Bud traits capture inter-species differences in bud temperature dynamics

Temperature is considered the primary driver of spring leaf phenology in temperate trees, playing a dual role by regulating dormancy release through winter chilling exposure and promoting budburst via the accumulation of forcing temperatures in late winter and spring. Accurately capturing the temperature experienced by buds during dormancy is therefore essential for predicting budburst dates. As air temperatures increase with climate change, spring phenological events are occurring earlier in the Northern Hemisphere, yet this advancement is not well captured by land surface models for reasons that remain poorly understood. One likely source of error is the use of air temperature instead of bud temperature, as bud meristems can differ from ambient air temperatures by several degrees, depending on time of day, climate conditions (e.g. bright sky) and also species-specific bud traits. Sunlight, and especially photoperiod, is often proposed as a secondary driver of spring leaf phenology, but its exact role is still debated. In particular, photoperiod alone cannot account for the energetic effects of solar radiation on bud temperature or for light-dependent biochemical processes in bud tissues. In this study, we tested the hypothesis that bud traits partly explain species-specific differences between bud and air temperature and their responses to sunlight. We further examined whether accounting for bud temperature and traits improves predictions of individual bud temperature sums across bud development stages.

To address this, we designed a common-garden experiment combining direct bud temperature measurements using fine-wire thermocouples inserted inside buds, with measurements of key morphological, radiometric and physiological bud traits across 12 temperate tree species. Using statistical models, we were able to weigh the influence of bud traits on the temperature difference between buds and the surrounding air (∆T_bud-air) and its sensitivity to sunlight. Our study revealed a strong, yet species-specific, relationship between incident shortwave radiation (SW_in) and ∆T_bud-air. We then developed a single model showing how interspecific variation in  ∆T_bud-air responds to SW_in. For instance, during morning hours, 46% of the sensitivity of ∆T_bud-air to SW_in was explained by differences in bud diameter, density, the presence of internal bristles, and reflectance in the visible spectrum. During night, variation in the amplitude of ∆T_bud-air was explained for 33% by bud diameter, gravimetric water content and reflectance in the middle infrared.

By linking simple bud trait metrics to bud–air temperature differences, this study provides new insights into the interspecific sensitivity of bud temperature to microclimate, thanks to simple bud traits, and into the combined roles of air temperature and sunlight in regulating spring phenology of temperate trees.