ABSTRACT Objective Biological monitoring is a major component of management decisions and operating budgets of many natural resource management agencies. Given the scientific and financial commitments to monitoring, it is critical to estimate the ability to detect trends through time (i.e., power). Methods The Georgia Department of Natural Resources has monitored reservoir sport fish populations since the 1980s. We estimated the power to detect simulated long-term (≥10 years) changes in relative abundance (CPUE) for Largemouth Bass (some of which are a potential genetic admixture of the recently described species Micropterus nigricans now known as Largemouth Bass and M. salmoides now known as Florida Bass) and Black Crappie Pomoxis nigromaculatus sampled with electrofishing and gill nets, respectively, across multiple reservoirs (n = 21). Reservoir-specific simulations were parameterized using 13 years (∼2010–2022) of monitoring data. Power was calculated as the proportion of simulations (n = 1,000) resulting in significant (P ≤ 0.1) temporal trends across a 10-year period. We considered power ≥0.8 (i.e., 80% of simulations with significant trends) as the threshold for sufficient power across reservoirs. For both species, we estimated power under three scenarios: (1) declining CPUE, (2) increasing CPUE, and (3) reduction to biennial sampling effort with a 50% decline in CPUE. Results Most reservoirs had sufficient power to detect either a 50% decline or a 100% increase in CPUE of Largemouth Bass across a 10-year period. Switching from annual to biennial sampling for Largemouth Bass reduced the number of reservoirs with sufficient power by half. Power was generally lower for Black Crappie until larger declines (75%) or increases (400%) were imposed. Conclusions We found that Largemouth Bass monitoring was generally near or beyond our reference threshold, but post hoc correlation analyses suggested that the power of Black Crappie data could be increased with more within-reservoir station replication. Overall, using data simulation to estimate power proved a valuable tool in assessing the potential ability of common monitoring approaches to detect change.
Simon et al. (Thu,) studied this question.