More and more, the wild bee species and their abundance are in dramatic decline due to the numerous anthropological activities, mainly because of local climate change caused by those activities. Understanding the behaviour of wild bees including the characteristics of their natural environment is one of the key points of a biodiversity study to propose a systematic way to protect future generations of wild bee species that take a crucial role in pollination services. In this study, the main research question looks for the answer of how the distribution of Earth surface patches affects the number of wild bee species and individual abundance. The study is taken on Biodiversity Exploratories in Germany, scattered in 3 different federal states. However, this study focuses on the Baden-Württemberg region only. Wild bee data is collected by a series of field work organized by the Plant-Insect-Interactions team in TUM School of Life sciences and, the Support Vector Machine algorithm is trained to perform classification to measure distribution of land surface patches with satellite images provided by Remote Sensing Database by the University of Marburg. Additionally, the ALKIS dataset, provided by “Landesamt für Geoinformation und Landentwicklung Baden-Württemberg“ was processed to create alternative classification result so that different results can compared to assess the quality of classification. After getting the required intermediate results, landscape heterogeneity index is calculated that represents how homogenous or heterogeneous the area is. Grassland experimental plot no.3 (AEG 03) is used as the test plot. Thereafter, the entire workflow is coded with python using ArcGIS’s python package (ArcPy), so that the code can be applied on rest of experimental plots in Alb region and can also be used by colleagues to perform the same study on other study regions (Thuringia, Brandenburg). Therefore, the python code is tested on every single plot in the Baden-Württemberg region. The code is following each step without an error and besides that, it extracts an online available dataset automatically by establishing connections to WCS and PostGIS server so that the end users will be assisted without any trouble of downloading the dataset. As a result of this study, plots with equally distributed land patches resulted in high heterogeneity index. When index values are correlated with data of wild bees, it’s discovered that the experimental plots in Alb region (amount of samples) were not enough to indicate that high index value areas contain both greater number of species and a greater number of individuals per species.     «

More and more, the wild bee species and their abundance are in dramatic decline due to the numerous anthropological activities, mainly because of local climate change caused by those activities. Understanding the behaviour of wild bees including the characteristics of their natural environment is one of the key points of a biodiversity study to propose a systematic way to protect future generations of wild bee species that take a crucial role in pollination services. In this study, the main rese...     »