"Water is our life": Assessing impacts of invasive Elodea spp. on aquatic resources and native livelihoods in Alaska

Alaska has one of the most rapidly changing climates anywhere on earth and is experiencing an accelerated rate of human disturbance. The combination of these factors increases the state’s vulnerability to biological... [ view full abstract ]

Alaska has one of the most rapidly changing climates anywhere on earth and is experiencing an accelerated rate of human disturbance. The combination of these factors increases the state’s vulnerability to biological invasion, including non-native aquatic plants, which can have acute negative impacts on ecosystem resilience and subsistence practices of resource-dependent communities. We assessed the threat posed by Elodea spp. (elodea) to aquatic resources and Native Alaskan subsistence livelihoods in the state. We created an elodea risk assessment using an ensemble of species distribution model (SDM) algorithms developed with current observed climate data at a 2km spatial resolution. Models were applied to future climate (2040-2059) using five general circulation models best suited for Alaska. Based on Native Alaskan and local land manager insight and concern, we focused our vulnerability assessment on Chinook salmon (Oncorhynchus tshawytscha) and whitefish species (Coregonus nelsonii) spawning and rearing habitat. Model evaluations indicated that our results had moderate to strong predictability, with Area Under the Receiver Operating Curve (AUC) values of 0.88 (generalized linear model), 0.96 (MaxEnt) and 0.88 (multivariate adaptive regression splines) and classification accuracies of 82%, 90% and 85%, respectively. Current and future ensemble results revealed different levels of relative invasion risk across the state, based on the interaction of dominant subsistence practices and elodea climate suitability. This risk includes current high risk in the Athabascan region of Interior Alaska and future high risk in the Yup’ik region of western Alaska by midcentury. Both regions to date have no recorded observations of elodea. Results of this study suggest such integrative modeling approaches can hold great utility for invasive species risk assessments, by facilitating the inclusion of resource-dependent community concerns in conservation planning and management efforts and leveraging an array of data sources to provide a robust targeted monitoring tool for invasion assessments across large landscapes.