LB 23-240 - Neighborhood light availability in forested landscapes and application for natural disturbance-based management simulations in Southern Appalachia

Thursday, August 18, 2022

5:00 PM – 6:30 PM EDT

Located in: ESA Exhibit Hall

Presenting Author(s)

AS

Andrew Shannon

Center for Geospatial Analytics, North Carolina State University, United States

Coauthor(s)

RS

Robert M. Scheller

Assc. Dean of Research
Department of Forestry & Environmental Resources, North Carolina State University, United States

Background/Question/Methods:

Species distributions will shift as climate changes, making an understanding of regeneration dynamics key for forest management. This requires quantifying how the transitional light gradient from open-canopy gap to closed-canopy benefits the long-term regeneration of moderately shade tolerant species, such as oak. We developed light availability metrics for an upland hardwood forest within the Bent Creek Experimental Forest of Southern Appalachia. We derived digital terrain and canopy height models from aerial LiDAR to compute subcanopy solar radiation during the growing season using a solar raycasting model. We tested methods of neighborhood analysis to assess the best representation of subcanopy solar radiation using the neighborhood Leaf Area Index (LAI) as a proxy for canopy structure. This algorithm was used to test management scenarios within a forest landscape simulation model, LANDIS-II. Previously, light availability for regeneration in the model was a function of LAI only within its cell. This approach did not consider neighborhood effects of surrounding canopy features, such as canopy gaps, and landscape topography on light availability. We used the complete model to simulate a harvest-based silvicultural practice that establishes small canopy gaps for the maintenance of the light environment preferential for oak regeneration and assessed its efficacy.

Results/Conclusions:

Preliminary results show that correlation between observed cell-level LAI and modeled light availability increased when adjusted for modeled neighborhood LAI. The optimal size of the neighborhood analysis was dependent on the spatial grain of rasterized LAI. Finer scale parametrization of LAI at 3- and 10-meter resolution improved the representation of the light gradient from open-canopy gap to closed-canopy compared to coarser scale parameterization (30-meter). Estimating transitional light environments is crucial to determining optimal regeneration for oaks and other tree species. The use of three dimensional structure models, such as those derived from LiDAR, and raycasting advance our understanding of the solar radiation dynamics in forests. Using these tools and our results on neighborhood LAI and modeled light availability, we can improve our estimation of species-specific probability of establishment as compared to empirical data. This improvement will enhance modeled outcomes of future species composition and age class distribution for simulations of the Bent Creek Experimental Forest. These results will be crucial for adapting management efforts to changing forest structure and regeneration conditions.