Main content area

A new process-based model for predicting autumn phenology: How is leaf senescence controlled by photoperiod and temperature coupling?

Lang, Weiguang, Chen, Xiaoqiu, Qian, Siwei, Liu, Guohua, Piao, Shilong
Agricultural and forest meteorology 2019 v.268 pp. 124-135
autumn, biogeography, carbon, carbon sequestration, climate change, color, ecosystems, herbaceous plants, indigenous species, leaves, models, phenology, photoperiod, prediction, temperature, time series analysis, trees, vegetation, winter, China
Autumn phenological variation of temperate and northern ecosystems plays a major but poorly defined role in the global carbon cycle. Constructing robust process-based autumn phenology models is key for improving current ecosystem models to accurately simulate ecosystem productivity and carbon sequestration. In this study, we developed a new process-based autumn phenology model. The model was fitted and validated using 67 leaf coloration and brown-down time series collected for 27 woody and herbaceous species at 18 stations on the Qinghai–Tibetan Plateau from 1981 to 2012. Then, we used the model to analyze the spatial and interspecific differentiation of driving factors of leaf senescence. Moreover, we compared fitting and validation precisions of the new model and previously published models. Results show that leaf senescence processes were triggered by photoperiod shortening in 61.2% of time series but by daily minimum temperature decrease in 38.8% of time series. Photoperiod control of the leaf senescence start occurs predominantly at stations with shorter annual maximum daylength (88.9%), while daily minimum temperature control of the leaf senescence start appears mainly at stations with longer annual maximum daylength (71.0%), especially for the native species. The new model reveals coupling effects of shortened photoperiod and decreased daily minimum temperature on leaf senescence processes. Comparing with the representative existing process-based autumn phenology models, the new model is a more general model and more robust in fitting and predicting leaf coloration and brown-down dates. Our study has provided a new insight on the understanding of leaf senescence mechanisms in winter deciduous trees and herbaceous plants, and significantly improved the ability to predict climate change impacts on vegetation growth and carbon balance in the sensitive biogeographical region of global climate change.