The leafhopper genus Xyphon, included in the sharpshooters, is a widely distributed group of insects whose species are vectors for various plant diseases. Xyphon has historically contained up to 9 species. These species have been poorly delimited in the past and their identification has been difficult using published keys. The genus is revised here based on a new species level phylogenetic assessment that incorporates both morphological and molecular data. The genus Xyphon was erected to contain leafhoppers that possessed a reticulated forewing apex but lacked both a median sulcus on the crown and a carinate anterolateral crown-face margin both of which are present in the closely related genus Draeculacephala. Young (1977) revised most of the genera included in Xyphon's containing subfamily. He did not attempt a revision of Carneocephala (the genus that formerly contained most Xyphon species), but noted the need for a revision of its species. This revision of the genus Xyphon is based on the examination of approximately 8,000 specimens and includes a phylogenetic analysis of the genus that includes data from one gene (NDI) and 47 morphological characters. A generalized model of each preliminary taxonomic unit was used to test the monophyly of each species. These tests resulted in the synonomization of 4 former species: Xyphon gillettei to include X. balli; and X. reticulatum to include X. diductum, X. dyeri, and X. sagittiferum. Parsimony and Bayesian techniques were used to infer relationships among species. These analyses resulted in almost identical tree topologies. In all analyses Xyphon was monophyletic and Draeculacephala was its sister genus although clade support for the genus was generally low. The analyses found that X. flaviceps and X. fulgidum form a basal clade within Xyphon, above which X. gillettei and X. n. sp. 1 (new species 1) form a clade that is sister to a third clade containing X. triguttatum, X. nudum, and X. reticulatum.

The tribe Leucophoropterini (Miridae: Phylinae) is a diverse assemblage of primarily Indo-Pacific and Australian bugs which are united by simple, small genitalia and a trend towards ant-mimetic body forms. Previous to this work, the relationship of the Leucophoropterini to the other tribes of Phylinae, as well as the generic relationships within the lineage, was unresolved. Further, the characters initially proposed to unite the tribe are brought into question with the addition of several recently discovered taxa from Australia. The Leucophoropterini is first re-evaluated within a phylogenetic analysis of the subfamily Phylinae, using a combined molecular and morphological dataset to test the monophyly of the lineage, re-test the character synapomorphies supporting it, and to determine the closest relatives to the tribe. The molecular dataset includes 4 genes (COII, 16S, 28S, and 18S), and 123 morphological characters for 104 taxa, which is analyzed in a parsimony analysis using Tree analysis using New Technology [TNT], a model-based analysis in RAxML, and a Bayesian analysis in Mr. Bayes. All three methods resulted in phylogenetic trees with nearly identical generic and tribal groupings, and a lineage containing Pseudophylus Yasunaga, Decomia Poppius and Tuxedo Schuh being sister-group to the Leucophoropterini. With the closest relatives to the Leucophoropterini determined for outgroup selection, a generic revision of the tribe including both Australian and Indo-Pacific taxa is accomplished using 137 morphological characters and is analyzed in an un-weighted and implied weighted parsimony analysis using TNT for 86 leucophoropterine taxa. The Indo-Pacific taxa of Leucophoropterini are found to be related to the Australian Leucophoropterini, and at least two genera within the tribe (Sejanus Distant, Leucophoroptera Poppius) were found to be paraphyletic. Lastly, taxa are revised within the context of the generic-level phylogenetic analysis, with new genera and species from Australia and the Indo-Pacific being described.

This dissertation has two major sections. In Chapter II, complete mitochondrial (mt DNA) genome sequences were used to construct a hypothesis for affinities of most major lineages of rodents that arose quickly in the Eocene and were well established by the end of the Oligocene. Determining the relationships among extant members of such old lineages can be difficult. Two traditional schemes on subordinal classification of rodents have persisted for over a century, dividing rodents into either two or three suborders, with relationships among families or superfamilies remaining problematic. The mtDNA sequences for four new rodent taxa (Aplodontia, Cratogeomys, Erethizon, and Hystrix), along with previously published Euarchontoglires taxa, were analyzed under parsimony, likelihood, and Bayesian criteria. Likelihood and Bayesian analyses of the protein-coding genes converged on a single topology that weakly supported rodent monophyly and was significantly better than the parsimony trees. Analysis of the tRNAs failed to recover a monophyletic Rodentia and did not reach convergence on a stationary distribution after fifty million generations. Most relationships hypothesized in the likelihood topology have support from previous data. Mt tRNAs have been largely ignored with respect to molecular evolution or phylogenetic utility. In Chapter III, the mt tRNAs from 141 mammals were used to refine secondary structure models and examine their molecular evolution. Both H- and L-encoded tRNAs are AT-rich with different %G and GC-skew and a difference in skew between H- and L-strand stems. Proportion of W-C pairs is higher in the H-strand and GU/UG pairs are higher in the L-strand, suggesting increased mismatch compensation in L-strand tRNAs. Among rodents, the number of variable stem base-pairs was nearly 75% of that observed across all mammals combined. Compensatory base changes were present only at divergences of 4% or greater. Neither loop reduction nor an accumulation of deleterious mutations, both suggestive of mutational meltdown (Muller's ratchet), was observed. Mutations associated with human pathologies are correlated only with the coding strand, with H-strand tRNAs being linked to substantially more of these mutations.

