Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Using genomic data to understand anthropogenic influences on Oomycete and Phytophthora communities, and… Dale, Angela 2018

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2018_may_dale_angela.pdf [ 5.84MB ]
JSON: 24-1.0365770.json
JSON-LD: 24-1.0365770-ld.json
RDF/XML (Pretty): 24-1.0365770-rdf.xml
RDF/JSON: 24-1.0365770-rdf.json
Turtle: 24-1.0365770-turtle.txt
N-Triples: 24-1.0365770-rdf-ntriples.txt
Original Record: 24-1.0365770-source.json
Full Text

Full Text

USING GENOMIC DATA TO UNDERSTAND ANTHROPOGENIC INFLUENCES ON OOMYCETE AND PHYTOPHTHORA COMMUNITIES, AND THE EVOLUTION OF AN ALIEN INVASIVE SPECIES RESPONSIBLE FOR SUDDEN OAK DEATH, PHYTOPHTHORA RAMORUM.  by ANGELA DALE  A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   April 2018  © Angela Dale, 2018ii  Abstract Emerging Phytophthora pathogens, often introduced, represent a threat to natural ecosystems. Phytophthora species are known for rapid adaptation and hybridization, which may be facilitated by anthropogenic activities. Little is known about natural Phytophthora and oomycete populations, or mechanisms behind rapid adaptation. We surveyed oomycete and Phytophthora communities from southwest B.C. under varying anthropogenic influences (urban, interface, natural) to determine effects on diversity, introductions and migration. We used DNA meta-barcoding to address these questions on oomycetes. We then focused on Phytophthora, adding baiting and culturing methods, and further sub-dividing urban sites into agricultural or residential. Finally, we studied an alien invasive species, Phytophthora ramorum responsible for sudden oak death, and how it overcame the invasion paradox, limited to asexual reproduction and presumed reduced adaptability. Anthropogenic activities increase oomycete and Phytophthora diversity. Putative introduced species and hybrids were more frequent in urban sites. Migration is suggested by shared species between urban and interface sites, and two known invasive species found in natural and interface sites. Different anthropogenic activities influence different communities. Abundance increased for some species in either residential or agricultural sites. Two hybrids appear to be spreading in different agricultural sites. In the invasive Phytophthora ramorum, mitotic recombination drives diversification of the four lineages (NA1, NA2, EU1 and EU2), generating runs of homozygosity. One genome region, enriched in putative plant pathogenicity genes and transposons, was fixed in NA1 and present in eight EU1 individuals, but affecting the opposite alleles. Longer lesions during initial colonization in inoculated larch and Douglas fir logs suggested a fitness advantage in these EU1 individuals. iii  Mitotic recombination breakpoints were associated with transposons and low gene density. Non-core and lineage-specific genomic regions were enriched in putative plant pathogenicity genes and transposons. Gene loss was observed in the EU2 non-core genome affecting all effectors. A two-speed genome, where regions enriched in transposons and plant pathogenicity genes evolve faster, appears to drive non-core genome divergence, and mitotic recombination resulting in population evolution. This may explain invasion success and adaptability in Phytophthora pathogens. These results highlight the importance of anthropogenic activities in the emergence of forest diseases.    iv  Lay Summary International trade can facilitate the transfer and introduction of non-native plant and forest pathogens into new environments, resulting in forest disease outbreaks. Urban activities like gardening or agriculture may also spread plant pathogens. This study investigated the effects of these activities by comparing a group of organisms found in soil and water in urban and natural environments. Several species from this group are forest and crop pathogens. We determined that populations of these organisms in urban areas were more diverse and had more introduced species. Natural areas close to urban centers were also more diverse and could act as a bridge for migration of introduced species into natural areas. Since these organisms are highly adaptable, we investigated DNA characteristics that may facilitate adaptation in an introduced invasive pathogen. DNA changes have led to increased population diversity which may explain the success of this pathogen in new environments.    v  Preface Chapters 2 and 3 of this dissertation are an original intellectual product of the author, Angela Dale, with some collaborative input as described below. Chapter 4 was conducted as an international collaboration using data generated for the Genome Canada funded Taiga project. I, Angela Dale, was the lead investigator on this paper, working under the guidance of the supervisory author Dr. Richard Hamelin. I was responsible for the research questions, hypotheses, and direction of the paper, with input from my supervisor and collaborators, especially Dr. Nicolas Feau, and Dr. Sydney Everhart. Original manuscript composition was by me, and the version presented here was reviewed and edited by me as well as by Dr. Nicolas Feau and my supervisor, Dr. Richard Hamelin. Contributions of collaborators and the author to specific analyses are as described below. Drafts of each of the sections completed by collaborators (methods, results and figures) were given to me and edited for inclusion into both the main document and the supplementary materials by me, Dr. Nicolas Feau, and Dr. Richard Hamelin. The supplementary materials (Appendix B) describe the details of the work performed, including the work performed by the author, Angela Dale. However, it was not included in the main body of the document due to the format required for the type of journal where we are planning to submit the work. All the work described in chapters 2, 3 and 4 was conducted by me, except where indicated in the following sections.  Chapter 2: All the work described was conducted by me, except for the following. The samples for DNA meta-barcoding were prepared (DNA extraction, PCR amplified, barcoded, checked, pooled) by me and sent to the laboratory of Dr. Jean Bérubé of the Canadian Forest Service vi  where they were cleaned and quantified prior to being sequenced by the McGill platform as described in the dissertation.   Chapter 3: The chi-square independence test and heat maps of UniFrac distance were done by Dr. Nicolas Feau.  Chapter 4: Genome sequencing: DNA extraction was done under the supervision of Dr. Guillaume Bilodeau at the CFIA. Genome sequencing was done at the Genome Sciences Centre under the supervision of Dr. Steve Jones.  Extraction of a SNP set: SNP extraction, filtering and testing filtering parameters was done by the author, Angela Dale with help/contributions from Dr. Nicolas Feau and Dr. Sydney Everhart.  De novo genome assemblies were done by Dr. Braham Dhillon and Dr. Nicolas Feau.  Phylogenetic analyses: Neighbor joining phylogenetic analysis was done by the author, Angela Dale. Maximum Likelihood phylogenetic reconstruction was done by Javier F. Tabima under the supervision of Dr. Niklaus Grünwald. Divergence time between lineages was done by Dr. Nicolas Feau.  Detection of runs of homozygosity (ROH) was done by the author, Angela Dale, with help in development of a python script by Dr. Nicolas Feau. Determining the effects on protein content vii  from conversion to homozygosity was done by the author Angela Dale, with help from Javier F. Tabima who reconstructed each DNA strand in regions with ROH.  Identification and characterization of a pathogenicity hotspot was done by the author, Angela Dale with advice from Dr. Nicolas Feau.  Effects of ROH on phenotype was done by the author, Angela Dale, with contributions from collaborators. Dr. Clive Brasier provided advice on the test design, and testing on Rhododendron leaves was done at the CFIA under the direction of Dr. Guillaume Bilodeau. Testing on Larch and Douglas-fir log sections was done by the author, Angela Dale, using the facilities at FPInnovations.  Genome-level copy number variation was done by Dr. Nicolas Feau.  Mitotic recombination breakpoints were identified by the author, Angela Dale. Surrounding gene content, gene density and comparison with random datasets was done by Angela Dale.   OrthoMCL analysis and evolution of gene family size was done by Dr. Nicolas Feau.  Core and non-core genome analysis was done by Dr. Nicolas Feau. An original version of this work was done by both the author, Angela Dale, and Dr. Nicolas Feau. At the request of the author, Dr. Nicolas Feau extracted the non-core genome regions which were then analyzed for gene content and compared between lineages by the author, Angela Dale. A final, more robust viii  analysis was done by Dr. Nicolas Feau to use for the final version to submit for publication. This also included work on lineage-specific proteins and evolution of gene family size done by Dr. Nicolas Feau.  Genes encoding effectors: RxLR effector extraction was done by Dr. Sydney Everhart. CRN or Crinkler effector extraction was done by Dr. Nicolas Feau. Extraction of a random protein set and the CEGMA gene set, protein filtering, conversion to DNA and analysis of positive selection was done by Dr. Nicolas Feau. Statistical comparisons between datasets was done by the author, Angela Dale.  This collaboration was led by Dr. Richard Hamelin of UBC. In addition to the author, Angela Dale, the collaborators at UBC included Dr. Nicolas Feau and Dr. Braham Dhillon. Members in addition to the UBC team: Dr. Niklaus Grünwald, Dr. Sydney Everhart and Javier F. Tabima of Oregon State University and the USDA Horticultural Crops Research Unit; Dr. Guillaume Bilodeau from the Canadian Food Inspection Agency; Dr. Steve Jones of the Genome Sciences Centre; Dr. Jean Bérubé of the Canadian Forest Service; Dr. Clive Brasier of The Forestry Commision, UK (Alice Holt Lodge).    ix  Table of Contents Abstract ..................................................................................................................................... ii Lay Summary............................................................................................................................ iv Preface ........................................................................................................................................v Table of Contents ...................................................................................................................... ix List of Tables ............................................................................................................................ xi List of Figures ......................................................................................................................... xiii Acknowledgements ................................................................................................................... xv Dedication .............................................................................................................................. xvii Introduction: Phytophthora, the plant destroyer genus .................................................................1 Chapter 1: Literature Review.......................................................................................................6 Oomycetes ...............................................................................................................................6 Phytophthora in natural ecosystems .........................................................................................8 Influence of anthropogenic activities on natural communities ................................................ 10 Phytophthora in British Columbia ......................................................................................... 11 Oomycete and Phytophthora monitoring ............................................................................... 12 Phytophthora biology ............................................................................................................ 14 Phytophthora genomics ......................................................................................................... 16 Emergence and evolution of new pathogens........................................................................... 19 Phytophthora ramorum: An emerging pathogen introduced through nursery trade ................. 23 Objectives of study.................................................................................................................... 34 Chapter 2: Anthropogenic activities increase diversity of oomycete communities in urban and urban/natural interface environments and are pathway for potential new pathogens to natural ecosystems ................................................................................................................................ 36 Introduction ........................................................................................................................... 36 Methods ................................................................................................................................ 40 Results ................................................................................................................................... 48 Discussion ............................................................................................................................. 53 Chapter 3: Anthropogenic activities introduce new species and increase the frequency of Phytophthora in urban habitats compared to native habitats ...................................................... 82 x  Introduction ........................................................................................................................... 82 Methods ................................................................................................................................ 87 Results ................................................................................................................................... 94 Discussion ........................................................................................................................... 100 Chapter 4: Mitotic recombination and a two-speed genome drive evolution in asexual lineages of the sudden oak death pathogen Phytophthora ramorum ........................................................... 128 Introduction ......................................................................................................................... 128 Methods .............................................................................................................................. 132 Results ................................................................................................................................. 137 Discussion ........................................................................................................................... 144 Chapter 5: Conclusions ........................................................................................................... 155 References .............................................................................................................................. 162 Appendices ............................................................................................................................. 185 Appendix A: Supplementary tables chapter 2 ....................................................................... 185 Appendix B: Chapter 4, supplementary material .................................................................. 187     xi  List of Tables Table 2.1: Details of sampling areas. ..................................................................................... 62 Table 2.2: Descriptive statistics for fall samples..................................................................... 64 Table 2.3: Descriptive statistics for spring samples ................................................................ 65 Table 2.4: ANOVA results for diversity comparisons on samples that were rarefied to 300 sequences per sample. ............................................................................................................ 66 Table 2.5: OTUs sharing low sequence identity to Phytophthora taxa. ................................... 67 Table 2.6: Number of OTUs in each genus that were found in both fall and spring or that were exclusive to only one season. ................................................................................................. 68 Table 3.1: Species that were significantly over or under represented by site type (urban, interface or natural) and by urban site type (residential or agricultural) ................................ 112 Table 3.2: Unknown or putative new taxa and the origin of their closest hits in a homology search on Genbank. ............................................................................................................. 113 Table S2.7: Sequencing, filtering and taxonomy assignment details for the two sampling periods. ................................................................................................................................ 185 Table S2.8: OTU species level taxonomic identifications using manual identification and automated identification ...................................................................................................... 186 Table S4.1: Phytophthora ramorum isolates sequenced in this study. ................................... 218 Table S4.2: Number and percent of homozygous and heterozygous SNPs in Phytophthora ramorum individuals with ROH and representatives of the general population without ROH. ............................................................................................................................................ 223 Table S4.3: Classification of 485 327 SNPs as polymorphic or fixed within Phytophthora ramorum lineages and as heterozygous or homozygous ....................................................... 225 Table S4.4: Phytophthora ramorum individuals with Chromosomal Copy Number Variation ............................................................................................................................................ 226 Table S4.5: Number of Phytophthora ramorum proteins with amino acid differences between allele retained and allele lost through conversion to homozygosity after mitotic recombination. ............................................................................................................................................ 227 Table S4.6: Proteins enriched in scaffold seven as determined by MGSA analysis ............... 228 Table S4.7: Comparison of intergenic distances on scaffolds with mitotic recombination breakpoints with ten sets of random scaffolds. ..................................................................... 229 xii  Table S4.8: Nucleotidic and gene content of the de novo assemblies obtained for the four Phytophthora ramorum lineages. ......................................................................................... 230 Table S4.9: Protein family content predicted in the non-core genome of the four Phytophthora ramorum lineages. ............................................................................................................... 231 Table S4.10: Shared and unique protein models found in Phytophthora ramorum lineages. . 232 Table S4.11: MGSA analysis for enriched GO terms in set of proteins unique to the Phytophthora ramorum lineages. ......................................................................................... 233 Table S4.12: Number of SNPs in Phytophthora ramorum before and after filtering at various coverage and quality values, and the percentage of validation SNPs, expected SNPs and false SNPs found after each round of filtering. ............................................................................. 234 Table S4.13: False call rate between sequencing runs of the same Phytophthora ramorum individual before and after applying SNP filters for coverage and quality. ........................... 236 Table S4.14: Assembly statistics and completeness for eight de novo assemblies generated for Phytophthora ramorum........................................................................................................ 237 Table S4.15: Mean posterior values for eight selected parameters estimated under the three molecular clock models. ...................................................................................................... 238 Table S4.16: Median divergence times and 95% confidence intervals for Phytophthora nodes estimated under the three molecular clock models................................................................ 239 Table S4.17: Enriched gene ontology for Phytophthora ramorum proteins affected by ROH as determined by MGSA analysis ............................................................................................ 240 Table S4.18: Growth and sporulation of isolates with and without ROH on sapwood agar. .. 241 Table S4.19: Number of Phytophthora ramorum and P. lateralis Crinkler effectors identified in the two approaches of the Crinkler identification pipeline. ............................................... 242   xiii  List of Figures Figure 2.1: Map of sampling sites. ............................................................................................. 69 Figure 2.2: Sample design showing the hierarchical arrangement of sites within geographic area and site types............................................................................................................................. 70 Figure 2.3: Oomycete community composition in southwest British Columbia for fall and spring collections. ................................................................................................................................ 71 Figure 2.4: Distribution of OTUs across samples. ...................................................................... 72 Figure 2.5: Influence of site type on species profile ................................................................... 73 Figure 2.6: Relative abundance of OTUs by genera ................................................................... 74 Figure 2.7: Number of shared and private oomycete OTUs between site types .......................... 75 Figure 2.8: Number of samples each OTU is found in each site type. ........................................ 79 Figure 2.9: Diversity by site type. .............................................................................................. 80 Figure 2.10: Phylogenetic analysis with unknown putative Phytophthora species. ..................... 81 Figure 3.1: Stream baiting using mesh traps with rhododendron leaves. .................................. 117 Figure 3.2: Percentage of Phytophthora positive samples by site type ..................................... 118 Figure 3.3: Total unique and shared species between the three site types, urban, interface and natural. .................................................................................................................................... 119 Figure 3.4: Frequency of each Phytophthora species by site type in fall and spring using baiting in the lab for water and soil and meta DNA-barcoding on soil ................................................. 120 Figure 3.5: Weighted UniFrac distance between sites. ............................................................. 121 Figure 3.6: Shannon’s diversity of sites. .................................................................................. 122 Figure 3.7: The relationship between urban site type (residential or agricultural), sample substrate (soil or water) and species presence and absence patterns. ........................................ 123 Figure 3.8: Representation of the abundance of Phytophthora clades (circle diameter) in urban sites that are in residential or agricultural areas. ....................................................................... 124 Figure 3.9: Phytophthora species present in south west BC and their frequency based on collection method. ................................................................................................................... 125 Figure 3.10: Unique and shared species by method ................................................................. 126 Figure 3.11: Comparison of Phytophthora positive samples and number of species by method and by substrate. DNA m ........................................................................................................ 127 Figure 4.1: Evolutionary relationships in Phytophthora ramorum ............................................ 151 xiv  Figure 4.2: Comparison of heterozygote density, gene content and gene location in scaffold seven of Phytophthora ramorum. ............................................................................................ 152 Figure 4.3: Genome characteristics around mitotic recombination breakpoints. ....................... 153 Figure 4.4: Core and non-core genomes predicted in the four Phytophthora ramorum lineages. ............................................................................................................................................... 154 Figure S4.5: Phylogenetic reconstruction of Phytophthora ramorum individuals of four lineages. ............................................................................................................................................... 243 Figure S4.6: Runs of homozygosity caused by mitotic recombination. .................................... 244 Figure S4.7: Read-depth analysis for copy number variation in the Phytophthora ramorum isolate P1563 (EU1) isolate. .................................................................................................... 245 Figure S4.8: Comparison of predicted secreted protein content for the first 25 largest scaffolds in P. ramorum. ............................................................................................................................ 246 Figure S4.9: Lesion length on larch and Douglas-fir four and eight weeks post inoculation of individuals with ROH on scaffold seven compared to isolates without ROH. .......................... 247 Figure S4.10: Lesions on larch four weeks four weeks post inoculation for two isolates, one with ROH and one without. ............................................................................................................. 248 Figure S4.11: Gene family contraction and expansion in Phytophthora spp. ............................ 249 Figure S4.12: PCA of codon usage on protein coding genes of the Phytophthora ramorum NA2 isolate 05-16845 ...................................................................................................................... 250 Figure S4.13: Example of degenerated protein sequence found in the non-core genomes of the NA2 and EU1 lineages. ........................................................................................................... 251 Figure S4.14: Distributions of dN/dS values obtained for the RxLR and the Crinkler gene sets compared with the CEGMA and random protein sets. ............................................................. 252 Figure S4.15: Phylogenetic relationships and evolutionary history of the CRN subfamily expanded in the P. ramorum EU1 and NA2 lineages. .............................................................. 253 Figure S4.16: Lesion length on Rhododendron leaves of isolates with ROH on scaffolds seven and 100 compared to isolates with no ROH. ............................................................................ 254 Figure S4.17: Complementary approaches used to identify P. ramorum and P. lateralis Crinkler effectors. ................................................................................................................................. 255 Figure S4.18: Count of each RxLR motif ................................................................................ 256 Figure S4.19: Proportion of each RxLR motif ......................................................................... 257 Figure S4.20: OrthoMCL clustering of Crinkler effectors found in the four P. ramorum lineages. ............................................................................................................................................... 258 xv  Acknowledgements I would like to offer my sincere thanks and appreciation to my supervisor, Dr. Richard Hamelin for his support, guidance, inspiration, and the opportunity to do this work. Also for all of the great scientific discussions; not only did they offer inspiration and helped to define my work, but they were moments of fun and laughter that created a great team environment that I will miss working in.   A special thanks to Dr. Nicolas Feau. It has been an amazing and inspirational journey. Thank you for your encouragement, guidance, advice, and for all the fun and laughter we shared over the past six years. Thank you for encouraging me to step outside of my comfort zone, to learn to code with Python, and to look beyond the surface of the data for the good stuff. Thank you for listening to my endless stream of thoughts and for help in turning my ideas into something concrete. And overall, thank you for your help in producing this work.  A special thanks to FPInnovations for the opportunity to do this work and for the support provided during the project. A special thanks to my colleagues for the inspiration, moral support, help and guidance over the years.   Thank you to my committee members, Dr. Kermit Ritland, Dr. Sue Grayston, and Dr. William Mohn for their guidance, inspiration and for their time.  xvi  A big thank you to all of the members of the UBC Hamelin lab, whose support was instrumental during the project, especially during the field season and in processing the hundreds of samples.  This project was done as part of the Taiga project, made possible by funding and collaboration from Genome Canada, Genome British Columbia, The University of British Columbia, University of Laval, FPInnovations, The Canadian Food Inspection Agency, Natural Resources Canada, Canada’s Michael Smith Genome Sciences Centre, and Boreal Genomics. Many thanks to the contributors for the opportunity to do this work.   A special thank you to my family and to my parents. Your never-ending support and your faith in me were instrumental in me being here today. Thank you for helping me through the tough times and for celebrating during the good times.   xvii  Dedication I dedicate this thesis to my children, Matthew, Nathan and Noah, to my sister, Lorelei, my niece Jorden, and to my parents. You are my inspiration. You make me strong and you continue to bring happiness into my life. You have taught me that you can do anything you put your mind to.   1  Introduction: Phytophthora, the plant destroyer genus The Phytophthora de Bary genus, first described in the late 1800s by Heinrich Anton de Bary , is considered one of the most destructive groups of plant pathogens in history. Phytophthora infestans (Mont.) de Bary caused the notorious Irish potato famine which resulted in more than (De Bary, 1876) one million deaths and the migration of more than two million people. Phytophthora species infect a wide range of hosts including economically valuable tree species, both angiosperm and gymnosperm, and many important agricultural species. Many known species of Phytophthora are invasive and have been introduced through nursery and agricultural trade.   In the United States, the recently emerged Phytophthora ramorum Werres, De Cock & Man in 't Veld has killed large numbers of oak trees along the West Coast and the range of the pathogen is increasing despite efforts to eradicate outbreaks (Kanaskie et al., 2007). The primary range of the epidemic is over 750 km from southern California to southern Oregon (Meentemeyer et al., 2008). Over 230 000 oak and tanoak were estimated to have been killed by 2005 in the Big Sur ecosystem of California, the heart of the epidemic (Meentemeyer et al., 2008). Phytophthora ramorum is believed to have been spread via the nursery trade and is invasive in California, Oregon and the UK (Grünwald et al., 2012). The origin of the pathogen remains unknown (Rizzo & Garbelotto, 2003; Grünwald et al., 2012). In the UK the pathogen has also rapidly spread despite intensive control measures and has recently jumped hosts from the previously reported known angiosperm hosts to Japanese larch (Larix kaempferi (Lamb.) Carr.) (Brasier & Webber, 2010). This is the first time that the pathogen has been reported to cause mortality and 2  widespread damage on conifers (Brasier & Webber, 2010). Previously, some conifer species were shown to be susceptible hosts in laboratory studies (Chastagner et al., 2004; Hansen et al., 2005; Denman et al., 2005). Infections were observed in heavily infected areas on some conifer species including Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), California red fir (Abies magnifica A. Murray bis), and coast redwood (Sequoia sempervirens (Lamb.) Endl.) (Davidson et al., 2002; Maloney et al., 2002; Brasier & Webber, 2010; Chastagner & Riley, 2010). These cases were associated with high inoculum loads, and infection appears to be limited to needles, shoots, and branches. The recent jump to a new conifer host and subsequent spread on a landscape scale raises concern in western North America both for forest health due to the relative abundance of susceptible conifers in forests, and for the forest industry that relies heavily on some of these species.  Other Phytophthora species attacking trees have also spread widely. Phytophthora lateralis Tucker & Milbrath, also an invasive pathogen, infects Port Orford Cedar (Chamaecyparis lawsoniana (A.Murray) Parl) and has spread throughout the natural range of the tree (Hansen, 2008). The epidemic has affected both the wood export market as this is a valued species in foreign markets as well as the nursery trade as it is also a valued horticultural species (Hansen, 2008). Phytophthora cinnamomi Rands has also been introduced and spread to many countries worldwide. Phytophthora cinnamomi has a very broad host range, from grasses and ferns to several different tree species. Many of the species affected by this pathogen are important horticultural and forest species and P. cinnamomi has been responsible for widespread epidemics in several forest species including Chestnut in the USA, cork oak in Europe, and in Australia the pathogen affects many species and is causing extensive damage to Jarrah forests and heathlands 3  (Hansen, 2008). The devastating effects of invasive Phytophthora species have resulted in trade restrictions being implemented in many countries in order to decrease the probability of further spread of these pathogens. Worldwide, millions of dollars have been spent on quarantine efforts, attempted control, and eradication programs.   In addition to economic costs, non-indigenous pathogens can also have wide ranging effects on ecosystems (Loo, 2009). The loss of a host species can change the dynamics of an ecosystem, especially when that species is a dominant one (Ellison et al., 2005). For example there can be changes in microclimate, hydrology, and nutrient fluxes (Ellison et al., 2005; Loo, 2009), and important food sources and habitat may be lost for some wildlife (Ellison et al., 2005) . Population reduction of Port-Orford cedar in Oregon and California by the pathogen P. lateralis could result in significant ecological changes including changes in soil calcium content, loss of habitat structure, as well as loss of stream structure and stability along riparian zones (Hansen et al., 2000; Loo, 2009). Extensive mortality of tanoak (Lithocarpus densiflorus (Hook. & Arn.) Rehd.) in Oregon and California forests due to P. ramorum increases fire risk in those forests, may affect animals dependent on the acorns as a food source, and could cause extended periods of chaparral vegetation in some areas (Hansen, 2008).  Introducing species to new environments and other anthropogenic activities such as agricultural activities may also have an effect on the evolutionary processes occurring in a community. For example bringing different species together could increase opportunities for hybridization between closely related species. Agricultural systems may also increase opportunities for these events as host plants are in high abundance and in close proximity to one another allowing for 4  increased contact between pathogens. Phytophthora xalni Brasier & S.A. Kirk 2004 is a hybrid species currently spreading throughout European ecosystems and responsible for alder decline. It has been suggested that it emerged after the introduction of Phytophthora uniformis Brasier & S.A. Kirk 2004 from North America followed by hybridization with Phytophthora xmultiformis Brasier & S.A. Kirk 2004 (Ioos et al., 2006; Aguayo et al., 2016).   The international trade of plants and trees has increased the global movement of plant and forest pathogens. Understanding the effect of this movement and of other forestry and agricultural activities on pathogen, population, and community evolution is important in order to mitigate negative effects on natural ecosystems. Improving our understanding of how activities alter pathogen communities and how that leads to disease can help in the development of management strategies to mitigate the effects of those activities. For instance if nursery activities lead to evolution of new species and spread into natural systems, then measures need to be taken to reduce or prevent pathogen escape from these systems.   So far the interest in Phytophthoras has stemmed from the extensive damage that members of this genus have inflicted on crops and forests. However little is known about the natural roles and communities of Phytophthora, and the oomycetes in general. Likewise, we have a poor understanding of the influence of anthropogenic activities on natural communities of oomycetes. This review will first give a general overview of the oomycetes, followed by a review of Phytophthora communities and the niches they occupy. The second part of the review will look at Phytophthora biology, genomics and evolutionary processes, specifically the importance of 5  interspecific hybridization. The third part of the review will look at focus on an example of an introduced and emerging pathogen, P. ramorum.   6  Chapter 1: Literature Review  Oomycetes  The oomycetes are commonly known as watermolds. They are part of the Superphylum heterokonts (also known as Stramenopiles) (Baldauf, 2003). They are superficially similar to fungi and share some key characteristics such as a mycelial growth habit, spore production, infection strategies and they are able to reproduce both sexually and asexually (Latijnhouwers et al., 2003). However, they are phylogenetically distinct from fungi (Baldauf, 2003) and differ in some key characteristics such as the presence of cellulose as the major constituent of their cell walls, instead of chitin, and they are diploid during the vegetative state (Latijnhouwers et al., 2003).  The oomycetes are a widespread group of organisms containing both saprobes and pathogens; however they are mostly known as plant pathogens. They are responsible for infamous crop diseases such as the potato late blight that was responsible for the Irish potato famine. In a ranking of the top ten plant pathogenic oomycetes, Phytophthora species (foliar blights, root and stem rots) made up the majority of the species (Kamoun et al., 2015). Other important species are the downy mildews such as Hyaloperonospora arabidopsidis and Plasmopara viticola, as well as the damping off and root rot pathogen, Pythium ultimum, and the white rust, Albugo candida (Kamoun et al., 2015). However oomycetes can also infect other organisms including animals and fish. Some of the most important diseases of fish are from the family 7  Saprolegniaceae (Noga, 1993; Torto-Alalibo et al., 2005; Phillips et al., 2008), including Saprolegnia parasitica which is especially important on salmon and trout species (Torto-Alalibo et al., 2005). Pythium is often considered an important disease of young plants in nurseries; however Pythium insidiosum is an important pathogen of mammals, including humans, causing the disease Pythiosis (Phillips et al., 2008; Gaastra et al., 2010).  Oomycetes display a range of lifestyles from obligate biotroph to necrotroph (Fawke et al., 2015). They can share an intimate relationship with a single host as in the case of obligate biotrophs, or they can exhibit a generalist lifestyle, infecting hundreds or thousands of different host species as in the forest pathogens P. ramorum or P. cinnamomi (Hardham, 2005; Cahill et al., 2008; Grünwald et al., 2012). Many are common inhabitants of aquatic and terrestrial environments. Natural dispersal is through wind and water (Fawke et al., 2015); however their frequent association with plant material, soil and roots makes them very amenable to human dispersal through trade, agricultural and horticultural activities (Desprez-Loustau et al., 2010). In general, we have a much better understanding of the species that cause important diseases with a large economic or ecological impact. Our knowledge of these microorganisms in natural habitats is more limited (Arcate et al., 2006). Several recent studies of oomycetes in soil environments show a dominance of the genus Pythium in both agricultural and natural environments (Arcate et al., 2006; Nelson & Karp, 2013; Sapkota & Nicolaisen, 2015; Riit et al., 2016; Esmaeili Taheri et al., 2017). In natural environments, oomycetes may play a role in driving succession (Dickie et al., 2017), nutrient cycling (Marano et al., 2016), and creating or maintaining forest diversity (Packer & Clay, 2000).  8  Phytophthora in natural ecosystems Phytophthora species have long been known as pathogenic species associated with agricultural activities. However, studies done since the mid-1990s suggest that a diverse community of Phytophthora species exist in natural ecosystems (Hansen et al., 2012a), many of which rarely cause noticeable disease or disturbance to the ecosystem (Hansen et al., 2012a). In part because of the emergence of new destructive Phytophthora species, there has been increased interest in studying Phytophthora communities in forest soils and streams (Hwang et al., 2008; Reeser et al., 2010, 2011; Hüberli et al., 2013; Oh et al., 2013; Shrestha et al., 2013; Brazee et al., 2016; Stamler et al., 2016), or to link tree declines with a possible Phytophthora pathogen (Greslebin et al., 2005, 2007; Balci et al., 2007; Vannini et al., 2013). These studies have uncovered many new Phytophthora species. The genus now contains over 100 described species, half of which have been described only in the past ten years (Kroon et al., 2012; Hansen et al., 2012a). The genus has a global distribution with some widely distributed species and others with a limited distribution. Phytophthora species have been found in waterways, forest soils and more recently, are being discovered in the tree canopies of forests (Hansen et al., 2012a, 2017).  Phytophthora communities from waterways appear to be dominated by species belonging to a monophyletic group designated as ITS clade 6, based on the sequence of the internal transcribed spacer of the ribosomal cluster (Jung et al., 2011). Species from this group are rarely associated with disease (Brasier et al., 2003). Several of the species appear to be saprophytic, often being found in riparian habitats or on submerged wood and plant material (Brasier et al., 2003; Jung et al., 2011). In North America Phytophthora gonapodyides (H.E. Petersen) Buisman, 9  Phytophthora lacustris Brasier, Cacciola, Nechwatal, Jung & Bakonyi, and Phytophthora chlamydospora (Brasier et al., 2003) are the species most frequently recovered in waterways (Hwang et al., 2008; Hulvey et al., 2010; Reeser et al., 2011; Hansen et al., 2012b). Although these species have a rather ubiquitous distribution in the northern hemisphere, they are limited in the southern hemisphere where other species dominate (Jung et al., 2011).  Soil communities show similar trends as waterway communities, except that invasive pathogens appear to be more prevalent. In the soil of declining oak forests in the Eastern US, nine different Phytophthora species were found but only P. cinnamomi was associated with disease symptoms (Balci et al., 2007). Phytophthora cinnamomi was the most frequently isolated species in forest soils in the eastern US, and the second most frequently isolated species in Oregon after P. ramorum (Balci et al., 2007; Reeser et al., 2010). Both of these species are believed to be introduced  (Rizzo et al., 2002, Hansen, 2008, Hansen et al., 2012a). Their presence in such high frequency in natural forest soils highlights the importance of invasive species in Phytophthora communities.  The forest canopy represents another niche where Phytophthora are found that has only recently been explored (Hansen et al., 2012a). Phytophthora ramorum is an aerially dispersed pathogen infecting foliage and stems of its hosts. Since its discovery, monitoring activities in the US have uncovered additional Phytophthora species in the forest canopy. Phytophthora nemorosa E. M. Hansen and Reeser and Phytophthora pseudosyringae T. Jung & Delatour were frequently found in California and Oregon forest canopies (Wickland et al., 2008; Reeser et al., 2010). In Oregon, P. gonapodyides and P. chlamydospora were also present in soils and canopy drip, but in lower frequency than in waterways (Reeser et al., 2010) suggesting not only a global distribution in 10  waterways for these two species, but possibly also in soils and forest canopies. Aerial dispersal can greatly increase the efficiency and length of dispersal and can generate rapid spread across and even between continents  (Brown & Hovmøller, 2002). This can be a serious problem for epidemics in plantations where a pathogen can quickly destroy an entire stand.  Influence of anthropogenic activities on natural communities Some Phytophthora species, such as Phytophthora cryptogea Pethybr. & Laff. or Phytophthora citricola Sawada, are found globally. Whether this ubiquitous distribution is natural or has taken place through trade and their association with the agriculture and food industry is not known. Hansen et al. (2006) report finding P. citricola, Phytophthora cambivora (Petri) Buisman, Phytophthora syringae (Berk.) Kleb. and Phytophthora cactorum (Lebert & Cohn) J. Schr. only occasionally in forest systems despite their dominance in urban and nursery settings. This pattern of distribution suggests that movement through trade has resulted in the ubiquitous nature of these agricultural pathogens, followed by subsequent escape or migration into forests. How much agricultural practices have changed the natural species composition of Phytophthora or of oomycetes in general is largely unknown. In the eastern US, Hulvey et al. (2010) compared isolates found in nurseries and in natural waterways and found three species common to streams and nurseries: P. citricola, Phytophthora hydropathica Hong, C. and Gallegly, M. and Phytophthora citrophthora (R.E. Sm. & E.H. Sm.) Leonian (the P. citricola complex has since been revised and P. citricola may be the wrong name). For the species from the P. citricola complex, an identical DNA fingerprint in nursery and stream samples suggest an escape (Hulvey et al., 2010). 11   Phytophthora in British Columbia Natural Phytophthora communities have not been surveyed in British Columbia and Canada. Similar to other countries, it is likely that BC forests harbour both native and introduced Phytophthora species. Some common forest pathogens have been found in irrigation water, gardens and parks and may have made their way into forests as well. In a study from BC in the 1960s, some common agricultural pathogens were found in irrigation water from orchards in the Okanagan and Similkameen valleys in the southern part of the province (McIntosh, 1966). Several irrigation sources sampled identified P. cactorum as the most common contaminant. In addition, P. cambivora, P. citricola, and Phytophthora megasperma Drechsler as well as some unidentified species were found. Identifications were made from morphological characteristics and there was some level of uncertainty associated with the identifications (McIntosh, 1966); furthermore many species have since been re-evaluated using molecular techniques, and isolates that were previously identified as one species, for example P. megasperma, turned out to be different species. In 1965, Atkinson looked at the causal agent of root rot in C. lawsoniana in nurseries and three gardens in southern BC in the Fraser Valley and in Victoria. He found P. lateralis, P. cinnamomi, and an unidentified Phytophthora species (Atkinson, 1965). More recently, P. lateralis has been found associated with diseased ornamental C. lawsoniana in the city of Victoria (Gorman, 1993) where infected trees had to be eradicated from parks and boulevards (Michelle Gorman, personal communication). Phytophthora ramorum has also been found in BC nurseries (Bilodeau et al., 2007). There are likely other pathogens present in agricultural systems that may have forest hosts as well, so there is a potential that these species 12  may have made their way into forest systems affecting the natural Phytophthora community, and the natural host community as well.  Oomycete and Phytophthora monitoring Oomycete and Phytophthora monitoring or surveying have long been conducted using stream baiting methods or by directly isolating the pathogen from infected plant tissues. Stream baiting involves using selected bait materials placed in some sort of container or bag which is anchored in a waterway allowing water to pass through the trap. The motile zoospores of oomycetes swim to the bait tissues, encyst and colonize. After a couple of days to weeks, the pathogen can be isolated from the necrotic tissues. Many variations of this technique have been used, ranging from orchard fruits placed in tin cans with holes punched into the sides for surveys of species found in orchard irrigation water (McIntosh, 1966), to leaves of target species placed in mesh bags and floated in streams to gain a general idea of the species present in an ecosystem (Hulvey et al., 2010; Reeser et al., 2011; Hüberli et al., 2013; Shrestha et al., 2013; Brazee et al., 2016). Rhododendron leaves are now commonly utilized as bait to attract a wide variety of species, but other plant tissues are sometimes used to target species that are more host specific or to match host and pathogen species that are found in particular ecosystems. For instance, Stamler and group used poplar and willow leaves as bait in stream surveys in the Southwestern USA as those were the common tree species in the ecosystems being surveyed (Stamler et al., 2016). Quercus robur leaves were used as bait in Phytophthora surveys of Oak systems to target species potentially involved in Oak decline (Balci et al., 2007). Many alternative techniques are now being commonly utilized to increase sampling efficiency. For instance baiting in the lab, where 13  soil and water samples are processing in the lab rather than in situ (Hansen et al. 2017; Balci et al., 2007; Oh et al. 2013). This has allowed surveying to be conducted on soil samples as well as water. Soil samples are placed in containers, flooded with water, and bait leaves are floated on top to attract the pathogen. Filtering is another method where water samples are passed through a filter and the pathogen can either be isolated from the filters, or directly identified using molecular techniques (Balci et al., 2007; Hwang et al., 2008; Reeser et al., 2011; Rollins et al., 2016). Finally, direct identification and characterization of oomycete and Phytophthora communities is being done using next generation sequencing techniques (Vannini et al. 2013; Català et al., 2015, 2017; Sapkota and Nicolaisen 2015; Prigigallo et al., 2016; Riit et al., 2016; Burgess et al., 2017; Dickie et al., 2017; Esmaeili Taheri et al., 2017)(Català et al., 2015, 2017; Burgess et al., 2017b).    Oomycete and Phytophthora species require water and favorable environmental conditions for dissemination of the motile zoospores. Like many other microorganisms, environmental conditions can dictate timing for various life cycle stages. Many survey and monitoring studies are conducted over different time frames to capture a wider sampling of the species present (Balci et al., 2007; Reeser et al., 2011). Time of year affected the diversity and species detected in streams over a year-long survey conducted in the USA (Hwang et al. 2008). Sampling in spring versus fall resulted in twice as many species in one study on Phytophthora in Oak stands (Balci et al., 2007). Selection of monitoring methods, season, habitat and geographic location can all affect results in Phytophthora or oomycete community studies which may complicate disease management or underestimate natural community diversity.   