Classification and comparison of upland hardwood and conifer ecosystems of the Cyrus H. McCormick Experimental Forest, upper Michigan

Publisher Website
Google Scholar
Abstract
An ecological method of forest site classification was used to classify and map local ecosystems of the Cyrus H. McCormick Experimental Forest near Marquette in the upper peninsula of Michigan, U.S.A. The field method of site classification was effective in distinguishing 11 ecosystem units that differed markedly in their vegetation as well as in their topographic and soil properties. Each ecosystem repeatedly occurred in a characteristic topographic position within the landscape. Ecosystems were identified in the field using combinations of biophysical properties such as slope, aspect, soil texture, soil drainage, and forest composition. Ecological species groups, groups of indicator plants, were especially helpful in distinguishing ecosystems in the field. Soils of ecosystems identified using the ecological method of classification differed significantly in their physical and chemical properties, and principal component analysis and numerical clustering confirmed that soils of the 11 upland ecosystems were different. Discriminant analyses combined with calculations of the probability of misclassification were used to compare the relative merit of physical factors (topographic and soil factors) and vegetation in classifying ecosystems. The simultaneous use of topographic, soil, and vegetal factors proved most efficient in classifying the upland ecosystems of the McCormick Experimental Forest.