Exploring the mechanisms of monodominance in species-rich tropical forests using process-based simulation models and percolation theory

热带雨林显示出极高的物种多样性，但有时包括仅由一种树种主导的大片区域。当超过60%的冠层树来自同一物种时，这种令人惊讶的现象被称为单优势。通常，单靠土壤因素不能解释这种单一优势地区的出现。已经描述了几种可以解释单显性的机制，例如大而强的种子的局部扩散，但到目前为止还没有全面的单显性理论，特别是到目前为止还没有机制模型可以让我们探索所有已知的单显性机制的相对重要性。我们将通过开发模块化的、空间上显式的模拟模型来缩小这一差距。与该模型相关的过程和结构将以文献为基础，与领域专家讨论，并以来自现有特征数据库的数据为基础，如Slot和Try。除了目前讨论的机制外，我们还将探索降水增加的作用，据报道，许多热带雨林都有这种情况，可能会导致越来越多的局部种子聚集传播，从而导致单一优势。模型验证将基于独立的预测，包括来自系统科学的既定原则。物种丰富的森林结构和单一优势森林结构的共存对应着一个所谓的临界点。在一个初步的模型研究中，我们能够证明这种共存只在很小范围的模型参数范围内是可能的。渗流理论将被用来探索这一参数范围，并推导出一般的、可测试的系统性质。例如，由于我们的模拟是在有限大小的网格上进行的，因此将使用统计力学中的有限大小缩放方法。该项目的申请者和联合PI已经建立了一个由从事单一支配地位、森林建模和基于模拟的理论物理工作的领先科学家组成的跨学科网络。该网络的成员将于2017年和2018年在莱比锡举行两次研讨会，综合他们关于单一优势的知识，并讨论适用于单一优势定量分析的经验和理论方法。系统论和生态学的结合将把我们的发现放在更大的背景下，从而有助于理解物种丰富和物种稀缺的森林是如何出现和共存的，以及人为因素如何影响森林结构。因此，这个项目将显著提高我们对生物多样性产生、稳定或侵蚀过程的理解。

Tropical rainforests show extreme high species diversity but sometimes include large areas dominated by just one tree species. This surprising phenomenon has been referred to as monodominance whenever more than 60% of all canopy trees are from one species. Usually edaphic factors alone cannot explain the emergence of such monodominant areas. Several mechanisms have been described which could explain monodominance, for example local dispersal of large and therefore strong seeds, but so far no comprehensive theory of monodominance exists.In particular, so far no mechanistic model exists that would allow us to explore the relative importance of all known monodominance mechanisms. We will close this gap by developing a modular, spatially explicit simulation model. Processes and structures relevant for the model will be based on literature, discussed with field experts, and underpinned with data from existing trait databases, such as sPlot and TRY. In addition to mechanisms discussed so far, we will also explore the role of increasing defaunation, which has been reported for many rainforests and could lead to increasingly local and thus clumped seed dispersal and, hence, monodominance.Model validation will be based on independent predictions, including established principles from systems science. The coexistence of species-rich and monodominant forest structures corresponds to a so-called critical point. In a pilot model study we were able to show that this coexistence is only possible for a narrow range of model parameters. Percolation theory will be used to explore this parameter range and deduce general, testable system properties. For example, since our simulations take place on grids of finite size, methods of finite size scaling from statistical mechanics will be used.The applicant and co-PIs of the project have already established an interdisciplinary network of leading scientists who are working on monodominance, forest modelling and simulation-based theoretical physics. Members of this network will meet in two workshops in Leipzig in 2017 and 2018, to synthesise their knowledge about monodominance and discuss suitable empirical and theoretical methods for quantitative analyses of monodominance.The combination of the systems theory and ecological perspective will put our findings in a larger context and thereby help to understand how both species-rich and species-poor forests emerge and coexist, and how anthropogenic factors affect forest structure. This project will thus significantly improve our understanding of processes that generate, stabilize, or erode biodiversity.