14  Phytophthora biology Life cycle, reproduction and dispersal  Phytophthora can be either homothallic (possessing the ability to self) or heterothallic (requiring individuals with different mating types for reproduction) (Brasier, 1992); however some outcrossing or self-fertilizing can be found in either mating system (Brasier, 1992; Goodwin, 1997). Phytophthora have a mixed mating system, possessing both a sexual and an asexual phase. Chlamydospores and sporangia are spores produced during the asexual phase and oospores during the sexual phase. Sporangia are released under favorable conditions which could be during periods of rainfall or extended leaf wetness in the rainy seasons for aerial species (Davidson et al., 2005; Hansen et al., 2008, 2012a), or in times of extended saturation in soils (Hansen et al., 2008). These spores and can either directly germinate on the host surface or germinate and release zoospores, a type of spores with flagella, which can swim to host tissue by chemotaxis. Once in contact with host tissue the zoospores encyst and initiate infection. Chlamydospores are asexual resting spores that can remain dormant in host tissue or soil and germinate under suitable conditions (Grünwald et al., 2012). During the sexual phase, the sexual reproductive organs (oogonia and antheridia) are produced, in the presence of the opposite mating type for heterothallic species, and fusion of the organs results in the production of an oospore (Grünwald et al., 2012). Oospores are thought to be resting spores that can withstand extreme conditions (Latijnhouwers et al., 2003; Grünwald et al., 2012).   15  Infection biology Phytophthora are hemibiotrophs possessing both biotrophic and necrotrophic phases (Latijnhouwers et al., 2003). They colonize and interact with living plant tissue and in the later necrotrophic phase when the plant cells die, they feed on the dead cells (Latijnhouwers et al., 2003). Once established on the host, Phytophthora produce appressoria that penetrate the host where the pathogen can then grow within the intercellular matrix (Kamoun, 2006; Birch et al., 2006). Once inside the host Phytophthora secrete proteins called effectors that manipulate host structure or function (Huitema et al., 2004; Kamoun, 2006). Effectors can be secreted into the plant apoplast or inside host cells through infection structures such as haustoria or through infection vesicles (Kamoun, 2006).   Apoplastic effectors include several classes of proteins involved in infection or in blocking plant defense processes enabling pathogens to overcome their hosts. Some proteins trigger host cell death such as the Nep1-like proteins (NLP) and small cysteine-rich proteins. The small cysteine rich proteins are specific to oomycetes (Jiang & Tyler, 2012) whereas the NLP proteins have been found in oomycetes, bacteria and fungi (Pemberton & Salmond, 2004). Phytophthora species contain much larger NLP proteins families in comparison to other closely related oomycetes (Jiang & Tyler, 2012).  Many effectors are under diversifying selection, most likely to evade host recognition, and are poorly conserved. However elicitins are a highly conserved group of proteins under purifying selection that are unique to Phytophthora and Pythium (Panabières et al., 1997; Jiang et al., 2006). Elicitins induce a host response called the hypersensitive response in plants (Jiang et al., 16  2006; Kamoun, 2006). Jiang et al. (2006) found 17 distinct clades in the elicitin gene family, 13 of which included proteins from three or more Phytophthora species suggesting that divergence in the gene family pre-dates species divergence.   The cytoplasmic effectors include the RXLR effectors and the CRN or crinklers and necrosis inducing effectors. RXLR effectors appear to suppress the host immune response that often results in host cell death (Jiang & Tyler, 2012). The originally identified CRN effectors cause cell death and leaf crinkling in tomato and Nicotiana species (Torto et al., 2003; Kamoun, 2006). The C-terminus of the proteins are highly divergent both from their homologs in other Phytophthora species as well as between paralogs within species (Jiang et al., 2008; Haas et al., 2009; Schornack et al., 2010). Jiang et al. (2008) found a large super family of RXLR effectors (avirulence homolog or Avh genes) that they believe has evolved from a single gene in a common ancestor of Phytophthora sojae Kaufm. & Gerd. and P. ramorum. Many RXLR proteins are under positive selection (Win et al., 2007; Jiang et al., 2008), and the majority of the genes are in frequently rearranged non-conserved regions of the genome, and do not form a homologous cluster based on sequence homology (Jiang et al., 2008).   Phytophthora genomics Whole genome sequencing has allowed for some important observations and inferences on the evolution of the Phytophthora genus as well as individual species within the genus. Genome size varies widely in the Phytophthora species sequenced to date. Phytophthora infestans has experienced large-scale genome expansion and is approximately 240 Mb, whereas the genomes 17  of other species are much smaller at 95 Mb for P. sojae and 65 Mb for P. ramorum (Tyler et al., 2006; Haas et al., 2009). Phytophthora infestans has a much higher amount of repetitive DNA and transposable elements, accounting for approximately 74% of the genome (Haas et al., 2009). Seidl et al. (2012) show large number of gene gains and duplications in the evolutionary history of the oomycetes and of the Pythium – Peronosporales clade, and a higher amount of duplications leading to the Phytophthora genus. There are large expansions of the protein groups involved in host pathogen interactions in oomycetes (Seidl et al., 2012) and there is copy number variation in genes between strains of a species, for example in P. sojae (Qutob et al., 2009). The differences in copy numbers of a virulence or avirulence gene can result in virulence differences between strains (Qutob et al., 2009). Copy number variation has been found in Phytophthora genomes, however has not been reported in fungal genomes (Raffaele & Kamoun, 2012).  Genome organization likely plays an important role in Phytophthora evolution. Comparative genome analyses of Phytophthora species show a distinct pattern of gene rich regions interspersed by gene sparse regions (Haas et al., 2009; Raffaele et al., 2010a). Gene dense regions have short intergenic distances and gene order is highly conserved. In gene sparse regions, intergenic regions are larger and gene order is not conserved (Haas et al., 2009; Raffaele et al., 2010a). This pattern is most pronounced in P. infestans which has much larger intergenic distances and a much higher repeat content compared to P. ramorum and P. sojae (Haas et al., 2009). The gene dense regions are rich in core orthologous genes and the gene sparse regions contain effector genes. In a core set of 7,113 orthologous genes sharing a 1:1:1 relationship between P. infestans, P. sojae, and P. ramorum, approximately 90% were located in the gene dense, conserved regions (Haas et al., 2009). These included genes involved in cellular processes 18  such as DNA replication, transcription, and translation. In contrast, effector or virulence genes are commonly found in the gene sparse regions (Raffaele et al., 2010b). This pattern was also found by Raffaele et al. (2010a) in a genome comparison of the clade 1c species – P. infestans, Phytophthora ipomoeae Flier & Grünwald, Phytophthora mirabilis Galindo & H.R. Hohl, and Phytophthora phaseoli Thaxt. Raffaele et al. (2010a) also found signs of accelerated evolution in the genes located in repeat rich, gene-sparse regions compared to those located in the gene-dense regions. Genes showed signs of positive selection and there were more structural polymorphisms in these regions. Many of the genes found in the repeat rich, gene-sparse regions were plant-induced genes, including many effectors Raffaele et al. (2010a). Haas et al. (2009) hypothesize that the mobile elements and dynamic nature of the repeat rich, gene-sparse regions could cause more recombination and gene loss and gene gain events in RXLR effector genes and may be a mechanism involved in the speed of evolution of P. infestans.   In addition to faster evolution in regions of the genome, genes involved in host interactions evolve more rapidly relative to the rest of the genome. The genomes of Phytophthora ramorum and P. sojae were sequenced by Tyler et al. (2006) and comparisons revealed rapid expansion and diversification of gene families associated with pathogenicity, in particular with the RXLR gene family. In addition they found a faster rate of evolution in the collective group of secreted proteins (secretome) relative to the proteome (Tyler et al., 2006).   19  Emergence and evolution of new pathogens Pathogen evolution can happen through many processes such as clonal speciation, host jumps, horizontal gene transfer and hybridizations (Brasier, 1995; Stukenbrock & McDonald, 2008; Stukenbrock, 2013, 2016). Some processes are slow and can take many generations to result in speciation; this is the case of allopatric speciation in a stable habitat. However, some events can result in the rapid evolution of new species. Sudden changes in a population or habitat can result in increased selection causing periods of rapid evolution (Brasier, 1995). There can be several causes of sudden change resulting in increased selection including massive habitat changes, species introductions, and host jumps. Species that are introduced from their native range to a new environment often confront new conditions and hosts. Successful introduced species adapt to the new environment and hosts and may result in clonal speciation or they may hybridize with existing species to create hybrid species able to establish in the new environment (Brasier, 1995, 2001; Stukenbrock & Bataillon, 2012; Gladieux et al., 2014).  Hybridization Hybridization may provide an opportunity for the evolution of a new species. In many cases, if hybrids are able to form, they introgress back to one or both parental species and do not persist as an independent species (Hegarty, 2012). However in some cases hybrids are able to occupy a new niche different from that occupied by the parental species. Transgressive segregation and increases in genetic variation or genome plasticity can generate novel gene combinations creating hybrid individuals that are pre-adapted for novel habitats (Rieseberg et al., 1999; Stukenbrock & Bataillon, 2012; Hegarty, 2012; Callaghan & Guest, 2015; Stukenbrock, 2016). 20  The novel niche and gene combinations can lead to rapid evolution, hybrid speciation or the emergence of new diseases (Rieseberg et al., 1999; Stukenbrock & Bataillon, 2012; Hegarty, 2012; Callaghan & Guest, 2015; Stukenbrock, 2016).  Several examples exist of pathogens that have emerged on new hosts or in new environments as a result of hybridization. Phytophthora andina Adler & Flier, a hybrid between P. infestans and an unknown Phytophthora parent (Goss et al., 2011a), is an emerging pathogen of Solanum species in the Andes. Although both pathogens infect Solanum hosts, the species they infect are different and there is little overlap (Oliva et al., 2010; Goss et al., 2011a) suggesting that hybridization could have led to a host range expansion or shift (Goss et al., 2011a). Phytophthora xalni was first discovered on Alder in Europe (Brasier et al., 1995). The pathogen resembled P. cambivora but was suggested to be a group of heteroploid hybrids (Brasier et al., 1999). Three hybrid sub-species were described as P. alni subsp. alni, P. alni subsp. uniformis Brasier & S.A. Kirk, and P. alni subsp. multiformis Brasier & S.A. Kirk (Brasier et al., 2004); and subsequently reclassified as P. xalni, P. uniformis, and P. xmultiformis (Husson et al., 2015). Phytophthora xalni is widespread across most of Europe, and it is also the most aggressive pathogen of Alder (Brasier et al., 2004). It is a hybrid between P. uniformis, and P. xmultiformis (Ioos et al., 2006). Phytophthora uniformis may have been generated by an ancient hybridization event between P. cambivora and another Phytophthora species and P. xmultiformis could have been generated interspecific hybridization of two unknown species (Ioos et al., 2006; Husson et al., 2015). Phytophthora xalni likely hybridized with P. uniformis after it was introduced from North America (Aguayo et al., 2016). It appears that these Phytophtora species are specific to 21  Alder, and also that Alder is not a host of P. cambivora (Petri) Buisman (Brasier et al., 1999) suggesting that hybridization has led to a host jump.   Hybridization can also be an important source of genetic exchange between closely related species. Introgression through backcrossing of hybrid species with parental species can introduce new genes from one parental species to the other. In Europe, the Dutch elm pathogen Ophiostoma novo-ulmi Brasier, which was initially introduced as a few clones of only one mating type which were spreading asexually, acquired the second mating type gene as well as other pathogenicity genes through introgression with Ophiostoma ulmi (Buisman) Nannf. (Brasier, 2012). The new genes and ability to sexually recombine would have increased the long term evolutionary potential of the pathogen evident by the sudden reduced impact of a virus that had been spreading through the pathogen population (Brasier, 2012). In cases where hybrids are unable to produce an F2 generation, they may still be able to backcross with parental species allowing gene flow back to the parental populations. In crosses between Phytopthora capsici Leonian and Phytophthora tropicalis (Aragaki & J.Y. Uchida) hybrids were unable to reproduce with each other however they were able to reproduce with the P. capsici parent (Donahoo & Lamour, 2008).  Influence of disturbances on hybridization  Hybridization between species has often been linked to human activities or disturbances in natural habitats (Callaghan & Guest, 2015; Burgess, 2015) . There are many examples of hybridization when an organism is introduced into a region that a closely related species inhabits. 22  Hubbs (1955), in his work with hybridization in fish species, made the important observation that the scarcity of one species and abundance of another closely related species often led to hybridization due to difficulties of the rare species in finding the proper mate. Introduced species hybridizing with native species may often be due to a lack of reproductive opportunities with conspecific mates in the introduced individuals. Hybridization can also occur as a result of human activities, habitat disturbance, or environmental changes if the changes bring together or increase the frequency of species that were once separated spatially or temporally (Hegarty, 2012). Nursery systems have very different environmental conditions than natural systems. The constant influx of plant material often brings pathogens together that would normally be separated in a natural habitat. In addition, there can be large numbers of host plants all within close proximity and abundant watering which are conditions that are conducive to pathogen development and that are usually not associated with natural habitats. These conditions may increase the opportunities for hybridization. There are several examples of hybridization of Phytophthora species in nurseries. Phytophthora alni subsp. alni is believed to have originated from a hybridization event in nurseries. Hybridization appears to be relatively common between Phytophthora nicotianae Breda de Haan and P. cactorum in hydroponic cultures in the Netherlands (Man in ’t Veld et al., 1998; Bonants et al., 2000). Hybrids were recovered from different hosts in different years and molecular analyses suggest that the hybrids were generated on multiple occasions. Some of the hosts that the hybrids were isolated from were not host species of the parental species suggesting a possible shift in host range in the hybrids (Bonants et al., 2000). Hybrids of P. cactorum and Phytophthora hedraiandra De Cock & Man in 't Veld, believed to be introduced, were also found in nurseries in the Netherlands as well as in a few parks suggesting movement through infected nursery material (Man in ‘t Veld et al., 2007). 23  These hybrids were also able to infect numerous hosts that the parental species could not infect (Man in ‘t Veld et al., 2007). These examples suggest that nursery environments may be conducive to hybridization events.  Hybridization in Phytophthoras appears to be relatively frequent, and may be an important process in the evolution and adaptability of these species (Bertier et al., 2013; Burgess, 2015). In some environments, hybridization appears to be a natural and frequent process (Burgess, 2015). How anthropogenic activities or niche preferences have influenced these processes is also unknown; however the more frequent finding of natural hybrids in nursery settings suggests that there is an influence.   Phytophthora ramorum: An emerging pathogen introduced through nursery trade Successful establishment into a new habitat does not necessarily require a pathogen to develop novel adaptations. If the environment or hosts are similar enough the pathogen may be able to establish and successfully spread with little to no adaptive changes. In other cases the pathogen may be able to establish, but selection pressures may be high favoring certain genotypes over others and may lead to clonal speciation (Brasier, 1995). Phytophthora ramorum is an invasive pathogen that has been introduced to North America and Europe and is spreading asexually throughout the range where it was introduced.  24  History of P. ramorum Phytophthora ramorum was discovered in the mid-1990s in both Europe (Werres et al., 2001) and the US (Rizzo et al., 2002). In the US, the pathogen was most likely first introduced to a nursery in California (Mascheretti et al., 2008), but spread to adjacent forests where it caused extensive mortality of coast live oak (Quercus agrifolia Nee.) and tanoak before it was noticed (Rizzo et al., 2002; Mascheretti et al., 2008). It is now found in natural forest ecosystems extending from California through Oregon (Rizzo & Garbelotto, 2003) and has been most recently isolated from waterways in Washington State as well as in some south eastern states (Chastagner et al., 2009). Phytophthora ramorum has been raising concerns due to the impact on the nursery trade and on valued oak species in the United States, and more recently on commercially important Japanese larch in Europe. In 2005, an estimated 20% of the oak and tanoak trees in the Big Sur Ecosystem of Coastal California had been killed by P. ramorum (Meentemeyer et al., 2008).   In Europe, P. ramorum was first found on woody ornamentals in Germany and the Netherlands (Werres et al., 2001). It is now found throughout many European countries (Grünwald et al., 2012) in nurseries and gardens largely affecting Rhododendron species with limited infection on trees (Brasier & Webber, 2010); however the pathogen has recently jumped to Japanese larch plantations in the UK where it is causing large-scale mortality. By early summer 2010, 1,900 hectares of Japanese larch had been affected by P. ramorum (Brasier & Webber, 2010).  In Canada, the pathogen has been found in several nurseries in British Columbia (Bilodeau et al., 2007). Plants originating from infected nursery beds were shipped between British Columbia and 25  West Coast US states several times, indicating the possibility for movement of the pathogen between the two countries (Goss et al., 2009). Phytophthora ramorum has not been reported in natural forests in Canada, however it has been found in some residential locations where eradication is underway (Kliejunas, 2011).   Biology and Life Cycle of P. ramorum The life cycle of P. ramorum follows the general life cycle of Phytophthora species (Grünwald et al., 2008) except that P. ramorum affects the above ground parts of the host species instead of the roots (Rizzo & Garbelotto, 2003; Hansen et al., 2008). Sporangia are produced on infected leaf tissue and released during periods of rainfall or periods of extended leaf wetness in the rainy seasons (Davidson et al., 2005; Hansen et al., 2008). Sporangia are most commonly dispersed by rain droplets (Davidson et al., 2005; Hansen et al., 2008) and wind driven rain usually within zero to ten meters and in some cases up to 15 m (Davidson et al., 2005); however dispersal (longer range dispersal) and infection patterns (top down infection in canopies) suggest that air dispersal is also an important means of spread (Hansen et al., 2008). Spores can also be dispersed through waterways; however this is thought to be a secondary source of infection suggested by the distribution and proximity of new infections (Hansen et al., 2008; Grünwald et al., 2012). Zoospores are produced and released from the sporangia (Davidson et al., 2005) and it is believed that once in contact with a suitable host the zoospores encyst and penetrate the host (Grünwald et al., 2008). Direct germination has not been documented from the sporangia; however this could occur similar to other Phytophthora species (Grünwald et al., 2008). Phytophthora ramorum also produces chlamydospores in infected plant tissue (Werres et al., 26  2001; Davidson et al., 2005). They can remain dormant in plant tissue or plant debris in the soil and cause infection when in contact with a host under suitable conditions (Shishkoff, 2007). Phytophthora ramorum is heterothallic, requiring both mating types for sexual reproduction to occur (Werres et al., 2001), however in the current known range sexual reproduction has not been observed (Goss et al., 2009; Prospero et al., 2009). Oospores are produced under laboratory conditions when the two mating types are grown together (Werres et al., 2001; Brasier & Kirk, 2004) suggesting that this part of the general Phytophthora life cycle should hold true for P. ramorum.  Phytophthora ramorum causes two distinct disease types: a bole canker type or a foliar blight (Rizzo et al., 2002; Grünwald et al., 2008). The canker type occurs on some oak species, tanoaks European beech (Fagus sylvatica L.) and on Japanese larch. The cankers spread under the bark, killing the phloem and girdling the tree (Grünwald et al., 2008) and the infection may also spread to the xylem (Brown & Brasier, 2007). The foliar blight occurs on a much broader range of hosts, over 100 different species (Grünwald et al., 2012), ranging from Rhododendron and Viburnum to conifer species including Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) (Grünwald et al., 2008). The foliar blight is characterized by leaf spots, or shoot tip dieback, and wilting (Grünwald et al., 2008). Phytophthora ramorum sporulation differs on the various hosts, ranging from no sporulation occurring in the cankers on Quercus species to prolific sporulation on the foliar tissue in some species such as California bay laurel (Umbellularia californica (Hook. & Arn.) Nutt.) (Davidson et al., 2005).   27  Population Characteristics and Migration Three distinct lineages of P. ramorum have been described in the known range of the pathogen, and a fourth newly described lineage (EU2) that has been found in Europe (Van Poucke et al., 2012). NA1 is the dominant lineage in the United States (Ivors et al., 2006; Mascheretti et al., 2008), and has been found in BC nurseries as well (Goss et al., 2011b). NA2 is the dominant lineage in British Columbia nurseries but has only been found in Washington and California nurseries in the United States, but not in forest stands (Goss et al., 2011b). The EU1 lineage was first found in Europe but has since spread to North America where it is found in nurseries in Canada and the US (Goss et al., 2011b). To date, all of the NA1 and NA2 isolates that have been found are the A2 mating type and EU1 has mostly been of the A1 mating type; however three isolates found in Belgium belonging to the EU1 lineage possessed the A2 mating type (Vercauteren et al., 2011). Analysis on the evolutionary history of NA1, NA2 and EU1 using nuclear sequence data suggests that the three lineages diverged approximately 165 000 to 500 000 years ago and remained isolated since their divergence (Goss et al., 2009); however the nuclear genes studied and the genome show high levels of heterozygosity suggesting that the founder population was sexually reproducing (Tyler et al., 2006; Goss et al., 2009). It is likely that the three lineages were initially introduced independently in North America and Europe (Goss et al., 2009) and that the current populations were generated by the introduction of only one or a few different isolates (Goss et al., 2009; Vercauteren et al., 2010). Some differences were found in laboratory tests in the aggressiveness of the three lineages on Rhododendron leaves between, with NA2 and EU1 showing higher levels of aggressiveness than NA1 (Elliott et al., 2011).  28  Overall genetic diversity was low in all three lineages of P. ramorum, reminiscent of clonally expanding populations (Prospero et al., 2009; Vercauteren et al., 2010; Goss et al., 2011b). In the NA1 and EU1 lineages there were a high number of genotypes; however they were based on only a few microsatellite loci and genotypes were closely related to one parent genotype (Prospero et al., 2009; Vercauteren et al., 2010). Very little genotypic diversity has been reported in the NA2 lineage; however this is based on microsatellite markers developed for the NA1 and EU1 lineages which could underestimate diversity in NA2 (Goss et al., 2011b). Differences between the different genotypes in all lineages can be explained by microsatellite mutation suggesting microsatellite mutation as the dominant process generating new genotypes rather than sexual recombination (Prospero et al., 2007, 2009; Vercauteren et al., 2010). The absence of sexual recombination could be from a lack of opportunity due to the presence of only one or the other mating type (Prospero et al., 2009), a partially dysfunctional reproductive system suggested by the difficulty in obtaining successful pairings in vitro (Brasier & Kirk, 2004; Garbelotto et al., 2006), or incompatibility between the three lineages due to an extended period of divergence (Garbelotto et al., 2006; Goss et al., 2009). In Belgium where the two mating types were found in the EU1 lineage, no signs of recombination were detected in the overall population (Vercauteren et al., 2010).   Population genetic studies suggest that the nursery trade is the predominant method of dispersal shaping the population structure of P. ramorum. Prospero et al. (2009) looked at genetic structure and migration patterns of P. ramorum in California, Oregon and Washington in both forests and nurseries using microsatellite markers. They found low overall genetic differentiation. Genetic distances were small between nursery populations from Oregon, 29  Washington and California, and there were very large geographical distances between populations with shared genotypes, much larger than the distances found for natural dispersal. In addition, there was no correlation between geographic distance and the number of shared genotypes with some of the populations sharing the highest numbers of genotypes separated by the largest geographic distances. Their data was consistent with trace back and trace forward surveys showing that infected plants were moved back and forth between the nurseries where the P. ramorum infections were found (Prospero et al., 2009). Other studies looking at the inter-continental migration patterns of P. ramorum have also highlighted the importance of the nursery trade in pathogen spread. There is continental migration from Europe to North America (Goss et al., 2011b). Shared genotypes were found in both continents; however they were found in Europe in years prior to being found in North America. It is unclear if migration from Europe occurred simultaneously into Canada and the US, or via secondary spread from Canada to the US or from the US to Canada. Movement of NA1 could be in a northward direction from the US to British Columbia suggested by a staggered temporal occurrence of one particular genotype, the lower occurrence of that lineage in BC, and trace forward and back surveys also showing infected material moving north (Goss et al., 2011b). The higher occurrence of the NA2 lineage in BC and the more infrequent occurrence in the US could indicate a southward migration from BC to the US for that lineage.  Studies in US forests and nurseries suggest that migration of P. ramorum genotypes occur from nurseries to forests (Prospero et al., 2007; Mascheretti et al., 2008). Higher genetic diversity was found in California nurseries than in forests using AFLP markers (Ivors et al., 2004) and microsatellite markers (Ivors et al., 2006) and forest populations appear to have been started 30  from the introduction of a single isolate (Ivors et al., 2006). In California the lack of genetic structure between some forest sites and nurseries reflects human mediated spread from nurseries to forests rather than natural spread (Mascheretti et al., 2008). In Oregon, population studies suggest the epidemic was started in Oregon nurseries and forests by introductions from outside nurseries and genotype flow does not occur from forests to nurseries (Prospero et al., 2007).   Center of Origin The origin of P. ramorum is currently unknown. Evidence suggests that in Europe and in North America the pathogen is introduced. The range of P. ramorum is limited in relation to its hosts (Rizzo & Garbelotto, 2003) and the dominance of single genotypes in California forests and in Europe as well as a clonal population structure suggests that the pathogen has been introduced to these locations (Rizzo & Garbelotto, 2003; Ivors et al., 2006; Vercauteren et al., 2010). In addition, it would be expected that both mating types would be found in high frequencies in a pathogen’s native range, which is not the case for P. ramorum, even though data suggests that these lineages were from a sexually reproducing parent population.  Disease management  The impacts of wide scale forest disease epidemics can be quite severe both economically and ecologically. Economic impacts can include costs from activities such as monitoring pathogen spread, enacting quarantine regulations and eradicating disease outbreaks. Income is lost due to the loss of marketable species or restrictions imposed on traded commodities. Phytophthora ramorum has had wide ranging impacts at many levels (Grünwald et al., 2012). Economic losses 31  have had a substantial effect on the nursery industry where multiple important species are hosts for P. ramorum. Losses result from the destruction of large numbers of infected plants, to the loss of trade opportunities due to quarantine regulations. In some nurseries, these losses have resulted in bankruptcy (Grünwald et al., 2012). Control efforts can cost governments millions of dollars. For example, the cost of eradication efforts for P. ramorum in Curry County, Oregon state between 2001 and 2008 was approximately 4.3 million US dollars (Kanaskie et al., 2010).  Early detection and eradication Avoiding introduction and early eradication of pathogens such as P. ramorum are two important management options but are dependent on early detection. Avoiding introduction of harmful pathogens can be far less costly than controlling an established pathogen (Mack et al., 2000). This has been attempted for various pathogens through quarantines, trade restrictions, and phytosanitary treatment requirements. If outbreaks are small and isolated then eradication is an option. However once a pathogen becomes established and widespread, eradication becomes difficult and costly (Mack et al., 2000).  Effective quarantine measures rely on accurate identification of pathogens, which can be problematic for pathogens that are difficult to identify or isolate. Genomic data can be used to create powerful tools to aid in identifying, screening for, and understanding the biology and epidemiology of important forest pathogens. Genomic data has the advantage of using data from the whole genome of an organism rather than small fragments of DNA and can therefore be used to create more efficient, specific and sensitive detection tools. Part of developing effective tools 32  will be to understand the target species its close relatives in order to avoid detection methods that give false positives. For example molecular tools were developed for the detection of P. lateralis (Winton & Hansen, 2001) prior to the discovery of P. ramorum. These tools showed a positive detection signal for both species and were not specific to P. lateralis (Sutton et al., 2008). A more complete understanding of Phytopthora genomics and the species in the genus will help to avoid these situations in the future. Understanding natural communities and the species present is also important for detection and monitoring for invasive pathogens. It would be ineffective to quarantine against pathogens that are natural in an ecosystem.  Controlling spread into natural environments It is important to determine if anthropogenic activities allow the movement of non-indigenous pathogens, virulent strains or hybrid strains into the natural environment. And if they do, the particular activities that pose the greater risks should be identified. For example if pathogens are transferred into natural systems through irrigation water, better practices or control measures can be developed to treat irrigation water before it runs into natural systems. Similarly gaining a better understanding of conditions or activities conducive to pathogen evolution may enable the development of practices to reduce evolutionary opportunities. This may include simple tactics such as appropriate methods of dealing with diseased plant waste. In other cases, the knowledge may assist in targeting eradication programs to deal with the highest risk localities. If hybridization is a likely scenario between Phytophthora pathogens then it may be an option to direct eradication efforts to populations likely to hybridize, for instance, different lineages of P. ramorum or P. ramorum and P. lateralis. 33   Management of established invasive pathogens If quarantine and eradication measures fail and invasive Phytophthoras become established in the environment then management efforts may need to be focused on controlling spread rather trying to halt spread. Spread of a disease can be slowed by limiting vehicle and recreational traffic in an area and by practicing sanitary procedures when leaving areas infected with a pathogen. These procedures have been used in Australia to prevent the spread of P. cinnamomi (Hardy et al., 2001), in Europe to prevent the spread of P. ramorum (Forestry Commision Great Britian, 2012) and in the US to prevent the spread of P. lateralis (Hansen, 2008). Reducing inoculum sources and spread has also been attempted by removing infected host plants for example Port Orford cedar in the range of P. lateralis in Oregon (Hansen, 2008), and understory Rhododendron as well as whole Japanese larch stands to reduce P. ramorum spread in Europe (Forestry Commision Great Britian, 2012). In areas where the pathogen is present or has been eradicated resistant species can be planted in order to avoid new outbreaks and host species should be planted in areas less prone to disease or where spread is less likely (Hansen, 2008). Fungicides can be applied in nursery or agricultural settings but is not always economically feasible or allowed in many countries to control disease in natural forests. Fungicides have been used to control P. cinnamomi in Australia (Hardy et al., 2001), and in California some oak trees on private lands are individually injected with a fungicide to control P. ramorum (Parke & Lucas, 2008); however this is not a viable option for controlling landscape level disease epidemics.  34  Objectives of study The overall objective of this current study was to improve our understanding of Phytophthora and of oomycete communities in general, and to assess the influence of anthropogenic activities on the evolution of these communities and the emergence of pathogenic species. In chapter two DNA meta-barcoding was used to compare the oomycete communities under various anthropogenic influences, from urban to natural environments. The hypothesis was that oomycete communities in urban areas would have a higher diversity and more introduced species than communities from natural environments due to influences from horticulture, gardening and agriculture.   In chapter three traditional culture based methods were used to further sample and characterize the Phytophthora community under various anthropogenic influences. Phytophthora communities from residential and agricultural areas were also compared. The hypothesis was that Phytophthora communities in urban areas are more diverse, have more introduced species, and have more hybridization events among species. It was also hypothesized that communities under different anthropogenic influence would show a different species profile.  In chapter four we focused on the evolution of an invasive alien species with a known record of anthropogenic dispersal, P. ramorum, by studying the variation at the genome level within and between the clonal lineages, following their initial introduction and subsequent invasion. Specifically, we wanted to test the hypothesis that P. ramorum populations are highly adaptable 35  due to a rapidly evolving genome, and that populations would show a genomic signature of rapid evolution.    36  Chapter 2: Anthropogenic activities increase diversity of oomycete communities in urban and urban/natural interface environments and are pathway for potential new pathogens to natural ecosystems  Introduction Anthropogenic activities are an important driver of the distribution and diversity of plant pathogens. One of the largest impacts is from international trade which has facilitated invasion of alien species into new regions, causing both widespread ecosystem damage and huge economic losses. There is a direct correlation between a country's gross domestic product (GDP) and the number of invasive plant pathogens introduced (Desprez-Loustau et al., 2010). Although some of this relationship is likely due to more research and more reporting of plant pathogens in countries with more resources (Scott et al., 2013; Hulbert et al., 2017), these same countries are also likely involved in more trade. The global plant trade is one of the major pathways for movement of exotic pathogens into new territories (Brasier, 2008). Climate change could compound these problems by creating favorable conditions or by increasing the range and distribution of invasive species. Both global trade and climate change have been implicated in forest health changes (Pautasso et al., 2010, 2015; Ramsfield et al., 2016). Horticultural and agricultural activities could add to these problems by facilitating the emergence of pathogens and providing a bridge to natural ecosystems. Urban trees may provide an opportunity for establishment and subsequent dispersal for introduced pests and pathogens (Paap et al., 2017). Agricultural ecosystems are less prone to environmental fluctuations, are more homogenous, and 37  undergo processes such as soil mixing, irrigation and fertilization which are not found in natural ecosystems and may all influence pathogen populations (Stukenbrock & McDonald, 2008).  The oomycetes are a group of eukaryotic microorganisms that resemble fungi and share a similar mycelial growth habit and infection strategies (Latijnhouwers et al., 2003) but belong to a different major subgroup, the Heterokonts or Stramenopiles (Baldauf, 2003). They can be found in plant foliage and forest canopies, in soil and in aquatic habitats. Sporangia can be dispersed through wind or water, and can either directly infect host tissue, or they can produce motile zoospores which are released in the presence of water and swim to host material (Tyler, 2002). Two orders within the oomycetes, Peronosporales and Pythiales, contain genera comprising some of the world’s most important plant pathogens: Phytophthora de Bary and Pythium Pringsh. Species in the genus Phytophthora are generally considered to be pathogens of plants and several species infect economically valuable tree species, both angiosperm and gymnosperm, and important agricultural species. Species in the genus Pythium include pathogens and saprotrophs that are widely distributed, diverse and are found in soil and water (Martin & Loper, 1999). Many Pythium species are also important crop pathogens and can cause extensive losses in agriculture.   Oomycetes are particularly prone to dissemination via the plant-for-planting trade since they can be propagated in the soil as resting spores or within plant tissues in an endophytic stage. DNA meta-barcoding of Phytophthora species in soil of nursery plants highlighted the role of nurseries and plant trade in the spread of Phytophthora: 25 Phytophthora phylotypes were found including two pathogens of conifers with a history of propagation outside nurseries: Phytophthora 38  ramorum, the causal agent of sudden oak death and Phytophthora lateralis the causal agent of Port Orford cedar root rot (Prigigallo et al., 2016). Molecular detection of Phytophthora using qPCR on the soil and foliage of nursery plants also revealed a high frequency of Phytophthora: the soil of 87% of the plants tested positive and 70% of the asymptomatic plants tested positive (Migliorini et al., 2015). In addition, a large proportion of first reports of Phytophthora species are associated with nurseries and urban areas also highlighting the importance of these activities in introducing and spreading these pathogens (Hulbert et al., 2017).  Additionally, oomycetes may be particularly affected by agricultural activities. Monoculture crops, the introduction of naive host species, and irrigation may facilitate spread or increase abundance in crop pathogens which could result in community changes and more frequent epidemics. Landscape structure can affect invasion success in disturbed habitats because of increased resource or habitat availability (With, 2002). Large, spatially connected habitats may increase the potential for successful spread (With, 2002). Agricultural crops provide an ideal habitat for invasive species as there is an abundance of resources and opportunity for spread via extensive corridors of connected host. In addition, oomycetes benefit from agricultural systems because irrigation facilitates the spread of zoospores which require water for dispersal and germination. Phytophthora and Pythium species are frequently associated with irrigation water, and several species of both genera have been implicated in disease outbreaks associated with infected water (Hong & Moorman, 2005).  Little is known about the diversity, abundance or ecological roles in native habitats of Phytophthora species or of the other genera within the oomycetes. Recent studies suggest that a 39  diverse community of Phytophthora species are present in natural ecosystems (Hansen et al., 2012a; Burgess et al., 2017b). One challenge with studies of Phytophthora diversity in natural ecosystems is that we are lacking baseline data on species presence and abundance. Surveys that attempt to describe species diversity often discover previously unknown or undescribed species. Many of the species discovered do not cause noticeable disease or impact to the ecosystem and appear to occupy a saprophytic niche in their natural habitats (Hansen et al., 2012a). Without knowledge or data on the native communities it is difficult to determine if these species are introduced, if they can cause potential damage, or if they are endemic and native but have not been described yet.   Few studies have been conducted using unbiased methods to compare diversity and distribution of oomycetes in environments with varying levels of anthropogenic influence. The objective of this study was to conduct a DNA meta-barcoding analysis to test the hypothesis that anthropogenically-impacted environments (i.e. urban centres, residential or agricultural lands) comprise a more diverse oomycete community and more non-native species than natural environments (i.e. forests) due to activities such as agriculture, exotic plant movement, and gardening that may facilitate the introduction of new species and/or new hosts to these environments. We also hypothesized that interface environments (i.e. natural forests or parks within or bordering urban centers) would be impacted by these activities, but to a lesser extent. In order to test these hypotheses, we designed our sampling to target three environment types: 1) urban areas in close proximity to nurseries, residential, agricultural areas and a park with extensive horticultural activities, 2) natural forests with little to no urban influence, and 3) urban/natural interface areas defined as natural environments/parks in urban settings. We used 40  natural environments with minimal urban influence to establish a baseline of the expected natural oomycete communities in southwest B.C. Urban areas were chosen around areas with activities that were expected to influence the oomycete community. Since many oomycete species are known or suspected to have been introduced through horticultural material and the nursery trade, and many others are associated with agriculture, it was hypothesized that signatures of anthropogenic influence would most likely be detected in areas geographically close to these activities. Interface areas provided a means to determine if migration was occurring. Since they were not directly associated with activities suspected to influence the community, migration would be needed to produce similar signatures of anthropogenic influence as those found in urban areas. We characterized and compared oomycete soil communities between the three site types using DNA meta-barcoding of the internal transcribed spacer one (ITS1). We chose to use soil since it likely reflects the most complete oomycete community encompassing both soil and foliar or canopy species due to fallen plant detritus, and dormant spores that remain protected over long periods of time.   Methods Field sampling Sampling was done in the south west coastal area of British Columbia, Canada in the vicinity of Greater Metropolitan Vancouver and southeast Vancouver Island near Victoria (Figure 2.1, Table 2.1). Four sites were sampled in each of the three environment types (urban, interface and natural): two in Vancouver and two in Victoria (Figure 2.2). These were further divided into the two dominant biogeoclimatic subzones in the area according to the biogeoclimatic ecosystem 41  classification system (Pojar et al., 1987): the coastal western hemlock very dry maritime subzone (CWHxm), and the coastal Douglas-fir moist maritime subzone (CDFmm). In addition, a fifth urban site in a highly agricultural area was sampled in the coastal western hemlock dry maritime subzone (CWHdm) (Figure 2.1). Each of the 13 total sites was given a two letter site code; full site names, two letter codes and details are found Table 2.1.   Sampling was conducted once in the fall of 2012 and a second time in spring of 2013 to obtain a more complete community profile and cover possible differences between seasons. At each site (or geographic region), ten soil samples (approximate volume 250-500 mls) were collected. Loose debris such as moss, leaves and twigs were removed and the top layer of soil was sampled with a trowel up to an approximate depth of 20 cm and stored in labeled Ziploc bags and brought to the lab for processing. The trowel was cleaned with water and a clean paper towel between each sample. Soil samples were frozen at -20 ºC until DNA extraction and DNA meta-barcoding for community analysis (below).  DNA meta-barcoding of Phytophthora communities Soil samples were mixed and one gram of soil was weighed from each sample and placed into one or two extraction tubes, depending on volume. DNA was extracted from soil samples using the FastDNA Spin Kit for Soil (MP Biomedicals, USA) following the manufacturer’s protocols. For samples that were split into two tubes, DNA from the two extractions were combined into a single sample. Samples were quantified using an Invitrogen Qubit 2.0 Fluorometer (ThermoFisher Scientific, Waltham, MA, USA) and diluted to 5 ng/ul. The ITS1 region between 42  the small subunit and the 5.8 subunit of the nuclear ribosomal RNA gene (rDNA) repeats was amplified using the primers ITS6 and ITS7 (Cooke et al., 2000). The ITS6 primer was amended on the 5’ end with Titanium adapter A (Roche Switzerland) and a unique multiplex identifier tag (MID (a unique 10 base fragment; Roche Switzerland)) for each sample and the ITS7 primer was amended with the Titanium adapter B (Roche Switzerland) on the 5’ end. Each sample was amplified in three duplicate 25 µl PCR reactions containing 0.5 units of Invitrogen Platinum Taq DNA polymerase (ThermoFisher Scientific, Waltham, MA, USA), 1X PCR buffer, 1.5 mM MgCl2, 0.2 mM Invitrogen dNTPs (ThermoFisher Scientific, Waltham, MA, USA), 1 µM of each primer, and 0.3 µg/µl bovine serum albumin (New England Biolabs, Ipswich, MA, USA) and 10 ng of DNA. Reactions were performed with an initial heating step at 94 °C for three minutes, followed by 35 cycles of a denaturation step at 94 °C for 30 seconds, an annealing step at 50 °C for 30 seconds, an elongation step at 72 °C for one minute. The reaction was finished with a 10 minute extension step. Amplicons were visualized on a 1% agarose gel stained with Ethidium Bromide. Samples were repeated if they failed and for samples that failed repeatedly, DNA concentrations of 7.5 and 5 ng/µl were tried.  Amplicons from the three duplicate reactions were combined and purified in two steps using the Agencourt Ampure XP magnetic PCR clean-up system (Beckman Coulter, Brea, CA, USA). Primer-dimers were eliminated using a 1.8:1 volume ratio of magnetic beads to amplicons and fragments smaller than 150-200 bp were eliminated with a 0.6:1 volume ratio. Samples were eluded in 26 µL EB buffer (Qiagen) and concentrations were measured using the Quant-iT Picogreen dsDNA assay kit (Invitrogen, Eugene, OR, USA). Samples were divided into five groups for the fall samples, with each group containing from 25 to 30 samples with a unique 43  multiplex identifier tag. The spring samples were divided into three groups each containing from 47 to 52 samples with a unique multiplex identifier tag. Amplicon libraries composed of samples in equimolar proportions were constructed and amplicon lengths were verified using an Agilent 2100 BioAnalyzer (Agilent Technologies, Santa Clara, CA, USA). Unidirectional sequencing was done on a Roche GS FLX+ system (Roche-454 Life Sciences, Branford, CT, USA) at McGill University and Genome Quebec Innovation Centre (Montreal, QC, Canada). Each library from the fall was sequenced on a quarter plate using the Titanium chemistry and each library from the spring was sequenced on half a plate using the FLX+ chemistry.   Data filtering and OTU assignment Sequence data was analyzed using the QIIME software package (Caporaso et al., 2010). Barcode sequences were used to assign reads to samples, and barcode and linker primer sequences were trimmed. Sequences were filtered for a minimum quality score of 25 using a 50bp sliding window, maximum homopolymer length of eight, maximum primer mismatch of two bases, and a maximum number of barcode errors of one. Data from the spring was processed and filtered as above; however due to erroneous additional bases found between the barcode and linker primer sequences, many sequences were filtered out. To fix this problem, prior to barcode assignment to samples, a custom Python script (Python Software Foundation. Python Language Reference, version 2.7. was developed to locate barcode sequences and linker primer sequences and eliminate any additional bases between the two. After filtering, the ITS1 sequence was extracted from all sequences (fall and spring) using the ITSx software which also screens for chimeric sequences (Bengtsson-Palme et al., 2013). The resulting sequences were 44  filtered for a minimum length of 160bp. To choose a minimum sequence length, sequences of Phytophthora were downloaded from the Phytophthora database (Park et al., 2013) and the ITS1 sequence was extracted using the ITSx software (Bengtsson-Palme et al., 2013). The minimum lengths of the known Phytophthora sequences were used to choose a minimum sequence length threshold.  Open reference operational taxonomic unit (OTU) picking was done using the QIIME software (Caporaso et al., 2010) with the UCLUST algorithm (Edgar, 2010) for sequence clustering at 97% and assignment of a representative sequence for each cluster. For taxonomic assignment, a database was created by downloading all Stramenopiles ITS sequences from NCBI. Additional ITS sequences of Phytophthora were downloaded from the Phytophthora database (Park et al., 2013) and added to the NCBI sequences. The ITS1 sequence was extracted using the ITSx software (Bengtsson-Palme et al., 2013). The resulting database was formatted to the specifications required in the QIIME software (Caporaso et al., 2010). A corresponding taxonomy file was created which contained the hierarchical taxonomic information for each of the sequences in the database file. For use with the Ribosomal Database Project (RDP) classifier 2.2 (Wang et al., 2007), the taxonomy file was created with six levels: 1) superphylum (=stramenopiles), 2) class, 3) order, 4) family, 5) genus and 6) species. Taxonomy was assigned to the OTUs with the RDP classifier 2.2 (Wang et al., 2007) implemented in the QIIME package (Caporaso et al., 2010), using the custom database as a reference and a confidence level cut-off of 0.5. Sequences with less than two copies (singletons), and samples with less than 100 total sequence reads were filtered out for downstream analysis. Sequences that could not be classified as oomycetes were eliminated for further analyses.  