Field sampling is stratified on the basis of biogeoclimatic units and soil moisture and nutrient regimes. We try to sample five or more plots representing each of the possible combinations of moisture and nutrients. Intensity of sampling varies according to the areal extent of the ecosystem, its apparent diversity, and its importance for forestry or range management. Sampling intensity also depends on available access and the nature and scale of the project.

Sample plots are selectively located in areas that are as uniform as possible. Heterogeneous, transitional, or disturbed sites are avoided. Plots are located so as to represent particular combinations of moisture and nutrients. Slope position, indicator plant species, relative tree growth, soil texture, seepage, and base status of parent materials are used as clues to moisture and nutrient regimes. The professional judgement of experienced fieldworkers in selecting representative ecosystems is an important part of the approach.

Plot size is usually 400 m2 in forest stands, but reduced for grassland, wetland, or alpine sampling. Plot shape can vary but is usually square or rectangular. Field procedures of the Ministry of Forests’ classification program originally followed those described in the manual Describing Ecosystems in the Field (Walmsley et al. 1980; Luttmerding et al. 1990), which has now been succeeded by Describing Terrestrial Ecosystems in the Field.

The data collected from the sample plots are stored on computer and then sorted and organized using various tabulation and multivariate analyisis programs to develop a classification. The programs do not produce a classification, but greatly simplify the process of rearranging the data, during development. Vegetation and selected soil, physical, and mensurational data are used to define units in the heirarchy at each level.

The classifier initially groups plots by tentative biogeoclimatic unit and by estimated moisture/nutrient regime. Then, relying largely on personal knowledge and judgement, reorders or groups plots that are floristically and environmentally similar. Similar groupings from other biogeoclimatic units are compared and then fitted into the vegetation hierarchy. Through this process of computer-assisted, experimental grouping, rearrangement, and refining, the plant and site associations are defined. Site associations may then be subdivided into site series, types, or phases, often on the basis of edaphic factors as summarized by the environment tables.