Late flights of British multivoltine butterflies preceded 8% higher next-year site abundance in 1976-2024, and the Wall showed no lost-generation decrease

We tested the lost generation hypothesis with UK Butterfly Monitoring Scheme site data for 1976-2024. The hypothesis states that warmth causes a late extra generation that dies before winter, so the population decreases in the next year. A late flight is a record on or after 18 September at a site that had records of any butterfly on or after 25 September. The within-site model held the current-year site index constant and removed site and region-year effects. After a late flight, the next-year site index of 9 other multivoltine species was 8.2% higher (95% interval 6.3% to 10.1%). All 9 species had positive estimates. The result repeated in southern and northern England and in 1976-2001 and 2002-2023. In 5 univoltine control species, a late record cannot be an extra generation, and the estimate was +1.1% (-6.6% to +9.4%). For the Wall, the main lost-generation candidate, the estimate was +4.5% (-6.5% to +16.8%; 339 sites). Thus decreases larger than about 7% per late-flight year are outside the interval. Late Wall flights became more frequent only in southern England (+7.2 percentage points per decade at fixed abundance). Results did not change with other day thresholds, abundance controls or latitude terms. Limitations: the public data give only a last-record day per site-year, not counts per brood. Late records also mark good years and depend on detection. The Scotland estimate was negative but uncertain. The association is not a proven causal effect.

butterflies · voltinism · lost generation · developmental trap · phenology · climate change · UKBMS · Lasiommata megera · observation bias · population growth