45   To extract the complete set of Phytophthora species as well as possible new species, OTU picking was done as above but with a similarity of 99% for the clustering step. Taxonomy was assigned to the representative OTUs (output file from QIIME where a single sequence represents a cluster of sequences in the full dataset) in two ways: 1) using the QIIME pipeline as above and 2) by searching each representative sequence from the OTU picking step against the NCBI nucleotide database using the BLASTn algorithm (Altschul et al., 1990). Results from both methods, and the spring and fall datasets were combined and sequences identified as Phytophthora were used to create a final set of representative Phytophthora sequences. The final set of representative Phytophthora sequences were crossed with the full data set to extract all sequences clustered with the representative Phytophthora sequences using a custom Python script (Python Software Foundation. Python Language Reference, version 2.7. The full QIIME pipeline was run again on the full dataset from both spring and fall. Clustering thresholds of 98%, 98.5%, and 99% were tested to find a threshold that reduced the resulting number of clusters with the same taxonomic assignments while maximizing those with different assignments. A final value of 98.5% was used. Final taxonomy assignments were checked manually; BLAST results were checked for alignment coverage and percent similarity. OTUs assigned to a species complex with closely related ITS sequences or with ambiguous results were aligned with known representatives of the complex or closely related species downloaded from the Phytophthora database (Park et al., 2013) or NCBI and searched for fixed nucleotide differences. OTUs were assigned to the best match based on fixed differences if possible; otherwise they were identified as a member of a species complex. OTUs with no close species matches were assigned to one of the ten previously defined Phytophthora 46  clades (as defined in (Cooke et al., 2000) and Blair et al. (2008)) or just as unknown Phytophthoras. OTUs with less than two sequence reads in a sample were filtered out of that sample to retain a more conservative Phytophthora dataset.  For sequences with a best match to a Phytophthora species but low sequence similarity, sequence similarities were calculated between aligned ITS1 sequences of the unknown OTUs and the ITS1 of sequences from the Phytophthora database (Park et al., 2013) and sequences of groups closely related to Phytophthora including Halophytophthora, Salisapilia, Phytopythium and Pythium. A multiple alignment was done using ClustalW (Thompson et al., 1994) in the BioEdit Sequence Alignment Editor (Hall, 1999), and a similarity matrix was calculated using the Biological Structure Analysis (bio3d) package (Grant et al., 2006) in the R statistical environment (R Core Team, 2017). Sequences were included as Phytophthora if the sequence similarities were within sequence similarities found between members of the Phytophthora genus, and if the sequence similarities were closer to members of the Phytophthora genus than members of other closely related groups. Read coverage of the unknown sequences was determined, and only sequences with a minimum coverage of five reads were included. Unknown sequences with over 99% identity to another unknown sequence were grouped together for read counts. In addition, the lengths of the ITS1 for the unknown sequences were compared to the length of the ITS1 for species found in the Phytophthora database (Park et al., 2013). For the phylogenetic analysis the full ITS of the representative Phytophthora and other closely related species (above) were aligned with MAFFT (Katoh & Standley, 2013) and a maximum likelihood phylogeny was reconstructed using default settings in the pipeline (Dereeper et al., 2008) with PhyML 3.0 (Guindon et al., 2010). 47   Data analysis Sequence abundance for each genera and OTU was calculated using QIIME (Caporaso et al., 2010). The number of OTUs and the number of samples each OTU was found in was calculated and compared for each major genus of the oomycetes for both fall and spring, and for each site type (urban, interface and natural). The number of Phytophthora positive samples from the manually curated Phytophthora dataset were calculated and compared between site types using a chi-square test. The number of OTUs as well as the number of Phytophthora and Pythium OTUs (97% sequence similarity) were compared between site types (urban, interface and natural) with a chi-square test. For comparisons of OTUs between site types, OTUs were filtered for a minimum of two reads per sample.  A canonical correspondence analysis (CCA) was done using the vegan package in R (Oksanen et al., 2016) to determine if there was an impact from site type on the OTU profile.   Samples were rarefied to 300 sequences and 500 sequences and number of OTUs and Shannon’s diversity index were calculated for each site and site type using QIIME (Caporaso et al., 2010). The difference between site type was compared using an analysis of variance (ANOVA).   48  Results Pythium and Phytophthora represent the most dominant genera in Oomycete soil communities. There were 291064 and 246641 ITS1 sequences after trimming and filtering the fall and spring datasets, respectively. A total of 266597 (91.6%) and 208128 (84.4%) of those sequences were identified as oomycetes by sequence homology in fall and spring, respectively. The percentage of sequences filtered at each step was similar in the datasets for the two sampling times (Appendix A, table S2.7). The number of Oomycete sequence reads ranged from 0 to 7273 per sample (averaging 2169 in fall and 1737 in spring; Tables 2.2 and 2.3). Within the oomycetes, Pythium was the genus represented by the highest number of sequences (79.2% in fall and 87.5% in spring; Figure 2.3A). Apodachlya, Saprolegnia and Phytophthora were less abundant, accounting for less than 5% each of the total Oomycete sequences (Figure 2.3A). However, even though Pythium sequences dominated the read counts, there was a higher OTU to read ratio for Phytophthora than for Pythium, and Phytophthora was represented by the second highest number of OTUs (Figure 2.3B). Out of 136 Oomycete OTUs in the fall and 112 in the spring sampling, 53 and 39 were Pythium, respectively, whereas 22 and 19 were Phytophthora OTUs. Even though Phytophthora sequences only accounted for 1.4% of the total reads in fall and 0.6% in the spring, they accounted for 16% of the total OTUs in the fall and 17% in the spring. Other oomycetes such as Saprolegnia and Aphanomycetes accounted for 10% or less of the OTUs for both sampling times (Figure 2.3B).   The Oomycete community was dominated by two Pythium species. The OTUs represented by the most sequences were identified as Pythium attrantheridium (55.5% of the Oomycete 49  sequences in fall and 69.6% in spring), followed by Pythium monospermum (17.1% of the Oomycete sequences in the fall and 6.3% of the sequences in the spring). These were also two of the most widespread species, found in 87% of the total samples (Figure 2.4). In contrast, most OTUs were not widespread (Figure 2.4); only 3% of the OTUs were found in more than 50% of the total samples. The majority of OTUs were rare with 78% of the OTUs found in less than 5% of the samples.   Anthropogenic activities influence Oomycete species composition and diversity. The site PR was significantly different in preliminary analyses (data not shown) and was removed from all analyses comparing site type for this chapter (data was retained for comparisons in chapter 3) to minimize any effects from having one extra urban site, or from one urban site in a biogeoclimatic zone (CWHdm) where there were no matching interface or natural sites. Site type (urban, interface or natural) had a significant influence on species composition in the canonical correspondence analysis (F = 0.1906, p = 0.001) (Figure 2.5). Urban sites were characterized by the highest number of genera (Figure 2.6), OTUs, OTUs that were exclusive to that site type (Figure 2.7) and low frequency OTUs (Figure 2.8). By contrast, natural sites had the fewest genera, OTUs, low frequency OTUs, and exclusive OTUs. Interface sites were closer to urban sites for these values. The difference in total number of OTUs between groups was significant (chi-square = 10.109, p = 0.006). There were more shared OTUs between urban and interface sites than between either and natural sites (Figure 2.7). This pattern was also observed within the two most abundant genera, Pythium and Phytophthora. There were 16, 20, and 27 Phytophthora OTUs in the natural, interface and urban groups, and 20, 35, and 45 Pythium OTUs in the natural, interface and urban groups respectively (Figures 2.8A, 2.8B). The difference in the 50  number of Pythium OTUs found in each site type was significant (chi-square = 6.969, p = 0.031).  Urban sites had the highest taxonomic diversity in both fall and spring. When samples were rarefied to the same number of sequences per sample (300), Shannon’s diversity and the number of OTUs were significantly higher (p<0.05) in urban than interface and natural sites (Figure 2.9; Tables 2.2, 2.3 and 2.4). When interface was removed from the comparison, urban sites were always significantly higher than natural sites (Table 2.4). Diversity was slightly lower in the spring than in the fall. Similar results were found with a sampling depth (rarefaction) of 500 (Tables 2.2 and 2.3). Shannon’s diversity and   Phytophthora are more frequent in urban than natural sites. Forty percent of the samples were positive for Phytophthora after combining the manually curated fall and spring samples. A higher proportion of the urban samples were positive for Phytophthora (51%) compared to interface (41%) or natural samples (28%); however the difference was not significant. The most widespread Phytophthora OTUs were P. syringae, and P. taxon Kelmania found in 14 and 10 samples respectively. Phytophthora syringae was found only in interface and urban samples, whereas P. taxon Kelmania was found in all site types, however only once in natural samples. The most widespread OTUs in natural sites were P. pseudotsugae, P. chlamydospora and P. cryptogea, which were only found in four samples each. Almost all Phytophthora OTUs were rare. All except P. syringae, and P. taxon Kelmania were found in less than five percent of the samples; P. syringae and P. taxon Kelmania were found in 6% of the samples. Over half of the OTUs were found in 1% or less of the total samples.  51  New Phytophthora species are more frequent in urban areas. Three putative new Phytophthora species were found in urban or in urban/interface samples; none were found in natural sites. These were found in the manual curation of the Phytophthora dataset where there were eighteen OTUs whose representative sequences did not have a high homology match in the NCBI database (less than 95% identity; Table 2.5) but had a closest match to a Phytophthora species. The percent identity between the sequences for eight of these OTUs and Phytophthora taxa were within the range found within the genus (54.3% to 100% similarity in the ITS1) (Table 2.5). Based on sequence homology, five of these OTUs were clustered into two taxa, unknown sp 1 and unknown sp 2, and another single OTU made the third taxa, unknown sp 3 (Table 2.5). The remaining two were rejected due to low read coverage. The species with the closest homology to unknown sp 1 were two uncultured Phytophthora clones (99% and 100% identity). It shared 83% sequence similarity with several unrelated species including Phytophthora clandestine, Phytophthora cactorum, Phytophthora asparagi and Peronospora sparsa. The ITS 1 was not informative enough to determine if this sequence was a Phytophthora species: in the phylogenetic analysis it groups with Peronospora sequences, within two major clades of Phytophthora and is closest to clade five, (Figure 2.10). Unknown sp 1 was found in three sites (SV, MD and HM) in both the fall and the spring sampling. The species with the closest homology to Unknown sp 2 is Phytophthora boehmeriae, (only 71% sequence identity). In the phylogenetic analysis it groups with clade nine (Figure 2.10). It was found in three sites (PS, PR and SV) in the fall sampling. Unknown sp 3 is closest to Phytophthora gallica (79%), and groups with clade 10 in the phylogenetic analysis, (Figure 2.10). It was found in four sites in spring (CV, PS, MD and SV). Further work using the both ITS 1 and 2 would be required to confirm 52  that these species are Phytophthora, if they could represent new clades, or if they could represent new genera.  The remaining ten unknown OTUs were rejected as potential Phytophthora species because the percent identity between these sequences and known members of the Phytophthora genus was lower than what was found between members of the genus (Table 2.5). In addition, the length of the ITS1 in these OTUs was much longer than what was found in Phytophthora (Phytophthora average ITS1 length = 209, min = 114, max = 248).   Manual curation of the Phytophthora dataset resulted in a total of 34 OTUs versus 36 from the automated taxonomic identification. Sixty-four percent of the OTUs identified to the species level through the automated pipeline were identified to the same species with the manual identification (Appendix A, Table S2.8).  Oomycete communities fluctuate with season. There was a larger and more diverse Oomycete community in the fall than in the spring. Eighty-four percent of the total OTUs were recovered in the fall sampling whereas 69% were recovered in the spring sampling period (Table 2.6). Fifty-three percent of the OTUs were recovered in both sampling periods. There was a high turnover in OTUs between seasons suggested by 31% and 16% of species recovered only in the fall and spring, respectively. This high turnover was also observed in the Phytophthora community; out of a total of 37 OTUs, only 35% were recovered in both seasons, whereas 35% and 27% were recovered in fall and spring only, respectively. There was a reduction in total Pythium OTUs in 53  the spring; out of a total of 59 OTUs, 56% were found in both seasons, 34% were found in the fall and only 10% were exclusive to the spring sampling period (Table 2.6).  Discussion  The anthropocene (a newly defined epoch in which human impacts are as important as natural processes) is changing the way we view species range and distributions. Globalization is resulting in human-mediated dispersal that is causing a breakdown of the normal biogeographic barriers (Capinha et al., 2015). This may result in a biotic homogenization across widespread similar environments which are socially and economically linked (Capinha et al., 2015). The impact of anthropogenic activities on plant pathogen communities is still poorly understood. For some prominent species, global distributions and genetic structure follow patterns for human-mediated dispersal rather than natural dispersal (Prospero et al., 2009; Goss et al., 2011b; Dutech et al., 2012; Grünwald et al., 2012) (Brasier 2001); however anthropogenic impacts are less documented at a community level for microorganisms. We demonstrate that anthropogenic activities impact Oomycete communities by increasing overall species diversity, the number and general presence of species in the plant pathogenic genus Phytophthora and the overall abundance of individual taxa. Furthermore, urban/natural interface environments likely act as a bridge for invasion into natural environments. This could impact both the natural biota and natural ecosystem processes. Phytophthoras introduced from nursery trade or spread by human movement have caused landscape level outbreaks in California, Australia, New Zealand and Chile (Wills, 1993; Garkaklis et al., 2004; Durán et al., 2008; Grünwald et al., 2012; Hansen et al., 2012a). 54   One likely explanation for our results is the migration of species into anthropogenically impacted ecosystems. Anthropogenic activities can influence the diversity of microorganisms through several mechanisms. Global trade, total population and population density have all been associated with patterns of invasion (Desprez-Loustau et al., 2010; Pautasso et al., 2010, 2015; Santini et al., 2013). Environmental disturbance, landscaping, forestry, and increased movement through dense road networks are typical activities associated with increased urbanization which may contribute to increased incidences of invasive species (Lake & Leishman, 2004; Alston & Richardson, 2006; Desprez-Loustau et al., 2010). Increase in diversity and abundance of exotic plant species has been linked to disturbances, in particular those that are accompanied by an increase in nutrient rich water (Lake & Leishman, 2004). Patterns of invasion in plant pathogens likely mimic those of invasive plants. Increases in potential hosts, or stress to native plant hosts could directly influence pathogen populations in invaded environments. Invasive plant species may also cultivate different pathogen communities which was suggested as a cause for the different Oomycete communities found around invasive and native reed populations (Nelson & Karp, 2013). Furthermore, microorganisms and plant pathogens associated with invasive plants, ornamental plants, species used in reforestation or other forest management activities can be carried into new habitats. High frequencies and a high diversity of Phytophthora species have been found in the soil and roots of nursery plants (Prigigallo et al., 2015; Jung et al., 2016); furthermore, a high frequency of first reports of Phytophthora species are associated with nursery plants highlighting the risks associated with horticulture activities (Hulbert et al., 2017). Higher rates of Phytophthora infestation in planted forests versus natural forests or seed generated forests, and a high frequency of Phytophthoras associated with planting stock 55  originating from nurseries all suggest that this is a very important pathway in the movement of plant pathogens (Jung et al., 2016).  Global trade results in the accidental introduction of organisms that hitchhike on traded goods and many crop and forest pathogens are introduced through the plant trade (Brasier, 2008; Burgess et al. 2016) . Invasive forest pathogens are often first introduced into agricultural or urban areas and then spread to natural forests (Paap et al., 2017). Our study clearly shows a pattern indicative of movement from urban to natural environments. There were ten times more shared species between urban and interface than between interface and natural environments. A similar pattern of decreasing diversity and frequency in Phytophthora populations going from urban to natural environments was found in Sweden (Redondo et al., 2018). Urban/natural interface environments are subject to many influences from urban centres such as increased human access, higher levels of disturbance, soil enrichment and introductions of alien species through horticulture (Alston & Richardson, 2006). Similar to the pattern shown here with oomycetes, interface environments reflect a pattern of increased diversity in alien plant species which is more pronounced with increasing proximity to urban areas (Alston & Richardson, 2006). The proximity of urban/interface areas to large sources of propagules makes them a sink for invasive species. This increased reservoir of inoculum could act as a bridge for invasive species into natural environments. Phytophthora ramorum is an invasive pathogen that has been introduced through the nursery trade and subsequently migrated into adjacent natural forests where it causes sudden oak death in the United States, and sudden larch death in the UK (Ivors et al., 2004; Prospero et al., 2007; Mascheretti et al., 2008; Brasier & Webber, 2010; Grünwald et al., 2012). Similarly, Phytophthora cinnamomi, an important agricultural and horticultural 56  pathogen, escaped from nursery material and spread to natural systems causing Jarrah dieback and ecosystem wide damage in Australia (Weste & Marks, 1987; Wills, 1993; Garkaklis et al., 2004; Hardham, 2005; Cahill et al., 2008).  Much of our understanding of pathogenic species in the oomycetes is focused on Phytophthora because of their long and visible history as crop pathogens. However, other groups and species in the oomycetes can also cause diseases and become invasive. The Peronosporales appear to be common invaders suggested by their abundance in invasion records (Desprez-Loustau et al., 2010). Our study suggests that anthropogenic activities likely facilitate the spread of many species in this important group of microorganisms, not only Phytophthora species. The trends observed in this study were not only present in the Phytophthora genus, but appeared to be present across the entire oomycete community.   Pythium was the dominant group of oomycetes in all habitat types examined in this study. Many Pythium species cause disease or damping off in young seedlings and can inhibit seed germination, but are only weakly pathogenic to mature plants, or can reduce growth without killing mature plants (Schroeder et al., 2013). Negative feedback on plant growth due to pathogens is hypothesized to maintain species diversity in plant communities by limiting growth in certain species, and allowing the growth of other competitors (Mills & Bever, 1998; Laliberté et al., 2015; Albornoz et al., 2017). This feedback was demonstrated with Pythium species preferentially limiting growth of Danthonia and Panicum in a plant community without causing mass mortality or obvious disease symptoms (Mills & Bever, 1998). Pythium attrantheridium, described in 2004, has been associated with both agricultural and natural environments in North 57  America (Packer & Clay, 2000; Allain-Boulé et al., 2004; Broders et al., 2007; Reinhart et al., 2010). Pythium attrantheridium has been suggested to play a role in shaping the distribution of trees in the ecosystem by inhibiting the establishment of younger trees near parent trees where it is associated with root systems (Packer & Clay, 2000; Allain-Boulé et al., 2004). Packer & Clay, (2000) tested the Janzen-Connell hypothesis (Janzen, 1970; Connell, 1971) that predators, in this case Pythium soil pathogens, maintain tree diversity by influencing tree distribution. They showed that Pythium species caused high mortality in seedlings close to parent trees, but had no effect on seedling survival further away from trees (Packer & Clay, 2000). Pythium monospermum has a ubiquitous distribution but little is known about its pathogenicity (Van der Plaats-Niterink A J, 1981). It may be pathogenic or associated with disease on some agricultural species (Robertson, 1973; Van der Plaats-Niterink A J, 1981). The high abundance of Pythium observed in natural or forest ecosystems in this study suggests that they may contribute to stand structure and maintenance of biodiversity in these mature ecosystems. Other studies have suggested that oomycetes play an important role in driving succession in young ecosystems but that they decline with ecosystem age (Dickie et al., 2017). In our study, the higher frequency of oomycetes in interface and urban sites likely reflects the disturbance from anthropogenic activities which increase diversity and abundance. These shifting patterns in Oomycete communities could interfere with natural ecosystem processes that these pathogens are a part of which may result in increases in disease or ecosystem declines.  Similarly, some Phytophthora species are likely natural inhabitants in healthy ecosystems. Recent studies have revealed many Phytophthora species in natural environments and have led to questions regarding their presence, role and potential impact (Hansen et al., 2012a; Burgess et 58  al., 2017b). A similar role to Pythium in the maintenance of biodiversity has been suggested and shown for native Phytophthora species (Albornoz et al., 2017). Overall, our data show that Pythium was the most abundant genus in Oomycete communities in all site types including natural environments, and that all other genera including Phytophthora are much less abundant. This is similar to what has been found in other studies. Low Phytophthora diversity was found in soil samples in Spain using pyrosequencing on environmental samples (Català et al., 2015). Most studies of Oomycete communities also find a much lower abundance of Phytophthora compared to Pythium or other Oomycete genera (Arcate et al., 2006; Nelson & Karp, 2013; Sapkota & Nicolaisen, 2015; Riit et al., 2016; Esmaeili Taheri et al., 2017). However, the reverse was true in the Pythium and Phytophthora populations around declining sweet chestnut forests where Phytophthora was known to be present (Vannini et al., 2013). Although there was a relatively low abundance of Phytophthora sequences recovered from the soil samples in our study, we found a relatively high species diversity suggesting a small but diverse population of Phytophthora likely resides in natural healthy environments. It is possible that the natural environment has been invaded by exotic species; however it is unlikely that all of the species present are exotic. Regardless of their indigenous or exotic status, the reservoir of inoculum is important because under conducive environmental conditions, host presence, or an imbalance in the ecosystem, a single species can cause devastating damage. In declining oak stands in Spain, metabarcoding of Phytophthora populations revealed an overall low species diversity but high frequency and dominance of a few species possibly causing disease (Català et al., 2017). Devastating pathogens such as Phytophthora cinnamomi can survive in soil for long periods in infected plant tissues or as chlamydospores or oospores and under the right conditions, disease develops (Hardham, 2005).  59   Culture-independent methods such as the DNA meta-barcoding used here are revolutionizing the way we conduct microbial community studies. Community DNA meta-barcoding was a useful and successful approach for exploring the Oomycete community in soils. A high percent of the sequence reads obtained in this study were identified as Oomycete or Stramenopiles; very few reads were identified from other groups. Sapkota and Nicolaisen (2015) tested the DNA meta-barcoding approach on samples with known amounts of Pythium and Phytophthora DNA, and although the number of sequences obtained did not match the actual abundance of DNA from the two groups, both groups were always detected, suggesting that the method is reliable and that the low frequency of Phytophthora in soils is biological.   The internal transcribed spacer region is a useful target for describing the taxa present in the environment. It can be used to identify species or complexes of closely related species as it is fast evolving giving a higher level of interspecific variation than more conserved regions of DNA. In addition, the vast number of barcoding studies done using the ITS has increased the number of sequences available for comparison in reference databases. This allowed us to identify most OTUs to the species level. Although there was good agreement between the manual and automated taxonomic identifications in the Phytophthora dataset, the differences between the two datasets highlights the importance of drawing conclusions made from species level taxonomic assignments. There are several Phytophthora species complexes that comprise species with greater than 99% sequence homology in the ITS. Manual identification using fixed differences allowed for a higher confidence in identifications; however using only the ITS1 region could still leave room for errors in this group of organisms. Further analysis using species 60  specific primers or culture-dependent methods would be needed to confirm results found using DNA meta-barcoding. Many regulatory agencies would require the presence of a culture to confirm positive findings of a quarantine organism. DNA meta-barcoding would be a useful tool for monitoring large geographic areas. More intensive sampling could be targeted to locations to confirm the presence of high risk pathogens or possible disease outbreaks.  The largest disadvantage of using the ITS is that the high divergence in this region makes it unsuitable for phylogenetic based diversity analyses. Differences in diversity between sites could be underestimated since only using species richness or counts of OTUs does not take into account how related or not the species are between the sites. Using primers targeted to specific genera could also be used to get a more in depth look at any particular genus. In this study and in those using more general primers, the diversity in less dominant genera could be underestimated, in particular for rare species.   The trends found in Oomycete communities were the same in the fall and spring sampling; however sampling in two seasons gave a higher coverage of the species present. There was a high turnover of OTUs between the two sampling seasons. This is likely due to both biological reasons and sampling methods. Phytophthora populations are known to fluctuate over seasons (Balci et al., 2007; Reeser et al., 2011). Sampling in multiple seasons and conditions gives a better look at the populations. In addition, when using DNA meta-barcoding, a very small amount of soil is used to make inferences about a much larger population. It is very likely that multiple samples would show different results. A larger sampling area and number of samples increases the confidence in the results found. Our study showed the same trends across seasons 61  and taxonomic groups, higher diversity in urban sites, more shared species in urban and interface sites, and similar relative abundance in different genera, giving us a high confidence in the conclusions drawn.  Overall, we have found a strong signature of anthropogenic activities on Oomycete communities. Further studies on important genera, in different ecosystems, and in healthy versus diseased forests will further help to understand how anthropogenic activities contribute to disease emergence, how anthropogenic activities influence this important group of organisms, and the role of these organisms in natural environments. Our study has provided important information of the composition and distribution of Oomycete communities in southwest British Columbia and contributes to our understanding of this important group of organisms worldwide.   62  Table 2.1: Details of sampling areas. Site Location Region Biogeoclimatic zone Site type Urban Site Type* Collection areas Dominant vegetation in sampled areas BW Bowen Island Bowan Island CWHxm Natural  forested areas Douglas-fir, alder, cedar, hemlock CR Chemainus River/ Cowichan Valley Vancouver Island CWHxm Natural  forested areas Alder, hemlock, cedar, Douglas-fir, salal, Abies fir GT Gowland Tod Park, Victoria Vancouver Island CDFmm Natural  forested areas Cedar, alder, Douglas-fir, ferns SA Saturna Island Gulf Islands CDFmm Natural  forested areas Cedar, hemlock, alder, Douglas-fir, big leaf maple, salal CV Campbell Valley Park, Langley Metro Vancouver CDFmm Interface  forests/meadows/ replanted areas Alder, birch, maple, blackberry, cedar GL Grant lake/ Cowichan Lake, Vancouver Island Vancouver Island CWHxm Interface  forested areas bordering residential areas Alder, cedar, hemlock, Abies fir, Douglas-fir, salal HM Cuthbert Holmes Park and Swan Lake ,Victoria Vancouver Island CDFmm Interface  forests/meadows/beside ponds/beside trails Willows, black berry, cedar, Douglas-fir, ivy, wild rose, other shrubs, oak PS Pacific Spirit Park, Vancouver Metro Vancouver CWHxm Interface  forested areas adjacent to trails Cedar, hemlock, alder, Douglas-fir 63  Site Location Region Biogeoclimatic zone Site type Urban Site Type* Collection areas Dominant vegetation in sampled areas PR Pitt River area, Pitt Meadows Metro Vancouver CWHdm Urban Agricultural areas adjacent to dykes, rivers, walking trails, agricultural fields Blackberry, grass, blueberry farms, willows, cottonwood, alder, cedar SV Steveston area, Richmond Metro Vancouver CDFmm Urban Agricultural areas adjacent to dykes, agricultural fields, roads, ditches Alder, blackberry, poplar, grass BP Beacon Hill Park, Victoria Vancouver Island CDFmm Urban Park flower beds, grassy meadow areas Rhododendron, cedar, Douglas-fir, oak, cypress, many horticultural plants, shrubs HV Happy Valley, Victoria Vancouver Island CWHxm Urban Residential public areas in residential neighborhoods: trails, sides of roads, forest patches Big leaf maple, cedar, Douglas-fir, alder, hemlock MD Marine Drive Area, Burnaby Metro Vancouver CWHxm Urban Residential public areas in residential neighborhoods: forest patches, trails, sides of roads edge of golf course and recreational fields Alder, Big leaf maple, blackberry, cedar, shrubs *Urban sites included areas influenced by different anthropogenic activities. The dominant influence based on a visual assessment of the area is listed.64  Table 2.2: Descriptive statistics for fall samples including minimum and maximum number of sequences, Shannon’s diversity index and number of OTUs for sequences classified as Oomycetes.  Site Type N Min readsc Max readsc Depth 300* Depth 500* Sa Ha Nb Sa Ha Nb CR Natural 9 29 3149 9 1.7 8 11 1.8 5 GT 10 662 5127 11 1.6 10 13 1.6 10 BW 11 29 2928 11 1.7 9 14 1.8 9 SA 9 1096 5882 14 1.8 9 16 1.8 9 GL Interface   9 736 3502 9 1.1 9 11 1.2 9 CV 9 485 3897 13 1.4 9 14 1.2 8 PS 7 119 2609 21 2.5 6 29 2.9 5 HM 10 1264 7273 12 1.6 10 14 1.6 10 PR Urban  9 461 1994 25 3.1 9 30 3.1 8 HV 10 466 4445 12 1.8 10 15 2.0 9 SV 10 935 3486 21 2.3 10 25 2.4 10 BP 10 602 5027 14 1.6 10 18 1.7 10 MD 10 341 2597 18 2.2 10 20 2.2 9 Natural 39 29 5882 11 1.7 36 14 1.7 33 Interface 35 119 7273 13 1.6 34 16 1.6 32 Urban (no PR) 40 341 5027 16 2.0 40 20 2.0 38 a S - number of species measured as the number of OTUs at 97 clustering, H - Shannon’s diversity index, Nb - adjusted sample size with samples less than the sampling depth excluded, c - minimum and maximum number of sequences after removing unclassified sequences, *sampling depth - the number of sequences randomly sampled for diversity calculations.     65  Table 2.3: Descriptive statistics for spring samples including minimum and maximum number of sequences, Shannon’s diversity index and number of OTUs for sequences classified as Oomycetes.  Site Type N Min readsc Max readsc Depth 300* Depth 500* Sa Ha Nb Sa Ha Nb CR Natural 9 0 2274 6 0.5 2 7 0.5 2 GT 10 91 2547 10 1.1 9 13 1.1 8 BW 11 126 2816 11 1.5 8 14 1.4 7 SA 7 680 5435 8 0.7 7 9 0.8 7 GL Interface  6 0 2749 10 0.5 2 13 0.5 2 CV 9 193 1827 16 1.6 8 20 1.5 7 PS 8 217 1729 12 1.9 7 15 2.0 6 HM 10 666 2569 14 1.3 10 17 1.3 10 PR Urban  10 20 1998 14 1.9 9 15 1.8 8 HV 10 1075 3042 12 1.1 10 14 1.1 10 SV 10 801 5560 17 1.7 10 20 1.7 10 BP 10 1512 3109 18 2.2 10 22 2.2 10 MD 10 1940 5371 20 2.0 10 24 2.0 10 Natural 37 0 5435 9 1.1 26 12 1.1 24 Interface 33 0 2749 14 1.5 27 17 1.4 25 Urban (no PR) 40 801 5560 16 1.7 40 20 1.7 40 a S - number of species measured as the number of OTUs at 97 clustering, H - Shannon’s diversity index, Nb - adjusted sample size with samples less than the sampling depth excluded, c - minimum and maximum number of sequences after removing unclassified sequences, *sampling depth - the number of sequences randomly sampled for diversity calculations.         66  Table 2.4: ANOVA results for diversity comparisons on samples that were rarefied to 300 sequences per sample. Group Dependant variable F-statistic p-value Fall: natural vs interface vs urban  Number of OTUs 5.730 0.004 Fall: natural vs urban Number of OTUs 15.055 0.000 Fall: natural vs interface vs urban Shannon’s Diversity 3.352 0.039 Fall; natural vs urban Shannon’s Diversity 5.127 0.026 Spring: natural vs interface vs urban Number of OTUs 14.547 0.000 Spring: natural vs urban Number of OTUs 28.991 0.000 Spring: natural vs interface vs urban Shannon’s Diversity 5.084 0.008 Spring: natural vs urban Shannon’s Diversity 10.886 0.002 67  Table 2.5: OTUs sharing low sequence identity to Phytophthora taxa.  Blast search results, read coverage, and range of percent identity to members of the Phytopthora genus.  Representative  sequence New  ID* Species with closest BLAST match Length (bp) Query coverage (%) Identity (%) # of reads  % Identity    to other Phytophthora  OTU2029 a Unknown sp 1 Phytophthora clandestina 212 99 83 61 55.0-87.7 OTU1617 ac Unknown sp 1 Phytophthora clandestina 217 99 81 31 61.6-85.9 OTU4328 ac Unknown sp 1 Phytophthora clandestina 215 99 82 206 61.6-86.3 OTU2987 ab Unknown sp 2 Phytophthora boehmeriae 234 100 71 2 59.2-81.4 OTU578 ab Unknown sp 2 Phytophthora boehmeriae 234 100 71 45 59.2-80.3 OTU3647  Phytophthora colocasiae 215 100 73 3 62.5-80.2 OTU1517 a Unknown sp 3 Phytophthora gallica 214 100 79 40 56.5-78.8 OTU1914  Phytophthora intercalaris 212 100 76 2 59.1-77.1 OTU2190  Phytophthora sp. REB326-69 335 96 89 33 43.4-70.8 OTU4062  Phytophthora cactorum 251 41 93 7 50.3-67.9 OTU2785  Phytophthora infestans 324 95 89 20 45.2-67.0 OTU2513  Phytophthora asparagi 327 92 84 2 45.3-66.7 OTU2551  Phytophthora megasperma/ P. asparagi 323 98 87 2 45.6-66.4 OTU1670  Phytophthora asparagi 323 91 87 514 45.0-66.4 OTU3579  Phytophthora cactorum 214 99 81 4 51.1-85.8 OTU4280  Phytophthora cactorum/ Peronospora sparsa 208 99 80 2 50.6-84.1 OTU5636  Phytophthora boehmeriae 232 100 71 4 50.3-68.8 OTU6144  Phytophthora quercina 217 99 81 2 51.6-85.0 a Sequences included as potential new Phytophthora species; b,c OTUs combined as same OTU. 68   Table 2.6: Number of OTUs in each genus that were found in both fall and spring or that were exclusive to only one season. Genus Shared in both seasons OTUs in fall only OTUs in spring only Total 89 53 27 Achlya 4 2 1 Albugo 1 0 0 Aphanomyces 5 1 1 Apodachlya 3 0 0 Brevilegnia 1 1 2 Dictyuchus 1 0 0 Globisporangium 1 0 1 Hyaloperonospora 0 1 0 Lagena 1 0 0 Lagenidium 2 1 0 Leptolegnia 1 0 1 Newbya 0 1 0 Other 7 1 1 Ovatisporangium 1 0 0 Paralagenidium 1 0 0 Peronospora 1 1 0 Phytophthora 13 13 10 Phytopythium 3 0 1 Plectospira 0 0 1 Protoachlya 0 0 1 Pythiogeton 2 1 1 Pythiopsis 1 3 0 Pythium 33 20 6 Saprolegnia 6 7 0 Thraustotheca 1 0 0   69   Figure 2.1: Map of sampling sites.  Natural, interface and urban sites are represented by markers with green, blue and red colors respectively; biogeoclimatic zone are represented by circles (CWHxm), triangles (CDFmm) or square (CWHdm)    70   Figure 2.2: Sample design showing the hierarchical arrangement of sites within geographic area and site types  71                 Figure 2.3: Oomycete community composition in southwest British Columbia for fall and spring collections. (A) Relative abundance of sequence reads by genera. (B) Relative abundance of OTUs by genera.   0%10%20%30%40%50%60%70%80%90%100%Fall SpringOtherPythiopsisLeptolegniaDictyuchusBrevilegniaPythiogetonGlobisporangiumPhytopythiumPeronosporaAchlyaLagenidiumLagenaAphanomycesPhytophthoraSaprolegniaApodachlyaPythium0%10%20%30%40%50%60%70%80%90%100%Fall SpringNewbyaHyaloperonosporaProtoachlyaPlectospiraOvatisporangiumDictyuchusThraustothecaParalagenidiumLagenaAlbugoPeronosporaPythiopsisLeptolegniaGlobisporangiumLagenidiumPhytopythiumApodachlyaPythiogetonBrevilegniaAchlyaAphanomycesSaprolegniaOtherPhytophthoraPythiumA B 72                  Figure 2.4: Distribution of OTUs across samples.  (A) Percentage of samples each higher frequency OTU was found in (higher frequency OTUs were those observed in at least 5% of the total samples; only OTUs assigned to the species level are shown). (B) Percentage of the OTUs that were found across each percentage category of the total samples.   0% 20% 40% 60% 80% 100%Phytophthora quercinaPythium destruensPythium rostratifingensPythium parvumPythium undulatumPhytophthora syringaePythium diclinumLagenidium humanumPythium junctumPythium sylvaticumAphanomyces stellatusSaprolegnia turfosaLagena radicicolaPythium heterothallicumSaprolegnia megaspermaPythium rostratumPythium glomeratumPythium balticumApodachlya minimaPythium intermediumPythium monospermumPythium attrantheridiumPercent positive samplesA B 1% 1%3%17%78%> 75%50-75%25-50%5-25%< 5%B 73    Figure 2.5: Influence of site type on species profile illustrated in a canonical correspondence analysis (CCA). Species profile is based on the relative abundance of species in the combined fall and spring samples.          74                 Figure 2.6: Relative abundance of OTUs by genera in each site type for the fall and spring Oomycete collections.          0%10%20%30%40%50%60%70%80%90%100%Natural Interface UrbanFall Spring 0%10%20%30%40%50%60%70%80%90%100%Natural Interface UrbanNewbyaHyaloperonosporaProtoachlyaPlectospiraOvatisporangiumDictyuchusThraustothecaParalagenidiumLagenaAlbugoPeronosporaPythiopsisLeptolegniaGlobisporangiumLagenidiumPhytopythiumApodachlyaPythiogetonBrevilegniaAchlyaAphanomycesSaprolegniaOtherPhytophthoraPythium75   Figure 2.7: Number of shared and private oomycete OTUs between site types (fall and spring sampling combined).   76                           0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%0% 20% 40% 60% 80% 100%Pythium utonaiensePythium selbyiPythium perplexumPythium lucensPythium kandovanensePythium iwayamaiPythium deliensePythium catenulatumPythium camurandrumPythium angustatumPythium tracheiphilumPythium pyrilobumPythium paddicumPythium longandrumPythium spiculumPythium ultimumPythium pachycaulePythium cryptoirregularePythium spinosumPythium sulcatumPythium anandrumPythium acanthicumPythium torulosumPythium dissimilePythium mamillatumPythium vanterpooliiPythium apiculatumPythium regularePythium minusPythium aquatilePythium irregularePythium grandisporangiumPythium macrosporumPythium oopapillumPythium volutumPythium sp_rooibosp2Pythium destruensPythium rostratifingensPythium parvumPythium undulatumPythium diclinumPythium junctumPythium sylvaticumPythium heterothallicumPythium rostratumPythium glomeratumPythium spPythium sp_balticumPythium intermediumPythium spPythium monospermumPythium attrantheridiumA 77                           0% 5% 10% 15% 20%Phytophthora alniPhytophthora sp_ohioensisPhytophthora idaeiPhytophthora hibernalisPhytophthora amnicolaPhytophthora mississippiaePhytophthora spPhytophthora hydropathicaPhytophthora irrigataPhytophthora sp_raspberryPhytophthora sp_Taxodium_1Phytophthora gallicaPhytophthora siskiyouensisPhytophthora cambivoraPhytophthora multivoraPhytophthora europaeaPhytophthora sp_xHennopsPhytophthora pseudosyringaePhytophthora uliginosaPhytophthora sansomeanaPhytophthora bilorbangPhytophthora plurivoraPhytophthora ramorumPhytophthora lateralisPhytophthora inundataPhytophthora chlamydosporaPhytophthora fragariaePhytophthora lacustrisPhytophthora nemorosaPhytophthora polonicaPhytophthora cactorumPhytophthora gonapodyidesPhytophthora pseudotsugaePhytophthora quercinaPhytophthora syringaePhytophthora sp0% 5% 10% 15% 20% 0% 5% 10% 15% 20%B 78    0% 10% 20% 30% 40% 50%Thraustotheca clavataNewbya dichotomaHyaloperonospora…Dictyuchus spAlbugo hohenheimiaLeptolegnia spPythiogeton spPythiogeton spParalagenidium karlingiiGlobisporangium spOvatisporangium spPeronospora spPeronospora corydalisBrevilegnia spBrevilegnia gracilisBrevilegnia minutandraBrevilegnia unispermaPythiopsis spPythiopsis humphreyanaPythiopsis intermediaPythiopsis terrestrisAchlya racemosaAchlya spAchlya stellataAchlya ambisexualisAchlya oligacanthaAchlya ornataPhytopythium litoralePhytopythium montanumPhytopythium vexansPhytopythium citrinumLagena radicicolaLagenidium humanumLagenidium spLagenidium spAphanomyces stellatusAphanomyces spAphanomyces spAphanomyces cladogamusAphanomyces iridisAphanomyces frigidophilusSaprolegnia spSaprolegnia megaspermaSaprolegnia turfosaSaprolegnia spSaprolegnia anisosporaSaprolegnia terrestrisSaprolegnia feraxSaprolegnia asterophoraSaprolegnia diclinaSaprolegnia torulosaSaprolegnia monoicaApodachlya minimaApodachlya spApodachlya brachynema0% 10% 20% 30% 40% 50% 0% 10% 20% 30% 40% 50%C a 79  Figure 2.8: Number of samples each OTU is found in each site type.  Pythium OTUs (A), Phytophthora OTUs (B) or other OTUs (C). Samples from natural shown in green, interface in blue and urban in red. OTUs were based on a sequence clustering threshold of 97% and minimum count of 2 sequences in a sample. a Hyaloperonospora nasturtii-aquatici. 80   Figure 2.9: Diversity by site type. Samples were rarefied to a depth of 300 sequences. Samples with less than 300 sequences were eliminated. Number of OTUs is right and Shannon’s diversity index is on the left. Error bars represent standard deviation. ANOVA results for comparisons are shown in Table 2.4.  81   Figure 2.10: Phylogenetic analysis with unknown putative Phytophthora species.  Phylogenetic reconstruction was done using the ITS 1 of representative Phytophthora species and close relatives. Clade 6 Clade 8 Clade 5 Clade 1 Clade 3 Clade 7 Clade 4 Clade 2 Clade 9 Clade 10 82  Chapter 3: Anthropogenic activities introduce new species and increase the frequency of Phytophthora in urban habitats compared to native habitats  Introduction Emerging plant and forest diseases threaten natural ecosystems, food supply, and the economies that depend on them. The most important drivers of emergent plant diseases are anthropogenic activities, in particular introductions of alien pathogens, climate change and agricultural practices (Anderson et al., 2017). Alien invasive species cause major environmental damage and economic losses (Pimentel et al., 2000). Emergent diseases of crops and trees are often caused by alien pathogens that benefit from attacking susceptible host species and from the lack of competitors. In new habitats, host species have not co-evolved with the pathogens and therefore have not evolved resistance (Anderson et al., 2017). Alien species can be introduced to new habitats through several mechanisms, both natural and human aided; however human aided introductions have far exceeded natural introductions since human migration and commerce began (Mack et al., 2000). Global trade has increased the risk of introductions by transporting pathogens on commodities such as living plant material, fruits or vegetables, wood products such as logs and lumber or on wood packaging materials.   In recent years invasive pathogenic Phytophthora species have impacted natural ecosystems, agriculture, forestry and horticulture industries and interrupted trade. The Phytophthora genus is often considered one of the most important and damaging groups of plant pathogens in history. 83  In the past few decades, an increasing number of emergent species have caused landscape level disturbances in natural and forest ecosystems (Hansen, 2015). Although the exact origin of those Phytopththora species is often not known, they are believed to have been introduced through international plant trade and the planting of infected nursery stock (Brasier & Jung, 2006; Jung et al., 2016)(Jung <i>et al.</i>, 2016)(Jung <i>et al.</i>, 2016)(Jung <i>et al.</i>, 2016). Phytophthora ramorum, Phytophthora lateralis and Phytophthora cinnamomi are examples of pathogens that were spread between geographic regions via trade and subsequently escaped into natural forests causing extensive damage to native hosts (Hansen et al., 2000; Brasier & Jung, 2006; Brasier, 2008; Cahill et al., 2008; Hansen, 2008, 2015; Brasier & Webber, 2010; Grünwald et al., 2012; Burgess et al., 2017a). Economic impacts include the costs of enacting quarantine regulations, monitoring pathogen spread, and eradicating disease outbreaks. Reduction of income is due to the loss of marketable species or restrictions imposed on traded commodities. Port Orford cedar (Chamaecyparis lawsoniana (A. Murr .) Parl.) is a highly valuable wood and ornamental species; however Phytophthora lateralis has caused the near complete shutdown of both the nursery and commodity trade for this tree (Hansen et al., 2000). Phytophthora ramorum has had wide ranging economic impacts (Grünwald et al., 2012). In some nurseries, the costs associated with controlling P. ramorum have resulted in bankruptcy (Grünwald et al., 2012). The cost of government eradication efforts for P. ramorum in Curry County, Oregon between 2001 and 2008 was approximately 4.3 million US dollars (Kanaskie et al., 2010). Furthermore, climate change may exacerbate the problems associated with many invasive species; P. cinnamomi is predicted to expand into new regions as the climate becomes more favorable (Burgess et al., 2017a).  84  Non-indigenous pathogens can have serious wide ranging effects on ecosystems (Loo, 2009). Loss of a dominant host species can result in disequilibrium because of changes in microclimate, hydrology, and nutrient fluxes or the loss of food sources and habitat for wildlife (Ellison et al., 2005; Loo, 2009). Reduction of Port-Orford cedar in Oregon and California by P. lateralis could result in significant ecological changes including changes in soil calcium content, loss of habitat structure, as well as loss of stream structure and stability along riparian zones (Hansen et al., 2000; Loo, 2009). Death of tanoak (Lithocarpus densiflorus) in Oregon and California forests due to P. ramorum increases fire risk, may affect animals dependant on acorns as a food source, and could cause extended periods of chaparral vegetation in some areas (Hansen, 2008; Metz et al., 2011).  Currently, we have very little knowledge of what native or invasive Phytophthora species are present in BC forests, or whether anthropogenic activities could influence Phytophthora community composition and distribution. Some common pathogenic species such as P. cactorum P. cambivora, P. citricola, and P. megasperma were reported in the irrigation water from orchards in the Okanagan and Similkameen valleys in Southern British Columbia in the 1960s (McIntosh, 1966); however these reports would have been based on morphological data and may not reflect updated species designations, and many of these species are now considered complexes of closely related sister species. Phytophthora ramorum has been found in nurseries in British Columbia (Bilodeau et al., 2007), and P. lateralis has been found in urban areas of Victoria (Gorman, 1993). It is unknown if these pathogens have escaped into natural ecosystems in British Columbia.   85  Phytophthora monitoring is often conducted using different techniques, from different niches (water, soil, or foliage) over different seasons. The choice of method, niche, or sampling period may result in different community profiles and raises question on which methods provide the most accurate or inclusive data. One of the most frequently utilized survey methods is stream baiting, using Rhododendron leaves in mesh traps submersed in streams, followed by isolation of necrotic bait leaf tissue on selective media, however other baits are commonly utilized including pear tissue (Hulvey et al., 2010, Reeser et al., 2011, Hüberli et al., 2013, Shrestha et al., 2013, Brazee et al., 2016). Additional survey methods are now becoming more frequently used including baiting in the lab where water and soil collected in the field are baited in bottles or other containers (soil samples are flooded with water), or filtration methods where water is filtered and Phytophthora are isolated from the filters (Balci et al., 2007; Hwang et al., 2008; Reeser et al., 2011; Oh et al., 2013; Hansen et al., 2017). DNA meta-barcoding on environmental samples is also starting to be used, however this technique is relatively new in Phytophthora studies (Vannini et al. 2013; Sapkota and Nicolaisen 2015; Prigigallo et al., 2016; Riit et al., 2016; Esmaeili Taheri et al., 2017). Methods employed, baits used in trapping methods, as well as substrate or niche, and seasonality all provide somewhat differing results (Balci et al. 2007; Reeser et al., 2011; Hansen et al., 2012a; Hüberli et al., 2013; Oh et al. 2013).  