Timing of cherry tree blooming: Contrasting effects of rising winter low temperatures and early spring temperatures

Abstract

Phenology reflects the interplay of climate and biological development. Early spring phenological phenomena are particularly important because the end of diapause or dormancy is related not only to heat accumulation in the early spring but also probably to winter low temperatures. Although a warmer winter can reduce overwintering mortality in many insects and plants, it also reduces the accumulation of chilling time that often triggers the end of diapause or dormancy. We examined a continuous 67-year time series of the first flowering date of cherry trees and compared three phenological models based on the temperature-dependent developmental rate: (i) the accumulated degree days (ADD) method, (ii) the number of days transferred to a standardized temperature (DTS) method, and (iii) the accumulated developmental progress (ADP) method. The ADP method performed the best but only slightly better than the DTS method. We further explained the residuals from the ADP method by an additive model using the mean winter minimum daily temperatures, the number of days with low temperatures (represented by daily minimum temperature) below a critical low temperature, and the minimum annual extreme temperature. These three temperature variables explained more than 57.5% deviance of the ADP model residuals. Increased mean winter low temperatures can delay the blooming of cherry trees by reducing the accumulation of chilling time, whereas reduced numbers of cold days can shift the blooming to become earlier. Overall, rising winter low temperatures will delay the flowering time, while rising early spring temperatures directly shift earlier the flowering time. The flowering time has been shifted to earlier, and the balance from the opposing effects of rising winter low temperatures and early spring temperatures explains this shift.