Don’t move a mussel: The role of key environmental drivers and management scale in assessing spatial variation in dreissenid spread risk in the Missouri River basin

• Non-native dreissenids pose substantial ecological and economic threats. • Spatially-heterogeneous risk indices inform targeting of prevention resources. • The use of state-wide versus basin-wide comparison can change the relative risk outcome. • A web tool allows invasive species managers to tailor risk index to local priorities. The spread of non-native freshwater mussels in North America is a growing threat that has already resulted in substantial ecological and economic damage to infested areas. A primary vector by which invasive mussels spread is watercraft that are transported over land from an infested waterbody to an uninfested waterbody. Management efforts such as watercraft inspection and detection programs that attempt to intercept infected watercraft can help limit this type of spread, but optimizing the effectiveness of these programs under limited resources is complicated. Studies have looked at developing watercraft inspection and decontamination program strategies that focus on policy-specific objectives such as maximizing the number of infested watercraft interceptions. However, there is limited work that has considered the heterogeneous impact of protection efforts across different regions and waterbodies. Knowledge about this heterogeneity can highlight regions that would benefit the most from protection as well as regions that would require less effort to protect, e.g., areas with naturally unsuitable water quality for dreissenids. To this end, we construct a composite relative risk index (CRR) for watersheds within the Missouri River Basin, a region in the United States on the front line of dreissenid spread. The CRR uses a model that mirrors an expected value model but uses relative indexing as a proxy for the model components. The CRR incorporates a wide array of data sets to account for the direct and indirect damages from a potential infestation along with the risk of an infestation occurring. Our results suggest that the relative priority of a specific watershed—measured through CRR—can depend on whether we consider the entire Missouri River Basin or just the watersheds in the same state. This also indicates substantial state-level heterogeneity in the CRR. Another contribution is that the CRR index includes user-specified weights for certain parameters so that a user can adjust the relative importance of various factors to match their specific context. An accompanying web tool allows users to view the CRR results and adjust multiple parameters to see the resulting impacts on the CCR for watersheds in the Missouri River Basin.