Habitat loss and fragmentation are primary factors in biodiversity decline. A major means by which fragmentation effects are thought to manifest at the population level is through impacts on dispersal, with consequences for colonization of new habitats, rescue effects of existing populations, and local-scale demography. However, there is little understanding of how fragmentation affects dispersal. Here, we focus on how fragmentation affects seed dispersal by wind, through influences on patch connectivity, shape (edge:area ratio), and distance to edge. Our work took place within eight replicated experimentally-fragmented landscapes; each containing patches of equal area, differing in shape and connectivity. In our system, wind speeds are higher in connected patches, patches with high edge:area ratios, along the long axis of patches, and at distances far from edges. We outplanted five primarily wind-dispersed grass/forb species into our landscapes at four edge distances. Prior to dispersal, we marked seeds of 728 individuals with fluorescent dye, allowed seeds to disperse, then relocated ~32,500 seeds post-dispersal. We measured the distance and direction that each seed dispersed from its parent (within a 5m radius). We hypothesized: (1) Patches with higher connectedness and edge:area have greater dispersal distances. (2) Individuals farther from an edge experience greater dispersal distances. (3) Dispersal along the long axis of a patch leads to increased dispersal distance. (4) The effect of dispersal direction is stronger at locations farther from edges.
Results/Conclusions
(1) Contrary to expectations, differences in connectivity and edge:area ratio had little effect on dispersal distance. (2) However, species did display a strong edge effect, generally dispersing greater distances when growing further from a habitat edge, in line with our predictions. (3) The direction of dispersal had an effect on dispersal distance, but not in line with the long axis of the patch. Rather, seeds dispersing in directions away from the edge typically dispersed farther. (4) Dispersal was more strongly directional with increased distance from the habitat edge, as predicted. Our results demonstrate that while some landscape factors such as connectivity appear not to significantly predict dispersal distance, other factors such as distance from an edge and dispersal direction play major roles. Our findings may be useful in reserve design in the face of increasing need for species dispersal in an increasingly fragmented world, as well as for decision making on how to maximize dispersal distances when planting native species in restorations.