Submitted Abstracts Names of presenting authors are underlined and listed alphabetically

Oral Presentations Groundtruthing next-gen sequencing for microbial ecology: is beta diversity robust to errors in pyrosequencing? Rachel I Adams1, Andrew J Rominger1, Sara Branco1, & Thomas D Bruns1

1University of California, Berkeley The use of nucleotide barcodes to tag differential environmental samples in high-throughput sequence techniques (the “community genomics” meaning of metagenomics) has increased sampling depth for community analysis by thousand order of magnitude, if not more. Numerous studies have shown that these new sequencing technologies, combined with the bioinformatics pipelines to process these large data sets, are imperfect tools for assaying the true biodiversity of environmental samples. Less is understood, however, how these biases in high-throughput sequencing (and indeed all nucleic acid-based approaches) affect conclusions of compositional similarities or differences across environmental samples. Here we take a simulation approach to explore different biases in nucleotide and bioinformatics analysis, and we investigate how these biases affects eventual results of community composition. Specifically, we explore how variation in the number of true OTUs, the true community dissimilarity, the underlying species-area distribution, sequencing error, and over-splitting of operational taxonomic units (OTUs) (i.e. an imprecise OTU clustering step) affects conclusions on relatedness when the true community is structured into two distinct groups. We report results of the simulation based on parameters informed from pyrosequencing of fungal communities, although other scenarios can be explored. Unsurprisingly, the difference between the true and simulated communities increases with sequencing error and over-splitting of OTUs. The difference also increases as more taxa are rare (the shape of the species-area distribution). On the other hand, communities with greater OTU richness are less biased than depauperate ones, and the magnitude of the bias decreases with increasing true dissimilarity. However, in all simulations, the underlying community structure of two divergent groups was maintained. We conclude that while absolute differences in community structure depend

on the richness of the community and biases in sequencing and bioinformatics processing, conclusions on the underlying community structure across environmental samples are robust to these biases. Regional Patterns in Grape Fungal Consortia Define Wine Fermentation Communities Nicholas A. Bokulich1, M. Ohta1, M. Ryazantseva1, C. Masarweh1, M. Zelin1, D.A. Mills1, 1Dept. of Viticulture & Enology, Dept. of Food Science & Technology, UC Davis, CA Fungal communities play a critical role in shaping the quality characteristics of wine and other fermented foods at all stages of production. Wine fermentations are exposed to a large area of specialized equipment surfaces within wineries, which may serve as important reservoirs for beneficial wine yeasts, as well as spoilage organisms. However, little is known about how the complete ecological contexts of vineyard and winery surfaces impact microbial inputs to wine fermentations. Using a quantitative, high-throughput sequencing workflow for analyzing fungal communities based on sequencing short amplicons of internal transcribed spacer 1, we profiled the fungal communities of winery surfaces, grape musts, and wine samples to monitor microbial traffic throughout wineries during wine fermentation. In grape and wine samples, unique fungal communities create regional and variety-specific signatures, linking regional origin to fermentation community composition. Additionally, the winery environment is an important reservoir for two-way transfer of microbes between wine fermentations. The combination of regional origin, varietal background, and winery surface interactions shapes the unique, traceable, microbial populations of wine fermentations, with strong implications for product quality. Fungi at a small-scale: spatial zonation of fungal assemblages around single trees Sara Branco1, Ian Singleton2, Tom Bruns1

1University of California, Berkeley 2Newcastle University Soil fungi are hyperdiverse and occur in complex and dynamic assemblages and it is unclear how taxa partition space, time, and resources even at a local scale. We studied the zonation patterns of soil fungi associated with trees in an attempt to understand small-scale spatial assemblage structure and its relationship to fungal growth habits and colonization abilities. We used ingrowth bags containing acid-washed sand to sample actively growing fungi under canopies and at drip lines of single pine trees at Point

Reyes National Seashore, CA. Our sample of 20 trees was partitioned between young (<15 years) and old (>30) trees. We used next generation sequencing to document the soil fungi present in the bags. We found significant differences between fungi living under the canopy versus the drip line and between old versus young trees. The non-mycorrhizal fungi were much more numerous than the mycorrhizal, exhibiting higher richness under the drip line, and showing a clear partition within and across trees of different ages. Mycorrhizal fungi, on the other hand, showed no pattern of spatial zonation; this is in contrast with earlier studies of fruiting and root tip colonization around single trees. Sequencing and assembly of diverse eukaryotes: like metagenomics, but much harder C. Titus Brown1,2

1Department of Computer Science and Engineering, Michigan State University 2Department of Microbiology and Molecular Genetics, Michigan State University Sequencing and de novo assembly of eukaryotes is challenging, because of material limitations, high polymorphism, large genomes, repetitive content, mixed samples, and a general lack of nearby reference genomes against which to validate assemblies. I will talk about our experiences in microbial metagenomics and also chordate genomics, as a guide to what the sequencing and assembly challenges are likely to be when sequencing diverse microbial eukaryotes from the environment. Microbial Euks from the Deep: Insights from DNA- & RNA-based gene libraries, single-cell whole genome amplicons, and high-pressure grazing experiments Peter D. Countway1, David A. Caron2

