Daylength predominates the bud growth initiation of winter deciduous forest trees in the monsoon region of China.

Climate warming has induced significant shifts in spring phenology over both temperate and boreal forests. The timing of bud growth resuming from dormancy is crucial for predicting spring phenology. However, the mechanisms by which environmental cues, other than chilling accumulation, initiate bud growth remains unclear. By constructing a revised process-based spring phenology model incorporating photoperiod and temperature triggers of bud growth, we simulated the first leaf unfolding and first flowering dates of four deciduous forest trees during 1981-2014 at 102 stations across China's monsoon regions. Then, we revealed spatial patterns of the two triggers. Moreover, we compared fitting precision and robustness of the revised model with three mainstream models. Results show that the revised models can effectively simulate all spring phenology time series. Growth initiation of foliar and floral buds was induced by photoperiod lengthening in 80.8% and 77.7% of time series, and by temperature increasing in remaining 19.2% and 22.3% of time series, respectively. The proportions of time series with photoperiod- and temperature-initiated bud growth significantly increase and decrease from northern to southern climatic zones, respectively. Chilling exposure controls the predominant bud growth triggers in different climate zones. Specifically, in regions with long and severe winters where chilling requirement is easily fulfilled, rising temperature in spring alleviates the cold constraint and initiate bud growth. Conversely, in regions with short and mild winters, prolonged daylength in spring compensates the lack of chilling exposure to initiate bud growth. These findings suggest that photoperiod may limit spring phenology response to temperature in low-latitudes. Overall, our model slightly outperforms other models in terms of efficiency, accuracy, and robustness in modeling leaf unfolding and flowering dates. Therefore, this study deepens our understanding of the mechanisms of spring phenology, and improves the predicting capability of spring phenology models in the face of ongoing global warming.