In Chapter 2, we used culture-independent approaches to generate oomycete community profiles and found higher diversity in urban than in natural sites. We also found more Phytophthora-positive samples and more Phytophthora species in urban than in natural sites. Our objective in this work was to use culture-dependant methods to bait, isolate and barcode Phytophthora species to obtain a more detailed analysis of species present in both water and soil. We again 86  targeted sampling to the same environments with varying anthropogenic impact: urban areas (around agricultural or residential areas and an urban park), interface areas (natural areas within or bordering urban areas), and natural areas (forests with minimal urban influence). As in Chapter 2, we tested the hypothesis that anthropogenic activities influence Phytophthora communities and increase the frequency of putative non-indigenous species. We hypothesized that urban environments contain higher Phytophthora diversity and more introduced species compared to natural environments. In addition, we investigated differences in Phytophthora communities under two different urban influences: agricultural or residential. We hypothesized that different activities would result in different communities. We surveyed the Phytophthora species present across the three site types (urban, interface and natural) using a baiting in the lab approach where soil and water samples were collected in the field, baited in the lab for Phytophthora using rhododendron leaves and pear tissue. Phytophthora species were isolated from necrotic bait tissue and barcoded using the internal transcribed spacers 1 and 2 (ITS1 and ITS2). Both soil and water were sampled over two seasons (fall and spring) to obtain a more inclusive sampling of the community; two bait types were also used for the same reason. We hypothesized that fall and spring communities would harbor different species due to climate and moisture differences. In addition, to further test the hypothesis that different urban influences (agricultural or residential) result in different communities, and to compare survey methods, we surveyed streams in the urban sites using an in stream baiting method with rhododendron leaves in mesh traps. We hypothesized that different methods and different substrates would result in different species profiles.  87  Methods Direct baiting and isolation of Phytophthora from streams In the fall of 2011, four Phytophthora traps were set in small to moderate waterways in each of the five urban sites (described in Chapter 2, Figures 2.1 and 2.2, Table 2.1). Traps were constructed following the design from Hwang et al. (2008). Mesh screen was wrapped around a wooden dowel and stapled to make four pockets for leaf baits. Total trap size was approximately 20 x 20 cm and each pocket was four to five centimeters wide. Rhododendron leaves were used for baits, with two entire leaves and two leaves wounded with small cuts. Traps were tied to a wooden stake which was anchored into the waterway (Figure 3.1). Traps were left for ten weeks from the end of August to the beginning of November 2011. Foliage baits were collected approximately every 2 weeks and traps were reset with fresh bait leaves. Collected leaves were wrapped in moistened towel and brought back to the lab for immediate processing.   Bait leaves were rinsed in sterile water and pieces of leaf tissue (approximately 5mm x 5mm) were aseptically cut from the green edge of symptomatic zones and partially embedded in Phytophthora selective PARPH-CMA media (Kannwischer and Mitchell) containing 17 g/L cornmeal agar (Sigma-Aldrich, St. Louis, MO, USA), 10 mg pimaricin preparation (Sigma-Aldrich, St. Louis, MO, USA), 250 mg ampicillin sodium salt (Sigma-Aldrich, St. Louis, MO, USA), 10 mg Rifamycin SV sodium salt (Sigma-Aldrich, St. Louis, MO, USA), 100 mg pentachloronitrobenzene 95% (Sigma-Aldrich, St. Louis, MO, USA), and 50 mg hymexazol (Alfa Aesar, Haverhill MA, USA) in 1.0 liter deionized water. For each bait leaf, five pieces of leaf tissue were embedded on a single plate. Petri plates were incubated at room temperature for 88  two to four days and between one to five individual isolates were selected based on differing morphology, growth rates and morphology resembling Phytophthora (white, and mycelial growth form). Each selected isolate was sub-cultured onto a new individual plate of PARPH- CMA and incubated at room temperature.  Isolation and identification of Phytophthora from soil and water samples Water and soil samples were collected from all 13 sites described in Chapter 2 in the fall of 2012 and in the spring of 2013 (Figures 2.1 and 2.2, Table 2.1). Ten soil samples were collected as described in Chapter 2, and ten water samples (one litre each) were collected in each stream or waterbody adjacent to the soil sampling location, placed in Ziplock bags and brought back to the lab for processing. Samples were chosen to cover the largest number of independent waterways available at each site. A total of 260 water samples and 260 soil samples were used for lab baiting and culture-based Phytophthora identification. Water samples used for lab baiting methods (below) were processed within 24 hours of collection, and soil samples used for lab baiting methods (below) were processed within nine days of collection.  Water samples were baited with pieces of Rhododendron leaves and pear placed in Ziploc bags. Soil samples were poured into 1 litre rectangular plastic containers, flooded with sterile water in an approximately 1:2 ratio of soil to water, covered with a layer of cheesecloth and baited with Rhododendron leaves and pear (Balci et al., 2007). Soil and water samples were incubated at room temperature for five to seven days after which time sections of the leaves and pear used for baiting (approximately 5mm x 5mm) were aseptically cut from the edge of symptomatic zones 89  and partially embedded in Phytophthora selective PARPH-CMA media. Subcultures of individual isolates were made as described above for the Phytophthora trapping methods to obtain pure cultures of each isolate.  Fresh subcultures of each individual isolate that yielded Phytophthora-like morphology (white, and mycelial growth form) on PARPH-CMA media in fall 2011, fall 2012 and spring 2013, were transferred onto 2% malt extract agar (MEA) and cultures were incubated at room temperature. Mycelium was scraped from the cultures and placed in a sterile 1.5 ml Eppendorf tube. The tissues were ruptured by placing tubes in liquid nitrogen and then in a Tissuelyser (Qiagen, Germany) for 3 minutes at maximum speed. Sterile water (1.2 ml) was added to the samples followed by a 5 minute boiling step at (95.9 °C). Samples were then diluted 1:100 and 1 µl was used in PCR. The internal transcribed spacer region (ITS) of the nuclear ribosomal RNA genes (rDNA) was amplified in PCR using the primers ITS6 (Cooke et al., 2000) and ITS4 (Bruns et al., 1990). PCR reactions were performed in 25 µl volumes and contained 0.5 units of Invitrogen Platinum Taq DNA polymerase (ThermoFisher Scientific, Waltham, MA, USA), 1X PCR buffer, 1.5 mM MgCl2, 0.2 mM Invitrogen dNTPs (ThermoFisher Scientific, Waltham, MA, USA), and 1 µM of each primer. Reactions were performed with an initial heating step at 94 °C for three minutes, followed by 35 cycles of a denaturation step at 94 °C for 30 seconds, an annealing step at 55 °C for 30 seconds, an elongation step at 72 °C for one minute. The reaction was finished with a 10 minute extension step. PCR products were visualized on a 1% agarose gel stained with Ethidium Bromide. PCR products were sent to the Plateforme de Séquençage et de Génotypage des Genomes, Centre de Recherche du CHUL (Quebec, Canada) for sequencing. Samples from 2011 were sequenced in both directions using both primers (ITS6 and ITS4), whereas samples 90  from 2012 and 2013 were sequenced in only one direction using the ITS4 primer. PCR reactions that failed were repeated using no dilution, and then again with a 1:10 dilution. Samples that failed 3 times were eliminated from the experiment.   For 2011 samples, ITS sequences were assembled in Geneious version (6.0.5) (, Kearse et al., 2012) using the default settings. All assembled and unassembled sequences were compared with sequence data from Genbank (US National Center for Biotechnology Information) using the Basic Local Alignment Search Tool (BLAST, Altschul et al., 1990). Species identification was based on sequence homology (a minimum of 99% sequence similarity) and phylogenetic analysis (as described below). Initially, one phylogenetic tree was built with all isolates from this study. Isolates that grouped tightly together with a strong homology with only one species in Genbank were assigned to that species designation. For sequences with a close homology to more than one species from a species complex, or with discordant results between the homology search and phylogenetic results, representative sequences from each of the species in the complex or closely related species within the Phytophthora phylogenetic clade (as defined in Cooke et al. (2000) and Blair et al. (2008)) were downloaded and aligned with the query sequence using ClustalW (Thompson et al., 1994) in the BioEdit Sequence Alignment Editor (Hall, 1999). Fixed nucleotide differences were used to assign species where possible. If no fixed differences separated the species, the sequence was assigned a species complex designation. A species was considered as a new taxon if there were no species with a close sequence homology and if the sequence did not group with others in the phylogenetic analysis with sequences of the closest available species. Sequences with many ambiguities were considered undetermined.  91   For phylogenetic analysis to determine putative species assignments, assemblies (ITS1 and ITS4 from 2011) and sequences matching Phytophthora based on homology were trimmed in Geneious 6.0.5 using the default settings. Sequences from 2011 that could not be assembled were aligned where possible. Assembled, aligned or single sequences were visually inspected for ambiguities and edited using the International Union of Pure and Applied Chemistry (IUPAC) nucleotide ambiguity codes (Cornish-Bowden, 1985). Sequences were aligned with representative sequences of closely related species obtained from Genbank, including where possible, sequences from publically available culture collections or from studies describing new species. Sequences were aligned using ClustalW (Thompson et al., 1994) in the BioEdit Sequence Alignment Editor (Hall, 1999). Maximum likelihood phylogenies were reconstructed using default settings in the pipeline (Dereeper et al., 2008) with PhyML 3.0 (Guindon et al., 2010), the approximate likelihood ratio test (aLRT, Anisimova et al. (2006)) for testing branch support, and TreeDyn for tree visualization (Chevenet et al., 2006).   Resulting species assignments were screened to eliminate duplicate species from a sample, leaving only one representative per species for each sample in order to eliminate the possibility of over sampling the same individual.   Sequences from the culture based methods from all three years were combined and a variable file and a taxonomy file (using the taxonomy assigned from phylogenetic analysis and homology searches described above) were created matching the format for mapping files and taxonomy files in QIIME (Caporaso et al., 2010). QIIME was used to cluster the sequences using a 98.75% 92  similarity cutoff, assign taxonomy, pick representative sequences and to create an OTU file as described in Chapter 2. Sequences were aligned in QIIME using MUSCLE (Edgar, 2004) with default settings. A maximum likelihood phylogeny was reconstructed as above using the pipeline (Dereeper et al., 2008). OTU tables from the culture based collection and the DNA meta-barcoding (Chapter 2) were exported as matrices and combined in Excel for downstream analyses. Prior to combining the DNA meta-barcoding, data was filtered for a minimum of 3 sequences per OTU, per sample, and data was converted to a presence (1)-absence (0) matrix.  Data analysis The number of samples that yielded Phytophthora (herein called Phythophthora-positives) was determined for each site and each site type (urban, natural, and interface) and compared with a Chi-square test. The number of species, unique species, and unknown species was calculated for each site type and the number of shared species was calculated between site types.   Species over- and under-representation in each site type was tested by comparing, for each species, the number of individuals found in each site type to the total number of individuals. The degree of over- or under-representation of individuals for a species at a site type relative to the total number of individuals at the three sites was assessed by applying chi-square independence test. For each species, the approach consisted in placing the data in a table of n columns (representing the three different site types i.e. urban, natural, and interface) and m rows for the 93  different species. The Aij value represent the expected number of individuals for site i and species j :    And then the chi-square value calculated as follow:   Shannon’s diversity index was determined for each site and averaged for each site type. Weighted UniFrac (Lozupone & Knight, 2005) implemented in QIIME (Caporaso et al., 2010) was used to compare the phylogenetic Beta diversity between sites (using only culture data with full ITS1 and ITS2 sequences). UniFrac values were averaged for sites of the same site type to compare site types.   To look for possible effects from different anthropogenic activities, four of the urban sites were sub-divided into agricultural or residential (Chapter 2, Table 2.1). Agricultural sites were near large-scale nursery or agricultural operations while residential sites were near residential housing. The number and frequency of Phytophthora species, alpha diversity (Shannon’s diversity index) and beta diversity (weighted UniFrac) were calculated as above but using agricultural versus residential site types. A canonical correspondence analysis using the vegan package (Oksanen et al., 2016) in the R statistical environment (R Core Team, 2017) was conducted to investigate the effect of urban site type and sample type (water or soil) on species 94  profiles. In addition, the package ggtree (Yu et al., 2017) in the R statistical environment (R Core Team, 2017) was used to compare species abundance profiles for the two site types.  Putative hybrid species were defined as those with multiple heterozygous SNPs in the ITS sequence. Sequences with only a few heterozygous SNPs were not considered. The relationships of putative hybrid species to other similar species were investigated by comparing heterozygous SNPs in the putative hybrids to the alleles present in the other taxa to try to determine the closest matching species. Where possible, the ITS sequences were phased using a putative parent species as a template to create one ITS haplotype, and the other was created by subtracting the one putative parent to leave the alternate set of alleles.  Results Anthropogenic activity increases the frequency and diversity of Phytophthora. There was a gradient of increasing Phytophthora abundance in samples from low anthropogenic influence (natural 23%), moderate influence (interface 33%) to high anthropogenic influence (urban 41% PR excluded, 39% PR included) (Figure 3.2). Since PR represents the only urban site in the biogeoclimatic zone CWHdm with no corresponding natural or interface sites, comparisons were made both with and without this site to account for possible increases in diversity or changes to the species profiles that could be due to sampling artefact.  The percentage of Phytophthora-positive samples was higher in urban than in natural and interface sites (chi-square = 17.542, p < 0.0001, PR excluded). This gradient was consistent between all methods used: baiting water and soil in the lab and with culture-independent DNA meta-barcoding on soil from Chapter 2 (Figure 95  3.2). There were a total of 38 Phytophthora species found in fall and spring using a combination of baiting in the lab and DNA meta-barcoding. The highest number of species was observed in urban sites, followed by interface and natural sites (Figures 3.3 and 3.4). Urban sites also had more unique species (found in only one site type) and more unknown species. Only 12 species (32% of the total) were found in all three site types (Figures 3.3 and 3.4). Urban and interface sites shared more species (26%) than either urban and natural (5%) or interface and natural sites (11%) (Figure 3.3). Six species were unique to urban sites whereas only two species were unique to natural interface sites (Figures 3.3 and 3.4). We found a total of six unknown or previously undescribed taxa using baiting in the lab and DNA meta-barcoding (three additional unknowns, one of which was a putative hybrid, were found using stream baiting in urban sites; described later) (Figure 3.4). All five unknown taxa were found in urban sites, four in interface sites and none were found in natural sites. Three of those unknown taxa were found exclusively through DNA meta-barcoding (unknown species 1, 2 and 3, Figure 3.4; these are described in chapter 2). Two of the unknown species were putative hybrids (described later) that placed phylogenetically within clade nine. One was found multiple times in urban and interface sites (P. Unknown_Clade9_sp.4 (putative hybrid)) and the other one were found only once in an urban sites (P. Unknown_Clade9_sp.5 (putative hybrid)). There was a third putative hybrid, P. x polonica, which was found in all site types (Figure 3.4; described later).   Only two species were significantly over or under represented in a particular site type. Phytophthora pseudotsugae was found six times, exclusively in natural sites, and P. cambivora was found seven times, six times in interface sites and once in a natural site (Table 3.1). Species from nine of the ten clades described by Cooke et al. (2000) and Blair et al. (2008) were found in 96  this study; only clade five was not represented. Clades six, eight and nine were most commonly represented (Figure 3.4). The most common species overall were P. gonopodyides, P. lacustris, and P. chlamydospora (clade 6), followed by P. plurivora (clade 2) and P. syringae (clade 8) (Figure 3.4). Phytophthora plurivora was found in all site types, but less frequently in natural samples and P. syringae was only found in urban and interface sites.   Beta diversity (weighted UniFrac) was highest among natural sites and lowest among interface sites (Figure 3.5). Beta diversity was also low between interface and urban sites (Figure 3.5). PR had the highest UniFrac distances (Figures 3.5). The UniFrac distances between site types were significant, both with and without PR (ANOVA with PR, UniFrac values log transformed: F = 3.045, p = 0.015; ANOVA without PR, no transformation: F = 3.030, p = 0.017). Natural sites had the lowest and urban sites had the highest alpha diversity (Shannon’s diversity index) (Figure 3.6). SV had the highest Shannon’s diversity. There was no significant difference in Shannon’s diversity index among site types (Figure 3.6).   Different anthropogenic activities drive differences in Phytophthora communities. Phytophthora communities differed in agricultural and residential sites. Species composition was significantly associated with urban site type (F = 3.787, p = 0.001) and substrate sampled (water or soil; F = 5.648, p = 0.001) in a canonical correspondence analysis, where the differences in soil and water communities were observed within both residential and agricultural sites (Figure 3.7). Agricultural sites had slightly more species (29 versus 26 in residential), and higher overall abundance of Phytophthora species than residential sites (Figure 3.8). Interestingly, taxa within clade nine were significantly more frequent in agricultural sites (five taxa/OTUs found 55 times) 97  than in residential sites (three taxa/OTUs found three times; Figure 3.8; Table 3.1). Two clade 9 putative hybrids, Phytophthora x polonica and P. unknown Clade 9 sp.4, as well as Phytophthora irrigata were significantly over-represented in agricultural samples (Table 3.1). Clade 2, specifically Phytophthora plurivora were significantly over-represented in residential samples as well as Phytophthora gonopodyides (clade 6; Table 3.1).   Beta diversity (weighted UniFrac) was highest amongst agricultural sites and between agricultural sites versus all other site types, largely due to PR (Figure 3.5). Both agricultural sites had high Shannon’s diversity values, as did one of the residential sites (Figure 3.6).  Several unknown, undescribed and putative hybrid species were found. We obtained cultures of five Phytophthora spp. that belong to unknown taxa and one additional isolate that was a putative hybrid of P. polonica. Three of the unknown/undescribed isolates, also from clade 9, had several heterozygous SNPs in the ITS sequences suggesting they could be hybrids. Phytophthora unknown Clade 9 sp. 4 had high ITS identity (96 - 99%) to several other undescribed species from Oregon, West Virginia, South Africa and Ecuador (Table 3.2). In a comparison to a hybrid isolate from South Africa, P. sp. xHennops, our isolate is a close match possessing either a heterozygous SNP with nucleotides of both P. hydropathica and P. sp. xHennops, or matches P. sp. xHennops where P. sp. xHennops and P. hydropathica differ. The isolate in our study was found multiple times, in particular at the PR site. Identity between sequences of our isolates ranged from 98.6 to 100% to each other; however some differences were from ambiguous bases or difficulties with base calling due to background noise in the sequence traces, so we could not determine if these were isolates of the same species or multiple 98  hybridization events. Two other isolates, which may also be hybrids, Phytophthora unknown Clade 9 sp.5 and unknown Clade 9 sp. 6 did not share a close identity (above 99%) to any other species; the closest described species were P. parsiana or P. hydropathica (Table 3.2). These isolates were both only found once; unknown Clade 9 sp. 6 was only found through stream baiting.   The putative hybrid, P. x polonica appeared to have fixed heterozygous SNPs shared between the isolates (not confirmed). There were no matches observed in NCBI, but isolates were approximately 98% similar to P. polonica. For every heterozygous SNP, one of the bases always matched P. polonica, suggesting that this taxon is a hybrid between an unknown species and P. polonica. When we reconstructed or phased the two putative versions of the ITS sequences (one with the P. polonica alleles and one with the alternate alleles), the closest match to the alternate ITS sequence was with an undescribed species isolated from a stream in Maryland in the United States (Balci and Goss, 2013, unpublished data; GenBank Accession Number KC479206); however there were still at least five nucleotide differences between the alternate allele sequence obtained in our study and this sequence.  The two other unknown species were from clade six, both found only once from stream baiting in urban sites. Both isolates matched other undescribed species in GenBank (Table 3.2).   Different Phytophthora detection methods show different community profiles. A total of 42 Phytophthora species were found over the three years of sampling with all methods: stream baiting, baiting in the lab, and DNA meta-barcoding (Figure 3.9). Twelve species (mostly from 99  clade 6 and 9) were found with all three methods (Figure 3.9 and 3.10). Six species were shared between lab baiting and DNA meta-barcoding methods, three between lab baiting and stream baiting, and only one between stream baiting and DNA meta-barcoding (Figure 3.9 and 3.10). DNA meta-barcoding recovered the most species; however all were found in low frequency. Clades 6 and 9 were the most frequent clades encountered using stream baiting and baiting in the lab, whereas clade 8 was the most frequent using DNA meta-barcoding (Figure 3.9).   Phytophthora were found in 32% of samples in all years using all three methods. Baiting in the lab on soil yielded the lowest percentage of Phytophthora positive samples (17%) whereas stream baiting yielded the highest percentage of positive samples (64%; Figure 3.11). However DNA meta-barcoding on soil yielded the highest number of Phytophthora species (34; Figure 3.11). Baiting water in the lab yielded the least number of species (16); 13 of those were found in the urban sites. Stream baiting in the same urban waterways yielded 20 different species (Figure 3.11).   The spring/summer sampling period yielded over double the percentage of Phytophthora positive samples compared to the fall sampling period (69% compared to 31%). In addition there were 32 species recovered in the spring/summer sampling period compared to 26 from the fall sampling (data not shown).  100  Discussion Anthropogenic disturbances and specifically alien invasive species are one of the most important drivers of emergent plant diseases (Anderson et al., 2017). Globalization and international trade are responsible for the introduction and emergence of many alien invasive plant pathogens (Desprez-Loustau et al., 2010). Environmental change, or changes in the pathogen can also lead to emergence (Stukenbrock & McDonald, 2008). We have found a strong signature of anthropogenic influences on Phytophthora populations in urban areas. Our data suggests a higher Phytophthora species diversity and frequency in urban and urban/natural interface areas likely resulting from both introductions of alien species and increases in natural populations due to environmental disturbance. Secondly, we find two putative hybrid Phytophthora species showing possible patterns of early emergence in agriculturally influenced environments. These results suggest anthropogenic influences not only shape Phytophthora communities, but likely also influence the frequency of disease emergence.  Phytophthora species are especially prone to introductions since they can easily be transported in soil of asymptomatic plants and be overlooked during inspections (Desprez-Loustau et al., 2010; Migliorini et al., 2015). Surveys of nursery plants have uncovered a relatively high frequency of Phytophthora species associated with soil (Dart & Chastagner, 2007; Parke et al., 2014; Jung et al., 2016; Prigigallo et al., 2016). Higher species diversity in urban habitats could be explained by the influx of non-indigenous Phytophthoras brought through both horticultural and agricultural activities. This pathway of introduction was strongly suggested by the results of a large-scale survey of Phytophthora diversity in nurseries producing stock for forests, 101  horticulture, ornamental, amenity and landscape plantings (source populations) as well as areas where stock would have been planted (sink populatlions) including forests, riparian zones, landscapes, ornamental and horticultural plantations (Jung et al., 2016). Similar infestation rates were found in nurseries producing planting material as in the areas where planting stock would have been destined. These rates of infestation were much higher than what had been found in mature forests (Jung et al., 2016).   Movement of Phytophthora from urban to natural environments is likely facilitated through interface environments. There was a gradient in the impact of the anthropogenic activities where the measures of abundance and diversity in interface sites were always intermediate between those in urban and natural sites. We also found the highest number of shared species between urban and interface sites. This indicates that the most anthropogenically affected sites are a repository of Phytophthora species diversity but that the interface sites could act as a bridge toward natural environments supporting previous suggestions that urban trees act as bridge-head for invasive species establishment and spread to natural environments (Paap et al., 2017). Under conducive conditions, a single Phytophthora species can cause an epidemic, a pattern that has been observed for some of the most damaging Phytophthora species, including two species that we recovered with DNA meta-barcoding. Phytophthora ramorum is believed to have been introduced through nursery material in California and Oregon and subsequently spread to adjacent forests where it caused extensive mortality of coast live oak (Quercus agrifolia Nee.) and tanoak (Rizzo et al., 2002; Mascheretti et al., 2008). It is now found in natural forest ecosystems extending from California through Oregon and Washington (Rizzo & Garbelotto, 2003; Chastagner et al., 2009; Strenge et al., 2016). In our study, P. ramorum DNA was found 102  using DNA meta-barcoding in urban, interface and natural sites; however we could not confirm results using a species specific quantitative PCR assay (data not shown), possibly because of the lower sensitivity in the PCR detection method versus DNA meta-barcoding, PCR inhibitors in soil, or a very low concentration of P. ramorum DNA in the sample. Nevertheless, these results suggest that this pathogen has escaped the nursery environment in BC, or that it is spreading from other sources in the Pacific Northwest. In spite of extensive monitoring and eradication from nurseries in BC, there are no previous reports of the disease occurring outside of nurseries in BC. Likewise, Phytophthora lateralis may be spreading from urban into natural sites. Although only found in natural and interface sites in our study, it is considered introduced in North America (Hansen et al., 2000) and has previously been found in urban areas of BC (Hamelin, unpublished). This pathogen causes mortality of Chamaecyparis lawsoniana in the natural range of the tree in Oregon (Hansen et al., 2000). C. lawsoniana does not occur naturally in BC, so our finding of P. lateralis in natural and interface sites in BC suggests that it must survive in natural environments on another host. Phytophthora lateralis has been found in the UK on western red cedar, and its host range extends to other members of the Cupressaceae family, including yellow cedar (Cupressus nootkatensis), Pacific yew and juniper, all of which occur naturally in BC (Hansen et al., 2000; Schlenzig et al., 2015, 2017). The possibility of P. ramorum and P. lateralis having spread to natural environments should be of concern as it could cause damage to natural ecosystems and to international trade.   A lack of knowledge on the biodiversity in a community or natural geographic range of a species can make it challenging to discern introduced species from native ones (Desprez-Loustau et al., 2007). Recent survey studies are recovering many new species, however it is difficult to know if 103  these species are native or introduced (Linzer et al., 2009). In addition to information on distributions, comparisons of the species composition and frequencies in natural versus urban environments may help to distinguish introduced from native species. We found several species only in urban areas that could represent introductions. For example, P. drechsleri was detected in only one urban sample in an agriculturally intense region. This pathogen causes disease on a variety of ornamental plants. In the eastern United States molecular evidence, specifically a single shared clone, suggests it was spread through several nursery facilities via movement of infected plants from a shared distributor (Lamour et al., 2003). Phytophthora hibernalis was detected in one sample from a region that contains several horticultural nurseries and ornamental plant stores. This species causes citrus brown rot but was recently found infecting Rhododendron plants in nurseries in California and Oregon (Blomquist et al., 2005; Yakabe et al., 2009; Reeser et al., 2011; Sims et al., 2015). Phytophthora multivora and P. gregata were all limited to urban sites in our study and are thought to originate from Western Australia (Jung & Burgess, 2009). Phytophthora multivora is mostly limited to nurseries, plantations and agricultural crops outside of Western Australia where it is widespread in natural ecosystems (Jung & Burgess, 2009). Phytophthora gregata also has a limited occurrence outside of Australia. This species was associated with dying plants in alpine heathland vegetation in Australia (Jung et al., 2011). Phytophthora gregata is from clade 6, which generally comprises weak or opportunistic pathogens; however it could still cause disease in other geographic ranges or on non-native hosts. Other species may be more difficult to determine natural versus introduced ranges. Several species were found in low frequency in our study, and were absent in natural sites; however they could still be endemic suggested by their occurrence in natural areas in other studies focused on 104  the Pacific Northwest. For example, Phytophthora riparia was found frequently in Oregon and Alaska but was not associated with disease (Reeser et al., 2011; Hansen et al., 2017).   Distribution alone may not be enough to determine the origin of a pathogen. Phytophthora pseudosyringae and P. nemorosa have been found frequently in Oregon and in California and, similar to P. ramorum, are associated with Tanoak, Bay Laurel and Coast live oak (Hansen et al., 2017). Phytophthora nemorosa has only been found rarely in nurseries and not outside of Oregon and California, while P. pseudosyringae has been also been found in Alaska and extensively in Europe (Murphy & Rizzo, 2006; Wickland et al., 2008; Hansen et al., 2017). Their geographic range, lower pathogenicity on local hosts, and variation in the internal transcribed spacer and mitochondrial cytochrome oxidase spacers could suggest the pathogens are indigenous in Oregon and California (Hansen et al., 2017). However low levels of genetic diversity and absence on herbarium specimens could also suggest they are introduced (Monahan et al., 2008; Linzer et al., 2009; Kozanitas et al., 2017). If P. pseudosyringae is introduced, the extensive distribution along the Pacific Northwest may be concerning as this pattern could suggest emergent behavior that could be due to either a particularly invasive species or that spread may be facilitated by another factor such as climate change.  Agricultural practices create ecosystems which are often conducive to pathogen spread. Densely planted and genetically uniform host populations, irrigation practices and tillage may all support pathogen spread, and the homogenous landscape may protect pathogen communities from environmental fluctuations that reduce population sizes (Stukenbrock & McDonald, 2008). These agro-ecosystems may also support evolutionary events such as hybridization, host jumps, 105  host-tracking or horizontal gene transfers leading to the emergence of new pathogens (Stukenbrock & McDonald, 2008). The results of our study suggest that Phytophthora communities are evolving in response to agricultural practices. Higher pathogen diversity in urban areas in our study may not only result from introductions, but also from an increase in the frequency of rare endemic species. We detected some pathogens that are likely present due to a close relationship with their agricultural hosts, specifically a species from the Phytophthora rubi/ P. fragariae complex. Both P. rubi and P. fragariae are prevalent in the agricultural industry in the Pacific Northwest and P. rubi is possibly endemic (Gigot et al., 2013; Stewart et al., 2014). Furthermore, sites with agricultural activities had very different communities among each other compared to sites in residential areas or parks. Frequency appeared to increase for different species or different clades depending on the dominant urban activity. This was particularly evident for clade nine in agricultural areas and for two of the putative hybrid species, unknown Clade 9 sp.4 and P. x polonica, which were locally abundant in agricultural sites (each in a different agricultural site), but were less abundant (or absent) in other sites. This pattern not only suggests rapid local spread possibly resulting from agricultural activities, but may also suggest that these putative hybrids have emerged in an anthropogenic ecosystem.   Hybridization can provide an evolutionary fast-track. It can lead to changes in pathogen properties and in some cases, the ability to infect new hosts leading the the emergence of a new pathogen (Stukenbrock & McDonald, 2008). Some invasive tree pathogens, such as dutch elm disease (Ophiostoma spp.) and poplar leaf rusts (Melampsora spp.), have been shown to have hybridized, providing the hybrids with abundant new genetic material that selection can act upon. Hybridization between Ophiostoma ulmi and Ophiostoma novo-ulmi resulted in the 106  transfer of genetic material to O. novo-ulmi including a mating type allele which was missing in the introduced population, leading to sexual reproduction, increased diversity and increased fitness in O. novo-ulmi populations (Brasier, 2001). Melampsora xcolumbiana, a hybrid of the poplar leaf rusts Melampsora medusae and Melampsora occidentalis, is not only pathogenic on the parental host species but has also emerged on hybrid poplar which is resistant to both parental rust species (Newcombe et al., 2000, 2001). Phytophthora alni, responsible for alder decline, emerged through multiple hybridization events leading to a more aggressive pathogen of alder than either of its parental species (Brasier et al., 2004; Ioos et al., 2006). We found evidence for at least three putative hybrids; one of which was limited to urban sites, another to urban and interface sites, and the third was widespread. For Phytophthora unknown Clade 9 sp.4 (locally abundant in one of the agricultural sites), we could not determine if there was a single hybridization event, or multiple hybridization events creating a hybrid swarm similar to what has been observed in the Phytophthora alni complex (Brasier et al., 2004; Ioos et al., 2006). Although it was difficult to determine the exact relationships of our unknown Clade 9 sp.4 isolates to several similar taxa reported in other studies without further genetic characterization, it appears as though they could either be the same species, or the result of a separate hybridization event between the same or similar parental species. The closely related hybrid, P. sp. xHennops from South Africa, was characterized by cloning the ITS gene region to reveal putative parents, P. hydropathica, and an unknown clade 9 species (Oh et al., 2013). We can use the distribution of this and other similar taxa in NCBI to attempt to hypothesize the origin and invasion route of this new species. In addition to the South African P. sp. xHennops, our isolate also matches isolates found in the rhizosphere of red alder in Oregon (Sims et al. 2015), from Rhododendron in a nursery in Oregon (Knaus et al., 2015), streams in West Virginia and Ohio 107  (Mansfield et al. 2008, unpublished data; GenBank accession numbers EU644712, EU644713, EU644715 and EU644718), and from the Cuyabeno Reserve in the Amazonian region of Ecuador (Coffey, 2012). This distribution suggests that this species could be from the southern hemisphere originating in the Amazon Forest, and that isolates found in South Africa, the US and Canada were introduced on imported plant material. The frequency of hybrids could indicate that P. hydropathica and the unknown parent, or parents, are capable of readily hybridizing when in contact, or that the hybrid is relatively fit, does well in the introduced environments and is spreading. The emergent pattern in this species could support the second scenario. Given the emergent pattern, further monitoring and pathogenicity testing would be warranted in this species.  The other species that was locally abundant in the other agricultural site was the putative hybrid Phytophthora x polonica, also from clade 9. The parent, P. polonica, could be European given its current known distribution and weak pathogenicity to alder in Poland (Belbahri et al., 2006; Sárándi-Kovács et al., 2016); however there are no reports of the hybrid there. Our isolates could be a clone spreading after a single hybridization event suggested by fixed heterozygous SNPs in all isolates. Phytophthora polonica could have been introduced to BC and hybridized with a local species; however, neither P. polonica nor any other closely matching isolates were recovered in our study. A hybridization event could have occurred with a local endemic and rare species that our survey did not uncover; however, the more widespread distribution of the hybrid in all site types and lack of either parental species makes this scenario less likely. In the Netherlands; Phytophthora cactorum hybridized with Phytophthora hedraiandra after P. hedraiandra was introduced (Man in ‘t Veld et al., 2007). The hybrids appear to be spreading 108  affecting multiple new hosts, possibly aided by anthropogenic activities. Interestingly, P. hedraiandra has only been detected once in the Netherlands suggesting it cannot successfully establish, or that it exists in very low frequency (Man in ‘t Veld et al., 2007). A similar situation could have occurred with the P. x polonica hybrid wherever it originates from. Hybridization appears to be common in Phytophthora, often resulting from closely related species being brought together under conducive conditions through anthropogenic activities (Brasier, 2001; Burgess, 2015); but may also occur in natural ecosystems as observed in Australia (Burgess, 2015). Furthermore, hybridization appears to be common amongst species from clades six and nine, both frequently found in waterways (Hüberli et al., 2013; Nagel et al., 2013; Oh et al., 2013; Burgess, 2015; Stamler et al., 2016). Although not all hybridization events result in viable species, the emergent nature of some hybrids resulting from a new host range or increased pathogenicity is concerning, in particular if these events are on the rise with globalization and increased international trade.  Overall, similar species were found in our study compared to other studies from the Pacific Northwest (Reeser et al., 2011; Sims et al., 2015). Eighteen different species were found in Oregon and Alaska, all of which, except “new species three”, were found in our study (Reeser et al., 2011). The most common species found in our study, P. gonapodyides, P. lacustris and P. chlamydospora (clade 6) were also the most common species in Oregon and/or Alaska streams (Reeser et al., 2011) and amongst the most common species found in riparian Alder stands in Oregon (Sims et al., 2015). These species were found in all site types in our study and could represent endemic species or species with a long history of establishment. It has been suggested that they are resident in streams, colonizing fresh plant litter and infecting riparian plants under 109  flooded conditions (Brasier et al., 2003; Jung et al., 2011). Many clade six species are ubiquitous, often not found associated with disease, or are opportunistic, causing disease on stressed plants or in overly wet conditions (Brasier et al., 2003; Jung et al., 2011; Hansen et al., 2012b, 2017; Stamler et al., 2016).   The three survey methods we used provide a complementary picture of Phytophthora communities and highlight that results are somewhat dependent on the method used. Our study suggests that soil harbors higher species diversity independent of the sampling method. The culture independent DNA meta-barcoding method (Chapter 2) yielded the highest number of species overall, as is expected. However some species, such as P. cambivora did not appear in our DNA meta-barcoding sampling but were readily found with our baiting approach. DNA DNA meta-barcoding yielded double the number of Phytophthora positive soil samples and nearly double the number of species. Similar results have been found in comparisons of baiting versus direct detection in soil. Comparisons of Peronomycete diversity from DNA extracted from hempseed baits versus DNA extracted directly from soil reveal higher diversity in soil suggesting not all species can be baited, that differences such as the vegetative state or spore types present in the soil influence detection of a species or that bait choice limits the species detected (Arcate et al. 2006). A higher number of Phytophthora species in soil were also found using DNA meta-barcoding compared to baiting by Vannini et al. (2013).   Similar to our study, Hansen et al. (2017) baited soil in the lab and also found a low frequency of Phytophthora species. Overall, Phytophthora was much less frequent in isolations from soils compared to canopy drip and plant foliage (Hansen et al., 2017). However, Balci et al. (2007) 110  found a slightly higher percentage of Phytophthora positive soil samples in Oak forests in the USA, and Oh et al. (2013) found a much higher percentage of Phytophthora positive soil samples in South Africa, Interestingly, even though we had a lower percentage of Phytophthora positive samples, we found a higher number of species considering only the lab baiting results compared to both of these studies. This could be due to differences amongst sites, or land use patterns in the study areas and suggests that even with the low number of positives, we likely have a relatively good sampling of the diversity present.   Stream baiting recovered more species from the same locations than baiting in the lab did. The disadvantage to the stream baiting method is the labor involved with leaving traps in the water and returning to sites to collect and replace the bait leaves. Stream baiting may also underestimate the true species diversity. Stream baiting was shown underestimate diversity compared to water filtration (Hwang et al., 2008; Brazee et al., 2016). In addition, culture dependent methods likely favor fast growing species or species better suited to lab conditions. The dominance of clade six species in streams is likely a result of the ease of recovering Phytophthora species from waterways, and also due to their overall abundance in aquatic environments. Additionally, the baits used may limit the species found and using only a single bait type likely results in underestimating the diversity in stream habitats (Hüberli et al., 2013).   Phytophthora are more frequent during the spring and early summer than in the fall, likely due to more available moisture and cooler temperatures. This was apparent in both the frequency of Phytophthora as well as the higher number of species found. Similar results were found by Balci et al. (2007) who also isolated twice as many Phytophthora in the spring compared to the fall. 111  Results from our study and the studies of others suggest that the most comprehensive picture of Phytophthora diversity is obtained by including multiple sampling substrates and environments, by using multiple methods, as well as by sampling over different times of the year to capture seasonal fluctuations.   One of the biggest challenges to surveying the Phytophthora community for invasive pathogens is how to determine if a species is introduced or native. Many species are largely unknown or new, and it can often be difficult to determine if a particular species is native or not. Studies such as this one not only uncover evolutionary patterns and mechanisms of pathogen emergence, but they serve to characterize Phytophthora communities and geographic ranges of Phytophthora pathogens, both natural and invasive. In addition, our data provides valuable information on putative invasive species in BC and will help to direct research and management efforts to high risk species.   112  Table 3.1: Species that were significantly over or under represented by site type (urban, interface or natural) and by urban site type (residential or agricultural) based on a chi-square analysis.  Comparison: Urban, Interface and Natural Species Urban Interface Natural Χ2 Urban Χ2 Interface Χ2 Natural P. pseudotsugae 0 0 6 2.907 1.921 19.868*** P. cambivora 0 6 1 3.391 6.307* 0.099 Comparison: Residential and Agricultural Species Agricultural Residential Χ2  Agricultural Χ2   Residential P. gonapodyides 28 56 7.905** 10.210** P. plurivora 2 18 7.628** 9.852** P. x polonica 20 1 5.631* 7.273** P. unknown clade 9 sp. 4 (putative hybrid) 17 1 4.631* 5.982* P. irrigata 12 0 4.054* 5.236* Clade 9 55 3 15.224 *** 19.665 *** Clade 2 4 21 7.226** 9.334** * Significance where * is significant at a probability lower than 0.05, ** is significant at a probability lower than 0.01, and *** is significant at a probability lower than 0.001.  113  Table 3.2: Unknown or putative new taxa and the origin of their closest hits in a homology search on Genbank.  Putative new taxa Closest blast hit % ID Accesion number Author Origin of closest hit P. unknown clade 9 sp. 4 (putative hybrid) Phytophthora parsiana  99.4% KJ666742.1 Sims,L.L., Sutton,W., Reeser,P.W. & Hansen,E.M. Oregon Alnus rubra rhizosphere  Phytophthora sp. Marion-2012  99.2% KM501456.1 Knaus,B.J., Fieland,V.J., Graham,K.A. & Grunwald,N.J. Rhododendron leaf in Oregon nursery  Phytophthora sp. MC5 99.0% EU644718.1 Mansfield,M.A., Balci,Y., Geiser,D. & Kang,S. stream-baiting in West Virginia and Ohio  Phytophthora sp. BF4  98.9% EU644715.1 Mansfield,M.A., Balci,Y., Geiser,D. & Kang,S. stream-baiting in West Virginia and Ohio  Phytophthora sp. TF16 98.9% EU644713.1 Mansfield,M.A., Balci,Y., Geiser,D. & Kang,S. stream-baiting in West Virginia and Ohio  Phytophthora sp. MC53  98.9% EU644712.1 Mansfield,M.A., Balci,Y., Geiser,D. & Kang,S. stream-baiting in West Virginia and Ohio  Phytophthora sp. xHennops  98.9% KC855182.1 Oh,E., Wingfield,M.J. & Burgess,T.I. Pietermaritzburg, Kwa Zulu Natal, South Africa; forest stream  Phytophthora sp. 1 EEO-2011 5 98.9% GU799642.1  Oh,E., Gryzenhout,M., Wingfield,B.D. & Wingfield,M.J. stream water in South Africa  Phytophthora sp. 1 EEO-2011  98.9% GU799640.1 Oh,E., Gryzenhout,M., Wingfield,B.D. & Wingfield,M.J. stream water in South Africa  Phytophthora sp. P16858 98.9% GU594789.1 Coffey,M.D., Brar,A.K., Xu,E. & Zhang,Y.H. Ecuador: Amazonian region, Cuyabeno Reserve  Phytophthora sp. P16855 98.9% GU594787.1 Coffey,M.D., Brar,A.K., Xu,E. & Zhang,Y.H. Ecuador: Amazonian region, Cuyabeno Reserve 114  Putative new taxa Closest blast hit % ID Accesion number Author Origin of closest hit …Continued from above Phytophthora sp. 1 EEO-2011  98.9% GU799639.1  Oh,E., Gryzenhout,M., Wingfield,B.D. & Wingfield,M.J. stream water in South Africa  Phytophthora sp. Maryland 5 98.7% KC479206.1 Balci,Y. & Goss,E. Maryland stream  Phytophthora hydropathica  96.8% MF959534.1 Shrestha,S.K., Zhou,Y. & Lamour,K. Tennessee, USA, stream water P. unknown clade 9 sp. 5 (putative hybrid) Phytophthora sp. x Kunnunara-like  97.5% KU682602.1 Jung,T., Chang,T.-T., Bakonyi,J., Seress,D., Perez-Sierra,A., Xiao,X., Hong,C., Scanu,B., Fu,C.-H., Hsueh,K.-L., Maia,C., Abad-Campos,P., Leon,M. & Horta Jung,M. Taiwan, river raft  Phytophthora sp. Hennops river  97.3% HQ292655.1 Oh,E., Gryzenhout,M., Wingfield,B.D. & Wingfield,M.J. South Africa: Gauteng river water  Phytophthora sp. Kunnunara 97.3% EF437222.1 Cunnington,J. Australia; water  Phytophthora parsiana  96.7% AY659741.1 Mostowfizadeh-Ghalamfarsa,R., Cooke,D.E. & Banihashemi,Z.  P. unknown clade 9 sp. 6 Phytophthora sp. Umtamvuna  97.5% KC855181.1 Oh,E., Wingfield,M.J. & Burgess,T.I. Umtamvuna, Kwa Zulu Natal, South Africa; forest stream  Phytophthora sp. 1 EEO-2011  97.1% GU799641.1 Oh,E., Gryzenhout,M., Wingfield,B.D. & Wingfield,M.J. stream water in South Africa  Phytophthora hydropathica  97.0% KC734446.1 Vydrzel,H., Prospero,S. & Schoebel,C.N. Switzerland 115  Putative new taxa Closest blast hit % ID Accesion number Author Origin of closest hit P. unknown clade 6 sp. 7 Phytophthora sp. forestsoil-like  99.7% KU682574.1 Jung,T., Chang,T.-T., Bakonyi,J., Seress,D., Perez-Sierra,A., Xiao,X., Hong,C., Scanu,B., Fu,C.-H., Hsueh,K.-L., Maia,C., Abad-Campos,P., Leon,M. & Horta Jung,M. Taiwan, river raft  Phytophthora sp. sylvatica  98.9% EF522140.1 Belbahri,L., Lefort,F., Calmin,G. & Oszako,T. Poland: Siestrzen; Alnus glutinosa rhizosphere P. unknown clade 6 sp. 8 Phytophthora sp. 1 KS-2013  100% KF286861.1 Milenkowic,I., Sikora,K., Borys,M., Oszako,T., Keca,N. & Nowakowska,J.A. Poland: Fraxinus excelsior soil  Phytophthora sp. 2 FFL-2008 100% EU594600.1 Lakatos,F. & Szabo,I. Hungary  Phytophthora sp. IBL/2011/5/1/1 99.7%  JX276039.1 Nowakowska,J., Milenkovic,I., Oszako,T., Borys,M., Sikora,K., Belbahri,L. & Kubiak,K. Polish and Serbian forests  Phytophthora sp. 92-207 99.7% EU106590.1 Lefort,F., Belbahri,L., Calmin,G., Bragante,F., Bolay,A., Gerber,B. & Viret,O. Switzerland; Ilex aquifolia  Phytophthora sp. AG43 99.7% EU000143.1 Lefort,F., Belbahri,L., Calmin,G., Bragante,F., Bolay,A. & Viret,O. Switzerland: Collonges-Bellerive; soil of infected Ilex aquifolium  Phytophthora sp. P1050 99.7% AF541905.1 Brasier,C.M., Cooke,D.E., Duncan,J.M. & Hansen,E.M. Sweden from root of infected raspberry             116  Putative new taxa Closest blast hit % ID Accesion number Author Origin of closest hit P. x polonica Phytophthora polonica 98.3% many P. polonica  many matches    Phytophthora polonica  97.9% GU258841.1 Coffey,M.D., Brar,A.K., Xu,E., Sarhan,E.A. &Cunningham,I.M. World Phytophthora collection Putative parental allele 1ab Phytophthora polonica  99.6% KX618506.1 Oszako,T., Tkaczyk,M., Nowakowska,J.A., Borys,M. & Tereba,A. Plant tissue, Alnus glutinosa Gaertn., Poland  Phytophthora polonica  99.6% KF358225.1 Yang,X. & Hong,C. Plant nursery in Mississippi Putative parental allele 2b Phytophthora sp. Maryland 8 98.9% KC479194.1 Balci,Y. & Goss,E. Stream from Maryland, USA amatches many other P. polonica; bputative parental alleles of P. x polonica. 117   Figure 3.1: Stream baiting using mesh traps with rhododendron leaves.  118   Figure 3.2: Percentage of Phytophthora positive samples by site type: Natural versus Interface or Urban; total of all methods (top left), and by each method where baiting and culturing was done in the lab and Meta DNA-barcoding was done on DNA extracts from soil samples.        