting the same species. Sequences separated by distances larger than this threshold were assumed to have originated from viruses that infect different phytoplankton species.  123  Predicting the identity of potential hosts using multivariate statistics Binary matrix of phytoplankton  Eukaryotic phytoplankton community composition was determined by D.L. Findlay at the Freshwater Institute, Winnipeg, Manitoba, Canada. Briefly, 125 mL of the integrated (0-16 m) water samples were fixed in Lugol's and 10 mL sub-samples were settled for 24 h in Utermal settling chambers (1958) as outlined by Nauwerck (1963). Cells were counted and identified using an inverted microscope with phase-contrast and the 10 x and 40 x objectives. The composition of the phytoplankton community on each sampling date was converted into a presence/absence binary matrix.  Binary matrix of viruses  Freshwater sequences from viruses thought to infect different hosts based upon the genetic distance analysis described above were traced back to the plugged DGGE bands, since each of the sequenced DNA pol fragments was associated with a band found at a particular gel location. On each sampling date, the presence or absence of each band was then converted into a binary matrix using GelCompar II' (Applied Maths).  Identity of potential hosts  A monotonic multiple dimensional scaling analysis (MDS) with the presence/absence data for phytoplankton species and viral sequence fragments was conducted to infer potential hosts. Phytoplankton species and viral pol sequences that occupied the same space in an MDS analysis co-occurred 100 % of the time. Co-occurrence of viral fragments and phytoplankton species strongly suggests that the two are associated but does not take into  124  account any time lag that may occur between virus and host and should, therefore, be considered a conservative estimate.  4.4 Results and Discussion Several significant results stem from the phylogenetic analysis of freshwater environmental Phycodnaviridae sequences. First, freshwater Phycodnaviruses form clusters that are largely distinct from both cultured isolates and their marine counterparts. Second, 99 % of the marine and freshwater environmental sequences were more closely related to viruses infecting Micromonas pusilla than other Phycodnaviridae isolates. Third, the genetic distance analysis indicated that the freshwater environmental sequences likely originated from viruses that infect at least nine different species of phytoplankton. The significance and ecological importance of these findings are discussed below.  Phylogenies of freshwater pol sequences While the DNA pol sequences from the four lakes at the ELA fall within the Phycodnaviridae, most sequences clustered into several strongly supported clades that were  distinct from the Phycodnavirus isolates (clusters I, II, III on Figure 4.1). The one exception was L227September2a which clustered with MpV; this sequence was likely a contaminant since it was 99.1 % similar to MpV-SP1, a virus that infects a marine phytoplankton. There was no obvious temporal or spatial pattern among the remaining freshwater sequences, aside from Cluster I which was mostly composed of sequences from Lake 239. Interestingly, the freshwater environmental sequences were not most closely related to sequences belonging to  125  Figure 4.1. Maximum likelihood tree of DNA polymerase inferred amino acids from fresh waters. The shown maximum likelihood (ML) tree topology of lake samples (n= 20) and Phycodnaviridae isolates was determined by quartet puzzling. Neighbour joining bootstrap values (rep =1000) and corresponding ML support values > 50 are indicated at the nodes (NJ/ML). The vertical line and roman numerals to the right of the tree indicate the three freshwater clades. Bold sequences indicated by L (i.e. L224, L227, L239 and L240) are from freshwaters, and the labels include the lake and collection date. Groups of Phycodnaviruses isolates are also indicated to the right of the tree and include: MpV = Micromonas pusilla virus; PBCV (NC64a) = Paramecium bursaria Chlorella virus infecting the NC64a strain of Chlorella-like algae; PBCV (Pbi) = Paramecium bursaria Chlorella virus infecting the Pbi strain of Ch/ore//a-like algae; PgV = Phaeocystis globosa virus; CbV = Chrysochromulina breviflum virus; EhV = Emiliania huxleyi virus; FSV = Feldmannia sp. virus; EsV = Ectocarpus siliculousus virus (see Table 4.1). ASFV = African Swine Fever Virus (Asfarviridae). The scale bar represents the number of amino acid substitutions per residue.  126  L240July13 L239June16b ^L239July1  ^100/9  1  L239Ju1y26c ^100/94^  L239June16a  L239June4  -/87  ■ L240August10  L239June16c  55/98  L239Ju1y26b  NM■  L224July20c  II  L239JuIy15 L239 L227Septem ber2b  100/98  L224July20a L224July20b L227Ju1y26a^  100/95  III  L240June29 L240June15  99/92  .111111=11  L240August24 MpVSP1 100/99 L227September2a  1MpV  MpVPL1 MpVSP2  /87  PBCV1SH6A PBCV1SC1B PBCV1AR932 PBCV1NYb1  100/99  PBCV1CH57 P BCV1 AL 1A 100/90 .1111■  88/57  100/99  PBCV1CA4B  ...1  ...4  PBCV (NC64a)  P BCV1 BCV1NYs1  PBCV1AR158  wit PBCV1NY2A PBCV1CWM 1  100/100  PBCV1CVB1  PBCV (Pbi)  PBCV1CVR1 PBCV1CVA1  PgV1OT PgVO4T PgV06T 100/92  100/90  PgV03T  PgV  PgV05T PgV07T 100/100  CbVPW3  CbVPW1 100/93 I  100/8  EhV208  16 EhV h  1 CbV  EhV  I^  FSV EsV ASFV  0.1  127  freshwater isolates in the PBCV group, but to sequences from viruses (MpV) infecting the cosmopolitan marine phytoplankton Micromonas pusilla. The freshwater DNA pol sequences were compared to AVS amplified marine environmental DNA pol fragments and metagenomic data from the Global Ocean Survey (GOS; Table 4.2; Figure 4.2). Both NJ and ML phylogenetic analyses showed similar tree topologies, with most of the freshwater samples falling into different clusters than the marine AVS amplified fragments (Figure 4.2). In contrast, while most the metagenomic sequences included from the GOS database clustered with the marine environmental AVS sequences of Chen et al. (1996), Short & Suttle (2002, 2003) and Cho et al. (unpublished), two clustered with the lake sequences (underlined sequences on Figure 4.2). However, both of these sequences were from Lake Gatun, Panama, the only freshwater station sampled on this leg of the GOS cruise (Rusch et al., 2007 and http://camera.calit2.net/index.php) . Blast searches using the freshwater sequences and the metagenomic data obtained from Lake Gatun, Panama revealed that only two of the 20 sequences with the conserved YGDTDS motif clustered with the Phycodnaviridae and both of these fall within a freshwater Glade (data not shown), providing further evidence of the differences between freshwater and marine Phycodnaviruses. There is also evidence that phage genes differ between marine and fresh waters. For example, cyanophage structural gene sequences (g20) from distant freshwater environments cluster into monophyletic groups (Short & Suttle, 2005). Similarly, freshwater cyanophage psbA genes that encode core photosynthetic proteins cluster into a separate Glade (Chenard, 2007). Together, these studies suggest an ancient divergence between marine and freshwater virioplankton.  128  Figure 4.2. Neighbour-joining tree of DNA polymerase (pol) sequences from fresh waters and other environmental samples. This neighbour-joining (NJ) tree topology included lake samples, some Phycodnaviridae isolates, marine environmental samples and GOS blast hits (see Tables 4.1 and 4.2). Neighbour-joining bootstrap values (reps = 1000) and corresponding ML support values > 50 are indicated at the nodes (NJ/ML). The freshwater sequences (bold) and cluster (roman numerals), Phycodnavirus isolates and the ASFV outgroup are as described in the caption for Figure 4.1, except MpV is boxed. Marine environmental sequences (italicized) are from several locations, including the Gulf of Mexico, and the Pacific and Southern Oceans (see Table 4.2). The top matches to each of the lake DNA pol sequences in the Global Ocean Survey (GOS) are indicated by their accession numbers. Underlined sequences are GOS sequences that cluster within a freshwater Glade. The scale bar represents the number of amino acid substitutions per residue.  129  BSA9975 BSA998 esAgge S0982 BSB994 BSA997a BSA997b ptilM831711gbIAACY020457048 99/83  S0984 0981 SIAM g813363441610b1AACY020076678 ESO2AV1 9813370240915b1AACY020008685  gill 31072082101AACY022626042 97/73 gil 3367 95/  g 0323213191g NAACY021388666  OTU2  CY020040066  OTU1 61'21n2.5F143.0m 9MbrA . CY020482121 ^ OTU3 MIB991 BSA992 981321819j349IgNAACY021524699  64/-  51/  JPays52 JPays51 JPays42 g81304971121gbIAACY023197486 90/72 01336936501g blAACY020017444 JPays37 9i1133704168101AACY020006926 JPays22 gil1336995041gb1AACY020011590 9i1133700740IghlAACY020010354 JPays38 BSA995 33730871gbIAACY020038007 g861297087971gbIAACY023984 140 gil1 30355916IgbIAACY023336016 gi1133641192 1gbIAACY020069902 981303295591gb1A9CY023362373 1/89 gil1 33840313lgb gil1304596431gbIA 8 CY023234611 MIB092 ^ 881336766111g NAACY020034483 L224July20a L227July26a 100/99^L224July2013 L240June29a L240June15 L240Auguat24 0131060011 jgbIAACY022638056 ESO2AV7 9032177557 IgbIAACY021532305 JPays53 96/66 981337020851g blAACY020009009 ESO2AV3 S0985 gil1338823351gb1AACY020028759 336974851AAACY020013609 ^ JPays55 91113045906719b1AACY023235187 981310727501gbIAACY022625378  III  74/62  70/  100/72  30/  86/65 60/53  75/  91 1 32010888IghlAACY021701173 gi11336970951981AACY020013999 BSA994 BSA991 BSB991 JPays64 JPays65 BSA993 gil1337082961gbIAACY020002798 ^ gipaM295101AACY020325924  ir  PS89112 PSB99I gi1130473990jgbIAACY023220264 ^ PSC992 L239June16a L239July28c L239June4 L239June16b L240July13 L239July1 L240Auguat10 L239June18c ^ ail 1335420601ab1AACY020166926 gi1133003601KNAACY020716371 100 ^ L239July2613 614■Es L224July20c ^ L239July15 100/93^L239  II  ^ L2275eptember2b ^ PBS993 • BSB992 PBCV1 PBCVINY2A PBCVICVA1 ^ 071/5 100/89 ^  ^. gVORPVPW3 FSV EsV ASFV  (los changes  130  Determining the representative amplification of Phycodnaviruses by AVS Most (99 %) of the environmental samples were more closely related to MpV than the other Phycodnavirus isolates (Figure 4.2). This topology could occur if the primers amplified MpV-like sequences preferentially to those from other Phycodnaviridae. This possibility was tested by blasting partial DNA pol fragments from the complete genomes of the Phycodnaviruses EhV and PBCV-1 (Wilson et al., 2005b; Dunigan et al., 2006) against the metagenomic GOS database. Not only do these sequences fall outside of the Glade containing MpV (see Figure 4.1) but they were also generated without AVS primers. If the AVS amplicons are representative of the natural richness of the Phycodnaviridae, then significant matches from the blast searches with the partial genomes should also be more closely related to MpV than other isolates. As was the case for the AVS environmental amplified sequences, the GOS sequences (n = 20) from the searches with the partial genome sequences were more closely related to MpV (100/99) than other Phycodnaviruses (Figure 4.3). Therefore, the primers are not responsible for the observed distribution of the environmental sequences and Figure 4.2 reflects the true genetic richness of Phycodnaviridae.  