Abstract:
Ecological processes are responsible for vegetation trajectory within the boreal forest landscape of interior Alaska. The reproductive response of boreal forest to disturbance controls vegetation trajectory. Boreal forest reproduction dynamics are influenced by both biotic and abiotic factors, acting upon the spatio-temporal dynamics of the landscape. Understanding these factors and how the boreal forest responds, both spatially and temporally, is critical for the development of accurate models of regional and global vegetation dynamics. I developed a geographic model of the early post-disturbance seedling regeneration pattern of upland white spruce ecosystems in interior Alaska. The model was developed and runs within a geographic information system (GIS). The model simulates the establishment patterns of white spruce, paper birch, and aspen across the landscape following fire. Seed production and dispersal, disturbance effects upon the seedbed, and the early establishment of both seedlings and vegetative stems are simulated. The model was used to simulate a 6 yr period (1983-1988) of seedling establishment at the Bonanza Creek Experimental Forest near Fairbanks, following the Rosie Creek fire. Correlation values between predicted and established seedlings were high, demonstrating the model's ability to simulate general establishment patterns. Sensitivity analysis revealed seed production, seed source location and orientation, and seedbed "receptivity" as important controls upon the early establishment success of white spruce seedlings following disturbance. Establishment patterns between a hypothetical clearcut, strip-cut, and residual tree islands cut were simulated and compared. Distance from the seed source was identified as a major limitation to adequate stocking levels in the clearcut. The residual islands cut provided the highest stocking levels, followed by the strip-cut and clearcut. The results suggest large clearcuts are not an efficient harvesting method in interior Alaska for successful natural regeneration and stocking levels. The model results warrant further development and identified a "real" potential use as a forest management tool.