0%20%40%60%Natural Interface Urban Urban noPR% Phytophthora positive samplesTotal Phytophthora postive0%20%40%60%Natural Interface Urban Urban noPR% Phytophthora positive samplesWater: baiting and culturing0%20%40%60%Natural Interface Urban Urban noPR% Phytophthora positive samplesSoil: baiting and culturing0%20%40%60%Natural Interface Urban Urban noPR% Phytophthora positive samplesSoil: Meta DNA-barcoding119   Figure 3.3: Total unique and shared species between the three site types, urban, interface and natural. Sampling occurred over both the fall and spring sampling periods, including both baiting in the lab and Meta DNA-barcoding on soil. Total samples for each group were 239 for urban, 229 for interface, and 234 for natural. PR excluded from counts.    120  Figure 3.4: Frequency of each Phytophthora species by site type in fall and spring using baiting in the lab for water and soil and meta DNA-barcoding on soil (stream baiting not shown). Clade nomenclature: Cooke et al. (2000) and Blair et al. (2008). unknown species 2unknown species 3unknown species 1P. gallicaP. intercalarisP. irrigataUnknown Clade9 _sp.5Unknown Clade9_sp.4P. hydropathicaP. x polonicaP. hibernalisP. lateralisP. sansomeanaP. ramorumP. taxon kelmaniaP. cryptogeaP. syringaeP. fragariaeP. cambivoraP. europaeaP. gregataP. amnicolaP. ripariaP. inundataP. megaspermaP. bilorbangP. chlamydosporaP. lacustrisP. gonapodyidesP. quercinaP. nemorosaP. pseudosyringaeP. piniP. siskiyouensisP. multivoraP. plurivoraP. pseudotsugaeP. cactorum0% 5% 10% 15%Natural0% 5% 10% 15%Urban0% 5% 10% 15%Interface1 2 3 4 6 7 8 9 10  Frequency of species in site type Clade 121   Figure 3.5: Weighted UniFrac distance between sites. Lower portion of the diagonal represents comparisons between sites within site types (natural, interface or urban). Values above the diagonal are the average of the sites with the standard deviation in brackets. a is the average and standard deviation without PR in the calculation. Urban sites are designated with Park, Res for sites within residential areas, and Agri for sites within agricultural areas.   122     Figure 3.6: Shannon’s diversity of sites. Values below site type are average Shannon’s diversity for each site type with standard deviation in brackets. Urban sites are designated with Park, Res for sites within residential areas, and Agri for sites within agricultural areas.   00.511.522.53CR SA GT BW GL HM CV PS BP HV PR MD SVShannon's diversity Natural Interface UrbanPark Res Agri AgriRes1.82 (0.39)2.19(0.35)2.32(0.47)123   Figure 3.7: The relationship between urban site type (residential or agricultural), sample substrate (soil or water) and species presence and absence patterns. Results are based on a Canonical Correspondence Analysis (CCA). Data includes samples from soil and water baited in the lab and meta DNA-barcoding on soil only. Both site type and substrate were significantly associated with species composition.     124                Figure 3.8: Representation of the abundance of Phytophthora clades (circle diameter) in urban sites that are in residential or agricultural areas. The dendrograms represent clades as defined using sequences in    Agricultural Residential Clade_1Clade_2Clade_3Clade_4Clade_8Clade_6Clade_7Clade_9Clade_10UnknownClade_1Clade_2Clade_4Clade_8Clade_6Clade_7Clade_9Clade_10Unknown125   Figure 3.9: Phytophthora species present in south west BC and their frequency based on collection method. Stream baiting was done in fall (2011) in urban sites only, lab baiting was done on water and soil collected in fall (2012) and spring (2013) and Meta DNA-barcoding was done on soil collected in fall (2012) and spring (2013).  0 20 40 60 80 100Meta barcoding0 20 40 60 80 100Lab baiting0 20 40 60 80 100P. borealisP. fragariaeP. hibernalisP. piniP. siskiyouensisP. unknown_clade_6_sp.1P. Unknown_clade_6_sp.2P. Unknown_Clade9_sp.2P. Unknown_Clade9_sp.3P. gregataP. lateralisP. nemorosaP. sansomeanaunknown_species_2P. multivoraP. ramorumunknown_species_1unknown_species_3P. inundataP. megaspermaP. pseudosyringaeP. gallicaP. pseudotsugaeP. ripariaP. cactorumP. quercinaP. amnicolaP. cambivoraP. kelmaniaP. europaeaP. hydropathicaP. intercalarisP. syringaeP. taxon OaksoilP. irrigataP. Unknown_Clade9_sp.1P. cryptogeaP. x polonicaP. plurivoraP. chlamydosporaP. lacustrisP. gonapodyidesStream baitingPercent of samples with each species 126   Figure 3.10: Unique and shared species by method: 1) Stream baiting, conducted only in urban sites in 2011, 2) Baiting soil and water samples in the lab, conducted on all samples in 2012 and 2013 and 3) DNA meta-barcoding, conducted on soil samples in 2012 and 2013. There were a total of 42 species found with all methods combined.           127                      Figure 3.11: Comparison of Phytophthora positive samples and number of species by method and by substrate. DNA meta-barcoding was done on soil only in 2012 and 2012, baiting in the lab was done on soil and water in 2012 and 2013, and stream baiting was done in five urban streams in 2011.  0%10%20%30%40%50%60%70%Metabarocoding Lab baiting Lab baiting Lab baitingurban onlyStream baitingurban onlyPhytophthora positive samplesSoil Water 0510152025303540Metabarocoding Lab baiting Lab baiting Lab baitingurban onlyStream baitingurban onlyNumber of Phytophthora speciesSoil Water128  Chapter 4: Mitotic recombination and a two-speed genome drive evolution in asexual lineages of the sudden oak death pathogen Phytophthora ramorum  Introduction The previous two chapters of this thesis suggest that anthopogenic activities are an important pathway of introduction for Oomycetes and Phytophthora species. However, it is unknown how many introduced species will become established and invasive. Furthermore, it is unknown if particular characteristics of Phytophthora species make them better invaders. In order to understand success in invasive pathogens and to predict risk of invasion it is important to understand the evolution of known successful invasive alien pathogens.   Invasive alien tree pathogens are responsible for some of the most devastating forest disease epidemics. Successful alien tree pathogens can spread over thousands of km within a few decades invading new habitats and attacking new hosts under new environmental conditions. There are numerous examples of invasive tree pathogens that became successfully established and created ecosystem-wide changes by nearly eliminating their hosts (Loo, 2009). But there is a genetic paradox to many of these invasions: pathogen populations often undergo genetic bottlenecks following founder events caused by the introduction of a limited number of propagules (Pérez et al., 2006, Gladieux et al., 2015). This can result in a loss of genetic diversity, often accompanied by the loss of sexual reproduction. In heterothallic species (that require two mating types to complete a successful sexual cycle) founder events reduce the 129  likelihood of sexual reproduction (Fry et al., 1992; Garbelotto, 2008; Brasier, 2012). Sometimes the absence of one mating type completely eliminates sexual reproduction, resulting in the proliferation of clonal lineages. The ability to reproduce sexually is considered an important life trait that can impact a pathogen’s ability to overcome host resistance (McDonald & Linde, 2002), adapt to new environments (North et al., 2011), and contribute to invasiveness (Philibert et al., 2011). Sexual reproduction can generate genotypic diversity, facilitate rapid dispersal of beneficial mutations, produce novel combinations of beneficial mutations and purge deleterious ones (Heitman, 2006; Heitman et al., 2013; Calo et al., 2013). Sexual reproduction can substantially speed up the rate of adaptation in diploid species, because sexual populations can produce the fittest homozygous genotype through segregation and mating of heterozygous individuals (Mandegar & Otto, 2007).  Despite the reduction in genetic diversity and frequent loss of sexual reproduction following an introduction, alien plant pathogens are often successful invaders and retain the ability to adapt, overcome host resistance and successfully establish in new habitats. One explanation proposed for this ability to adapt in invasive species is the generation of novel genotypes without sexual recombination. Mitotic recombination (MR) is one mechanism that can uncover previously hidden genetic variation without sexual genetic segregation and could help explain the genetic paradox of invasions. In an asexually reproducing diploid lineage, two rare mutational events must occur before a beneficial mutation can become fixed: one producing a heterozygous carrier and the second, at the same site, converting a heterozygous to a homozygous carrier. In simulation studies, MR increased the spread of beneficial mutations in asexual populations, allowing adaptation to proceed as fast as in sexual populations (Mandegar & Otto, 2007).  130   A second mechanism that could help explain the genetic paradox of invasion is the presence of a two-speed genome in invasive alien pathogens (Raffaele & Kamoun, 2012; Dong et al., 2015). Genes encoding virulence effectors that modulate host plant processes are often associated with genome regions enriched in repetitive sequences and transposable elements, which can serve as a cradle for adaptive evolution (Dong et al., 2015). These genome regions are often gene-sparse, but rich in effector genes in contrast to regions that are gene-dense and rich in core orthologous genes (Haas et al., 2009; Raffaele et al., 2010b,a). These genome regions have been found to evolve faster than the gene dense regions and may have facilitated host jumps and rapid adaptation in pathogens (Haas et al., 2009; Raffaele et al., 2010b,a).  Phytophthora (Oomycetes, Pythiaceae) is a genus of that comprises some of the most destructive invasive plant pathogens affecting crops and forests. Phytophthora infestans (Mont.) de Bary caused one of the worst agricultural disease outbreaks in human history and has been responsible for a famine that resulted in large-scale human migration. Phytophthora ramorum Werres, De Cock & Man in 't Veld is an invasive pathogen with a broad host range that causes sudden oak death, sudden larch death and Ramorum blight on horticultural plants and trees (Rizzo et al., 2002; Brasier & Webber, 2010). Four lineages (EU1, EU2, NA1 and NA2; named after the continent they were first observed in) with low genotypic diversity and apparent lack of sexual reproduction have spread in the USA, Europe and Canada (Prospero et al., 2009; Grünwald et al., 2009, 2012; Vercauteren et al., 2010; Goss et al., 2011b). Since P. ramorum is diploid and heterozygosity is observed within lineages, it is suspected that the ancestral source population 131  reproduced sexually (Goss et al., 2009). However, the absence of recombination in the extant lineages raises questions about adaptation and diversification in P. ramorum.   Mitotic recombination and two-speed genomes have been reported in Phytophthoras (Haas et al., 2009; Raffaele et al., 2010b,a; Lamour et al., 2012; Kasuga et al., 2016). However, the extent to which these factors play a role in genome evolution and adaptation in P. ramorum lineages is not known. In order to determine how genome features and mechanisms drive evolution and adaptation in asexual populations, we sequenced the genomes of 107 P. ramorum samples representing the four known lineages and known geographical range of the pathogen. We found evidence that extensive MR is reshaping the genomic profiles of the asexual lineages generating large-scale non-synonymous changes that could impact the phenotypic outcome. Mitotic recombination can lead to the expression of recessive or new alleles which are subject to selection. We show evidence for a selective sweep of a genomic region rich in putative pathogenicity genes that had undergone an MR event. Mitotic recombination events are associated with transposons and low gene density suggesting that they originate in fast evolving genomic regions. We also observed non-core and lineage specific regions of the genome which were enriched in genes associated with host-pathogen interactions and transposable elements. These non-core regions are characterized by rapid gene loss, a distinct nucleotide composition, and a high repeat and transposon content also suggesting faster evolving genomic regions. A two speed genome appears to drive lineage divergence and population evolution in P. ramorum.   132  Methods Genome resequencing A total of 107 individuals (Table S4.1) were selected for genome resequencing using the Illumina sequencing technology (Appendix B, DNA extraction, genome sequencing, SNP extraction and filtering). We selected representatives of the four lineages: 38 individuals of NA1, 17 individuals of NA2, 46 individuals of EU1 and 6 individuals of EU2. The individuals selected covered the chronological range (1995-2012) and the geographical range (Canada, the United States, and Europe) of the epidemics. Illumina reads were mapped onto the P. ramorum reference genome (version 1.1; DOE Joint Genome Institute, Walnut Creek USA) (Tyler et al., 2006) and a SNP set was obtained for the full set of individuals (Appendix B, DNA extraction, genome sequencing, SNP extraction and filtering). The illumina reads of two representative individuals of each of the four clonal lineages were assembled for inter-lineage comparisons (Appendix B, De novo genome assemblies).   Phylogenetic analysis A neighbor-joining phylogenetic analysis based on Euclidean distances was done with all individuals using the full SNP set to determine inter and intra-lineage diversity patterns in P. ramorum (Appendix B, Phylogenetic analysis). In addition, a set of genes under neutrality determined with Tajima’s D values (Tajima, 1989) was used to reconstruct a maximum likelihood phylogeny to observe evolutionary patterns in the coding genome (Appendix B, Phylogenetic analysis). Divergence time between the lineages was estimated using BEAST v1.7.5 (Drummond et al., 2012) (Appendix B, Divergence time between lineages). 133   Runs of homozygosity We searched for runs of homozygosity (ROH) in each individual and within lineages. Using VCFtools (Danecek et al., 2011), the multi-sample VCF file was separated into independent files for each individual. Using a custom python script (Python 2.6), the VCF files were searched for stretches of homozygous sites uninterrupted by heterozygotes. The length of these stretches was compared between individuals within a lineage to identify those with extensive ROH relative to the population and to identify scaffolds with extensive ROH relative to other scaffolds. For individuals that had extensive ROH, VCFtools (Danecek et al., 2011) was used to extract observed number of heterozygotes and homozygotes for each SNP site for individuals with and without ROH for comparison. The percent of sites homozygous for either allele and the percent of heterozygous sites were calculated for individuals with ROH and representative individuals without ROH for each lineage. Individuals were also checked for genome-level copy number variation including aneuploidy (Appendix B, Genome-level copy number variation).  Effects on protein content from conversion to homozygosity To determine the potential effects of conversion to homozygosity, the genes on the scaffolds with the ROH pattern were reconstructed for each strand of each individual (Appendix B, Phylogenetic analysis). Using a custom Python script, the genes were translated to proteins and the protein translations were compared for individuals with and without ROH to determine the number of amino acid changes between alleles for each group. Gene Ontology (GO) analysis was done on proteins found in the ROH regions using the model-based gene set analysis 134  approach (MGSA) in R (R Core Team, 2017), which analyzes all inferred GO terms at once by embedding them in a Bayesian network and provides summarized views of core biological processes (Bauer et al., 2011).   Identification of a pathogenicity gene hot spot A long run of homozygosity that occurred in scaffold seven in all NA1 and eight EU1 individuals was further investigated to identify protein candidates potentially involved in host-pathogen interactions. The protein set was extracted from the reference genome (version 1.1) (Tyler et al., 2006) and re-annotated using Blast2GO (Conesa et al., 2005). Putative effectors (van Damme et al., 2012) and other proteins potentially involved in pathogenicity (carbohydrate active enzymes, some transporters, peptidases), and transposons (or transposon-like elements) were counted and mapped along the scaffold using GenoPlotR (Guy et al., 2010) in R (R Core Team, 2017). MGSA (Bauer et al., 2011) was used to identify protein sets that were enriched in scaffold seven and a chi-square was used to test if putative effectors were located on scaffold seven more often than expected.  Mitotic recombination breakpoints To determine possible causes or mechanisms for mitotic recombination, we searched for scaffolds with putative recombination breakpoints. Mitotic recombination breakpoints were identified as regions where a switch from heterozygosity to homozygosity occurred. The region around each recombination breakpoint was scanned for gene content using annotations from the P. ramorum reference genome (version 1.1) (Tyler et al., 2006). Intergenic distances (5’ to 5’ 135  and 5’ to 3’) were calculated on the scaffolds with potential breakpoints and graphed on a heatmap using a custom Python script (Saunders et al., 2014). For comparison, ten data sets, each with nine 15 kbp genomic regions with no potential breakpoints, no ROH, and no stretches of ambiguous nucleotides were randomly generated. Gene content and intergenic distances were determined for the 15 kbp regions. An analysis of variance (ANOVA) was used to compare the observed and random data sets using the package RCommander (Fox, 2005) in R (R Core Team, 2017).   Testing for phenotype differences in individuals with ROH To test for phenotypic differences resulting from ROH, we compared the fitness of eight EU1 individuals with ROH in scaffold seven, to eight EU1 individuals without ROH (Table S4.2). We assessed the pathogen’s ability to cause lesions on woody stems of two hosts: Pseudotsuga menziesii (Mirb.) Franco (Douglas-fir) and Larix kaempferi (Lamb.) Carr. (Japanese larch). We also tested the ability to infect detached Rhododendron sp. leaves, and to grow on sapwood agar (Appendix B, Effects of ROH on phenotype).   Core and non-core genomes of Phytophthora ramorum lineages To identify core and lineage-specific genomic regions, we used the mpileup function of SAMtools (Li et al., 2009) to assess coverage depth over 1 kb sliding windows of all 107 re-sequenced individuals over two de novo genome assemblies of each of the four lineages (Appendix B, De novo genome assemblies). A window was considered as non-core when a contiguous region representing more than 90% of the window size was missing in 50% of the 136  individuals of at least one of the four P. ramorum lineages (thresholds of 75% and 99% were also tested in order to obtain a measure of variability in windows coverage [data not shown]). Lineage-specific, or non-core genomic regions shared between only two or three lineages were identified by reciprocal mapping to the de novo assemblies of each lineage. Groups of homologous non-core windows were identified with reciprocal BLASTn searches (E-value < 1e-20) within and between lineages followed by graph clustering using a TCL implementation of the Deep-First Search algorithm (identity cutoff = 40%; overlap cutoff = 100nt) (Kozik et al., 2003; Joly et al., 2010).  A neighbor-joining tree was reconstructed from a distance matrix of similarity based on non-core genome regions (as in Appendix B, Divergence time between lineages: Phylogenetic analysis of one-to- one orthologs). The extent of gene gain or loss was determined by attributing costs to gain and loss events and minimizing total cost (Maximum parsimony criterion, (Mirkin et al., 2003)). GC content, transposon and repeat content were compared for the non-core and core genome.  Lineage-specific proteins and evolution of gene family size  Genome assemblies obtained for the four P. ramorum lineages and P. lateralis (CBS_168.42) (Feau et al., 2016) were annotated using AUGUSTUS ver. 2.7 (Keller et al., 2011) trained with models from P. sojae, P. infestans and P. ramorum available on the Joint Genome Institute (JGI) server (Grigoriev et al., 2012). The protein models obtained were combined with those of five Phytophthora genomes and submitted to a clustering process with OrthoMCL (Li et al., 2003) 137  (Appendix B, OrthoMCL analysis and evolution of gene family size). The resulting OrthoMCL clusters were filtered to minimize the confounding effect of truncated proteins and gene mis-predictions and CAFE (Computational Analysis of gene Family Evolution)(De Bie et al., 2006) was used to infer gene family expansions/contractions (Appendix B, OrthoMCL analysis and evolution of gene family size). Genes encoding secreted proteins were predicted using the pipeline described in (Dhillon et al., 2015). Putative effectors were identified using EffectorP (Sperschneider et al., 2016).   Adaptation in the effector repertoire of Phytophthora ramorum lineages RxLR and Crinkler effector genes were extracted from the de novo genome assemblies of each of the four lineages (Appendix B, Genes encoding effectors). Effector groups were clustered with OrthoMCL (Li et al., 2003), aligned and dN/dS was estimated with Codeml implemented in PAMLV4.0 (Yang & Nielsen, 1998) as described in Appendix B (Genes encoding effectors). Observed dN/dS were compared to dN/dS calculated on a random set of genes and the set of sequence homologs of the 458 proteins of the core eukaryotic CEGMA dataset (Parra et al., 2007) from the P. ramorum reference genome (Appendix B, Genes encoding effectors).   Results Mitotic recombination generates diversity and novel phenotypes Extensive genotypic diversity in clonal P. ramorum lineages. The four lineages of P. ramorum are well resolved and divergent based on a neighbor-joining tree (Fig. 4.1) using 485 327 bi-138  allelic SNPs (Appendix B, DNA extraction, genome sequencing, SNP extraction and filtering; Table S4.3) and a maximum likelihood phylogenetic analysis (Appendix B, Phylogenetic analysis; Fig. S4.5). We estimate that these lineages diverged approximately between 0.75 million years ago (MYA) for the split between EU1 and NA1 and 1.3 MYA for the split between EU2 and the other lineages (Appendix B, Divergence time between lineages). We observed an unexpected amount of intra-lineage diversity; several individuals were divergent from the main group including a cluster of eight individuals in EU1 (Fig. 4.1; Fig S4.5). These individuals were characterized by a high number of loci that were homozygous in one or more individuals but heterozygous in the rest of the members of the lineage (Table S4.2 and S4.3). These homozygous sites were clustered in runs of homozygosity (ROH) ranging in size from 60 to 339 Kb and were most abundant within the EU1 lineage (30% of the individuals) (Appendix B, Effects of ROH on genotype). The ROHs were not randomly distributed throughout the genome; rather entire scaffolds or large portions of scaffolds were entirely homozygous in some individuals, interrupted by only a few heterozygous loci (Fig. S4.6). In addition, four individuals also exhibited plasticity in chromosome numbers, potentially contributing to the observed diversity (Appendix B, Genome-level copy number variation, Table S4.4 and Fig. S4.7). One NA1 individual with CCNV was amongst the more diverged individuals in the neighbor-joining tree (Fig. 4.1).   Protein content differences within clonal P.ramorum lineages. The ROHs observed have the potential to alter the protein content within clonal lineages if individuals with ROH lost one of two haplotypes with non-synonymous differences. We observed this pattern in 52% of the 5 172 genes located within scaffolds that contained ROHs (Appendix B, Effects of ROH on genotype; 139  Table S4.5). In 21% of those genes, there were five or more non-synonymous heterozygous nucleotide positions in individuals without ROH compared to those with ROH and up to 35 amino acid changes were observed in some genes (Table S4.5). Such large scale differences have the potential to generate substantial phenotypic changes in individuals with ROH.   ROH can uncover recessive mutations and be subject to selection. Scaffold seven had the longest ROH, covering approximately 650 Kb. All 38 NA1 individuals have this ROH (Fig. 4.2A) suggesting a selective sweep. Interestingly eight (out of 38) EU1 individuals possess the alternate haplotype suggesting that a similar selection event could be occurring in that lineage. This ROH was absent in all NA2 and EU2 individuals. The secretome differed on scaffold seven compared to the rest of the genome: the highest percentage of secreted proteins occurred on scaffold seven (24.6%) when compared to the 25 largest scaffolds (Fig. S4.8; chi-square = 45.127, p < 0.001). Fifteen percent of the genes on that scaffold were annotated as putative effectors, 5% of which encode necrosis inducing proteins or NPP1s; this represents the highest proportion of NPP1s on any scaffold, and is significantly higher than expected across the genome (chi-square = 79.261, p < 0.001). The NPP1 encoding genes occur in a cluster within the homozygous region, as do most of the putative effectors. Scaffold seven also comprises several genes that act in some plant pathogens as virulence factors such as carbohydrate active enzymes, peptidases and sugar transporters (Appendix B, Effects of ROH on genotype, Fig. S4.8, Table S4.6).   Transposable elements and low gene density associated with mitotic recombination. Mitotic recombination breakpoints were identified as locations where there was a transition from 140  heterozygosity to homozygosity in the regions affected by ROH. Transposable elements (TEs) were found in close proximity to mitotic recombination breakpoints in seven out of nine instances, which was four times more frequent compared with ten randomly generated datasets (Fig. 4.3, Table S4.7). The mitotic recombination breakpoints occurred in regions with a lower gene density than the rest of the genome. In scaffolds with mitotic recombination breakpoints, the intergenic distances measured as the 3’ flanking intergenic region (FIR) and the 5’ FIR were significantly larger than the intergenic distances in random sets of scaffolds (Table S4.7). Five percent of the genes had a 5’ FIR greater than 1 x 104, and seven percent of the genes had a 3’ FIR greater than 1 x 104 (Fig. 4.3), compared to two to four percent for the 5’ FIR and three to six percent for the 3’ FIR in the random data sets (Fig. 4.3, Table S4.7).  Phenotypic characterization of individuals affected by ROH. Lesion size on larch and Douglas-fir caused by EU1 individuals with ROH on scaffold seven were larger than lesions caused by EU1 individuals without ROH. Individuals with ROH produced longer lesions on larch log sections (average 57.4 mm ± versus 38.0 mm ±; Fig. S4.9 and S4.10; Kruskal-Wallis chi-square test= 4.412; p = 0.036) and slightly longer lesions on Douglas-fir log sections four weeks post-inoculation than those produced by non-ROH individuals (average 56.6 mm ± versus 54.7 mm; p = 0.753). The trends were reversed on both host species eight weeks post inoculation (Appendix B, Effects of ROH on phenotype).  141  The core and non-core genomes differ markedly among Phytophthora ramorum lineages  Reduced non-core genome in EU2. The four lineages of P. ramorum share a core genome representing 98.2% to 99.2% of the full content, leaving a non-core ranging from 0.34 Mbp (EU2) to 0.90 Mbp (NA2) (Fig. 4.4A). The non-core genome shared between EU2 and the other lineages was smaller (0.007 Mbp) than the non-core genome shared among the other lineages (1.7 Mbp; Fig. 4.4B). This corresponds to 0.16% to 0.37% of the EU2 non-core genome that was not found in the genomes of the three other lineages and 1.24% to 1.68% of the EU1, NA1 and NA2 non-core genomes that is missing in the EU2 genome (Fig. 4.4C). The neighbor-joining tree based on lineage similarity of non-core regions was topologically incongruent with the P. ramorum lineage phylogeny reconstructed from the core gene set (Fig. 4.4C and S4.11). This strong discordance between the two phylogenies suggests differences in the rate and extent of genome fluctuation (regions gain and losses) among lineages in the non-core genome, or a two-speed genome. We estimated that most of the fluctuations were the result of region losses in EU2 (68.6%), whereas they resulted from gains in the other lineages (Fig. 4.4C).   Evidence for a two-speed genome. The nucleotide and gene composition of the non-core genome in the four P. ramorum lineages were clearly distinct from the rest of the genome. The non-core region had a significantly lower G+C content (49.4% for non-core vs. 54.0% for core averaged over the four lineages; paired t-test value = 25.79, P < 0.001), a higher repeat content (46.3% for non-core vs. 15.5% for core averaged over the four lineages) coupled with a significant enrichment in TE-like gene sequences (χ2 = 19.3, P < 0.001), four to seven time less 142  genes per Mb of DNA, and a higher number of unannotated genes (Table S4.8). Genes in the non-core genome were shorter and had a lower G+C content than those from the core genome (average = 54.9% vs. 57.9%; t = 5.14, P = 0.01) and genes found in repetitive regions (57.3%; t = 3.87, P < 0.05) (Table S4.8). A large number of genes showed strong bias in codon usage (from 82% in NA2 to 100% in EU2), reflecting a clear preference towards codons ending with T or A nucleotides (Fig. S4.12). In addition, one third to half of the genes found in the non-core regions (35% in EU1 and EU2, 41% in NA1 and 50% in NA2) share paralogs with genes in the core genome and show high mutation loads making them likely degenerated copies (Appendix B, Core and non-core genomes; Fig. S4.13).   Among the protein-coding genes identified in non-core genome regions (17 in EU2 to 44 in NA2), we did not find significant enrichment in genes encoding secreted proteins (χ2 = 0.15, p = 0.98) or effectors. However, we found presence/absence patterns for effector-like proteins that are part of the RxLR effector family (5 RxLR gene models; Table S4.9) and for genes with functions related to degradation of the plant cell wall (proteases and ester hydrolases; Table S4.9). Remarkably, none of the 17 non-core protein models predicted in EU2 encoded effector-like proteins (Table S4.9).   The non-core proteome of P. ramorum is enriched in plant cell wall modification enzymes. We identified a core proteome of 10,197 clusters representing 44,295 shared proteins between the four P. ramorum lineages, that have likely been conserved since the lineages split (Table S4.10). About 5.3% (541 clusters) of this core proteome is not shared with any other Phytophthora species and represents P. ramorum unique gene families (Table S4.10). We 143  identified only slight protein family expansion within each P. ramorum lineage followed by rapid divergence, resulting in proportions of 2.5 to 3.2% lineage-specific proteome (i.e. not shared with any other Phytophthora species or between P. ramorum-lineages) (Fig. S4.11). GO terms associated with peptidase activities (lineages EU1, EU2 and NA1 ) and pectin and glucan modification (EU1, NA1 and NA2), were enriched in the unique protein models suggesting that rapidly diverging proteins unique to each of the four P. ramorum lineages are enriched in functions related to plant cell wall modification and/or degradation (Table S4.11).  Evidence of host driven adaptation in Phytophthora ramorum lineages  Phytophthora ramorum effectors under positive selection. Host driven adaptation between the lineages was suggested by positive selection in 20% of the RxLR and 38% of the Crinkler effector genes (CRN) compared to only 1% of the CEGMA gene set and only 3% in a random set of genes in the P. ramorum genomes (Fig. S4.14). CEGMA (72%) and random genes (70%) were under strong negative selection (dN/dS < 0.3). The distribution of dN/dS ratios was significantly different between the effector and the CEGMA gene sets (F-value = 376.93, p < 0.001) as well as the between the effector and the random gene sets (F-value = 298.82, p < 0.001). We identified one CRN family that likely diversified within the evolutionary history of the P. ramorum lineages through duplication, recombination and episodes of accelerated nucleotidic evolution (Appendix B, Genes encoding effectors; Fig. S4.15).   144  Discussion  Phytophthora ramorum provides a classic example of a successful emerging pathogen with a small effective population size, limited to asexual reproduction. The clonal lineages are without apparent meiotic recombination either due to the uneven distribution of the mating types or the presence of reproductive barriers (Brasier & Kirk, 2001; Ivors et al., 2006; Goss et al., 2009). We highlight two mechanisms that can generate diversity and adaptation in these asexual lineages. First, we observed a large amount of variation within lineages that could impact the pathogen’s ability to adapt and strong evidence that mitotic recombination is at the origin of this variation, potentially affecting hundreds of genes simultaneously. We found compelling evidence that a sweep in one long ROH has affected a genome region enriched in putative pathogenicity-related genes. Remarkably, this appears to have occurred in two independent events in NA1 and EU1. Secondly, we observed accelerated evolution in lineage specific regions affecting genes putatively involved in host-pathogen interactions. These lineage-specific regions have a distinct nucleotidic composition, are repeat-rich, gene-poor and comprise a large number of unannotated proteins suggesting rapid evolution. It is likely that this two-speed genome is contributing to phenotypic differences observed among the clonal lineages, as suggested for other Phytophthora pathosystems (Haas et al., 2009; Raffaele et al., 2010a; Dong et al., 2015).  Mitotic recombination appears to be a dominant mechanism driving evolution in the asexual P. ramorum lineages. This process is expected to cause a loss of heterozygosity in all loci distal from the breakpoint creating runs of homozygosity (Goodwin, 1997; Tischfield, 1997; Llorente et al., 2008). ROHs resulting from mitotic recombination are much longer than meiotic 145  recombination, and not uniformly distributed throughout the genome, the pattern observed in this study (Lee et al., 2009; Magwene et al., 2011). Mitotic recombination has been suggested in P. ramorum (Ivors et al., 2006; Goss et al., 2009; Kasuga et al., 2016) and in other Phytophthora species (Chamnanpunt et al., 2001; Dobrowolski et al., 2003; Lamour et al., 2012; Hu et al., 2013); however the extent of the effect shown here is much larger than previously reported. LOH was suggested as a contributing factor in generating diversity and adaptation in the broad-host-range pathogen Phytophthora capsici (Lamour et al., 2012). LOH was found in progeny from crosses as well and in clonal individuals affecting up to 36.5% of the genome; the LOH affected more than 11 000 sites and was associated with changes in pathogenicity and in mating type. However only 17% of those sites were in coding regions and only 264 genes were predicted to be affected by non-synonymous changes (Lamour et al., 2012), a far lower number than the 5172 genes with non-synonymous changes that were predicted in our study. In P. sojae, mitotic gene conversion occurred in a genomic region coding for an avirulence/virulence gene potentially leading to virulence differences in new pathogen races (Chamnanpunt et al., 2001). In P. ramorum, mating type switches observed in EU1 individuals could be explained by mitotic gene conversion (Vercauteren et al., 2011). Mitotic recombination is also observed in fungi and has been associated with increases in diversity or with adaptation to environmental changes. For example, mitotic recombination resulting in loss of heterozygosity was the cause of high genotypic variation in the clonal species Batrachochytrium dendrobatidis responsible for the amphibians chytridiomycosis epidemic (James et al., 2009). In Candida albicans LOH rates increase proportionally to exposure to oxidative stress, heat stress, and antifungal drugs (Forche et al., 2011). In Saccharomyces cerevesiae a trade-off between outcrossing and mitotic recombination seems to play a role in shaping genome architecture in response to nutrient stress 146  (Magwene et al., 2011). The variation generated by mitotic recombination in P. ramorum could be a response to host or environmental stresses and could influence adaptation to novel environments. The high number of genes with nonsynonymous changes in regions affected by ROH in P. ramorum could be the result of phenotypes that were selected for on different hosts, and are retained in the population.  One of the proposed advantages of mitotic recombination is that it facilitates the rapid evolution of asexual eukaryotic microorganisms facing changing environments (Forche et al., 2011; Magwene et al., 2011). Conversion of genomic regions to homozygosity may result in the expression of new or recessive alleles, and can improve the efficiency of selection on both beneficial and deleterious alleles (Mandegar & Otto, 2007). This can be particularly important for a pathogen that is reproducing clonally as new mutations or recessive alleles can remain masked by dominant alleles. Mandegar & Otto (2007) theorized that mitotic recombination can drastically increase the rate of adaptation in asexual populations by reducing the time between the emergence of beneficial mutations in a heterozygote and the appearance of a mutant homozygote; under some scenarios adaptation can occur at the same rate as in sexual populations, or even faster if the rate of mitotic recombination is very high. In lab experiments with diploid isogenic individuals of Aspergillus nidulans, increased rates of evolution were found in individuals that built up recessive mutations in diploid genomes, which were later expressed after individuals reverted to a haploid state (Schoustra et al., 2007). A similar result was found with an individual which remained diploid, but had likely undergone a loss of heterozygosity in part of the genome, likely due to mitotic crossing over or non-disjunction.   147  We identified a genome region that appears to be a mitotic recombination hotspot in P. ramorum. ROH on scaffold seven affected all members of the NA1 lineage. The same region was affected in eight individuals in EU1 lineage, but the opposite haplotype was present. Two separate ROH events affecting the same genomic region suggests that genes on scaffold seven are likely under selection and that resulting phenotypes could have a selective advantage. To support this hypothesis, we showed that scaffold seven has a higher number of pathogenicity-related genes than expected and that LOH fixed non-synonymous mutations for a substantial proportion of these genes. Changes resulting from ROH in this genomic region could have direct effects in the interactions of the pathogen with its hosts. As P. ramorum is a generalist pathogen with over 100 host species, changes to the gene content or alleles could provide advantages in response to different environments. The slightly larger lesions caused by individuals with ROH in the early stages of infection on Japanese larch and Douglas-fir indicates faster colonization suggesting that these individuals could have a fitness advantage in overcoming host defences. There was no indication of loss of fitness associated with ROH on infected Rhododendron leaves or in artificial media. Further investigations into other host interactions or on sporulation efficiency could uncover other phenotypic differences associated with ROH.  Genotypic diversity and effective population sizes are important features for the successful establishment, persistence and adaptability of an invading population (Sakai et al., 2001; Forsman, 2014; Gladieux et al., 2015; Blackburn et al., 2015). Rapid evolution in plant pathogens may be related to genome characteristics such as high transposon content and activity, genome size, or the physical arrangement of genes within the genome (Haas et al., 2009; Raffaele et al., 2010b,a; Judelson, 2012; Raffaele & Kamoun, 2012; Jiang & Tyler, 2012; Moller 148  & Stukenbrock, 2017). We hypothesize that the effective population size of P. ramorum has increased due to mitotic recombination triggered by TE activity and low gene density. The higher frequency of mitotic recombination breakpoints in regions with low gene density and physically associated with transposons suggests that these elements could be associated with a mechanism generating genotypic diversity in P. ramorum populations. In the asexual plant pathogen Verticillium dahliae, pathogenicity has been hypothesized to evolve through recombination between highly similar copies of retrotransposons which may generate diversity by facilitating mitotic crossover events and creating new beneficial genetic combinations (de Jonge et al., 2013; Faino et al., 2016). TEs have been suggested as the driving force behind adaptive evolution of V. dahliae (Faino et al., 2016). TEs have been associated with large chromosomal rearrangements in the wheat pathogen Zymoseptoria tritici also suggesting their in role adaptation (Plissonneau & Stürchler, 2016; Moller & Stukenbrock, 2017).  Non-core genome regions in pathogenic micro-organisms generally contain effectors that may be involved in ecological adaptation (Hatta et al., 2002; de Jonge et al., 2013). These effectors are often adapted specifically to interact with host proteins leading to host or niche specialization and often result in speciation. As a consequence, the selective pressures acting on effectors can be especially efficient since having the wrong set of genes has a high fitness cost (Giraud et al., 2010). Effectors that are no longer needed following a host jump are either eliminated or degenerated (Dong et al., 2015). The positive selection and evolutionary history in effector genes in the P. ramorum lineages are further evidence of a two-speed genome where genes and regions of the genome involved in host-pathogen interactions are evolving faster than the rest of the genome. In addition, the P. ramorum EU2 lineage showed a striking reduction in gene content, 149  particularly in genes putatively involved in host-pathogen interactions. Although the non-core genomic regions were not significantly enriched in putative effectors, most of those found were missing in EU2 indicative of niche differentiation and speciation. Specialization of EU2 to Japanese larch (Larix kaempferi) (Brasier & Webber, 2010; Van Poucke et al., 2012) and adaptive differences observed for this lineage could be related to the reduced non-core genome. Localization of race specific effectors in lineage-specific regions was suggested as a contributing factor in niche adaptation and virulence in Verticillium dahliae (de Jonge et al., 2013). Lineage-specific regions were also not enriched in effectors compared to the non-core genome; however, plant-induced genes were upregulated in the lineage specific transcriptome (de Jonge et al., 2013). Genes involved in pathogenicity are often found clustered together in less conserved, gene poor regions of the genome such as subtelomeric regions usually enriched in repeated elements or transposons (Haas et al., 2009; Raffaele et al., 2010a; Rouxel et al., 2011; Bopp et al., 2013; Dutheil et al., 2016). In our study, not only were non-core regions enriched in genes associated with plant-pathogen interactions, they were also enriched in TE-like gene sequences, some of which were found in tandem with effector genes. This suggests that the non-core genome plays an important role in genome evolution and adaptation, and in shaping differences among the lineages. TEs were also associated with putative effectors in Leptosphaeria maculans, and lineage-specific effector genes in Zymoseptoria tritici (Rouxel et al., 2011; Plissonneau & Stürchler, 2016). In Magnaporthe oryzae, transposons were linked to multiple translocations of an effector in different strains of the pathogen (Chuma et al., 2011).  Several species of fungi possess genome defense mechanisms thought to limit the accumulation of transposable elements. The best known is repeat-induced point mutation or RIP. The RIP 150  process was described in Ascomycete fungi and consists of inactivating repeated sequences by introducing point mutations in CpG sites, resulting in mutational loads favoring GC to AT mutational changes (Galagan et al., 2003; Galagan & Selker, 2004). RIP has never been demonstrated in Phytophthora species; however we found characteristics in the non-core genome that resembled a genome defense mechanism similar to RIP. The non-core genome was clearly distinct from the core genome in the P. ramorum lineages with a lower GC content, strong bias in codon usage favoring codons ending in AT, a high number of degenerated paralogs and enrichment of transposons or transposon like genes. Similarly, sequences exclusive to individuals found in Z. tritici also showed a mutational load likely resulting from the effect of a genome defense mechanism similar to RIP (Plissonneau & Stürchler, 2016). The existence of a genome defense mechanism that can deactivate transposable elements would restrict genome reshuffling during low-stress periods or could potentially lead to a more stable genome after TE activation during periods of stress or change.   Pathogen populations limited to asexual reproduction face a long term threat of extinction if they cannot adapt to changes in the environment or host. Mitotic recombination could promote the long term existence of diploid asexuals by increasing the rate of adaptation (Mandegar & Otto, 2007). Invading pathogens can carry some of the parental population’s genetic diversity in the genome through heterozygosity. Mitotic recombination can release the diversity back into the population thereby increasing fitness as it will be able to respond to increased selection pressure. Several Phytophthora species are successful and adaptable invaders. Mitotic recombination may be an important contributing factor to the adaptability of these species.   151   Figure 4.1: Evolutionary relationships in Phytophthora ramorum. Neighbor-joining tree constructed from Euclidean distances between individuals using 485 327 bi-allelic SNPs. EU1 individuals with mitotic recombination patterns shown.    EU1 EU1 with individual mitotic recombination events EU1 with mitotic recombination affecting scaffold 7 EU2 NA2 NA1 80.ROH CCNV ROH and CCNV 152   Figure 4.2: Comparison of heterozygote density, gene content and gene location in scaffold seven of Phytophthora ramorum. (A) Heterozygote density of representative isolates of each lineage. ROH pattern displayed for NA1 is observed in all 38 NA1 individuals. EU1 individuals possessed two distinct patterns on scaffold seven: an ROH pattern shared by eight individuals and a normal pattern in the remaining 38 individuals. (B) Gene map showing approximate location of genes along scaffold seven. (C) Piechart showing the percentage of each category of putative plant pathogenicity genes. ABC transporter CAZY Putative effector NPP1-effector Other Peptidase Sugar transporter Transposon like 10%5%1%10%5%5%2%62%Gene ContentPR_07_192_NA1Pr_518_NA1PR_07_101_NA104_0002_NA1PR_09_167_NA1PR_09_065_NA1PR_01_004_NA1Pr_1546_NA1P1559_EU1P1614_EU1P1959_EU1P1563_EU1P2688_EU1P1549_EU1P1562_EU1P2600_EU1BBA_26_02_EU133_EU104_17031_EU1PR_08_028_EU1P1856_EU111_301_EU1P2561_EU2P2111_EU2P2461_EU2P2586_EU2P2460_EU2P2566_EU204_20470_NA2PR_10_120_NA211_1291_NA210_3885_NA208_0688_NA206_0012_NA2NA1 EU1 EU1 EU2 NA2 Scaffold 7 heterozygote density A B 153   Figure 4.3: Genome characteristics around mitotic recombination breakpoints. (A) Transposable element content in the nine regions with ROH breakpoints in comparison to random regions of ten random datasets; (B) intergenic distances in nine scaffolds with ROH breakpoints compared to one of the random datasets with no recombination breakpoints; (C) intergenic distances represented by 5’ and 3’ distances between genes.    i-1 Protein i i+1 5’ -FIR i-1 Protein i i+1 3’ -FIR 0123456781 2 3 4 5 6 7 8 9 10 11Random datasets Regions with breakpoints Number of TEs or TE-like  A C Random scaffolds, no runs of homozygosity Scaffolds with ROH breakpoints 2% 3% 5% 7% B 154   Figure 4.4: Core and non-core genomes predicted in the four Phytophthora ramorum lineages.  (A) Proportion of core (present in all lineages) and non-core (missing in at least 50% of the individuals of a lineage) genomes; values above the bars are the proportion of core genome relatively to the size of the reference genome. (B) Number of non-core 100bp windows unique and shared between lineages. (C) Non-core regions shared between lineages. Neighbor Joining-tree based on genome similarity in terms of non-core regions (presence/absence of 100bp non-core region in the four lineages); pie-charts are proportional to the number of non-core regions that changed in each genome as predicted by CAFE, with expansions in blue and contractions in green. Proportions of lineage non-core genome (rows) that are found in the genomes of the other three lineages (columns); e.g. top right corner indicates that 1.68% of the non-core genome of NA2 was found in EU2 and bottom left corner shows that 0.37% of the non-core genome of EU2 was found in NA2.    155  Chapter 5: Conclusions  The objective of this work was to look at the influence of anthropogenic activities on oomycete communities, and on the pathogenic genus Phytophthora. We hypothesized that anthropogenic activities facilitate the introduction of alien species, the emergence of pathogens, and the post-introduction secondary migration of alien species to natural environments. Secondly, we hypothesized that season would affect diversity, as well as the habitat (water versus soil) and the method used to survey the community. For this reason, we chose to survey both soil and water in the spring and fall to obtain a more inclusive sampling of the community. Furthermore, we employed multiple approaches (DNA meta-barcoding soil, lab baiting soil and water, and stream baiting) to survey the population for the same reason. These additional approaches allowed us to compare seasonality in oomycete and Phytophthora communities as well as to compare the results obtained using multiple sampling methods. Lastly, we wanted to determine how invasive Phytophthora species overcome the invasion paradox after they are introduced into new environments. We hypothesized that Phytophthora species are highly adaptable due to a rapidly evolving genome, and that we would find genomic signatures of adaptation in an invasive population.   The first objective was addressed by looking on a broad scale at the oomycetes. There have been few studies characterizing oomycete communities in natural environments and none have previously been conducted in B.C. Furthermore, few studies have investigated Phytophthora communities in B.C. The impact of anthropogenic activities on these communities was an 156  important aspect of this study given the increasing number of emerging diseases believed to be linked to globalization. There are concerns regarding the potential of Phytophthora species being introduced or spread by human activities due to the recent emergence of a few devastating forest pathogens. In chapter 2, we compared natural oomycete communities to those in urban and interface environments to determine if those with varying levels of anthropogenic influence show a signature in the overall oomycete community. In chapter 3, we looked more in depth at the Phytophthora community, expanding on the methodologies used, to revisit that question, to assess the impact of different anthropogenic activities on species profiles, and to determine if anthropogenic activities increase the opportunity for inter species hybridizations.  Our results showed that anthropogenic activities influenced both urban oomycete as well as Phytophthora communities by increasing species diversity. This increase in oomycete and Phytophthora species diversity is likely caused in large part by the introduction of new species via the nursery and agricultural industry. Communities in urban sites were more diverse and several species were more abundant than in natural and interface sites, suggesting that anthropogenic activities also produce an environment conducive to population growth and spread. Furthermore, our results suggest that interface environments act as a bridge between urban and natural environments, and that there is a directional migration from urban sites to interface sites that could eventually lead to establishment of alien species in natural sites. The Phytophthora genus made up a relatively small portion of the oomycete community which was dominated by Pythium. However, the Phytophthora community was still quite diverse and the introduction of even just a single species that is not indigenous can have a devastating outcome.   