1Bigelow Laboratory for Ocean Sciences 2  University of Southern California The deep-sea in most parts of the World’s oceans is sparsely populated by microbial eukaryotes compared to their typical abundances in overlying euphotic zones. Although a portion of the microbial eukaryote community in the deep-sea is comprised of taxa that sink to these depths from the overlying water column, our previous work suggests that a large fraction of the phylotypes found below 500 m are unique to the deep-sea. Deep-sea microbial mats located near hydrothermal vents appear to be hot-spots for active communities of single-celled eukaryotes and microzooplankton, some of which appear to graze on large filaments of sulfur oxidizing bacteria and bacterioplankton from the vent ecosystem. Isolation of single free-living ciliates from Guaymas Basin microbial mats and 9-North East Pacific Rise detrital communities has provided key

linkages between several deep-sea morphotypes, their genetic identities, and their ecological functions. Several approaches were utilized to investigate the genetic diversity and ecological roles of mat-associated microbial eukaryotes including; whole community 18S rDNA and rRNA sequencing, whole genome amplification of single-cell isolates, and incubations of re-pressurized mat communities to investigate bacterivory by deep-sea ciliates. RNA vs. DNA sequence comparisons indicated dramatically different community compositions within the mats, while single-cell amplification revealed the identities of some of the more conspicuous and active microbial eukaryotes, and in several cases, the bacteria they had recently ingested. Many of the microbial eukaryotes that are active at deep-sea hydrothermal vents are only distantly-related to taxa from shallower depths. Overall, our study points to the important role of ciliates in hydrothermal vent ecosystems as active grazers of microbial biomass, and their potential to serve as a link between bacterial primary production and higher trophic levels. Eukaryotic metagenetic biodiversity assessment: current status, challenges and future opportunities Simon Creer1

1Bangor University, Wales, UK Understanding patterns and processes governing biodiversity facilitates understanding between biodiversity and ecosystem function in all ecosystems. However, our understanding of microbial eukaryotic communities is severely impeded by the level of the taxonomic challenge associated with cataloguing microbial diversity and the number of taxonomists that are able to simultaneously identify inter-phylum level diversity. To a lesser extent, similar problems also pervade the task of macro-invertebrate community identification. Over the past few years we have been developing and optimising an emerging field of eukaryotic biodiversity assessment, that uses second-generation sequencing (environmental metagenetics) to identify microbial eukaryotic (metazoan) and macro-invertebrate richness on a scale that is likely to empower ecosystem level understanding of eukaryotic diversity. Focusing on methodology and macro-ecological questions, this talk will overview the development of the field and showcase some of the major findings derived from analyses of alpha and beta diversity of the meiofaunal (44um-500um) size fraction, including protist groups, from the European marine benthic biosphere.

Phaeocystis antarctica spatiotemporal diversity, activity and its relation to prokaryotes and viruses using physiological and ‘omic approaches Tom O Delmont1, Anne-Carlijn Alderkamp2, Gert L. van Dijken2, Kevin R. Arrigo2 and Anton F. Post1

1 Josphine Paul Bay center, Marine biological Laboratory, Woods Hole, MA 2Department of Environmental Earth System Science, Stanford University, Stanford, CA 94305, USA The haptophyte algal genus Phaeocystis is a key component of the phytoplankton community in highly productive waters. In particular, P. antarctica is known to recurrently dominate phytoplankton blooms in the Southern Ocean and Antarctic coastal polynyas during the austral summer. Blooms are dominated by the colonial form, where cells are embedded in a mucuous matrix to form sphere-shaped colonies. The blooms contribute significantly to CO2 drawdown and subsequent carbon export below the euphotic zone. Blooms are usually limited by iron and/or light, leaving high concentrations of nitrate, phosphate and silicate that are not utilized. However, several aspects of the blooms (e.g., intra-species diversity, activity limitations, role of viruses and prokaryotes) are still unknown. Moreover, the Phaeocystis/Diatom spational distribution patterns remain unexplained in specific locations. Here, we use a combination of field- and laboratory-based studies to identify the environmental factors that control the distribution of P. antarctica in the Southern Ocean and to identify the specific adaptive responses to these environmental factors. The recently sequenced genome of a reference strain (160 Mbp and about 30 000 reconstructed genes) was used to determine the functional potential of this organism. Acclimation experiments (specifically, acclimation to high and low light and Fe conditions) were performed to produce large transcriptomic datasets in combination with physiological parameters (e.g., photosynthetic properties, cellular constituents and pigment composition). Results are expected to refine metatranscriptomic experimental designs that will take place in open ocean and the Ross Sea and Amundsen polynyas during the two next austral summers. In parallel, bloom samples dominated by P. antarctica and collected during a previous cruise in the Amundsen coastal polynyas are being studied using both metagenomics and 16S/18S amplicon sequencing approaches to study their spatiotemporal diversity, functional pool and relation to prokaryotes and viruses. Results will be used to better understand the life cycle (especially, transition from single cells to colonies), adaptation and acclimation mechanisms of this remarkable organism. Preliminary results from this inter-disciplinary project will be presented.

Towards building a reliable meiofauna reference database to link OCTUs to species names

Sofie Derycke1, Guilini K1, Moens T1

1Ghent University, Biology Department, Marine Biology Section The huge amount of data generated by high throughput sequencing platforms provides an opportunity to investigate biogeographical patterns and ecological interactions of microbial communities to a scale that was not possible before. Recent metagenetic studies have illustrated widespread distributions of meiofaunal taxa, but interpretation of the data is hampered by the lack of a good reference database that allows linking OCTU’s to biological species. Therefore, we have generated MoMentuM, a project aiming at morphological identification of meiofaunal species through experts, and simultaneously linking these identifications to a morphological back-up (pictures/ videos) and Sanger sequences of the 18S SSU and mitochondrial COI gene. The reference database so far contains around 250 marine nematode specimens belonging to over 70 different species. Although COI is less easily amplified in meiofaunal species, it provides resolution at the species level and allows investigation of phylogeographic patterns. The presence of DNA samples that are linked to species enables the amplification of any marker that would prove to be better suited for diversity assessment than the markers currently used. The need for a phylogeny driven genomic encyclopedia of eukaryotes Jonathan Eisen1,2 1Department of Evolution and Ecology, University of California, Davis, California, United States of America 2Department of Medical Microbiology and Immunology, University of California, Davis, California, United States of America Phylogeny driven genome sequencing has been shown to be useful for many groups of organisms including vertebrates, haloarchaea, bacteria and archaea broadly, and fungi. Benefits from such phylogeny driven genome sequencing include improved annotation of (other) genomes, high rate of discovery of novel gene families, construction of reference catalogs for interpreting metagenomic data, and more. I will discuss prior experiences with such projects (especially the JGI Genomic Encyclopedia of Bacteria and Archaea) and propose a model for how to do this for eukaryotic diversity broadly.

