AUTHORS: Adam C. Jones, US Geological Survey; Jesse R. Fischer, US Geological Survey; Josey L. Ridgway, US Geological Survey.
ABSTRACT: There are over 40,000 dams in the Mississippi River Basin, and concentrations of invasive Silver Carp (Hypophthalmichthys molitrix) below them are well documented as populations attempt to migrate, reproduce, and establish in upstream ecosystems. Consumer-grade sonar and image analysis techniques provide an efficient and cost-effective approach to evaluate and monitor the density of invasive carp populations in these unique and intensely managed areas of interest. Current automated image analysis tools are reliable and used throughout the Mississippi River Basin to count, measure, and understand the behavior of invasive carp populations in a diversity of habitats. However, these tools are ineffective in quantifying targets within high density groups (i.e., schools), prompting the need for additional image analysis techniques and an improved understanding of the three-dimensional distributions of invasive carp individuals and concentrations. Overall, our objective is to evaluate fine-scale and reach-wide abundance dynamics through time (i.e., day, night, and season) and environmental conditions (i.e., water discharge, velocity, temperature). We conducted repeated sonar surveys using side-scan and down-imaging sonar in the Kentucky Lake and Lake Barkley tailwaters (Tennessee River and Cumberland River respectively) throughout 2023 and 2024. Four parallel transects were conducted for each 5-km reach downstream of each dam. Surveys began at the dams and moved downstream at ~8 km/h to ensure consistent images. Current data processing involves the manual identification of fish groups (i.e., area) within side-scan sonar images and the use of simultaneously recorded down-imaging sonar images to quantify depth distributions of high-density schools to estimate volumetric abundances of invasive carp. Manual processing will be used to train machine learning models similar to approaches being used to estimate individual fish targets. These results will inform future research objectives to improve survey design as well as management efforts, such as deterrence and removal actions in these critical bottleneck habitats.
