Count slopes were larger for late than early flight tails in four English butterfly samples in 1985-2024

We tested whether peak-count slopes differed between late and early scaled flight tails. We used 32,999 site-species-years for Meadow Brown and Ringlet in England and Wales in 1985-2024. Models controlled site and calendar year. All four English contrasts were positive, from 0.126 to 0.298 flight-date standard-deviation units per count doubling. Their intervals adjusted for eight tests were above zero. All four Welsh intervals included zero. A simulation gave unequal tail slopes when only sampled counts changed on a fixed skew flight curve. Thus unequal tail slopes do not by themselves identify a biological change. This is a direct contrast on records used in earlier campaign work, not independent confirmation. Missing visits, shared measurement errors, selected records and few Welsh sites limit inference. No causal climate or pollution effect is estimated. The package includes source extracts, exact raw-source hashes, code, all model results, simulations and an independent numerical check.

phenology · observation bias · butterflies · flight tails · sampling · count dependence · ecological timing