Large scale phylogenomic analyses of Blastocystis subtype 1 reveals a major role for lateral gene transfer in adaptation to parasitism of the human gut Laura Eme1, Eleni Gentekaki1 and Andrew J.Roger1

1Department of Biochemistry and Molecular Biology, Dalhousie University, Halifax, Nova Scotia, Canada. The unicellular stramenopile Blastocystis is the most common eukaryotic parasite reported to colonise the human gut affecting 10-20% of the population in the first world and up to 80% in parts of the developing world. However, it remains unclear whether this microbe causes disease or is a harmless commensal. although some studies have identified a high Blastocystis infection rate in people with Irritable Bowel Syndrome (IBS). The issue is further confused by the existence of numerous subtypes (STs) within the genus Blastocystis that infect both humans and animals. So far, a genome sequence is only available from Blastocystis subtype 7 (ST7). We generated, sequenced and assembled transcriptome data from Blastocystis sp. Nand II (subtype 1). Our analyses reveal that the genes encoded by this strain are surprisingly distantly related to ST7 orthologues, sharing on average ~50% identity at the protein level. Preliminary analyses allowed us to detect ~1000 genes in ST1 that have no homologue in Blastocystis sp. ST7. Furthermore, we have developed a robust large-scale phylogenomic protocol to exhaustively catalogue lateral gene transfers (LGTs) in this organism, enumerating the numbers and nature of the donor organisms implicated in the transfer events. Our analyses consisted of: i) Reconstructing maximum likelihood phylogenies of all 12000 transcript clusters from Blastocystis sp. ST1 and their close homologues. ii) Identification of LGT candidates and their taxonomic origin from these phylogenetic trees; and iii) functional annotation of LGT candidate protein families. From these analyses we identified 250 horizontally acquired genes, including an unexpectedly large number of eukaryote-to-eukaryote LGTs (~150). Interestingly, the broad functional categories to which these genes belong differ according to the taxonomic origin of the LGT candidate. Most genes of eukaryotic origin seem to be involved in cellular processes and signaling mechanisms whereas genes of prokaryotic origin are predominantly involved in metabolic processes such as energy production, amino acid and carbohydrate transport and metabolism. Our results also revealed a dozen of genes that underwent extensive rounds of gene duplications after having been transferred to Blastocystis. These genes seem to be specifically involved in adaptation to parasitism, such as in adaptation to the gut environment or in resistance to host immune responses. Among others, we identified the LGT of an enzyme that metabolizes toxic aldehydes produced during an immune reaction. Another example is the acquisition of a gene encoding an Na+/H+ antiporter thought to help Blastocystis resisting alkaline stress encountered in the gut environment.

Addressing microbial eukaryotic biogeography in home, hospitals and the world. Jack Gilbert1,2

1Argonne National Laboratory, Chicago, Illinois 2University of Chicago The understanding of Earth’s climate and ecology requires multi-­‐‑scale observations of the biosphere, of which microbial life, including bacteria, Archaea and eukaryotes, are a major component. However, to acquire and process physical samples of soil, water and air that comprise the appropriate spatial and temporal resolution to capture the immense variation in microbial dynamics, would require a herculean effort and immense financial resources dwarfing even the most ambitious projects to date. To overcome this hurdle we created the Earth Microbiome Project, a crowd-sourced effort to acquire physical samples from researchers around the world that are, importantly, contextualized with physical, chemical and biological data detailing the environmental properties of that sample in the location and time it was acquired. The EMP leverages these existing efforts to target a systematic analysis of microbial taxonomic and functional dynamics across a vast array of environmental parameter gradients. The EMP uses the data standards format to capture the environmental gradients, location, time and sampling protocol information about every sample donated by our valued collaborators. Physical samples are then processed using a standardized DNA extraction, PCR, and shotgun sequencing protocol to generate comparable data regarding the microbial community structure and function in each sample. One of the key goals of the EMP is to map the spatiotemporal variability of microbial communities to capture the changes in important processes that need to be appropriately expressed in models to provide reliable forecasts of ecosystem phenotype across our changing planet. In addition to the EMP, we will also discuss the Home and Hospital Microbiome Projects, which represent ambitious efforts to characterize a predictive roadmap for microbial transmission in these environments; these studies are linked to the EMP to enable intercomparability. All three studies will be explored in the context of their impact on our understanding of microbial eukaryotic biogeography, transmission and structural changes, including challenges and limitations for microbial eukaryotic data analysis. Diversity of diversity: an assessment of the differences between bacterial and protist communities in marine environments Jean-David Grattepanche1