The research presented here deals with inferring phylogenetic trees and their use to study the evolution of functional traits in diatoms (Heterokontophyta: Bacillariophyceae). Two chapters are concerned with the phylogeny of a mainly freshwater group, the Cymbellales, with a convoluted taxonomic history and classification. I generated a multi-gene dataset to test the monophyly of the Cymbellales and reconstruct the relationships within the order. The molecular data were equivocal with respect to the monophyly of the Cymbellales, especially when taking into account some problematic taxa like Cocconeis and Rhoicosphenia. Aside from the problem with their monophyly, my work shows that the current genus- and family-level classification of the Cymbellales is unnatural, arguing for the need of nearly wholesale re-classification of the group. The two following chapters make use of phylogenetic trees to model the evolution of functional traits. I explored the evolution of cell size across the salinity gradient finding that the opposing selective forces exerted by marine and fresh waters select for different optimal cell sizes -- larger in the oceans and smaller in lakes and rivers. Thereafter, I modelled the evolutionary histories of habitat preference (planktonic-benthic) and growth form (solitary-colonial) across the diatoms. These traits exhibit markedly different evolutionary histories. Habitat preference evolves slowly, is conserved at the level of large clades, and its evolution is generally uniform across the tree. Growth form, on the other hand, has a more dynamic evolutionary history with frequent transitions between the solitary and colonial growth forms and rates of evolution that vary through time. I hope that these empirical studies represent an incremental advancement to the understanding of the evolution diatom species and functional diversity. / text

This dissertation focuses on the systematics of the Paussinae, a lineage of approximately 775 beetle species all of which produce explosive defensive chemicals and many which are obligate associates of ants (myrmecophiles). This dissertation includes six discrete studies, which taken together span the spectrum of systematic research. It includes taxonomic projects on both larval and adult paussines. It also includes phylogenetic investigations, based on DNA sequence data and morphological data, that address the evolutionary relationships among the major lineages within Paussinae as well as the evolutionary position of the Paussinae within the beetle suborder Adephaga. The inferred trees are used to study character evolution of physiological and behavioral traits of these organisms in a phylogenetic framework. This work sets the stage for a lifetime of work on this fascinating group.

I report the results of the first detailed phylogenetic study of the ant subfamily Ponerinae, a diverse lineage whose properties make it an attractive model system for investigating social and ecological evolution in ants. Molecular sequence data were obtained from seven nuclear genes for 86 ponerine taxa, representing all three ponerine tribes, 22 of the 28 currently recognized genera, and 14 of the 18 informal subgenera of Pachycondyla, a genus strongly suspected to be non-monophyletic. Divergence dates, historical biogeography, and the origin of reproduction by gamergate workers were estimated on the inferred phylogeny. The most important results of this study include: (1) the monophyly of Ponerinae; (2) the paraphyly of Ponerini; (3) the extensive non-monophyly of Pachycondyla; (4) a rapid basal radiation in Ponerini; (5) resolution of many deeper relationships within Ponerini; (6) a rapid and rich basal radiation in the Odontomachus group; (7) an origin for Ponerinae in the upper Cretaceous, with a major burst of diversification near the K/T boundary; (8) a history of rich and continual ponerine diversification during the Cenozoic; (9) a history of regionalized radiations and frequent faunal exchange between major biogeographic provinces; and (10) support for multiple origins of gamergates in Ponerinae. Ponerine taxonomic classification is revised to reflect both these phylogenetic results and a reappraisal of ponerine morphological diversity. The monogeneric tribe Thaumatomyrmecini is newly synonymized under Ponerini, and the diverse genus Pachycondyla is fragmented into 15 genera, largely along the lines of its junior synonyms, though two new genera are erected: Buniapone (gen. nov.) and Mayaponera (gen. nov.). Some junior synonyms of Pachycondyla are transferred to junior synonym status under other genera: Wadeura as a junior synonym of Cryptopone (syn. nov.), Xiphopelta as a junior synonym of Pseudoponera (syn. nov.), and both Termitopone and Syntermitopone as junior synonyms of Neoponera (syn. nov.). Molecular and morphological justifications for these taxonomic changes are given alongside discussions of phylogenetic relationships. Keys to the world genera of Ponerinae are provided, and morphological diagnoses and species lists are given for each genus. Finally, the available information on ponerine ecology and behavior is reviewed and synthesized.

