Abstract
Cherry blossoms, an icon of spring, are celebrated in many cultures of the temperate region. For its sensitivity to winter and early spring temperatures, the timing of cherry blossoms is an ideal indicator of the impacts of climate change on tree phenology. Here, we applied a process-based phenology model for temperate deciduous trees to predict peak bloom dates (PBD) of flowering cherry trees (Prunus×yedoensis ‘Yoshino’ and Prunus serrulata ‘Kwanzan’) in the Tidal Basin, Washington, DC and the surrounding Mid-Atlantic States in response to climate change. We parameterized the model with observed PBD data from 1991 to 2010. The calibrated model was tested against independent datasets of the past PBD data from 1951 to 1970 in the Tidal Basin and more recent PBD data from other locations (e.g., Seattle, WA). The model performance against these independent data was satisfactory (Yoshino: r2 = 0.57, RMSE = 6.6 days, bias = 0.9 days and Kwanzan: r2 = 0.76, RMSE = 5.5 days, bias = −2.0 days). We then applied the model to forecast future PBD for the region using downscaled climate projections based on IPCC's A1B and A2 emissions scenarios. Our results indicate that PBD at the Tidal Basin are likely to be accelerated by an average of five days by 2050 s and 10 days by 2080 s for these cultivars under a mid-range (A1B) emissions scenario projected by ECHAM5 general circulation model. The acceleration is likely to be much greater (13 days for 2050 s and 29 days for 2080s ) under a higher (A2) emissions scenario projected by CGCM2 general circulation model. Our results demonstrate the potential impacts of climate change on the timing of cherry blossoms and illustrate the utility of a simple process-based phenology model for developing adaptation strategies to climate change in horticulture, conservation planning, restoration and other related disciplines.
Highlights
Warming associated with climate change has been shown to alter ecosystem processes including phenology – the timing of organism development [1,2]
The two-step model used in this study was capable of explaining most of the variability in peak bloom dates (PBD) accurately for Yoshino (r2 = 0.78, mean absolute error (MAE) = 2.4 days, root mean square error (RMSE) = 2.7 days, bias = 0.9) and Kwanzan (r2 = 0.89, MAE = 2.1 days, RMSE = 2.0 days, bias = 21.2) cherry trees in the Tidal Basin over the period of 1991 through 2010 (Fig. 2)
Model testing Applying the new parameter estimates for Yoshino and Kwanzan cherry trees, we tested the model performance against temporally and spatially independent data sets that included PBD observations from 1951 to 1970 at the Tidal Basin, from 1994 to 2011 at the University of Washington campus in Seattle, WA, and from the Project BudBurst database recorded in 2008 at the four locations described in Methods section
Summary
Warming associated with climate change has been shown to alter ecosystem processes including phenology – the timing of organism development [1,2]. Considerable shifts in tree phenology have been reported in the temperate regions; these shifts are likely to be a response to the changing climate. Mynei et al [3] and Parmesan and Yohe [4] demonstrated that the growing season of trees has increased by 2.3 days in the past 40 years. Richardson et al [5] reported a similar increase of 2.1 days in temperate tree species. In Washington, DC area, 89 of 100 plant species surveyed, including flowering cherry trees, exhibited a significant advance of 4.5 days in first-flowering over the 30 years from 1970 to 1999 [6]. Shorter and warmer winters can reduce the cold hardening of trees, leaving them vulnerable to frost injury [9]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