Number of potential host species Since Phycodnavirus isolates infecting the same host species cluster in monophyletic groups (see Figure 4.1, Chen & Suttle, 1995; Brussaard et al., 2004), the number of host species should be reflected by the number of discrete clades (Suttle, 2003), which have a genetic distance greater than that which separates viruses that infect the same species. Distances 'within' and 'between' the monophyletic groups of Phycodnavirus isolates were  131  Figure 4.3. Testing the representative amplification of Phycodnaviruses by AVS. A maximum likelihood tree of sequences obtained by blast hits of the DNA polymerase fragments from the genomes of Phycodnaviruses against the GOS database. PBCV-1 (U42580) and EhV (AJ890364) were used in the analysis. The shown maximum likelihood (ML) tree topology of GOS sequences and Phycodnaviridae isolated was determined by quartet puzzling. Neighbour-joining bootstrap values (rep = 1000) and corresponding ML support values > 50 are indicated at the nodes (NJ/ML). Phycodnavirus isolates and the ASFV outgroup are as described in the caption for Figure 4.1 and Table 4.1. The GOS sequences are indicated by their accession numbers. The scale bar represents the number of amino acid substitutions per residue.  132  10 /  MpVSP1 gi11332580981gbIAACY02046212  98  MpVPL1 4MpVSP2  MpV  92/64  a gi[1336995041gbIAACY020011590  it  911133700740IgbIAACY0200010354 gill 303559161gbIAACY023336016  6,...E01332631711g . b IAACY020457048 /72 73/8^011336710281gbIAACY020040066  gil 1336344161gbIAACY020076678 56/91^• g i ll  collr  303295591g blAACY023362373  911133641192Ig blAACY020069902 gill 30459643jgbIAACY023234611 gill 33693650101AACY020017444  gill 337041681gbIAACY020006926  100/99  ^ gi11336766111gbIAACY020034483 6^gi11304590671gbIAACY023235187  gill 336970951gbIAACY020013999  .194011336974851g blAACY020013609 gill 337020851gbIAACY020009009 gi11337082961gbIAACY020002798 ^ gi11304739901gbIAACY0232220264 011333942951gbjAACY02035924 PBCV1SH6A 53/7f  PBCV1AR932 PBCV1SC1B PBCV1 PBCV1CA4B PBCV1NYb1  PBCV (NC64a)  PBCV 1AL1A 100/9^ PBCV1CH57  PBCV1AR158  t  100/9 72/61  PBCV1NYs1 PBCV1NY2A  inciaia .PBCV1CWM 1 PBCV1CVB1  PBCV1CVR1  PBCV (Pbi)  PBCV1CVA1 PgV1OT 100/90  y, PgV06T  74/59  PgV03T PgVO4T 100/8  PgV  PgV05T PgV07T CbVPW3  loom' E hV2081  Ehv  CbVPW1I  CbV  EhV I ^FSV EsV AS FV  133  determined from branch lengths (Figure 4.4a). Since there was a significant difference between the 'within' and 'between' distance groups (Figure 4.4b, c; ANOVA, F 1 , 19 = 43.966and p < 0.0001), both groups were used as predictors in a discriminant analysis. Overall, 'unknown' distances were placed in the correct group 95 % of the time; however, the 'within' distances were always (100 %) placed correctly. As a result, the upper 95 % confidence interval around the mean 'within' distance (0.081 amino acid substitutions per residue) was used as the threshold to distinguish among discrete clades of viruses. Therefore, if the distance between two viral sequences was less than 0.081 substitutions per amino acid, the DNA pol sequences were assumed to have originated from viruses that infect the same host species. Applying this threshold to the environmental sequences indicated that the viral sequences came from viruses infecting nine different freshwater and 20 different marine hosts (Figure 4.5). Although the 'within' distances are similar among all six isolate groups (see Figure 4.4), it should be noted that the threshold distance was based upon the extant Phycodnaviridae sequences and it may be necessary to adjust the distance as more sequence data becomes available. Nonetheless, this analysis should provide a reasonable approximation of the number of hosts that are infected by the viruses from which these sequences were derived.  Species identity of potential hosts In an attempt to identify some of the nine potential freshwater hosts resolved by the analysis of genetic distance, multivariate statistics were used to assess co-occurrence between the viral clades and phytoplankton species. A multiple dimensional scaling analysis (MDS) with both the phytoplankton and virus presence/absence data (phytoplankton data available  134  a  1.5  CbV  EhV  PgV  PBCV (NC64a)  PBCV (Pbi)  CbV  0.01846  PgV  0.19729  0.01199  EhV  1.08917  1.0129  0.01212  PBCV (Pbi)  1.21224  1.13597  1.13175  0.03074  PBCV (NC64a)  1.612868  1.65970  1.65548  0.73471  0.11830  MpV  1.03029  0.95402  0.94980  0.75331  1.27704  b  C  g to  0 a)  C  • 0.0  0.03753  `Within' Groups  `Between' Groups  DA (%)  100  93  Mean  0.038  1.094  S.D.  0.041  0.383  S.E.  0.017  0.099  95 % Upper CI  0.081  1.306  95 %  -0.004  0.882  ANOVA, F119= 44.0, p < 0.0001  e0) C  MpV  Lower CI 'Within' Groups' Between' Groups Figure 4.4. Genetic distance Within' and 'between' Phycodnavirus groups. a) Distances either Within' a group of Phycodnavirus isolates infecting the same host or 'between' the different host groups were determined from branch lengths on a ML tree. 'Within' groups distances are bold. b) Box plots of 'within' and 'between' host group distances; means, quartiles and outliers are indicated, as are the ANOVA statistics from a comparison between the groups. c) Discriminant Analysis (DA) results and basic statistics on each group, including mean, standard deviation (S.D.), standard error (S.E.), and the upper and lower 95 % confidence interval around the mean. The threshold value used in further analysis is bolded.  135  Figure 4.5. Number of host species. The neighbour-joining tree of DNA polymerase gene fragments from fresh waters and other environmental samples (see Figure 4.2) was modified using the threshold genetic distance of 0.081 amino acid substitutions per residue to identify viral clades that were assumed to infect different hosts. Potential freshwater hosts are indicated by bold upper case letters to the right on the tree; while marine hosts are labeled with italic lower case letters. GOS sequences were not included in this analysis unless the sequences clustered with environmental AVS sequences.  136  99/83  BSA997c BSA 998 BSA998 S0982 BSB994 BSA997a BSA997b 9i113328317118b1AACY020457048 Pt C991 50984 It) 0981 S1A991 gil1338344181gbIAACY020078878 ESO2AV1 gil133702491gbIAACY020008885  a  IC  ai1131072082leblAACY022828042 1323213191abiAACY021388888  97/7  OTU2  4  33eggilAA ui CY02111. 0068 95/-  gil1332580981gbIAACY020482121 L227September2a 01 3  1  m/B9g,^  h  N IAAC Y021524899  BSA992 g i1132181.9L 84/-  e  JPays52 JPays51 JPays42 911130497112IgNAACY023197486 90/72^911133893850IghlAACY020017444 JPays37 gi11337041881gbIAACY020008928 JPays22 gi11338995041gbIAACY020011590 9 1 1 133 7 0074 0 Ig bIAACY020010354 JPays38 BSA 995 01338730871g blAACY020038007 gil1297087971gbIA8CY023984140 gill 30355918IgbIAACY023338018 0111338411921gbIAACY020089902 gil130329559IgbIAACY023382373 1/89^gill 338403131gIAAACY020070781 migi1193204498431gbIAACY023234811  k  5l/  9 , 113387881119bIAACY020034483 L224July20a L227Ju1y213a 100/99^L224July20b L240June29a L240June15 L240August2 gi11310800111gblAACY022838058 ESO2AV7 91113217755719bIAACY021532305 JPays53^ 96/6"' 0/11337020851gbIAACY020009009  ^  B  177  SEMA " • 9i11338823351gbIAACY020028759 911133897485101AACY020013609 JPays55.411^ 82 ^ 91113045908710IAACY023235187 gill 3107275019 blAACY022825378  74/-  70/5  ^  6  0/9  100/72  88/85 80/53  psagn  gil1=3992941g6IAACY023220284 L239June113a L239July213c L239June4 L239June16b L240July13 L239July1^• L240August10^ L239June18c 9i1133542080101AACY020188928 941330038011g blAACY020718371 I13b LJuy 239 12 L224Ju1y20c L239July154-100/9 , L239  E G  00 /8  11  L227September2b 1I BSB992 Is PBCV1 PBCV1NY2A PBCV1CVA1 OTU54- t ^ cs,i,VPW3  d PBS993  100/89 ^  r  D  r  75/-  0  9111320108881961AACY021701173 903369709519 b1AACY020013999 BSA894 BSA991 BSBI791 JPays84 1 JPays65 BSA993 gi1133708298IghlAACY02000279 g 13339429519 blAACY020325924 PS■9994 PS8992 g  I  rsV  EsV  ASFV 0.05 changes  137  from the Freshwater Institute; while virus data can be found in Appendices 1,2,3 and 5) revealed several potential phytoplankton hosts, including Mallomonas sp., Chrysosphaerella sp., Monoraphidium sp., Synedra sp. (now called Fragilaria sp.), Cyclotella sp., Trachelomonas sp. and Peridinium sp. Despite the conservative co-occurrence value used to  identify potential hosts, which does not take into account any potential lag between viruses and host populations, this analysis recognized seven of the nine hosts predicated from the distance analysis. These results are interesting because hosts from several different and ecologically important phytoplankton groups were identified, including dinoflagellates, chlorophytes, chrysophytes and diatoms. Being able to infer the number and identity of phytoplankton infected by specific groups of viruses should help unravel the complex interactions between hosts and viruses in nature.  Implications The results of this study have evolutionary and ecological implications for the field of aquatic virology. This first phylogenetic analysis of environmental freshwater Phycodnaviridae DNA pol sequences further emphasizes the fundamental difference between  the viruses present in lakes and oceans. This difference is evolutionary interesting as it represents an ancient divergence between the two virioplankton communities. Further, although the relationship between phytoplankton and their viruses remain elusive, knowing the number and identity of hosts that are susceptible to viral infection could be helpful. This information can be used to focus research on particular phytoplankton species, which will not only reveal the complex interactions between viruses and hosts but may further our understanding of the impact viruses have on structuring phytoplankton communities.  138  4.5 Acknowledgments We gratefully acknowledge the staff of the Experimental Lakes Area and the Freshwater Water Institute (Winnipeg, MB, Canada), particularly Dave L. Findlay for his phytoplankton data. Thanks to past and present Suttle lab members for numerous helpful discussion and especially Caroline Chënard, Johan Vande Voorde and Chris Payne for reviewing and editing drafts of this manuscript. Financial support came from the Natural Sciences and Engineering Research Council of Canada through a graduate student fellowship to J.L.C. and a Discovery Grant to C.A.S.  139  4.6 References Bergh, 0. , Borsheim, K. Y. , Bratbak, G. & Heldal, M. (1989) High abundance of viruses found in aquatic environments. Nature 340, 467-468.  Bratbak, G. , Egge, J. K. & Heldal, M. (1993) Viral mortality of the marine alga Emiliania huxleyi (Haptophyceae) and termination of alga blooms. Marine Ecology Progress Series 93,  39-48.  