Landscape-scale influence of topography on organic layer accumulation in paludified boreal forests

Abstract

The aim of this study was to quantitatively investigate the relationship between topographic variables and organic layer thickness (OLT) and to use these relationships for mapping OLT distributions at the landscape scale within the paludified boreal forests of eastern Canada. Topography was quantified by a set of predictor variables (slope, elevation, aspect, mean curvature, plan curvature, and profile curvature) that were extracted from a LiDAR-derived digital terrain model (DTM) with four resolutions (1, 5,10, and 20 m). OLT was collected from field measurement (n = 1,600) across the landscape and varied from 5 to 150 cm. Weak correlations between OLT and individual topographic variables were obtained at the landscape scale. Stratification by aspect did not significantly improve these correlations. Consequently, regression tree analysis divided the data into six different landscape units, based on slope, aspect, and mean curvature. The resulting landscape units delimited the major patterns of OLT and elucidated three spatial relationships between OLT and topographic variables: greater OLTs (mean = 62 cm) were confined to gentle slopes (<1.8%), whereas lower OLTs (mean = 27 cm) were found in steeper slopes (slope >3.2%); OLTs were deeper on south- and west-facing than on north- and east-facing slopes; and the most accurate results were obtained by the LiDAR-derived DTM at 10- and 20-m resolutions. A thematic productive map of the distribution of the resulting six landscape units showed good matching (71%) with both vulnerable and promising areas for forest management. This study confirmed the fact that topographic variables influence OLT at the landscape scale, which had been previously reported at the plot scale within the Clay Belt. © 2014 Society of American Foresters.