Characteristics associated with differences between undisturbed and industrially-disturbed soils

Abstract

Disturbed soils from south and central Alberta that had been affected by oil well drilling, surface mining of coal and oil sands, hydrocarbon spills and pipeline construction were compared with undisturbed soils from similar areas. Biological and conventional soil assay methods were studied to determine their usefulness in reclamation management. Biological measurements were related to nutrient cycling (alkaline phosphatase, arylsulfatase, protease, arginine deaminase, mineralizable N and nitrification potential), microbial activity (dehydrogenase, invertase, biomass C and basal respiration rate) and organic matter content (extractable organic C and extradable color index). Conventional measurements included pH, electrical conductivity, organic C, CEC, mineral N, SO 4 2−, P, saturation percentage and particle size distribution. Means of four characteristics were statistically different between disturbed and undisturbed soils. Electrical conductivity and SO 4 2− were higher in disturbed soils while dehydrogenase and arylsulfatase activities were lower. The best multivariate subset of variables was associated with a discriminant function weighted most heavily by electrical conductivity, extractable organic C, alkaline phosphatase, P, CEC, extractable color index and arginine deaminase. Of the undisturbed soils 86% and 70% of the disturbed soils were correctly classified using canonical variate scores calculated from this subset. Most errors in classification occurred in the Dark Brown and Brown Chernozemic soil sub-groups; no errors in classification were observed for the Luvisolic soils. These results suggest that using this combination of biological and conventional measurements may provide a practical degree of assurance when determining the extent to which industrially-disturbed sites may have responded to remediation.