Brussaard, C. D. P. , Short, S. M. , Frederickson, C. M. & Suttle, C. A. (2004) Isolation and phylogenetic analysis of novel viruses infecting the phytoplankton Phaeocystis globosa (Prymnesiophyceae). Applied and Environmental Microbiology 70, 3700-3705.  Brussaard, C. P. D. (2004) Viral control of phytoplankton populations-a review. The Journal of Eukaryotic Microbiology 51, 125-138.  Castberg, T. , Lasen, A. , Sandaa, R. A. , Brussaard, C. P. D. , Egge, J. K. , Heldal, M. , Thyrhaug, R. , van Hannen, E. J. & Bratbak, G. (2001) Microbial population dynamics and diversity during a bloom of the marine coccolithophorid Emiliania huxleyi (Haptophyta). Marine Ecology Progress Series 221, 39-46.  Chen, F. & Suttle, C. A. (1995) Amplification of DNA polymerase gene fragments from viruses infecting microalgae. Applied and Environmental Microbiology 61, 1274-1278.  140  Chen, F. , Suttle, C. A. & Short, S. M. (1996) Genetic diversity in marine algal virus communities as revealed by sequence analysis of DNA polymerase genes. Applied and  Environmental Microbiology 62, 2869-2874.  Chenard, C. (2007). Phylogenetic analysis of genes encoding photosynthesis proteins in cyanophage isolates and natural virus communities. Earth and Ocean Sciences. Vancouver, University of British Columbia. Master's.  Cleugh, T. R. & Hauser, B. W. (1971) Results of initial survey of experimental lakes area, Northwestern Ontario. Journal of the Fisheries Research Board of Canada 28, 129-137.  Dunigan, D. D. , Fitzgerald, L. A. & Van Etten, J. L. (2006) Phycodnaviruses: A peek at genetic diversity. Virus Research 117, 119-132.  Gobler, C. J. , Hutchins, D. A. , Fisher, N. S. , Cosper, E. M. & Sanudo-Wilhelmy, S. A. (1997) Release and bioavailability of C, N, P, Se and Fe following viral lysis of a marine chrysophyte. Limnology and Oceanography 42, 1492-1504.  Mayer, J. A. & Taylor, F. J. R. (1979) A virus which lyses the marine nanoflagellate  Micromonas pusilla. Nature 281, 299-301.  Nagasaki, K. , Ando, M. , Itakura, S. , Imai, I. & Ishida, Y. (1994) Viral mortality in the final stage of Heterosigma akashiwo (Raphidophyceae) red tide. Journal of Plankton Research 16, 1595-1599. 141  Nauwerck, A. (1963) Die beziehungen zwischen zooplankton and phytoplankton in See Erken. Symbolae Botanicae. Upsaliensis 17, 1-163.  Rusch, D. B. , Halpern, A. L. , Sutton, G. , Heidelberg, K. B. , Williamson, S. , Yooseph, S. , Wu, D. , Eisen, J. A. , Hoffman, J. M. , Remington, K. , Beeson, K. , Tran, B. , Smith, H. , Baden-Tillson, H. , Stewart, C. , Thorpe, J. , Freeman, J. , Andrews-Pfannkoch, C. , Venter, J. E. , Li, K. , Kravitz, S. , Heidelberg, J. F. , Utterback, T. , Rogers, Y.-H. , Falc , n, L. I. , Souza, V. , Bonilla-Rosso, G. , Eguiarte, L. E. , Karl, D. M. , Sathyendranath, S. , Platt, T. , Bermingham, E. , Gallardo, V. , Tamayo-Castillo, G. , Ferrari, M. R. , Strausberg, R. L. , Nealson, K. , Friedman, R. , Frazier, M. & Venter, J. C. (2007) The Sorcerer II Global Ocean Sampling Expedition: Northwest Atlantic through Eastern Tropical Pacific. Public Library of Science Biology 5, 398-431.  Shearer, J. A. (1978) Two devices for obtaining water samples integrated over depth. Canadian Fisheries and Marine Service Technical Reports 772, 1-10.  Short, C. M. & Suttle, C. A. (2005) Nearly identical bacteriophage structural gene sequences are widely distributed in both marine and freshwater environments. Applied and Environmental Microbiology 71, 480-486.  Short, S. M. & Suttle, C. A. (1999) Use of the polymerase chain reaction and denaturing gradient gel electrophoresis to study diversity in natural virus communities. Hydrobiologia 401, 19-32.  142  Short, S. M. & Suttle, C. A. (2002) Sequence analysis of marine virus communities reveals that groups of related algal viruses are widely distributed in nature. Applied and Environmental Microbiology 68, 1290-1296.  Short, S. M. & Suttle, C. A. (2003) Temporal dynamics of natural communities of marine algal viruses and eukaryotes. Aquatic Microbial Ecology 32, 107-119.  Suttle, C. A. (2000). Ecological, evolutionary, and geochemical consequences of viral infection of cyanobacteria and eukaryotic algae. Viral Ecology. C. J. Hurst. London, Academic Press: 247-296.  Suttle, C. A. (2003). Viral diversity and its implications for infection in the sea. CIESM. Ecology of marine viruses, CIESM Workshop Monographs n°21.  Suttle, C. A. (2005) Viruses in the sea. Nature 437, 356-361.  Suttle, C. A. , Chan, A. M. & Cottrell, M. T. (1991) Use of ultrafiltration to isolate viruses from seawater which are pathogens of marine phytoplankton. Applied and Environmental Microbiology 57, 721-726.  Utermiihl, H. (1958) Zur Vervollkommnung der quantitativen Phytoplankton-Methodik. Mitteilung Internationale Vereinigung fuer Theoretische unde Angewandle Limnologie 9, 1-  38.  143  Van Etten, J. L. , Burbank, D. W. , Kuczmarski, D. & Meints, R. H. (1983) Virus infection of culturable Chlorella-like algae and development of a plaque assay. Science 219, 994-996.  Van Etten, J. L. , Graves, M. V. , Mullerm, D. G. , Boland, W. & Delaroque, N. (2002) Phycodnaviridae — large DNA alga viruses. Archives of virology 147, 1479-1516.  Wilhelm, S. W. & Suttle, C. A. (1999) Viruses and nutrient cycles in the sea. Bioscience 49, 781-788.  Wilson, W. H. , Van Etten, J. L. , Schroeder, D. S. , Nagasaki, K. , Brussaard, C. P. D. , Delaroque, N. , Bratbak, G. & Suttle, C. A. (2005b). Phycodnaviridae Virus Taxonomy: Classification and Nomenclature of Viruses, Eighth Report of the International Committee on the Taxonomy of Viruses. C. M. Fauquet, M. A. Mayo, J. Maniloff, U. Desselberger and L. A. Ball. San Diego, Elsevier Academic Press: 163-175.  Wommack, K. E. & Colwell, R. R. (2000) Virioplankton: Viruses in aquatic ecosystems. Microbiology and Molecular Biology Reviews 64, 69-114.  144  Chapter five: An evaluation of the utility of the modified dilution experiment to estimate viral-mediated phytoplankton mortality  A version of this chapter will be submitted for publication. Clasen, J.L.,. & Suttle, C.A. An evaluation of the utility of the modified dilution experiment to estimate viral-mediated phytoplankton mortality  145  5.1 Summary The high abundance of viruses in aquatic ecosystems emphasizes the need to quantify viral-mediated mortality of phytoplankton and bacteria, in order to understand their impact on community composition, nutrient cycling and genetic exchange. In one approach that has been used to estimate viral mortality of phytoplankton, the impact of viral lysis is reduced in a series of dilutions. The present study uses both a simple modeling exercise and field studies to evaluate the utility of the dilution experiment to estimate viral-mediated phytoplankton mortality. An analysis of the model outputs indicated that experimental results are highly sensitive to variations in burst size, host growth rate, length of the lytic cycle and the fraction of host cells infected at the beginning of the experiment. Experiments conducted in the Experimental Lakes Area, Ontario failed to satisfy the assumptions of the dilution experiment, resulting in the inability to accurately calculate apparent phytoplankton growth rates. This failure, coupled with the model analysis, suggests that the modified dilution experiment should be used and interpreted carefully.  146  5.2 Introduction Viruses are abundant and dynamic members of aquatic ecosystems and are significant mortality agents of bacteria and phytoplankton in oceans and lakes (see Fuhrman, 1999; Wommack & Colwell, 2000; Suttle, 2005; Suttle, 2007). Through lysis, viral infections influence nutrient cycling, community composition and horizontal gene transfer (see Fuhrman, 1999; Paul, 1999; Wilhelm & Suttle, 1999; Wommack & Colwell, 2000; Suttle, 2005; Suttle, 2007). As a result, quantitative estimates of viral-mediated mortality are necessary to understand the role of viruses in aquatic ecosystems. Since phytoplankton occupy the base of most aquatic food webs, there is particular interest in quantifying viralmediated mortality of phytoplankton. Two approaches that have been used include determining the frequency of visibly infected cyanobacteria by TEM (Proctor & Fuhrman, 1990) and measuring the effects of viral enrichment on primary production (Suttle, 1992). However, these approaches have limited applicability (Suttle, 2005) and are difficult to adapt to eukaryotic phytoplankton. Recently, a dilution approach used to estimate grazing rates by microzooplankton (Landry & Hassett, 1982) was modified to infer viral lysis rates of Micromonas spp. (Evans et al., 2003) and Phaeocystis globosa (Baudoux et al., 2006). In this approach, four dilutions  are made by mixing whole water (i.e. unfiltered lake water) with grazer-free (0.2 pm filtered) water to create a dilution series; a second dilution series is also made using water in which both the grazer and viruses are removed by ultrafiltration (virus+grazer-free). As the fraction of whole water decreases with dilution, the relative abundance of phytoplankton grazers and/or viruses also decreases. This decrease leads to fewer encounters between phytoplankton and their grazers or viruses, which results in an increase in phytoplankton  147  growth rates. In each dilution series, the growth rate of phytoplankton in each dilution fraction is determined and regressed against the fraction of whole water present. The slope coefficient of each regression is the phytoplankton mortality rate caused by grazing or grazing and viral lysis while the y-intercepts are the phytoplankton growth rates. The difference between the regression slope coefficients of the two dilution series is an estimate of the mortality caused by viruses (Figure 5.1). The percent of phytoplankton mortality attributed to viruses and grazers is then determined from these mortality rates and the potential phytoplankton production rates; while daily turnover is calculated from the mortality rates. There are a number of assumptions implicit in the original dilution experiment including a constant and equal rate of exponential phytoplankton growth in all the dilutions, a grazing rate proportional to the dilution fraction, an abundance of grazers that remains constant over the experiment and a statistically significant linear regression between growth rate and dilution fraction (Landry & Hassett, 1982; Landry et al., 1995). Intuitively, the addition of a virus+grazer-free dilution series to the experimental design requires a modification of these assumptions. Additionally, since viral abundance can be altered by parameters of viral infections, and not just host abundance, a new level of complexity is introduced with the additional dilution series. For example, growth of phytoplankton in the experiment may be influenced by the length of the lytic cycle or by differences in the proportion of phytoplankton cells infected prior to the experiment (Jacquet et al., 2005). Given this information, the goal of this study was to evaluate the utility of the modified dilution experiment to estimate viral-mediated phytoplankton mortality using both theoretical and empirical analyses. Along with evaluating the assumptions of the dilution  148  1.0 Virus+grazer-free regression y = m g ,X + u k  0.5  0.1 0.0^0 .2^0 .4^0 .7  1.0  Fraction of Whole Water Figure 5.1. Stylized regressions from a modified dilution experiment. A representation of the regressions generated from each dilution series in the modified dilution experiment. The slope coefficients are the mortality rates of phytoplankton in the virus+grazer-free dilution series (m g ,) and the grazer-free dilution series (m g ); the difference between them in the mortality rate associated with viruses (m v = m gv - m g ).  