INVASION OF RUSTY CRAYFISH, ORCONECTES RUSTICUS, IN THE UNITED STATES: NICHE SHIFTS AND POTENTIAL FUTURE DISTRIBUTION

Invasive species are among the foremost threats to freshwater ecosystems. Predicting their spread is important, especially if the species is associated with undesirable effects on recipient ecosystems. Ecological niche modeling allows for the assessment of invasion potential of non-native species. Here, we used native and invasive occurrence locations and the Maxent niche modeling algorithm to predict the invasion potential of rusty crayfish, Orconectes rusticus Girard, 1852, in the United States. We built and compared three models based on: 1) native occurrences only, 2) invasive occurrences only, and 3) all occurrence points. We found that the model using native occurrences did not accurately predict the current invasion of O. rusticus, as it omitted 58% of the known invaded locations. Furthermore, the model based on invasive occurrences failed to accurately predict the native range of O. rusticus. Predicted suitable areas for O. rusticus closely matched the known distribution when all occurrences were used to train the Maxent model. Differences in models based on native and invasive occurrence points are likely due to a niche shift of O. rusticus during invasion. We recommend the use of multiple sources of information to better understand the invasion potential of invasive species, as solely using native or invasive occurrence information ultimately may provide inaccurate predictions about the spread of invasive species.