Articles
Projecting almond bloom dates in California with the PhenoFlex framework
Article number
1406_64
Pages
455 – 464
Language
English
Abstract
Climate change is expected to impact fruit tree bloom dates, including delayed bloom or bloom failure, especially where warming winters reduce the accumulation of winter chill, which is needed to overcome dormancy.
Modeling fruit tree bloom is challenging, as the mechanisms of chill and heat accumulation are not well understood, and progress through the dormancy period is difficult to monitor.
We fitted the process-based phenology model PhenoFlex to long-term observations (1948-2008; 1984-2008, depending on cultivar and location) of the 10% bloom dates of three almond (Prunus dulcis) cultivars (‘Mission’, ‘Nonpareil’ and ‘Sonora’) in two locations (Chico and Modesto) in the Central Valley of California.
Using a weather generator and climate scenarios produced by seven climate models included in the CMIP6 model intercomparison project, we generated plausible weather scenarios for current conditions and for an ensemble of future climate scenarios, focusing on two points in time (2050, 2085) and four future greenhouse gas emission scenarios based on four shared socioeconomic pathways (SSP126, SSP245, SSP370, SSP585). PhenoFlex adequately captured the dynamics of 10% almond bloom, with a root mean square error of 2.7-4.4 days for calibration (34-49 observations) and 4.0-5.3 days for validation (15-22 observations). Comparison of historic and future simulated bloom dates showed that the median bloom dates of ‘Mission’ remained unchanged compared to the baseline, whereas ‘Sonora’ and ‘Nonpareil’ showed slightly later bloom dates (1-2 days) in the near future.
In the more distant future and under pessimistic warming scenarios, ‘Mission’ and ‘Sonora’ bloom was delayed by 6-14 days on average, with ‘Mission’ occasionally exhibiting strong delays by up to 30 days.
Given the need for pollinizer-cultivar overlap in most almond production, this work demonstrates that the cultivar-specific response of bloom to climate change could pose a risk to production that warrants strong consideration.
Modeling fruit tree bloom is challenging, as the mechanisms of chill and heat accumulation are not well understood, and progress through the dormancy period is difficult to monitor.
We fitted the process-based phenology model PhenoFlex to long-term observations (1948-2008; 1984-2008, depending on cultivar and location) of the 10% bloom dates of three almond (Prunus dulcis) cultivars (‘Mission’, ‘Nonpareil’ and ‘Sonora’) in two locations (Chico and Modesto) in the Central Valley of California.
Using a weather generator and climate scenarios produced by seven climate models included in the CMIP6 model intercomparison project, we generated plausible weather scenarios for current conditions and for an ensemble of future climate scenarios, focusing on two points in time (2050, 2085) and four future greenhouse gas emission scenarios based on four shared socioeconomic pathways (SSP126, SSP245, SSP370, SSP585). PhenoFlex adequately captured the dynamics of 10% almond bloom, with a root mean square error of 2.7-4.4 days for calibration (34-49 observations) and 4.0-5.3 days for validation (15-22 observations). Comparison of historic and future simulated bloom dates showed that the median bloom dates of ‘Mission’ remained unchanged compared to the baseline, whereas ‘Sonora’ and ‘Nonpareil’ showed slightly later bloom dates (1-2 days) in the near future.
In the more distant future and under pessimistic warming scenarios, ‘Mission’ and ‘Sonora’ bloom was delayed by 6-14 days on average, with ‘Mission’ occasionally exhibiting strong delays by up to 30 days.
Given the need for pollinizer-cultivar overlap in most almond production, this work demonstrates that the cultivar-specific response of bloom to climate change could pose a risk to production that warrants strong consideration.
Authors
L. Caspersen, K. Jarvis-Shean, E. Luedeling
Keywords
warm winters, flowering, chilling requirement
Groups involved
Online Articles (66)