Considerations for using tag returns to monitor targeted removal of invasive fishes

ABSTRACT Objective Targeted removals are used for management of some invasive fish populations. Tag–return studies are one approach that can be used to assess the efficacy of targeted removals. However, there are many decisions to make when designing a tag–return study. We used simulation modeling to outline general guidelines for consideration when designing efficient tag–return studies to measure annual removal rates of invasive fish, particularly invasive carps. Methods We simulated data sets using scenarios with varying numbers of fish tagged per year, removal rates, tag reporting rates, tag retention rates, and study durations. We generated the data sets under a set of “known” parameters with added stochasticity; we then fitted the simulated data sets to a Bayesian tag–return model and measured the precision and accuracy of the model-estimated removal rates. Results We found that the model was able to predict removal rates without bias for most of the scenarios. However, we did find patterns in the precision of the predictions that could help to inform tag–return studies. When the proportion of the population removed through harvest was constant, the proportion of the population removed per year and the probability that harvested tags were reported had the largest effect on precision. The number of tags released per year and the study duration also had moderate effects. For scenarios testing the ability of the model to predict removal rates in stochastic populations, the precision of the model was primarily influenced by the number of fish tagged, the underlying nature of the stochasticity, and whether fish were tagged during the year of the prediction. Conclusions Based on our simulations, we outline how study objectives, the underlying population variability, and the tolerance range for error can guide decisions regarding the number of fish to tag, how to monitor tag return rates, and how long to conduct a study.