149  experiments, a mathematical model was constructed to examine the sensitivity of the approach to several parameters of viral infections that can directly affect viral abundance. The results from the models were then used to comment on the accuracy of the mortality rates calculated from field experiments conducted in two lakes at the Experimental Lakes Area, Ontario, Canada.  5.3 Methods Empirical: Modified dilution experiment Sampling On four occasions (June 7 & 23 and July 24 & 29, 2004), ca. 50 L of water was collected from 0.5 m in Lakes 227 and 239 in the Experimental Lakes Area in Northwestern Ontario (49 ° N 93 °W; Cleugh & Hauser, 1971) by submerging pre-rinsed (10 % HC1 and lake water) polyethylene carboys. These lakes differ in trophic status largely due to the fertilization of Lake 227 since 1969 with nitrogen and/or phosphorus (Findlay et al., 1994). This water was analyzed (as described below) for chlorophyll a concentrations, as well as the abundance of unicellular cyanobacteria, and then was used to set up the modified dilution experiments (see experimental set up section below).  Abundance estimates Unicellular cyanobacterial abundance (hereafter cyanobacteria, cells mL -I ) was determined by filtering 15-30 mL of sample onto a black 0.2 pm pore-size polycarbonate filter (Millipore, Billerica, MA, USA) until dry. The filter was mounted on a glass slide with  150  100 % glycerol and frozen until enumerated. More than 200 autofluorescent cells were counted using epifluorescence microscopy and wide-green excitation (510-550 nm). Chlorophyll a concentrations (pg L -1 ) were determined by gently (< 250 mm Hg) filtering 50 to 100 mL of sample onto a 0.7 pm pore-size glass-fiber GF/F filter (Whatman, Florham Park, NJ, USA). Filters were frozen and extracted overnight in 90 % acetone within four months of collection, and chlorophyll a concentrations (gg L -1 ) were determined fluorometrically (Wetzel & Likens, 1991).  Experimental set up Grazing or grazing+viral infection rates on phytoplankton were reduced in a triplicate series of dilutions (0.33, 0.50, 0.66 and 1.0 fraction of whole water) with either 0.22-pm (Durapore; Millipore) or 10-kDa (S1Y10; Millipore) filtered water, respectively. Nitrate and phosphate (final concentration of 3.2 pM and 16 MM, respectively) were added to each of the 2.5-L bottles to prevent nutrient limitation. The bottles were incubated in situ for 48 h. Phytoplankton growth in each dilution was estimated from changes in cyanobacterial abundances or chlorophyll a concentrations over the incubation period (see calculation section below). Additionally, unmanipulated lake samples were collected during the experiment to determine in situ phytoplankton growth rates from changes in the cyanobacterial abundance or chlorophyll a concentrations over 48 hours, using k = ln(N48/No)/2 (see Table 5.1).  Calculations The phytoplankton growth rates estimated from the dilutions are hereafter referred to as 'apparent' (a.k.a. net) growth rates (4) , following Landry & Hassett (1982). The addition 151  Table 5.1. Terms used in the modified dilution experiment. Term k mg mg " mv Ilk p, pc  <C>  Definition apparent growth rate grazer-mediated mortality  In situ abundance of phytoplankton cells Potential production  Description / Equation = Ln(N2/Ni)/t = (Slope of grazer-free regression)(-1) = (Slope of virus+grazer-free regression)(-1) = mg "-mg = y-intercept from regressions = Ln(N2/N 1 )/t = Pk ± (difference between p i and the average k in the undiluted incubations (1.0)) = Cyanobacteria counts or Chlorophyll a concentrations = (p c )(<C>)  Production grazed  = (m g )(<C>)  Production lysed  = (m v )(<C>)  % of phytoplankton mortality due to grazing  = [(Production grazed)/(Potential production)] x 100 = [(Production Lysis)/(Potential production)] x 100 = (1-e-mg) x 100  % (1 -1  = (1-e-m") x 100  % d -1  virus+grazer-mediated mortality viral-mediated mortality growth rate in situ growth rate corrected growth rate  % of phytoplankton mortality due to viral lysis % of phytoplankton turnover due to grazing % of phytoplankton turnover due to viral lysis  Units d-1 d-1  cr' c1-1 c1 -1 d-1 d-1  Cells mL -1 or p g L-1  Cell mL -1 d -1 or fig L -1 d .1 Cell mL -1 d - ' or pg L -1 d -1 Cell mL -1 d -1 or pg L-1 d -1 %  %  152  of nutrients to the bottles enhances phytoplankton growth and this terminology clearly distinguishes these altered rates from both the in situ (t) and corrected growth rates (Re ) which are used to determine phytoplankton mortality (see Table 5.1). The apparent growth rate (k) of phytoplankton was determined from changes in cyanobacterial abundance or chlorophyll a concentration using: k= ln(N2/No)/t where No and N2 are the abundances of cyanobacteria or chlorophyll a concentrations at the onset of the experiment and after 48 h (t = 2 d), respectively. For each dilution series (virus+grazer-free and grazer-free), the average apparent growth rate at each dilution was regressed against the proportion of whole water present in that dilution fraction (i.e. 0.33, 0.50., 0.66 or 1). The slope coefficient of the regression is the mortality rate (m), while the y-intercept is the growth rate (t). Therefore, the difference between the virus+grazer-free and grazer-free slope coefficients is the viral-mediated mortality rate (Figure 5.1). ANOVA (SYSTAT v.11) was used to test whether each individual regression was significantly different from zero (Sokal & Rohlf, 1995) and an ANCOVA (Prism 5, Graphpad) was used to test for a significant difference between the virus+grazer-free and grazer-free regression slopes generated in each experiment (Zar, 1984). When the difference between slopes was significant, the number of phytoplankton cells lysed by viruses (production lysed) was calculated by multiplying the experimentally determined mortality rate by the in situ phytoplankton abundance (see Table 5.1). The percent of phytoplankton mortality due to viruses is then determined by diving the production lysed by the potential production, which is calculated from the corrected phytoplankton growth rates (see Table 5.1 and Evans et al., 2003). Corrected growth rates (p c ) were calculated by adding or subtracting the difference  153  between in situ phytoplankton growth (p i ) and that average apparent growth rate in the whole water dilutions (1.0 dilution fractions) from the regression growth rate (p. k ), this is done to compensate for enhanced growth in the experimental bottles.  Theoretical: Model analysis Description of the models The sensitivity of the modified dilution experiment to parameters of infections that influence viral abundance was theoretically examined by constructing a model using equations modified from Bratbak et al. (1998). Two iterations of the model, one for cyanobacteria and cyanophages, and one for eukaryotic phytoplankton and their viruses, were constructed to test the sensitivity of the modified dilution experiment to several key infection parameters that can affect viral abundance, including the number of viruses produced per cell lysed (burst size), host growth rate, length of the lytic cycle and the fraction of initially infected host cells. In each model run, the abundances of phytoplankton, infected phytoplankton and viruses were calculated at 1 hour steps for 200 hours to assess the stability of the model. In general, phytoplankton abundance (P) is a function of phytoplankton growth less the cells that become infected, while the abundance of infected phytoplankton cells (I) is described by the number of newly infected cells less the number of infected cells that lyse. Finally, viral abundance is a function of the infected phytoplankton that lyse releasing new viral particles less the viral particles that adsorb to cells or decay. The abundances can be described by the following equations: dP/dt = pP — c6162VP — iP o dI/dt = c6162VP-C6I62Vt-T  Pt-T  iP0  154  dV/dt = mea l cr2V t _T Pt _T-ccriV(P+I) — yV + m(iPo) where p = host growth rate (d -1 ), P = phytoplankton abundance (cells mL-1 ), c = host clearance rate (a variable involved in determining contact rate between host and viruses; mL 11 1 ), al = fraction of viruses adsorbed to particles, cr 2 = fraction of infectious viruses, V = viral abundance (mL -1 ), i = fraction of phytoplankton cells infected at the onset of the experiment (The model assume an equal proportion of these cells lyse hourly over the length of the first lytic cycle) , I = infected phytoplankton abundance (cells mL -I ), Po = phytoplankton abundance at t = 0 h, t-T (T is the time between infection and lysis of a cell) = length of the lytic cycle in hours and thereafter is called 1, m = burst size and y = viral decay rate (11 1 ). The cyanobacteria-cyanophage and phytoplankton-virus iterations of the model were constructed and run in Microsoft Excel 2003 using the parameters in Table 5.2. Each run of the model included four fractions in which phytoplankton and viruses were initially diluted by 0, 30, 50 and 70 %. No other parameters (such as host growth rate) were changed in these dilutions, and all the values were held constant throughout the experiment, fulfilling key assumptions of the original dilution experiments (Landry & Hassett, 1982). The model was run several times to investigate the effect of variations in burst size (m), host growth rate (p), length of the lytic cycle (1) and fraction of initial host cells infected (i) on the apparent phytoplankton growth rate. This was achieved by rerunning the model and changing the value of one parameter in all the dilutions, but keeping all the other parameters constant. The values of each of the four investigated parameters were chosen to represent the ranges reported in the literature (Table 5.2).  Model analysis  To determine the sensitivity to each parameter of interest, the apparent growth rate (k) 155  7:1 O  a)  Cl  N  a.)  O z  Cl a) ".5.•  0  71.)1  =  v,--..) ,  4  -  c;  tr)7  'a —  :0 M  CU c`  0 4-)  6 "^)(7) ..._, ,-,  N  C3 ,--i  4 0 4 In  7 n:i1.)1 0  et  w  :.6  O. o  ,,-  ire  —  cn ,..  1■1  2  73 ■•• U  a)  0 0 1.-1 ..;  Cin 0 CIA  (4  ;■( 0  7:))  /11  qsz,Q  CI C1) 2 a) .,_,  '4 . ,../  CI $..1  4-1 0  ci)  a.)  CA  .-.:  rCI)  ,--, 6 .6 Ce  o . 71-  o  "a' t  o 7,- C7 9,4  (  6 7t-' (9 3 -NI a* ,-  cn ,...