Next-Generation Approaches to Understanding the Diversity and Evolution of Marine Fungi

Abstract

Fungi are among the most diverse extant eukaryotic lineages, with estimates of total global diversity projecting millions of species that have yet to be cataloged. Though fungi from all phyla and habitats await discovery, marine fungi are particularly poorly understood. Historical surveys of fungi in marine habitats, which relied primarily on direct culturing or observation of fruiting bodies and other structures on incubated shore detritus, suggest that marine fungi are unexceptional in their diversity and frequency—and, therefore, unimportant, one might be tempted to conclude. However, with the increasing adoption of environmental sequencing as a primary tool for exploring fungal diversity and ecology across disparate habitats, the discovery of novel phylotypes representing new species—and in some cases, even new phyla—demands a reappraisal of fungal diversity in marine habitats using modern molecular methods. This dissertation represents an attempt to advance our understanding of the breadth of fungal diversity, to establish a broader evolutionary context for marine fungi, and to provide some molecular tools to better study marine fungi that have, until now, eluded our detection.In Chapter 1, I investigate the diversity and spatio-temporal distribution of unicellular eukaryotes in the surface waters of the English Channel using high-throughput sequencing of 18S ribosomal DNA. In addition to characterizing the taxonomic and phylogenetic diversity of planktonic protists, I also estimate the niche breadth of the taxa observed and infer ecological roles and trophic modes to examine if and how functional guilds change through space and time. I find that while the community observed at any given time is likely to be dominated by only a handful of species, many of them members of the Stramenopiles, Alveolates, and Rhizaria, most protistan taxa are rare specialists. I also find that the relative abundance of an individual taxon is not indicative of a specialist or generalist habit. Interestingly, I also find that fungi comprise a significant fraction of the microbial community, but the most prominent fungal taxa observed do not closely resemble phylogenetically any circumscribed species of marine fungi. Rather, they are nested within the enigmatic Cryptomycota, a recently described phylum at the base of the fungal tree. Thus, while this is the only chapter of my dissertation which does not focus primarily on fungal diversity or evolution, the results in many ways set the stage for my subsequent work.In Chapter 2, I turn my sails back toward land to challenge long-standing mycological lore by re-examining the diversity of coastal marine fungi. I focus on four marine habitats in coastal North Carolina (surface water, persistent wetlands, intertidal sand flats, and marine benthos) from which I sample water and sediments over the course of a year. Using primers designed to amplify across the entire fungal kingdom, I use the Ion Torrent platform to sequence amplicons from 28S ribosomal DNA and evaluate how successfully extant reference databases identify novel fungal sequences to both high and low taxonomic ranks. I find that marine fungi are far more varied than previously thought, with some early diverging fungi, and the Chytridiomycota in particular, proving to be diverse and ubiquitous. I also find that curated reference databases struggle to assign robust taxonomic identities to novel sequences across all fungal phyla, but are particularly ill-equipped to identify the marine representatives of non-Dikarya fungi. In Chapter 3, I aim to address the deficiencies of curated reference databases used in taxonomic assignment by generating high-quality reference sequences from complex environmental samples. Using a third-generation sequencing technology, PacBio circular consensus sequencing, which trades the high coverage of other sequencing platforms for much longer read lengths, I target an approximately 2kb fragment of the ribosomal DNA operon that contains both the full fungal internal transcribed spacer (ITS) region and over 1kb of the 28S ribosomal subunit. Using a mock community approach and successive rounds of filtering, I calculate the average sequencing error for PacBio amplicons by comparing them to known sequences from axenically cultured, circumscribed species. I then revisit my samples from Chapter 2, generate amplicon sequences to be used for phylogenetic inference, and compare the accuracy of taxonomic assignments made by reference databases curated for individual loci (i.e., ITS vs. 28S) for different rDNA regions from the same operational taxonomic unit. I find that stringent quality filtering of PacBio sequence data produces consensus sequences approaching the quality of MiSeq and 454, which can be used in phylogenetic analyses to provide improved taxonomic assignments. Furthermore, many of the fungal taxa observed belong to known marine lineages, while others are only distantly related to reference accessions and may represent new lineages.