Infodoc : Sciences, Techniques et Culture

The aim of the GlobNET meta-analysis is to uncover soil forest trophic groups turnover spatial codependencies, we want to reveal how they respond to an environmental gradient or a disturbance at a biogeographic scale. In this context, I focus my analysis on a single dataset based on Norway and try to build an efficient approach to retrieve these ecological information. With the use of environmental DNA, we can build soil trophic groups and compute diversity metrics like species turnover. Thanks to non-paranormal transformation, we can use these trophic groups turnover as input in a Gaussian Graphical Model called graphical lasso. The resulting partial correlation net-work provides us ecological information on the spatial codependence of soil trophic groups. Using network metrics, we w[...]

The aim of the GlobNET meta-analysis is to uncover soil forest trophic groups turnover spatial codependencies, we want to reveal how they respond to an environmental gradient or a disturbance at a biogeographic scale. In this context, I focus my analysis on a single dataset based on Norway and try to build an efficient approach to retrieve these ecological information. With the use of environmental DNA, we can build soil trophic groups and compute diversity metrics like species turnover. Thanks to non-paranormal transformation, we can use these trophic groups turnover as input in a Gaussian Graphical Model called graphical lasso. The resulting partial correlation net-work provides us ecological information on the spatial codependence of soil trophic groups. Using network metrics, we will then be able to compare different networks and draw conclusions on how spatial codependencies vary in a biogeographic context.