Quantifying Cumulative Effects of Harvesting on Aspen Regeneration through Fuzzy Logic Suitability Mapping

Abstract

Vigorous aspen (Populus tremuloides Michx.) regeneration immediately following a harvesting event is important to ensuring the continued health and productivity of the future forest. This study aimed to examine the potential of using unoccupied aerial vehicle, multispectral remote sensing, and GIS mapping techniques to develop a comprehensive approach for predicting aspen regeneration success at the harvest block scale. Three winter harvested blocks were studied at Duck Mountain Provincial Park in east-central Saskatchewan, Canada. Ten regeneration predictor variables (number of skidder passes, percentage slash coverage, topographic wetness index, slope, aspect, slope position, and four vegetation indices: green normalized vegetation index [GNDVI], normalized red-edge index [NDRE], simple RED to NIR ratio [SR], and chlorophyll index green [CIG]) were determined for 168 measurement plots 1 yr after harvest. Principal component analysis, principal component regression, fuzzy logic analysis, and GIS mapping techniques, were combined for the first time in this study to determine cumulative effects on aspen regeneration. On average, low suitability areas had significantly more skidder traffic (34 passes) compared to below average (17), above average (10), and high (7) suitability areas. Low suitability areas also had significantly more slash coverage (13.1%) compared to below average (8.49%) or high suitability land (7.18%). High suitability areas had significantly higher GNDVI, NDRE, SR, and CIG indices, compared to low and below average suitability land. Not only does this method of analysis help to assess how a combination of factors may influence aspen regeneration, it can also act as a decision support system tool for industry, or government, to improve aspen regeneration assessments.