Different numerical methods used to distinguish between organic soil types are evaluated. The research was initiated by the suggestion that acid leaching from mining wastes could be prevented by capping the tailings with a self-renewing methane-producing muskeg bog, in order to prevent the penetration of oxygen to the wastes. Thirty organic soils from bogs in the mining districts of Elliot Lake, Sudbury, and Timmins, Ontario, and Noranda, Quebec, were sampled and 28 soil characteristics were measured. These characteristics, whose values are normally or lognormally distributed, were analyzed by several different statistical methods. Some characteristics indicate the existence of two populations, and others are bivariantly correlated. Canonical discriminant analysis was more successful than cluster analysis in separating the bogs into well-defined geographical groups. However, principal component analysis proved best at grouping the organic soils according to their organic and inorganic components, and we suggest that this is a suitable method for the general discrimination of organic soil types. Methane was present in all the 17 bogs tested for it, and in two very wet bogs more than 2 mmol of methane per liter were extracted. Key words: Muskeg bog, organic soils, soil characterization, principal component analysis
Organic–Inorganic Hybrid Materials
