Combining fixed-location count data and movement data to estimate abundance of a lake sturgeon spawning run: a framework for riverine migratory species

Abstract

Estimating abundance of migrating fishes is challenging. While sonars can be deployed continuously, improper assumptions about unidirectional migration and complete spatial coverage can lead to inaccurate estimates. To address these challenges, we present a framework for combining fixed-location count data from a dual-frequency identification sonar (DIDSON) with movement data from acoustic telemetry to estimate spawning run abundance of lake sturgeon (Acipenser fulvescens). Acoustic telemetry data were used to estimate the probability of observing a lake sturgeon on the DIDSON and to determine the probability that a lake sturgeon passing the DIDSON site had passed the site previously during the season. Combining probabilities with DIDSON counts, using a Bayesian integrated model, we estimated the following abundances: 99 (42–215 credible interval, CI) in 2017, 131 (82–248 CI) in 2018, and 92 (47–184 CI) in 2019. Adding movement data generated better inferences on count data by incorporating fish behavior (e.g., multiple migrations in a single season) and its uncertainty into abundance estimates. This framework can be applied to count and movement data to estimate abundance of spawning runs of other migratory fishes in riverine systems.