Hybridization has been repeatedly put forward to explain the invasiveness of Rhododendron ponticum L. in the British Isles. The present study investigates the pattern of ecotypic differentiation and hybridization among native North American R. catawbiense and R. maximum, native R. ponticum from Georgia and Spain, and invasive R. ponticum from Ireland and aims to assess the contribution of hybridization for Rhododendron invasion in the British Isles. Six populations per taxon were analyzed with AFLP markers for genetic dissimilarity, subjected to germination and growth experiments, and tested for frost hardiness. We assessed variation in morphological and ecological characteristics to identify traits displaying evidence of hybridization, thus, promoting invasiveness. Molecular marker analyses revealed a clear distinction between North American R. catawbiense and R. maximum on the one hand, and all R. ponticum populations on the other hand, displaying a complete intermixture of native Spanish and invasive Irish populations. Multivariate analyses of traits revealed leaf length–width ratio, relative growth rates (RGRs) in leaf length, root biomass, and shoot–root ratio to significantly discriminate between the different taxa and unequivocally assigned invasive Irish R. ponticum to the Spanish phenotypes. While the Irish R. ponticum had similar growth traits as conspecific native R. ponticum provenances, germination and biomass allocation were more similar to North American R. catawbiense and R. maximum. Hybridization did not contribute to explaining invasiveness of R. ponticum in Ireland. The similarity in germination and biomass allocation of invasive Irish R. ponticum and North American species has evolved independently and can more probably be attributed to an independent shift within the Ponticum cluster in Ireland.

No abstract

No abstract

Habitat fragmentation decreases plant population size and increases population isolation, as well as altering patterns of plant–animal interactions, all of which may reduce plant fitness. Here, we studied effects of habitat fragmentation (in terms of population size and isolation) and soil quality on the reproduction of two rare legume species, Genista anglica (13 populations) and Genista pilosa (14 populations), confined to remnants of acidic and nutrient‐poor Calluna heathlands. Single individuals of the Genista plants are impossible to distinguish; population size was therefore estimated according to the area occupied (referred to as population size hereafter). We collected soil samples in all heathland sites to determine content of soil water, C, N, P, Ca, K and Mg. In both species values of soil pH and C/N ratio, as well as content of soil P and base cations, reflected the highly acidic and nutrient‐poor environment of the heathlands. Population sizes were unrelated to soil quality. Although the two Genista species are similar in morphology and ecology, effects of explanatory variables on reproduction were largely inconsistent across species. In G. anglica, population size had a positive impact on all reproductive variables except germination rate, which, in contrast, was the only variable affected positively by population size in G. pilosa. In both species, mean total reproductive output, calculated as the product of total seed mass per shoot and total germination, increased with increasing water content and decreased with increasing P. In G. anglica, we found positive effects of the C/N ratio on all reproductive variables except mean single and total seed mass per shoot. In summary, in both species reproductive success per shoot decreased with increasing soil nutrient availability in the heathland sites. The infestation of two large populations of G. pilosa with the pre‐dispersal, seed‐predating weevil Apion compactum had no significant effect on reproduction of the populations.