OOS 48-4 - Using GEDI metrics to map keystone cavity excavator distributions and associated biodiversity hotspots in western US forests

Forest wildlife species respond to both horizontal and vertical heterogeneity in forest structure. Spatial habitat modeling efforts that incorporate vertical structure better reflect forest characteristics that can influence habitat quality and biodiversity patterns. Satellite-based sensors that capture patterns of forest structure at near-global extents, such as GEDI are likely to provide important vertical structural data at extents that are both difficult to obtain with airborne lidar yet relevant to landscape management for many animals and overall biodiversity. In this study, we used products created through the fusion of GEDI information with additional remote sensing data to explore how ecosystem structure influences animal species, using hotspots of biodiversity as a measure of habitat quality. Our focal species include keystone species (e.g., woodpeckers), whose excavation of tree cavities in western US forests supports high levels of diversity. The diversity of this group of keystone species has been suggested to be an indicator of overall bird diversity, while simultaneously supporting a wide variety of cavity-dependent mammal species. We used eBird data to conduct a random forest analysis of species-specific encounter rates for the cavity excavator guild (n=11) within Colorado, Idaho, Montana, Oregon, Washington, and Wyoming; developed predictive maps of species distributions; and identified hotspot areas where multiple woodpecker species occurred. In particular, we identified hotspots with a greater degree of diversity in body size of cavity excavators, as heterogeneity in cavity sizes is expected to support a greater diversity of other cavity-using species. Here we show that these species of cavity excavators respond strongly to vertical structure metrics characterized by GEDI, such as foliage height diversity, percentage of canopy cover, and canopy height. Model performance for individual species was relatively high, with AUCs between 0.76 and 0.86. We further present maps of hotspots for cavity excavators. In conclusion, the products from this analysis can be used by forest managers for species-specific efforts (e.g., Pileated woodpecker management) and may be of particular use in multi-species, multiple-use forest planning.