Climatic factors controlling plant sensitivity to warming

Plant sensitivity to warming can be expressed as β or the number of days of advance in leafing or flowering events per 1 °C of Mean Annual Temperature (MAT) change. Many local studies demonstrate that β estimates for spring flowering species are usually larger than estimates for plants flowering in summer or fall. Until now, however, neither observational nor experimental estimates of this parameter were considered to be climate or geographically dependent. Here we question this paradigm through reanalysis of observational β estimates and mathematical modeling of the seasonal warming signal. Statistical analysis of a large number of bulk (averaged over species) estimates of β derived from the Pan European Phenology Data network (PEP725) revealed a positive spatial correlation with MAT, as well as a negative correlation with the Seasonal Temperature Range (STR). These spatial correlations of bulk β values as well as interseasonal variability in β were explained using a simple deterministic model of the Thermal Growing Season (TGS). More specifically, we found that the geographic distribution of bulk plant sensitivity to warming as well as the seasonal decline of β were controlled by the seasonal patterns in the warming signal and by average soil thermal properties. Thus, until recently, plants managed to keep pace with climate warming by shifting their leafing and flowering events by the same number of days as the length of the period of weather suitable for their growth. Our model predicts, however, an even greater increase in the TGS for subsequent increases in MAT. Depending on how they interact with other factors such as changes in precipitation and increased temperature variability, these longer thermal growing seasons may not be beneficial for plant growth.