Prior summer rain gave no clear repeated link to paired butterfly flight spread in England and Wales in 1985-2024

We tested whether prior summer rain was linked to the next year's Ringlet minus Meadow Brown flight-date spread at the same survey sites. Four country and period models used 4,102 paired site-years at 163 sites in 1985-2024. Site effects removed stable site differences. Annual models controlled prior and current summer temperature, current summer rain, spring temperature and linear time. Slopes ranged from -3.21 to +0.43 days per 100 mm more prior summer rain. All four intervals corrected for four tests included zero. Early negative estimates did not establish a repeated response. Only three Wales sites passed the record rules, and its early slope changed greatly without the survey-boundary filter. Coarse weather, missing visits, selected detected populations and short annual series limit inference. This does not show no rain effect or identify a causal climate or pollution effect. Code, source hashes, aggregate estimates and figures are included. External records and derived databases remain outside the signed package.

phenology · butterflies · rainfall · delayed weather · paired sites · observation bias · null result