Individual‐based models forecast the spread and inform the management of an emerging riverine invader

Abstract

AbstractAimMounting ecological impacts of invasive species on freshwater ecosystems are among the greatest challenges confronting ecologists and decision‐makers in conserving biodiversity and ecosystem function. Tools to slow the proliferation of aquatic invasive species are still needed to guide the allocation of limited resources more effectively and efficiently once a species is already established. Here we develop mechanistic models to recreate the invasion history of the rusty crayfish Faxonius rusticus in the John Day River (JDR) basin, forecast its future distribution, and evaluate the management efficiency of, and trade‐offs among, population control actions.LocationJohn Day River Basin, Oregon.MethodsThe spread and control of rusty crayfish in the JDR was simulated with a spatially explicit individual‐based model (SEIBM) whereby the life history of each crayfish in the population is modelled in response to environmental conditions that vary across space and time. The model was calibrated by comparing modelled rusty crayfish spread throughout the JDR to known occurrences according to three comprehensive surveys.ResultsOur model accurately reproduced historical rusty crayfish distribution data for 2005, 2010, and 2016 with a specificity and sensitivity of ~80%. Leveraging this realistic model of the spread of rusty crayfish, we show that rapid management actions to the initial invasion would have resulted in an opportunity to slow the spread of rusty crayfish. We instead predict that rusty crayfish will reach the mainstem of the Columbia River by 2025, at which our model predicts that the crayfish population will number on the order of 108 individuals over more than 1100 km of river throughout the watershed and progress into critical salmon spawning habitat.Main conclusionThis study demonstrates that SEIBMs can provide unique insight into the future distribution of aquatic invasive species and concretely support decision‐makers in choosing an optimal control strategy.