„ (4 0 •— ___: tri N ca. —t _- ct  Cl  V)  I-- o o  .....^,..4 '40 In C7 1 C 0 ,...s. tn c. 1 ,--,  ,--. -4a)a'  0  0. G  C' C•1■1  C---4'  .E  77' C tf ) cn d 7r. 6 .• N 0  0  ,2 0  vn i■4  cn  1..  .-1 75 0 "CS  a)  C.)  2  ,•", P■1  c.) "C  Ci9 .,  .5  4-1  CID •■••1  ti.) 7.1  0  0  z  c.) --)  ai  0 ›-b  4-1  N •5 4■  I.) tj a)  w V.1  C.)  CI^L e'', ;■■•1  1)  tO  W  ,7,' ;.,  1.0  0 Z  ,.. C1 •^,.  wa.)(-)ct 0 1=3 0 t%•••1 c49 c) 0^a.,) ,5 tI) -t) "C  ,..  ,4  U  ,..■1  0 •,•-•  ,-  ,•-■  Cl •  5 .-5  C3 •!:-1  1■1 41 t 0 0  ,.sz Cl  at > •;=I 4:1 •;=1 " (4-4 C1-1 ,.., C1■1 41.)00 ,-4 4... 0  0 Ct I.-1  Z Cl .0 ;-• C.3  ‹.)  6 (31 > —  E.•4=1  :" E —  (...) w w > w .5 ..1 z ¢ >  CI) `—.  1-1 1-1 C) C)  ., N •,-, .. ■•C `—' (,) s---' 0 t-4 N 0 ct  d  14-)  ^  c; tr)  ,--1  4— ,-. --, d ::..-%  C  • CD  ,—•  6, —' a) 4-) ir) Cae a%  ^  0 71- .1: d C)  4  ;•4  7/ )' 0  'CS)  w —0 -4 r..... x cf.) '—' _. inn, ci ,--1 ,••••:1 ,--1 ,--1 --4 co 7._-,  r \•4 tr) ,444),  CA i.e •■I  =  14, ::  0 •.4 •i-) (1) •4-4 c w a,  c%) 0  ab ,..o t4-, ct  cL) ,4 as = _ '0  ;.., a) a)  a:E z. a., 04 0.4  156  of the phytoplankton host (P) in the model was determined after 48 h, using k = ln(P48/F'0)/2• For each run of the model, the calculated growth rates in each of the dilutions were regressed against the fraction of whole water present (0.3, 0.5, 0.7. 1.0), exactly as is done in the modified dilution experiment. For each parameter, the regressions generated from each run of the model were then statistically compared using an ANCOVA (Zar, 1984) (Prism 5, Graphpad). If the regression slopes generated from each parameter value were statistical different from each other, then the mortality rates are sensitivity to that parameter.  5.4 Results Empirical: Modified dilution experiments Two modified dilution experiments were conducted in each lake during the summer of 2004. The daily turnover of cyanobacteria or chlorophyll a containing cells due to viral lysis was calculated from the difference between the apparent growth rate regressions (Figure 5.2, Table 5.3). In all four experiments, the differences between the virus+grazer-free and grazer-free regressions were not significant at a = 0.05 or 0.10 levels (Table 5.3). Although not statistically significant, the experiments in Lake 227 were used to calculate viralmediated mortality rates of cyanobacteria and cells containing chlorophyll a. From the apparent growth rate regressions, it was estimated that 36 % and 21 % of the daily turnovers of cyanobacteria and chlorophyll a biomass, respectively was due to by viruses (Table 5.4).  157  Lake 227 0.8 0.7 •  a 1^•  0  Co  Lake 239 0.05  Grazer free diluent  Grazer free diluent  •  0.00  Viruargrazer free diluent  .r`  0.6 •  •  Virus-tctrazer free diluent  -0.05  (1)^0.5 -0.10  U •  0.4  Om  -0.15 0.3  1  a -0.20 -  4 ^0.2  Co  -0.25 -  >rot^0.1  -0.30  0.0  az^0.4^0.6^0.8 Fraction  1.0  1.2  0.2  0.30 -^ 0.25  0.35  C  •  Grazer free diluent Virua*grazer free diluent  a3 b 0.20 =  2 O  0.30 o  0.25  •  0.20  °' 0.10  2'  0.15  0.05  a  0.1 0  0.15  0.00  0.4^0.6^0.8  1.0  1.2  Fraction of whole water  of whole water  8  Grazer free diluent Virus+grazer free diluent  0.05  -0.05 ^0.00 0.2^0.4^0.6^0.8 Fraction of whole water  . ^ 1.0^1.2^0.2^0.4^0.6^0.8  1.0  1.2  Fraction of whole water  Figure 5.2. Modified dilution experiments. Regressions analyses of cyanobacteria (a, b) or chlorophyll a containing phytoplankton cells (c, d) from modified dilution experiments conducted in eutrophic Lake 227 (a, c) and oligotrophic Lake 239 (b, d) at the Experimental Lakes Area. In each panel, the apparent growth rate (k) in each of the dilutions was regressed against the fraction of whole water present. The closed circles represent the dilution series in which the host cells were diluted with the grazer-free water; the open circles represent the virus+grazer-free dilution series. Each data point represents the average apparent growth rate at each dilution and the error bars are + 1 S.E., when not visible the error was less then the width on the symbol. For individual regression statistics see Table 5.3. Note different scales were used in the panels.  158  ^  •▪•• Cr■  N  1 ;-.  0  0  "c) a) (1 0  C 4  N a) cd  n C.) 0  "Cj  0 C..) to V)  al  c 6, 6, a) to  .  0  1..1  ..  .  a.)^. .4 .4'" ,s,^= . ',4 • 0 EA ^0 15 GA  o ©  o o  C1-1 Ce)  o o c) c (6 o N N c c (P, P, • --. ,-1  0^c.)^z.. 71- 71- • N N  zi^=^ -s= A c as C.) V 0 0 N N C2 CL1 00  6  N o  6  ,-1  C)  6 6  cu i-, cs ,•••., 6.r a DI, .0 ;I:$  6 , -0  In tr)  N kr) 7i-  N .--■  6 6 6 6 c,5 'c  tn ---I  6  a)  ct  .0  -  a)  5  •  • c:; c=> 6  N 71- M N t---  a.) a) :-. 1 CD 3-■ cll a) ..47,1 r..4 ,  4 4  a.) a) ;.-.  1-4  a)  Li"  (I) (1)  I-4  1 N i-4 Ct S-1 Q) t:k1) N  a) t) ;-.4 I-) a)  a) ,4-.,4  64  '474  N 1 CI ;-■ S4 CI) 0,0 N  -I- m + czs  5 ED 5 E 6  et  R3 S-1 Q) ;-.1 00 N  E .,5.,  ,_, .4 0, 0 ;•-■ 0 ,--1 ,4 U (1) 0 ct .-0 o cz1 ›, C..) >, C.)  CA N  __, .4 04 0 ;-4 o ,--4 ,U  CT N N NNm N N N  1.„) ci cz$ 0  . ,. 5,  v'  + czi +  ri.) N 0.4 co t...1 , ;-■ 7.4 CL) CIO N  a) ;.-. c't  6 6 6 6 6 6 6 6  ■ocNiclocq,——.,©  N N  6 6 6  N O■ 0 C) ■0 0 ,—i 710 0 N 1--1 00 Cr) Cr) N 0 c)  6 6 6 6 6 c=> 6 6  O cr) 0 oo ,--1 cr) In v:) 0 0  ,,  t^4:  a.)  0.-4^o^..... nt :  3  0,^0.,  Q,,  .,---,  ,4-"N  0. hz  0 0 6 4 6  GI  •-.4  ,.x  ....) rn 0 Z  ot a..) g  '  -1-,^2 -f ^a.)^ cA a) v) 0  a)  ,  O O  a  0 H bJD cu  ^c) a) 4e4  2 •  O  • •  4 '  "4-■  a) a.) tt a 7:)‘-• ci cu  ct  a.) a;  ,  • N E  • U •  :u  E^••  ,  O t) 3 o ,L) • •--1 c,  a)  a.)  cci  E; 3  4  ti) • r:)., re)^  E-1  159  7  e  ).) CI)^cn .--, S^0) .", 0^w) = CP 0 "0 VD c**) ir^0^;•+  —,  —■ CV  0 N  v:,  ,--,  if-;  If)  0 •-c U  z = 0.. 0  6  N  71-  (-,1 o  6  00 0  C'  N  al . c. ..... 0 ct ..0 0  6  71-  N  6  N  R I —  0  o o to tn  0 kr) 0  in 4.)^by ,--, 0.. 49^0 _. p 0^•■-, gg ..„, = CD N ).)^0^CI e ' ' 0 "1:1^au ,...,---, 0^,Ltz 0 , CO^44 1—i 0^cl)^E 1-.3^P.1 O^cf) 6. 0.•^Ti  0.0 CA &.i^,...,  ,—, 0^,Ltz = ^c3 7, Q^s1 4) ,-4 :^e% ^0^6, 0.4^4).)., ,.,.---, 0 ,Ltz ITS © = = :^CJ 14  te CP pg E  0) cp C.)  ,,, cm 0.,  ©^az ....  a  -at ^,5, •-■-' co^1/4:$ cl.)^P. )-i^0^^ 6^L.,^  C...)  g  ,,t, r•••■  == •'^:3 N'1 ., Co c 5 c,.,  Co Tei t  ;...^,E^--..  -4... ci) 0 =  c ct >, u  160  Theoretical: Model analysis of the modified dilution experiment The abundances of phytoplankton (P), infected phytoplankton (I) and viruses (V) varied over time in each run of the model, but always achieved a stable state (Figure 5.3; a typical model run). Additionally, all three abundances behaved as expected (for example, an increase in burst size resulted in a quicker decline in phytoplankton abundance), adding creditability to the constructed models. Further, the shapes of the abundance curves became more gradual in the higher dilution fractions, indicating both a smaller and slower affect of viruses. Apparent growth rate regression in both iterations of the model (cyanobacteriacyanophage and phytoplankton-virus) were sensitive to burst size (m), host growth rate (m), length of the lytic cycle (1) and the fraction of cells initially infected (i). In the cyanobacteria-cyanophage iteration, the regressions generated with each parameter value of the model were significantly different from each other regardless of whether the parameter under investigation was m, p., 1 or i (Figure 5.4a, c, e, g, respectively). Likewise, the phytoplankton-virus iteration regressions were also significantly different among each of the parameter values (Figure 5.4b, d, f, g). In both model, burst size and length of the lytic cycle were extremely significant (p < 0.0001), suggesting that the modified dilution experiment is highly sensitive to these parameters.  5.5 Discussion A theoretical examination of the assumptions and robustness of the modified dilution approach and an empirical test indicated that the experimental method was largely inadequate  161  Figure 5.3. Model outputs. One cyanobacteria iteration of the model used to assess the sensitivity of the modified dilution experiment (m = 100, p = 0.5 d -1 , I = 12 h, i = 0.1). Each panel represents a different dilution (a = 1.0, b = 0.7, b = 0.5 and d = 0.3) of whole water). In each case, the black closed circles is the abundance of phytoplankton (P) at each hour over the 200 hours the model ran. The dark grey open squares is the abundance of infected phytoplankton cells (I) and the light grey crosses is the viral abundance (V).  162  ▪  12000  1 8,6 1.8.+8  10000 —41— Phytopterrkton —.4— Infected phytoplankton ---- Viruses  8000  1,40+0 120+6 -g  2 2  8000  =  a  100+6 80,5 806+5 -j er  -5- 2000  4 0,5 2 0,5 00 20^40^60^80^100^120^140^180^180^200  Time (hours) 1 8. 4-6  2000 —11— Phyloptankton Infec1e9 phyloplanklon  10000  1 6.46  • - Viruses 411  "8 t  8800 2a 6  aoco  1 04/46 ; 8 0.5.5  4000  50.+5 §. 9. 2000 4.0005 2 0,-5 00  2000 0^20^40^60^80^100^120^140^1611^180^200  Time (hours)  86+0  12000  c 0.5 Dilution fraction  Phytoplankton —+0— labeled phytoplankton — - Viruses  10000  80 , 5 40+8 1 20+6 -g  316D°D  100+6 5 8 Oe+5  i  4Q°  a 0.5+5  S 2000  C.  3 5  4.00+5  0  2000  2 00 4 5 00  0  20^40^60^60^100^120^140^100^180^200  Time (hours)  12000  1 84+5  10900  -  Phytoplarkton  -  tntected phytoplankton VIruses  1.64+8 1.40+6 1.20 4 6  3  1 16000  04 , 6 5 80.45 Ge+5  12000 40005 0 06+5 00  2000 0^20^40  60^60^100^120^140^160^180^200  Time (hours)  163  Figure 5.4 Model analysis of the modified dilution experiment. Each panel shows the apparent growth rates (k) regressions generated from each value of the investigated parameter. The regressions for each model iteration (cyanobacteria-cyanophage: a, c, e, g and phytoplankton-virus: b, d, f, h) are shown. The four parameters investigated included, burst size (a, b), host growth rate (c, d), length of the lytic cycle (e, f) and the fraction of initial cells infected (g, h). In each case, the differences between the regressions were statistical determined using an ANCOVA and the F and p values are shown. Note different scales were used in the panels.  164  ^ •  Chlorophyll a model  Cyanobacteria model Burst size (m)  0.3 0 2^0.4^0.6^0.8  1.0^1 2^0 2^0.4^0.8^0.8  Fraction of whole water  1.0  12  Fraction of whole water  2  1.6  o^F38 = 5, p = 0.031  1.4  Host growth rate^i ^(u)^3  1.2 1.0  • ^  u = 0.32 u = as  •  u=1.0  0.8  -2 '  a 6  -4  a6 0.4 0.2 -  5 0 2^0.4^0.6^0.8  0,0 1.0^1 2^0 2^0.4^0.6^0.8  Fraction of whole water  1.0  12  1.0  12  Fraction of whole water 0.44 ^ 0.42  Length of lytic cycle (I)  0.40  e  0.38 0.36  -  2 0.34 cn t 0.32  k 0.30 -  Q  0.28 0.28 0.24 -  0 2^0.4^0.6^0.8  1.0  0.22 ^ 0 2^0.4^0.6^0.8  12  Fraction of whole water  Fraction of whole water  0.5  0.45  Fraction 00 of Initial^-0 5 cells^I _,0_ Infected 1 (I) -1.5 -  0.40  1 0.35 2 m 0.30 • 1=0.01 ^ 1=0.05 • 1= 0 1 6 i=0.15  0.25  2.0 02  • ^  i=0.01 0.03  • 1=0.oe a 1=0.1  a4^0.6^0.8 Fraction of whole water  1.0  ^  0.20 12  0 2^0.4^0.6^0.8  1.0  12  Fraction of whole water  165  at estimating viral-mediated mortality rates of phytoplankton in Lakes 227 and 239. The reasons for this conclusion are detailed below, starting with the theoretical analyses.  