Revision of the New World Clusiidae (Diptera: Schizophora), with a phylogenetic analysis of the family using morphological and molecular data

The first half of this thesis is an investigation of the New World Clusiidae (Diptera: Schizophora), of which there are nearly 370 species (most are described here as new) in ten genera. All New World genera are revised with the exception of Chaetoclusia Coquillett, 1904, Procerosoma Lonsdale & Marshall, 2006a, Craspedochaeta Czerny, 1903 and the Sobarocephala flaviseta species group, because revisions of these genera have been published and are available from the author upon request. Furthermore, Amuroclusia syn. nov. and Paraclusia syn. nov. are included as junior synonyms of Clusia, Tranomeringia syn. nov. is included as a junior synonym of Heteromeringia, and the new genus Melanoclusia is erected. The second half of this thesis is a genus-level revision of the Clusiidae, providing the first phylogenetic hypothesis of the family since Frey (1960) divided the family into the subfamilies Clusiinae and Clusiodinae. The Clusiidae (Diptera: Schizophora: Opomyzoidea) is defined on the basis of derived morphological characters, and relationships within the family are reviewed using morphological and molecular data. The biology and distribution of the main lineages of Clusiidae are reviewed and a generic key is provided. Molecular data sets include genomic DNA sequences from the mitochondrial genes COI and COII, the large ribosomal nuclear subunit 28S and the nuclear protein-coding CAD (or "rudimentary"). The genes were each analyzed separately, in combination with each other, and in combination with the morphological data set. The phylogenetic signals provided by the molecular data often conflicted with one another (as indicated by decay values) but the topology of the cladograms produced from each data set alone or in combination was largely similar. Although the molecular results could not recover ambiguous basal divergences and most genus-level relationships were not convincingly resolved, the data revealed one unexpected phylogenetic relationship not evident from the morphological data: Peratochaetus (formerly of the subfamily Clusiinae) is here assigned to the Sobarocephalinae on the basis of strong nodal support from all molecular analyses, revealing remarkable character convergence and reversal in that subfamily. The results illustrate the benefits of utilizing multiple independent data sets for phylogenetic reconstruction in order to verify and refine existing classifications.