COS 48-1 - Quantifying and restoring stand-level spatial pattern in dry mixed-conifer forests of Eastern Washington

Tuesday, August 7, 2012: 8:00 AM
C120, Oregon Convention Center
Derek J. Churchill, School of Environmental and Forest Sciences, University of Washington, Seattle, WA and Andrew J. Larson, College of Forestry and Conservation, University of Montana, Missoula, MT
Background/Question/Methods

Pre-settlement reference conditions provide much of the scientific basis for the ongoing restoration of millions of hectares of fire-prone, dry mixed conifer forests in the interior western United States. There is increasing recognition that structural targets for fuels reduction treatments should incorporate within-stand spatial pattern to increase resilience. Tree spatial pattern has been shown to influence key aspects of ecological resilience such as fire behavior, regeneration, insect behavior, nutrient cycling, snow retention, habitat, and understory plant diversity and abundance. Past research has shown that these forest types were historically complex mosaics of openings, clumps, and individual trees. Yet methods to quantify and explicitly incorporate spatial information from reference conditions in the design and/or monitoring of restoration and fuels reduction treatments are not widely used. A primary reason is the pervasive use of global, or stand average, spatial analysis methods that do not report the within-stand spatial variability of forest in reference plots. We present new methods based on percolation based clustering and local hot spot analysis that quantify within-stand structure in terms of lower-level patch types in an intuitive and ecologically relevant manner. We use these new methods in conjunction with widely used global methods to characterize the range of within-stand pattern and structure in pre-settlement dry forests at two scales in the Eastern Cascades of Washington State. We present data from three new 5ha reconstruction plots and compare these to data from 32 previously published 0.5 ha plots.

Results/Conclusions

Results show wide variation in pattern at both the 0.5 and 4.5-5.5 scales, but with clear upper and lower thresholds. Structure and pattern on some plots indicate past occurrence of mixed severity fire, and edaphic or other factors appear to create non-stationary patterns. Our results indicate that restoration thinning prescriptions should incorporate openings and large clumps, as well as widely spaced individual trees and small clumps. We then demonstrate how our methods can be directly translated into restoration thinning and monitoring guidelines. Results from case studies implementing this method show that it is an operationally efficient method of managing stands as patch mosaics with pattern targets tied to reference conditions. Finally, we present information on how to adapt the method in situations where significant climatic shifts are anticipated.