Assumptions of the dilution experiments to estimate phytoplankton mortality The classic dilution experiment of Landry & Hassett (1982) requires that a) the growth rate of the phytoplankton is exponential and constant, b) grazing is directly proportional to the dilution fraction and the grazer density does not change during the experiment and c) the apparent growth rate versus dilution relationship is described by a significant linear regression. The virus+grazer-free dilution series added to estimate mortality rates caused by viral lysis must also satisfy versions of these assumptions. This necessity is discussed below in the context of both the original and modified experimental approaches.  a) Phytoplankton growth is constant and exponential The growth rate of phytoplankton must be independent of dilution and thus unaffected by cell abundance. If phytoplankton growth rate varies with dilution, then changes in the apparent growth rate inferred from the regression analysis cannot be attributed solely to grazing. If the phytoplankton cells are not growing exponentially, growth rate can vary because competition for nutrients diminishes as the abundance of cells decreases with dilution (Landry & Hassett, 1982; Kimmance et al., 2007). As well, differences in growth rate between dilutions can affect viral infection dynamics, including burst size and length of the lytic cycle (Proctor et al., 1993). Consequently, inorganic Nitrogen and Phosphorous are often added to prevent nutrient limitation, and ensure exponential growth. The apparent growth rate is then corrected using estimates of in situ growth (see Table 5.1). 166  Adding nutrients to ensure constant growth rates can have other consequences, such as increasing burst size (Wilson et al., 1996; Clasen & Elser, 2007), reducing the length of the lytic cycle (Proctor et al., 1993) and causing prophage to enter the lytic cycle (Wilson et al., 1996). These processes would increase the abundance of viruses and influence infection  rates by increasing encounters and can not be corrected for. As well, the addition of nutrients may affect the composition of the phytoplankton community. Since viruses are host specific (Suttle & Chan, 1993; Waterbury & Valois, 1993) any change in the composition of the phytoplankton community could affect mortality rates. In theory, the impact of nutrients of viral abundance should be consistent across all the dilutions, but this may be an unjustified assumption and needs to be tested.  b) Loss rates are constant and proportional to dilution A key assumption of the dilution experiments is that phytoplankton loss rates are reduced in proportion to the dilution fraction. However, both high and low prey abundances can suppress microzooplankton grazing and result in non-linear changes in grazing rates with dilution (Gallegos, 1989; Evans & Paranjape, 1992). Also, any changes in grazer density over the duration of the experiment can cause the phytoplankton loss rates to be non-linear. Landry et al. (1995) suggested that relative grazing rates in each dilution should be determined and used in the apparent growth rate regression analysis. However, for microzooplankton grazing experiments these issues are typically dealt with by keeping the experimental duration short. Satisfying this assumption is more problematic for viruses, because viral abundance can not be solely explained by host density. For example, an unknown proportion of the phytoplankton will be infected prior to the beginning of the experiment and this proportion will be independent of the dilution fraction (i.e. the total 167  phytoplankton abundance will decrease with dilution but the percent of the population infected does not; Jacquet et al., 2005). These cells will lyse as the experiment progresses, increasing the rates of infection over the duration and may adversely influence the observed apparent growth rates.  c) Regressions are both linear and significant  The apparent growth rates must be directly proportional to dilution if the assumptions described above are met. Since phytoplankton growth rate is constant, but grazer abundance and grazing rates are proportional to dilution, the net change in phytoplankton abundance should be directly dependent upon the dilution fraction. Additionally, the linear regression should be significantly different from zero; otherwise there is no affect of grazing on phytoplankton abundance. Significant linear relationships are an indication that the assumptions of the experiment have been met. Finally, the regressions generated for each dilution series in the modified dilution experiment should be significantly different from each other or viruses are not a significant source of phytoplankton mortality.  Theoretical considerations: Model analysis Besides the abundances of phytoplankton host cells, there are several parameters associated with viral infection that could influence mortality rates by altering the abundance of viruses. Analysis with a simple model indicated that viral-mediated phytoplankton mortality rates determined using the modified dilution experiment are heavily influenced by these parameters. The apparent growth rate regressions are crucial for determining mortality rates; hence, the differences between the regressions generated from each model run were used to assess the influence of infection parameters on mortality rates. Mortality rates 168  determined from the model iterations using either cyanobacteria or chlorophyll a containing cells as hosts were sensitive to changes in burst size (m), host growth rate (p), length of the lytic cycle (1) and the fraction of initially infected cells (i) (Figure 5.4). The degree to which each parameter affected the regressions varied, but burst size and length of the lytic cycle were extremely significant for both host models (p < 0.0001, Figure 5.4). All four parameters investigated in this analysis vary substantially in aquatic ecosystems. For example, the burst size of phytoplankton viruses can range from 72 for viruses of Micromonas pusilla (Waters & Chan, 1982; Cottrell & Suttle, 1991) to > 10 5 for RNA  viruses of Heterosigma akashiwo (Lawrence et al., 2001; Tai et al., 2003), both of which are phytoplankton viruses that can occur within an environmental sample. Burst sizes, host growth rates, fraction of initially infected cells and the length of lytic cycles will vary with each virus-host pair and with environmental conditions. Thus the sensitivity of the apparent growth rate regressions to the inherent environmental variations in infection parameters, suggests that the modified dilution approach will likely yield inaccurate viral-mediated phytoplankton mortality rates in natural samples where multiple virus-host pairs occur; as a consequence results should be interpreted cautiously.  Empirical analysis: Modified dilution experiments Further evidence of the difficulty in using the modified dilution approach for estimating viral-mediated mortality comes from the statistical examination of linear regressions from four field experiments. Most of the individual regressions were not significantly different from zero (Table 5.3), and some were better described by inverse exponential rather than linear fits (i.e. Figure 5.2a,b). These results suggest that the assumptions of the dilution model were not met. Moreover, the difference between the 169  regressions generated from the two different dilution series was not significant (a > 0.05), however the two experiments conducted in Lake 227 were almost significant at a = 0.1. As a result, only the data from Lake 227 were used to estimate viral-mediated phytoplankton mortality rates, which indicated that 21 to 36 % of the daily phytoplankton turnover was due to viral lysis. Similar estimates have found in marine environments. For example, the mortality of  Micromonas spp. lysed by viruses in a mesocosm experiment conducted in Norway ranged from 9 to 25 % c1 -1 (Evans et al, 2003). Other experimental approaches have yielded mortality rates similar to the ones in this study. Cottrell & Suttle (1995) reported  Micromonas pusilla mortality rates of 2 to 10 % determined from viral decay rates, and Bratbak et al. (1993) estimated that 25 to 100 % of Emiliania huxleyi mortality was attributed to virus in a mesocosm study. While Proctor & Fuhrman (1990) and Suttle (1994) used TEM and viral decay rates, respectively and estimated that 5 to 24 % of cyanobacterial cells were lysed by viruses. However, as a eutrophic lake, Lake 227 likely experiences high host growth rates and burst size, which are both parameters that the model analysis indicated heavily influenced the modified dilution experiment. Therefore, the calculated mortality rates for Lake 227 should be interpreted carefully.  Utility of a modified dilution experiment to estimate viral-mediated phytoplankton mortality The lack of strong linear relationships between apparent phytoplankton growth and dilution fraction and significant differences between the dilution series in the experiments conducted in Lakes 227 and 239 makes it difficult to use the dilution approach to infer viralmediated mortality rates of the natural phytoplankton communities within these lakes. There 170  are several likely explanations of the non-significant linear regressions, including insufficient experimental replication which can be determined a posteriori using statistics (Kimmance et al., 2007) and sensitivity of mortality rates to viral infection parameters.  Although very a simple model were used in this analysis, it is clear that the dilution approach is very sensitive to changes in viral infection parameters that vary substantially among viruses, hosts and environments. Moreover, many of these parameters are linked. For example, host growth rate and length of the lytic cycle are positively correlated (Jacquet et al., 2005). Although, the modified dilution approach was of limited use in this study, it  may be appropriate to use in bloom situations where viral parameters such as burst size and length of the lytic cycle are relatively well constrained. Nonetheless, more complex models will likely be required before results under these conditions can be interpreted with confidence.  