Modelling the causes and consequences of intraspecific plant chemodiversity in a changing world

植物化学多样性是由一系列复杂的分子、生态和进化过程决定的。第一个资助期的目标是开始为植物化学多样性的进化、出现和维持建立一个量化模型框架。与P5合作，我们回顾了现有模型的文献，这些模型解释了植物化学多样性是如何产生和维持的。虽然语言模型很多，但到目前为止，数学模型和计算机模拟的研究还很少。为了填补这一空白，我们开发了两种互补的种内植物化学多样性进化建模方法。首先，对于P3和P8，我们开发了一个简单的概念验证模型，其中代谢物的缺乏或存在直接在基因组中编码。我们表明，如果代谢产物存在的抗草食活性等位基因的遗传显性大于它们的成本遗传显性，那么植食压力的时间变化可以维持相当大的种内化学多样性，无论是在植物内部还是在植物之间。其次，我们开发了一个更机械化的基于个体的模拟模型，其中植物个体的代谢物分布来自其基因组、蛋白质组和反应网络的动力学的详细模型。我们再次发现，时间变化在选择压力中起到了很大作用，也有很大的历史偶然性。最后，利用P5、P6和P8，我们开发了虚拟Tanacum的第一个版本，这是一个模拟模型，综合了目前所有关于叶萜类化学多样性的遗传基础以及这种化学多样性与食草动物和传粉者相互作用的信息。第一个虚拟茄子(包括P3、P4)也将在第一个供资阶段准备好。在拟议的第二个资助期，我们的目标是扩展我们的化学多样性成因模型，并探索植物化学多样性在多变的生物和非生物世界中的后果。首先，我们将探索在不同的传粉者和食草动物环境中，化学多样性对植物种群反应的作用。其次，我们将研究气候变化下植物种群的种内化学多样性的作用(来自P10和COR的投入)。对于这两个工作包，我们假设存在投资组合或缓冲效应，即在不断变化的环境中，更多不同化学物质的种群波动不那么强烈，平均而言也可能更大。第三，我们将利用新产生的基因组数据、新的杂交试验和常见的花园数据，开发虚拟杨树，并扩展虚拟杨树和虚拟茄子。第四，我们将通过将短期实验结果输入我们的虚拟植物模型来为常见的化学多样性-可塑性实验(COR)做出贡献，以推导出在反复发生的干旱胁迫和食草性事件下化学多样性的影响的长期预测。

Plant chemodiversity is shaped by a complex set of molecular, ecological, and evolutionary processes. The goal within the first funding period was to start building a quantitative modelling framework for the evolutionary emergence and maintenance of plant chemodiversity. In collaboration with P5, we reviewed the literature for existing models that explain how plant chemodiversity arises and is maintained. While there are many verbal models, there are so far few mathematical models and computer simulations. To fill this gap, we developed two complementary modelling approaches for the evolution of intraspecific plant chemodiversity. First, with P3 and P8, we developed a simple proof-of-concept model where the absence or presence of metabolites is directly encoded in the genome. We showed that temporal variation in herbivore pressure can maintain substantial intraspecific chemodiversity, both within and between plants, if the genetic dominance of the metabolite presence alleles for anti-herbivore activity is larger than their genetic dominance for costs. Second, we developed a more mechanistic individual-based simulation model where a plant individual’s metabolite profile emerges from a detailed model of its genome, its proteome and the kinetics of the reaction network. We again found a large role for temporal variation in selection pressures, and also substantial historical contingency. Finally, with P5, P6 and P8, we developed a first version of the virtual Tanacetum, a simulation model synthesising all currently available information on the genetic basis of leaf terpenoid chemodiversity and on the interactions of this chemodiversity with herbivores and pollinators. A first virtual Solanum (with P3, P4) will likewise be prepared in the first funding phase. In the proposed second funding period, we aim to extend our models for the causes of chemodiversity and explore the consequences of plant chemodiversity in a variable biotic and abiotic world. First, we will explore the role of chemodiversity for the responses of plant populations in a variable pollinator and herbivore environment. Second, we will investigate the role of intraspecific chemodiversity for plant populations under climate change (with input from P10 and COR). For both work packages, we hypothesise that there is a portfolio or buffering effect whereby in a changing environment more chemodiverse populations fluctuate less strongly and are potentially also larger on average. Third (with P1-P8), we will develop the virtual Populus and extend the virtual Tanacetum and the virtual Solanum using newly generated data from the genomes, new crossing experiments and common garden data. Fourth, we will contribute to the common chemodiversity-plasticity experiment (COR) by feeding the short-term experimental results into our virtual plant models to derive long-term predictions for the effects of chemodiversity under recurrent episodes of drought stress and herbivory.