Soil fertility advances spring phenology of deciduous trees across temperate European forests

Phenology affects tree growth, as well as ecosystem dynamics such as the carbon, water and nutrient cycles. As phenology represents a plastic response of trees to environmental changes, leaf phenology of temperate trees has been intensively investigated in the last three decades in the context of global change. Accordingly, most research has been focused on the relationship between phenology and its environmental and climatic drivers such as air temperate, light, elevated CO₂, etc. Little attention has been given to the impact on phenology of non-climatic factors, in particularly the impact of soil nutrient availability. Here, we present a new analysis showing that soil fertility has a small but significant effect in advancing spring phenology of temperate deciduous forest trees. The analysis was based on data from monitoring programs (i.e. ICP forests and RENECOFOR) combining long-term phenological observations to soil physical and chemical properties for 121 European sites. First, we built meteorological models explaining a large portion (80-90%) of the inter-site budburst variability. Second, we related the residuals of the meteorological models to site fertility derived from a validated fertility index based on soil organic carbon, C:N ratio and pH. Third, we studied the effect of site fertility on chilling and forcing. Spring phenology was investigated using budburst date (50% of buds at budburst) but also the ancillary variables of budburst start (5% budburst) and budburst end (95% budburst). We found that more fertile sites showed an advanced spring phenology up to 3-4 days. This was most clear when sites were aggregated but it was also significant at species level (for the model species Fagus sylvatica, Quercus petraea, and Quercus robur) when considering budburst start and budburst end. Furthermore, we observed a significant interaction between fertility and chilling requirements indicating that trees at lower fertility sites show heightened avoidance of a premature budburst and require increased forcing. Our results suggest differences in the adaption of spring phenology to varying nutrient availability. Whilst small, the effect of fertility on budburst can be crucial in determining late frost risk. Furthermore, the effect of soil fertility on the interplay between chilling and forcing temperatures to initiate budburst suggests that its impact on spatial variability might increase under climate change. This warrants implementation in dynamic global vegetation models, and further encourages the community to study the impact of non-climatic drivers of forest phenology.