Forest structure in unharvested old-growth: understanding the influence of soils on variability of long-term tree dynamics and fire history
Hasstedt, Steven C. M., author
Binkley, Dan, advisor
Rocca, Monique, committee member
Sibold, Jason, committee member
Martin, Patrick, committee member
Western North American forests adapted to frequent fires have been fundamentally altered and fragmented as a result of fire exclusion, the transportation infrastructure, development patterns, and other landscape level changes. In order to enhance the resiliency, diversity, and productivity of forest ecosystems it is essential to encourage collaborative, science-based restoration of forest landscapes and to develop a public understanding of the dynamic nature of forests. The USDA's Collaborative Forest Landscape Restoration Program (CFLRP) provides guidelines for community stakeholders and private advocacy groups to engage with federal, state, and local agency stewards to develop monitoring and assessment goals for forest restoration projects. The Uncompahgre Plateau (UP) was identified as one of ten initial CFLRP locations nationwide and has very large, old heritage trees scattered across never-harvested areas within the National Forest boundaries. Understanding how soil depth influences fire behavior and canopy development and determining the nature of historical fire regime on the UP are key elements for the development of local forest restoration prescriptions. Heritage trees appeared to occur more often on rocky soils, and I expected this could result from different fire behavior in landscape patches where soils are too thin to support dense forest cover and fuels. I tested the influence of soil depth on forest composition and the fire history of the study area in three phases. First, I recorded soil depth and site characteristics for 80 randomly selected plots and included targeted site sampling of all heritage trees (≥80 cm diameter) on two unroaded, never-harvested mesas to examine the relationship between soil depth, stand basal area, and the presence of heritage trees. The development of forest canopies (which influence fire regimes and tree survival) appeared to relate to soil characteristics (particularly rock cover). Single factor analysis revealed that soil depth alone only accounts for about 10% of stand basal area variation, but locations with soil > 30 cm deep had almost twice the basal area of locations with < 15 cm of soil depth. Comparing the observed to expected occurrence of heritage trees in four soil depth categories revealed a disproportionately greater presence of old-growth heritage trees on the locations with shallow soils. These results indicate that simple soil depth measurements can be used by restoration planners to develop stand-level spatial patterns. In the second phase, I used standard dendrochronology techniques to age trees from random plots, targeted heritage trees, and aspen transects to determine if historical stand structure patterns revealed forest age caps in concert with known landscape level historical fires. The spatial pattern of pre-1880 trees revealed that landscape level fires in the 1800s were likely not intense enough to kill all conifers over large areas, but were intense enough to kill most aspen stems on two of four sampled mesas. The mixed-severity nature of historical fires indicates that forest managers should have the leeway to plan for a spectrum of low to high severity fire effects within their restoration treatments. The third phase of my research tested the validity of my age-cap sampling methodology. I applied my sampling protocols to results from thirty systematic grid sampling locations composed of mixed conifer and spruce fir forests on the Kaibab Plateau, located on the North Rim of the Grand Canyon. My methodology targeted the largest trees and identified the oldest trees in the mixed conifer plots 97% of the time (in all but one plot). While a complete census of all trees would provide perfect information on the presence of an age cap, the validation of my targeted sampling approach provides a high confidence method to characterize the oldest trees in sampling locations with a substantial savings in the amount of time and resources expended. The subsequent ability to characterize the historical fire regime at the scale of the sampling design provides forest managers with another tool to inform restoration prescriptions. Applying the knowledge gained from the unroaded, never-harvested mesas to similar forest types on the Uncompahgre Plateau will guide landscape-scale treatment planning designed to restore ecosystem structure, composition, and function while reintegrating and managing wildfire as a natural component to reduce the risk of unnaturally severe crown fires.
Includes bibliographical references.
Includes bibliographical references.