Latest publication

Lee-Yaw Julie A, Fracassetti Marco, Willi Yvonne (2017) Environmental marginality and geographic range limits: A case study with Arabidopsis lyrata ssp. lyrata. Ecography, in press


It is commonly assumed that geographic range limits of species reflect ecological niche limits and that species experience increasingly marginal conditions towards the edge of their ranges. Using spatial data and ecological niche models we tested these hypotheses in North American Arabidopsis lyrata. We tested whether range limits coincide with predicted niche limits and whether the suitability of sites declines towards the range edge. We further explored patterns of environmental change towards the edge of the range and asked whether genome-wide patterns of genetic diversity decline with increasing peripherality and environmental marginality. 

Results suggest that latitudinal range limits coincide with niche limits. Populations experienced increasingly marginal environments towards these limits—though patterns of environmental change were more complex than most theoretical models for range limits assume. Genomic diversity declined towards the edge of the species’ range and with increasing distance from the estimated centre of the species’ niche in environmental space, but not with the suitability of sites based on niche model predictions. 

Thus while latitudinal range limits in this system are broadly associated with niche limits, the link between environmental conditions and genetic diversity (and thus the adaptive potential of populations) is less clear.




The distribution of Arabidopsis lyrata sub. lyrata in North America. Black circles in all panels indicate known localities (thinned dataset), with the species’ range limits represented by the minimum convex polygon around these points (outer polygon). Panel a shows the centroid of the MCP of the species’ entire range (red star), genomic sampling sites (blue circles), the boundaries (inner polygons) and centroids (red triangles) of the western and eastern genetic groups described by Griffin & Willi (2014) and the location and orientation of the range-edge transects (numbered grey lines) used in Fig. 3. Panels b-d show the predicted distribution of suitable conditions for A. l. lyrata across the study region based on niche models built using MAXENT. Continuous estimates of suitability are shown in shades of blue in panel b and represent the average prediction from ten rounds of model calibration. Maps of suitable habitat were generated from the continuous prediction surface using c) the lowest and d) the 5th percentile of the average suitability scores of the locality data as cutoffs for defining suitable habitat.