| dc.description.abstract | The prediction of growth onset is particularly complicated, because this phenophase is the end-point of a series of poorly understood physiological changes that are not directly observable, occurring while the bud is dormant. However, it is becoming more and more important to understand how tree growth responds to changes in environmental conditions, and predict how species will be impacted by global warming. In order to gain more understanding of the environmental control of dormancy, ecophysiological studies are necessary. The objectives of this study were to (i) quantify the effects of temperature, photoperiod, light intensity and nutrient deficiency on the percentage and timing of budburst of Betula pubescens, Fagus sylvatica, Salix aurita, Salix x smithiana and Tilia cordata in a series of experiments, (ii) identify mathematical relationships describing the response of the timing of budburst to these environmental factors, (iii) use the information gained during the experimental study to develop a mechanistic model of budburst. The experiments highlighted the presence of a common mechanism controlling dormancy, based mostly on temperature and a photoperiod signal. The degree to which the two triggers contributed to the progress of dormancy varied among species. In F. sylvatica and
B. pubescens daylength could substitute for chilling completely or almost completely, while S. aurita, S. x smithiana and T. cordata had a stricter requirement for chilling. The phenological differences observed among species might be related to differences in their ecology and life strategy. Dominant, long lived species such as T. cordata and F. sylvatica showed low rates of budburst, high chilling requirements and responsiveness to light intensity, while opportunistic, pioneer species such as B. pubescens, S. aurita and S. x smithiana had high rates of budburst, low chilling requirements and were not affected by light intensity. In addition, B. pubescens and S. x smithiana were more responsive to high forcing temperatures than T. cordata and F. sylvatica. These results suggest that the timing of growth onset in B. pubescens and S. x smithiana is regulated through a less conservative mechanism than in T. cordata and F. sylvatica, and that these species trade a higher risk of frost damage for the opportunity of vigorous growth at the beginning of spring, when more light is available. B. pubescens was selected for a more com prehensive study that would facilitate the development of a physiologically based model of budburst. As this species was found to be particularly sensitive to photoperiod, an attempt was made at incorporating photoperiod into the Unified model (Chuine, 2000) and developing a model which could account for the experimental data and increase the understanding of the environmental control of dormancy. The effect of photoperiod was integrated through a photoperiod-dependent change in the rate of forcing. The model was applied to experimental and field observations and compared with the Unified model. Both models showed high internal validities when calibrated on the field observations, but only the UnPhot model showed high internal validity when applied to the experiments in which photoperiod was manipulated. This suggests that the integration of photoperiod enabled the model to explain a portion of variance unaccounted for by temperature based models. However, when calibrated on field observations and used to predict external field observations, both models showed a low external validity. It is concluded that the model needs to be calibrated on a large dataset containing both experimental and field observations before its framework can be validated (or falsified). | |
