Profiling as method to identify relevant metabolites for phenological modelling purposes

点击翻译按钮获取中文摘要

After decades of active research on dormancy of woody plants and the subdivision of this phase into para-, endo- and ecodormancy, these phases are still not adequately considered in phenological models. Semi-mechanistic models are still used to describe plant development in which the model parameters are optimised on phenological observations. Although this allows to find a model adapted to the data, its physiological plausibility is not guaranteed. In a completed DFG-project, we were able to show that models for the flowering of sweet cherries, which were not tested for physiological plausibility, failed under warmer climatic conditions. On the basis of easily determinable parameters (fresh, dry matter, water content of the cherry buds) and in combination with classical approaches for the determination of endodormancy, we were able to determine for the first time the invisible phases para-, endo- and ecodormancy for the sweet cherry cultivar 'Summit' under field conditions. The knowledge of the length of these phases opens new possibilities for the validation of semi-mechanistic models and enables the development of new mechanistic approaches. In this regard, it was recently examined whether metabolites are related to the dormancy phases of woody plants. For sweet cherries, we were able to show that the courses of abscisic acid and sucrose are suitable markers to determine the end of endodormancy. However, these substances had significant annual variations in their concentration. In addition, phenological modelling requires multi-year data of physiologically relevant markers to account for the variability of environment in the models. The aim of this project is to systematically investigate representatives of different groups (e.g. carbohydrates, carotenoids, amino acids) in cherry buds with regard to their involvement in the dormancy phases. Untargeted metabolite profiling should be used to get an overview of which groups of metabolites show significant changes in content during dormancy. For physiological relevant metabolites, which show significant changes in the transition of the different phases, it is necessary to proof whether a relation to the environment (air temperature, global radiation, etc.) exists, a basic requirement for phenological modelling. For those biomarkers that show a physiological plausibility and environmental control, in the second step a targeted metabolite profiling must be done, in order to determine the metabolite contents in buds over 9 years. Data for the years 2012/13-2017/18 are used to generate a mechanistic 3-phase model that considers endo-, ecodormancy and ontogenetic development in a sequential order. For model validation, data for the years 2018/19-2020/21 will be used at Berlin for internal and multi-year data at the sites Dresden and Geisenheim for external model validation. Additionally, the transferability of the modelling approach for the varieties 'Regina' and 'Karina' will be examined.