1Smith College, Northampton, Massachusetts Microbiologists studying bacteria or eukaryotes use different concepts of species and, more broadly, units of biodiversity. These differing concepts raise problems for analyses as assumptions and pipelines built for one community type may not apply to the other. For example, ideas on the biodiversity of protists have been influenced by the

morphospecies concept and by the observation of high similarity among isolates of many protist morphospecies. At the same time, other protists including many ciliates show patterns of cryptic diversity with multiple genetic species underlying a single morphospecies. Such concepts are not easily applicable to bacteria. In contrast, pipelines to handle high-throughput data from bacteria include a denoising step that aims to minimize the impact of both the considerable error associated with pyrosequencing plus the high levels of diversity within bacterial operational taxonomic units (OTUs). We show that use of such denoising steps leads to a substantial underrepresentation of microbial taxa as multiple distinct morphospecies become collapsed into single OTUs when applying parameters appropriate for bacteria. To further complicate matters, many protists and especially ciliates have high copy number of genes like the small subunit rDNA, which confounds attempts to estimate abundance. Thus, care must be taken when using approaches developed for bacteria to get a ‘good’ estimation of the eukaryote diversity. Identifying Loci of Functional Evolution in Ancestral Genes Victor Hanson-Smith1 and Alexander Johnson1 1Department of Microbiology and Immunology University of California, San Francisco Gene duplication events are rampant across the eukaryotic tree of life, and have been associated with many important functional transitions in eukaryotic history. A central challenge is to identify the specific sequence changes that were mechanistically responsible for shifting protein functions from their ancestral to their descendant paralogous forms. In order to address this problem, we developed a new algorithm to predict the set of mutations that historically modified protein functions. Our approach uses phylogenetic ancestral sequence reconstruction as a starting point. We scanned reconstructed ancestors for shifts in residue conservation, specifically looking for primary changes to amino acid state and secondary changes to protein structural properties. By incorporating ancestral inference, our method identified cryptic evolutionary patterns that may not seem obvious when examining only the conservation of modern-day sequences. Using our approach, we predicted the functional genetic determinants for several protein families in eukaryotes, and then tested the effects of these predictions using manipulative molecular experiments in vivo. We observed that highly-ranked predicted substitutions indeed produced significant shifts in protein function, strongly suggesting that our method made valid predictions. We also used our algorithm to analyze gene families with previously-published ancestral function-mutation screens, including yeast genes involved in mating and cellular acidification, and vertebrate genes involved in dim-light vision and steroid- hormone response. Our method highly-ranked those historical substitutions that had been previously shown to be necessary for conferring derived protein functions. Finally, we used our method to analyze gene families lacking any knowledge of ancestral function, including yeast genes involved in white/opaque cell-type switching and pathogenic biofilm development. In both cases, our algorithm identified putative origins of transcription cofactor

interactions. To the extent that eukaryotic diversity is driven by changes to protein function, our approach should be of general interest to anyone investigating the molecular mechanisms of functional evolution. Exploring the deep-sea foraminiferal diversity by multiplexed RNA and DNA-based ultra-deep sequencing and ultra-fast bioinformatic analyses Franck Lejzerowicz1, Philippe Esling1, Jan Pawlowski1

1University of Geneva, Department of Genetics and Evolution, Geneva The estimation of extent biodiversity and the description of correlated biogeographic patterns are particularly challenging for single-celled microbial eukaryotes. In marine sediments, the foraminiferans are among the most diverse and abundant living representatives of all eukaryotic lineages. Therefore, they are well suited for studying the largest and least known ecosystem on Earth: the deep sea. Morphology-based approaches are limited by the number of samples required for large-scale studies and are problematic for assessing the viability of foraminiferal tests. These drawbacks are especially pronounced for the numerous, inconspicuous and tiny soft-walled species that are dominant in deep-sea sediments. Environmental DNA sequencing approaches based on next-generation technologies can overcome this sampling bias. However, a thorough investigation of living cells can only be performed by targeting RNA molecules. For that purpose, we used Illumina sequencing on multiplexed metagenetic and metatranscriptomic libraries from deep-sea sediments. The 37f hypervariable region of the foraminiferal SSU rRNA gene was PCR amplified with tagged primers in order to multiplex the samples. In order to analyze the resulting multi-million reads dataset, we wrote an ultra-fast analysis pipeline in the C programming language including paired-end reads assembly, sample de-multiplexing and de-replication based on a novel hashing tree function. We also introduce a new taxonomy-based approach to sequence assignment and clustering. We describe the observed alpha- and beta-diversity patterns by multivariate ordination analyses. We further discuss aspects of sequencing errors stemming from the extremely large number of reads generated by the Illumina platform. Finally, we also provide a comparison with the same samples analyzed using Sanger sequencing and provide considerations suggesting the optimal size of metagenetic datasets.

Next-generation sequencing reveals the impacts of inundation and drought on eukaryote biodiversity in semi-arid floodplain soils Matthew J. Morgan1, Darren S. Baldwin2,3, Mattthew J. Colloff1, Gavin N. Rees2,3, Anthony A. Chariton4, Garth O. Watson2, Leon N. Court1, Diana M. Hartley1, Andrew J. King1, Jessica S. Wilson3, Michael Hodda1, Christopher M. Hardy1

1CSIRO Ecosystem Sciences, GPO Box 2200, Canberra, ACT 3602, Australia. 2CSIRO Land and Water and the Murray-Darling Freshwater Research Centre, PO Box 992, Wodonga, Vic. 1689, Australia. 3La Trobe University and the Murray-Darling Freshwater Research Centre, University Drive, Wodonga, Vic. 1690, Australia. 4CSIRO Land and Water, Kirrawee, NSW 3313, Australia . Floodplain ecosystems are characterized by alternating wet and dry phases and periodic inundation defines their ecological character. A unique series of climatic events spanning a decade in semi-arid southern Australia enabled us to explore the combined long-term effects of duration since flood and a decade-long drought on floodplain soil biodiversity. Managed flooding events across a lowland floodplain have combined to create a mosaic of flood histories, the extent of events separated only by narrow levee banks. High-throughput barcoded 18S rRNA amplicon pyrosequencing was used to determine the soil eukaryotic community structure on this floodplain, from locations inundated for over a year to nine years since last inundation. Soil biodiversity and community structure on this floodplain showed a reproducible pattern of transitional change from sites inundated at the time of sampling through to sites not flooded for nine years. Interflood communities were dominated by persistent rather than either aquatic or dry-specialist organisms. Almost half the biota were present in wet and long-term dry soils, and relatively little difference was observed between sites dry for two years and nine years. This suggests that most of the floodplain biota are adapted to regular, frequent flooding, but can persist long-term if drought conditions prevail. Incorporating microbial eukaryotes into analyses of the human microbiome Laura Wegener Parfrey1 and Rob Knight1,2

