Concordant genetic breaks, identified by combining clustering and tessellation methods, in two co-distributed alpine...

Natural genetic breaks may indicate limitations to gene flow or the presence of contact zones of previously isolated populations. Molecular evidence suggests that genetic breaks have aggregated in distinct geographical areas. We propose a new application of well-established statistical methods for analysing multilocus genetic data to identify intraspecific genetic breaks. The methodological approach combines Bayesian clustering with a spatially explicit maximum-difference algorithm to visualize and quantify breaks between clusters. We used amplified fragment length polymorphism data of two co-distributed, silicicolous alpine plant species, Geum montanum and Geum reptans, exhaustively sampled on a consistent, regular grid over their entire range of the European Alps and the Carpathians. We found a distinct and highly similar allocation of genetic breaks in both species. The pattern of breaks did not conform to scenarios of glacial refugial survival, alternatively in peripheral refugia or on nunataks, as expected because of the species' ecologies (late-successional low alpine vs. early-successional high alpine). Our findings rather substantiated the postulate of a general pattern of genetic breaks in alpine plants and corresponded well with biogeographical regions formerly described. Our approach could serve as a tool not only for landscape genetics and comparative phylogeography, but also for floristics or faunistics to compare biogeographic with phylogeographic breaks.