March 13 2014

5:00 pm 154 BSRB

EcoEvoPub Series

Graduate Student Presentations


Caitlin Fong
Department of Ecology, Evolution, and Marine Biology


Greer Dolby
Department of Ecology and Evolutionary Biology

"Assessing GIS Estuarine Habitat Predictions since the Last Glacial Maximum Using a New Statistical Approach for Testing Genetic Signatures of Postglacial Recolonization"

Coastal species faced arguably the greatest habitat variability from glacial-interglacial cycles as sea levels fluctuating 140 meters along the coastline shifted, created, and destroyed habitat in southern and Baja California repeatedly from ca. .8 Ma to present. Yet most work modeling glacial habitat and corresponding genetic patterns focuses on terrestrial species, which are not influenced directly by eustatic changes and are benefited in their compensatory ability to follow suitable habitat as it shifts through time unlike coastal species that are limited to unidirectional movement along the coast. Using ArcGIS, we developed time-series habitat models since the last glacial maximum (LGM) for three estuarine fish species which predict two glacial refuges and habitat origination times throughout southern and Baja California. These predictions were tested with extensive microsatellite data analyzed using a new statistical approach we developed to uncover the postglacial spatial distribution of refugial alleles. The results significantly support habitat predictions. The results illustrate that limited resolution imagery can still yield informative findings, and the discriminant alleles approach to testing admixture may prove fruitful in other studies. Finally, a multidisciplinary approach like this could be instructive for how to address the geological-biological interplay at the heart of evolution.















































































































































































































































































































































































































































































































this is idtest: