| Abstract Detail
Phylogeography of northern North America with insights from paleontological, geological, and molecular data Joly, Simon [1], Laliberte, Etienne [2]. Spatial modeling in biogeography: an application of principal coordinates of neighbour matrices (PCNM) for investigating biogeographic patterns in Rosa blanda. Phylogeography has become a very popular tool for investigating biogeographic patterns at the species level. Although phylogeography was originally defined as a descriptive approach, several quantitative methods have been recently described to formally test biogeographic hypotheses and facilitate comparative studies at the species level. This presentation will introduce a novel method in spatial modeling, principal coordinate of neighbour matrices (PCNM), and will demonstrate its usefulness for identifying and interpreting biogeographic patterns at the species level. The PCNM method essentially consists in extracting from a pre-determined spatial connectivity matrix the eigenvectors that maximize spatial autocorrelation between sampling sites (e.g. invidivuals in biogeography). The resulting eigenvectors can then be used in regression to model spatial structures at different scales. Eigenvectors that model a significant amount of spatial autocorrelation among individuals can then be interpreted in terms of environmental features. To illustrate the applicability of the PCNM method, we apply it to decipher biogeographic patterns in the North American Rosa blanda. The eigenvectors that are found to significantly explain the spatial structure of genetic information among individuals (amplified fragment length polymorphisms, AFLP) are correlated to geomorphological features and climatic variables.
Log in to add this item to your schedule
1 - Massey University, Allan Wilson Centre for Molecular Ecology and Evolution, Private Bag 11222, Palmerston North, 4442, New Zealand
2 - University of Canterbury, School of Forestry, Private Bag 4800, Christchurch, New Zealand
Keywords:
Biogeography
landscape genetics
Spatial modeling
Principal Coordinates of Neighbour Matrices (PCNM).
Presentation Type: Symposium or Colloquium Presentation
Session: S2
Location: Room 4/Woodward
Date: Monday, July 28th, 2008
Time: 9:15 AM
Number: S2004
Abstract ID:251 |