5.6 Acknowledgments We gratefully acknowledge the staff and students of the Experimental Lakes Area and the Freshwater Water Institute (Winnipeg, MB, Canada), particularly Mark Lyng, Dave Findlay, Corben Bristow, Shelley Brule, Ken Sanidlands and Justin Shead. Thanks to past and present Suttle lab members for discussions about the modified dilution experiment, especially Alice Ortmann, Emma Hambly and Andre Comeau. Chris Payne, Johan Vande Voorde and Caroline Chenard greatly assisted by reviewing and editing drafts of this manuscript. J.L.C. thanks Phillippe Tortell for the encouragement and Jon Shruin (UBC), Catherine Johnson (DFO), Andy Ridgwell (UBristol) and Joe Mahaffy (UCSD) for fruitful  171  discussion and assistance with the sensitivity analysis section of this manuscript. Financial support came from a graduate student fellowship from the Natural Sciences and Engineering Research Council of Canada to J.L.C. and a discovery grant to C.A.S.  172  5.7 References Baudoux, A. C. , Noordeloos, A. A. M. , Veldhuis, M. J. W. & Brussaard, C. P. D. (2006) Virally induced mortality of Phaeocystis globosa during two spring blooms in temperate coastal waters. Aquatic Microbial Ecology 44, 207-217.  Bratbak, G. , Egge, J. K. & Heldal, M. (1993) Viral mortality of the marine alga Emiliania huxleyi (Haptophyceae) and termination of algal blooms. Marine Ecology Progress Series 93, 39-48.  Bratbak, G. , Jacobsen, A. , Heldal, M. , Nagasaki, K. & Thingstad, F. (1998) Virus production in Phaeocystis pouchetii and its relation to host cell growth and nutrition. Aquatic Microbial Ecology 16, 1-9.  Clasen, J. L. & Elser, J. J. (2007) The effect of host Chlorella NC64A carbon: phosphorus ratio on the production of Paramecium bursaria Chlorella Virus-/. Freshwater Biology 52, 112-122.  Cleugh, T. R. & Hauser, B. W. (1971) Results of initial survey of experimental lakes area, Northwestern Ontario. Journal of the Fisheries Research Board of Canada 28, 129-137.  Cottrell, M. T. & Suttle, C. A. (1991) Wide-spread occurrence and clonal variation in viruses which cause lysis of a cosmopolitan eukaryotic marine phytoplankter, Micromonas pusilla. Marine Ecology Progress Series 78, 1-9.  173  Cottrell, M. T. & Suttle, C. A. (1995) Dynamics of a lytic virus infecting the photosynthetic marine picoflagellate Micromonas pusilla. Limnology and Oceanography 40, 730-739.  Evans, C. , Archer, S. D. , Jacquet, S. & Wilson, W. H. (2003) Direct estimates of the contribution of viral lysis and microzooplankton grazing to the decline of a Micromonas spp. population. Aquatic Microbial Ecology 30, 207-219.  Evans, G. T. & Paranjape, M. A. (1992) Precision of estimates of phytoplankton growth and microzooplankton grazing when the functional response of grazers may be nonlinear. Marine Ecology Progress Series 80, 285-290.  Findlay, D. L. , Kasian, S. E. M. , Hendzel, L. L. , Regehr, G. W. , Schindler, E. U. & Shearer, J. A. (1994) Biomanipulation of Lake 227 in the Experimental Lakes Area (ELA): effects of phytoplankton and nutrients. Canadian Journal of Fisheries and Aquatic Sciences 51, 2794-2807.  Fuhrman, J. A. (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399, 541-548.  Gallegos, C. L. (1989) Microzooplankton grazing on phytoplankton in the Rhode River, Maryland: nonlinear feeding kinetics. Marine Ecology Progress Series 57, 23-33.  174  Jacquet, S. , Domaizon, I. , Personnic, S. , Sriram, A. , Ram, P. , Heldal, M. , Duhamel, S. & Sime-Ngando, T. (2005) Estimates of protozoan-and viral-mediated mortality of bacterioplankton in Lake Bourget (France). Freshwater Biology 50, 627-645.  Kimmance, S. A. , Wilson, W. H. & Archer, S. D. (2007) Modified dilution technique to estimate viral versus grazing mortality of phytoplankton: limitations associated with method sensitivity in natural waters. Aquatic Microbial Ecology 49, 207-222.  Landry, M. R. & Hassett, R. P. (1982) Estimating the grazing impact of marine microzooplankton. Marine Biology 67, 283-288.  Landry, M. R. , Kirshtein, J. & Constantinou, J. (1995) A refined dilution technique for measuring the community grazing impact of microzooplankton, with experimental tests in the central equatorial Pacific. Marine Ecology Progress Series 120, 53-63.  Lawrence, J. E. , Chan, A. C. & Suttle, C. A. (2001) A novel virus (HaNIV) causes lysis of the toxic bloom-forming alga Heterosigma akashiwo (Raphidophyceae). Journal of Phycology 37, 216-222.  Paul, J. H. (1999) Microbial gene transfer: an ecological perspective. Journal of Molecular Microbiology and Biotechnology 1, 11-26.  Proctor, L. M. & Fuhrman, J. A. (1990) Viral mortality of marine bacteria and cyanobacteria. Nature (London) 343, 60-62. 175  Proctor, L. M. , Okubo, A. & Fuhrman, J. A. (1993) Calibrating estimates of phage-induced mortality in marine bacteria: Ultrastructural studies of marine bacteriophage development from one-step growth experiments. Microbial Ecology 25, 161-182.  Sokal, R. R. & Rohlf, F. J. (1995) Biometry. W.H. Freeman and Company, New York.  Suttle, C. A. (1992) Inhibition of photosynthesis in phytoplankton by the submicron size fraction concentrated from seawater. Marine Ecology Progress Series 87, 105-112.  Suttle, C. A. (1994) The significance of viruses to mortality in aquatic microbial communities. Microbial Ecology 28, 237-243.  Suttle, C. A. (2005) Viruses in the sea. Nature 437, 356-361.  Suttle, C. A. (2007) Marine viruses-major players in the global ecosystem. Nature Review 5, 801-812.  Suttle, C. A. & Chan, A. C. (1993) Marine cyanophages infecting oceanic and coastal strains of Synechococcus: abundance, morphology, cross-infectivity and growth characteristics. Marine Ecology Progress Series 92, 99-109.  Tai, V. , Lawrence, J. W. , Lang, A. S. , Chan, A. C. , Culley, A. I. & Suttle, C. A. (2003) Characterization of HaRNAV, a single-stranded RNA virus causing lysis of Heterosigma akashiwo (Raphidophyceae). Journal of Phycology 39, 343-352.  176  Waterbury, J. B. & Valois, F. W. (1993) Resistance to co-occurring phages enables marine Synechococcus communities to coexist with Cyanophages abundant in seawater. Applied and Environmental Microbiology 59, 3393-3399.  Waters, R. E. & Chan, A. T. (1982) Micromonas pusilla virus: The virus growth cycle and associate physiological events within the host cells host range mutation. Journal of General Virology 63, 199-206.  Wetzel, R. G. & Likens, G. E. (1991) Limnological Analysis. Springer-verlag, New York.  Wilhelm, S. W. & Suttle, C. A. (1999) Viruses and nutrient cycles in the sea. Bioscience 49, 781-788.  Wilson, W. H. , Can, N. G. & Mann, N. H. (1996) The effect of phosphate status on the kinetics of cyanophage infection in the oceanic cyanobacterium Synechococcus sp. WH7803. Journal of Phycology 32, 506-516.  Wommack, K. E. & Colwell, R. R. (2000) Virioplankton: Viruses in aquatic ecosystems. Microbiology and Molecular Biology Reviews 64, 69-114.  Zar, J. H. (1984) Biostatistical analysis. Prentice-Hall, Englewoods Cliffs, N.J.  177  Chapter six: Conclusion Summary, suggestions for future research directions and final thoughts  178  6.1 Summary This dissertation began with a simple idea, re-examine a comparison made between freshwater and marine environments with a much larger and more robust data set. Maranger & Bird (1995) found a difference between viral abundances in freshwater and marine environments and suggested that viruses infecting phytoplankton were a more abundant and important part of the virioplankton in lakes than oceans. When this comparison was reexamined with a large data set (Chapter two), not only was there a significant relationship between viral abundance and chlorophyll a concentrations, as originally observed by Maranger & Bird (1995), but also between viral and bacterial abundances. Interestingly, in my study the relationship between viral and bacterial abundances differed in marine and freshwater environments. There was a significantly steeper regression in freshwaters, suggesting that there could be more phytoplankton derived viruses in lakes than in oceans. These results supported the claims initially suggested by Maranger & Bird (1995), that there is a fundamental difference between freshwater and marine environments and that viruses infecting phytoplankton may be more important in lakes. My dissertation work has greatly extended our knowledge of viruses that infect phytoplankton in lakes. By using algal virus specific (AVS) primers (Chen & Suttle, 1995; Chen et al., 1996) combined with DGGE, I determined that Phycodnavirus richness varied temporally and spatially in several lakes at the Experimental Lakes Area in Northwestern Ontario (Chapter 3). Viral richness patterns were influenced by both trophic status and regional climatic conditions, with higher richness occurring in more eutrophic lakes and in the spring and early summer months. These results are ecologically interesting because eutrophication, temperature, and the timing and intensity of the spring phytoplankton bloom  179  are all predicted to change in the next few decades (Mugnuson et al., 1997). The effects of these alterations on the richness of viral communities are presently unknown; however, since richness can directly affect phytoplankton composition through host-virus interaction, the effect of climate change seems like an important area for future investigations. Genetic analysis suggested that freshwater environmental Phycodnavirus DNA polymerase (pol) sequence fragments fall into different phylogenetic groupings than their marine counterparts (Chapter 4), implying that there are different evolutionary histories of algal viruses in lakes and oceans. Genetic distance analysis indicated that the 20 freshwater DNA pol fragments came from Phycodnaviruses infecting nine different host species. Subsequent analysis revealed that these co-occurred with representative species from ecologically important groups of phytoplankton (green algae, diatoms and dinoflagellates), suggesting these may be the hosts for these viruses. Interestingly, this analysis identified no potential cryptophyte host. These small, naked, flagellates are thought to be important members of lake phytoplankton communities (Wetzel, 2001). The apparent lack of cooccurring hosts could be the result of under-sampling or, perhaps suggests that cryptophytes are not infected by Phycodnaviridae. Presently, quantifying viral-mediated phytoplankton mortality is problematic. Attempts to use modified dilution experiments (Evans et al., 2003) at the Experimental Lakes Area produced results that must be interpreted carefully (Chapter 5) since they failed to meet the assumptions implicit in the method. A simple model demonstrated that the dilution approach is extremely sensitive to parameters associated with viral infection, including host growth rate, burst size, length of the lytic cycle and the fraction of cells initially infected, and all of these parameters vary substantially within environments. It is possible that the approach is  180  more suited to specific host-virus systems that have well constrained parameters, such as might occur in a bloom situation.  6.2 Future research directions  6.2.1 Isolation of new phytoplankton viruses In the context of my dissertation work, the isolation of new host-virus systems is necessary to assign clades on the environmental DNA polymerase tree (see Figure 4.2) to specific viral taxa, allowing the phylogenetic tree to be further resolved. However, isolation may reveal other groups and families of viral pathogens that infect freshwater phytoplankton, including viruses that can not be amplified by the AVS primers. This may result in the development of new (or more specific) primer pairs. Additionally, further isolation will allow detailed investigations into infection kinetics which will continue to expand our understanding of phytoplankton viruses.  