Implications of seed banking for recruitment of southern Appalachian woody species

Seed dormancy is assumed to be unimportant for population dynamics of temperate woody species, because seeds occur at low densities and are short lived in forest soils. However, low soil seed densities may result from low seed production, and even modest seed longevity can buffer against fluctuating seed production, potentially limiting density-dependent mortality and ensuring that seeds are available for germination when recruitment success is likely. To investigate whether seed banking affects woody seedling dynamics in the southern Appalachians, we monitored seed rain, seed bank, and seedling densities to (1) determine the prevalence of seed banking among southern Appalachian woody species, (2) quantify annual seed mortality rates for three seed-banking species using a Bayesian statistical approach, and (3) assess whether or not the ability to,seed bank affects recruitment rates. We found that the seeds of eight woody taxa (Acer rubrum, Betula spp., Liriodendron talipifera, Nyssa sylvatica, Robinia pseudoacacia, Rubus spp., Sassafras albidum, and Vitis sp.) remain viable in the soil for more than one year. Seeds of six taxa (Amelanchier spp., Acer pennsylvanicum, Carya spp., Quercus prinus, Quercus rubra, and Tsuga canadensis) were never found in the soil seed bank, despite high seed production and germination. For three species, a substantial proportion of seeds available for germination came from dispersal events two or more years in the past (Acer rubrum 12-37%, Betula spp. 59-73%, Liriodendron tulipifera 40-76%), even though annual seed mortality was high (Acer rubrum 70-98%, Betula spp. 21-81%, Liriodendron tulipifera 12-59%). In years when no seeds fall in local microsites (approximately one in five years), seed banks are the only source of seedling recruitment for these species. Comparing our results to those of previous studies led to valuable insights: first, that seeds of Acer rubrum and Betula spp. suffer high mortality while being incorporated into the seed bank; and second, that seed decay varies greatly over relatively small spatial scales (i.e., within a watershed). Taken together, these results demonstrate that seed banking may play a critical role during woody seedling recruitment in temperate forests.