THE WORLD ACCORDING TO GARS: THE MOLECULAR SYSTEMATICS AND COMPARATIVE PHYLOGEOGRAPHY OF LIVING GARS (ACTINOPTERYGII: LEPISOSTEIDAE)

Author

Justin Todd Sipiorski, Southern Illinois University CarbondaleFollow

Date of Award

8-1-2011

Degree Name

Doctor of Philosophy

Department

Zoology

First Advisor

Burr, Brooks

Second Advisor

Krajewski, Carey

Abstract

There are seven living species of gars in two genera (Lepisosteiformes: Actinopterygii). In my first chapter, I estimate phylogenetic relationships among six of them using DNA data generated from four complete mitochondrial loci (cytb, CR, 12S, and 16S) and a single, partial nuclear locus--recombination activating protein 1 (RAG1) intron. A single outgroup taxon, the bowfin (Amia calva), was included in my analyses. Regardless of optimality criterion (genetic distance, parsimony, maximum likelihood, and Bayesian inference), a single, well-supported phylogeny estimate emerged. Both sister genera (Atractosteus and Lepisosteus) were monophyletic. Within Atractosteus, A. spatula paired with A. tropicus in the absence of data from A. tristoechus. Within Lepisosteus I recovered two sister pairings of species: L. platyrhincus and L. oculatus; and L. ossues and L. platostomus. I estimated the phylogenetic position of the seventh gar species, A. tristoechus, using DNA data generated from several partial mitochondrial loci and 105 morphological characters. In this analysis, within a monophyletic Atractosteus I recovered a well-supported sister pairing between A. spatula and A. tristoechus, with A. tropicus sister to the pair. My molecular phylogenies largely agree with recent morphologically-based phylogenies aside from the positions of L. ossues and L. platostomus. Using dates associated with the best fossil data available (244 Ma minimal age of the node joining bowfins to gars, and 100 Ma for Lepisosteidae), I estimated divergence dates under a relaxed molecular clock model in a Bayesian framework. Estimated ages for lineages ranged from approximately 50 Ma for Lepisosteus, 23 Ma for the split between L. oculatus and L. platyrhincus, 28 Ma on the L. osseus and L. platostoms divergence, and 30 Ma for the node uniting A. spatula and A. tropicus--dates ranging from the Early Eocene to the Early Miocene. In my second chapter, I conducted basic phylogeographic investigations into the geographical structuring of gene genealogies built with mitochondrial D-loop sequences from 115 individual gars, belonging to five species (A. spatula, L. osseus, L. platostomus, L. platyrhincus and L. oculatus) and collected from a broad array of localities. Across species, across localities, I found some phylogeographic structuring in L. osseus and L. oculatus. I found one unique set of haplotypes confined to L. osseus in the Pee Dee River drainage of North Carolina. The level of molecular divergence between these and other L. osseus haplotypes was similar to that among gar species. I found evidence that there might be mixing of haplotypes across a contact zone between L. platyrhincus and L. oculatus in the Florida Panhandle. I found significant molecular divergence among populations of L. osseus and L. oculatus distributed among drainages to the Gulf of Mexico. However little genetic divergence was detected among populations of L. osseus, L. platostomus and L. oculatus within the Mississippi River basin. In my third chapter I present pilot work into characterizing the origins of a population of morphologically unusual gars in eastern Wisconsin. The longnose gar, Lepisosteus osseus, and the shortnose gar Lepisosteus platostomus are native to Wisconsin. In the Fox and Wolf River systems in eastern Wisconsin there is a third form that superficially resembles the spotted gar, L. oculatus (not previously reported to occur in Wisconsin) in that is exhibits heavy head and body pigmentation and a relatively short, broad snout. After initial molecular phylogenetic analyses showed that these gars did not belong to L. oculatus, results of more detailed molecular investigations, coupled with simple morphological findings, are consistent with the hypothesis that these unusual gars may be hybrids of L. platostomus and L. osseus. Further molecular and morphological investigations must be conducted to definitively infer hybrid status for these unusual gars.