PREDICTING SEED/SEEDLING DYNAMICS: THE NON-TAROT CARD APPROACH. Robert Cosgriff1, Yao Yin2, Theresa Blackburn3, Thad Cook4, Heidi Langrehr5, and Megan Moore6 1Great Rivers Field Station, Illinois Natural History Survey, 8450 Montclair Avenue, Brighton, IL 62012 2USGS, Upper Midwest Environmental Sciences Center, 575 Lester Avenue, Onalaska, WI 54650 3Iowa Department of Natural Resources, Mississippi River Monitoring Station, 206 Rose Street, Bellevue, Iowa 52031 4Illinois River Biological Station, Illinois Natural History Survey, 704 North Schrader Avenue, Havana, IL 62644 5Wisconsin Department of Natural Resources, Onalaska Field Station, 575 Lester Avenue, Onalaska, WI 54650 6Minnesota Department of Natural Resources, Lake City Field Station, 1801 South Oak Street, Lake City, MN 55041. Predicting future floodplain forest composition is an arduous task that requires an understanding of how the annual flood pulse, various forms and degrees of herbivory, and inter/intraspecific competition effects seed germination and seedling survivorship. A seed/seedling dynamics study was conducted from 1996- 2001 on five key pools of the Upper Mississippi River System (UMRS). Twenty- four species were identified on the five pools in which Populus deltoides, Betula nigra, Acer saccharinum, Fraxinus pennsylvanica, Ulmus americana, and Acer negundo were the most common species encountered. River pool, forest community, species, and month were significant predictors of seed density. Year was not a significant predictor of seed density. This is due to the dominance of light-seeded species, which typically do not have masting years. River pool and germination year were significant predictors of seedling density, whereas forest community, germination month, and species were not. It is likely that the dominance of a few light-seeded species at all of the forest sites played an integral role in decreasing the effect of forest community and species in determining seedling density. Although B. nigra produced the greatest number of seeds and had high reproductive power, this species had poor representation in the seedling cohorts of the sites studied (germination ranged from 0.0-0.15%). Likewise, P. deltoides had no seedling germination in any of the sites studied. The percent germination of A. saccharinum varied from 0.35% in the wet forests of the La Grange Pool to 218.2% in the wet forests of Pool 13. Greater than 100% germination rates for A. saccharinum seeds may be explained by the additional deposition of seeds following annual flooding. Fraxinus pennsylvanica and U. americana seed germination ranged from 0.0-25.3% and 0.92-85.0%, respectively. Hard-mast species (Quercus and Carya spp.) had much lower germination rates than what was expected (0.0-25.3%). This is likely related to high mortality in seed crops due to predation. Due to low numbers of seedlings in all but five species, survivorship analysis was conducted for F. pennsylvanica, A. saccharinum, U. americana, Rhamnus cathartica, and Zanthoxylum americanum. Seedling survival was strongly related to species, river reach, and day of mortality. Fraxinus pennsylvanica, Z. americanum, and U. americana had greater long-term survivorship than did A. saccharinum and R. cathartica. Survivorship within a species also varied according to forest community. Fraxinus pennsylvanica had a greater chance of surviving on dry forests than wet (Chi-square=32.74, df=1, p<0.001) whereas A. saccharinum had a greater chance of survivorship on wet forests (Chi- square=41.76, df=1, p<0.0001). All river reaches other than Reach 4 and the La Grange Reach showed significant differences in seedling survivorship curves. Likewise, there were significant differences in survival curves between forest communities. Dry forests had greater seedling survivorship than wet forests. Chance of survivorship increased as seedling age increased. The ability of Fraxinus pennsylvanica to survive long-term on the floodplains of the UMRS, coupled with abundant conspecific adults, adequate seed production, and adequate germination rates, indicate that this species will dominate the dry forests and be a major component of the wet forests. For the same reasons, A. saccharinum will dominate the wet forests. Quercus and Carya spp. need special attention as these species are failing to regenerate into the seedling and sapling cohorts. Planting might be the only solution to preserving Quercus- Carya forests long-term. Keywords: floodplain forest, seed dynamics, seedling dynamics, Upper Mississippi River System