Abstract Methods: Field data Increased tree establishment in subalpine parkland and an upward movement of treeline are expected in a warming climate, and may already be occurring in some ecosystems, but other factors, such as changing disturbance regimes, have received less attention. We examined the effects of fire on subalpine parkland with dominant tree species of Abies lasiocarpa, Larix lyallii, Pinus albicaulis, and Picea engelmannii in the northern Cascade Range of Washington. We used field data from two burned areas, 17 and 25 years after fire, to investigate relationships between fire severity and post-fire herbaceous species composition and tree regeneration across the subalpine parkland ecotone. We sampled 68 and 45 plots within 1994 Butte Creek and 1985 Hubbard Creek fire perimeters, respectively, to assess herbaceous species composition across a range of fire severities. We used plant functional groups, classified based on growth form, to investigate how the pre-fire setting (closed forest, open forest, alpine woodland, krumholtz, alpine treeless vegetation) interacts with the severity of fire to affect the abundance of plant functional groups. We also examined the influence of pre-fire overstory tree structure and fire severity on post-fire seedling density and composition, across 151 and 100 plots within Butte Creek and Hubbard Creek, respectively. Data was collected in 2012 within the 1985 Hubbard Creek fire (n=100) and the 1994 Butte Creek fire (n=151). Plant functional groups differed between burned and unburned sites and along the gradient from closed forest to treeless alpine vegetation. We found significantly more cushion plants in unburned sites than burned sites. Sites that burned with high and very high severity had significantly more graminoids, herbaceous perennials, and tall shrubs. Likewise, closed forest and open forests had significantly more graminoids in burned than unburned areas. Seedling density was positively associated with burn severity in areas where there was a closed canopy forest before the fire, but showed inconsistent relationships with fire severity and forest structure in open forests and alpine parkland. Our initial results indicate that post-fire succession dynamics in A. lasiocarpa forests in the eastern Cascades may be strongly influenced by heterogeneous patterns of severity resulting in heterogeneous survival of seedlings and overstory trees. Results Figure 2. Map of Hubbard Creek and Butte Creek study areas 2. Post-fire seedling establishment Plot measurements Used in this analysis: • Burn severity (see below) • Structural classes (see below) • Seedling trees (<1.4m height) • Herbaceous vegetation transect Will be used in future analysis: • Overstory trees (>1.4m height) • Distance to nearest tree seed sources Box-plots show post-fire seedling densities (per ha) across all species combined: • No significant difference in seedling densities between burned and unburned areas, burn severity classes. • In combined models (severity * pre-fire structural class) differences in pre-fire structure were stronger and more significant than any differences between burned and unburned plots of severity classes. • In combined models there were significant species-specific responses to fire severity within open forest and parkland structural classes. Table 1. Burn severity classes Class Class description Unburned No evidence of fire anywhere Table 2. Structural classes Very Low Low burn severity; only small portion of area effected Class Class description Low Surface fire; few, if any trees killed Closed forest > 40% tree canopy cover Moderate Surface fire; 20-70 % trees killed Open forest 10-40% tree canopy cover High 50-100% of trees killed; >75% of surface fuels consumed Parkland 5-10% tree canopy cover; trees often in clumps Very high 100% of trees killed; >90 surface fuels consumed; soil altered Alpine No overstory tree cover. Krumholtz tress may be present Pre-fire Fire pattern Post-fire Figure 2. “State-transition models” provide a conceptual framework for this research. The probability of fire impacting different parts of the alpine treeline ecotone is influenced by the prefire structural class and their position in the landscape. Likewise the probability of a persistent transition to a different structural type is influenced by the severity of the fire and the position on the landscape (e.g. environmental stress at that location). 3. Structural class transitional probabilities Table 3. Hubbard Creek (1985) structural transitional probabilities Pre-fire Structural class Research questions 1. 2. 3. How likely are different pre-fire structural classes to be impacted by fire of burn at different severities? How likely are different pre-fire structural classes transition to a different structural class after fire? How do seedling densities after fire differ among fire severity classes and pre-fire structural classes Closed forest Alpine 0.98 0.02 0 0 Parkland 0.32 0.68 0 0 Parkland 0.39 0.59 0.02 0 Open forest 0.31 0.34 0.34 0 Open forest 0.37 0.09 0.54 0 Closed forest 0.26 0.22 0.3 0.22 Closed forest 0.36 0.09 0.09 0.45 Future research We will conduct similar analysis using data-driven classification of pre-fire and post-fire plant structure and a continuous burn severity index. Additional topographical variables will be derived from DEMs and used as predictors in models. We will expand the analyses using field data from additional fires in the Cascade Range (n=2) and Northern Rockies (n=2). Models of burn severity probabilities and structural class transitional probabilities will be expanded using a larger data set of classified pre-and post-fire aerial photography. Acknowledgements Graminoids have a positive response to fire, as frequency increased after high severity fire in all forest classes. Figure 1. Climate change will have direct impacts on tree establishment and survival in the alpine treeline ecotone, as well as the annual area bend in forested ecosystems. This research addresses uncertainties regarding the direct effects of fire and post-fire successional dynamics in modern warming climates in alpine treeline ecotones. Open forest 0 Plant species that have a cushion or mat-forming growth form are less frequent in severely burned plots, primarily within the parkland structural class. GRAMINOIDS HERBACEOUS PERENNIALS Parkland 0 Dwarf shrub species occur at low frequency in closed forest and alpine environments; we could attribute differences in frequency due to forest structural class, but not fire severity or occurrence. Alpine 0 Results: Herbaceous functional groups DWARF SHRUBS Closed forest Post-fire structural class 1 Background Open forest Pre-fire Structural class Parkland Figure 4. Examples of pre-fire structural components: alpine, parkland, open forest, closed forest CUSHION/MAT Post-fire structural class Alpine Alpine Climate change may increase area burned and frequency of fire in highelevation ecosystems, which could have long-term impacts on ecosystem functions such as wildlife habitat use, hydrology, and nutrient cycling. Our results provide a starting point for understanding the effects of fire on herbaceous vegetation and post-fire forest succession in subalpine parkland. Table 4. Butte Creek (1994) structural transitional probabilities Field & office assistants: J. Andreychuk, S. Hiebert, E. Banks, E. Driskill, S. Erickson, E. Snoozy, E. Fales, K. Kingery, M. Tjoelker Funding provided by US Forest Service, Pacific Northwest Research Station, through a cooperative agreement with the University of Washington, School of Forest Resources, the Western Mountain Initiative, and Joint Fire Science Program Graduate Research Innovation Award Author affiliations We did not observe any difference in the frequency of occurrence for herbaceous perennials due to fire or structural class. 1 School of Forest Resources, University of Washington, Seattle, WA 98195-2100. 2 Pacific Wildland Fire Sciences Lab, USDA Forest Service, 400 N 34th Street, Suite 201, Seattle, WA 98103. Figure 5. Box plots of frequency of each functional group within each severity class For more information contact acansler@uw.edu Website: www.cfr.washington.edu/research.FME/people/alina.htm