Feeding the fire: Invasion in a forest-mosaic landscape

Rebecca Lemons, Claire Tortorelli, Alex W. Dye, Ty Nietupski and Meg A. Krawchuk, Forest Ecosystems and Society, Oregon State University, Corvallis, OR, Becky K. Kerns and John B. Kim, Pacific Northwest Research Station, USFS, Corvallis, OR, Kevin Vogler, Pyrologix, Missoula, MT, Karin Riley, Missoula Fire Sciences Lab, USFS, Missoula, MT, Nicole Vaillant, Rocky Mountain Research Station, USFS, Bend, OR

Background/Question/Methods
Invasive annual grasses are problematic where they increase fine fuels and facilitate larger and more frequent fires in previously fuel-limited desert, shrub-steppe, and woodland ecosystems. Closed canopy forests have remained relatively unimpacted by annual grass invasion, however, forests could become susceptible to invasion impacts if altered fire behavior in adjacent invaded ecosystems influences fire in nearby forests. In the Blue Mountains of eastern Oregon, forested ecosystems are highly heterogeneous, composed of a patchwork of forest and sparsely vegetated, fire-resistant “scab-flats”. Until recently, these forest mosaics have been relatively resistant to grass invasion and subsequent impacts. However, an invasive annual grass, Ventenata dubia, has been able to colonize and heavily invaded many of the previously sparsely vegetated scabland patches. Ventenata has increased fuel loading and continuity, potentially altering landscape-scale fire behavior. We examined how the V. dubia invasion into primarily non-forest patches impacted simulated fire behavior across a forest mosaic landscape in the Blue Mountains Ecoregion. We simulated fire behavior for invaded and uninvaded landscapes using the Large Fire Simulation System (FSim). We altered fuel models in the invaded fuelscape to represent V. dubia invasion using expert opinion and a map of V. dubia invasion throughout the ecoregion.
Results/Conclusions
Annual grass invasion altered simulated fire behavior throughout the ecoregion. Across the entire landscape, V. dubia invasion increased mean burn probability 9.4% overall and 25% in historically sparsely vegetated, fire-resistant scab-flats. Mean burn intensity increased 7.6% across the entire landscape and 12% in scab-flats with invasion. Within only invaded areas, mean burn probability increased 40% and mean burn intensity increased 133.6% compared to the same areas without invasion. Additionally, fires ignited in scab-flats transitioned into and burned 205 hectares more (350%) closed canopy forest under the invaded scenario than in the uninvaded scenario. Our study demonstrates how annual grass invasion can influence landscape-scale and forest fire behavior, despite primarily invading non-forested areas, and highlights invasion as an important management issue in a forest mosaic ecosystem.