The influence of episodic reproduction by red spruce on long-term forest dynamics in the Cranberry Wilderness, West Virginia

Episodic reproduction ("masting") by shade-tolerant tree species could be a significant influence on long-term forest dynamics. A recent episode of high seed production by red spruce (Picea rubens Sarg.) in second-growth (64-71 yr old) forests of the Cranberry Wilderness, West Virginia provided an opportunity to examine seedling mortality, the development of seedling spatial patterns, and possible effects of reproductive episodes on long-term forest dynamics. Canopy conditions in the spruce forests were quantified to infer their future influences on regeneration. Spruce comprised 81% of the live basal area and 65% of live stems. Most live spruce were in codominant crown positions, while dead spruce were mostly small, suppressed trees. One third of live spruce, mostly intermediate or suppressed trees, showed >10% needle loss and had below-average crown ratios. Over half of live spruce showed some loss of recent needles, and 64% showed narrowed or one-sided crowns. Little evidence of fungal or insect pathogens was found. These data suggest typical post-logging stand development. Continued dominance by spruce in the canopy should favor shade-tolerant spruce seedlings. A large seed crop by spruce in Cranberry generated 2,944,000 new seedlings per hectare by August 1990. Changes in log_e(density) of this cohort were adequately explained by linear and negative exponential survivorship curves. A power function fit the data less well. These models gave widely differing predictions of seedling density after 5 yr. The linear function suggested little effect of this cohort on seedling pool size, the negative exponential model suggested that pool size would triple, and the power function predicted that pool size would increase 8-fold. First-year mortality was inversely correlated with initial density, indicating that conditions favoring germination also favored early survival. Seedling associations with rooting substrates changed erratically, perhaps because of the shallow rooting of seedlings. A strong association of seedlings with bryophyte-covered surfaces developed early and persisted. Declining seedling mortality, implicit in the negative exponential and power models, should allow this cohort to dominate the seedling pool for several years. The long-term effects of such episodic recruitment were assessed with a forest gap model. An existing model was modified to simulate semi-periodic recruitment of spruce in West Virginia forests under several scenarios: logging, varying available soil nitrogen, a 50% spruce growth decline, and varying nitrogen combined with spruce decline. Episodic recruitment reduced the importance of yellow birch (Betula alleghaniensis Britton), even after logging. Varying soil nitrogen reduced birch importance with or without spruce masting. A spruce growth decline allowed birch to become more important, but masting moderated decline effects. Varying nitrogen combined with spruce decline resulted in low spruce and birch basal areas and high densities of spruce and red maple (Acer rubrum L.). The addition of masting to this scenario nearly eliminated birch. Masting appeared to act as a stabilizing influence on spruce dominance by reducing recruitment of its competitors. Highly variable seedling recruitment may noticeably influence long-term forest dynamics.