Large aspen stands occur 2.42 times more often at the edge of a wildfire than inside the burn, slowing the fire and lowering its severity, even during Canada’s worst seasons on record. That is according to Flavie Pelletier, a recent PhD graduate in natural resource sciences at McGill University and lead author of a study in Forest Ecology and Management, co-authored with McGill’s Jeffrey A. Cardille and Joanne C. White of the Canadian Forest Service.
A single aspen burns fast on its thin bark, yet a large patch does the opposite, holding more moisture in leaf and trunk than spruce or pine and carrying none of the resin that helps conifers burn hot. Its canopy also starts higher up the stem, so a ground fire climbs the conifers around it first, and at scale, those traits make aspen a barrier to wildfire.
“The bigger the aspen patch, the lower in severity the fire tends to be,” Pelletier said when McGill released the findings on 7 July, with the pattern holding through the historic 2023 season, Canada’s worst on record for area burned. Researchers expected leafless spring stands to burn more vigorously, yet found the difference between leafed and leafless aspen mixed stands rather inconclusive.
Daily burned area fell from a median of 717 hectares to 646 hectares as aspen cover rose from under 10 per cent to more than a quarter of the ground. Aspen also burned at significantly lower severity than spruce in the two ecozones where the broadleaf is most abundant.
To reach that scale, the researchers built an algorithmic tool that turns satellite imagery from NASA’s Landsat and the European Space Agency’s Sentinel-2 into near-real-time maps of forest disturbance, then matched them with new Canadian Forest Service data on dominant tree species. The analysis covered four forested ecozones spanning 275 million hectares across three active fire years from 2021 to 2023, and Pelletier said the service may take up the tool to sharpen its coarse daily fire maps.
Commercial forestry works in the other direction, treating aspen as competing vegetation and controlling it each year to release the pine and spruce that mills want. Forestry companies across Ontario apply herbicide over thousands of hectares of Crown land to suppress broadleaf regrowth, most of it glyphosate, and Pelletier warned that reducing aspen, whether by brushing or herbicide, raises the fire risk those forests carry over the long term.
Not everyone reads it that way, with Fred Pinto, an adjunct forestry professor at the University of Toronto and a former executive director of the Ontario Professional Foresters Association, saying broadleaf trees return to a treated site within 5 to 10 years. Only a small share of Ontario’s boreal forest is harvested in any year, Pinto said, and many of the fires burning now sit in remote northwestern country never harvested nor treated.
It comes as fires burned through more than 3.9 million acres in the United States this season, with America’s top forester testifying before Congress on Wednesday over a crisis lawmakers linked to falling timber harvests on federal land. Ontario is meanwhile enduring the worst fire season in its history, with smoke from the province’s northwest reaching far south of the fire lines.
The result adds to a growing body of work on deciduous cover and fire, echoing a January finding that more birch and aspen means far less carbon lost when boreal forests burn. Species is not the only factor; a separate study last year found that private timber forests face a fire threat 1.5 times greater than public land, a gap put down to management.
Pelletier’s team measured 2.42 times more aspen along a fire’s edge than inside it, a margin that held through Canada’s harshest seasons and one the researchers say argues for keeping the broadleaf standing around towns and critical infrastructure.
For more information: Pelletier, F., Cardille, J. A., & White, J. C. (2026). Populus tremuloides as a natural fire barrier in Canada’s boreal forest under a changing climate. Forest Ecology and Management, 610, 123671. https://doi.org/10.1016/j.foreco.2026.123671
