Very high resolution digital elevation models (VHR DEMs) and multiscale landscape genomics analysis applied to an alpine plant species

Abstract

In landscape genomics, choosing the appropriate environmental parameter is primordial. While most studies are looking at large scale associations between genetic markers and environmental variables, adaptation of plants can also happen on a fine scale. However, the predictive ability of large scale predictors is weak when considering microclimates and landscape heterogeneity in mountainous areas. In that context, DEM's derivatives have been proven useful but remain confidential. This research focuses on the alpine herb Biscutella laevigata subsp. laevigata L., an autotetraploid perennial plant, which occurs in small patches in a heterogeneous landscape. The objectives are firstly, to explore correlations between environmental variables, exclusively derived from two DEMs of different spatial resolutions, respectively 25 and 2 meters, and genetic markers. Secondly, to analyze the effect of different spatial resolutions on significant correlations. The most important result obtained is the large congruency between results from a population genomics and a spatial coincidence approach. In addition, three variables show interesting results that is altitude, duration of insolation and morphometric protection index, which is rarely used. And finally, better significances were found most of the times with the 25m DEM, which raises questions about the most appropriate scale of computation of geomorphological and environmental processes.