Moose alter the rate but not the trajectory of forest canopy succession after low and high severity fire in Alaska

• In severely burned areas, moose stunted aspen apical growth for up to 10 years.
• 20 years after low severity fire, moose had no effect on aspen growth.
• Spruce growth showed no direct or indirect browsing effects.
• No change in the initial trajectory of forest canopy dominance.
• Heterogeneously burned forests could provide forage for up to 50 years.

Mammalian herbivory on palatable trees affects tree growth, forest composition, and forest succession. Antecedent effects of herbivores can be identified through remnants of dead stems and altered tree morphology as well as changes in tree ring patterns and growth. Increases in fire severity, particularly surface fuel combustion, in the boreal forest of western North America can cause a shift in the successional trajectory from coniferous to deciduous dominated forests, which may alter plant–animal interactions. We measured height and tree ring growth of the two dominant canopy tree species, trembling aspen (Populus tremuloides Michx.) and black spruce (Picea mariana (Mill.) BSP), in sites that experienced stand-replacing fire with deep versus shallow surface fuel combustion 20 years ago. We also classified individual trees into a category of browsing damage based on external features and morphology. We hypothesized the effects of browsing to be contingent on fire severity. Using linear mixed effect models, we investigated the main and interactive effects of fire severity and browsing intensity on aspen growth. We also developed tree ring chronologies to test for growth releases in aspen and black spruce.Effects of moose browsing on aspen growth depended on fire severity (surface fuel combustion), with negative effects in high severity sites and no effects in low severity sites. Spruce growth showed no direct or indirect browsing effects, indicating moose have not altered the potential for spruce to reach the forest canopy. Aspen in severely burned sites showed abrupt growth releases in tree rings corresponding to changes in herbivore pressure and density. Height-growth projections indicated that moose slowed the rate of aspen growth and canopy dominance in severely burned areas by ten years, through repeated stunting of apical growth in aspen, without affecting the initial trajectory to an aspen-dominated canopy. Lightly burned areas, with their larger proportion of spruce biomass, slower aspen growth, and reduced browsing pressure during the first 20 years after fire, will likely provide accessible aspen forage until >50 years post-fire as sites return to spruce dominance. Heterogeneously burned forests could thus sustain high rates of available moose forage for a much longer period than previously reported. Our study highlights the importance of including fire severity when considering the impacts of large herbivores on tree growth and forest structure.