SPECIATION ALONG THE PERUVIAN ANDES-AMAZON TRANSITION ZONE: PHYLOGEOGRAPHY AND LANDSCAPE GENETICS OF THE AMEEREGA PETERSI GROUP (FAMILY: DENDROBATIDAE)

Author

Connor Martin French, Southern Illinois University CarbondaleFollow

Date of Award

8-1-2018

Degree Name

Master of Science

Department

Zoology

First Advisor

Brown, Jason

Abstract

The Amazon rainforest along the Andean foothills contains exceptionally high diversity, much of it recent. The complex geology of the Andes and paleoclimate fluctuations preclude complex dispersal scenarios. This, in turn, has contributed to idiosyncratic speciation modes among shallowly-diverged Amazonian taxa. The poison frog genus Ameerega recently radiated throughout the Andes and Amazon (MRCA ~8.7 mya), with some taxa diverging as recently as the late Pliocene and early Pleistocene. Some species-level relationships remain poorly resolved, especially among recently diverged taxa. Here, I define ancestral populations and address the phylogenetic relationships among three recently diverged Peruvian Ameerega species (A. cainarachi, A. petersi, and A. smaragdina), using multiple species tree methods, including one that accounts for reticulate evolution. I complement species tree inference with assessments of behavioral divergence and niche overlap to better resolve species boundaries. I further explore the phylogeographic history of these species of Ameerega with demographic inference, considering evidence for population expansions. These analyses provide the basis to address speciation hypotheses in the Andean lowlands, including the refugial hypothesis and dispersal-vicariance hypothesis. I find support to synonymize A. smaragdina with A. petersi, and that divergent and convergent reticulation processes and historical range expansion impacted the A. petersi group’s speciation history. In addition, I use species distribution modeling (SDM) to infer the A. petersi group’s range dynamics since the mid-Pleistocene (785 kya). SDMs reveal periods of range expansion, contraction, and shifts, tracking climate fluctuations during the Pleistocene. In order to explicitly consider the relative roles of climate and geography in structuring genetic diversity at different time periods, I use a landscape genetics approach and consider 21 isolation-by-resistance hypotheses. These hypotheses include climatic resistance layers from five time periods in the Pleistocene, a stability layer, two geographic layers that reflect the two species’ natural history (distance-from-river and mid-elevation resistance), and composite layers that pair geographic and environmental layers. I find that climate stability and river proximity best explain gene flow. I find that phylogeographic, niche modeling, and landscape genetic evidence supports a dispersal-vicariance model of speciation in the A. petersi group.