Abstract
BACKGROUND: Freshwater ecosystems are some of the most affected by biological invasions due, in part, to the introduction of invasive carp worldwide. Where carp have become established, management programs often seek to limit further range expansion into new areas by reducing their movement through interconnected rivers and waterways. Lock and dams are important locations for non-physical deterrents, such as carbon dioxide (CO(2)), to reduce unwanted fish passage without disrupting human use. The purpose of this study was to evaluate the behavioral responses of common carp (Cyprinus carpio) to non-physical deterrents within a navigation structure on the Fox River, Wisconsin. Acoustic telemetry combined with hidden Markov models (HMMs) was used to analyze variation in carp responses to treatments. Outcomes may inform CO(2) effectiveness at preventing invasive carp movement through movement pinch-points. METHODS: Carbon dioxide (CO(2)) was recently registered as a pesticide in the United States for use as a deterrent to invasive carp movement. As a part of a multi-component study to test a large-scale CO(2) delivery system within a navigation lock, we characterized the influence of elevated CO(2) and forced water circulation in the lock chamber on carp movements and behavior. Through time-to-event analyses, we described the responses of acoustic-tagged carp to experimental treatments including (1) CO(2) injection in water with forced water circulation, (2) forced water circulation without CO(2) and (3) no forced water circulation or CO(2). We then used hidden Markov models (HMMs) to define fine-scale carp movement and evaluate the relationships between carp behavioral states and CO(2) concentration, forced water circulation, and temperature. RESULTS: Forced water circulation with and without CO(2) injection were effective at expelling carp from the lock chamber relative to null treatments where no stimulus was applied. A portion of carp exposed to forced water circulation with CO(2) transitioned from an exploratory to an encamped behavioral state with shorter step-lengths and a unimodal distribution in turning angles, resulting in some carp remaining in the lock chamber. Whereas carp exposed to forced water circulation only remained primarily in an exploratory behavioral state, resulting in all carp exiting the lock chamber. CONCLUSION: Our findings illustrate the potential of forced water circulation, alone, as a non-physical deterrent and the efficacy of CO(2) injection with forced water circulation in expelling carp from a navigation lock. Results demonstrate how acoustic telemetry and HMMs in an experimental context can describe fish behavior and inform management strategies.