W283
Comparative Genetics Of Populus angustifolia and P. trichocarpa Across Their Similar Latitudinal Ranges

Date: Sunday, January 13, 2013
Time: 9:10 AM
Room: Sunrise
Luke Evans , West Virginia Univeristy, Morgantown, WV
Gancho T. Slavov , Aberystwyth University, Aberystwyth, United Kingdom
Gery Allan , Northern Arizona University
Thomas Whitham , Northern Arizona University
Gerald Tuskan , Oak Ridge National Laboratory, Oak Ridge, TN
Wellington Muchero , BioEnergy Science Center, Oak Ridge National Laboratory, Oak Ridge, TN
Joel Martin , DOE Joint Genome Institute
Wendy Schackwitz , DOE Joint Genome Institute
Stephen DiFazio , West Virginia University, Morgantown, WV
Comparative studies of tree species with similar distributions and ecological functions provide insight into what influences molecular and phenotypic genetic variation.  Populus species are ecologically and economically important riparian trees.  Along the interior Rocky Mountain and the Pacific Coast and Sierra Nevada Mountain Ranges, P. angustifolia and P. trichocarpa, respectively, have narrow longitudinal, but large latitudinal distributions.  We compared patterns of molecular diversity and phenotypic variation using range-wide collections.  Three main findings emerged.  First, strong genome-wide neutral structure is prevalent in both species, generally structured north to south, and could represent the legacy of the effects of past climate on effective population size.  Second, phenology candidate genes show some evidence of stronger latitudinal clines than control loci, but the strong neutral structure complicates tests of selection.  Lastly, phenotypic variation in both species suggests climate driven selection and local adaptation, as well as additional genetic variation not related to climate.  Cumulatively, we found that genetic variation is largely structured along a latitudinal gradient, which is also the primary axis of climate variation along which climate-driven selection should act.  Thus, for other forest trees with similar distributions (e.g., along the Coast Range, Sierra Nevada, Rocky Mountains, or Appalachian Mountains), patterns of genetic variation may largely result from demographic factors (e.g., population size and migration) rather than the impacts of selection.  Our findings suggest that in these cases, careful accounting of latitudinal clines and fine-scale structure will be necessary for tests of natural selection and association mapping.