Genetically divergent cytotypes of Vaccinium uliginosum co-occurring in the central Eastern Alps can be distinguished based on the morphology of their flowers

Abstract

The cloudberries (Vaccinium uliginosum s. lat.) are among the most frequent constituents of dwarf shrub communities in the Alps. Recently, we have shown that (i) both diploid and tetraploid cytotypes are widespread in the Eastern Alps and sometimes occur in mixed “populations” and that (ii) these two cytotypes correspond to two non-overlapping AFLP groups, which are in almost perfect congruence with two plastid lineages. Further (iii), we revealed that diploids and tetraploids show some degree of niche separation, but co-occur in low-alpine heath vegetation. Finally, (iv), in spite of the clear separation of diploids and tetraploids in the molecular data sets, we failed to detect consistent morphological differences. Here, we focused on patterns of genetic, ecological and morphological divergence between cytotypes at mixed-ploidy sites as well as on the underlying processes. Our study is based on a sampling of a hundred individuals each at three sites with previously determined co-occurrence of diploids and tetraploids in the Central Alps of Tirol, Austria. We employed an integrative approach combining flow cytometry, plastid DNA sequencing, AFLP fingerprinting, morphometry of leaf and flower characters as well as a characterization of ecological niches based on the vegetation surrounding the individual Vaccinium plants, in order to address the following questions. What is the small-scale distribution of diploid and tetraploid V. uliginosum s. lat. at the three sites investigated? Are the molecular genetic (AFLP and plastid DNA) relationships reflecting the previously detected divergence governed by differences in ploidy level or is there evidence of across-cytotype gene flow in areas of immediate co-occurrence of diploid and tetraploid V. uliginosum s. lat.? Is there evidence for ecological or morphological separation between cytotypes? More specifically, do flower characters allow for a better discrimination between cytotypes than the previously used leaf characters? We show that mixed-ploidy sites are clearly secondary contact zones as based on AFLP markers individuals cluster by cytotype and not by locality. We found no evidence for recent hybridization across the ploidy boundary but the intersection of biparentally inherited recombining AFLPs with maternally inherited plastid DNA sequences provided strong evidence for past inter-ploidy gene flow as one fifth of the tetraploids exhibited a haplotype otherwise restricted to diploids. Further, small-scale ecological segregation between cytotypes was weak, but statistically significant. Leaf characters indicated little differentiation between the cytotypes, whereas flower characters had relatively strong discriminative power. Overall, from a taxonomic point of view, we suggest that clear genome-wide differentiation combined with consistent morphological differences in flower size do not preclude the differentiation of two distinct species (diploid V. gaultherioides and tetraploid V. uliginosum s. str.).