Leaf phenology in 22 North American tree species during the 21st century

Abstract Recent shifts in phenology are the best documented biological response to current anthropogenic climate change, yet remain poorly understood from a functional point of view. Prevailing analyses are phenomenological and approximate, only correlating temperature records to imprecise records of phenological events. To advance our understanding of phenological responses to climate change, we developed, calibrated, and validated process‐based models of leaf unfolding for 22 North American tree species. Using daily meteorological data predicted by two scenarios (A2: +3.2 °C and B2: +1 °C) from the HadCM3 GCM, we predicted and compared range‐wide shifts of leaf unfolding in the 20th and 21st centuries for each species. Model predictions suggest that climate change will affect leaf phenology in almost all species studied, with an average advancement during the 21st century of 5.0 days in the A2 scenario and 9.2 days in the B2 scenario. Our model also suggests that lack of sufficient chilling temperatures to break bud dormancy will decrease the rate of advancement in leaf unfolding date during the 21st century for many species. Some temperate species may even have years with abnormal budburst due to insufficient chilling. Species fell into two groups based on their sensitivity to climate change: (1) species that consistently had a greater advance in their leaf unfolding date with increasing latitude and (2) species in which the advance in leaf unfolding differed from the center to the northern vs. southern margins of their range. At the interspecific level, we predicted that early‐leafing species tended to show a greater advance in leaf unfolding date than late‐leafing species; and that species with larger ranges tend to show stronger phenological changes. These predicted changes in phenology have significant implications for the frost susceptibility of species, their interspecific relationships, and their distributional shifts.