Ecosystem classification in the Central Rocky Mountains, Utah

Ecosystem classification in the Central Rocky Mountains, Utah

Introduction Societal demand for sustainable natural resources: Manage ecosystems rather than individual resources; ecosystem-specific approach Lack of comprehensive ecosystem classification for the Central Rocky Mountains (ID, UT, CO, NV); lack of ecological information Landscape division; ecologically specific criteria Landscape: complex, heterogeneous, hierarchical system Organize Classify

Introduction Land Classifications Vegetation: Climatic: Combined: Braun-Blanquet taxonomic classification, 1921 Habitat type classification, Daubenmire1952, 1968, Mauk & Henderson 1984, Mueggler 1988), e.g. Picea engelmannii/vaccinium scoparium h. t. Köppen classification modified by Trewartha, 1968 Map of Potential Natural Vegetation; Küchler, 1964 Ecoregions; Bailey, 1976 Different purposes - spatial scales

Introduction Land Classifications Idealized landscape Vegetation zones as elevation belts, La Sal Mountains, UT; Kusbach, 2009

Introduction Land Classifications Vegetation zones of the Rocky Mountains; modified Daubenmire, 1943; telescoping, interfingering, discontinuity inversion Cross-section: P, DF, S-F San Francisco Peaks, AZ; Merriam, 1890 Cross-section: S-F, DF, P, J-P

Introduction 1. Theoretical premise ECOSYSTEM CONCEPT : a landscape segment uniform in the five basic components: climate, soil, vegetation, animals, and microorganisms and relationships among them (Pojar et al., 1987) Macroclimate Topoclimate Biota Soil Site + vegetation = readily observable components, reasonable basis for land classification Ground water Bedrock Bailey, 1988 Comprehensive ecosystem (ecological) classification: environmental components + at least one biotic component (vegetation)

Introduction 2. Theoretical premise HIERARCHY THEORY + THEORY OF SCALE Organism S-F DF J Organ DF DF Focal stand level, Plant community Soil Cell DF DF Two dimensional (space and hierarchical organization) concept with three-level (triadic) approach

Objectives Objectives 1. Organize the landscape i.e., to build up a landscape hierarchy based on fundamental environmental gradients 3. Classify the landscape hierarchy

Study area 15,000 ha Rough, mountainous area in North Wasatch Range M331 D - Southern Rocky Mountains Steppe - Open Woodland - Coniferous Forest - Alpine Meadow, Overthrust Mountain Section; USDA, 2007 Franklin Basin Watershed Elevations:1600 3000 m Temp: 1 to 7 o C; SNOTEL Precip: up to 1250 mm T.W. Daniel Experimental Forest,

Data collection 163 plots sampled, summers 2006, 2007 Stratified fixed sampling design Vegetation data: Env. data: Morphometric variables : (e.g. elevation, slope, aspect, sl. position) Soil variables : (e.g. O horizon, parent material, texture, ph) Lab nutrient analysis: static nutrient availability index dynamic nutrient availability index, PRS probes Not focused on biotic interactions and historic events

Analysis 1. Principal Component Analysis (PCA) on original dataset (163 observations, 44 variables) Find levels of organization and principal components (PC) 2. PCA on reduced dataset (18 observations, 42 variables) Differentiate between the levels 3. Random Forest on reduced dataset (18 observations, 26 variables) Determine important variables 4. ANOVA Differentiate between ecosystem groups

Analysis Reduction of dataset based on ZONAL (CLIMATIC CLIMAX) CONCEPT : Stable late-seral plant communities with intermediate edaphic conditions best reflect the influence of regional climate. These communities grow on zonal sites. Bailey, 1988 Bailey, 1988 Stable late-seral plant communities : Engelmann spruce, subalpine fir, Douglas-fir, Rocky Mountain juniper, aspen. Intermediate edaphic conditions: mild, subdued terrain, loamy soils, coarse rock fragment content<35% vol., no growing season water table.

Results and Discussion PCA on entire complex dataset (163 observations, 44 variables) PCA on reduced dataset (18 observations, 43 variables) Microclimatic PCs 19, 18, 11, 7 % Mesoclimatic PC Elevation=surrogate 30 % or ecologically meaningful 58 %

Results and Discussion Landscape Organization Mesoclimate Climatic climax ecosystems Elevation L+1 Micro-climate Topo-climax ecosystems PC2 Topography/SMR, PC3 Microbial activity L Soil Nutrient Regime Edaphic-specific ecosystems PC1 Soil fertility, PC4 Soil development L-1

Results and Discussion Landscape Organization Meso-climate L+1 Micro-climate L Soil Nutrient Regime L-1 + Cleland et al. 1997, Winthers et al. 2005

Results and Discussion Random Forest PCA on reduced dataset (18 observations, 43 variables) Mesoclimatic PC Elevation=surrogate 30 % 58 % RandomForest suggests calcium, elevation, min. nitrogen, magnesium, and manganese (error rate = 11.1%)

Results and Discussion ANOVA, Tukey Grouping 5 classes 3 classes Subalpine Montane The same pattern for all important variables: Ca, Mg, N, K, Mn

Conclusions Next step: Based on our results: Confirmed environmental heterogeneity in the study area Organized the landscape; built up a landscape hierarchy based on ecologically significant environmental gradients Suggested vegetation climatic zonation (L+1 level) Ecosystem classification at (L and L-1 levels)

Thank you! Questions? Acknowledgments : Karel Klinka, professor emeritus, UBC, project supervisor Jim Long, Helga Van Miegroet, advisors Committee members: Leila Shultz, Janis Boettinger Western Ag Innovations, Saskatoon, SK, Canada USDA Forest Service, Intermountain Region All my helpers, especially diggers!