1Biofrontiers Institute, University of Colorado - Boulder 2HHMI, University of Colorado - Boulder In recent years, high-throughput analyses of microbial communities have yielded staggering insights into human health, ecosystem function, and biodiversity. As we turn these tools toward the eukaryotic component of the human microbiome exciting insights into variability of this community across broad geographic and socioeconomic scales. Placing microbial community data within the phylogenetic context of the eukaryotic tree of life lends increased power to elucidate ecological and evolutionary processes that have shaped communities. A comprehensive, well curated, and informatics compatible

database is central to all of these efforts. To achieve this goal, we established the Eukaryotic Taxonomy Working Group in collaboration with Pelin Yilmaz, the curator of the key Silva database for ribosomal DNA sequences. The effort brings together taxonomic and computational experts and we have released the curated database as a resource for the microbial ecology community. Integration of eukaryotic analyses into the Qiime software package will also be discussed. Quantifying taxonomic and functional diversity of metagenomes from next-generation sequencing data Katherine S. Pollard1,2

1Gladstone Institutes, University of California, San Francisco 2Department of Epidemiology & Biostatistics, Institute for Human Genetics, and Institute for Computational Health Sciences, University of California, San Francisco Micro-eukaryotes are important, but poorly understood, components of microbial communities that play roles ranging from primary production to regulation of community composition via predation. Shotgun sequenced environmental DNA, known as metagenomics, provides a means to assay the taxonomic and functional composition of microbial assemblages and may help to shed light on the roles of eukaryotes in these communities. However, the fragmentary, non-overlapping nature of metagenomic sequence data and the limitations of current databases in terms of their coverage of microbial diversity present challenges for the analyses of metagenomes. To address these issues, we designed novel statistical phylogenetic methods for de novo identification of operational taxonomic units (OTUs) and protein families (MRC) in metagenomes, plus automated methods for clustering proteins into families (SFams). Two key features of our approach are the use of probabilistic models of gene family evolution (e.g., profile hidden Markov models and stochastic context free grammars) and the generation of phylogenetic trees in which each leaf is a metagenomic sequencing read from a gene family. To test the performance of these and other tools, we developed error detection methods and a simulation pipeline. We then quantified the contributions of experimental parameters, current knowledge, and analysis methods to metagenome analyses. We discovered some expected patterns, such as the value of longer sequencing reads and diverse reference genomes. Unexpectedly, we also found that profile models do not perform as well as pair-wise search for classifying short reads into protein families and that phylogenetic methods can perform well even when there are many errors in trees built from metagenomic data. Micro-eukaryotes pose some specific challenges for metagenomics due to the relatively low number of sequenced genomes from this domain, their high proportion of non-coding DNA, and their genome size and structure. However, our success in using metagenomic data to identify novel bacteria and reveal global patterns of bacterial diversity suggests that with appropriate bioinformatics tools this approach may also provide substantial insight into the phylogenetics, physiology, and biogeography mico-eukaryotes.

Discrimination of microscopic pests of agricultural importance with NGS Dorota L Porazinska1,2, M Hodda2, LN Court2, MJ Morgan2, CM Hardy2.

1Fort Lauderdale Research and Education Centre, University of Florida, 3205 College Ave, Fort Lauderdale, FL 33314 2The Commonwealth Scientific and Industry Research Organization, Ecosystem Sciences, GPO Box 1700, Canberra, Australian Capitol Territory 2601, Australia In the face of climate change, ever-growing human population, and global transport of agricultural commodities, fast and accurate diagnostic tools for agricultural pests are a vital part of food security. Currently employed identification tools may be appropriate for known and suspected pathogens, but fall short for those that are unknown and unsuspected. We assessed the potential of pyrosequencing of ~400bp SSU rDNA as an alternative to current methods which gets past these limitations. In a designed experiment, we used four soils (replicated 8 times) representing a gradient of nematode pest complexity, ranging from no known pathogens to a few known to a complex suite of poorly known nematode pathogens. To improve the understanding of the extent of false-negative and false-positive identifications, half of the replicates were amended with three known nematode pathogens in known densities. Two of these, Pratylenchus neglectus and P. thornei, are close relatives and were amended as vermiform individuals, while the third was the distantly-related Heterodera avenae which was added as a single cyst (a survival-life form). Because detection and identification of pests must be accurate but potentially can be confounded by sequencing errors and the choice of bioinformatics, all samples were sequenced twice and processed using two different pipelines: OCTUPUS (Operationally Clustered Taxonomic Units for Parallel-tagged Ultra Sequencing) and APDP (Amplicon Pyrosequence Denoising Program). The amendment resulted in an increase of the initial nematode pathogen richness across all soil sources by ~3 species. P. neglectus and P. thornei were detected from amended samples only and in predicted abundances by both pipelines from both sequencing runs. The detection of H. avenae, however, was less consistent showing false-negatives and false-positives that were dependent on the sequencing run as well as the bioinformatics pipeline. In addition to the expected three species, many other pathogens (nematodes and fungi) were detected with their identities and abundances specific to the source of soil but generally irrespective of the sequencing run and the pipeline. Overall, the ecometagenetic approach was very useful for the detection of pathogens. Under all four scenarios of varying degree of nematode complexity, the three “known and suspected” pests were detected and identified with Pratylenchus species diagnosed correctly 100% of the time. Diagnostics of species from cysts or similar life-forms should be further investigated. Most importantly, a wide suite of pathogens that were unexpected and unknown were also identified warranting this approach for monitoring of

pathogenic communities in agricultural and biosecurity settings. Horizontal gene transfer and the evolution of the 'fungal lifestyle' Thomas Richards1

