Abstract
Phenological shifts in alpine plants, which are sensitive to global warming, are one possible indicator for detecting the effects of warming at early stages. The responses of flowering time to cumulative ground temperature and long-term trends in flowering phenology were examined for Callianthemum miyabeanum in the dry alpine grasslands of Mt. Apoi. The date of first flowering and daily mean temperature at the ground surface were monitored in three populations from 2005 to 2008. To estimate the date of first flowering, the accuracy of a temperature-sum model was assessed by changing the threshold temperature and starting date for accumulation using 10 data sets. The starting date was assumed to be the day following a period of time during which the temperature consistently rose above the base point for bud break. The mean values of cumulative temperatures for flowering onset over the 10 data sets were calculated for every combination of the assumed threshold temperatures and starting dates, after which dates of first flowering were estimated. As a result of comparisons between estimated and observed dates of first flowering, a threshold base of 1℃ and a starting date of the day following that on which the temperature reached 6℃ were chosen as the optimum model. The dates of first flowering from previous years were estimated based on the optimum model using ground surface temperature estimated from the air temperature at Urakawa Town. Results indicated that the date of first flowering has advanced by 7.6 days during the last 100 years.