I aimed to expand our understanding of community assembly and species co-existence by examining the implications of phylogenetic robustness on metrics describing phylogenetic community structure, as well as the phylogenetic patterns of co-occurring insect species in Churchill, MB. Using a variety of tree reconstruction methods, I found that cytochrome c oxidase subunit I (COI) was able to accurately estimate phylogenetic community structure metrics calculated from a multi-gene phylogeny when using more biologically realistic approaches. This included incorporating known phylogenetic relationships among families, and methods that employ best-fit models of molecular evolution (i.e. Bayesian inference). My second study examined the phylogenetic community patterns of freshwater insects. Overall communities were phylogenetically clustered suggesting environmental filtering, but community structure varied with time, habitat, taxonomic group, and water chemistry (particularly pH and turbidity). My thesis has suggested more robust techniques for calculating phylogenetic community structure, and described patterns of phylogenetic community composition in subarctic freshwater insects. / Natural Sciences and Engineering Research Council of Canada (NSERC), International Barcode of Life (iBOL), Genome Canada, Ontario Genomics Institute, Canadian Foundation for Innovation, Ontario Ministry of Research and Innovation, Churchill Northern Studies Centre, and Aboriginal Affairs and Northern Development Canada.

The objective of this thesis was to assess the diversity and community structure of fungi in a forest plot in Ontario using a variety of field sampling and analytical methods. Three broad questions were addressed: 1) How do different measures affect the resulting view of fungal diversity? 2) Do fruiting bodies and soil rDNA sampling detect the same phylogenetic and ecological groups of Agaricomycotina? 3) Will additional rDNA sampling resolve the phylogenetic position of unclassified fungal sequences recovered from environmental sampling? Generally, richness, abundance, and phylogenetic diversity (PD) correspond and identify the same dominant fungal groups in the study site, although in different proportions. Clades with longer branch lengths tend to comprise a larger proportion of diversity when assessed using PD. Three phylogenetic-based comparisons were found to be variable in their ability to detect significant differences. Generally, the Unifrac significance measure (Lozupone et al., 2006) is the most conservative, followed by the P-test (Martin, 2002), and Libshuff library comparison (Singleton et al., 2001) with our dataset. Fruiting body collections and rDNA sampling recover largely different assemblages of fungi at the species level; however, both methods identify the same taxonomic groups at the genus-order level as well as ectomycorrhizal fungi as the dominant functional type of Agaricomycotina. This work also shows that the Soil Clone Group I (SCGI) clade is widespread in soils of diverse origins and represents a novel subphylum of Ascomycota.

The biology of the sub-aquatic lichen, Dermatocarpon luridum, was investigated. This lichen is sparsely distributed within the temperate climatic zones around the world, colonizing rock along watercourses that regularly experience water level fluctuations. Specimens collected from Canada and Austria were cultured using standard growth media. Brightfield microscopy was used for algal species identification, while fungal ITS, algal ITS and algal actin gene sequences were used for phylogenetic and population genetic evaluation. Results were: 1) axenic cultures of the photobiont were successful, while those of the mycobiont were not successful; 2) Diplosphaera chodatii is the photobiont associated with D. luridum var. luridum and allies, suggesting algal sharing between mycobiont species; 3) genetic diversity is high, and gene flow was high within local populations, but low between continental populations. Diplosphaera chodatii may be a keystone species contributing to the survival of D. luridum var. luridum along with other sub-aquatic, aquatic and terrestrial lichens.

Accurate identification of pathogens and vectors is essential in epidemiological studies of mosquito-borne pathogens. However, the members of the communis and punctor complexes are difficult to distinguish because they are highly cryptic species, with little to no species-specific morphological characters. The objective of this thesis is to develop molecular tools, including RFLP and DNA barcoding using cytochrome oxidase I (COI), internal transcribed spacer 2 (ITS2) and the intron of ribosomal protein S12 (RPS12) to facilitate identification of the members of these two complexes in Manitoba. A distinct interspecific distance for COI was found between the members of the communis complex included here, and diagnostic RFLP profiles were developed for Oc. communis and Oc. churchillensis. Relatively low average interspecific genetic distances using COI, ITS2 and RPS12 were observed between the members of the punctor complex, indicates no discernable boundaries between these species based on DNA barcoding.