First-pass catches gave greater mean taxon composition change than three-pass catches at 633 English river sites in 2000–2025

We tested whether lower catch size explains the effect of electrofishing pass number on recorded fish composition change. We used 1,734 repeat-survey pairs at 633 English river sites in 2000–2025. Pairs had the same method, similar survey seasons, and similar fished areas. Mean squared distance of taxon proportions was 0.1544 for first-pass catches and 0.1313 for pooled three-pass catches. The difference was 0.0232, or 17.6% of the three-pass mean. A nominal 95% range from resampling catchments and shared three-year blocks was 0.0057 to 0.0357. An exact finite-population calculation gave 0.1455 as the expected distance for random catches with the first-pass fish counts. Lower catch size therefore reproduced 61.4% of the mean excess. The remaining difference was 0.0089, with a range of -0.0051 to +0.0191. Bray–Curtis distance and all planned scope checks also had remaining-difference ranges that included zero. Catchment-only resampling gave a narrower positive range, so shared years matter to the inference. The result shows a survey-design effect on measured temporal change in selected catches. It does not establish random pass labels, equal capture rates, or change in the full river community. Three passes are a conditional reference, not a complete fish census. The package preserves source files, provenance, selection ledgers, exact expectations, simulation checks, resampling draws, figures, and offline code.

river fish · electrofishing · community composition · temporal change · sample size · finite population · biodiversity monitoring