Thermal time, chill days and prediction of budburst in Picea sitchensis.
Published and unpublished records of dates of budburst of lateral shoots on 2- to 10-yr-old trees collected on 14 occasions at sites near meteorological stations in Britain were used to produce an empirical model. Thermal time to date of budburst (day degrees more than 5°C accumulated from Feb. 1) declined exponentially with chill days (number of days ≤5°C from Nov. 1). This relation accounted for 92% of variation in thermal time. Eight-month-old seedlings were placed outside in Nov. 1979 on 2 sites in Scotland, transferred to constant environment cabinets (with day/night temp. 12°/5°C) in Jan., Feb. or March 1980, and budburst recorded. Results agreed with the linear portion of the model. A trial in which 8-yr-old potted trees, kept outside, were given a longer photoperiod from the end of Feb. 1980, using floodlamps, showed that this had no effect on budburst compared with control trees. Further aspects of the model discussed included: the small effect of provenance; linearity in the relation between bud growth rate and temp.; the omission of day length and soil temp.; and the choice of starting dates. The model was used to predict budburst at Eskdalemuir for 1912-82.