157  In chapter 3 we were able to refine the analysis by focusing on the Phytophthora community, an important group because of their worldwide impact on crop and forest health. We used a culture-dependant approach coupled with DNA barcoding to allow for a deeper characterization of the taxa, and the discovery of putative hybrids. We observed again increased diversity in both urban and interface communities and more shared species between interface and urban sites. This convergence of our findings in chapters 2 and 3, using a completely different methodology, brings strong support to our initial hypothesis of the impact of anthropogenic activities on these communities. In addition, we identified several species in low frequency in urban environments that could represent recent introductions from other areas suggested by data on native ranges. Furthermore, we found two invasive pathogens, P. ramorum and P. lateralis, previously determined to be introduced in North America, in all site types suggesting that both introductions of new species and spread of invasive pathogens into natural environments is occurring.   Our sampling also allowed us to compare the impact of the different anthropogenic activities. Agricultural activities appear to create diverse Phytophthora communities that could be facilitating increased interspecific hybridization. Furthermore, agricultural activities may be supporting population growth of particular species which may lead to disease emergence. Two agricultural areas had an increased abundance of two separate putative hybrid species. At least one of these species could have spread through the global plant trade. Alternatively, one of the parental species could be spreading through global plant trade and hybridizing with local species in the locations it is introduced in. The second putative hybrid could also be introduced and emerging as suggested by the absence of any putative parental species.  158  This study has likely underestimated both the Oomycete and Phytophthora diversity present in British Columbia. Each method employed has limitations; however using multiple methods over multiple time frames has resulted in one of the most comprehensive surveys of Phytophthora communities. Continued work could include more sampling timeframes, sampling the canopy and foliage in addition to water and soil and by using DNA meta-barcoding on all sample types in addition to traditional culture based methods. Further work using DNA meta-barcoding could utilize both oomycete and Phytophthora primers. Genus specific Phytophthora primers would allow us to detect more rare species and provide a higher confidence in the results for low frequency species. Furthermore, more intensive sampling in different forest types would help to develop a better understanding of natural oomycete communities. Sampling both the oomycete diversity, as well as the above ground plant community would provide insight into ecosystem processes. How plant species might cultivate pathogen or microbial communities, and how microbial communities might influence forest diversity and stand structure.    In chapter 4 we looked more in depth at an invasive Phytophthora pathogen that has been introduced through the nursery trade. Phytophthora ramorum has emerged several times in the USA, Europe and Canada. Despite being restricted to clonal reproduction and a reduced gene pool, in each region where the pathogen has emerged, it has successfully invaded the new territory. We suspected that the unique genome features of Phytophthora pathogens enabled P. ramorum populations to rapidly evolve and overcome the invasion paradox. Our results strongly support the hypothesis that P. ramorum populations are diversifying, which is likely being driven by rapidly evolving, gene sparse and transposon rich genomic regions causing mitotic recombination. In addition, we also found evidence that genome structure is driving the 159  divergence between the lineages, possibly leading to observed niche differences, particularly in the EU2 lineage which showed a striking pattern of gene loss compared to the other three lineages. These findings suggest that the hypothesized two speed genome in Phytophthora species could also drive population evolution and species evolution.   This research is a significant contribution to our knowledge of oomycete and Phytophthora communities and how those communities are impacted by anthropogenic activities. Studies on oomycete communities are in their infancy and we are only beginning to understand community structure and possible roles of the different species in natural environments. A growing body of evidence is suggesting that native Phytophthora species are present in all habitats and many do not cause widespread disease. Many oomycetes including some Phytophthora likely play a role as saprotrophs. Plant pathogenic oomycetes likely contribute to succession patterns, maintaining diversity, or creating stand level structure in healthy forests. Our study will help in the global understanding of this group of microorganisms, and how we unintentionally facilitate the emergence of new forest diseases. Devastating forest diseases are often caused by alien invasive pathogens and understanding how these pathogens move into natural environments will allow us to improve our practices to reduce international spread, as well as to reduce or change activities that spread pathogens from urban to natural environments. This could include improved sanitation methods for exports in trade, improved inspection of material crossing borders including molecular testing for invasive pathogens, and pathogen tracking to uncover dispersal routes and stop the movement of invasive species. At a local scale, pathogen spread can be limited by reducing runoff and treating waste water in agricultural and nursery settings, as well as treating recycled water before reusing in nursery settings. In addition, public education programs 160  could highlight the effects of invasive species and encourage the public to properly dispose of garden waste and other plant material. Programs could teach the public not to plant alien species in natural environments.   Our study also provides new knowledge and a baseline to BC and Canada regarding the Phytophthora species present, including invasive species and species that are likely introduced. This is important information for forest and pest managers who may be able to use it for further monitoring of invasive species and possible next steps including eradication if necessary.   In addition, this work has provided valuable information on methods to monitor and track invasive pathogens or new emerging pathogens including meta-barcoding which can be used as a first step to monitor a wide range of areas and a large number of samples. Follow up monitoring can be done in locations where invasive pathogens are detected with more thorough methods to retrieve living cultures of pathogens to confirm their presence. The results of this study have already been shared with the Ministry of Forests, Land and Natural Resources who have used our findings to initiate a Phytophthora monitoring program including monitoring in areas where invasive alien species were detected.   This work has also shown that mechanisms important in the evolution of Phytophthora species are facilitating rapid adaptation in P. ramorum populations. This not only contributes to our scientific understanding of invasive pathogens, population evolution and Phytophthora, but it provides valuable information that may help in the management of the disease. One follow-up study already planned is to conduct a broader assessment of the impact of genomic structural 161  rearrangements on pathogenicity and other fitness traits. As this pathogen continues to invade and emerge in new places, a better understanding of how it adapts will support decisions and programs to stop spread. This may include the creation of better tracking and detection tools or eradication programs where feasible. Already a team of scientists funded by a Genome Canada grant are following this up by creating new genomic tools for monitoring and risk assessment of Phytophthora species. That work will be greatly facilitated by the work presented in this thesis.   162  References  Abascal F, Zardoya R, Posada D. 2005. ProtTest: selection of best-fit models of protein evolution. Bioinformatics 21: 2104–5. Aguayo J, Halkett F, Husson C, Nagy ZÁ, Szigethy A, Bakonyi J, Frey P, Marçais B. 2016. Genetic diversity and origins of the homoploid-type hybrid Phytophthora ×alni. Applied and Environmental Microbiology 82: 7142–7153. Alamouti SM, Haridas S, Feau N, Robertson G, Bohlmann J, Breuil C. 2014. Comparative genomics of the pine pathogens and beetle symbionts in the genus Grosmannia. Molecular biology and evolution 31: 1454–1474. Albornoz FE, Burgess TI, Lambers H, Etchells H. 2017. Native soilborne pathogens equalize differences in competitive ability between plants of contrasting nutrient-acquisition strategies. Journal of Ecology 105: 549–557. Allain-Boulé N, Tweddell R, Mazzola M, Bélanger R, André Lévesque C. 2004. Pythium attrantheridium sp. nov.: Taxonomy and comparison with related species. Mycological Research 108: 795–805. Alston KP, Richardson DM. 2006. The roles of habitat features, disturbance, and distance from putative source populations in structuring alien plant invasions at the urban/wildland interface on the Cape Peninsula, South Africa. Biological Conservation 132: 183–198. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. Journal of molecular biology 215: 403–410. Anderson PK, Cunningham AA, Patel NG, Morales FJ, Epstein PR, Daszak P. 2017. Emerging infectious diseases of plants: Pathogen pollution, climate change and agrotechnology drivers. Trends in Ecology & Evolution 19: 535–544. Anisimova M, Gascuel O, Sullivan J. 2006. Approximate likelihood-ratio test for branches: A fast, accurate, and powerful alternative. Systematic Biology 55: 539–552. Arcate JM, Karp MA, Nelson EB. 2006. Diversity of Peronosporomycete (Oomycete) communities associated with the rhizosphere of different plant species. Microbial Ecology 51: 36–50. Atkinson RG. 1965. Phytophthora species inciting root rot of Chamaecyparis lawsoniana and other ornamentals in coastal British Columbia. Canadian Journal of Botany 43: 1471–1475. Bailey TL, Williams N, Misleh C, Li WW. 2006. MEME: discovering and analyzing DNA and protein sequence motifs. Nucleic Acids Research 34: W369–W373. Balci Y, Balci S, Eggers J, Macdonald WL, Plant D, Sciences S, Gottschalk KW, Service UF. 2007. Phytophthora spp . associated with forest soils in Eastern and North-Central U . S . oak ecosystems. Plant Disease 91: 705–710. Baldauf SL. 2003. The deep roots of Eukaryotes. Science 300: 1703–1706. 163  De Bary A. 1876. Researches into the nature of the potato fungus Phytophthora infestans. Journal of the Royal Agricultural Society of England 212: 239–269. Bauer S, Robinson PN, Gagneur J. 2011. Model-based gene set analysis for Bioconductor. Bioinformatics 27: 1882–3. Becq J, Churlaud C, Deschavanne P. 2010. A benchmark of parametric methods for horizontal transfers detection. PLoS ONE 5: e9989. Belbahri L, Moralejo E, Calmin G, Oszako T, García JA, Descals E, Lefort F. 2006. Phytophthora polonica, a new species isolated from declining Alnus glutinosa stands in Poland. FEMS Microbiology Letters 261: 165–174. Bendtsen JD, Nielsen H, von Heijne G, Brunak S. 2004. Improved prediction of signal peptides: SignalP 3.0. Journal of molecular biology 340: 783–95. Bengtsson-Palme J, Ryberg M, Hartmann M, Branco S, Wang Z, Godhe A, De Wit P, Sánchez-García M, Ebersberger I, de Sousa F, et al. 2013. Improved software detection and extraction of ITS1 and ITS2 from ribosomal ITS sequences of fungi and other eukaryotes for analysis of environmental sequencing data. Methods in Ecology and Evolution 4: 914–919. Bertier L, Leus L, D’hondt L, de Cock AWAM, Höfte M. 2013. Host adaptation and speciation through hybridization and polyploidy in Phytophthora. PLoS ONE 8: e85385. De Bie T, Cristianini N, Demuth JP, Hahn MW. 2006. CAFE: a computational tool for the study of gene family evolution. Bioinformatics 22: 1269–71. Bilodeau GJ. 2008. Détection et génomique de Phytophthora ramorum agent causal de la mort subite du chêne (l’encre des chênes rouges). Thesis, Université Laval, Québec. Bilodeau GJ, Lévesque CA, Cock AWAM de, Duchaine C, Brière S, Uribe P, Martin FN, Hamelin RC. 2007. Molecular detection of Phytophthora ramorum by real time-PCR using TaqMan, SYBR®Green and molecular beacons. Phytopathology 97: 632–642. Birch PRJ, Rehmany AP, Pritchard L, Kamoun S, Beynon JL. 2006. Trafficking arms: oomycete effectors enter host plant cells. Trends in Microbiology 14: 8–11. Birky-Jr. CW. 1996. Heterozygosity, heteromorphy, and phylogenetic trees in asexual eukaryotes. Genetics 144: 427–437. Blackburn TM, Lockwood JL, Cassey P. 2015. The influence of numbers on invasion success. Molecular Ecology 24: 1942–1953. Blair JE, Coffey MD, Park SY, Geiser DM, Kang S. 2008. A multi-locus phylogeny for Phytophthora utilizing markers derived from complete genome sequences. Fungal Genetics and Biology 45: 266–277. Blomquist C, Irving T, Osterbauer N, Reeser R. 2005. Phytophthora hibernalis: A new pathogen on Rhododendron and evidence of cross amplification with two PCR detection assays for Phytophthora ramorum. Plant Health Progress DOI: 10.1094/PHP-2005-0728-1001-HN. Bonants PJM, Hagenaar-de Weerdt M, Man in ’t Veld WA, Baayen RP. 2000. Molecular characterization of natural hybrids of Phytophthora nicotianae and P. cactorum. Phytopathology 90: 867–874. 164  Bopp SER, Manary MJ, Bright AT, Johnston GL, Dharia N V, Luna FL, McCormack S, Plouffe D, McNamara CW, Walker JR, et al. 2013. Mitotic evolution of Plasmodium falciparum shows a stable core genome but recombination in antigen families. PLoS Genetics 9: e1003293. Brasier CM. 1992. Evolutionary Biology of Phytophthora. Annual Review of Phytopathology 30: 153–170. Brasier CM. 1995. Episodic selection as a force in fungal microevolution, with special reference to clonal speciation and hybrid introgression. Canadian Journal of Botany 73: 1213–1221. Brasier CM. 2001. Rapid evolution of introduced plant pathogen via interspecific hybridization. Bioscience 51: 123–133. Brasier CM. 2008. The biosecurity threat to the UK and global environment from international trade in plants. Plant Pathology 57: 792–808. Brasier CM. 2012. Rapid evolution of introduced tree pathogens via episodic selection and horizontal gene transfer. Proceedings of the fourth international workshop on the genetics of host-parasite interactions in forestry: Disease and insect resistance in forest trees. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, Pages 133-142. Brasier CM, Cooke DEL, Duncan JM. 1999. Origin of a new Phytophthora pathogen through interspecific hybridization. Proceedings of the National Academy of Sciences of the United States of America 96: 5878–5883. Brasier CM, Cooke DEL, Duncan JM, Hansen EM. 2003. Multiple new phenotypic taxa from trees and riparian ecosystems in Phytophthora gonapodyides-P. megasperma ITS Clade 6, which tend to be high-temperature tolerant and either inbreeding or sterile. Mycological Research 107: 277–290. Brasier CM, Jung T. 2006. Recent developments in Phytophthora diseases of trees and natural ecosystems in Europe. Progress in Research on Phytophthora Diseases of Forest Trees. Proceedings, 3rd Int. IUFRO Working Party 7: 11–17. Brasier CM, Kirk SA. 2001. Comparative aggressiveness of standard and variant hybrid alder phytophthoras, Phytophthora cambivora and other Phytophthora species on bark of Alnus, Quercus and other woody hosts. Plant Pathology 50: 218–229. Brasier C, Kirk S. 2004. Production of gametangia by Phytophthora ramorum in vitro. Mycological Research 108: 823–827. Brasier CM, Kirk S, Delcan J, Cooke DEL, Jung T, Man in’t Veld WA. 2004. Phytophthora alni sp. nov. and its variants: designation of emerging heteroploid hybrid pathogens spreading on Alnus trees. Mycological research 108: 1172–1184. Brasier CM, Rose J, Gibbs JN. 1995. An unusual Phytophthora associated with widespread alder mortality in Britain. Plant Pathology 44: 999–1007. Brasier C, Webber J. 2010. Sudden larch death. Science 329: 824–825. Brazee NJ, Wick RL, Hulvey JP. 2016. Phytophthora species recovered from the Connecticut river valley in Massachusetts, USA. Mycologia 108: 6–19. 165  Broad Institute ( 2014. ‘Phytophthora parasitica Assembly Dev initiative’. : Broders KD, Lipps PE, Paul PA, Dorrance AE. 2007. Characterization of Pythium spp. associated with corn and soybean seed and seedling disease in Ohio. Plant Disease 91: 727–735. Brown A V, Brasier CM. 2007. Colonization of tree xylem by Phytophthora ramorum, P. kernoviae and other Phytophthora species. Plant Pathology 56: 227–241. Brown JKM, Hovmøller MS. 2002. Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science 297: 537–541. Browning SR, Browning BL. 2007. Rapid and accurate haplotype phasing and missing-data inference for whole-genome association studies by use of localized haplotype clustering. American journal of human genetics 81: 1084–1097. Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA Genes for phylogenetics. In: M A Innis, D H Gelfand JJS and TJW, ed. PCR Protocols: A Guide to Methods and Applications. Academic press, London, 315–322. Burgess TI. 2015. Molecular characterization of natural hybrids formed between five related indigenous Clade 6 Phytophthora species. PLoS ONE 10: e0134225. Burgess TI, Scott JK, Mcdougall KL, Stukely MJC, Crane C, Dunstan WA, Brigg F. 2017a. Current and projected global distribution of Phytophthora cinnamomi, one of the world ’ s worst plant pathogens. Global Change Biology: 1661–1674. Burgess TI, White D, McDougall KM, Garnas J, A.Dunstan W, Catala S, Carnegie AJ, Worboys S, Cahill D, Vettraino A-M, et al. 2017b. Distribution and diversity of Phytophthora across Australia. Pacific Conservation Biology 23: 1–13. Cahill DM, Rookes JE, Wilson BA, Gibson L, McDougall KL. 2008. Phytophthora cinnamomi  and Australia’s biodiversity: Impacts, predictions and progress towards control. Australian Journal of Botany 56: 279–310. Callaghan S, Guest D. 2015. Globalisation , the founder effect , hybrid Phytophthora species and rapid evolution : new headaches for biosecurity. Australasian Plant Pathology 44: 255–262. Calo S, Billmyre RB, Heitman J. 2013. Generators of phenotypic diversity in the evolution of pathogenic microorganisms. PLoS Pathogens 9: e1003181. Capinha C, Essl F, Seebens H, Moser D, Pereira HM. 2015. The dispersal of alien species redefines biogeography in the Anthropocene. Science 348: 1248–1251. Caporaso G, Kuczynski J, Stombaugh J, Bittinger K, Bushman F, Costello E, Fierer N, Peña A, Goodrich J, Gordon J, et al. 2010. QIIME allows analysis of high-throughput community sequencing data. Nature Methods 7: 335–336. Català S, Pérez-sierra A, Abad-campos P. 2015. The use of genus-specific amplicon pyrosequencing to assess Phytophthora species diversity using eDNA from soil and water in Northern Spain. PLoS ONE 10: e0119311. Català S, Pérez-Sierra A, Abad-campos P. 2017. Metabarcoding and development of new real-time specific assays reveal Phytophthora species diversity in holm oak forests in eastern Spain. Plant Pathology 66: 115–123. 166  Chamnanpunt J, Shan W, Tyler BM. 2001. High frequency mitotic gene conversion in genetic hybrids of the oomycete Phytophthora sojae. Proceedings of the National Academy of Sciences 98: 14530–14535. Chastagner GA, Hansen EM, Riley KL, Sutton W. 2004. Susceptibility of conifer shoots to infection by Phytophthora ramorum. Phytopathology 94: S16. Chastagner G, Oak S, Omdal D, Ramsey- A, Coats K, Valachovic Y, Lee C, Jeffers S, Elliott M. 2009. Spread of P . ramorum from nurseries into waterways — Implications for pathogen establishment in new areas. Fourth Sudden Oak Death Science Symposium. Santa Cruz, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station, Pages 22-26. Chastagner GA, Riley KL. 2010. First report of Phytophthora ramorum infecting California red fir in California. Plant Disease 94: 1170. Chevenet F, Brun C, Bañuls A-L, Jacq B, Christen R. 2006. TreeDyn: towards dynamic graphics and annotations for analyses of trees. BMC Bioinformatics 7: 439. Chuma I, Isobe C, Hotta Y, Ibaragi K, Futamata N, Kusaba M, Yoshida K, Terauchi R, Fujita Y, Nakayashiki H, et al. 2011. Multiple translocation of the AVR-Pita effector gene among chromosomes of the Rice Blast fungus Magnaporthe oryzae and related species. PLoS Pathogens 7: e1002147. Coffey MD. 2012. The World Phytophthora Genetic Resource Collection (WPC). Conesa A, Götz S, García-Gómez JM, Terol J, Talón M, Robles M. 2005. Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics 21: 3674–3676. Connell JH. 1971. On the role of natural enemies in preventing competitive exclusion in some marine animals and rain forest trees. Dynamics of population. PUDOC, 298–312. Cooke DEL, Drenth A, Duncan JM, Wagels G, Brasier CM. 2000. A molecular phylogeny of Phytophthora and related Oomycetes. Fungal Genetics and Biology 30: 17–32. Cornish-Bowden A. 1985. Nomenclature for incompletely specified bases in nucleic acid sequences: recommendations 1984. Nucleic Acids Research 13: 3021–3030. van Damme M, Cano LM, Oliva R, Schornack S, Segretin ME, Kamoun S, Raffaele S. 2012. Evolutionary and functional dynamics of Oomycetes effector genes. In: Martin F,,  In: Kamoun S, eds. Effectors in plant-microbe interactions. Chichester, UK: John Wiley & Sons, Inc., 426. Danecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, Handsaker RE, Lunter G, Marth GT, Sherry ST, et al. 2011. The variant call format and VCFtools. Bioinformatics 27: 2156–2158. Darriba D, Taboada GL, Doallo R, Posada D. 2012. jModelTest 2: more models, new heuristics and parallel computing. Nature Methods 9: 772. Dart NL, Chastagner GA. 2007. High Recovery Rate of Phytophthora from Containerized Nursery Stock Pots at a Retail Nursery Highlights Potential For Spreading Exotic Oomycetes. 167  Plant Health Progress. Davidson JM, Garbelotto M, Koike ST, Rizzo DM. 2002. First Report of Phytophthora ramorum on Douglas-Fir in California. Plant Disease 86: 1274. Davidson JM, Wickland AC, Patterson HA, Falk KR, Rizzo DM. 2005. Transmission of Phytophthora ramorum in mixed-evergreen forest in California. Phytopathology 95: 587–596. Denman S, Kirk SA, Brasier CM, Webber JF. 2005. In vitro leaf inoculation studies as an indication of tree foliage susceptibility to Phytophthora ramorum in the UK. Plant Pathology 54: 512–521. Dereeper A, Guignon V, Blanc G, Audic S, Buffet S, Chevenet F, Dufayard J-F, Guindon S, Lefort V, Lescot M, et al. 2008. robust phylogenetic analysis for the non-specialist. Nucleic Acids Research 36: W465–W469. Derrien T, Estellé J, Sola SM, Knowles DG, Raineri E, Guigó R, Ribeca P. 2012. Fast computation and applications of genome mappability. PLoS ONE 7: e30377. Desprez-Loustau M-L, Courtecuisse R, Robin C, Husson C, Moreau P-A, Blancard D, Selosse M-A, Lung-Escarmant B, Piou D, Sache I. 2010. Species diversity and drivers of spread of alien fungi (sensu lato) in Europe with a particular focus on France. Biological Invasions 12: 157–172. Desprez-Loustau M-L, Robin C, Buée M, Courtecuisse R, Garbaye J, Suffert F, Sache I, Rizzo DM. 2007. The fungal dimension of biological invasions. Trends in Ecology & Evolution 22: 472–480. Dhillon B, Feau N, Aerts AL, Beauseigle S, Bernier L, Copeland A, Foster A, Gill N, Henrissat B, Herath P, et al. 2015. Horizontal gene transfer and gene dosage drives adaptation to wood colonization in a tree pathogen. Proceedings of the National Academy of Sciences 112: 3451–3456. Dickie IA, Wakelin A, Martinez-Garcia L, Richardson SJ, Makiola A, Tylianakis JM. 2017. Oomycetes during 120,000 years of temperate rainforest ecosystem development. bioRxiv DOI: Dobrowolski MP, Tommerup IC, Shearer BL, O’Brien PA. 2003. Three clonal lineages of Phytophthora cinnamomi in Australia revealed by microsatellites. Phytopathology 93: 695–704. Donahoo RS, Lamour KH. 2008. Interspecific hybridization and apomixis between Phytophthora capsici and Phytophthora tropicalis. Mycologia 100: 911–920. Dong S, Raffaele S, Kamoun S. 2015. The two-speed genomes of filamentous pathogens: waltz with plants. Current Opinion in Genetics & Development 35: 57–65. Drummond AJ, Suchard MA, Xie D, Rambaut A. 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29: 1969–1973. Durán A, Gryzenhout M, Slippers B, Ahumada R, Rotella A, Flores F, Wingfield BD, Wingfield MJ. 2008. Phytophthora pinifolia sp. nov. associated with a serious needle disease of Pinus radiata in Chile. Plant Pathology 57: 715–727. Dutech C, Barrès B, Bridier J, Robin C, Milgroom MG, Ravigné V. 2012. The chestnut blight fungus world tour: Successive introduction events from diverse origins in an invasive plant 168  fungal pathogen. Molecular Ecology 21: 3931–3946. Dutheil JY, Mannhaupt G, Schweizer G, M.K. Sieber C, Münsterkötter M, Güldener U, Schirawski J, Kahmann R. 2016. A tale of genome compartmentalization: The evolution of virulence clusters in smut fungi. Genome Biology and Evolution 8: 681–704. Eddy SR. 2008. A probabilistic model of local sequence alignment that simplifies statistical significance estimation. PLoS Computational Biology 4: e1000069. Edgar RC. 2004. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC bioinformatics 5: 113. Edgar RC. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461. Elliott M, Sumampong G, Varga A, Shamoun SF, James D, Masri S, Grünwald NJ. 2011. Phenotypic differences among three clonal lineages of Phytophthora ramorum. Forest Pathology 41: 7–14. Ellison AM, Bank MS, Clinton BD, Colburn EA, Elliott K, Ford CR, Foster DR, Kloeppel BD, Knoepp JD, Lovett GM, et al. 2005. Loss of foundation species: consequences for the structure and dynamics of forested ecosystems. Frontiers in Ecology and the Environment 3: 479–486. Esmaeili Taheri A, Chatterton S, Gossen BD, McLaren DL. 2017. Metagenomic analysis of oomycete communities from the rhizosphere of field pea on the Canadian prairies. Canadian Journal of Microbiology 63: 758–768. Excoffier L, Laval G, Schneider S. 2005. Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online  1: 47–50. Faino L, Seidl MF, Shi-Kunne X, Pauper M, Van Den Berg GCM, Wittenberg AHJ, Thomma BPHJ. 2016. Transposons passively and actively contribute to evolution of the two-speed genome of a fungal pathogen. Genome Research 26: 1091–1100. Farrer RA, Henk DA, Garner TWJ, Balloux F, Woodhams DC, Fisher MC. 2013. Chromosomal copy number variation, selection and uneven rates of recombination reveal cryptic genome diversity linked to pathogenicity. PLoS Genetics 9: e1003703. Fawke S, Doumane M, Schornack S. 2015. Oomycete interactions with plants: Infection strategies and resistance principles. Microbiology and Molecular Biology Reviews 79: 263–280. Feau N, Taylor G, Dale AL, Dhillon B, Bilodeau GJ, Birol I, Jones SJM, Hamelin RC. 2016. Genome sequences of six Phytophthora species threatening forest ecosystems. Genomics Data 10: 85–88. Forche A, Abbey D, Pisithkul T, Weinzierl MA, Ringstrom T, Bruck D, Petersen K, Berman J. 2011. Stress alters rates and types of loss of heterozygosity in Candida albicans. mBio 2: e00129-11. Forestry Commision Great Britian. 2012. Phytophthora ramorum in larch trees – update. : Forsman A. 2014. Effects of genotypic and phenotypic variation on establishment are important for conservation, invasion, and infection biology. Proceedings of the National Academy of 169  Sciences 111: 302–307. Fox J. 2005. The R Commander: A basic statistics graphical user interface to R. Journal of Statistical Software 14: 1–42. Fry WE, Goodwin SB, Matuszak JM, Spielman LJ, Milgroom MG, Drenth A. 1992. Population genetics and intercontinental migrations of Phytophthora infestans. Annual Review of Phytopathology 30: 107–130. Gaastra W, Lipman LJA, De Cock AWAM, Exel TK, Pegge RBG, Scheurwater J, Vilela R, Mendoza L. 2010. Pythium insidiosum: An overview. Veterinary Microbiology 146: 1–16. Galagan JE, Calvo SE, Borkovich K a, Selker EU, Read ND, Jaffe D, FitzHugh W, Ma L-J, Smirnov S, Purcell S, et al. 2003. The genome sequence of the filamentous fungus Neurospora crassa. Nature 422: 859–868. Galagan JE, Selker EU. 2004. RIP: The evolutionary cost of genome defense. Trends in Genetics 20: 417–423. Garbelotto M. 2008. Molecular analysis to study invasions by forest pathogens : examples from Mediterranean ecosystems. Phytopathologia Mediterranea 47: 183–203. Garbelotto M, Ivors K, Hüberli D, Bonants P, Wagner A. 2006. Potential for sexual reproduction of Phytophthora ramorum in Washington state nurseries. Proceedings of the Sudden Oak Death Second Science Symposium: The State of Our Knowledge. Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station, Pages 129-130. García-Alcalde F, Okonechnikov K, Carbonell J, Cruz LM, Götz S, Tarazona S, Dopazo J, Meyer TF, Conesa A. 2012. Qualimap: evaluating next-generation sequencing alignment data. Bioinformatics 28: 2678–2679. Garkaklis MJ, Calver MC, Wilson BA, Hardy GESJ. 2004. Habitat alteration caused by an introduced plant disease, Phytophthora cinnamomi: A potential threat to the conservation of Australian forest fauna. Conservation of Australia’s Forest Fauna. Royal Zoological Society of New South Wales, 899–913. Gigot J, Walters TW, Zasada IA. 2013. Impact and occurrence of Phytophthora rubi and Pratylenchus penetrans in commercial red raspberry (Rubus ideaus) fields in northwestern Washington. International Journal of Fruit Science 13: 357–372. Giraud T, Gladieux P, Gavrilets S. 2010. Linking the emergence of fungal plant diseases with ecological speciation. Trends in Ecology & Evolution 25: 387–395. Gladieux P, Feurtey  a., Hood ME, Snirc  a., Clavel J, Dutech C, Roy M, Giraud T. 2015. The population biology of fungal invasions. Molecular Ecology 24: 1969–1986. Gladieux P, Ropars J, Badouin H, Branca A, Aguileta G, de Vienne DM, Rodríguez de la Vega RC, Branco S, Giraud T. 2014. Fungal evolutionary genomics provides insight into the mechanisms of adaptive divergence in eukaryotes. Molecular ecology 23: 753–73. Goodwin SB. 1997. The population genetics of Phytophthora. Phytopathology 87: 462–473. Goodwin SB, Ben M’barek S, Dhillon B, Wittenberg AHJ, Crane CF, Hane JK, Foster AJ, Van der Lee T a J, Grimwood J, Aerts A, et al. 2011. Finished genome of the fungal wheat pathogen Mycosphaerella graminicola reveals dispensome structure, chromosome plasticity, and 170  stealth pathogenesis. PLoS genetics 7: e1002070. Gorman M. 1993. Tree and ornamental disease control in the city of Victoria. B.C. Pest Monitor 2. Goss EM, Carbone I, Grünwald NJ. 2009. Ancient isolation and independent evolution of the three clonal lineages of the exotic sudden oak death pathogen Phytophthora ramorum. Molecular Ecology 18: 1161–1174. Goss EM, Cardenas ME, Myers K, Forbes G a, Fry WE, Restrepo S, Grünwald NJ. 2011a. The plant pathogen Phytophthora andina emerged via hybridization of an unknown Phytophthora species and the Irish potato famine pathogen, P. infestans. PloS ONE 6: e24543. Goss EM, Larsen M, Vercauteren A, Werres S, Heungens K, Grünwald NJ. 2011b. Phytophthora ramorum in Canada: Evidence for migration within North America and from Europe. Phytopathology 101: 166–171. Grant BJ, Rodrigues APC, ElSawy KM, McCammon JA, Caves LSD. 2006. Bio3d: an R package for the comparative analysis of protein structures. Bioinformatics 22: 2695–2696. Greslebin AG, Hansen EM, Sutton W. 2007. Phytophthora austrocedrae sp. nov., a new species associated with Austrocedrus chilensis mortality in Patagonia (Argentina). Mycological Research 111: 308–316. Greslebin AG, Hansen EM, Winton LM, Rajchenberg M. 2005. Phytophthora species from declining Austrocedrus chilensis forests in Patagonia, Argentina. Mycologia 97: 218–228. Grigoriev I V, Nordberg H, Shabalov I, Aerts A, Cantor M, Goodstein D, Kuo A, Minovitsky S, Nikitin R, Ohm R a, et al. 2012. The genome portal of the Department of Energy Joint Genome Institute. Nucleic Acids Research 40: D26-32. Grünwald NJ, Garbelotto M, Goss EM, Heungens K, Prospero S. 2012. Emergence of the sudden oak death pathogen Phytophthora ramorum. Trends in Microbiology 20: 131–138. Grünwald NJ, Goss EM, Ivors K, Garbelotto M, Martin FN, Prospero S, Hansen E, Bonants PJM, Hamelin RC, Chastagner G, et al. 2009. Standardizing the nomenclature for clonal Lineages of the Sudden Oak Death pathogen, Phytophthora ramorum. Phytopathology 99: 792–795. Grünwald NJ, Goss EM, Press CM. 2008. Phytophthora ramorum: a pathogen with a remarkably wide host range causing sudden oak death on oaks and ramorum blight on woody ornamentals. Molecular Plant Pathology 9: 729–740. Guindon S, Dufayard J-F, Lefort V, Anisimova M, Hordijk W, Gascuel O. 2010. New algorithms and methods to astimate maximum-likelihood phylogenies: Assessing the performance of PhyML 3.0. Systematic Biology 59: 307–321. Guy L, Roat Kultima J, Andersson SGE. 2010. genoPlotR: comparative gene and genome visualization in R. Bioinformatics 26: 2334–2335. Haas BJ, Kamoun S, Zody MC, Jiang RHY, Handsaker RE, Cano LM, Grabherr M, Kodira CD, Raffaele S, Torto-Alalibo T, et al. 2009. Genome sequence and analysis of the Irish potato famine pathogen Phytophthora infestans. Nature 461: 393–398. Hall TA. 1999. BioEdit: A user-friendly biological sequence alignment editor and analysis 171  analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41: 95_98. Hansen EM. 2008. Alien forest pathogens: Phytophthora species are changing world forests. Boreal Environ. Res. 13: 33–41. Hansen EM. 2015. Phytophthora species emerging as pathogens of forest trees. Current Forestry Reports 1: 16–24. Hansen EM, Goheen DJ, Jules ES, Ullian B. 2000. Managing Port-Orford-Cedar and the introduced pathogen Phytophthora lateralis. Plant Disease 84: 4–14. Hansen EM, Kanaskie A, Prospero S, McWilliams M, Goheen EM, Osterbauer N, Reeser P, Sutton W. 2008. Epidemiology of Phytophthora ramorum in Oregon tanoak forests. Canadian Journal of Forest Research 38: 1133–1143. Hansen EM, Parke JL, Sutton W. 2005. Susceptibility of Oregon forest trees and shrubs to Phytophthora ramorum: A Comparison of Artificial inoculation and natural infection. Plant Disease 89: 63–70. Hansen EM, Reeser PW, Sutton W. 2012a. Phytophthora beyond agriculture. Annual Review of Phytopathology 50: 359–378. Hansen EM, Reeser PW, Sutton W. 2012b. Phytophthora borealis and Phytophthora riparia, new species in Phytophthora ITS Clade 6. Mycologia 104: 1133–1142. Hansen EM, Reeser PW, Sutton W. 2017. Ecology and pathology of Phytophthora ITS clade 3 species in forests in western Oregon, USA. Mycologia 109: 100–114. Hansen EM, Rizzo D, Garbelotto M. 2006. Phytophthora species from oak and tanoak forests in California and Oregon. Sudden Oak Death Second Science Symposium: The State of Our Knowledge. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, Pages 63-65. Hardham AR. 2005. Phytophthora cinnamomi. Molecular Plant Pathology 6: 589–604. Hardy GESJ, Colquhoun IJ, Shearer BL, Tommerup I. 2001. The impact and control of Phytophthora cinnamomi in native and rehabilitated forest ecosystems in Western Australia. Forest, Snow Landscape Research 76: 337–343. Hatta R, Ito K, Hosaki Y, Tanaka T, Tanaka A, Yamamoto M, Akimitsu K, Tsuge T. 2002. A conditionally dispensable chromosome controls host-specific pathogenicity in the fungal plant pathogen Alternaria alternata. Genetics 161: 59–70. Hegarty MJ. 2012. Invasion of the hybrids. Molecular Ecology 21: 4669–4671. Heitman J. 2006. Sexual reproduction and the evolution of microbial pathogens. Current Biology 16: R711–R725. Heitman J, Sun S, James TY. 2013. Evolution of fungal sexual reproduction. Mycologia 105: 1–27. Hong CX, Moorman GW. 2005. Plant pathogens in irrigation water: challenges and opportunities. Critical Reviews in Plant Sciences 24: 189–208. Hu J, Diao Y, Zhou Y, Lin D, Bi Y, Pang Z, Trout Fryxell R, Liu X, Lamour K. 2013. Loss of heterozygosity drives clonal diversity of Phytophthora capsici in China. PLoS ONE 8: e82691. 172  Hubbs CL. 1955. Hybridization between fish species in nature. Systematic Zoology 4: 1–20. Hüberli D, Hardy GESJ, White D, Williams N, Burgess TI. 2013. Fishing for Phytophthora from western Australia’s waterways: a distribution and diversity survey. Australasian Plant Pathology 42: 251–260. Huitema E, Bos JIB, Tian M, Win J, Waugh ME, Kamoun S. 2004. Linking sequence to phenotype in Phytophthora–plant interactions. Trends in Microbiology 12: 193–200. Hulbert JM, Agne MC, Burgess TI, Roets F, Wingfield MJ. 2017. Urban environments provide opportunities for early detections of Phytophthora invasions. Biological Invasions 19: 3629–3644. Hulvey J, Gobena D, Finley L, Lamour K. 2010. Co-occurrence and genotypic distribution of Phytophthora species recovered from watersheds and plant nurseries of eastern Tennessee. Mycologia 102: 1127–1133. Hunt M, Newbold C, Berriman M, Otto TD. 2014. A comprehensive evaluation of assembly scaffolding tools. Genome Biology 15: R42. Husson C, Aguayo J, Revellin C, Frey P, Ioos R, Marçais B. 2015. Evidence for homoploid speciation in Phytophthora alni supports taxonomic reclassification in this species complex. Fungal Genetics and Biology 77: 12–21. Hwang J, Oak SW, Jeffers SN. 2008. Detecting Phytophthora ramorum and other species of Phytophthora in streams in natural ecosystems using baiting and filtration methods. Proceedings of the Sudden Oak Death Third Science Symposium. Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station, Pages 55-58. Ioos R, Andrieux A, Marçais B, Frey P. 2006. Genetic characterization of the natural hybrid species Phytophthora alni as inferred from nuclear and mitochondrial DNA analyses. Fungal genetics and biology 43: 511–29. Ivors K, Garbelotto M, Vries IDE, Ruyter-Spira C, Hekkert BTE, Rosenzweig N, Bonants P. 2006. Microsatellite markers identify three lineages of Phytophthora ramorum in US nurseries, yet single lineages in US forest and European nursery populations. Molecular Ecology 15: 1493–1505. Ivors KL, Hayden KJ, Bonants PJM, Rizzo DM, Garbelotto M. 2004. AFLP and phylogenetic analyses of North American and European populations of Phytophthora ramorum. Mycological Research 108: 378–392. James TY, Litvintseva AP, Vilgalys R, Morgan JAT, Taylor JW, Fisher MC, Berger L, Weldon C, du Preez L, Longcore JE. 2009. Rapid global expansion of the fungal disease Chytridiomycosis into declining and healthy amphibian populations. PLoS Pathogens 5: e1000458. Janzen DH. 1970. Herbivores and the number of tree species in tropical forests. The American Naturalist 104: 501–528. Jiang RHY, Tripathy S, Govers F, Tyler BM. 2008. RXLR effector reservoir in two Phytophthora species is dominated by a single rapidly evolving superfamily with more than 700 members. Proceedings of the National Academy of Sciences 105: 4874–4879. 173  Jiang RHY, Tyler BM. 2012. Mechanisms and evolution of virulence in Oomycetes. Annual Review of Phytopathology 50: 295–318. Jiang RHY, Tyler BM, Whisson SC, Hardham AR, Govers F. 2006. Ancient origin of elicitin gene clusters in Phytophthora genomes. Molecular Biology and Evolution 23: 338–351. Joly DL, Feau N, Tanguay P, Hamelin RC. 2010. Comparative analysis of secreted protein evolution using expressed sequence tags from four poplar leaf rusts (Melampsora spp.). BMC Genomics 11: 422. Jombart T. 2008. Adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24: 1403–1405. de Jonge R, Bolton MD, Kombrink A, van den Berg GCM, Yadeta K a, Thomma BPHJ. 2013. Extensive chromosomal reshuffling drives evolution of virulence in an asexual pathogen. Genome Research 23: 1271–1282. Judelson HS. 2012. Dynamics and innovations within oomycete genomes: Insights into biology, pathology, and evolution. Eukaryotic Cell 11: 1304–1312. Jung T, Burgess TI. 2009. Re-evaluation of Phytophthora citricola isolates from multiple woody hosts in Europe and North America reveals a new species, Phytophthora plurivora sp. nov. Persoonia 22: 95–110. Jung T, Orlikowski L, Henricot B, Abad-Campos P, Aday AG, Aguín Casal O, Bakonyi J, Cacciola SO, Cech T, Chavarriaga D, et al. 2016. Widespread Phytophthora infestations in European nurseries put forest, semi-natural and horticultural ecosystems at high risk of Phytophthora diseases. Forest Pathology 46: 134–163. Jung T, Stukely MJC, Hardy Ges, White D, Paap T, Dunstan WA, Burgess TI. 2011. Multiple new Phytophthora species from ITS Clade 6 associated with natural ecosystems in Australia: Evolutionary and ecological implications. Persoonia : Molecular Phylogeny and Evolution of Fungi 26: 13–39. Jurka J, Kapitonov V, Pavlicek A, Klonowski P, Kohany O, Walichiewicz J. 2005. Repbase Update, a database of eukaryotic repetitive elements. Cytogenetic and genome research 110: 462–467. Kamoun S. 2006. A catalogue of the effector secretome of plant pathogenic Oomycetes. Annual Review of Phytopathology 44: 41–60. Kamoun S, Furzer O, Jones JDG, Judelson HS, Ali GS, Dalio RJD, Roy SG, Schena L, Zambounis A, Panabières F, et al. 2015. The Top 10 oomycete pathogens in molecular plant pathology. Molecular Plant Pathology 16: 413–434. Kanaskie A, Goheen E, Osterbauer N, Mcwilliams M, Hansen E, Sutton W. 2007. Eradication of Phytophthora ramorum from Oregon forests : Status after 6 years. Proceedings of the Sudden Oak Death Third Science Symposium. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, Pages 15-17. Kanaskie A, Hansen E, Goheen EM, Mcwilliams M, Laine J, Savona S, Timeus H, Woosley B, Sutton W, Reeser P, et al. 2010. Detection and eradication of Phytophthora ramorum from Oregon Forests , 2001-2008. Proceedings of the Sudden Oak Death Fourth Science Symposium. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of 174  Agriculture, Pages 5-7. Kasuga T, Bui M, Bernhardt E, Swiecki T, Aram K, Cano LM, Webber J, Brasier C, Press C, Grünwald NJ, et al. 2016. Host-induced aneuploidy and phenotypic diversification in the Sudden Oak Death pathogen Phytophthora ramorum. BMC Genomics 17: 385. Kasuga T, Kozanitas M, Bui M, Hüberli D, Rizzo DM, Garbelotto M. 2012. Phenotypic diversification is associated with host-induced transposon derepression in the Sudden Oak Death pathogen Phytophthora ramorum. PLoS ONE 7: e34728. Katoh K, Standley DM. 2013. MAFFT Multiple Sequence Alignment Software Version 7 : improvements in performance and usability. Molecular Biology 30: 772–780. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper A, Markowitz S, Duran C, et al. 2012. Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28: 1647–1649. Keller O, Kollmar M, Stanke M, Waack S. 2011. A novel hybrid gene prediction method employing protein multiple sequence alignments. Bioinformatics 27: 757–763. Kliejunas JT. 2011. Sudden oak death; ramorum blight. A risk assessment of climate change and the impact of forest diseases on forest ecosystems in the Western United States and Canada. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, Pages 18-24. Knaus BJ, Fieland VJ, Graham KA, Grünwald NJ. 2015. Diversity of foliar Phytophthora species on Rhododendron in Oregon nurseries. Plant Disease 99: 1326–1332. Kozanitas M, Osmundson TW, Linzer R, Garbelotto M. 2017. Interspecific interactions between the Sudden Oak Death pathogen Phytophthora ramorum and two sympatric Phytophthora species in varying ecological conditions. Fungal Ecology 28: 86–96. Kozik A, Chan B, Michemore R. 2003. BLAST parser, distance matrix file and protein sequence clustering. How to obtain a distance matrix file from BLAST search results and use it for sequence clustering analysis : Kroon LPNM, Brouwer H, de Cock AW a M, Govers F. 2012. The genus Phytophthora anno 2012. Phytopathology 102: 348–64. Lake JC, Leishman MR. 2004. Invasion success of exotic plants in natural ecosystems: The role of disturbance, plant attributes and freedom from herbivores. Biological Conservation 117: 215–226. Laliberté E, Lambers H, Burgess TI, Wright SJ. 2015. Phosphorus limitation , soil-borne pathogens and the coexistence of plant species in hyperdiverse forests and shrublands. New Phytologist 206: 507–521. Lamour KH, Daughtrey ML, Benson DM, Hwang J, Hausbeck MK. 2003. Etiology of Phytophthora drechsleri and P. nicotianae (=P. parasitica) diseases affecting floriculture crops. Plant Disease 87: 854–858. Lamour KH, Mudge J, Gobena D, Hurtado-gonzales OP, Kuo A, Miller NA, Rice BJ, Raffaele S, Liliana M, Bharti AK, et al. 2012. Genome sequencing and mapping reveal loss of 175  heterozygosity as a mechanism for rapid adaptation in the vegetable pathogen Phytophthora capsici. Molecular Plant-Microbe Interactions 25: 1350–1360. Latijnhouwers M, de Wit PJGM, Govers F. 2003. Oomycetes and fungi: Similar weaponry to attack plants. Trends in Microbiology 11: 462–469. Lee PS, Greenwell PW, Dominska M, Gawel M, Hamilton M, Petes TD. 2009. A fine-structure map of spontaneous mitotic crossovers in the yeast Saccharomyces cerevisiae. PLoS Genetics 5: e1000410. Li H, Durbin R. 2009. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25: 1754–1760. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R. 2009. The sequence alignment/map format and SAMtools. Bioinformatics 25: 2078–9. Li L, Stoeckert CJ, Roos DS. 2003. OrthoMCL: Identification of ortholog groups for eukaryotic genomes. Genome Research 13: 2178–2189. Linzer RE, Rizzo DM, Cacciola SO, Garbelotto M. 2009. AFLPs detect low genetic diversity for Phytophthora nemorosa and P. pseudosyringae in the US and Europe. Mycological Research 113: 298–307. Llorente B, Smith CE, Symington LS. 2008. Break-induced replication: What is it and what is it for? Cell Cycle 7: 859–864. Loo JA. 2009. Ecological impacts of non-indigenous invasive fungi as forest pathogens. Biological Invasions 11: 81–96. Lozupone C, Knight R. 2005. UniFrac: A New Phylogenetic Method for Comparing Microbial Communities. Applied and Environmental Microbiology 71: 8228–8235. Mack RN, Simberloff D, Mark Lonsdale W, Evans H, Clout M, Bazzaz FA. 2000. Biotic invasions: Causes, epidemiology, global consequences, and control. Ecological Applications 10: 689–710. Magwene PM, Kayıkçı Ö, Granek JA, Reininga JM, Scholl Z, Murray D. 2011. Outcrossing, mitotic recombination, and life-history trade-offs shape genome evolution in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences 108: 1987–1992. Maloney PE, Rizzo DM, Koike ST, Harnik TY, Garbelotto M. 2002. First report of Phytophthora ramorum on oast redwood in California. Plant Disease 86: 1274. Man in ‘t Veld WA, de Cock AWAM, Summerbell RC. 2007. Natural hybrids of resident and introduced Phytophthora species proliferating on multiple new hosts. European Journal of Plant Pathology 117: 25–33. Man in ’t Veld WA, Veenbaas-Rijks WJ, Ilieva E, de Cock AWAM, Bonants PJM, Pieters R. 1998. Natural hybrids of Phytophthora nicotianae and Phytophthora cactorum demonstrated by isozyme analysis and random mmplified polymorphic DNA. Phytopathology 88: 922–929. Mandegar MA, Otto SP. 2007. Mitotic recombination counteracts the benefits of genetic segregation. Proceedings of the Royal Society B: Biological Sciences 274: 1301–1307. 176  Marano A V, Jesus AL, de Souza JI, Jerônimo GH, Gonçalves DR, Boro MC, Rocha SCO, Pires-Zottarelli CLA. 2016. Ecological roles of saprotrophic Peronosporales (Oomycetes, Straminipila) in natural environments. Fungal Ecology 19: 77–88. Martin FN, Loper JE. 1999. Soilborne plant diseases caused by Pythium spp.: Ecology, epidemiology, and prospects for biological control. Critical Reviews in Plant Sciences 18: 111–181. Mascheretti S, Croucher PJP, Vettraino A, Prospero S, Garbelotto M. 2008. Reconstruction of the Sudden Oak Death epidemic in California through microsatellite analysis of the pathogen Phytophthora ramorum. Molecular Ecology 17: 2755–2768. Matari NH, Blair JE. 2014. A multilocus timescale for oomycete evolution estimated under three distinct molecular clock models. BMC Evolutionary Biology 14: 101. McDonald BA, Linde C. 2002. Pathogen population genetics, evolutionary potential, and durable resistance. Annual Review of Phytopathology 40: 349–379. McIntosh DL. 1966. The occurrence of Phytophthora spp. in irrigation systems in British Columbia. Canadian Journal of Botany 44: 1591–1596. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, et al. 2010. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Research 20: 1297–1303. Meentemeyer RK, Rank NE, Shoemaker DA, Oneal CB, Wickland AC, Frangioso KM, Rizzo DM. 2008. Impact of sudden oak death on tree mortality in the Big Sur ecoregion of California. Biological Invasions 10: 1243–1255. Metz MR, Frangioso KM, Meentemeyer RK, Rizzo DM. 2011. Interacting disturbances: wildfire severity affected by stage of forest disease invasion. Ecological Applications 21: 313–320. Migliorini D, Ghelardini L, Tondini E, Luchi N, Santini A, Richardson DM. 2015. The potential of symptomless potted plants for carrying invasive soilborne plant pathogens. Diversity and Distributions 21: 1218–1229. Miller MA, Pfeiffer W, Schwartz T. 2010. Creating the CIPRES Science Gateway for Inference of Large Phylogenetic Trees. 2010 Gateway Computing Environments Workshop (GCE), New Orleans, LA: 1–8. Mills KE, Bever JD. 1998. Maintenance of diversity within plant communities: Soil pathogens as agents of negative feedback. Ecology 79: 1595–1601. Mirkin BG, Fenner TI, Galperin MY, Koonin E V. 2003. Algorithms for computing parsimonious evolutionary scenarios for genome evolution, the last universal common ancestor and dominance of horizontal gene transfer in the evolution of prokaryotes. BMC Evolutionary Biology 3: 2. Möller EM, Bahnweg G, Sandermann H, Geiger HH. 1992. A simple and efficient protocol for isolation of high molecular weight DNA from filamentous fungi, fruit bodies, and infected plant tissues. Nucleic acids research 20: 6115–6116. 177  Moller M, Stukenbrock EH. 2017. Evolution and genome architecture in fungal plant pathogens. Nature Reviews Microbiology 15: 756–771. Monahan WB, Tse J, Koenig WD, Garbelotto M. 2008. Mycological Research News1. Mycological Research 112: 757–758. Murphy SK, Rizzo DM. 2006. Incidence of Phytophthora ramorum, P . nemorosa and P . pseudosyringae in three coastal California forest communities. Proceedings of the sudden oak death second science symposium: the state of our knowledge. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, Pages 69-71. Nagel JH, Gryzenhout M, Slippers B, Wingfield MJ, Hardy GESJ, Stukely MJC, Burgess TI. 2013. Characterization of Phytophthora hybrids from ITS clade 6 associated with riparian ecosystems in South Africa and Australia. Fungal Biology 117: 329–347. Nelson EB, Karp MA. 2013. Soil pathogen communities associated with native and non-native Phragmites australis populations in freshwater wetlands. Ecology and Evolution 3: 5254–5267. Newcombe G, Stirling B, Bradshaw HD. 2001. Abundant pathogenic variation in the new hybrid rust Melampsora ×columbiana on hybrid poplar. Phytopathology 91: 981–985. Newcombe G, Stirling B, McDonald S, Bradshaw HD. 2000. Melampsora × columbiana, a natural hybrid of M. medusae and M. occidentalis. Mycological Research 104: 261–274. Noga EJ. 1993. Water mold infections of freshwater fish: Recent advances. Annual Review of Fish Diseases 3: 291–304. Normark BB, Judson OP, Moran NA. 2003. Intraclonal genetic variation: ecological and evolutionary aspects. Biological Journal of the Linnean Society 79: 69–84. North A, Pennanen J, Ovaskainen O, Laine A-L. 2011. Local adaptation in a changing world: the roles of gene-flow, mutation, and sexual reproduction. Evolution 65: 79–89. Nowell RW, Green S, Laue BE, Sharp PM. 2014. The extent of genome flux and its role in the differentiation of bacterial lineages. Genome biology and evolution 6: 1514–1529. Oh E, Gryzenhout M, Wingfield BD, Wingfield MJ, Burgess TI. 2013. Surveys of soil and water reveal a goldmine of Phytophthora diversity in South African natural ecosystems. IMA Fungus 4: 123–131. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin P, O’Hara RB, Simpson GL, Solymos P, Stevens, M.H.H, Wagner H. 2016. Vegan: Community ecology package. R package version 2.3-5. Oliva RF, Kroon LPNM, Chacón G, Flier WG, Ristaino JB, Forbes GA. 2010. Phytophthora andina sp. nov., a newly identified heterothallic pathogen of solanaceous hosts in the Andean highlands. Plant Pathology 59: 613–625. Paap T, Burgess TI, Wingfield MJ. 2017. Urban trees : bridge-heads for forest pest invasions and sentinels for early detection. Biological Invasions 19: 3515–3526. Packer A, Clay K. 2000. Soil pathogens and spatial patterns of seedling mortality in a temperate tree. Nature 404: 278–281. Panabières F, Ponchet M, Allasia V, Cardin L, Ricci P. 1997. Characterization of border 178  species among Pythiaceae: several Pythium isolates produce elicitins, typical proteins from Phytophthora spp. Mycological Research 101: 1459–1468. Paradis E. 2010. pegas: an R package for population genetics with an integrated–modular approach. Bioinformatics 26: 419–420. Paradis E, Claude J, Strimmer K. 2004. APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20: 289–290. Park B, Martin F, Geiser DM, Kim H-S, Mansfield MA, Nikolaeva E, Park S-Y, Coffey MD, Russo J, Kim SH, et al. 2013. Phytophthora database 2.0: Update and future direction. Phytopathology 103: 1204–1208. Parke JL, Knaus BJ, Fieland VJ, Lewis C, Grünwald NJ. 2014. Phytophthora community structure analyses in Oregon nurseries inform systems approaches to disease management. Phytopathology 104: 1052–1062. Parke JL, Lucas S. 2008. Sudden oak death and ramorum blight. The Plant Health Instructor DOI: 10.1094/PHI-I-2008-0227-01. Parra G, Bradnam K, Korf I. 2007. