A species-specific approach to predicting the timing of deciduous leaf emergence in the United States

Abstract

This research seeks to better understand deciduous leaf phenology – the timing of current biological events like leaf emergence and coloration – and its relationship with climate, through new seasonal cycle data from the USA National Phenology Network. Specifically, it looks at how temperature, date of snowmelt, and photoperiod relate to date of leaf emergence, and how differences in species characteristics explain both the different dates of seasonal processes and their relative sensitivities to climate. NPN data provides a unique opportunity to explore large-scale, species-level phenology patterns and to compare individual species’ responses. Present results show that models based on photoperiod, temperature, or snowmelt date alone cannot adequately explain spring phenology either on a regional or national scale. While a model based on snowmelt performs better than a temperature-dependent model in a single region, on the national scale the temperature-driven model exhibits the lowest errors. A qualitative comparison between species characteristics and model parameters supports earlier budburst by early successional species, species of the families Aceraceae and Rosaceae, and diffuse-porous species. Comparison with model errors shows larger errors for diffuse-porous and shallow-rooted species, suggesting potential water limitation as an additional driver of spring phenology. These results can help better inform projections of how deciduous tree phenology will respond to climate change, which in turn will affect numerous other fluxes, notably carbon, water, and heat.