1The Natural History Museum, London, UK Horizontal or lateral gene transfer (HGT) is recognised as an important factor in prokaryotic genome evolution and poses a series of problems when assembling metagenomic data. The role of ‘bona-fide’ HGT in the evolution of eukaryotes, separate from endosymbiotic gene transfer, remains unclear. In this talk I will outline data showing that HGT plays a minor role in terms of number of genes transferred into eukaryotic genomes but that these transferred genes can have significant consequences in terms of the functional biology of the recipient cells. These consequences seem to be important for the ecology of fungi and fungal-like protists who feed by the process of osmotrophy. HGT has therefore shaped osmotrophic phenotypes, and related to this, the public goods games in which these organisms engage. Is everything everywhere? - Metagenomic analysis of bacteria communities Søren Sørensen1

1Institute of Biology, University of Copenhagen, Denmark Next-generation sequencing (NGS) is an excellent tool for investigating microbial diversity and function without the biases of cultivation dependent methods. The dramatic change in price and output of NGS offers a number of new applications of sequencing technology. I will present examples of different applications of NGS technologies in microbiology with main emphasis on a recent study of artic bacterial communities. The global ocean covers more than 70% of Earth’s surface and has an average depth of ~3700 m, making it the largest ecosystem on the planet. Microorganisms inhabit every corner of the sea, from the deepest hydrothermal vent to the surface of the arctic oceans and comprise the largest biomass in the ocean. We have investigated marine archaeal and bacterial communities from Arctic and Antarctic regions at three different depths (surface, medium and deep) to estimate the community composition and vertical and horizontal dispersion in the ocean. A phylotype analysis of 28 marine communities indicates that the vertical distance, rather than geographical location (horizontal distance), directs the community composition. Furthermore, the analysis reveals location specific phylotypes that characterize different environments through the water column. These phylotype data are correlated with metagenomic data to further investigate the functions of the different environments. The ability to use tag-encoded amplicons in MiSeq Illumina sequencers enables simultaneous sequencing of multiple samples in the same run and improves the cost-benefit. Finally, I will show examples of

how 16S rcDNA sequencing enables investigation of the diversity of metabolic active bacteria. Host species and environmental effects on bacterial communities associated with Drosophila in the laboratory and in the natural environment Fabian Staubach1, John F. Baines2, Sven Kuenzel2, Elisabeth M. Bik3, and Dmitri A. Petrov1

1Department of Biology, Stanford University, 371 Serra St, Stanford, California, 94305-5020, United States of America, 2Max Planck Institute for Evolutionary Biology, August-Thienemann- Str. 2, 24306 Plön, Germany, 3Department of Microbiology & Immunology, Stanford School of Medicine, 269 Campus Drive, Stanford, California, 94305, United States of America The fruit fly Drosophila is a classic model organism to study adaptation as well as the relationship between genetic variation and phenotypes. Although associated bacterial communities might be important for many aspects of Drosophila biology, knowledge about their diversity, composition, and factors shaping them is limited. We used 454-based sequencing of a variable region of the bacterial 16S ribosomal RNA gene to characterize the bacterial communities associated with wild and laboratory Drosophila isolates. In order to specifically investigate effects of food source and host species on bacterial communities, we analyzed samples from wild Drosophila melanogaster and D. simulans collected from a variety of natural substrates, as well as from adults and larvae of nine laboratory-reared Drosophila species. We find no evidence for host species effects in lab-reared flies, instead lab of origin and stochastic effects, which could influence studies of Drosophila phenotypes, are pronounced. In contrast, the natural Drosophila–associated microbiota appears to be predominantly shaped by food substrate with an additional but smaller effect of host species identity. We identify a core member of this natural microbiota that belongs to the genus Gluconobacter and is common to all wild-caught flies, but absent from the laboratory. This makes it a strong candidate for being part of what could be a natural Drosophila core microbiome. Metagenetic examinations of benthic eukaryl estuarine communities Sarah Stephenson1

1CSIRO Land and Water, Lucas Heights, AUSTRALIA Australia is surrounded by diverse range of coastal ecosystems. These environments support a range of unique biota, providing numerous ecosystem services. However, in general there is paucity of information with regards to the region’s endemic biota, with taxonomic information often unavailable. Australia’s coastal environments are being

increasingly altered by a myriad of stressors, including: mining, urbanization, aquaculture, contamination and climate change. Traditionally, the health of benthic coastal ecosystems is measured by monitoring a small number (<50 taxa) of taxa, including polychaetes, crustaceans and molluscs. This approach is favoured due to the size of the taxa and their ease of sampling. The underpinning limitation of this approach is that it focuses on a minute fraction of the system’s true diversity. The primary issue is that decisions regarding the ecological condition of coastal environments are being made using an underrepresented view of benthic ecology. The genomic revolution provides an opportunity to expand our ecological knowledge. One approach, is to use next-generation sequencing (NGS) platforms to obtain amplicon sequences which potentially captures the true diversity of a sample (>1000 OTU outputs). While still in infancy, metagenetics provides an opportunity to bolster traditional ecological monitoring programs and produce data in scenarios where traditional means may not be practical. In this presentation I will provide a brief synopsis of some projects where we are currently applying metagenetics to address some regionally important ecological issues. For example, the effects of urbanization in South East Queensland and increasing our ecological understanding of Kakadu, a World Heritage region in Australia’s Top-End. In addition, I will discuss some the advantages and limitations associated with metagenetics from the perspective of obtaining robust and relevant ecological information. Searching for the Phylogenetic Place of Eukaryotes, Using Data from STORI Joshua G. Stern1,2 and Eric A. Gaucher1,2,3

