Can changes in autumn phenology facilitate earlier green-up date of northern vegetation?

Climate warming has induced substantial advances in the onset of vegetation green-up in the northern hemisphere during recent decades. To date, however, the temporal changes in green-up date have not been adequately explained by the statistical relationships between green-up date and climatic factors, posing challenges in the attribution and prediction of phenological change. In this study, we thus turned to focus on autumn phenology, a critical biotic factor that is likely to affect the subsequent spring phenology of vegetation. Using satellite-retrieved start and end of growing season (SOS and EOS) over the period from 1982 to 2015, we examined the association between the EOS and the SOS in the following year in northern middle and high latitudes (north of 25 degrees N). Interannual changes in SOS were significantly (P < 0.05) related to changes in EOS in the previous year in 26.4% of the total pixels, mostly in the boreal region, with a 1-day advance of EOS generally resulting in about a 0.5- to 1.0-day advance of the following SOS, suggesting that the advanced SOS may be associated with the advanced EOS. In temperate ecosystems, however, SOS showed a weak negative partial correlation with previous year's EOS (significant for 10.3% of the total pixels), suggesting that the delayed EOS may have limited contribution to the advanced SOS. Our analysis further revealed that changes in the EOS contributed little to the changes in the number of subsequent chilling days in temperate ecosystems and that the sum of forcing temperatures was weakly related with the number of the chilling days in the boreal region, suggesting that EOS may affect SOS through other mechanisms such as changes in the timing when the chilling requirement is ma as well as in carbohydrate and nutrient economy. This study suggested that the timing of EOS may explain some of the temporal changes in SOS in the following year in 36.7% of the study region, but further studies are needed to identify the exact mechanisms.