All four adjusted warmth intervals included zero for horse chestnut flowering progression in two Swiss networks in 1996-2023

We tested whether April-May warmth was linked to the time from flowering start to 50 percent flowering in horse chestnut. We used 504 paired records from 18 fixed Swiss stations in 1996-2023. Western and eastern station sets were tested in two 14-year periods. Annual models controlled March temperature and a linear time trend. All four warmth slopes were negative, from -0.19 to -0.99 days per C. All four intervals adjusted for four tests included zero. Some model variants gave clear negative intervals in later West or earlier East. One-day reporting-delay bounds permitted either sign in three main tests. Short series, unknown visits, selected stations and temperature after flowering limit the result. The tests do not establish a common temperature effect, no effect, total flower duration, or pollinator mismatch. The package contains paired evidence, code, source hashes, figures and numerical checks.

phenology · horse chestnut · flowering progression · temperature · observation bias · Switzerland