Variation in the date of budburst in <Emphasis Type="Italic">Quercus robur</Emphasis> and <Emphasis Type="Italic">Q. petraea</Emphasis> across a range of provenances grown in Southern England

The widely observed advance in spring budburst across a range of temperate forest species due to climatic warming has received considerable attention. This change in forest phenology has important implications for the choice of species and provenances currently being planted, which need to be suited to both current and future climatic conditions. Using a provenance trial in the south of England, this study assessed variation in the timing of budburst across 23 different European provenances of oak (Quercus robur L. and Q. petraea (Matt.) Liebl.) over 8 years of observations. The order in which the different provenances reached budburst was related to provenance source latitude: the southern provenances were always earlier than those from more northerly latitudes. The statistical technique partial least squares regression was used to identify critical periods of both chilling and warming. A General Linear Model and three-dimensional temperature response surfaces were used to analyse the temporal trends in budburst. There was a negative correlation between the date of budburst and mean daily air temperature in both the chilling and warming periods for all provenances, which was statistically significant for a majority. Spring warming had a larger effect on budburst than winter chilling with a mean spring temperature-driven advance of 3.61 days/°C (standard error = 0.17 days/°C) and mean winter period temperature-driven advance of 0.99 days/°C (standard error = 0.17 days/°C). Surprisingly, there was no statistically significant interaction between mean air temperatures during the chilling and warming phases on the date of budburst.