Methods for Estimating Adult Pacific Salmon Abundance and Uncertainty from Resistivity Counters

Accurate estimates of population size are fundamental to understanding population dynamics and essential for effective management and conservation. Fish counter technologies are increasingly being used to enumerate salmon populations, however, there are no standard methods or guidelines for validating fish counter data. Using resistivity fish counters as a model technology, we developed validation methods for incorporating the uncertainty in observation error in estimating adult salmon population sizes. Our framework incorporates the uncertainty in counter accuracy and species identification into abundance estimates using Monte Carlo simulations. We illustrate this framework using three case studies that vary in sensor type, river conditions, abundance, and species composition. When possible, we compared abundance estimates from the resistivity fish counter to abundance estimates from more conventional methods (i.e., helicopter overflights and fence counts) that were conducted concurrently. We also explore the trade-off between validation effort (i.e., cost) and the accuracy, precision, and bias in abundance estimates in a simulation study. We found diminishing returns of accuracy, precision and bias with increased validation effort. Our validation framework provides a standard methodology that incorporates uncertainty into abundance estimates from fish counter data, improving the quality of information communicated to managers.