Prior-minus-future summer temperature slopes were above zero after correction in three of six British butterfly tests, 1980-2023

We compared previous-summer and following-summer temperature slopes in abundance indices for nine butterfly species in England, Scotland and Wales. Six separate tests used 1980-2001 and 2002-2023. Models controlled current temperature, rainfall at all three times and quadratic time trends. The prior-minus-future slope was above zero after six-test correction in early Scotland, later Scotland and later England. The other three intervals included zero. Adding monitoring-coverage measures left only early Scotland above zero under the same correction. The result did not meet our rule for a common pattern across all six tests. Negative following-summer slopes in later samples warn against a simple causal reading. Countries share weather and methods. Short records, changing survey coverage, coarse exposure and source-index error limit inference. Future weather is an imperfect timing control. We did not test pollution, daily extremes or a causal climate-change effect. The package contains raw snapshots from a verified prior SCI publication, exact source hashes, code, all fits and offline checks.

butterfly · temperature · temporal specificity · negative control · observation bias · regional checks · abundance