River network architecture, genetic effective size and distributional patterns predict differences in genetic structure across species in a dryland stream fish community.

Abstract

Dendritic ecological network (DEN) architecture can be a strong predictor of spatial genetic patterns in theoretical and simulation studies. Yet, interspecific differences in dispersal capabilities and distribution within the network may equally affect species' genetic structuring. We characterized patterns of genetic variation from up to ten microsatellite loci for nine numerically dominant members of the upper Gila River fish community, New Mexico, USA. Using comparative landscape genetics, we evaluated the role of network architecture for structuring populations within species (pairwise FST ) while explicitly accounting for intraspecific demographic influences on effective population size (Ne ). Five species exhibited patterns of connectivity and/or genetic diversity gradients that were predicted by network structure. These species were generally considered to be small-bodied or habitat specialists. Spatial variation of Ne was a strong predictor of pairwise FST for two species, suggesting patterns of connectivity may also be influenced by genetic drift independent of network properties. Finally, two study species exhibited genetic patterns that were unexplained by network properties and appeared to be related to nonequilibrium processes. Properties of DENs shape community-wide genetic structure but effects are modified by intrinsic traits and nonequilibrium processes. Further theoretical development of the DEN framework should account for such cases.