Summer daytime cooling, not nighttime cooling, induces earlier bud set and leaf senescence in European beech

Forecasting autumn phenology remains challenging partly because many models rely on daily mean or daily minimum temperature (Tmin) as the primary thermal cue. However, a growing body of evidence suggests that daytime cooling, i.e. low values of daily maximum temperature (Tmax), can exert disproportionate control on end-of-season transitions, implying that models based on Tmin or daily means may miss key mechanisms. Here we ask: Does late-season nighttime cooling ever induce bud set and leaf senescence when daytime temperatures remain relatively warm? Using 330 European beech (Fagus sylvatica) saplings, we ran a controlled-environment manipulation experiment targeting autumn phenology. Across two treatment windows (Jul 3 – Aug 15 and Aug 15 – Sep 25), we imposed a nighttime cooling gradient (night temperatures of −1, 8, or 14 °C [control]) while holding daytime conditions constant. Additionally, we included an all-day cooling treatment (8 °C for 24 h) to contrast “night-only” versus “full-day” cooling. To test whether developmental state modulates responsiveness, plants were stratified into early- vs. late-leafing groups reflecting faster versus slower spring development. In the first window, no level of nighttime cooling alone induced earlier bud set or leaf senescence; in contrast, full-day cooling (and, by inference, reduced daytime temperatures) had a consistent advancing effect. Evidence from the second window indicates that nighttime cooling can contribute only at the very end of the growing season, when photoperiod is shortest and trees are most responsive. These results support the idea that daytime cooling (low Tmax) can be the decisive thermal signal for autumn transitions, and they motivate phenology models that explicitly separate day vs. night temperature effects rather than relying on Tmin or daily means alone.