贝叶斯框架下的叶片光合模型-数据整合方法研究

Leaf photosynthesis is limited by stomatal conductance, photosynthetic metabolism and mesophyll conductance. Quantitative and mechanistic researches of mesophyll conductance went slowly. The main reason is that mesophyll conductance is affected by leaf anatomy and metabolism simultaneously. Leaf anatomy affects distribution of light absorption and CO2 inside the leaf, and therefore affects the photosynthetic metabolism inside. Classic C3 leaf photosynthesis model, Farquhar model assumes leaf is a uniform media, which is facing difficulties in applying to quantitative research about the relationship between leaf anatomical features and photosynthetic rate. This proposal learns from the frontier model-data integration methods in physics and geoscience, meanwhile it focuses on the key question of heterogeneous leaf photosynthesis. Based on a 3D model of leaf photosynthesis we developed, we will parameterize the model according to specific rice cultivars, and we will develop new methods to achieve a quantitative dissection of the anatomical effects on leaf photosynthesis of features such as leaf thickness and cell size. Moreover, this project will incorporate the Bayesian framework of uncertainty analysis, we will first estimate reliability of the results predicted above. This uncertainty analysis can also predict contribution of uncertainty due to imperfect model and experimental data, which can suggest directions of future research for a more reliable prediction. In all, the model, algorithm and framework developed in this project will provide a powerful tool for a systematic dissection of the key anatomical and biochemical factors controlling leaf photosynthesis, it will also promote related physiological and ecological studies.

气孔导度、光合代谢和叶肉导度是影响叶片光合效率的三大因素，其中叶肉导度的定量及机理研究相对进展缓慢，主要由于叶肉导度同时受到叶片结构及代谢的影响。叶片结构影响光线和CO2在叶片内部的分布，进而影响代谢过程。当前研究C3叶片光合生理所用的经典Farquhar模型假设叶片是均一介质，因而难以用于叶片结构变异和光合效率关系的定量研究。本项目借鉴物理学和地学中前沿的模型-数据整合方法，聚焦异质性叶片结构下的光合效率这一关键问题，基于我们已发展的叶片光合三维模型，对特定水稻品种系统参数化，发展全新算法，实现定量解析不同结构特征，如叶厚和细胞大小，对叶片光合效率的影响。同时结合贝叶斯概率的理论框架，定量分析这种不同因素影响光合预测的不确定性，揭示更可靠预测的模型发展方向和最佳测量参数选择。本项目发展的模型、算法和理论将为系统解析控制叶片光合的关键结构和代谢因素提供有力工具，促进相关生理生态研究。

叶片结构在光合作用中起着重要作用，然而对于任一结构参数和叶片光合速率的定量影响关系以及因果关系仍有待阐明。本项目从自底向上的机理模型和自顶向下的经验模型两个角度研究了这一问题。一方面我们开发了针对水稻的三维叶片光合模型，并使用了环境二氧化碳和升高的二氧化碳条件下生长的IR64水稻叶片的一系列成像数据和生理生化测量来参数化模型，然后询问该模型以获得各种结构因素和生化因素对于叶片光合的量化影响关系。我们发现虽然光合代谢是升高二氧化碳条件下叶片光合变化的主因，但各种结构因素也做出了或正或负的贡献。在模型中更大地变化几个重点关注的结构因素，我们发现叶片光合对于叶肉孔隙率和叶肉细胞凸起个数的变化不敏感，同时对于叶肉细胞接触空气面积的变化敏感，且在高二氧化碳下敏感性会降低。另一方面我们通过贝叶斯方法系统评估了Farquhar模型中光合参数估计的不确定性，并开发了一种更稳定可靠的数据收集和参数估计的方法流程。同时通过观察到的实际测量和预测的不符，指出实现更细化的光合特征估算的实验发展方向和模型发展方向。这些模型进展和分析结果对于理解叶片结构对光合的影响，以及改造叶片结构以提高作物光合作用等问题都具有重要意义。

内容获取失败，请点击重试