Molecules, morphology and monophyly : resolving pleuronectiform phylogeny and investigating why it has been so difficult to do.

xii, 157 leaves : illustrations (some color)Although adult flatfishes (order Pleuronectiformes) start out in life as bilaterally symmetrical larvae, they undergo a remarkable metamorphosis, where one eye of the symmetrical larva migrates to the opposite side of the cranium, resulting in highly asymmetrical juvenile and adult forms. Because all flatfishes exhibit this bizarre morphology and variation, both the degree of asymmetry and handedness (direction of eye migration) exists within the order, this group provides multiple tests of hypotheses regarding the evolution of bilateral asymmetry and underlying mechanisms. Unfortunately, undertaking such studies has been elusive because of three major issues confounding pleuronectiform phylogenetics: 1) relationships of the major groups within the order remain mostly unresolved, 2) the sister group of flatfishes is unknown, and 3) monophyly of the assemblage is weakly supported. To resolve these issues in pleuronectiform phylogenetics, my dissertation research has focused on: 1) evaluating the effects non-neutral markers on phylogeny estimation, principally rhodopsin1 (rho), 2) rigorously testing both flatfish monophyly and sistergroup hypotheses and 3) resolving relationships within the order, re-examining characters of adult morphology and comparing them to often overlooked larval characters in light of new phylogenetic hypotheses. In the first study, Chapter II, I use previously published sequence data from 78 iv acanthomorph (including flatfishes) taxa for rho, rnf213, irbp and mll, perform tests for neutrality, and compare neutral versus non-neutral markers for congruence using tree distance metrics and topology testing. I find that while the signal provided by rho may be discordant with the others, neutrality alone does not predict congruence and therefore should not be used as a justification to omit data. In the second study, Chapter III, I optimize new molecular markers and sequence them along with rho and rnf213 for 58 flatfishes and 90 putative outgroups to test monophyly, intraordinal relationships and sister-group hypotheses. Those sequences along with data from a previous study are analyzed to determine possible causes for gene tree incongruence or phylogenetic error. I discover that the new markers are variable, providing large amounts of data, while being conserved so that alignment is unambiguous. When combined with the others and analyzed simultaneously, these markers provide overwhelming support for a monophyletic Pleuronectiformes. Additionally, I demonstrate that abundant missing data is likely the cause of low resolution, validate the importance of investigating substitution saturation as a cause of error and discuss asymmetrical taxonomic distribution as a cause of low resolution at the base of Carangimorpha. Finally, in the third study (Chapter IV) I infer an ultrametric tree, recode a previously published matrix of characters of adult morphology, combine those with new larval characters and test whether life history is correlated with phylogenic signal. Further, I investigate the accuracy of ML ancestral character state estimation (ACE) to determine if these morphological characters provide additional support for hypotheses of relationships among major pleuronectiform groups. My results suggest that larval v characters should not be treated as a source of independent data, but do provide resolution and additional support for novel relationships within Pleuronectiformes, although they may be in violation of the condition of low rates on ML ACE. Lastly, I show that because larval characters are mostly pleisiomorphic for the order, and that larval morphology is similar to that of putative sister groups, these characters are a potential source of evidence needed to resolve the placement of this lineage within Acanthomorpha.Richard Gilder Graduate School at the American Museum of Natural History