Ongoing Spruce Beetle Outbreak in Southwest Colorado Increases Water Yields in Snow-Dominated Catchments

Abstract:

Over the last 20 years, increasingly large bark beetle outbreaks have resulted in widespread conifer mortality throughout North America. Most beetle impacted areas are located in snow-dominated headwater catchments, which generate a disproportionately large fraction of surface water resources. However, significant uncertainty remains in how beetle-killed forests alter water yields and downstream water availability. In recent years, a spruce beetle (Dendrochtonous ruffipenis) outbreak in Southwestern Colorado has affected over 1.8 million acres of spruce forest in headwater catchments of the Rio Grande and Colorado Rivers. Impacts on water yields from this outbreak have yet to be investigated.

This study addresses this uncertainty through a series of empirical analyses using publicly available streamflow and climate data from eight snow-dominated study basins. We employ three separate methods to assess whether relationships between runoff and hydrological forcings differ between periods before and after widespread beetle-kill. For all methods, we found consistent 15-47% increases in water yield, despite clear warming trends throughout the region. We also found decreased low flows in control catchments. We suggest that these results, along with higher snowmelt-driven peak flows, indicate that beetle-kill in Southwestern Colorado spruce forests increases streamflow by altering snowpack and associated runoff processes, rather than a change in evapotranspiration losses in the growing season. In the area affected by this outbreak, increased flows relating to beetle-kill may be buffering a systematic decrease in water yields caused by increased atmospheric water demand. Additionally, these findings contrast with nearby mountain pine beetle effects on streamflow, highlighting the dramatic heterogeneity of forest-streamflow interactions and the importance of snowpack and vapor loss processes to future water resources.