Gene flow and genetic structure of a mountain riparian tree species, Euptelea pleiospermum (Eupteleaceae): how important is the stream dendritic network?

Riparian landscapes are dendritic in nature. However, much attention on genetic structure of riparian plants has been paid to linear models of connectivity while studies that investigate the influence of dendritic landscape are scarce. In this study, we used nuclear microsatellite markers to investigate genetic diversity, gene flow, and genetic structure of a streamside tree species (Euptelea pleiospermum) in a natural stream dendritic network in the Shennongjia Mountains, central China. We tested the following hypotheses: (1) genetic diversity is higher at confluence than that at headwater populations and (2) genetic structure within the stream dendritic network was determined by in-stream dispersal or out-of-stream dispersal. Contrary to our prediction, we found that both genetic diversity and effective population size are congruent at headwater and confluence populations. We found symmetrical gene flow in most (four out of six) headwater–confluence pairs and asymmetrically downstream gene flow in the other two headwater–confluence pairs. Analysis of molecular variance (AMOVA) detected significant differentiation at two scales (among streams within catchments, among populations within stream) and did not reveal significant structure among catchments. STRUCTURE analysis clustered individuals from different catchments into the same genetically homogeneous group. There was no significant isolation by distance (IBD) with Euclidean, stream, or overland distance. Our results suggest that E. pleiospermum populations within the stream dendritic network did not present a hierarchical genetic structure probably because of extensive out-of-stream dispersal.