Using herbarium specimens to study evolutionary change related to climate warming - novel uses of natural history collections through specimen and label images

该项目的目的是利用植物标本馆标本记录过去130年来17种欧洲树种的种间和种内落叶时间，并推断其物候策略与气候相关的微观和宏观进化变化。在SPP 1991中，该项目属于研究重点3，通过分析标本和标签图像，包括计算机视觉/机器学习，新颖地使用自然历史收藏品。我们将使用来自焦点物种范围内的数千个标本的（图像）来量化(i)沿纬度和经度的种群变化，（ii）随时间的变化，以及（iii）已知系统发育关系的物种之间的变化。方案中提供了试点数据。由于在任何特定纬度，日长增长最快的时间发生在每年的同一时间，因此光周期敏感物种的空间和年际变化预计比光周期不敏感物种要小。然而，到目前为止，还没有研究对这两个假设进行测试，因为成年树的光线很难控制，而且气候室里的树苗不能使用，因为它们与成年树在不同的时间长叶。基于Unger， Merhof和Renner （BMC evolution）的工作。医学杂志。，我们也想应用他们的支持向量机代码，这是为13个相同的物种开发的。植物标本室研究落叶时间的方法可以为树木物候提供长时间序列，并允许回溯时间来推断树木对气候变暖的反应的幅度和方向。

The objective of this project is to use herbarium specimens to document inter- and intra specific leaf-out times of 17 European tree species over the past c. 130 years and to infer micro- and macro-evolutionary changes in their phenological strategies that relate to climate. Within the SPP 1991, this project falls into research focus 3, novel uses of natural history collections through analyses of specimen and label images, including computer vision/machine learning. We will use (images of) thousands of specimens from throughout the range of the focal species to quantify change in (i) populations along latitude and longitude, (ii) over time, and (iii) between species of known phylogenetic relationships. Pilot data are presented in the proposal. Because the fastest increase in day length at any particular latitude occurs at the same time each year, photoperiod-sensitive species are expected to show lower spatial and inter-annual variation than photoperiod-insensitive species. Up to now, no study has tested these twin hypotheses, however, because light is difficult to control for adult trees, and saplings in climate chambers cannot be used because they leaf-out at different times than adult trees. Building on work of Unger, Merhof, and Renner (BMC Evol. Biol., in press), we also want to apply their Support Vector Machine code, which was developed for 13 of the same species. The herbarium approach to studying leaf-out times can provide long time series for tree phenology and allows going back in time to infer the magnitude and direction of tree responses to climate warming.