1School of Biology, 2Parker H. Petit Institute for Bioengineering and Biosciences, and 3School of Chemistry, Georgia Institute of Technology, 310 Ferst Dr., Atlanta, GA 30332, USA Selectable Taxon Ortholog Retrieval Iteratively (STORI) is a novel algorithm and tool. STORI retrieves orthologous protein sequences and phyletic patterns for a set of user-defined taxa and genes. We demonstrate the role of this algorithm in a phylogenetic pipeline. We also discuss how STORI may help place and resolve Eukaryotes in the Tree of Life.

Back to the future: Large insert clones and building reference metagenomes W. Kelley Thomas1

1Hubbard Center for Genome Studies, University of New Hampshire, Durham, NH The advent of modern sequencing technologies has driven a sea-change in genome-enabled biology. We are now capable of generating shotgun-based genomic resources for large numbers of non-model organisms. These organisms are typically ones where we know and understand specific biological processes. Most importantly, these organisms are often ideal for elucidating the underlying biological mechanisms once the genomic information is available. One area of biology where this logic is most striking is metagenomics, where we must understand the structure and function at the community level and in the context of the environment. Metagenomics resources for non-traditional species, like genomics resources for individual organisms, have been dramatically expanded by the next generation sequencing revolution. For both marker-based and true metagenomics, we now have the capacity to drill very deep into the communities and gene catalogues from complex environments. Nevertheless, like genomics of individual species, the limitations are now focused on the quality of the information, specifically on the assembly of genomes based on cost-effective (i.e. cheap) short read sequencing. The problem with the effective and accurate assembly of short read datasets typical of High ThroughPut (HTP) Sequencing lies mostly in two issues. First, repetitive sequences interfere with assembly within and between organisms. Second, in complex communities with closely related species, heterozygosity (or closely related homologs of different species) interferes with the assembly when combined in a metagenomic dataset. Here we are proposing to take a technological step backward and help address these problems by sequencing pools of large-insert fosmid clones. The approach has two potential advantages. First, by sequencing (using next gen approaches) a sub-fraction of the diversity in a metagenome the opportunity for interference by repeats is greatly reduced. Second, as clones the fosmids are homozygous and can be assumed to represent a single genomic fragment. Although, fosmid cloning represents a step backward in the march toward simple and cheap HTP sequencing, the relative ease of fosmid-based cloning, and pooling coupled with HTP sequencing may represent a valuable approach to generating higher quality reference metagenomes. Comparative genomic study of the origin and evolution of eukaryotes Xianfa Xie1, Alonzo Anderson1

1Virginia State University Though it has been widely accepted that eukaryotes originated from symbiosis of bacteria and archaea, two main questions still remain: First, what bacteria and archaea have given rise to the mitochondrial and nuclear genomes, respectively. In other words,

what living bacterial and archean speices are the closest relatives of those that gave rise to eukaryotes through symbiosis? Second, is it possible that different lineages of eukaryotes might have originated from different symbiosis events? Though phylogenetic tree could depict all eukaryotes originated from the same ancestor, just like all eukaryotes are often depicted as splitting from the same common ancestor with archea, the origin and evolution of different lineages of eukaryotes may be much more complex. These questions would be best answered through comparative and evolutionary genomic analyses of the diverse lineages of eukaryotes, particularly protists, and related bacterial and archean groups that are known to have contributed to the origin of eukaryotes via symbiosis processes. Some evidences suggest the possibility of multiple origins of eukaryotes. In the presentation we will describe the approach that can be used to further such studies and present some preliminary findings in answering the above questions.

Poster Presentations Rhizosphere fungal microbiome response to moisture and genetic resistance to herbivory

Andrew Krohn1, DL Patch1 Catherine Gehring1

1Northern Arizona University Root-associated fungi provide critical ecosystem services for maintaining the health and long-term fitness of many plant species. Previous studies in pinyon pine (Pinus edulis) have found certain fungal taxa associate with certain tree genotypes, but have mainly focused on characterization of mycorrhizal taxa. We used Illumina sequencing to: 1) determine whether the composition of the rhizosphere fungal microbiome varies when pinyon pine seedlings are subjected to drought-like conditions and are either genetically susceptible or resistant to attack by a major herbivore, the pinyon tip moth (Dioryctria albovittella); and 2) assess whether the taxonomic composition of greenhouse soils, where seedlings were cultivated before transplantation, contained different fungal communities than those found in the field. Our studies revealed: 1) considerable taxonomic diversity including expected functional groups (e.g. saprotrophic and ectomycorrhizal taxa; 2) no major differences in taxonomic composition among treatment groups; and 3) little to no influence of greenhouse fungi on experimental plots in the field. These results provide the first comprehensive view of the rhizosphere fungal microbiome in pinyon pine and are an important step toward understanding microbial function and plant health in this important southwestern plant species.

