An ecological approach to the investigation of a beech forest in Vestfold, SE Norway

The vegetation in a beech forest, Fritzøehusparken, is analysed using one hundred sample plots, each 25 m 2 , distributed by a restricted random procedure. Percentage cover and frequency in subplots are used for quantification of species amounts, and their relative performance are evaluated. Advantages of choosing frequency in subplots are discussed. Thirteen environmental variables are measured from each of the sample plots. The vegetation is divided into four topographical/ecological types. This classification served as a reference in the further treatment. Ecological gradients corresponding to the most important vegetational gradients in the beech forest are identified by means of DCA ordination followed by statistical interpretation. Two main complex gradients were recognized; (1) the gradient in nutrient conditions, and (2), the gradient inlitter‐wind conditions. The advantages of using DCA and statistical interpretation with integration of ecological measurements are emphasized.