Adding species terms weakened the flight-date link to butterfly index correlations at 38 English sites in 2005–2024

We tested whether butterfly species with similar adult flight dates had more correlated later annual abundance-index changes. We used the UKBMS 2024 release. Earlier dates came from 62,855 records in 1985–2004. The test used 14 species, 38 English standard transects, 261 complete population records, and 899 local species pairs in 2005–2024. With site terms and equal total weight per site, the slope was −0.221 per 30-day date gap on a Fisher correlation scale. Its conditional 95% range was −0.437 to −0.009. A global species-date shuffle gave a one-sided fraction of 0.0428. Adding species terms reduced the slope size by 68%, to −0.071, with a range of −0.338 to +0.158 and a shuffle fraction of 0.2469. The date gap then accounted for 0.17% of residual variation. Removing red admiral also weakened the slope. The adjusted checks therefore did not show a clear general timing relation beyond species average correlation levels. This does not prove no timing effect. The study uses selected long-record populations, earlier adult means, and estimated abundance indices. It does not test a weather cause. Ranges condition on training dates and source indices and use shared three-year and 100-km spatial blocks. The package has source archives, rights, provenance, code, data panels, all results, figures, and offline repeat instructions. Separate calculations matched the selection, dates, correlations, and slopes.

butterflies · phenology · population synchrony · UKBMS · species effects · reproducibility