Fertility grouping of Montreal Lowlands soil mapping units based on selected soil characteristics of the plow layer

• Published in: Canadian journal of soil science 1989-08, Vol.69 (3), p.525-541
• Main Authors: Nolin, M.C, Wang, C, Caillier, M.J
• Format: Article
• Language: English
• Subjects: lowlands; multivariate analysis; soil classification; soil fertility; soil properties; soil survey and mapping
• ISSN: 0008-4271; 1918-1841
• DOI: 10.4141/cjss89-054

Ten soil properties, including texture, pH, available phosphorous, cation exchange capacity, exchangeable cations and organic carbon, were determined from the Ap horizons of 91 map units of two Montreal Lowlands counties. Multivariate analysis techniques were used to classify the soils into fertility groups. Nine fertility groups were derived from cluster analysis. Principal component analysis yielded two components which accounted for more that 75% of the variance. These principal components were related to soil texture and fertilizer management factors. They confirmed the grouping obtained by cluster analysis. Multiple discriminant analysis distinguished these fertility groups and allocated new soil mapping units to one of them using discriminant functions. Clay content, pH, exchangeable Ca and available P were selected as the more potent and significant variables in descriminating groups. This work demonstrated that multivariate analysis has considerable promise in deriving meaningful interpretive grouping from selected soil parameters. Key words: Fertility, interpretive grouping, soil quality, multivariate analysis