Bayesian multistate models for measuring invasive carp movement and evaluating telemetry array performance

Understanding the movement patterns of an invasive species can be a powerful tool in designing effective management and control strategies. Here, we used a Bayesian multistate model to investigate the movement of two invasive carp species, silver carp (Hypophthalmichthys molitrix) and bighead carp (H. nobilis), using acoustic telemetry. The invaded portions of the Illinois and Des Plaines Rivers, USA, are a high priority management zone in the broader efforts to combat the spread of invasive carps from reaching the Laurentian Great Lakes. Our main objective was to characterize the rates of upstream and downstream movements by carps between river pools that are maintained by navigation lock and dam structures. However, we also aimed to evaluate the efficacy of the available telemetry infrastructure to monitor carp movements through this system. We found that, on a monthly basis, most individuals of both species remained within their current river pools: averaging 76.2% of silver carp and 75.5% of bighead carp. Conversely, a smaller proportion of silver carp, averaging 14.2%, and bighead carp, averaging 13.9%, moved to downstream river pools. Movements towards upstream pools were the least likely for both species, with silver carp at an average of 6.7% and bighead carp at 7.9%. The highest probabilities for upstream movements were for fish originating from the three most downstream river pools, where most of the population recruitment occurs. However, our evaluation of the telemetry array's effectiveness indicated low probability to detect fish in this portion of the river. We provide insights to enhance the placement and use of these monitoring tools, aiming to deepen our comprehension of these species' movement patterns in the system.