Utility of Single-Channel Electronic Resistivity Counters for Monitoring River Herring Populations

Automated electronic counting systems are used to estimate river herring populations throughout New England. Measuring the accuracy of these counting systems is essential to understanding the limitations of counter-derived estimates, which comprise a vital component of river herring stock assessments. Our objective was to conduct a direct visual-electronic comparison of passage to evaluate the accuracy and biases of using single-channel electronic resistivity counting systems to monitor river herring. From 1999 to 2012, we conducted 1,189 direct comparisons. Of these, 555 occurred when river herring were actively migrating through the system. The counting system consistently underestimated passage, with 34-87% of seasonal observations recording fewer fish than were known to pass. Mixed-effects logistic regression indicated that the probability of an inexact count increased with increasing passage rate, which we attribute to simultaneous passage. We explored two techniques to correct for the biases observed during our sampling period and produce corrected estimates of run size. Although our results have limitations, the consistency of observed bias provides evidence that single-channel counters yield the resolution needed to infer population trend. Future monitoring should endeavor to correct the biases in single-channel counters and to explore alternative technologies that provide greater accuracy.