Publication

In many forests, advance regeneration represents an important 'seedling bank' for replacing overstory trees after canopy disturbance. However, long-term spatial and temporal dynamics of understory tree seedlings are poorly understood, particularly in topographically complex southern New England mixed-deciduous forests. From 1996 to 2005 we tracked the recruitment, growth and survival of seedlings of masting tree species across different topographic positions. Species included the largely valley associated sugar maple (Acer saccharum) and white ash (Fraxinus americana), the primarily ridge associated white oaks (Quercus alba and Q. prinus), and the site generalist red oaks (Q. rubra and Q. velutina) and red maple (A. rubrum). Four I m 2 circular plots were randomly established in nine different sites, which included three replicates of each topographic position (ridge, midslope and valley). Seedling recruits were tagged and measured annually. Densities varied largely between species, both over time and by topographic position. Ridge and midslope positions had similar pattems, with high densities of red oak (>200,000 seedlings/ha) declining over time, and other species generally fluctuating at much lower densities (<10,000 seedlings/ha). Trends differed in the valley, where higher white ash and sugar maple densities resulted in overlap with red oak. Overall declines in density were largely driven by a lack of significant recruitment events of red oak during the study, suggesting that seedling dynamics are driven by longer (decadal) cycles, likely reflecting the interaction of temporal variation in masting and conditions for establishment. Total recruitment for all species was positively related to overstory composition, as expected given topographic associations between species. Survival varied between species, but differed among topographic positions. White ash and sugar maple had relatively low survival in midslope and ridge positions relative to the red and white oaks, but higher survival than red oaks in valleys. Red maple had low overall survival, and only persisted beyond age 10 in the midslope position. Seedling heights generally increased with age for all species, but growth was slow, and many seedlings exhibited dieback. These survival patterns likely reflect effects of light limitation (in the valley) and moisture limitation (in the ridge) associated with topographic position. Overall, while seed inputs likely vary topographically related to differences in overstory abundance, differences in survival clearly reinforce these distributional patterns. In management terms, these survival patterns have significant site-specific implications for the type and timing of silvicultural interventions aimed at stand regeneration, particularly given infrequent events of seedling recruitment. (c) 2007 Elsevier B.V. All rights reserved.