Increasing overlap of chilling and forcing sensitivity periods in sweet cherry flowering under warming conditions

Many key phenological stages of temperate fruit trees are strongly controlled by environmental conditions. This includes the timing of dormancy release and flowering which are regulated by exposure to winter chilling and subsequent mild temperatures (forcing), processes that are essential to ensure high fruit yield and quality. Warmer winters due to global warming are associated with advanced flowering phenology and higher risks of frosts in the early spring, as well as delays and reductions in chill accumulation, which in turn can cause poor flowering and fruit set, with dramatic consequences for fruit production. In this context, a major challenge is to understand and predict the impacts of climate change on flowering, and ultimately, to breed fruit trees adapted to future climatic conditions.

In this study, we used flowering observations from a wide range of sweet cherry (Prunus avium L.) cultivars grown at the INRAE experimental station located in southwestern France, spanning the period from 1981 to 2024. Since cultivars were observed over different periods, resulting in a heterogeneous dataset, flowering dates were corrected for year effect using an ANOVA model to perform pairwise comparisons among cultivars and construct similarity clusters based on flowering precocity. We defined six classes from very early to very late flowering. Partial Least Squares Regression analyses were then used to identify the sensitivity periods when flowering dates are mostly regulated by chilling and forcing for the different flowering precocity classes through time (~4 decades). We aimed (i) to evaluate how the periods of sensitivity to forcing and chilling have shifted over the last decades, and (ii) to test whether distinguishable patterns could be observed between flowering precocity classes.

Our results suggest that, during the earliest years of observation (from 1981 to 2000), the periods of sensitivity to chilling and forcing did not overlap. However, over the last two decades, the period of sensitivity to chilling has been delayed, while the sensitivity period to forcing has advanced, resulting in an overlapping and more complex response to low and warm temperatures. This evolution could be attributed to longer periods of chill accumulation due to milder fall and winter temperatures whereas warmer winter and spring temperatures may lead to earlier response to forcing, with a more pronounced shift for early flowering cultivars. These results reveal that phenology models based on strictly sequential chilling and forcing phases may no longer be sufficient to predict flowering of cherry cultivars under warming conditions and models that account for overlapping sensitivity period should be preferred.