prilagođeno pretraživanje po punom tekstu

ŠUMARSKI LIST 3-4/2018 str. 31     <-- 31 -->        PDF

in several species (Worral 1983, Billington and Pelham 1991, von Wuehlisch et al. 1995). The leaf unfolding dates of three species (birch, oak and beech) have been shown to advance significantly with warming in two experimental years (Fu et al. 2013). An earlier correlative study revealed that Q. Robur flushing in the UK had advanced and that temperatures in the February to April time window were strongly negatively correlated with the first leaf unfolding date (Sparks and Carey, 1995). Altitudinal transplant experiments of Q. petraea, a close relative of Q. robur, have revealed high levels of temperature-mediated plasticity. For example, a 1 °C rise in mean spring temperature caused leaf unfolding to advance by approximately six (Vitasse et al. 2010) to seven days (Phillimore et al. 2013).
A phenological model for Quercus petraea was formulated to predict the date of budburst and the increase in leaf area as a function of cumulative temperature and of day-length. Budburst occurs when the temperature sum of the past 10 days exceeds a given threshold, which is a decreasing function of day-length (Nizinski and Saugier 1988).
Accumulated GDD is often assumed to be effective for predicting leaf unfolding in trees in present climate conditions (Chuine et al. 2000, Chuine and Beaubien 2001). In this research on the influence of variables, such as precipitation, day of year and insolation, on GDD values, insolation is found to be the main factor that best determined the real value of GDD. Insolation predicted the GDD value with an accuracy of 74.1 to 89.1 % (Figure 2). Other parameters do not have as strong an impact on GDD, and only day of year variable contributed significantly, with the values of 1.5 % (for intermediate) and 7.9 % for the late flushing group. Therefore, separate prediction GDD models are used for each of the leaf unfolding groups. Further research should perhaps examine the possibility of using only the insolation as the main variable in the estimation of GDD. The leaf unfolding process is very sensitive to environmental drivers, including both temperature and photoperiod (Basler and Körner 2012, Heide 1993). Comparing the performance of two methods (time-window and GDD) in the prediction of the spatiotemporal variation in the timing of first flushing will provide congruent insights into the cues and processes that underpin geographic variation in the phenology of Quercus robur (Phillimore et al. 2013).
As a part of this investigation, we decided to apply the GDD method to facilitate future predictions for the leaf unfolding. Natural stands and clonal seed orchards are in constant environmental struggle with various pathogens that to some extent undermine their stability. Thus, the application of GDDs could in the future contribute to better and easier predictions of how phenoforms and pathogens are related, based on their biology and ecology. Late leaf flushing among Pedunculate Oaks serves as an antiherbivore defence, via the development of leaves that is asynchronous with the eclosion of caterpillars. This trait avoids the costs of defoliation, which may provide benefits later in the growing season (Hunter 1992, Wesolowski and Rowiński 2006). Large intraspecific variation in the timing of budburst of the Quercus robur may have a negative effect on herbivores, such as Operophtera brumata (Tikannen and Julkunen-Tiltto 2003). Late flushing forms are not harmed by late spring frost, which may result in lower damage and better quality trees. The main goal of the testing of the GDD model in this example of oak phenology is to determine the average values of the studied parameters associated with GDD that are required by each phenological group, specifically for the leaf unfolding trait.
The duration of the growing season may increase the vegetation period. Due to climate change, this duration is an important feature for carbon sequestration (Delpiere et al. 2009) and to the growth response (Kramer et al. 2000). Several studies have documented advances in the leaf unfolding and flowering dates of 2 to 3 days per decade on average during the last 64 years (Menzel 2000). Certain experimental results showed that future warming should continue to advance the tree species leaf unfolding dates, but other results also suggest that the phenological response of three European oak species to increased temperatures is nonlinear and can vary strongly among species (Morin et al. 2010).
A negative relationship is known to exist between required chilling and forcing, and Murray et al. (1989) reported that a negative exponential equation was the best fit across all fifteen investigated woody species. The findings of Fu et al. (2013) suggest that a constant forcing requirement could be used to predict the timing of leaf unfolding in oaks only when the chilling requirement is either far too low or fully reached. In between, a linear (negative) relationship between chilling and forcing needs to be applied. The relationship between GDD and chilling units (CU) was sigmoidal; GDD was not correlated with CU both at very low (< 40 CU) and very large (> 80 CU) chilling unit values. When chilling was less than 40 CU, the forcing requirement of oaks no longer increased and even appeared to become uncoupled from the accumulated chilling units. The breakdown of the negative linear relationship between forcing requirement and chilling at very low chilling units may be related to the influence of photoperiod. A long photoperiod may partially compensate for a lack of sufficient chilling (Caffarra et al. 2011, Garber 1983, Nienstaedi 1966, Wareing 1953). The values obtained by chilling could be used for the prediction of the required amount of forcing temperature and the occurrence of the leaf unfolding phase, which may indicate accurate protection from pests/diseases in clonal seed orchards or in natural stands of Pedunculate Oak.