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics 23: 1061–1067. Parra G, Bradnam K, Ning Z, Keane T, Korf I. 2009. Assessing the gene space in draft genomes. Nucleic acids research 37: 289–97. Pautasso M, Dehnen-Schmutz K, Holdenrieder O, Pietravalle S, Salama N, Jeger MJ, Lange E, Hehl-Lange S. 2010. Plant health and global change - Some implications for landscape management. Biological Reviews 85: 729–755. Pautasso M, Schlegel M, Holdenrieder O. 2015. Forest health in a changing world. Microbial Ecology 69: 826–842. Pemberton CL, Salmond GPC. 2004. The Nep1-like proteins—a growing family of microbial elicitors of plant necrosis. Molecular Plant Pathology 5: 353–359. Pérez JE, Nirchio M, Alfonsi C, Muñoz C. 2006. The biology of invasions: The genetic adaptation paradox. Biological Invasions 8: 1115–1121. Philibert A, Desprez-Loustau M-L, Fabre B, Frey P, Halkett F, Husson C, Lung-Escarmant B, Marçais B, Robin C, Vacher C, et al. 2011. Predicting invasion success of forest pathogenic fungi from species traits. Journal of Applied Ecology 48: 1381–1390. Phillips AJ, Anderson VL, Robertson EJ, Secombes CJ, van West P. 2008. New insights into animal pathogenic oomycetes. Trends in Microbiology 16: 13–19. Pimentel D, Lach L, Zuniga R, Morrison D. 2000. Environmental and economic costs of nonindigenous species in the United States. BioScience 50: 53–65. Van der Plaats-Niterink A J. 1981. Monograph of the genus Pythium. Studies in Mycology 21: 1–244. Plissonneau C, Stürchler A. 2016. The evolution of orphan regions in genomes of a fungal pathogen of wheat. mBio 7: e01231-16. Pojar J, Klinka K, Meidinger D V. 1987. Biogeoclimatic ecosystem classification in British 179  Columbia. Forest Ecology and Management 22: 119–154. Van Poucke K, Franceschini S, Webber JF, Vercauteren A, Turner JA, McCracken AR, Heungens K, Brasier CM. 2012. Discovery of a fourth evolutionary lineage of Phytophthora ramorum: EU2. Fungal Biology 116: 1178–1191. Prigigallo MI, Abdelfattah A, Cacciola SO, Faedda R, Sanzani SM, Cooke DEL, Schena L. 2015. Metabarcoding analysis of Phytophthora diversity using genus-specific primers and 454 pyrosequencing. Phytopathology 106: 305–313. Prigigallo MI, Abdelfattah A, Cacciola SO, Faedda R, Sanzani SM, Cooke DEL, Schena L. 2016. Metabarcoding analysis of Phytophthora diversity using genus-specific primers and 454 pyrosequencing. Phytopathology 106: 305–313. Prospero S, Grünwald NJ, Winton LM, Hansen EM. 2009. Migration patterns of the emerging plant pathogen Phytophthora ramorum on the west coast of the United States of America. Phytopathology 99: 739–749. Prospero S, Hansen EM, Grünwald NJ, Winton LM. 2007. Population dynamics of the sudden oak death pathogen Phytophthora ramorum in Oregon from 2001 to 2004. Molecular Ecology 16: 2958–2973. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, Maller J, Sklar P, de Bakker PIW, Daly MJ, et al. 2017. PLINK: A tool set for whole-genome association and population-based linkage analyses. The American Journal of Human Genetics 81: 559–575. Quinlan AR, Hall IM. 2010. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26: 841–842. Qutob D, Tedman-Jones J, Dong S, Kuflu K, Pham H, Wang Y, Dou D, Kale SD, Arredondo FD, Tyler BM, et al. 2009. Copy number variation and transcriptional polymorphisms of Phytophthora sojae RXLR effector genes Avr1a and Avr3a (FM Ausubel, Ed.). PLoS ONE 4: e5066. R Core Team. 2017. R: A language and environment for statistical computing. Raffaele S, Farrer R a, Cano LM, Studholme DJ, MacLean D, Thines M, Jiang RHY, Zody MC, Kunjeti SG, Donofrio NM, et al. 2010a. Genome evolution following host jumps in the Irish potato famine pathogen lineage. Science 330: 1540–1543. Raffaele S, Kamoun S. 2012. Genome evolution in filamentous plant pathogens : why bigger can be better. Nature Reviews Microbiology 10: 417–430. Raffaele S, Win J, Cano LM, Kamoun S. 2010b. Analyses of genome architecture and gene expression reveal novel candidate virulence factors in the secretome of Phytophthora infestans. BMC Genomics 11: 637. Rambaut A, Lam TT, Max Carvalho L, Pybus OG. 2016. Exploring the temporal structure of heterochronous sequences using TempEst (formerly Path-O-Gen). Virus Evolution 2: vew007. Ramsfield TD, Bentz BJ, Faccoli M, Jactel H, Brockerhoff EG. 2016. Forest health in a changing world: Effects of globalization and climate change on forest insect and pathogen impacts. Forestry: An International Journal of Forest Research 89: 245–252. Ravenhall M, Škunca N, Lassalle F, Dessimoz C. 2015. Inferring horizontal gene transfer. 180  PLOS Computational Biology 11: e1004095. Redondo MA, Boberg J, Stenlid J, Oliva J. 2018. Functional traits associated with the establishment of introduced Phytophthora spp . in Swedish forests. Journal of Applied Ecology: 1–15. Reeser P, Sutton W, Hansen E. 2010. Phytophthora species in tanoak trees, canopy drip, soil, and streams in the sudden oak death epidemic area in southwestern Oregon. The Fifth Meeting of the International Union of Forest Research Organizations (IUFRO), Working Party 7.02.09. Reeser PW, Sutton W, Hansen EM, Remigi P, Adams GC. 2011. Phytophthora species in forest streams in Oregon and Alaska. Mycologia 103: 22–35. Reinhart KO, Tytgat T, Van der Putten WH, Clay K. 2010. Virulence of soil-borne pathogens and invasion by Prunus serotina. New Phytologist 186: 484–495. Rice P, Longden I, Bleasby A. 2000. EMBOSS: The european molecular biology open software suite. Trends in Genetics 16: 276–277. Rieseberg LH, Archer MA, Wayne RK. 1999. Transgressive segregation, adaptation and speciation. 83: 363–372. Riit T, Tedersoo L, Drenkhan R, Runno-Paurson E, Kokko H, Anslan S. 2016. Oomycete-specific ITS primers for identification and metabarcoding. MycoKeys 14: 17–30. Rizzo DM, Garbelotto M. 2003. Sudden oak death: Endangering California and Oregon forest ecosystems. Frontiers in Ecology and the Environment 1: 197–204. Rizzo DM, Garbelotto M, Davidson JM, Slaughter GW, Koike ST. 2002. Phytophthora ramorum as the cause of extensive mortality of Quercus spp. and Lithocarpus densiflorus in California. Plant Disease 86: 205–214. Robertson GI. 1973. Pathogenicity of Pythium spp. to seeds and seedling roots. New Zealand Journal of Agricultural Research 16: 367–372. Rollins L, Coats K, Elliott M, Chastagner G. 2016. Comparison of five detection and quantification methods for Phytophthora ramorum in stream and irrigation water. Plant Disease 100: 1202–1211. Rouxel T, Grandaubert J, Hane JK, Hoede C, van de Wouw AP, Couloux A, Dominguez V, Anthouard V, Bally P, Bourras S, et al. 2011. Effector diversification within compartments of the Leptosphaeria maculans genome affected by Repeat-Induced Point mutations. Nature Communications 2: 202. Sakai AK, Allendorf FW, Holt JS, Lodge DM, Molofsky J, With KA, Baughman S, Cabin RJ, Cohen JE, Ellstrand NC, et al. 2001. The population biology of invasive species. Annual Review of Ecology and Systematics 32: 305–332. Salinas NR, Little DP. 2014. 2matrix: A utility for indel coding and phylogenetic matrix concatenation. Applications in Plant Sciences 2: apps.1300083. Santini A, Ghelardini L, De Pace C, Desprez-Loustau ML, Capretti P, Chandelier A, Cech T, Chira D, Diamandis S, Gaitniekis T, et al. 2013. Biogeographical patterns and determinants of invasion by forest pathogens in Europe. New Phytologist 197: 238–250. 181  Sapkota R, Nicolaisen M. 2015. An improved high throughput sequencing method for studying oomycete communities. Journal of Microbiological Methods 110: 33–39. Sárándi-Kovács J, Nagy L, Lakatos F, Sipos G. 2016. Sudden Phytophthora dieback of wild cherry trees in Northwest Hungary. Acta Silvatica et Lignaria Hungarica 12: 117. Saunders DGO, Win J, Kamoun S, Raffaele S. 2014. Two-dimensional data binning for the analysis of genome architecture in filamentous plant pathogens and other Eukaryotes - Plant-pathogen interactions: Methods and protocols. In: Birch P,, In: Jones JT,,  In: Bos JIB, eds. Totowa, NJ: Humana Press, 29–51. Savory F, Leonard G, Richards TA. 2000. The role of horizontal gene transfer in the evolution of fungi. Annual Review of Phytopathology 38: 325–363. Schlenzig A, Campbell R, Chard J. 2015. Phytophthora species infecting hardy ornamentals in nurseries and the managed environment in Scotland. Journal of Phytopathology 163: 686–689. Schlenzig A, Campbell RB, Roberts AMI. 2017. The susceptibility of selected conifer foliage to infection with Phytophthora lateralis. Forest Pathology 47: e12333–n/a. Schmieder R, Edwards R. 2011. Quality control and preprocessing of metagenomic datasets. Sciences-New York: 3–5. Schornack S, van Damme M, Bozkurt TO, Cano LM, Smoker M, Thines M, Gaulin E, Kamoun S, Huitema E. 2010. Ancient class of translocated oomycete effectors targets the host nucleus. Proceedings of the National Academy of Sciences  107: 17421–17426. Schoustra SE, Debets AJM, Slakhorst M, Hoekstra RF. 2007. Mitotic recombination accelerates adaptation in the fungus Aspergillus nidulans. PLoS Genetics 3: e68. Schroeder KL, Martin FN, de Cock AWAM, Lévesque CA, Spies CFJ, Okubara PA, Paulitz TC. 2013. Molecular detection and quantification of Pythium species: Evolving taxonomy, new tools, and challenges. Plant Disease 97: 4–20. Scott P, Burgess T, Hardy G. 2013. Globalization and Phytophthora. In: Lamour K, ed. Phytophthora: A Global Perspective. CABI, 226–232. Seidl MF, Van den Ackerveken G, Govers F, Snel B. 2012. Reconstruction of Oomycete genome evolution identifies differences in evolutionary trajectories leading to present-day large gene families. Genome Biology and Evolution 4: 199–211. Shishkoff N. 2007. Persistence of Phytophthora ramorum in soil mix and roots of nursery ornamentals. Plant Disease 91: 1245–1249. Shrestha SK, Zhou Y, Lamour K. 2013. Oomycetes baited from streams in Tennessee 2010–2012. Mycologia 105: 1516–1523. Simão FA, Waterhouse RM, Ioannidis P, Kriventseva E V., Zdobnov EM. 2015. BUSCO: Assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics 31: 3210–3212. Simpson JT, Wong K, Jackman SD, Simpson JT, Durbin R, Salzberg SL, Phillippy AM, Zimin A, Simpson JT, Wong K, et al. 2012. ABySS : A parallel assembler for short read sequence data structures assemblies and assembly algorithms. Genome Research 19: 1117–1123. 182  Sims LL, Sutton W, Reeser P, Hansen EM. 2015. The Phytophthora species assemblage and diversity in riparian alder ecosystems of western Oregon, USA. Mycologia 107: 889–902. Snel B, Bork P, Huynen MA. 1999. Genome phylogeny based on gene content. Nature Genetics 21: 108–110. Soanes D, Richards T a. 2014. Horizontal gene transfer in eukaryotic plant pathogens. Annual review of phytopathology 52: 583–614. Sperschneider J, Gardiner DM, Dodds PN, Tini F, Covarelli L, Singh KB, Manners JM, Taylor JM. 2016. EffectorP: Predicting fungal effector proteins from secretomes using machine learning. New Phytologist 210: 743–761. Stamatakis A, Hoover P, Rougemont J. 2008. A rapid bootstrap algorithm for the RAxML Web servers. Systematic biology 57: 758–771. Stamler RA, Sanogo S, Goldberg NP, Randall JJ. 2016. Identification of Phytophthora species in rivers and streams of the Southwestern United States. Applied and Environmental Microbiology 82: 4696–4704. Stewart JE, Kroese D, Tabima JF, Larsen MM, Fieland VJ, Press CM, Zasada IA, Grünwald NJ. 2014. Pathogenicity, fungicide resistance, and genetic variability of Phytophthora rubi isolates from raspberry (Rubus idaeus) in the western United States. Plant Disease 98: 1702–1708. Stipes RJ, Campana RJ (Eds.). 1981. Brasier Clive M,. Compendium of Elm diseases. St Paul, Minnesota: American Phytopathological Society, 76–79. Stoletzki N, Eyre-walker A. 2011. The positive correlation between dN/dS and dS in mammals is due to runs of adjacent substitutions. Molecular Biology and Evolution 28: 1371–1380. Strenge D, Elliott M, Chastagner G, Sclar C, Stern D. 2016. Managing Phytophthora ramorum at the Bloedel Reserve. Proceedings of the Sudden Oak Death Sixth Science Symposium.43–44. Stukenbrock EH. 2013. Natural selection and genome plasticity in play: a genomic view on speciation in fungal plant pathogens. New Phytologist: 895–907. Stukenbrock EH. 2016. The role of hybridization in the evolution and emergence of new fungal plant pathogens. Phytopathology 106: 104–112. Stukenbrock EH, Bataillon T. 2012. A population genomics perspective on the emergence and adaptation of new plant pathogens in agro-ecosystems. PLoS Pathogens 8: 1–4. Stukenbrock EH, McDonald BA. 2008. The origins of plant pathogens in agro-ecosystems. Annual Review of Phytopathology 46: 75–100. Sung JM, Jin KS, Park JS. 1983. Identification and pathogenicity of Pythium spp. associated with seedling damping off of rice. Korean Journal of Mycology 11: 27–32. Sutton W, Hansen EM, Reeser P, Kanaskie A. 2008. Comparing Phytophthora ramorum diagnostic protocols for the national sudden oak death stream monitoring program. Proceedings of the Sudden Oak Death Third Science Symposium. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, Pages 461-466. 183  Tajima F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585–595. Thompson JD, Higgins DG, Gibson TJ. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673–4680. Tischfield JA. 1997. Loss of heterozygosity or: how I learned to stop worrying and love mitotic recombination. American Journal of Human Genetics 61: 995–999. Torto-Alalibo T, Tian M, Gajendran K, Waugh ME, van West P, Kamoun S. 2005. Expressed sequence tags from the oomycete fish pathogen Saprolegnia parasitica reveal putative virulence factors. BMC Microbiology 5: 46. Torto T, Li S, Styer A, Huitema E, Testa A, Gow NA, van West P, Kamoun S. 2003. EST mining and functional expression assays identify extracellular effector proteins from plant pathogen Phytophthora. Genome Res 13: 1675–1685. Tyler BM. 2002. Molecular basis of recognition between Phytophthora pathogens and their hosts. Annual Review of Phytopathology 40: 137–167. Tyler BM, Tripathy S, Zhang X, Dehal P, Jiang RHY, Aerts A, Arredondo FD, Baxter L, Bensasson D, Beynon JL, et al. 2006. Phytophthora genome sequences uncover evolutionary origins and mechanisms of pathogenesis. Science 313: 1261–1266. Vannini A, Bruni N, Tomassini A, Franceschini S, Vettraino AM. 2013. Pyrosequencing of environmental soil samples reveals biodiversity of the Phytophthora resident community in chestnut forests. FEMS Microbiology Ecology 85: 433–442. Venables WN, Ripley BD. 2002. Modern Applied Statistics with S. New York, NY.: Springer. Vercauteren A, De Dobbelaere I, Van Bockstaele E, Maes M, Heungens K. 2011. Genotypic and phenotypic characterization of the European A2 isolates of Phytophthora ramorum. European Journal of Plant Pathology 129: 621–635. Vercauteren A, De Dobbelaere I, Grünwald NJ, Bonants P, Van Bockstaele E, Maes M, Huengens K. 2010. Clonal expansion of the Belgian Phytophthora ramorum populations based on new microsatellite markers. Molecular Ecology 19: 92–107. Wang Q, Garrity GM, Tiedje JM, Cole JR. 2007. Naïve bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology 73: 5261–5267. Wernersson R, Pedersen AG. 2003. RevTrans: multiple alignment of coding DNA from aligned amino acid sequences. Nucleic Acids Research 31: 3537–3539. Werres S, Marwitz R, Man In’t veld WA, De Cock AWAM, Bonants PJM, De Weerdt M, Themann K, Ilieva E, Baayen RP. 2001. Phytophthora ramorum sp. nov., a new pathogen on Rhododendron and Viburnum. Mycological Research 105: 1155–1165. Weste G, Marks GC. 1987. The biology of Phytophthora cinnamomi in australasian forests. Annual Review of Phytopathology 25: 207–229. Wheeler TJ, Eddy SR. 2013. nhmmer: DNA homology search with profile HMMs. Bioinformatics 29: 2487–2489. 184  Wickland AC, Jensen CE, Rizzo DM. 2008. Geographic distribution, disease symptoms and pathogenicity of Phytophthora nemorosa and Phytophthora pseudosyringae in California, USA. Forest Pathology 38: 288–298. Wills RT. 1993. The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18: 145–159. Win J, Morgan W, Bos J, Krasileva K V, Cano LM, Chaparro-Garcia A, Ammar R, Staskawicz BJ, Kamoun S. 2007. Adaptive evolution has targeted the C-Terminal domain of the RXLR effectors of plant pathogenic Oomycetes. The Plant Cell 19: 2349–2369. Winton LM, Hansen EM. 2001. Molecular diagnosis of Phytophthora lateralis in trees, water, and foliage baits using multiplex polymerase chain reaction. Forest Pathology 31: 275–283. With KA. 2002. The landscape ecology of invasive spread. Conservation Biology 16: 1192–1203. Yakabe LE, Blomquist CL, Thomas SL, MacDonald JD. 2009. Identification and frequency of Phytophthora species associated with foliar diseases in California ornamental nurseries. Plant Disease 93: 883–890. Yang Z, Nielsen R. 1998. Synonymous and nonsynonymous rate variation in nuclear genes of mammals. Journal of Molecular Evolution 46: 409–418. Yoshida K, Schuenemann VJ, Cano LM, Pais M, Mishra B, Sharma R, Lanz C, Martin FN, Kamoun S, Krause J, et al. 2013. The rise and fall of the Phytophthora infestans lineage that triggered the Irish potato famine. eLife 2013: 1–25. Yu G, Smith DK, Zhu H, Guan Y, Lam TT-Y. 2017. ggtree: An R package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods in Ecology and Evolution 8: 28–36. Zdobnov EM, Apweiler R. 2001. InterProScan--an integration platform for the signature-recognition methods in InterPro. Bioinformatics 17: 847–848.    185  Appendices Appendix A: Supplementary tables chapter 2 Table S2.7: Sequencing, filtering and taxonomy assignment details for the two sampling periods.  Fall Spring Number of samples 124 122 Total number of 454 reads 620355 760186 Filtering   sequences too short 158917 193267 Poor quality 74118 93666 Ambiguous sequences and homopolymers 874 2563 primer or barcode mismatch 52518 36927 Filtered from ITSx 42201 74939 Number of sequences after quality  filtering 291727 247053 Singletons 663 412 Number of seqs minimum count 2 291064 246641 Descriptive data   Range of sequences per sample 12 to 7273 15 to 5592 Number of OTUs (total) 1311 1692 number of OTUs not assigned 138 99 Stramenopile reads 285746 240541 Stramenopile OTUs 1173 1593 Oomycete reads 266609 208128 Oomycete reads (minimum reads 2) 92 84 Oomycete OTUs 609 1015 Range of Oomycete sequences per sample 12 to 7273 0 to 5560 Mean number of Oomycete sequences per sample 2150 1706 Range of Oomycete OTUs per sample 2 to 78 0 to 58 Mean number of Oomycete OTUs per sample 24 22 Number of Oomycete OTUs identified to Genus 427 710 Number of Oomycete OTUs identified to Species 319 468   186  Table S2.8: OTU species level taxonomic identifications using manual identification and automated identification with clustering at 97% similarity.  OTUs identified with both manual and automated methods Addition OTUs from manual taxonomic identification Addition OTUs from automated taxonomic identification Phytophthora amnicola Phytophthora Unknown_Clade9_large_group_parsiana_like Phytophthora sp_xHennops Phytophthora cactorum Phytophthora cryptogea Phytophthora alni Phytophthora chlamydospora Phytophthora gregata Phytophthora bilorbang Phytophthora europaea Phytophthora intercalaris Phytophthora cambivora Phytophthora rubi/fragariae Phytophthora megasperma Phytophthora idaei Phytophthora gallica Phytophthora pini/citricola Phytophthora irrigata Phytophthora gonapodyides Phytophthora taxon_kelmania Phytophthora mississippiae Phytophthora hibernalis Phytophthora taxon_Oaksoil Phytophthora Other Phytophthora hydropathica unknown sp_1 Phytophthora s_ Phytophthora inundata unknown sp_2 Phytophthora sp_ohioensis Phytophthora lacustris unknown sp_3 Phytophthora sp_raspberry Phytophthora lateralis  Phytophthora sp_Taxodium_1 Phytophthora multivora  Phytophthora uliginosa Phytophthora nemorosa   Phytophthora plurivora   Phytophthora polonica   Phytophthora pseudosyringae   Phytophthora pseudotsugae   Phytophthora quercina   Phytophthora ramorum   Phytophthora sansomeana   Phytophthora siskiyouensis   Phytophthora syringae     187  Appendix B: Chapter 4, supplementary material  DNA extraction, genome sequencing, SNP extraction and filtering Methods DNA extraction. DNA was extracted from P. ramorum isolates cultured on potato dextrose agar (PDA) media at 20°C under a 12-hour photoperiod for 7 to 10 days. We used a chloroform method (Möller et al., 1992) with the following modification: tissue was collected from mycelia grown on a cellophane membrane (GE Healthcare Bio-Sciences Corp., Piscataway Township, NJ) placed on top of the agar. The mycelia was ground and placed in a 2ml tube with 500µl of TES Buffer and 4ul of Proteinase K (20mg/ml) and incubated for 30 min at 55°C, with occasional gentle mixing. One hundred and forty microlitres NaCl (5M) and 65µl 10% CTAB was added and samples were incubated for 10 min at 65°C. Seven hundred microlitres chloroform:isoamyl alcohol (24:1) was added and samples were vortexed and centrifuged for 5 min at 13 000rpm. The supernatant was transferred to a fresh 1.5ml tube and 3µl RNase (10mg/ml) was added. Samples were incubated for 30 min at 37°C, centrifuged for 1 min at 13 000 rpm. Supernatant was transferred to a fresh 1.5ml tube and 150µl ammonium acetate (3M) and 500µl isopropanol was added. Samples were placed on ice for 30 min and centrifuged 5 min at 13 000 rpm, 4°C. The supernatant was aspirated off and the pellet washed with cold 70% EtOH (~250µl) and centrifuged for 5 min at 13 000 rpm, 4°C. The pellet was air dried for 10 min and suspended in 50µl TE buffer. DNA concentration was quantified with a Qubit Fluorometer (Life Technologies Inc., Grand Island, NY). 188   Genome Sequencing. Genome sequencing was done at Canada's Michael Smith Genome Sciences Centre or GSC (Vancouver Canada). Libraries were constructed using one of two methods (method for each isolate in Table S4.1). The majority of the libraries were constructed on a SPRI-TE robot (Beckman-Coulter, USA) according to the manufacturer’s instructions (SPRIworks Fragment Library System I Kit, A84801). Briefly, 1ug of genomic DNA in a 60uL volume, and 96-well format, was fragmented by Covaris E210 sonication for 30 sec using a “Duty cycle” of 20% and “Intensity” of 5. Up to 10 paired-end genome sequencing libraries were prepared in parallel using the SPRI-TE 300-600bp size-selection program. Following completion of the SPRI-TE run the adapter ligated library templates were quantified using a Qubit fluorometer. Five ng of adapter ligated template was PCR amplified using Phusion DNA Polymerase (Thermo Fisher Scientific Inc. USA) and Illumina’s PE indexed primer set, with cycle conditions: 98˚C for 30 sec followed by 10 cycles of 98˚C for 15 sec, 62˚C for 30 sec and 72˚C for 30 sec, and a final amplicon extension at 72˚C for 5min.  For the remaining samples, 2ug of genomic DNA in a 96-well format was fragmented by Covaris E210 sonication for 30 sec using a “Duty cycle” of 20% and “Intensity” of 5. The paired-end sequencing library was prepared following the BC Cancer Agency’s Genome Sciences Centre 96-well Genomic ~350bp-450bp insert Illumina Library Construction protocol on a Biomek FX robot (Beckman-Coulter, USA). Briefly, the DNA was purified in a 96-well microtitre plate using Ampure XP SPRI beads (40-45uL beads per 60µL DNA), and was subject to end-repair, and phosphorylation by T4 DNA polymerase, Klenow DNA Polymerase, and T4 polynucleotide kinase respectively in a single reaction, followed by cleanup using Ampure XP SPRI beads and 189  3’ A-tailing by Klenow fragment (3’ to 5’ exo minus). After cleanup using Ampure XP SPRI beads, picogreen quantification was performed to determine the amount of Illumina PE adapters used in the next step of adapter ligation reaction. The adapter-ligated products were purified using Ampure XP SPRI beads, then PCR-amplified with Phusion DNA Polymerase (Thermo Fisher Scientific Inc. USA) using Illumina’s PE indexed primer set, with cycle conditions: 98˚C for 30 sec followed by 6 cycles of 98˚C for 15 sec, 62˚C for 30 sec and 72˚C for 30 sec, and a final extension at 72˚C for 5 min.  For both library construction methods, PCR products were purified using Ampure XP SPRI beads, and quantified with Caliper LabChip GX for DNA samples using the High Sensitivity Assay (PerkinElmer, Inc. USA). PCR products of desired size range were purified using gel electrophoresis (8% PAGE or 1.5% Metaphor agarose gels in a custom built robot) and DNA quality was assessed and quantified using an Agilent DNA 1000 series II assay and Quant-iT dsDNA HS Assay Kit using Qubit fluorometer (Invitrogen), then diluted to 8nM. The final concentration was verified by Quant-iT dsDNA HS Assay prior to Illumina Sequencing. For sequencing, 100bp paired-end tagged (PET) reads were prepared. Twelve individuals were pooled per lane and sequencing was performed on the Illumina HiSeq 2000 (Illumina Inc). Reads were provided in BAM format for each individual.  Mapping and filtering reads. PRINSEQ v0.20.3 (Schmieder & Edwards, 2011) was used to filter the bam files for redundant reads, reads containing one or more ‘N’s, and to trim low quality bases from the ends of the reads. Trimming was done by discarding the last ten nucleotides if the average quality fell below 20 (average quality was calculated on windows of 190  10bp). Trimmed reads were mapped onto the P. ramorum reference genome (Pr-102 isolate) (Tyler et al., 2006) available through the Department of Energy-Joint Genome Institute (DOE-JGI, Walnut Creek USA) (version 1.1) with the Burrows-Wheeler Aligner (BWA) (Li & Durbin, 2009) using the default parameters. Mapping coverage, the percent of reads mapped and mapping quality statistics were calculated using Qualimap v.0.7.1 (García-Alcalde et al., 2012).  Extracting SNPs and filtering. Variant sites were obtained in a mulitvcf format using the mpileup program of Samtools and the Bayesian variant calling models implemented in BCFtools (Li et al., 2009). Single nucleotide polymorphisms (SNPs) were filtered using VCFtools (Danecek et al., 2011) with the parameters determined from testing to minimize false positives and maximize true positives (Appendix B, Testing SNP filtering parameters and false call rates); a maximum mean depth of 90 and a minimum mean depth of 10, and for each SNP, a minimum quality of 30, a minimum depth of 10, a minimum mapping quality of 30, and a minimum distance to a gap of 10bp. For each genotype, we set a minimum depth of 4 reads, a maximum of 240, and a minimum genotype quality of 20. We then applied a filter for a minimum allele count of two (an allele needed to be found in at least two individuals in the data set) and for no missing data.  Testing SNP filtering parameters and false call rates. To set the final vcf filtering parameters, various values for each parameter were applied to a subset of genomes for 92 isolates using VCFtools (Danecek et al., 2011). The resulting SNP datasets were screened for a set of true positive SNPs and false positive SNPs. For true positive SNPs, Sanger sequenced SNPs from previous studies (Bilodeau, 2008; Goss et al., 2009) were used. For false positives, genes with 191  close paralogs in the JGI reference genome (NA1 lineage) were aligned to their paralogs, and differing sites between the two genes were used as heterozygous false positives. NA1 isolates were then screened for the false positives. For a maximum depth, we tested a mean maximum depth filter (the mean coverage of all individuals) set at just below the mean average mapping coverage of 92 genomes, just above the mean average coverage and approximately two times the mean average coverage. We then tested SNP filters for minimum coverage, minimum variant quality, minimum mapping quality, and distance of the SNP to a gap using VCF-annotate from the VCFtools package (Danecek et al., 2011). Finally we tested genotype filters for minimum and maximum depth at an individual SNP, minimum genotype quality, and a site filter for minimum average coverage (using the mean coverage of all individuals). Mean minimum and maximum site filters allowed us to accept SNPs with low or high coverage in some individuals provided the mean values did not go below or above the set threshold. The number of SNPs filtered, the number of true positive SNPs found, and the number of false calls was compared for the different parameters tested to choose a combination of filters that maximized the number of true positives found and minimized the number of false calls.  To assess the rate of false SNPs generated during the sequencing process, one individual was sequenced twice and one individual three times in independent runs using the same conditions for library preparation and sequencing. These individuals were trimmed and mapped to the reference as above and SNPs extracted for each set of isolates relative to the JGI reference using the mpileup function in Samtools as above. The resulting BCF file was converted to a VCF and the total number of discordant SNPs between duplicate individuals in independent runs was obtained 192  for the raw data set. The VCF was filtered following the above parameters and the total number of discordant SNPs relative to the total number of SNPs was used to calculate a false call rate.  SNP counts and population diversity. To compare SNP content between lineages, for each SNP in the dataset, we first determined if a given SNP was fixed in all isolates within a lineage (but different between lineages), or if it was polymorphic in some of the individuals in a lineage. We then determined if the SNP was homozygous or heterozygous for each of the two categories (fixed and polymorphic).  Results and discussion Genome Sequencing and mapping statistics. The P. ramorum reference genome (Pr-102 ; NA1 lineage) sequenced at the DOE-JGI was estimated at approximately 65Mbp with 15 743 genes (Tyler et al., 2006). For the 107 isolates re-sequenced in this project, we achieved an average mapping coverage of 49.4X with a range of 16.7X to 80.9X. The average percent of the total sequenced reads that mapped to the DOE-JGI reference was 92.2% (ranged from 62.6% to 95.9%; however, all isolates except one were above 85%). Mapping of isolates from the different lineages to the NA1 reference genome showed similar ranges in the percent of reads mapped with the highest mapping occurring in the NA1 lineage. The EU2 isolates had the least variance in the percent of reads mapped, however the maximum percent of reads that mapped to the NA1 reference was lower (NA1: 62.6 to 95.9%; EU1: 85.8 to 94.9%; NA2: 87.2 to 94.8%; EU2: 92.7 to 92.9%).  193  SNP extraction and filtering. The total number of SNPs extracted from the initial set of 92 isolates used to test filter parameters was 912 622 (prior to filtering). No false positive validation SNPs were found (out of the set of false SNPs created from close paralogs) prior to or after applying any of the filters (Table S4.12). At all of the depth filtering parameters nearly twice the average coverage (mean depth filters 80, 90 and 100; mean depth 49.4X; Table S4.12), greater than 90% of the Sanger sequenced validation SNPs were found and only one unexpected, possibly false positive was found (TT expected and CA found). The only other discrepancies between the Sanger data and the Illumina data were cases where a SNP was called with  one allele difference from the expected (TT expected, CT found) which could be due to either a false call or a new variant. The final set of filtering parameters used were a minimum average SNP coverage over all individuals of 10 reads, with a minimum of four reads covering each SNP for each individual;  a maximum average SNP coverage over all individuals of 90 reads with no individual exceeding a coverage of 240X at a SNP; a minimum site quality of 30 with each individual reaching a minimum quality of 20; a minimum mapping quality of 30 around each SNP; and a minimum distance of 10 bases between each SNP and the closest gap. With this set of filtering parameters, 434 504 SNPs remained in the set of 92 test isolates (431 112 loci excluding singleton SNPs). Ninety-seven percent of the validation SNPs were found and 96.5% were the expected call. For the full data set of 107 isolates, the total number of variants including indels prior to filtering was 1 085 718. After filtering there was a total of 895 274 SNPs. Excluding singletons and missing data, there were a total of 485 327 bi-allelic sites.   The rate of false calls between independent runs of the same individuals ranged from 4.6% (two runs) to 8.3% (three runs). After applying the final filtering parameters, this rate dropped to 0.7% 194  to 1.8% (Table S4.13). The false call rate dropped to less than 0.2% when the SNP sites that were filtered out of the main data set for missing data and singletons were removed from the controls.  From these results, the final filtering parameters were chosen (above) and the data set with a minimum allele count of two (no singleton SNPs) and no missing data were used in the population analyses.  SNP counts and lineage genetic diversity. The majority of the SNPs discovered by mapping the Illumina reads onto the reference NA1 genome were between the four clonal lineages rather than within, where the individuals within each clonal lineage were fixed for the alleles (either homozygous (72% to 80%) or heterozygous (20% to 30%)) (Table S4.3). The fixed heterozygous loci accounted for 81% to 92% of the total heterozygous loci in each lineage (% fixed heterozygous loci = 88% for EU1, 91% for NA1, 81% for NA2 and 92% for EU2). Genome wide SNP profiles of the isolates within lineages were very similar. The majority of the SNPs that were polymorphic (varied between the individuals within a lineage) were heterozygous, and the highest number was found in the NA2 lineage (Table S4.3). Nucleotidic diversity was calculated for each lineage and was highest in the EU1 lineage and lowest in the EU2 lineage (Pi = 0.00116 for EU1, 0.00108 for NA1, 0.00108 for NA2 and 0.00077 for EU2).  De novo genome assemblies Methods Two representative isolates of each of the four clonal lineages was selected for de novo genome assembly. For each isolate,  Illumina sequencing reads were filtered by trimming the barcode 195  sequence using PRINSEQ v0.20.3 (Schmieder & Edwards, 2011) and assembled into contigs and pre-scaffolded using ABySS (Simpson et al., 2012) with k-values ranging from 32 to 96. Final scaffolding was completed with SSPACE v.3.0 (Hunt et al., 2014) and the best assembly was then selected based on genome size (~50.0Mb) and contiguity (best N50 and length of longest scaffold). Completeness of the de novo-genome assemblies was assessed using BUSCO (Benchmarking Universal Single-Copy Orthologs) (Simão et al., 2015). The de novo-assemblies were then repeat masked with RepeatMasker (A.F.A. Smit, R. Hubley & P. Green RepeatMasker at using 2,101 Phytopthora spp. repeats available in Repbase (Jurka et al., 2005) and annotated using AUGUSTUS ver. 2.7 (Keller et al., 2011) trained with models from P. sojae, P. infestans and P. ramorum available on the JGI server (Grigoriev et al., 2012). Predicted protein models were functionally annotated with Blast2GO (Conesa et al., 2005) following homolog searches using BLASTp (Altschul et al., 1990) against NR (E-value cut-off of 1e-05) and a protein domain search using Interproscan (Zdobnov & Apweiler, 2001).  Results and discussion Genome assembly statistics were in the range of those obtained for other Phytophthora and Oomycete de novo assemblies generated from short Illumina re-sequencing reads; completeness reached 82% of the BUSCO eukaryotic set, a value similar to those obtained for other Oomycete de novo assemblies and slightly higher than the 81.6% value obtained for the JGI Pr-102 genome assembly (Tyler et al., 2006; Feau et al., 2016) indicating these Illumina de novo assemblies can be considered complete(Table S4.14).   196  Phylogenetic analysis Methods Neighbor joining phylogenetic reconstruction. To reconstruct the phylogenetic relationships of the isolates using genome wide SNP patterns, the VCF was converted to a Plink format using VCFtools (Danecek et al., 2011) and Plink v1.07 (Purcell et al., 2017) was used to recode the data for use in the package Adegenet (Jombart, 2008). Euclidean distances were calculated using the stats-package version 3.4.0 in R (R Core Team, 2017) and used to construct a neighbor-joining tree using the APE package (Paradis et al., 2004). Arlequin 3.0 (Excoffier et al., 2005) was used to calculate pairwise FST values between the lineages.  Maximum Likelihood phylogenetic reconstruction. To assess evolutionary dynamics on P. ramorum, we reconstructed the phylogenetic relationships of all the isolates using a selection of genes under total neutrality. We extracted the complete set of alternate gene calls for the 107 isolates. Data from homozygous isolates were used to phase the heterozygous isolates using Beagle4 (Browning & Browning, 2007). Alternate genome sequences were generated using the FastaAlternateReferenceMaker script from the GATK suite (McKenna et al., 2010), using the VCF file as the source of variants. We used BEDTools (Quinlan & Hall, 2010) to extract the genes from their genomic coordinates. Two Fasta formatted files per gene, one for each strand, were obtained for each of the isolates. Each alternate gene model was translated to its amino acid sequence. Each gene model and amino acid translation was aligned using RevTrans (Wernersson & Pedersen, 2003). RevTrans aligns the amino acid sequences and uses this alignment as a scaffold to construct the gene model alignment. After alignment, Tajima’s D was measured using 197  the package pegas (Paradis, 2010). We selected the gene models with a Tajima’s D value with a significance greater than 0.05 (genes under neutrality).  For each gene, models of nucleotide substitution were calculated in jModelTest 2 (Darriba et al., 2012) for 500 random neutral genes across the genome. All genes were concatenated and partitioned using 2matrix (Salinas & Little, 2014). The partitioned matrix was ran in RAxML 8.1.24 (Stamatakis et al., 2008) to reconstruct the phylogenetic tree using Maximum likelihood and node supports using non-parametric bootstrap resampling using the CIPRES Science gateway (Miller et al., 2010). One thousand bootstrap repetitions were used and the bipartition tree (best tree with bootstrap support values) was obtained. The tree was midpoint rooted to determine the most divergent clade inside P. ramorum.  Results and discussion Maximum likelihood phylogenetic reconstruction. In the phylogenetic reconstruction, a total of 5588 genes were found under neutrality, 2050 genes under positive selection and 1984 genes were non-segregating. The phylogenetic reconstruction using a subset of 500 neutral genes yielded a topology similar to the neighbor-joining tree (Fig. S4.1), where all clonal lineages were clustered and highly supported monophyletic clades. In the case of NA1, two subclades can be observed, but lack bootstrap support of this splitting event. In the phylogenetic analysis, the two strands of phased haplotypes are clustered as expected for an asexual species (the Meselson  effect; (Birky-Jr., 1996)), except for a few individuals, which were the same individuals forming clusters or diverging in the neighbor-joining analysis. These were isolates with runs of 198  homozygosity (ROH; Appendix B, Effects of ROH on genotype) which possessed two identical or near identical alleles for each gene.    Divergence time between lineages Methods Phylogenetic analysis of one-to-one orthologs. Sequences from the 4929 one-to-one ortholog clusters obtained for seven Phytophthora species, including P. lateralis and the four lineages of P. ramorum (Appendix B, OrthoMCL analysis and evolution of gene family size) were aligned with MAFFT (Katoh & Standley, 2013) and filtered for sequence alignments > 100 amino-acids with no gaps. The resulting 492 alignments were then submitted to the PAML package of the program CODEML, to estimate the ratio ω = dN/dS and test for positive selection by comparing the “site-specific” models M1a/ M2a and M7/M8 (Katoh & Standley, 2013) as described in (Alamouti et al., 2014). This resulted in 49 polymorphic alignments (≥4 SNPs between the P. ramorum lineages) with no evidence of positive selection that were retained. Each alignment was submitted to ProtTest v. 2.4 (Abascal et al., 2005) and then concatenated into different partitions according to the best protein evolution model fitted (DayHoff, WAG or JTT), resulting in a super-alignment of 13 879 amino-acids. A maximum likelihood tree search followed by a rapid bootstrap analysis with 1000 replicates was then performed under partitioned data mode with RAxML 8.0.12 (Stamatakis et al., 2008).  Divergence time between P. ramorum lineages. We used the Bayesian clock method implemented in BEAST v1.7.5 (Drummond et al., 2012) to estimate divergence times between 199  the four P. ramorum lineages under three distinct clock models with the alignment generated for the phylogenetic analysis. Analyses were run with the strict clock, random local clock and UCLD relaxed clock models used in (Matari & Blair, 2014). For each run, partitions were treated separately under WAG, JTT or DayHoff model depending on the protein evolution model fitted, and the RAxML tree obtained in the previous analysis was used as a user-specified starting tree; a Yule speciation process was assumed with a uniform distribution on the birthrate (0–100; initial value 0.01). The 3 clock models were set with prior parameters defined in (Matari & Blair, 2014). To calibrate the molecular clock analyses, normal distributions were specified for Phytophthora species of the phylogenetic clades seven and eight (P. sojae, P. cinnamomi, P. lateralis and P. ramorum; mean = 22.3 My, SD = 0.61), clades one and two (P. infestans, P. capsicii and P. parasitica; mean = 21.5 My, SD = 0.67) and the radiation of the Phytophthora genus (mean = 26.6 My, SD = 0.67), according to the results of the strict clock analysis of (Matari & Blair, 2014).  For each dataset, the three clock models were tested under three independent runs with 50 million generations each. LogCombiner v1.7.5. was used to combine log and tree files from the three independent runs, for each model. Tracer v1.6 ( was used to evaluate convergence, estimate the appropriate burn-in for each run, and calculate Bayes factors for model comparisons. Trees were visualized in FigTree v1.4 .2. (  200  Results and discussion Phylogenetic analysis. The RAxML phylogenetic analysis conducted on a 13 879 amino-acids alignment, resulted in a maximum likelihood tree with a likelihood value of –ln 83715.663913 and a tree topology congruent with Phytophthora phylogenies reconstructed from smaller sequence datasets (Blair et al., 2008; Matari & Blair, 2014). The tree topology is strongly supported with bootstrap values of 100% for six out of the eight nodes. The relationship among the four P. ramorum lineages was well supported, with values varying between 82% (node NA2 with NA1/EU1) and 92% (node NA1/EU1) (Fig. S4.7).  Divergence time estimates. The low coefficient of variation obtained from a root-to-tip analysis of the RaxML phylogenetic tree with TempEst V1.5 (Rambaut et al., 2016) indicated a low degree of deviation of this dataset from the strict molecular clock hypothesis. However, posterior distributions of parameters were consistent across all three runs, and clock/ucld parameter estimates consistent across the three models (Table S4.15). Similarly, node divergence estimates were consistent across the three models with mean values obtained under one model generally falling within the 95% confidence intervals obtained under the other two models (Table S4.16). Estimated time points suggest a divergence between P. lateralis and P. ramorum at 6.38 My (SD, 5.91-6.84) and divergence between the lineages within P. ramorum at 1.31 My (SD, 1.14-1.48) (EU2 from the three other lineages), 1.06 (0.91-1.20) (NA2) and 750,000 (0.61-0.88) years (split between EU1 and NA1) (Fig. S4.7; Table S4.16). These estimates are in the same order of magnitude as those obtained from nuclear sequence data, with reproductive isolation between P. lateralis and P. ramorum estimated between 1.5 and 5.4 My, and a divergence between the NA1, NA2 and EU1 lineages ranging from 0.5 to 0.17 My (Goss et al., 2009).  201   Genome-level copy number variation Methods Genome-level chromosomal copy number variation (CCNV) was detected by using two methods based on the relative number of mapped reads. First, we used the read depth approach implemented in Control-FREEC software for high resolution CCNV quantification. A mappability profile was generated from the P. ramorum reference genome (Tyler et al., 2006) with the gem-do-index and gem-mappability programs of the GEM library (Derrien et al., 2012). Copy-number was then estimated for each 100bp non-overlapping window of the reference genome by running Control-FREEC on the BWA-mapped reads (bam format) and the mappability profile files with the following parameters : window = 100, telocentromeric = 0, ploidy = 2 to 6n.   We also used the measure read-depth ratio of alleles at heterozygous sites as described in Farrer et al. (Farrer et al., 2013) and Kasuga et al. (Kasuga et al., 2016). For each re-sequenced genome, the SAMtools (Li et al., 2009) mpileup and Bcftools programs were run on each scaffold to generate allele counts at heterozygous loci. Only heterozygous loci with read depth ≥ 20 and mapping quality ≥ 30 were retained; loci with read ratios of heterozygous alleles equal or larger than eight were also excluded. Output files from both methods were then visualized by the way of a custom Python program (available upon request), providing plots of CCNV across scaffold and average ratio of reads of heterozygous alleles in sliding non-overlapping windows of 10Kb across scaffold. 202   Results and discussion Plasticity in chromosome numbers i.e. aneuploidy is a common feature of asexual micro-organism genomes (Normark et al., 2003). Polyploidy resulted in major evolutionary implications for some modern asexual lineages of the potato late blight pathogen Phytophthora infestans (Yoshida et al., 2013). Similarly, CCNV was recently reported in P. ramorum with some NA1 isolates showing atypical non wild-type colony morphology (nwt) isolated from oaks. Oaks were described as a dead-end host for P. ramorum, as sporulation of the pathogen from oak trunk cankers has never been reported. Based on their observations, Kasuga et al. (Kasuga et al., 2012) hypothesized that oak defense mechanisms could induce chromosomal aberrations resulting in the nwt phenotype change in P. ramorum. Similarly, the same authors observed CCNV and the nwt phenotype in EU1 isolates infecting the rare host Lawson cypress (Chamaecyparis lawsoniana).  We looked for CCNV and aneuploidy as a possible cause for the observed ROH in our isolates, as well as a potential driver of population diversity within the lineages.  We observed chromosomal differences in a few NA1 and EU1 isolates, with inferred ploidy varying from 2n to 5n depending on the scaffold analyzed (Table S4.4; Fig. S4.3). CCNV was not responsible for the ROH pattern; although two of the isolates with CCNV had ROH, it was not observed on the same scaffolds as the ROH was. In addition, no signs of aneuploidy were observed in isolates with ROH. Similar to ROH, CCNV could affect phenotype, in particular by increasing the expression of genes with increased copies. It could also relax the selection pressure on the extra gene copy allowing for adaptation to occur.  203  Effects of ROH on genotype Results and discussion Runs of homozygosity and population evolution. ROH was observed in individuals in all four lineages producing homozygous scaffolds in single individuals which were heterozygous in the general population (Fig. S4.2); lengths ranged from 60 to 339Kb in EU1, 76 to 120Kb in EU2, 131 to 236Kb in NA1 and 90 to 99Kb in NA2. In one EU1 individual, 13.5% of the loci that were heterozygous in the rest of the EU1 population were homozygous in the affected individual (Table S4.2). In comparison, 0.2% to 0.3% of the total SNP loci which were heterozygous in the general EU1 population were homozygous in individuals without the runs of homozygosity (Table S4.2).   In individual isolates, there were between 53 and 1587 genes in the affected scaffolds, and between 24 to 893 proteins with amino acid differences between the alleles. ROH resulted in a loss of approximately 14 to 880 alleles (Table S4.5).   The ROH pattern was slightly different in the NA2 lineage. In contrast to the other lineages, some of the ROHs in the NA2 lineage were often interrupted by heterozygous loci, or were not completely contiguous over the full region of the scaffold affected. Some isolates had short ROH alternating with heterozygous regions on affected scaffolds and in some cases this resulted in a general loss of heterozygosity (LOH) with no long ROH patterns. In addition, the NA2 lineage had more ROHs shared between all of the isolates than the other lineages.   204  There were also less allele differences resulting from ROHs in the individuals and scaffolds with non-contiguous ROHs in NA2, which could suggest that these are older ROHs. In one individual, Pram_04_20470, scaffolds 12 and 40 had 86 and 41 proteins respectively with amino acid differences between the alleles; however only 21 and 3 proteins were affected by ROH. In scaffold five, which comprised several ROHs or a general loss of heterozygosity shared by all NA2 individuals, there were 26 proteins (out of a total of 144 in the affected region; Table S4.5) with amino acid differences, about half of what the other lineages had on scaffold five (NA1 = 79, EU1 = 73, EU2 = 56). An older sexual recombination event, possibly in the source population, followed by asexual propagation could produce the observed pattern of non-continuous ROHs and high fixed heterozygosity.  GO terms associated with protein binding, transferase activity, oxidoreductase activity, hydrolase activity and RNA-directed DNA polymerase activity were the most highly ranked terms in the gene ontology analysis for genes affected by ROH (Table S4.17).   Amino acid differences in scaffold 7 ROH. In the shared ROH region on scaffold seven, 63 alleles (36%) with amino acid differences were lost in the eight EU1 isolates with ROH (Table S4.5). In the region affected in NA1 between 73 and 85 alleles were lost (using the other three lineages for comparison). Fifteen proteins had five or more amino acid differences between alleles in the affected region of EU1, and 20 proteins had five or more amino acid differences between alleles in the affected region of NA1 (Table S4.5). In one protein, one allele had a premature stop codon. Isolates with the ROH in EU1 had the allele with the stop codon mutation so may no longer have a functioning copy of this gene.  205   Enrichment of putative plant pathogenicity genes on scaffold 7. GO terms associated with lyase activity and hydrolase activity were enriched on scaffold seven (Table S4.6). Hydrolysis of glycosyl compounds and bonds were both GO terms associated with the carbohydrate active enzymes. There was also a moderate enrichment in terms associated with the breakdown of pectin (aspartyl esterase activity and pectinesterase activity) (Table S4.6).   Effects of ROH on phenotype Methods Growth of isolates with ROH in culture. Isolates were grown on 1.5% MEA prior to the experiment (Table S4.2). Agar plugs 5 mm in diameter were cut from the margin of the culture and placed onto the center of a petri plate containing 18ml of larch sapwood agar or Douglas-fir sapwood agar (Stipes & Campana, 1981). Sapwood agar was prepared by peeling twigs less than 0.5 cm in diameter from freshly fallen trees and homogenizing the tissues to obtain a sapwood pulp which was incorporated into the agar prior to autoclaving. Three replicates were made of each isolate on each agar type. Cultures were incubated in the dark at 25°C and growth was measured four, seven, 10 and 14 days post-inoculation. The number of spores present was counted at four, seven and 11 days on Douglas-fir agar and at seven and 11 days for larch agar, by observing the bottom of the petri plate under a dissecting microscope at two diagonal locations beginning from the inoculation point for days four and seven, and at four points 5 mm from the plug in each 90 degree direction for day 11.    206  Growth of isolates with ROH on Rhododendron leaves. Seventeen leaves of Rhododendron sp. showing no signs of disease were collected from each of three different plants. Leaves were surface sterilized in 10% bleach for 30 sec, rinsed in a sterile water bath for one minute, and in a second sterile water bath for an additional minute. Leaves were air dried and then pricked with a sterile pin on the left, abaxial surface, leaving an equal distance between the wound and the leaf edge, and between the wound and the mid-rib. A 7 mm diameter plug of mycelium, taken from a 7 day old culture grown on 15% V8-agar (15mls V8 juice, 2 g CaCO3, 15 g agar in 1 L water) was placed mycelium side down onto the wound of each leaf. Inoculated leaves were incubated in clean plastic bins lined with 200 g autoclaved vermiculite and 100ml sterile distilled water. Each bin was placed in a plastic bag, and incubated in a growth chamber at 20 °C in the dark. Area of each lesion was measured daily from day three to seven, and day 10 to 12, with a final measurement on day 17. Sterile water was added once on day five to moisten the vermiculite. Leaves were checked visually for any signs of sporulation.   