Rhizosphere bacterial microbiome response to moisture and genetic resistance to herbivory DL Patch1, Andrew L Krohn1, G Allan1, CA Gehring1 1Northern Arizona University Root-associated microbial communities (rhizosphere microbiome) are critical for maintaining the health and long-term fitness of many plant species. Prior rhizosphere studies in pinyon pine (Pinus edulis) have succeeded in identifying only a small proportion of the total community. Here, we use quantitative PCR (qPCR) and next generation Illumina sequencing to: 1) assess both the abundance and taxonomic composition of the rhizosphere microbiome in pinyon pine; 2) determine whether the composition of the microbiome varies when pinyon pine seedlings are subjected to drought-like conditions and are either genetically susceptible or resistant to attack by a major herbivore, the pinyon tip moth (Dioryctria albovittella); and 3) assess whether the taxonomic composition of greenhouse soils, where seedlings were cultivated before transplantation, contained different microbial communities than those found in the field. Our studies revealed: 1) considerable taxonomic diversity for both bacterial and archaeal communities; 2) no major differences in taxonomic abundance or composition among treatment groups; 3) a general trend in higher microbial biomass for both moth resistant and well-watered trees; and 4) little to no influence of greenhouse microbes on experimental plots in the field. These results provide the first comprehensive view of the rhizosphere microbiome in pinyon pine and are an important first step toward understanding microbial function and plant health in this important southwestern plant species. Composition of the Maize Endophytic Microbiome is Correlated with Maize Genotype Surya Saha1, Alice C.L. Churchill1, Santiago X. Mideros1, Peter Balint-Kurti2, Rebecca J. Nelson1,3

1Department of Plant Pathology and Plant-Microbe Biology, Cornell University, Ithaca, NY 14853 2USDA-ARS Plant Science Research Institute, Department of Plant Pathology, North Carolina State University, Raleigh, NC 27695 3Department of Plant Breeding and Genetics, Cornell University, Ithaca, NY 14853 Abstract text (400 words max): All plants contain endophytes that have the potential to provide fitness benefits to their hosts by increasing tolerance to environmental stressors, boosting plant nutrition and growth, and providing increased resistance or tolerance to insect pests and plant pathogens. We are characterizing endophytic populations inhabiting aboveground maize tissues with the goal of associating maize genetic variation with the diversity, structure and constitution of maize- associated

microbial communities. Nine maize lines, representing a diverse subset of the founders of the NAM (Nested Association Mapping) population, were field-grown and assayed for culturable endophytic bacteria and fungi. Two distinct seed sources for each maize line were planted in a randomized experimental design, and three replicates per seed source were assayed, representing a total of 54 samples. Leaf pieces were harvested just prior to pollination for each maize line, surface sterilized using standard endophyte isolation methodologies, and ground leaf extracts were cultured on four media that select for slow- and fast-growing fungi and copiotrophic, diazotrophic, and oligotrophic bacteria. Approximately 65% of the samples contained one or more phenotypically distinct, culturable bacteria, 28% contained one or more fungi, 22% contained both bacteria and fungi, and endophytes were undetectable in 28% of the samples. Interestingly, a greater number and diversity of fungi were cultured from tropical maize lines than from temperate lines. Bacteria were isolated from all maize lines, with some lines exhibiting significantly greater microbial community diversity than others. Several phenotypically similar bacteria and fungi were isolated from multiple maize lines. Microbial cataloging of unculturable endophytes via 16S and ITS sequencing, as well as identification of novel endophytes via whole genome metagenomic sequencing, is in progress. In a parallel analysis, whole genome shotgun sequences generated for the nine maize lines were selected from the HapMap2 dataset for in-silico taxonomic classification of the microbial population. Identification of the bacterial microbiome is underway using FCP (naïve Bayes) and Phymm (Hidden Markov Models). Characterization of fungal endophytes is being done using a read-mapping algorithm and custom ITS databases. We are particularly interested in identifying members of the microbiome that modulate disease symptoms caused by maize leaf and ear pathogens. Hence, future studies will focus on in vitro and in planta endophyte-pathogen interactions. Hunting the unrevealed protozoan reservoir of Francisella Andreas Sjödin1,2, Johanna Thelaus1, Mats Forsman1 1CBRN Defence and Security Department, Swedish Defence Research Agency, Umeå, Sweden 2Computational Life Science Center (CLiC), Umeå University, Umeå, Sweden The natural reservoir of Francisella tularensis, the causative agent of tularaemia, is yet to be identified 100 years after the pathogen was first detected. Scandinavian countries now reports the highest number of cases per country in the world. Tularemia is emerging and expanding, occurring in new areas in over Sweden. The disease is especially prevalent in late summer and autumn in Sweden. The reasons for the geographical distribution, the seasonal occurrence and mechanisms behind outbreaks remain largely unexplained. It has been hypothesized that nutrient-rich environments, with strong protozoan predation, favour the occurrence of the tularaemia bacterium. The role of natural waters in the ecology of the organism is, however, not well characterized, largely due to the fact that the bacterium is difficult to culture from environmental

samples. Vector and field studies so far relied entirely on molecular techniques that detect the bacteria on subspecies level but that are unable to detect specific genotypes within mixed populations. Although there are results in support of an association between Francisella tularensis type B and a potential protozoan reservoir, further in-depth characterization of the natural aquatic habitat is needed in order to clarify the role of potential vectors. Here we used a combination of large volume ultrafiltration of water, metadata water parameters and amplicon sequencing of 16S, 18S and specific Francisella phylogenetic markers to identify key factors promoting the persistence and the dynamics behind tularemia outbreaks. We are able to correlate occurrence of Francisella tularensis with specific species composition of the ambient prokaryotic and eukaryotic communities to identify biomarkers predictive for occurrence of Francisella tularensis type B in the environment.

Meeting sponsorship generously provided by: