Elevated Rocky Mountain elk numbers prevent positive effects of fire on quaking aspen (Populus tremuloides) recruitment

• We study the roles of wildfire, elk and natural refugia on aspen regeneration.
• Elk reduced aspen regeneration by 95% in the presence, but not the absence of fire.
• At the landscape level, elk browsing negatively correlated with aspen height.
• Further, that relationship was stronger in areas burned by wildfire.
• We did not find evidence that natural refugia protect aspen from elk herbivory.

Quaking aspen (Populus tremuloides) is the most widespread tree species in North America and has supported a unique ecosystem for tens of thousands of years, yet is currently threatened by dramatic loss and possible local extinctions. While multiple factors such as climate change and fire suppression are thought to contribute to aspen’s decline, increased browsing by elk (Cervus elaphus), which have experienced dramatic population increases in the last ∼80 years, may severely inhibit aspen growth and regeneration. Fires are known to favor aspen recovery, but in the last several decades the spatial scale and intensity of wildfires has greatly increased, with poorly understood ramifications for aspen growth. Here, focusing on the 2000 Cerro Grande fire in central New Mexico – one of the earliest fires described as a “mega-fire” - we use three methods to examine the impact of elk browsing on aspen regeneration after a mega-fire. First, we use an exclosure experiment to show that aspen growing in the absence of elk were 3× taller than trees growing in the presence of elk. Further, aspen that were both protected from elk and experienced burning were 8.5× taller than unburned trees growing in the presence of elk, suggesting that the combination of release from herbivores and stimulation from fire creates the largest aspen growth rates. Second, using surveys at the landscape level, we found a correlation between elk browsing intensity and aspen height, such that where elk browsing was highest, aspen were shortest. This relationship between elk browsing intensity and aspen height was stronger in burned (r = −0.53) compared to unburned (r = −0.24) areas. Third, in conjunction with the landscape-level surveys, we identified possible natural refugia, microsites containing downed logs, shrubs etc. that may inhibit elk browsing by physically blocking aspen from elk or by impeding elk’s ability to move through the forest patch. We did not find any consistent patterns between refuge elements and aspen size or canopy cover suggesting that natural refugia are not aiding in aspen recruitment and that all young aspen were susceptible to browsing. In much of their normal range, aspen are not growing to large size classes, which threatens the future of this iconic species and calls into question the ability of ecosystems to recover from mega-fires. Our results highlight the importance of considering multiple interacting factors (i.e. fire and increased elk browsing) when considering aspen management and regeneration.