Soil-site relationships for young white spruce plantations in north central Ontario / Richard R. LaValley
LaValley, Richard R.
DisciplineForestry and the Forest Environment
SubjectForest site quality Ontario
White spruce Ontario
Site index (Forestry) Ontario
MetadataShow full item record
Site index of white spruce (Picea glauca (Moench) Voss) in North Central Ontario was related to features of soil and topography using both multiple regression techniques and principal component analysis. Two different measures of site index, TOTSI25 (site index is total height at total age of 25 years) , and BHSI15 (breast height site index is total height at 15 years breast-height age) were used as dependent variables; 81 soil and topographic values were independent variables considered for analysis. Preliminary regressions computed from 54 plots indicated poor relationships between TOTSI25 and soil and topographic variables. Preliminary regressions also indicated that the correlations were much stronger using BHSI15. Correlations also were much stronger when the plots were stratified into three landform types as opposed to unstratified regressions. Three final regression equations were based on the relationship between BHSI15 and lacustrine, morainal, and glaciofluvial landform groups, and explained 77, 73, and 65 percent of the variation in BHSI15, respectively. The final regression equation for the lacustrine landform included the type of clay deposit (CLAY) , the depth to a root restricting layer (DRRL), and the hue of the C horizon (HUEC). The final regression equation for the morainal landform included the natural logarithm of the depth to a root restricting layer (LNDRRL), and the pH of the C horizon (PHC). The final regression equation for the glaciofluvial landform included the drainage class of the soil (DRAIN). The ability of the final regression equations to predict BHSI15 was tested on 14 independent test plots; these tests showed close agreement between actual site index based on stem analysis and site index predicted from the regression equations.