Spatial arrangement and genetic structure in Gentianella aspera in a regional, local, and temporal context

Abstract

Gentianella aspera is a biennial plant of various nutrient-poor grasslands that has become rare in the landscapes outside the Alps of eastern Austria. Using AFLP fingerprinting we investigated: (1) effects of spatial structure on genetic structure in a large vineyard population that is confined to the embankments separating the grapevines; (2) temporal variation in genetic diversity and structure in this population; (3) relationships with other populations in a regional context. On the regional scale, moderate isolation by distance among populations was revealed by a Mantel test. Bayesian analysis of population structure indicated three spatially distinct gene pools and an additional one within the vineyard population. Within this population, spatial autocorrelation analysis revealed a positive correlation between genetic and spatial distance up to 50 m. Patterns found by PCoA were not in line with a priori defined subpopulations and indicated substantial gene flow across embankments. AMOVA revealed low differentiation among both the subpopulations that were found on the linear embankments and among two local groups of these subpopulations. We found, however, striking differences in the among-group variation between the 2 years, i.e., between two local groups within the generations and between those groups among generations. This was due to the highly variable larger group of the younger generation, in which an additional gene pool was identified by Bayesian analysis of population structure. Based on these results we discuss scenarios of local and regional dynamics within and among G. aspera populations.