6.2.2 Development of a robust and reliable way to estimate viral-mediated phytoplankton mortality rates In order to understand the impact viruses have on community composition, nutrient cycling and maintaining diversity, reliable estimates of viral-mediate mortality rates are necessary. However, there is no experimental approach currently available to accurately determine viral mortality rates in phytoplankton communities. As indicated in Chapter 5, more complex modeling exercises may determine the key parameters necessary to  181  confidently apply the modified dilution experiment. Molecular approaches have been used to determine species-specific mortality rates, however given the paucity of non-degenerate primers that target phytoplankton viruses, this approach has limited utility. So until better methods (or primer pairs) are developed, mortality rates estimated from a combination of experiments, including the dilution experiment (Evans et al., 2003), the viral reduction approach (Wilhelm et al., 2002), the enzymatic digestion assay (Agusti & Sanchez, 2002) and the esterase assay (Agusti et al., 1998) may help to ascertain an accurate range of viralmediated phytoplankton mortality rates.  6.2.3 Impact of climate change on viral richness In my dissertation, the richness of the Phycodnavirus community was indirectly influenced by climate (through phytoplankton abundance and composition), however, I also suggested that richness may be directly affected by aspects of climate, such as temperature and UV radiation. I believe that the effect of climate change on viral communities should be investigated. First, the direct affect of climate of viral richness and abundance must be clearly demonstrated and then investigators can look into how this direct affect translates into ecosystem processes, such as mortality rates and maintaining diversity. Since viruses influence nutrient cycling (Fuhrman, 1999; Wilhelm & Suttle, 1999; Suttle, 2005; Suttle, 2007), any changes in the amount of phytoplankton lysed by viruses that can be attributed to climatic conditions may ultimately alter the efficiency of the biological pump (Suttle, 2007) and should, therefore, be incorporated into both climate change models and our understanding of ecosystem function.  182  6.3 Final thoughts This report of Phycodnaviridae in lakes suggests that they are rich, genetically distinct and heavily influenced by aspects of their environment, including physical and chemical parameters. Additionally, viruses may be a substantial source of phytoplankton mortality in lakes, but further research is required to confirm this. Given the role of viruses in mortality, nutrient cycling and genetic exchange, more research is clearly needed to understand the influence that viruses have on structuring phytoplankton community composition in lakes. Hopefully, my dissertation sheds some light on the viral dynamics occurring in lakes and will stimulate further research into freshwater Phycodnaviruses.  183  6.4 References Agusti, S. & Sanchez, M. C. (2002) Cell Viability in natural phytoplankton communities quantified by a membrane permeability probe. Limnology and Oceanography 47, 818-828.  Agusti, S. , Satta, M. P. , Mura, M. P. & Benavent, E. (1998) Dissolved esterase activity as a tracer of phytoplankton lysis: evidence of high phytoplankton lysis rates in the Northwestern Mediterranean. Limnology and Oceanography 43, 1836-1849.  Chen, F. & Suttle, C. A. (1995) Amplification of DNA polymerase gene fragments from viruses infecting microalgae. Applied and Environmental Microbiology 61, 1274-1278.  Chen, F. , Suttle, C. A. & Short, S. M. (1996) Genetic diversity in marine algal virus communities as revealed by sequence analysis of DNA polymerase genes. Applied and Environmental Microbiology 62, 2869-2874.  Evans, C. , Archer, S. D. , Jacquet, S. & Wilson, W. H. (2003) Direct estimates of the contribution of viral lysis and microzooplankton grazing to the decline of a Micromonas spp. population. Aquatic Microbial Ecology 30, 207-219.  Fuhrman, J. A. (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399, 541-548.  184  Maranger, R. & Bird, D. F. (1995) Viral abundance in aquatic systems: a comparison between marine and fresh waters. Marine Ecology Progress Series 121, 217-226.  Mugnuson, J. J. , Webster, K. E. , Assel, R. A. , C.J., B. , Dillon, P. J. , Eaton, J. G. , Evans, H. E. , Fee, E. J. , Hall, R. I. , Mortsch, L. R. , Schindler, D. W. & F.H., Q. (1997) Potential effects of climate changes on aquatic systems: Laurentian Great Lakes and precambrian shield region. Hydrological Processes 11, 825-871.  Suttle, C. A. (2005) Viruses in the sea. Nature 437, 356-361.  Suttle, C. A. (2007) Marine viruses-major players in the global ecosystem. Nature Review 5, 801-812.  Wetzel, R. G. (2001) Limnology: Lake and river ecosystems. Academic Press, San Diego.  Wilhelm, S. W. , Brigden, S. M. & Suttle, C. A. (2002) A dilution technique for the direct measurement of viral production : A comparison in stratified and tidally mixed coastal waters. Microbial Ecology 43, 168-173.  Wilhelm, S. W. & Suttle, C. A. (1999) Viruses and nutrient cycles in the sea. Bioscience 49, 781-788.  185  Appendices  186  Appendix 1. Presence or Absence of particular AVS amplified bands on Lake 227 DGGE. Bold indicates sequenced AVS DGGE bands used in a multivariate analysis (MDS) to identify potential phytoplankton hosts. AVS Band V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 V17 V18 V19 V20 V21 V22 V23 V24 V25 V26 V27 V28 V29 V30 V31 V32 V33 V34 V35 V36 V37 V38 V39 V40 V41 V42 V43 V44 V45 V46  May 26 1 1 0 0 0 0 0 0  1  Jun 09 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  0 0 0 0 0 0 0 1 0 0  1 0 0 0 1 0 0 1 1  0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0  1  Jun 23 1 1 0 0 0 0 0 0 0 1 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 1 0 1 1 1 1 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0  Jul 07 1 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1  Jul 21 1 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 1 0 0 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 0 1 1 0 0  Aug 18 1 0 0 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  Sep 02 1 0 0 0 0 0 0 1 0 0 0 1 1 0 0 1 1 1 1 1 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  Sep 15 1 1 0 0 0 0 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1  1  1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  Sep 29 1 1 1 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  Oct 13 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  187  V47  1  0  0  0  0  0  0  0  0  0  V48  0  0  0  1  1  0  1  0  0  0  V49  0  0  0  1  0  1  1  0  0  0  V50  0  0  1  0  1  1  0  0  0 0  V51  0  0 1  0  0  0  0  0  1  0  V52  0  0  0  0  0  1  1  0  0  0  V53  0  0  0  1  0  1  1  1  0  0  V54  0  0  0  0  0  0  1  1  1  0  V55  0  0  1  0  1  0  0  0  1  0  188  Appendix 2. Presence or Absence of particular AVS amplified bands on Lake 239 DGGE. Bold indicates sequenced AVS DGGE bands used in a multivariate analysis (MDS) to identify potential phytoplankton hosts. AVS Band V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 V17 V18 V19 V20 V21 V22 V23 V24 V25 V26 V27 V28 V29 V30 V31 V32 V33 V34 V35 V36 V37 V38  May 20 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0  Jun 04 1 0 0 0 0 0 0 1 1 1  1  1 1 0 1 1 1 1 1 0 0 0 0 0 0 1 1 1 0 1 0 0 1 1 0 1 0 0  Jun 16 1 0 0 0 0 0 0 1 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 1 0 0 0  Jul 01 0 1 0 0 0 0 1 1 0 1 0 0 0 0 1 0 1 1 1 1 1 1 1 1 0 0 0 0 0 1 1 1 0 0 1 0 0 1  Jul 15 1 0 0 0 0 0 0 1 0 0  Jul 26 0 0 0 0 0 0 0 1 0 0  0 1 0 0 1 1 1  0 0 0 1 1 1 1  1 1 0 0 0 0 0 0 0 0 0 1 0 1 0 0 1 1 0 0 0  1  1 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 1 0 0 0  Jul 29 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0  Aug 12 0 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0  Aug 26 0  Sep 09 0  0 1 0 0 0 1 0 0 0 0 0 0 0  0 0 1 1 0 0 0 0 0 0  1  1  0 0 0 0 0 0 0 0 0 0 0 0 0 0  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  0 1 0 0 0 0 0 0 0  0 0 0 0 0 0 0 1 1  Sep 22 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  Oct 04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  189  Appendix 3. Presence or Absence of particular AVS amplified bands on Lake 240 DGGE. Bold indicates sequenced AVS DGGE bands used in a multivariate analysis (MDS) to identify potential phytoplankton hosts. AVS Band V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 V17 V18 V19 V20 V21 V22 V23 V24 V25 V26 V27 V28 V29 V30 V31 V32 V33 V34 V35 V36 V37 V38 V39 V40 V41 V42 V43 V44 V45  Jun 15 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 1 1 1 1 1 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1  Jul 13 1 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 1 0 1 1 1 0 1 1 1 1 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 1 1 0 1  Aug 10 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 1  1  0 1 1  Sep 08 1 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1  Oct 06 1 1 1 0 0 0 0 1 1  0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 1 0 0 0 1 1 0  190  V46  1  1  1  1  V47  0  1  1  1  1  V48  0  1  1  1  1  1  V49  0  0  1  0  0  V50  0  0  0  0  1  V51  0  0  0  0  1  191  1.6e+7 1.4e+7  3.0e+6  1.2e+7 -  E  2.5e+6  1.0e+7 2.0e+6  cri  8.0e+6 1.5e+6 6.0e+6  N  1.0e+6 -0  4.0e+6 -  5.0e+5^a) co cO  03 2.0e+6 0.0^.... , . May^Jun  Jul  Aug^Sep  Oct  Nov  Appendix 4. Temporal variation in bacterial abundances. Bacterial abundance (cells mL -1 ) in each of the three lakes determined from samples collected mid-May to mid-October.  192  Appendix 5. Presence or Absence of particular AVS amplified bands on Lake 224 DGGE. Bold indicates sequenced AVS DGGE bands used in a multivariate analysis (MDS) to identify potential phytoplankton hosts. AVS Band V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 V17 V18 V19 V20 V21 V22 V23 V24 V25 V26 V27 V28 V29 V30 V31 V32 V33 V34  Jul 06 1 0 0 0 0 0  1  1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 1  1  1 1 0 0  Jul 20 0 0 0 0 1 1 0 1 0 0 1 1 0 0 0 1 1 1 1 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0  Aug 07 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 1 0 0 0  Aug 17 0 0 1  Sep 14 0 1 0 0 0 0 1  Sep 28 0 1 0 0 0 0 0 1 0 1 0  Oct 12 0  0 0 0 0 0 0  0 0 1 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0  1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 1  Aug 31 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1  1  0  0  1 1 1 1 0 0 0  0 0 0 0 0 0 0  0 0 0 0 0 0 0  0 0 1 0 0 0 1  1  1  0 0 0 1 1 1 0 0 0  1  0 0 0 0 1 1 0 1  1  0 0 0 0 1 0 0 1  1  0 0 0 0  0 1 1 0  1  193  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0053315/manifest

Comment

Related Items