Dreissena distribution in commercial waterways of the U.S.: using failed invasions to identify limiting factors

Abstract

Preventing the introduction of an invasive species and managing the ecological and economic effects of successful invasions often rely on accurate predictions of the potential distribution of the species in its new habitat. Instances of failed invasion can highlight factors other than dispersal that limit range expansion in an invasive species. The commercially available waterways (CAWs) of the U.S. are a network of interconnected rivers that support a regular traffic of commercial and maintenance vessels. As such, they are at extremely high risk of infestation from Dreissena, yet not all the rivers are infested. We used traffic, water chemistry, and impoundment characteristics of the rivers in this system to develop a set of logistic-regression models to predict the occurrence of Dreissena. Four single variable models correctly classified 20 out of 24 (83%) river systems. The best two-variable models had only one misclassified river, for an overall accuracy of 96%. Among rivers to which Dreissena has successfully dispersed, permanent populations are most likely to establish in systems that have moderate ionic strength and that also have some impoundment areas to increase long-term population persistence. We use this hierarchy of factors to extend predictions to four systems that currently lack the high dispersal intensity of the CAWs.