Common factors drive adaptive genetic variation at different spatial scales in Arabis alpina

Authors

  • S. MANEL,

    1. Laboratoire d’Écologie Alpine, CNRS UMR 5553, Université Joseph Fourier, BP 53, 2233 Rue de la Piscine, 38041 Grenoble Cedex 09, France
    2. Laboratoire Population Environnement Développement, UMR 151 UP/IRD, Université de Provence, 3 place Victor Hugo, 13331 Marseille Cedex 03, France
    Search for more papers by this author
  • B. N. PONCET,

    1. Laboratoire d’Écologie Alpine, CNRS UMR 5553, Université Joseph Fourier, BP 53, 2233 Rue de la Piscine, 38041 Grenoble Cedex 09, France
    Search for more papers by this author
  • P. LEGENDRE,

    1. Département de sciences biologiques, Université de Montréal, C.P. 6128, succursale Centre-ville, Montréal, Québec, Canada H3C 3J7
    Search for more papers by this author
  • F. GUGERLI,

    1. WSL Swiss Federal Research Institute, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland
    Search for more papers by this author
  • R. HOLDEREGGER

    1. WSL Swiss Federal Research Institute, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland
    Search for more papers by this author

Stéphanie Manel, Fax: 0476514279; E-mail: stephanie.manel@ujf-grenoble.fr

Abstract

A major challenges facing landscape geneticists studying adaptive variation is to include all the environmental variables that might be correlated with allele frequencies across the genome. One way of identifying loci that are possibly under selection is to see which ones are associated with environmental gradient or heterogeneity. Since it is difficult to measure all environmental variables, one may take advantage of the spatial nature of environmental filters to incorporate the effect of unaccounted environmental variables in the analysis. Assuming that the spatial signature of these variables is broad-scaled, broad-scale Moran’s eigenvector maps (MEM) can be included as explanatory variables in the analysis as proxies for unmeasured environmental variables. We applied this approach to two data sets of the alpine plant Arabis alpina. The first consisted of 140 AFLP loci sampled at 130 sites across the European Alps (large scale). The second one consisted of 712 AFLP loci sampled at 93 sites (regional scale) in three mountain massifs (local scale) of the French Alps. For each scale, we regressed the frequencies of each AFLP allele on a set of eco-climatic and MEM variables as predictors. Twelve (large scale) and 11% (regional scale) of all loci were detected as significantly correlated to at least one of the predictors (inline image > 0.5), and, except for one massif, 17% at the local scale. After accounting for spatial effects, temperature and precipitation were the two major determinants of allele distributions. Our study shows how MEM models can account for unmeasured environmental variation in landscape genetics models.

Ancillary