Growth of isolates with ROH on Douglas-fir and larch logs. Three Douglas-fir and three Japanese Larch trees between 10 cm and 25 cm diameter were freshly felled from the Malcolm Knapp Research Forest in Maple Ridge, British Columbia. The logs were cut into 0.5 m to 0.7 m bolts and sealed with epoxy-resin (Intergard 740, International Paint, Houston, TX, USA). Four isolates were inoculated on each bolt by placing a mycelial plug under the bark at the depth of the cambium and sealing it with a cotton plug and aluminum foil secured with tape (Brasier & Kirk, 2001). The log bolts were incubated at room temperature in a plastic bag for eight weeks. The length and width at the longest and widest points of each lesion were measured four and eight 207  weeks post inoculation (wpi). Differences between the groups were compared with a Kruskal-Wallis test. Statistical significance was assumed at a probability value of less than 0.05.  Results and discussion Growth of isolates with ROH in culture. There were no indications of any gain or loss of fitness in isolates with ROH when grown in culture. The isolates without ROH had a slightly larger diameter than the isolates with the ROH in both the larch and Douglas-fir agar; however the differences were not significant (Table S4.18). The differences were largely due to one isolate, Pram_P2599_EU1, in the group without ROH which grew faster than all other isolates on both media. When this isolate was removed, the groups had average diameters that were not statistically different at all time periods (data not shown). No differences were found in the number of spores produced by each group of cultures in each of the collection times (Table S4.18).   Growth of isolates with ROH on Rhododendron leaves. No signs of gain or loss in fitness were observed for isolates with ROH on detached Rhododendron leaves. There were no significant differences between lesion length in the individuals with and without ROH. The average of the group was slightly higher for the individuals without ROH for the first 12 days, and slightly higher for the individuals with ROH by 17 days post inoculation; however the range of lengths were very similar for each group throughout the 17 days of the experiment (Fig. S4.12).  208  Growth of isolates with and without ROH on larch and Douglas-fir log bolts. In Japanese larch bolts eight weeks post inoculation, several lesions had grown together and in some of the bolts the entire cambium had turned necrotic reducing the number of reps from different log bolts, as well as the overall sample size of each group (number of isolates with ROH = 5, number of isolates without ROH = 6). The average lesion length of the remaining isolates with no ROH was longer (272.7 mm ±) than isolates with ROH (235.7 mm ±); however the difference was not significant (Kruskal-Wallis chi-squared = 0.833, p = 0.3613) (Fig. S4.5).In Douglas-fir eight weeks post inoculation, the average lesion length was higher in the isolates without ROH (155.3 mm versus 140.4 mm), however the difference was not significant (Kruskal-Wallis chi-squared = 0.6562, p = 0.4179) (Fig. S4.5).  OrthoMCL analysis and evolution of gene family size Methods The protein models obtained for the de novo assemblies EU1 030002 (14 095 models), EU2 P2586 (14 028), NA1 PR09-175 (14 213), NA2 0438813 (14 186) and P. lateralis CBS_168.42 (17 463) (Appendix B, De novo genome assemblies) were combined with those from the sequenced genomes of P. infestans (18 140 models) (Haas et al., 2009), P. sojae (26 584 models) (Tyler et al., 2006), P. capsicii (19 805 models) (Lamour et al., 2012), P. parasitica (27 942 models) (Broad Institute (, 2014) and P. cinnamomi var. cinnamomi (26 131 models) (Grigoriev et al., 2012) and a tentative clustering of one-to-one orthologs was carried out using OrthoMCL (Li et al., 2003) (BLASTp search with E-value cut-off of 1e-05, coverage of at least 50% of the query sequence and identity of at least 30%; OrthoMCL inflation value of I = 4) 209  and then automatically aligned with MAFFT ver. 7.123b (Katoh & Standley, 2013). The resulting list of 51,451 OrthoMCL clusters was then submitted to two filters to minimize the confounding effect of truncated proteins resulting from fragmented de-novo assemblies and gene mis-predictions (Nowell et al., 2014). First, redundant OrthoMCL clusters were eliminated by searching the full 51 451 OrthoMCL clusters with 1- and 2-member OrthoMCL clusters; clusters with a significant hit (BLASTn cut-off of 80% amino acid identity over a 80% alignment) to another OrthoMCL cluster were considered spurious and eliminated. The process was repeated iteratively over increasing member size classes (3-members, 4-members, etc.) until reaching an error rate of <0.05% (Nowell et al., 2014). Second, OrthoMCL clusters containing multiple protein fragments were corrected by eliminating protein sequences from the same genome with an average pairwise amino acid identity of ≥ 80%. We performed GO enrichment analysis of the unique protein models using MGSA (Bauer et al., 2011).  To reconstruct a neighbor-joining (NJ) tree representing genome content similarity, we calculated a distance matrix based on the gene presence/absence matrix inferred from the OrthoMCL analysis in which the distances measured between pairs of taxa are inversely proportional to the number of gene they share (Snel et al., 1999). The tree was reconstructed by using the fneighbor program of the PHYLIP Ver. 3.696 package, with 1000 bootstrap replicates sampling columns of the presence/ absence matrix. The maximum likelihood model of CAFE (Computational Analysis of gene Family Evolution) (Snel et al., 1999) was then used to study gene family expansions/contractions while taking into account the one-to-one ortholog phylogeny reconstructed for Phytophthora spp. (Appendix B, Divergence time between lineages - Phylogenetic analysis of one-to-one orthologs).  210   Core and non-core genomes Results and discussion For each lineage, in addition to the bias in codon usage for the non-core genome, there was also a bias for the genes located on repetitive DNA, but in lower frequencies than in the non-core(e.g. 72.7% of the genes found in the repetitive genome of NA2 showed the bias; Fig. S4.8) (Table S4.8). Different compositional feature such as G+C content and codon usage biases are usually diagnostic of gene regions with different origin than the rest of the genome (Becq et al., 2010; Soanes & Richards, 2014; Ravenhall et al., 2015). For example, strong compositional differences between the dispensable chromosomes and the core genome of the wheat fungal pathogen Zymoseptoria tritici suggested an origin by ancient horizontal gene transfer (Goodwin et al., 2011). In addition, genome analyses in Oomycetes have demonstrated that HGT has had major impacts on the evolution of these organisms, specifically in the plant pathogens of the Phytophthora genus (Savory et al., 2000). To test the HGT-origin hypothesis we searched for the presence of foreign top blast hits for the genes of the non-core genomes in the four P. ramorum lineages. Proportions of non-oomycetes hits were low and not significantly different from those observed in the full proteome (χ2 = 0.35, P = 0.95), indicating that most of the genes found in the non-core genome were likely inherited vertically from Oomycete ancestors, ruling out the possibility of a recent HGT at the origin of the non-core genome of the P. ramorum lineages. Rather, the characteristics of the non-core genome suggest that genes in this region may be affected by a degenerative process that likely results in the strong differences observed with those of the core regions.  211   Genes encoding effectors  Methods RxLR effectors. RxLR protein annotation was performed on de novo assemblies and also on the P. ramorum reference genome (Pr-102 isolate) (Tyler et al., 2006) available through the Department of Energy-Joint Genome Institute (DOE-JGI, Walnut Creek USA) (version 1.1) (Appendix B, De novo genome assemblies).  The latter was performed as a means of evaluating optimal strategies for identifying and filtering candidate Avh proteins.  Annotation of Avh effectors for each lineage was comprised of two major parts, with each part optimized to produce the most number of candidates, which were manually inspected at the end.  The first part used a large database of Avh family genes predicted from genomes of 23 Phytophthora species, including the P. ramorum genome, to predict candidate Avh genes within each de novo assembly.  In order to refine and expand this list of candidates, the predicted Avh gene set was used to recursively search the de novo assembly.  In addition, the second step was repeated, with the addition of Avh proteins known from the P. ramorum genome.  Results from each recursive step were then combined with duplicates removed and manually reviewed.  Details of each step in this approach are provided below.  Part 1 – Predicting Avh candidates.  First, genomic sequences were translated to six-frame open-reading frame using Emboss, which translates regions between stop codons in the forward and reverse complement directions with the user-specified minimum length of 90 amino acid residues (Rice et al., 2000). These protein sequences were trimmed to start with the m-signal peptide.  A 212  database of previously identified Avh (avirulence homolog) proteins was used to search de novo assemblies using Hmmer version 3.1b (Wheeler & Eddy, 2013). Protein homology is determined using probabilistic models called profile hidden Markov models, which assign position specific scoring for substitutions, insertions, and deletions, and gain power from use of a database that contains multiple instances from a sequence family. The database of Phytophthora Avh proteins contained 9,779 proteins identified from genomes of 23 Phytophthora species (P. cajani, P. europaea, P. foliorum, P. hibernalis, P. litchii, P. megakarya, P. melonis, P. parvispora, P. pistaciae, P. syringae, P. uliginosa, P. vignae, P. cinnamomi var. robiniae, P. taxon niederhauserii, P. pisi, P. cinnamomi, P. rubi, P. fragariae, P. palmivora, P. parasitica, P. sojae, P. ramorum, P. infestans)( provided by Danyu Shen and Brett Tyler, Oregon State University).  Candidate proteins with an HMM score less than 20 and those lacking both a RxLR and dEER motif were removed, following which, candidate proteins identified more than once by the search were reduced to single instance for each protein sequence. Presence / location of a signal peptide cleavage site was predicted for each protein sequence using Signal-P 3.0 (Bendtsen et al., 2004), where only sequences with probability of a signal peptide greater than or equal to 0.8 were retained.  To maximize the number of candidates recovered during recursive protein prediction, no other filtering of protein sequences was performed at this step.   Part 2 – Recursive search.  Avh protein candidates identified in part 1 were converted into a HMM database.  The first step was to create multiple sequence alignment using MUSCLE version 3.8.31 (Edgar, 2004) that was used to build a HMM model of aligned amino acid sequences using Hmmer.  Candidate proteins were trimmed to start with the m-signal peptide, where those with an HMM score less than 10 and those lacking both the RxLR and dEER motif 213  were removed.  Candidate proteins identified more than once by the search were reduced to a single instance and only those candidates with a signal peptide score greater than or equal to 0.8 were retained.  This recursive search was repeated using a HMM database comprised of Avh protein candidates identified in part 1 and 370 previously identified in the P. ramorum genome (Jiang et al., 2008).  Resultant candidate proteins identified in both recursive searches were combined, and those identified more than once were removed.  To aid manual curation, common motifs among the combined candidate proteins were identified using MEME version 4.9.1 (Bailey et al., 2006), with user specified setting that the motif have a minimum of four amino acids and maximum of eight.  The top three motifs identified by MEME were typically the m-signal peptide region, RxLR and dEER, where strength of the m-signal peptide region was likely increased due to trimming of reads to this position.   Part 3 – Optimization and manual curation.  Effector annotation was performed on all de novo assemblies and also on the P. ramorum reference genome.  Each candidate protein with an RxLR motif identified by MEME was subsequently categorized according to actual RxLR sequence.  The 370 Avh proteins previously identified from the P. ramorum genome were similarly processed by MEME to generate a list of 312 that had a predicted RxLR motif, which was used as a baseline for optimizing filtering to avoid exclusion of true positives.  It should be noted that including the additional 370 predicted P. ramorum Avh proteins in the recursive search step yielded an additional 19 RxLR candidates when applying this method to the P. ramorum reference genome, but yielded fewer candidates for some de novo assemblies. Combining results using both recursive search strategies resulted in the most number of candidates.  214   Crinkler protein effectors (CRN). These effectors were identified by using two complementary approaches similar to those developed in Haas et al. (Haas et al., 2009) (Fig. S4.13). A total of 552 previously reported CRNs from P. infestans (196 effectors), P. ramorum (19), P. sojae (100) and P. capsici (237) (CRNdb) were aligned with MAFFT (Katoh & Standley, 2013). The recombination domain containing a LxKLAK motif in the first 60 aa of the alignment and the HVLVVVP motif were then used to set and train two HMM models with the hmmbuild and hmmcalibrate commands of HMMER v3 (Eddy, 2008). The whole proteomes predicted for the four P. ramorum lineages and P. lateralis (see Core and non-core genomes) were searched for these two models with HMMER v3 (Eddy, 2008) (cut-off value of 1e-0.05). Candidate CRN identified through this first approach were then aligned with other CRN of the CRNdb, before training a new HMM model based on full-length CRN sequences.   In the second approach, sequences of the CRNdb were searched against the genomes of P. lateralis and the four P. ramorum lineages using the TBLASTn algorithm (e-value cut-off of 1e-04). The coordinates of matches were captured and all matches overlapping genes found in the first approach were removed. For the other matches, the corresponding DNA sequence was retrieved and translated in ORFs by using the EMBOSS package getorf (minimum size cut-off of 100 nt and a maximum size cut-off of 6000 nt) (Rice et al., 2000). Predicted ORFs were then submitted to an HMMsearch (-T 0) for the full-length CRN model developed in the previous approach. Additional editing was then carried out on the positive hits by checking the presence of both LxKLAK and HVLVVVP domains. This set of CRN was then merged with the one obtained with the first approach to generate a final non-overlapping set of CRN-like proteins. 215   Analysis of positive selection on P. ramorum RxLR and CRN effectors. Sequences of the RxLR and CRN protein datasets were individually clustered with OrthoMCL (Li et al., 2003) using the following parameters: BLASTp search with E-value cut-off of 1e-05, coverage of at least 50% of the query sequence, identity of at least 50% and an OrthoMCL inflation value of I = 1.5. Protein clusters were filtered for truncated proteins as described in Appendix B, OrthoMCL analysis and evolution of gene family size. For each sequence cluster, tBLASTn was used to retrieve DNA sequence homologs in the P. ramorum de novo assemblies used for the RxLR search. To generate a set of proteins not expected to be under positive selection for comparison, sequence homologs of the 458 proteins of the core eukaryotic CEGMA dataset (Parra et al., 2009) were also retrieved by using the same tBLASTn approach. In addition, a random set of 500 proteins with sequence homologs was extracted and used for comparison. Each sequence dataset was then aligned with MAFFT ver. 7.123b (Katoh & Standley, 2013). Only sequence alignments of high quality (less than 5% ambiguous data) and with at least 3 synonymous sites were retained to reduce statistical bias in the estimation of the dN/dS value (Stoletzki & Eyre-walker, 2011). This resulted 255, 51, 315 and 292 alignments for the RxLR, CRN, CEGMA and random datasets, respectively. For each alignment, the average dN/dS value was estimated by fitting the sequence alignment with the basic maximum likelihood model M0 of Codeml (CodonFreq option set to F3X4) implemented in PAMLV4.0 (Yang & Nielsen, 1998). For statistical comparisons, RxLR and CRN datasets were grouped into effectors and a Box-Cox transformation was done on the full dataset using the MASS package (Venables & Ripley, 2002) in R (R Core Team, 2017) and the transformed data sets were compared with an ANOVA.   216  Results and discussion RxLR effectors. Applying this annotation pipeline to the P. ramorum reference genome identified 391 candidate proteins, where 331 contained a predicted RxLR motif and 391 contained the dEER motif. Among the 331 identified RxLR motifs, 291 were an exact match of 8 amino acids to the RxLR motif in the P. ramorum Avh database, and 309 matched one of 44 different RxLR patterns (17 were a perfect RxLR string, 27 were imperfect, such as RYLK, KLLR, RFSR, etc.). As mentioned by Jiang et al. (Jiang et al., 2008), it is possible that some of the Avh proteins in the P. ramorum Avh database and those identified using the present approach are pseudogenes or inactive alleles.  In order to categorize the diversity of RxLR patterns present within and between lineages, the Phytophthora Avh database of 9,779 proteins was analyzed by MEME to identify the RxLR motif. Perfect RxLR motifs are summarized in Fig. S4.14 and S4.15.   Crinkler protein effectors. HMM searches for Crinkler effectors (CRN) specific domains resulted in a total of 40 (EU2) to 50 (EU1) CRN-like proteins having both the LxKLAK and HVLVVVP domains (Table S4.19). These values are slightly below the number of CRN proteins found in the NA1 reference genome (61 models; (Haas et al., 2009)), and are probably dependent on the quality of the assemblies obtained with these Illumina genomes. Based on a Chi-square test, significant enrichment was not observed in any of the P. ramorum lineages. A first OrthoMCL clustering analysis including these CRN protein models and those predicted for P. lateralis (this study; Table S4.19) and P. sojae (Tyler et al., 2006), resulted in 59 clusters, among which 11 are unique to P. lateralis and/or P.sojae (i.e. all the other clusters share at least one protein with at least one of the P. ramorum lineage), 11 are unique to the group P. ramorum/P. 217  lateralis (i.e. not shared with P. sojae) and 22 are unique to P. ramorum. More than half of the clusters unique to P. ramorum (12) are balanced with one ortholog from each lineage, indicating that novelty in CRN proteins stayed conserved among lineages within P. ramorum.   A second clustering with OrthoMCL between Crinklers from the four lineages resulted in 59 multigene clusters and 12 singlets (Fig. S4.16). Half of the clusters (28) were found to be symmetrical with equal contribution (i.e. one Crinkler protein) from each lineage, indicating no preferential gene expansion from one lineage compared to another. The EU1 lineage showed the highest number of unique Crinklers (7), compared to the other lineages having only one (EU2) or two (NA1 and NA2) unique. These ‘unique’ proteins may have not been assigned to any other cluster due to their higher divergence with Crinklers from the same or other lineage. Only one cluster showed an expansion in the EU1 (2 protein models) and NA2 (3 protein models) lineages. Genes of the cluster have only one single ortholog copy in P. lateralis, suggesting that the diversification happened after divergence between this species and P. ramorum. Branch specific estimates of synonymous and non-synonymous substitution (ω) along the phylogeny showed a significant difference in evolutionary rates between paralog copies of this Crinkler subfamily, indicating an increase of non-synonymous mutations for some members of this subfamily (Fig. S4.11A). Furthermore, a RDP analysis allowed the detection of 4 inter-lineages recombination breakpoints (Fig. S4.11B). This suggests that the history of this CRN effector subfamily was likely shaped by duplication events (gains) followed by recombination before the emergence of the different lineages.   218  Table S4.1: Phytophthora ramorum isolates sequenced in this study. Isolate ID Location Year Host Lib. Prep.1 # of QC passed reads # of mapped reads (%) EU1 P1549 SW England, UK 2003 Fagus sylvatica 1 32 572 330 29 933 640 (91.9%) P1559 SW England, UK 2003 Rhododendron sp. 1 35 073 128 31 629 489 (90.18%) P1562 NA 2003 Quercus cerris 1 32 555 546 29 801 130 (91.54%) P1563 NA 2003 Castanea sativa 1 40 256 606 37 763 140 (93.81%) P1614 NA 2004 Nothofagus sp. 1 33 518 942 30 151 372 (89.95%) P1856 NA 2004 Acer platanoides 1 29 707 842 27 710 875 (93.28%) P1898 NA 2005 Fagus sylvatica 1 26 664 156 24 511 964 (91.93%) P1959 NA 2005 Quercus cerris 1 32 083 066 29 294 756 (91.31%) P2098 NA 2007 Cinnamomum camphora 1 34 939 704 32 201 270 (92.16%) P2673 UK 2009 Larix kaempferi 1 30 161 994 27 754 864 (92.02%) P2677 UK 2009 Larix kaempferi 1 31 571 516 29 388 052 (93.08%) P2687 UK 2009 Larix kaempferi 1 30 543 134 28 447 493 (93.14%) P2688 UK 2009 Larix kaempferi 1 34 649 664 32 313 089 (93.26%) P2707 UK 2009 Larix kaempferi 1 40 594 090 37 496 546 (92.37%) P2738 UK 2009 Larix kaempferi 1 38 177 036 35 280 954 (92.41%)  03-0002  BC, Canada 2003 Rhododendron sp. 1 35 783 200 33 174 125 (92.71%)  03-0107 BC, Canada 2003 Rhododendron sp. 1 34 576 636 30 615 115 (88.54%)  04-17031 BC, Canada 2004 Rhododendron sp. 1 36 986 074 33 755 002 (91.26%)  07-13013 BC, Canada 2007 Rhododendron sp. 1 48 525 936 46 064 371 (94.93%)  08-2933 BC, Canada 2008 Pieris japonica 1 42 129 748 39 296 121 (93.27%)  10-5341b BC, Canada 2010 Rhododendron sp. 1 32 166 628 29 176 869 (90.71%) 219  Isolate ID Location Year Host Lib. Prep.1 # of QC passed reads # of mapped reads (%)  11-301 BC, Canada 2011 unknown 1 36 467 348 32 524 682 (89.19%) 6N0408-2 France 2006 Rhododendron Cheer 1 42 197 568 38 865 912 (92.10%)  10-2278a BC, Canada 2002 Rhododendron Germania 1 59 386 418 55 967 132 (94.24%)  10-2807a BC, Canada 2003 Viburnum tinus 1 30 177 970 28 388 606 (94.07%)  10-3884 BC, Canada 2003 Rhododendron Scyphocalix 1 37 484 444 34 424 704 (91.84%)  10-3885 BC, Canada 2002 Viburnum tinus 1 45 128 834 41 500 703 (91.96%)  10-43896 BC, Canada 2004 Rhododendron sp. 1 41 733 280 39 476 814 (94.59%)  10-4389a BC, Canada 2005 Rhododendron sp. 1 38 796 700 36 498 575 (94.08%)  BBA 26/02 Belgium 2002 Viburnum bodnantense 1 34 521 624 32 479 615 (94.08%)  BBA 14/98-a Germany 1998 Rhododendron catawbiense 1 35 420 988 33 246 850 (93.86%)  BBA 9/95 Germany 1995 Rhododendron catawbiense 1 44 229 884 40 781 496 (92.20%)  CSL 1620 UK NA Viburnum sp. 1 31 039 742 27 959 446 (90.08%)  P2599 England 2011 Larix kaempferi 2 35 011 528 32 853 974 (93.84%)  P2600 England 2011 Larix kaempferi 2 33 617 038 31 660 288 (94.18%)  P2633 England 2009 Larix kaempferi 2 34 166 566 32 015 694 (93.70%)  PD 98/8/6743 Netherlands 1998 Rhododendron sp. 1 51 419 912 47 277 383 (91.94%)  BBA 35/01-01 Germany 2001 Rhododendron sp. 1 38 819 524 35 225 656 (90.74%)  1369017-11 CA, USA 2006 Osmanthus heterophyllus 1 34 512 168 31 666 736 (91.76%)  BBA CSL 1604 UK 2002 Rhododendron grandiflora 1 47 359 894 44 890 851 (94.79%)  BBA CSL 1560 UK 2002 Viburnum tinus 1 37 675 962 33 675 842 (89.38%)  09-82-68 OR, USA 2009 Pieris japonica 1 37 013 252 34 206 433 (92.42%)  09-617-17  OR, USA 2009 Rhododendron sp. 1 41 073 932 38 554 871 (93.87%)  10-923-6-1j OR, USA 2010 Rhododendron sp. 1 40 869 198 35 045 986 (85.75%) BBA 16/99 Germany 1999 Viburnum bodnantense 1 43 800 238 41 055 238 (93.73%) 220  Isolate ID Location Year Host Lib. Prep.1 # of QC passed reads # of mapped reads (%) EU2  P2111 North of Ireland 2007 Quercus robur 2 29 685 302 27 602 287 (92.98%)  P2460 North of Ireland 2010 Larix kaempferi 2 29 213 436 27 107 877 (92.79%)  P2461 North of Ireland 2010 Larix kaempferi 2 27 427 386 25 428 118 (92.71%)  P2561 Scotland 2011 Larix kaempferi 2 37 013 894 34 332 585 (92.76%)  P2566 North of Ireland 2011 Rhododendron ponticum 2 25 240 228 23 437 003 (92.86%)  P2586 Scotland 2011 Larix kaempferi 2 32 741 190 30 384 114 (92.80%) NA1  04-0002 BC, Canada 2004 Camellia sp. 1 45 995 292 42 621 657 (92.67%)  04-2231 BC, Canada 2004 Camellia japonica 1 29 964 586 28 093 396 (93.76%)  04-22572 BC, Canada 2004 Rhododendron sp. 1 60 933 030 57 884 344 (95.00%)  04-35009 BC, Canada 2004 Salal sp. 1 39 248 998 35 098 663 (89.43%)  10-6342a  BC, Canada 2010 Rhododendron sp. 1 42 709 422 40 676 077 (95.24%)  11-263 BC, Canada 2011 Rhododendron sp. 1 28 370 272 26 811 793 (94.51%) 1020.1 OR, USA 2001 Lithocarpus densiflorus 1 47 690 832 45 355 262 (95.10%) 2092 OR, USA 2002 NA 1 40 185 274 35 685 804 (88.80%)  06-81-2 OR, USA 2006 Camellia japonica 1 39 112 726 36 377 329 (93.01%)  33-WA-RD WA, USA 2005 Soil sample 1 29 931 574 28 220 328 (94.28%)  52-WA-A WA, USA 2007 NA 1 49 022 376 45 131 177 (92.06%)  1428631-1 CA, USA 2007 Camellia japonica 1 38 225 732 35 037 704 (91.66%) 1418886 CA, USA 2004 Camellia x ‘Coral Delight’ 1 51 195 010 47 572 883 (92.92%)  GA-28C GA, USA 2004 Camellia sasanqua 1 41 105 758 38 727 544 (94.21%)  WSDA-1107 WA, USA 2004 Camellia sp. 1 37 457 570 33 582 416 (89.65%)  WA 1107.1 WA, USA 2004 NA 1 31 237 818 27 197 217 (87.07%) 221  Isolate ID Location Year Host Lib. Prep.1 # of QC passed reads # of mapped reads (%)  08-SNJ-GO/NC-5B NC, USA 2008 Rhododendron sp. 1 44 628 632 41 812 161 (93.69%)  SPN-S1-C SC, USA 2009 Rhododendron catawbiense 1 41 499 194 37 687 693 (90.82%)  Lakeland IV, G-6-b MS, USA 2009 water 1 41 013 618 39 337 004 (95.91%)  SOD-GA-741W3 S1587 GA, USA NA NA 1 33 521 948 31 310 504 (93.40%)  09-SNJ-SC.09-0353a SC, USA 2009 Rhododendron catawbiense 1 44 662 740 40 203 242 (90.02%)  263AL GA, USA 2005 Kalmia latifolia 1 54 023 504 51 463 597 (95.26%)  304F GA, USA 2005 Camellia japonica 1 43 207 376 40 890 039 (94.64%)  RWN-P20-AR SC, USA 2010 Soil sample 1 33 751 716 31 483 350 (93.28%)  SPN5-P2-2d SC, USA 2010 Kalmia latifolia 1 38 574 618 35 910 371 (93.09%)  SPN5-P2-3c SC, USA 2010 Kalmia latifolia 1 46 473 738 43 289 205 (93.15%)  Pr 733 CA, USA 2010 Water sample 1 41 303 016 38 177 582 (92.43%) 9488 OR, USA 2006 Lithocarpus densiflorus 1 40 766 538 38 577 598 (94.63%) 11372 OR, USA 2011 Lithocarpus densiflorus 1 46 264 218 44 237 245 (95.62%) 1365048 CA, USA 2011 Umbellularia californica 1 51 671 892 49 402 199 (95.61%) 1496607 CA, USA 2011 Camellia japonica 1 52 540 946 49 908 631 (94.99%)  1560191-36 CA, USA 2011 Pieris sp. 1 51 864 454 49 032 929 (94.54%)  1621926-8 CA, USA 2011 Laurus nobilis 1 34 038 516 31 469 874 (92.45%) 1625769 CA, USA 2011 Molinadendron sinaloense 1 44 207 710 41 880 104 (94.73%)  PC#11-1366 CA, USA 2011 Cinnamomum camphora 1 34 162 180 31 046 716 (90.88%) Pr 1546 CA, USA 2012 Quercus agrifolia 2 27 022 916 25 514 680 (94.42%) Pr 420 CA, USA 2004 Quercus agrifolia 2 28 587 464 26 714 242 (93.45%) Pr 518 CA, USA 2009 Quercus agrifolia 2 20 147 362 12 620 067 (62.64%)        222  Isolate ID Location Year Host Lib. Prep.1 # of QC passed reads # of mapped reads (%) NA2  04-20470 BC, Canada 2004 Rhododendron sp. 1 38 656 650 34 610 518 (89.53%)  04-38813 BC, Canada 2004 Viburnum tinus 1 33 812 788 30 697 228 (90.79%)  05-16845 BC, Canada 2005 Ardisia sp. 1 51 967 210 48 406 956 (93.15%)  05-9611 BC, Canada 2005 Viburnum sp. 1 34 762 888 31 976 355 (91.98%)  06-0012 BC, Canada 2006 Soil sample 1 44 616 902 40 658 776 (91.13%)  06-4942 BC, Canada 2006 Distyllium Myricoides 1 33 730 186 31 071 533 (92.12%)  07-17204 BC, Canada 2007 Gaultheria sp. 1 43 194 974 39 904 438 (92.38%)  07-46063 BC, Canada 2007 Rhododendron sp. 1 25 742 572 23 772 653 (92.35%)  08-0688 BC, Canada 2008 Rhododendron sp. 1 53 952 202 50 236 079 (93.11%)  09-4390 BC, Canada 2009 Camellia sp. 1 42 129 748 39 296 121 (93.27%)  09-4415 BC, Canada 2009 Rhododendron sp. 1 42 129 748 39 296 121 (93.27%)  10-3885 BC, Canada 2010 Leucothoe fontanesia 1 44 798 706 41 455 703 (92.54%)  10-4389a BC, Canada 2010 Rhododendron sp. 1 45 198 056 42 833 947 (94.77%)  11-1291 BC, Canada 2011 NA 1 37 164 490 32 387 192 (87.15%)  wsda 3765 WA, USA 2004 Rhododendron sp. 1 42 866 926 37 913 912 (88.45%)  MR31 WA, USA 2004 Rhododendron sp. 1 61 592 742 54 726 800 (88.85%)  1470614-10 CA, USA NA Rhododendron sp. 1 38 106 050 35 534 364 (93.25%)  1586251-1 CA, USA 2010 Camellia japonica 1 94 614 546 84 555 561 (89.37%) 1Library preparation and sequencing protocol : 1, SPRI-TE 300-600bp library HiSeq 100 base PET; 2, plate-based large gap, HiSeq 100 base PET    223  Table S4.2: Number and percent of homozygous and heterozygous SNPs in Phytophthora ramorum individuals with ROH and representatives of the general population without ROH.    Number of SNPs % of SNPs   ROH Total loci [0/0] [0/1] [1/1] [0/0] [0/1] [1/1] HoL1 Notes* EU1 03_0002 Yes 159 218 10 804 137 744 10 670 6.8 86.5 6.7 13.5  PR_99_002 Yes 159 218 6193 146 051 6974 3.9 91.7 4.4 8.3  P1856 Yes 159 218 7357 145 926 5935 4.6 91.7 3.7 8.3 Tested phenotype 07_13013 Yes 159 261 1758 154 576 2927 1.1 97.1 2927 1.8  P1559 Yes - scaf 7 158 852 647 156 403 1802 0.41 98.5 1.1 1.5 Tested phenotype P1614 Yes - scaf 7 158 852 651 156 403 1798 0.41 98.5 1.1 1.5 Tested phenotype P2600 Yes - scaf 7 158 852 657 156 389 1806 0.41 98.5 1.1 1.6 Tested phenotype P1549 Yes - scaf 7         Tested phenotype P1562 Yes - scaf 7         Tested phenotype P1563 Yes - scaf 7         Tested phenotype P1959 Yes - scaf 7         Tested phenotype P2688 Yes - scaf 7         Tested phenotype P2673 No 158 852 85 158 761 6 0.05 99.9 0.0 0.1 Tested phenotype P1898 No         Tested phenotype P2098 No         Tested phenotype P2599 No         Tested phenotype P2633 No         Tested phenotype P2677 No         Tested phenotype P2707 No         Tested phenotype P2738 No 158 852 68 158 780 4 0.04 99.9 0.0 0.0  224     Number of SNPs % of SNPs   ROH Total loci [0/0] [0/1] [1/1] [0/0] [0/1] [1/1] HoL1 Notes* PR_08_034 No 158 852 90 158 756 6 0.06 99.9 0.0 0.1  10_5341b No 159 218 361 158 852 5 0.2 99.8 0.0 0.2  04_17031 No 159 218 204 158 907 107 0.1 99.8 0.1 0.2  EU2 P2586 Yes 99 266 3868 91 962 3436 3.9 92.6 3.5 7.4  P2561 No 99 266 299 98 906 61 0.3 99.6 0.1 0.4  P2460 No 99 266 316 98 881 69 0.3 99.6 0.1 0.4  P2566 No 99 266 293 98 921 52 0.3 99.7 0.1 0.3  NA1 PR_07_191 Yes 146 840 825 145 271 744 0.56 98.9 0.5 1.1  PR_01_004 Yes 146 840 3709 139 788 3343 2.53 95.2 2.3 4.8  Pr_420 Yes 146 840 2034 142 903 1903 1.39 97.3 1.3 2.7  PR_09_167 No 146 840 72 146 749 19 0.05 99.9 0.0 0.1  PR_11_018 No 146 840 97 146 700 43 0.07 99.9 0.0 0.1  PR_11_011 No 146 840 154 146 552 134 0.10 99.8 0.1 0.2  NA2 10_3885 Yes 147 086 6857 133 049 7180 4.7 90.5 4.9 9.5  04_20470 Yes 147 086 6336 136 572 4178 4.3 92.9 2.8 7.1  05_16845 No 147 086 1770 143 441 1875 1.2 97.5 1.3 2.5  06_0012 No 147 086 3356 141 029 2701 2.3 95.9 1.8 4.1  04_38813 No 147 086 605 144 828 1653 0.4 98.5 1.1 1.5  09_4415 No 147 086 576 144 865 1645 0.4 98.5 1.1 1.5  1HoL, total homozygous loci; *Tested phenotype - isolates used to inoculate Douglas-fir, larch and Rhododendron and for growth studies on sapwood agar 225  Table S4.3: Classification of 485 327 SNPs as polymorphic or fixed within Phytophthora ramorum lineages and as heterozygous or homozygous. Numbers exclude singletons and missing data.  Lineage Fixed within lineage (variable between lineage) Polymorphic within lineage Homozygous Heterozygous Homozygous1 Heterozygous2 Both3 NA1 338 354 133 089 16 13 836 32 NA2 337 857 118 845 5 28 574 46 EU1 325 883 129 762 13 16 872 12 797 EU2 386 023 91 125 4 8168 7 1SNPs that are homozygous and different between individuals (ex AA in some TT in others); 2SNPs that are homozygous in some and heterozygous in others (ex AA in some and AT in others); 3SNPs with all three states (e.g. AA, AT, and TT).     226  Table S4.4: Phytophthora ramorum individuals with Chromosomal Copy Number Variation (CCNV).  Lineage Isolate Origin (Host) Inferred Ploidy1 CCNV EU1 04-17031 Canada (Rhododendron sp.) 2n Scaffold 5: 2n and 3n EU1 P1563 UK (Castanea sativa) 3n Scaffold 12: 2n and 3n Scaffold 19: 2n EU1 PR-09-106 USA (Pieris japonica) 4n Scaffolds 1, 2, 5, 7 to 10: 5n NA1 PR-05-055 USA (Forest) 3n Scaffold_1, 6 and 7: 4n Scaffold 4 : 2n 1As infered by the Control-FREEC software.   227  Table S4.5: Number of Phytophthora ramorum proteins with amino acid differences between allele retained and allele lost through conversion to homozygosity after mitotic recombination. Lineage Isolate Number of genes In scaffolds with ROH With amino acid difference between alleles With allele loss due to ROH Genes with ≥5 differences between alleles EU1 03_0002 1587 893 880 202 EU1 07_13013 349 186 175 46 EU1 P1856 879 516 507 128 EU1 PR_99_002 875 514 501 127 EU2 P2586 905 444 387 41 NA11 PR_01_004 641 328 298 47 NA11 PR_07_191 204 108 55 9 NA11 PR_11_011 280 150 14 2 NA11 Pr_420 220 147 146 36 NA2 04_20470 650 303 195 33 NA2 10_3885 1202 638 615 104 NA1 scaffold_72 175 84 73 20 NA1 scaffold_1002 53 24 24 1 EU1 scaffold_73 175 94 63 15 EU1 scaffold_1003 53 25 24 2 NA2 Scaffold_52 144 56 26 2 All4 All 51724 29604 26984 5554 1Excludes scaffold 7 and 100 which were shared by all NA1 individuals; 2Shared by all individuals in the lineage; 3Shared by 8 individuals in the EU1 lineage; 4Total across all four lineages and all affected individuals where the same gene affected in two or more individuals is only counted once.   228  Table S4.6: Proteins enriched in scaffold seven as determined by MGSA analysis  (posterior probability estimate ≥ 0.25).  GO term # in Genome # in scaffold 7 Estimate Std error Biological function GO:0016849 16 6 0.95 0.003 phosphorus-oxygen lyase activity GO:0004553 255 14 0.55 0.010 hydrolase activity, hydrolyzing O-glycosyl compounds GO:0016798 265 14 0.29 0.007 hydrolase activity, acting on glycosyl bonds GO:0045330 11 2 0.24 0.004 aspartyl esterase activity GO:0030599 13 2 0.20 0.004 pectinesterase activity GO:0008474 19 2 0.14 0.004 palmitoyl-(protein) hydrolase activity GO:0098599 19 2 0.14 0.004 palmitoyl hydrolase activity GO:0030248 20 4 0.12 0.012 cellulose binding GO:0016790 22 2 0.11 0.003 thiolester hydrolase activity    229  Table S4.7: Comparison of intergenic distances on scaffolds with mitotic recombination breakpoints with ten sets of random scaffolds.  Number of genes ANOVA test mean Std error F-value p 3'FIR > 10 000 5'FIR > 10 000 Both > 10 000 Scaffolds with breakpoints 49 (7%) 38 (5%) 8 (1%) 3`FIR 4018 4149 - - 5'FIR 2425 4144 - - random set 1 26 (5%) 15 (3%) 1 (0%) 3`FIR 3596 3805 3.20 0.074 5'FIR 1776 3226 8.43 0.004 random set 2 19 (3%) 13 (2%) 1 (0%) 3`FIR 3001 3420 23.8 0.000 5'FIR 1340 2706 31.41 0.000 random set 3 18 (4%) 8 (2%) 0 (0%) 3`FIR 3497 3555 4.62 0.032 5'FIR 1610 2634 13.03 0.000 random set 4 28 (5%) 23 (4%) 1 (0%) 3`FIR 3670 4612 2.1 0.148 5'FIR 2012 4567 2.99 0.084 random set 5 30 (3%) 22 (2%) 2 (0%) 3`FIR 3192 3110 21.47 0.000 5'FIR 1511 2800 28.54 0.000 random set 6 23 (4%) 13 (2%) 0 (0%) 3`FIR 3224 2945 16.2 0.000 5'FIR 1581 2594 19.65 0.000 random set 7 22 (4%) 15 (3%) 0 (0%) 3`FIR 3323 3335 10.15 0.002 5'FIR 1672 3330 11.93 0.001 random set 8 28 (6%) 18 (4%) 3 (1%) 3`FIR 3569 3652 3.61 0.057 5'FIR 1799 3432 7.31 0.007 random set 9 27 (3%) 15 (2%) 2 (0%) 3`FIR 3365 3023 12.38 0.000 5'FIR 1543 2685 24.43 0.000 random set 10 26 (4%) 18 (3%) 1 (0%) 3`FIR 3397 3285 9.942 0.002 5'FIR 1619 2949 18.11 0.000  230  Table S4.8: Nucleotidic and gene content of the de novo assemblies obtained for the four Phytophthora ramorum lineages.  Size (Mb) % repetitive %GC # of genes (Genes/Mb-1) Av. transcript size (bp) # of gene with BLASTp hit # of gene with Oomycete hit # of gene with Blast2GO annotation # of TE-like genes Full genome EU1 49.1 16.06 53.98 14 095 (287) 1448 13 847 (98.2%) 13 754 (97.6%) 7976 (56.6%) 76 (0.54%) EU2 47.6 15.70 54.02 14 028 (295) 1458 13 785 (98.3%) 13 684 (97.6%) 7912 (57.4%) 22 (0.16%) NA1 50.7 14.32 53.96 14 213 (280) 1438 13 900 (97.8%) 13 811 (97.2%) 7758 (55.8%) 88 (0.61%) NA2 48.5 16.10 53.94 14 186 (292) 1439 13 876 (97.8%) 13 744 (96.9%) 7871 (56.7%) 118 (0.83%) Repeats EU1 7.88 100.0 52.88 500 (63) 1002 491 (98.2%) 472 (94.4%) 210 (42.0%) 18 (3.6%) EU2 7.47 100.0 52.89 463 (62) 962 441 (95.2%) 412 (88.9%) 203 (43.8%) 80 (17.3%) NA1 7.26 100.0 52.75 414 (57) 999 400 (96.6%) 389 (94.0%) 191 (46.1%) 15 (3.6%) NA2 7.81 100.0 52.11 392 (50) 902 375 (95.7%) 357 (91.1%) 180 (45.9%) 27 (6.8%) Non-core EU1 0.63 49.33 50.01 26 (41) 859 21 (80.7%) 20 (95.2%) 4 (15.4%) 3 (11.5%) EU2 0.24 37.75 49.35 17 (71) 952 14 (82.4%) 12 (85.7%) 5 (29.4%) 6 (35.3%) NA1 0.64 50.95 49.93 34 (53) 864 31 (91.2%) 30 (96.8%) 8 (23.5%) 8 (5.8%) NA2 0.85 47.32 49.86 44 (52) 824 37 (84.1%) 34 (91.8%) 9 (20.5%) 7 (9.0%)   231  Table S4.9: Protein family content predicted in the non-core genome of the four Phytophthora ramorum lineages. Putative effector families according to (van Damme et al., 2012) are bolded.   EU1 EU2 NA1 NA2 Ester hydrolases - - 1 - Peptidases 1 - 2 4 RxLR 2 - 1 2 ABC transporter 1 - 1 1 Kinases - - - 1 Methyltransferases 3 1 1 2 Glycoside transferases 1 - - - Helicases - 2 1 - Transposons 3 6 8 7 Other hypothetical proteins 10 5 17 22 Putative proteins without homologs 5 3 2 5 Total 26 17 34 44    232  Table S4.10: Shared and unique protein models found in Phytophthora ramorum lineages.  # of clusters (# of protein sequences) Before filtering After filtering1 All  56 560 (192 586) 51 458 (177 030) Core (shared among 10 Phytophthora protein sets) 6201 (73 502) 6200 (69 770) One-to-one orthologs2 3934 (39 340) 4929 (49 340) Lineage specific sets P. ramorum + P. lateralis 238 (1229) 238 (1206) P. ramorum 541 (2262) 541 (2197) EU1 616 (635–4.50%3) 349 (353–2.47%) NA1 713 (739–5.02%) 427 (439–3.09%) EU2 717 (764–5.11%) 437 (452–3.12%) NA2 738 (762–5.37%) 453 (460–3.19%) 1Filters applied to eliminate redundant OrthoMCL clusters and correct those having multiple protein fragments; 2Only one copy of the gene in each of the taxa considered 3The proportion (%) of unique protein is given for the four P. ramorum lineages   233  Table S4.11: MGSA analysis for enriched GO terms in set of proteins unique to the Phytophthora ramorum lineages. Five most represented terms shown.  Lineage Biological functions Estimate SD EU1 GO:0030570, pectate lyase activity 0.50 0.03 GO:0016837, carbon-oxygen lyase activity, acting on polysaccharides 0.30 0.01 GO:0005544, calcium-dependent phospholipid binding 0.50 0.03 GO:0004523, ribonuclease H activity 0.40 0.02 GO:0008239, dipeptidyl-peptidase activity 0.20 0.01 EU2 GO:0008474, palmitoyl-(protein) hydrolase activity 0.49 0.04 GO:0070001, aspartic-type peptidase activity 0.44 0.02 GO:0004190, aspartic-type endopeptidase activity 0.42 0.03 GO:0003810, protein-glutamine gamma-glutamyltransferase activity 0.37 0.03 GO:0016790, thiolester hydrolase activity 0.25 0.01 NA1 GO:0005200, structural constituent of cytoskeleton 0.88 0.01 GO:0004035, alkaline phosphatase activity 0.72 0.02 GO:0070006, metalloaminopeptidase activity 0.43 0.02 GO:0030570, pectate lyase activity 0.33 0.02 GO:0052862, glucan endo-1,4-beta-glucanase activity  0.27 0.02 NA2 GO:0005200, structural constituent of cytoskeleton,  0.78 0.02 GO:0003864, 3-methyl-2-oxobutanoate hydroxymethyltransferase act.  0.50 0.01 GO:0004448, isocitrate dehydrogenase activity,  0.36 0.01 GO:0043168, anion binding 0.20 0.02 GO:0030570, pectate lyase activity 0.11 0.01   234  Table S4.12: Number of SNPs in Phytophthora ramorum before and after filtering at various coverage and quality values, and the percentage of validation SNPs, expected SNPs and false SNPs found after each round of filtering.    Mean depth SNP depth SNP quality1 Gap2 Mac3 Total SNPs before filter4 Total SNPs after filter % validation SNPs found after filter % expected SNP % False SNP Min Max Min Max Q MQ GQ n/a n/a n/a n/a n/a n/a n/a n/a n/a 912 622 N/A 97.2 96.1 0  40        912 622 262 702 0 N/A 0  60        912 622 596 132 57 98.5 0  70        912 622 761 989 86 95.5 0  80        912 622 839 974 97.2 96.1 0  90        912 622 872 609 97.2 96.1 0  100        912 622 887 393 97.2 96.1 0  90   20 20  10  872 609 848 193 97.2 96.1 0  90   30 20  10  872 609 839 102 97.2 96.1 0  90   20 30  10  872 609 848 193 97.2 96.1 0  90   30 30  10  872 609 839 102 97.2 96.1 0 10 90   30 30  10  839 102 811 527 97.2 96.1 0 10 90 5  30 30  10  811 527 629 579 97.2 96.1 0 10 90 4  30 30  10  811 527 647 405 97.2 96.1 0 10 90 3  30 30  10  811 527 665 856 97.2 96.1 0 10 90 2  30 30  10  811 527 685 939 97.2 96.1 0 10 90  150 30 30  10  811 527 545 725 96.9 96.1 0 10 90  200 30 30  10  811 527 743 609 97.2 96.1 0 10 90  240 30 30  10  811 527 801 298 97.2 96.1 0 10 90   30 30 30 10  811 527 398 814 96.9 96.5 0 10 90   30 30 20 10  811 527 441 034 96.9 96.5 0 235  Mean depth SNP depth SNP quality1 Gap2 Mac3 Total SNPs before filter4 Total SNPs after filter % validation SNPs found after filter % expected SNP % False SNP Min Max Min Max Q MQ GQ 10 90   30 30 15 10  811 527 465 216 96.9 96.5 0 10 90 4  30 30 20 10  811 527 438 385 96.9 96.5 0 10 90 4  30 30 20 10 2 811 527 434 966 96.9 96.5 0 10 90 4 240 30 30 20 10  811 527 434 504 96.9 96.5 0 10 90 4 240 30 30 20 10 2 811 527 431 112 96.9 96.5 0 10 90 4 200 30 30 20 10  811 527 398 589 96.9 96.5 0 10 90 4 200 30 30 20 10 2 811 527 395 388 96.9 96.5 0 10 90 4  30 30 30 10  811 527 397 846 96.9 96.5 0 10 90 4  30 30 30 10 2 811 527 395 110 96.9 96.5 0 1Q, SNP quality value; MQ, Minimum Mapping quality at the SNP; GQ, minimum genotype quality; 2Gap, number of base pairs between the SNP and the closest gap; 3Mac, minimum allele count; 4Indels removed.   236  Table S4.13: False call rate between sequencing runs of the same Phytophthora ramorum individual before and after applying SNP filters for coverage and quality.   Before Filtering After Filtering No missing, no singletons1 Individual Total sites Discordant sites False call rate Total sites Discordant sites No of masked sites False call rate Discordant sites False call rate Pram_06_4942_run 1 vs 2 476 646 39 428 8.3% 353 152 6434 36 1.8% 541 0.2% Pram_06_4942_run 1 vs 3 481 739 38 261 7.9% 335 629 5819 23 1.7% n/a n/a Pram_04_17031 418 014 19 052 4.6% 266 020 1819 15 0.7% n/a n/a 1Filters applied across the full dataset to remove SNPs with missing data in some individuals and SNPs found in only one individual. 237  Table S4.14: Assembly statistics and completeness for eight de novo assemblies generated for Phytophthora ramorum.  Lineage Isolate name – Host Total size (Mbp) # of scaffolds N50 (bp) Length of longest scaffold (bp) BUSCO coverage1 EU1 03-0002 – Rhododendron sp. 49.1 4846 24 101 261 007 352 (82.1%)  07-13013 – Rhododendron sp. 48.2 5224 22 468 204 338 351 (81.8%) EU2 P2586 – Larix kaempferi 47.6 3876 31 609 286 734 353 (82.3%)  P2111 – Quercus robur 45.9 3802 31 672 298 197 353 (82.3%) NA1 PR-09-175 – Camellia japonica 50.7 5202 22 339 261 091 353 (82.3%)  PR-11-009 – Umbellularia californica 48.3 4942 22 480 261 530 353 (82.3%) NA2 05-16845 – Ardisia sp. 48.5 4983 23 557 261 946 353 (82.3%)  04-38813 – Viburnum tinus 49.3 4940 27 888 299 169 352 (82.1%) 1Assessed over 429 single-copy ortholog genes common to Eukaryotes (Simão et al. 2015; Bioinformatics, 31:3210).  238  Table S4.15: Mean posterior values for eight selected parameters estimated under the three molecular clock models.  Strict clock UCLD relaxed clock Random local clock Likelihood -84 847.60 -84 795.49 -84 799.32 Posterior -84 890.00 -84 834.71 -84 862.12 Yule.birthrate 0.0755 0.0756 0.0765 Clock.rate 0.0043 - 0.0044 ucld.mean - 0.0043 - ucld.stdev - 0.18 - Coefficient of variation - 0.169 0.141 Rate change count - - 1.985     239  Table S4.16: Median divergence times and 95% confidence intervals for Phytophthora nodes estimated under the three molecular clock models. Divergence time estimates in My.  Nodes Strict clock UCLD relaxed clock Random local clock (EU1,NA1) 0.75 (0.61,0.88) 0.77 (0.54,1.00) 0.76 (0.62,0.90) (NA2,(NA1,EU1)) 1.06 (0.91,1.20) 1.10 (0.84,1.39) 1.07 (0.91,1.23) (EU2,(NA2,(NA1,EU1))) 1.31 (1.14,1.48) 1.39 (0.65,2.04) 1.33 (0.91,1.73) (P. lateralis, P.ramorum) 6.38 (5.91,6.84) 6.70 (5.07,8.34) 6.53 (6.00,7.04) (P. sojae, P. cinnamomi) 11.81 (11.15,12.52) 12.3 (9.71,14.84) 12.78 (11.73,13.85) (Clade 7, Clade 8)1 21.77 (20.95,22.62) 22.45 (20.20,24.78) 22.71 (21.65,23.81) (P. infestans, P. parasitica) 14.65 (13.89,15.41) 13.93 (11.94,15.81) 12.22 (11.02,13.36) (Clade 1, Clade 2) 23.09 (21.11,24.83) 22.12 (19.46,24.59) 22.02 (20.96,23.14) (Clade 1,(Clade 2,(Clade 7,Clade 8)) 25.64 (23.89,27.48) 25.8 (24.00,28.70) 25.65 (23.70,27.38)  1Clade members: Clade 1 = P. parasitica and P. infestans; Clade 2 = P. capsici; Clade 7 = P. sojae and P. cinnamomi; Clade 8 = P. lateralis and P. ramorum.    240  Table S4.17: Enriched gene ontology for Phytophthora ramorum proteins affected by ROH as determined by MGSA analysis (posterior probability estimate ≥ 0.25).  GO accession inPopulation inStudySet Estimate Std error Biological function GO:0005515 1758 326 0.50 0.05 protein binding GO:0016746 176 50 0.49 0.05 transferase activity, transferring acyl groups GO:0016491 791 140 0.42 0.04 oxidoreductase activity GO:0003964 326 4 0.39 0.05 RNA-directed DNA polymerase activity GO:0016835 85 3 0.33 0.03 carbon-oxygen lyase activity GO:0004553 255 52 0.32 0.03 hydrolase activity, hydrolyzing O-glycosyl compounds GO:0005198 146 5 0.26 0.03 structural molecule activity     241  Table S4.18: Growth and sporulation of isolates with and without ROH on sapwood agar.   Day 4 Day 7 Day 10 Media type Group Mean growth (SD) Mean sporulation (SD) Mean growth (SD) Mean sporulation (SD) Mean growth (SD) Mean sporulation (SD) Larch ROH 22.4 (1.4) n/a 37.0 (2.1) 5.6 (1.9) 52.9 (2.7) 2.3 (0.9) Larch no ROH 23.1 (1.4) n/a 39.9 (6.5) 6.0 (3.4) 56.4 (9.2) 1.8 (0.8)  F-value 0.849 n/a 1.455 0.093 1.070 1.615  prob 0.372 n/a 0.248 0.765 0.318 0.224 Douglas-fir ROH 25.5 (1.2) 0.7 (0.8) 1 42.8 (2.8) 1.1 (0.6) 1 58.5 (4.1) 4.8 (0.7) 2 Douglas-fir no ROH 25.9 (1.6) 0.6 (0.5) 1 44.8 (6.2) 1.2 (0.8) 1 63.3 (10.1) 4.8 (1.6) 2  F-value 0.256 0.255 0.670 0.004 1.517 0.022  prob 0.621 0.622 0.427 0.953 0.238 0.884 1Counts done at two diagonal locations beginning from inoculation point; 2Counts done on day 11 at four locations 0.5 cm from center of inoculation point.       242  Table S4.19: Number of Phytophthora ramorum and P. lateralis Crinkler effectors identified in the two approaches of the Crinkler identification pipeline. Lineage (# of proteins) 1st approach 2nd approach Final count    (Incomplete proteins2) LFLAK HVLVVVP LFLAK + HVLVVVP1 NA1 (16 752) 21 15 14 28 42 (47) NA2 (16 187) 27 16 15 32 47 (52) EU1 (16 183) 32 25 18 32 50 (55) EU2 (15 539) 21 14 12 28 40 (47) P. lateralis (20 040) 12 9 3 43 46 (70) 1Both a LFLAK and a HVLVVVP domain were identified in the proteins; 2Incomplete proteins are Crinkler like proteins with only a LFLAK or a HVLVVVP domain.    243   Figure S4.5: Phylogenetic reconstruction of Phytophthora ramorum individuals of four lineages. Maximum likelihood tree reconstructed between individuals using the sequence of 500 genes that are presumed to be neutral based on Tajima’s D. The four lineages are located in highly supported clades, while most of the secondary clades have no support (Bootstrap values < 70)    244  Figure S4.6: Runs of homozygosity caused by mitotic recombination. Heterozygote density on five scaffolds of two individuals of the EU1 lineage: one with long runs of homozygosity (ROH) caused by mitotic recombination (left) and one without ROH (right). The first three scaffolds of the isolate with the ROH pattern show no loss of heterozygosity, whereas heterozygotes are nearly absent in two affected scaffolds (19 and 24). In the individuals unaffected by MR (represented by the individual on the right) there is no loss of heterozygosity on the same two scaffolds.    Isolate with runs of homozygosity Basepairs  X 1000 1 2 3 19 24 Number of heterozygote sites Isolate with no runs of homozygosity  245    Figure S4.7: Read-depth analysis for copy number variation in the Phytophthora ramorum isolate P1563 (EU1) isolate.    246   Figure S4.8: Comparison of predicted secreted protein content for the first 25 largest scaffolds in P. ramorum.    0510152025301 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25% of total protein contentScaffoldSecreted protein content by scaffold 247    Figure S4.9: Lesion length on larch and Douglas-fir four and eight weeks post inoculation of individuals with ROH on scaffold seven compared to isolates without ROH.     248   Figure S4.10: Lesions on larch four weeks four weeks post inoculation for two isolates, one with ROH and one without.   P1559_EU1 with ROH P2677_EU1 NO ROH  249    Figure S4.11: Gene family contraction and expansion in Phytophthora spp. Left: a CAFE analysis was conducted on the 51 458 OrthoMCL gene families and the maximum likelihood phylogenetic tree obtained for the four P. ramorum lineages, P. lateralis and five other Phytophthora species. The size of pie charts is proportional to the total number of gene families that changed; values in blue within the pie chart indicate the proportion of expansions; values in black at node are bootstrap support for the RAxML tree. Right: neighbor joining-tree reconstructed from a distance matrix based on genome similarity in terms of gene content (presence/absence of 51 458 OrthoMCL gene families). Values above nodes are statistical bootstrap support obtained from 100 resampling with replacement of gene families; scale bar indicates 1% genomic dissimilarity i.e. percentage difference in terms of number of shared genes.         250    Figure S4.12: PCA of codon usage on protein coding genes of the Phytophthora ramorum NA2 isolate 05-16845 showing a strong bias toward usage of codons ending with ‘A’ or ‘U’. Results are shown for genes included in the core (top left), repetitive (top right) and non-core (bottom left) genome; color scales on the right indicate protein density; percentage value on top right corner are the proportion of genes found on the right part of the PC1 axis. Bottom right graph represents the contribution of each codon variable to the different components; codons ending with ‘G’ or ‘C’ are labelled in black whereas those ending with ‘A’ or ‘U’ are in green.     251   Figure S4.13: Example of degenerated protein sequence found in the non-core genomes of the NA2 and EU1 lineages. The non-core protein is aligned with its non-degenerated closest homolog found in the core genome; homologous amino-acids are represented with a dot. Waterman-Eggert similarity and coverage values are indicated under each alignment. An asterisk represents a stop codon.    252   Figure S4.14: Distributions of dN/dS values obtained for the RxLR and the Crinkler gene sets compared with the CEGMA and random protein sets. RxLR gene set composed of 255 genes, Crinklers 51 genes, CEGMA set 315 genes, and random protein set 292 genes. Comparison with CEGMA set in top graph and random protein set on bottom graph.     253   Figure S4.15: Phylogenetic relationships and evolutionary history of the CRN subfamily expanded in the P. ramorum EU1 and NA2 lineages. (A) CodeML branch test indicated a significantly increased dN/dS rate for five members of the CRN subfamily. (B) Recombination breakpoints observed among members of this CRN subfamily; segments of different colors represent genomic region between two breakpoints.     254   Figure S4.16: Lesion length on Rhododendron leaves of isolates with ROH on scaffolds seven and 100 compared to isolates with no ROH. 4 5 6 7 10 11 12 17lesion length (mm)days post inoculationP1549No convP1856Isolates ROH  Isolates no ROH  255   Figure S4.17: Complementary approaches used to identify P. ramorum and P. lateralis Crinkler effectors.    256   Figure S4.18: Count of each RxLR motif,where the top row shows the number present within the Avh database, the bottom row is the annotation pipeline applied to the reference genome, and all others between are the total obtained from each de novo assembly.  The most abundant motifs (85.4%) were RFLR, RLLR, RSLR, and RRLR (Left), secondary in abundance (13.1%) were RHLR, RQLR, RMLR, RALR, RYLR, RNLR, and RDLR (Middle), and least abundant (1.55%) were RILR, RVLR, RKLR, RGLR, RWLR, and RVLR (Right).    257     Figure S4.19: Proportion of each RxLR motif, where the top row shows the relative ratio of present in the Avh database, the bottom row is the annotation pipeline applied to the reference genome, and all others between are the relativized values for each de novo assembly.  The most abundant motifs (85.4%) were RFLR, RLLR, RSLR, and RRLR (Left), secondary in abundance (13.1%) were RHLR, RQLR, RMLR, RALR, RYLR, RNLR, and RDLR (Middle), and least abundant (1.55%) were RILR, RVLR, RKLR, RGLR, RWLR, and RVLR (Right).    258    Figure S4.20: OrthoMCL clustering of Crinkler effectors found in the four P. ramorum lineages.                


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items