COS 157-4 - Using legacy structure to guide ecological restoration in the Crater Lake Panhandle, Oregon

Thursday, August 10, 2017: 2:30 PM
B113, Oregon Convention Center
Sean M.A. Jeronimo, Van R. Kane, Derek J. Churchill and Jerry F. Franklin, School of Environmental and Forest Sciences, University of Washington, Seattle, WA

We developed a method for estimating the number, sizes, and spatial pattern of large and old trees across a dry mixed-conifer forest landscape in Crater Lake National Park, Oregon and used these data to constitute the basis of an ecological forest restoration treatment. We used airborne LiDAR individual tree detection to pick out tree crowns across the 200 ha study area. Since many tree crowns are hidden from aerial view by the dominant canopy layer, we considered the detected trees to be tree-approximate objects (TAOs). TAOs are typically dominated by a single tree, but often also encompass a few subordinate trees. We used TAOs to identify “backbone trees,” representing large and old trees that are contributing disproportionately to ecological functioning and are keystone structures in terms of ecosystem resistance and resilience.

We compared the current conditions measured with TAOs to various reference conditions characterizing the historical (ca 1880-1920) structure and pattern of climatically similar forests. We compared metrics of density, basal area, distribution of tree clump sizes, and distribution of inter-tree open space. Lastly, we used these comparisons to inform restoration prescriptions anchored around the backbone trees.


The study area was generally at a higher density than the envelope of historical conditions, but was within the historical envelope in terms of basal area. Current conditions had more very large clumps (>30 trees), fewer small and medium sized clumps (2-9 trees), and fewer large openings >9 m radius compared to historical conditions. Focusing on only the backbone trees, density and basal area were both low compared to the historical envelope, while spatial pattern was more regular: more individual trees and small clumps (<5 trees) and fewer medium and large clumps (5-15 trees). We prescribed variable density thinning treatments with explicit clumping targets for young-tree retention and opening creation.

We concluded that TAOs can be used to assess density, stocking, clumping, and opening patterns in a way that is sufficient to assign restoration prescriptions to treatment units. TAOs provide an excellent means for understanding the largest trees in a stand, which are the biological anchors guiding ecological restoration. Restoration prescriptions still require ground-based reconnaissance and inventory; however, the TAO analysis provides a landscape-level context that is generally not available by other means.