In order to estimate flowering dates of fruit trees by the developmental rate (
DVR) and the developmental index (
DVI), the relations between the
DVR and temperature were experimentally examined, and the
DVR from rest break to flowering was modeled.
Cultivars of Japanese pear (
Pyrus serotina Rehd.), ‘Housui’ and ‘Kousui’ planted on pots were exposed to eight constant temperatures (31°C, 26°C, 22°C, 18°C, 16°C, 13°C, 10°C and 6.8°C) from January 24, 1989. We assumed that
DVI of the initiation of temperature treatment was 0 and of flowering 1.0. The measured
DVR at each temperature can be calculated in the form of reciprocals of the number of days from the initiation of temperature treatment to the flowering. The relations between natural logarithms of the measured
DVR and reciprocals of the absolute temperatures became linear, but the parameters in the model changed around 20°C (Fig. 1).
The models and their parameters obtained from the experiment was validated to the data observed in the open field. The difference between the observed and the estimated days to flowering was less than 3 days. The estimated time course of
DVI was shown in Fig. 2.
To study the locational difference in the model parameters, the twigs of Japanese pear (‘Housui’ and ‘Kousui’) collected from 13 prefectures were exposed to several different constant temperatures. The
DVR in this experiment was determined from the date of germination to the flowering date. The result did not show a clear difference of the model parameters (Fig. 4, Table 3).
In order to compare our model with the model of effective accumulated temperature, the latter was applied to measured
DVR. MSR (mean square residual) of the former was less than of the latter.
抄録全体を表示