In urban riparian restoration projects, trees and shrubs are typically planted, while understory herbaceous plants are often left to colonize on their own. There is little evidence, however, that desired herbaceous species recover on their own; dispersal-limitation may constrain the potential for this approach to be successful. While streams are known to connect otherwise isolated plant populations, more flashy hydrology and channelized morphology of urban streams may make them less effective dispersal vectors. The purpose of this study was to quantify seed deposition by small streams across a gradient of watershed urbanization. It was hypothesized that as urbanization increases, the overall number of seeds deposited by streams would decrease; it was further hypothesized that, (a) large seeds, (b) seeds from herbaceous plants, and (c) seeds with no special dispersal structure would become more limited. Seed traps (n=12) consisting of turf mats were installed on stream banks in ten riparian forests along a gradient of watershed urbanization, ranging from 1% to 41% total impervious area, in the Portland, Oregon metropolitan area. Funnel traps collecting wind-deposited seeds were installed throughout study sites (n=20), allowing for comparison with stream-deposited samples. Samples were collected from seed traps in March, June, and October of 2015.
In all sites, more seeds were collected in turf traps than in funnel traps in all three sample periods, indicating that more seeds are deposited by water than by wind for a given location along the stream corridor. In all sites, the highest total numbers of seeds were deposited in the winter months when stream flows were the highest. The total number of seeds deposited by streams decreased as watershed urbanization increased. The largest difference in seed deposition across the gradient was seen in winter months (R² = 0.51, p < 0.05). The number of tree seeds collected by traps did not vary across the urbanization gradient, however, consistent with hypotheses, there were fewer shrub and herbaceous seeds collected in urban sites than in rural sites. While streams may facilitate passive restoration by depositing more seeds than wind alone, the reduction in seed deposition in the most urban sites suggests that a passive approach will not be as successful here as in rural areas. Results suggest that urban streams are not likely to allow shrub and herbaceous species to overcome seed limitation, and managers should consider more active means for understory plant restoration in these sites.