Complete 2005–2024 English butterfly indices did not show a clear rise in species correlation after spatial averaging

Do different butterfly species have more similar annual changes after changes are averaged across sites? We used complete UKBMS records for 16 species at 55 English transects in 2005–2024. Each of 71 species pairs used the same 10–19 sites at both scales. Mean correlation was 0.200 within sites and 0.279 between spatial means of standardized annual log index changes. The difference was +0.079. Its nominal 95% interval was −0.008 to +0.145 from shared three-year blocks. Thus, the main test did not show a clear increase. Different-site covariance terms supplied 85% of the mean spatial correlation, but their number is much larger than the number of same-site terms. A check with at least 15 sites gave a positive difference interval in a smaller pair set. A 2010-start check had too many constant-series resamples for a reliable interval. The result applies to fixed complete records and assumes short year blocks can represent annual variation. It does not measure a climate cause or extinction risk. The package has source data, rights, hashes, code, all planned checks, figures, independent calculations and repeat instructions.

butterflies · UKBMS · species synchrony · spatial aggregation · complete records · reproducibility