The seed banks of Mid-Atlantic forest fragments: Influx of non-native species and implications for forest regeneration
Fragmented forest landscapes surrounded by a matrix of non-forest vegetation contain more non-native species than continuous forests. Increased abundance of non-native species creates more competition for native species and may reduce native populations. Long-term persistence of native species in forest fragments is uncertain because it is unclear whether non-native species will continue to increase in abundance. The purpose of this study was to determine the prevalence of native and non-native species in the persistent seed bank of eastern deciduous forest fragments, while also assessing the potential for native tree regeneration. The seed banks of three forest fragments, Gordon Natural Area (GNA), Mount Cuba Center (MC), and White Clay Creek State Park (WC), were studied in the Mid-Atlantic region. Each site contained five plots at which 20 soil cores were collected at three depths (0-5, 5-10, 10-15cm). Trees > 5 cm dbh within 0.2 ha of plot centers were measured and identified. Soils were composited, sieved, and placed on trays in a greenhouse where they were watered daily for 5.5 months. Germinated plants were identified and tallied. Two-way analyses of variance were used to determine whether the number of native and non-native species and individuals varied by site or depth interval.
A total of 1026 individual plants germinated, 428 (42%) were native while 598 (58%) were non-native. Sixty species were identified, of which 34 (57%) were native and 26 (43%) were non-native. The GNA seed bank was comprised of 82% non-native individuals and 48% non-native species. In terms of individual plants, the seed banks of WC and MC were approximately 20% non-native, while on a species basis these seed banks were 34% and 40%, respectively, non-native. Only two native tree species germinated (Liriodendron tulipifera and Betula lenta), despite diverse native canopy cover at all sites (11 to 14 species per site). The number of germinated native species did not vary across sites (p > 0.05), but the number of native individuals varied significantly by site (p = 0.007), with fewer native individuals in the GNA. The number of non-native species and non-native individuals also varied significantly across sites (p < 0.0001 and p = 0.003, respectively). Although there are more native species than non-native species, non-native individuals outnumber native individuals overall. These findings show potential for shifts in forest regeneration caused by increasing numbers of non-native individuals.