Seasonal and long-term phytoplankton trait dynamics during trophic change and a regime shift in phytoplankton biomass

该项目将数据分析和基于数据的建模相结合，研究群落特征的季节性和长期变化对浮游植物生产力和优势食草动物表现的影响。利用康斯坦茨湖40多年来的数据，研究了浮游植物群落特征随季节和营养变化的变化。利用统计模型将群落特征和功能群特征与环境条件联系起来，分析其统计分布的时间变化。特别关注的是浮游植物生物量的制度转变，这是对康斯坦斯湖营养变化的响应（Jochimsen et al. 2013）。我们将探讨性状动力学在制度转变发生中的作用，并考察制度转变之前、期间和之后性状分布的时刻。在统计分析浮游植物群落特征变化的基础上，对浮游植物群落进行了数值模拟，比较了考虑具有时间常数特征的官能团的模型方法和根据数据确定具有时变特征的聚合官能团的模型方法。通过该模型，我们验证了功能群特征与环境条件之间的机制联系是否导致了观测到的生物量，并评估了具有时变特征的功能群聚集的潜力，以描述同一系统中不同营养状态下浮游植物群落动态的季节性发展。此外，这些模型是垂直分解的，从而提供了对社区特征垂直分布的见解。建模得益于康斯坦斯湖的优秀数据集，该数据集允许直接从数据中提取大多数强迫因子，从而减少状态变量的数量。通过数值实验研究不同营养状态下群落性状的变化对浮游植物和食草动物的影响，以及群落性状的变化是否可以解释康斯坦斯湖营养变化过程中的补偿动态。

The project combines data analysis and data based modeling to study the implications of seasonal and long-term changes in community traits for the productivity of phytoplankton and the performance of the dominant grazers. Based on over 40 years of data from Lake Constance, we investigate how the traits of the phytoplankton community change seasonally and with trophic change. We use statistical models to relate community traits and traits of functional groups to environmental conditions and analyze the temporal changes in the statistical distributions of the traits. Particular focus is on the regime shift in phytoplankton biomass that occurred in response to trophic change in Lake Constance (Jochimsen et al. 2013). We will investigate the role of trait dynamics for the occurrence of the regime shift, and examine moments of the trait distribution before, during and after the regime shift. The statistical analysis of the changes in the traits of the phytoplankton community is complemented by numerical modeling of the phytoplankton community comparing model approaches that consider functional groups with time constant traits and model approaches using aggregated functional groups with time varying traits determined from data. With the modeling we test whether the mechanistic connection between the traits of the functional groups and environmental conditions results in the observed biomasses and assess the potential of aggregation of functional groups with time varying traits to describe the seasonal development of phytoplankton community dynamics at different trophic states in the same system. Further, the models are vertically resolved and thus provide insights into the vertical distribution of the community traits. The modeling profits from the excellent data set from Lake Constance that allows taking most of the forcing factors directly from the data thus reducing the number of state variables. Numerical experiments will be performed to investigate how changes in community traits at different trophic states affect the performance of phytoplankton and grazers and whether the compensatory dynamics during trophic change in Lake Constance can be explained by changes in the community traits.