Abstract
Environmental heterogeneity influences ecosystem structure and function in part by governing the distribution and availability of forage resources for herbivores. Herbivory, in turn, directly affects a wide variety of processes, and large herbivores in particular often have an outsized influence on plant growth, productivity, and species composition. Accordingly, understanding the factors that govern plant-herbivore interactions is critical for optimizing efforts to manage large herbivores and their habitat. In the western United States, suppression of natural and anthropogenic disturbances such as fire and logging has led to the maturation of many forests into closed-canopy stands, reducing the availability of early-seral habitats critical for large herbivores like elk (Cervus canadensis). These changes have contributed to declines in elk populations in some regions and potentially to density-dependent limitations on forage production and regeneration in remaining early seral habitats. To address these challenges, we evaluated the effects of herbivory by elk on shrub growth and assessed the nutritional landscape available to elk in a managed forest in northern Idaho (the Grandad Elk Mitigation Area, GEMA). Using long-term elk exclosures (n = 5) and a before-after-control-impact experimental design, we examined how winter herbivory influenced shrub performance and whether shrub size influenced compensatory regrowth. We found that shrubs inside exclosures produced significantly more biomass and attained greater height and volume. Consistent with herbivore optimization theory, shrubs responded positively when browsing was reintroduced by producing longer shoots and concomitantly more biomass of currently annual growth, but only after surpassing minimum size thresholds (93 cm in height and/or 14,545 cm3 in volume) prior to being browsed. In parallel to our experimental work, we conducted intensive vegetation surveys and used generalized additive modeling to quantify forage production and identify key spatial drives of variation in forage availability for elk within the GEMA. Our dynamic foodscape model explained 54% (adjusted R2 = 0.54) of the variation in suitable forage biomass (i.e., biomass of forage that together met or exceeded nutritional quality thresholds for supporting lactation), identifying canopy cover, slope, heat index, and temperature as key predictors. Forest canopy cover in particular was strongly and negatively related to suitable forage biomass. Simulation analyses using our top model indicating that thinning or clearcutting could increase forage production by up to 150%, particularly in winter. Collectively, our results highlight the need for management strategies that expand early seral and shrubland habitats, such as prescribed fire, timber harvest, and temporary fencing to protect young shrubs. These strategies could enhance forage availability, reduce density dependent limitations, and improve habitat quality for large herbivores. Our findings provide valuable insights and tools for wildlife managers seeking to optimize forest management for large-herbivore populations.