Patterns and drivers of biotic disturbance hotspots in western United States coniferous forests

Abstract

Globally, forest disturbances caused by herbivorous insects and plant pathogens (i.e. biotic disturbances) have increased since the 1990s, a trend linked in part to climate warming. With increases in biotic disturbance activity, an emerging ecological phenomenon has been documented: biotic disturbance ‘hotspots’, or areas where two or more biotic disturbance agents co-occur in space and time. Biotic disturbance hotspots may have important implications for forest resilience, particularly if they erode mechanisms of post-disturbance forest recovery. The factors leading to hotspot occurrence, however, remain poorly understood. We characterized the patterns and drivers of biotic disturbance hotspots occurring from 2000 to 2020 across three broad forested regions in the western United States (US; the Southern Rockies, Middle Rockies, and Cascades). Using Bayesian spatio-temporal models, we evaluated whether hotspots can be predicted from predisposing factors expected to increase forest susceptibility to biotic disturbance (i.e. forest composition, topography, and average climate), as well as inciting factors known to trigger individual bark beetle and pathogen outbreaks (i.e. annual weather). Biotic disturbance hotspots exhibited distinct spatio-temporal patterns and trends within each region. Forest structure and composition were the strongest and most consistent drivers of hotspots. Other factors varied in their importance by region, reflecting regional differences in biophysical context. Relative to the predictor variables included in our models, estimated spatio-temporal random effects were more closely correlated with model predictions, suggesting that dynamic factors such as outbreak spread strongly shape patterns of biotic disturbance hotspots. Our results illustrate the widespread nature of biotic disturbance hotspots across western US coniferous forests and demonstrate the importance of forest structure and regional outbreak dynamics in anticipating hotspots at regional scales. These findings provide a deeper understanding of interacting forest disturbances and have important implications for the resilience of forests during a period marked by continued increases in disturbance activity.