Towards a new autumn phenology model integrating seasonal productivity, climate, and day length

Predicting the timing of autumn leaf senescence in northern trees remains challenging because the seasonal interplay in the effects of day length, climate, and plant productivity is not well understood. This severely limits our ability to forecast vegetation activity and carbon uptake in temperate and boreal ecosystems. Here we present a new framework for predicting autumn senescence dates based on the idea that day length mediates the effects of climate on autumn phenology, with early-season (pre-solstice) growth and late-season temperatures constituting antagonistic forces. To test these predictions, we used a combination of satellite-derived vegetation productivity across Northern Hemisphere forests, ground-sourced European phenology observations of four widespread tree species, and a climate-manipulation experiment on European beech. Our results reveal important constraints on the late-season carbon uptake potential of northern trees, improving our understanding of vegetation dynamics in response to climate change.