Reconstruction of recent and Holocene fire chronologies and associated changes in forest composition: a basis for forest landscape management

Abstract

Simulations were carried out in the Fire Behavior Prediction (FBP) and the BEHAVE systems for spring and summer seasons to study the importance of vegetation and climate on fire behavior in the mixedwood boreal forest. Stand were characterized as deciduous, mixed- deciduous, mixed-coniferous, or coniferous stands, if their conifer basal area was respectively less than 25%, between 25 and 50%, between 50 and 75%, or greater than 75% of stand basal area. Sampled fuel loads (litter, duff, woody debris, herbs, and shrubs), and local weather conditions around Lake Duparquet (Quebec, Canada) were input in the simulation systems. The predicted fire behavior variables were Rate of Spread (ROS), Head Fire Intensity (HFI), and burned area. Results from Anovas on ranks showed that the two simulators attribute differently the maximum explained variance to the different factors. Climate is the most important factor for all fire behavior variables with the FBP whereas BEHAVE gives the greatest part of variance to the vegetation factor. Spring values are always higher than those from summer simulations. Deciduous stands record the lowest values. Mixed stands present intermediate values while conifer stands record the highest values for all variables. These differences related to stand composition exist within each fire risk (FWI), even for extreme weather risk. This implies that the widely accepted idea that during extreme FWI days all vegetation types would burn without difference is not applicable to southeastern boreal forest. Moreover, three prescribed burnings from Ontario, used to compare both systems, revealed that BEHAVE was not adapted to the mixedwood boreal region whereas FBP predictions were close to observed prescribed values. Climate and vegetation impacts on fire behavior and forest mosaic are discussed.