Forest regeneration is limited by, among other factors, dispersal from nearby existing forest species (seed rain). Understanding seed rain patterns will give insight into mechanisms of regeneration and inform management strategies where accelerated regeneration is required. Many floodplain forests of the Upper Mississippi River have declined as a result of catastrophic floods in 1993 that resulted in prolonged inundation and tree mortality. Subsequently tree canopy openings facilitated invasion of Phalaris arundinacea, a perennial rhizomatous grass. Efforts to clear P. arundinacea result in bare ground adjacent to forest edge, and rapid arrival of tree seed may reduce P. arundinacea reinvasion. These forests and adjacent areas present a model system for investigating seed rain patterns. We assessed seed rain from eight floodplain forests in National Wildlife Refuges in the Upper Mississippi River System from April to October 2008. We used 27 raised seed traps at each site placed within the forest and at stratified distances from the forest edge into an adjacent unforested area. Seed trap contents were recovered bi-weekly from April to November 2008. We asked: (1) how does seed rain differ with increasing distance from seed source? and (2) what temporal pattern does seed rain follow?

Results/Conclusions

Across sites, seed rain consisted largely of light-seeded species (Populus deltoides, Acer sacharinum, Fraxinus pennsylvanica, Betula nigra). Seeds of mature forest floodplain species (e.g. Quercus spp.) were not detected in this study. Time of dispersal and length of dispersal period differed with species, e.g. F. pennsylvanica dispersed over a 3 month period, but P. deltoides dispersed over a two week period. Seed rain was highly variable across the study region, but similar sites had similar seed rain. Within the forest, seed rain totals reached a maximum of 3600 seeds/m²; beyond the forest edge, seed rain totals reached a maximum of 1000 seeds/m². All species declined in seed rain density as distance from the forest edge increased. Study results suggest that there at some sites there may be an opportunity to exploit natural seed rain for revegetation efforts at distances of 25 m or less from the forest edge, where seed rain is dense enough to compete with P. arundinacea propagule pressure.