Use of Plant Strategy Ordination,DCA and ANOVA to elucidate relationships among habitats of Salix planifolia and Salix monticola

Abstract. Detrended Correspondence Analysis (DCA) was used on a data set of 112 species from 14 stands dominated by Salix planifolia and/or S. monticola. Environmental variables were determined which might be important to explain differences in stand vegetation composition, and willow stature (height), in willow carrs in the Rocky Mountain National Park. Correlation of environmental variables with DCA stand scores indicate that peat depth and soil redox potential are highly related to the primary DCA axis. Soils in stands of S. planifolia are more reduced in the early summer and have greater peat depths, than soils in stands of S. monticola. The second DCA axis is related to a water chemistry gradient of hydrogen ion, sulphate, and ammonium concentration. Four environmental variables selected by correlation analysis were regressed on stand scores from the first two DCA axes. Regression coefficients of August soil redox potential and soil water pH were significantly non-zero on the primary DCA axis. In addition to overall vegetation relationships elucidated by DCA, the relationships between willow stature and environmental variables were tested by simultaneous ANOVAs. ANOVA results and DCA ordinations indicate that S. planifolia and S. monticola plants achieve the greatest stature in the least reduced soils, and in shallow peat. Competitive, stress tolerant, and ruderal plant strategy theory applied to the vegetation data produced an arrangement of stands highly correlated to the DCA Axis 1 stand arrangement. These correlations indicate that the primary axis of vegetation structure can be extracted either by species composition, or species morphology/physiology relationships. Concurrent use of DCA and plant strategy theory is suggested as a tool for the prediction of community composition.