Abstract This study deals with populations of the European-South-Siberian geoelement Adenophora liliifolia (L.) A. DC. in the Czech Republic, Slovakia, Hungary, Romania, and Poland, where this species has its European periphery distribution. We studied the population size, genetic variability, site conditions, and vegetation units in which A. liliifolia grows. Recent and historical localities of A. liliifolia were ranked into six vegetation units of both forest and non-forest character. A phytosociological survey showed differences in the species composition among localities. Only a weak pattern of population structure was observed (only 22% of total genetic variation present at the interpopulation level, AMOVA analysis), with moderate values for gene diversity (Hj = 0.141) and polymorphism (P = 27.6%). Neighbor-joining and Bayesian clusterings suggest a similar genetic background for most of the populations from Slovakia, the Czech Republic, and Poland, contrary to the populations from Hungary, Romania, as well as two populations from Central and South Slovakia. This might be explained by a relatively recent fragmentation of the A. liliifolia populations in Central Europe. Nevertheless, it seems that several populations in Romania, South Hungary, and Slovakia were isolated for a longer period of time and their genetic differentiation is more evident.

Abstract Viola uliginosa (bog violet) is a declining species throughout its range due to – mostly anthropogenic – drying out of the wet habitats it occupies. Using AFLP markers, we aimed to estimate the genetic diversity in Polish populations, that may give an insight into the situation of plant populations facing rapid loss of natural habitats. Bog violet from several dispersed Polish populations is generally characterized by very low genetic diversity (HT = 0.048), even lower than several other endangered violets; therefore, we suggest that it should preserve at least EN rank in the red lists/red data books. The mean gene diversity within all populations (HS) was much lower than gene diversity (GST) between populations (0.020 versus 0.583, respectively) which supports the prevalence of clonal propagation of the species (mainly by stolons) but may also point to some significance of autogamy in cleisto- and chasmogamous flowers. A high FST value and the Mantel test for all populations revealed significant isolation by distance. Geographically neighboring pairs of populations formed genetic clusters supported by all (in the case of two closest populations) or most statistical analyses applied. Special attention should be paid to the locus classicus of the species in Rząska, consisting of a small number of individuals, forming a genetically distinct group, revealing very low gene diversity (Hj = 0.009) and the longest genetic distance to the remaining populations. Our results can contribute to planning future protection measures for the species at this and other locations. Genetic structure of the studied populations suggests local affinities of populations but does not generally support hypothesized recent continuity of V. uliginosa range along the river valleys of southern Poland; this view may, however, be altered with widening of the scope of studied populations and chosen molecular markers.