Windthrow modelling in old-growth and multi-layered boreal forests

Windthrow is a recurring disturbance process in regions that are influenced by maritime climates, including boreal forests. Within the circumboreal region, forest management approaches have been adapted to emulate natural disturbance in order to promote biodiversity and ecosystem resilience. Few studies have evaluated windthrow outcomes in natural or managed old-growth boreal mixed-species stands, and these studies are primarily empirical. The objective of this study was to adapt the hybrid empirical mechanistic ForestGALES_BC model, to investigate windthrow dynamics in natural and managed boreal old-growth stands, under various wind regimes. ForestGALES_BC was updated by adding biomechanical data for balsam fir (Abies balsamea (L) Mill.) and black spruce (Picea mariana (Mill.) B.S.P.). The model has the ability to simulate damage propagation during wind events by recalculating wind loading on a subject tree after failure of upwind trees. The number of iterations of this recalculation provides insights into the potential for propagation during longer duration storms. A simulation space made up of 500, 20 m x 20 m cells, was created using tree-lists from silviculture systems experiments in north eastern Quebec. The tree-lists for cells within the simulation space were edited to represent the plot conditions at the Quebec experimental sites, allowing simulation of a range of cell-(plot) and matrix-level (landscape) partial harvesting regimes. Above-canopy winds of various speeds were applied in order to test the initiation and propagation of damage within the simulated forest. Simulated outcomes were compared to observations of windthrow 6-7 years after partial harvesting at the field experimental site. Observed damage at the experimental site ranged from 1 to 40% of plot total basal area (mean = 14%, se = 1.96). For above-canopy wind speeds within the range expected for the experimental site, simulated windthrow levels in partial cuts ranged from 7 to 32% (mean =12%, se = 1.94). When the combined effects of wind speed and number of iterations (similar to event duration) were investigated, event duration was more important for wind speeds over 20 m/s. By enabling simulation of outcomes at the tree, plot (cell) and stand (simulation space) levels, the model allows investigation of a wide range of harvesting strategies, and sets the stage for inclusion of wind disturbance in ecosystem succession models and other stand or landscape-level decision-support tools for forests in windy climates. (C) 2016 Elsevier B.V. All rights reserved.