Recreation, Fire and Disease Create a Mosaic of Threats for Golden Eagles in Southwestern Idaho

Anthropogenic stressors have resulted in ecosystem impoverishment and biodiversity loss worldwide. As the strength and reach of the human footprint increases, investigation of the additive or interactive effects of synergistic stressors on the landscape is imperative for conserving ecosystems and species within them. Apex predators can reflect how stressors impact ecosystems because of bottom-up effects. Golden eagles (Aquila chrysaetos) are apex predators of North American sagebrush-steppe ecosystems that are impacted by a suite of stressors, including wildfire, outdoor recreation disturbance, and habitat loss. We investigated whether multiple threats had additive or interactive effects on golden eagle occupancy, reproduction, and diet. We used a beforeafter-control-impact (BACI) design to study the effects of fire and recreation on eagle reproduction at 22 historical territories in southwestern Idaho. In 2015, the Soda wildfire burned 14 historical eagle territories, and 8 territories were unburned. We collected data on recreation and eagle territory occupancy, confirmed egg-laying rates, young fledged per egg-laying pair, and diet in 2017 and 2018 and compared these data to pre-fire levels of recreation and eagle reproduction in 2013 and 2014. Off-road vehicle (ORV) use, as well as total use, increased in unburned areas after the fire and remained the same in burned areas. ORV use was negatively associated with eagle territory occupancy, regardless of whether the time period was before or after the fire, or whether an area had burned. Conversely, early season pedestrian use decreased in burned areas after the fire and the effect of early season pedestrian use depended on fire. Before the Soda fire, pedestrian use was negatively associated with the rate of confirmed egg-laying. In burned territories after the Soda fire, pedestrian use decreased and the rate of confirmed egglaying increased, suggesting that the decrease in pedestrian use had an interactive, positive effect on eagle reproduction. Diet composition differed between burned and unburned territories, but overall diet diversity and prey delivery rates were similar across fire and recreation gradients. In burned areas, eagles brought less leporid prey (rabbits and hares) and more sciurid prey (ground squirrels and marmots) to nests than in unburned areas. Additionally, eagle diets included more leporid prey and less rock pigeons (Columba livia) in areas with higher recreation use. This result may indicate diet shifts in areas with less leporids to eating more rock pigeons, which are vectors for disease. Combined our results suggest that recreation is a significant threat to eagle occupancy and reproduction, even compared to large-scale wildfires that can have massive effects on shrub-steppe ecosystems. Together, these results reveal a mosaic of stressors that threaten eagles across the southwestern Idaho landscape. It is therefore imperative that we understand the additive or interactive effects of synergistic stressors acting on ecosystems so that we can best manage lands and conserve biodiversity in a time of rapid global change