Henrik von Wehrden (Lüneburg): Statistical analyses are increasingly being used in current ecology and conservation. A high number of approaches are available today, making it difficult for scientists to identify the proper methods for their specific analysis. More importantly, while scientists are expected to keep track of numerous methodological developments, there is little consent on the ideal and best fitting analysis.
In my talk, I will give an introduction on which statistical approaches are most frequently used and under which circumstance these are appropriate, both from a statistical and an ecological perspective. In a second step, I will outline common problems and pitfalls in current statistical approaches. Using examples from the literature and my own research I will illustrate how the choice for statistical tests can often be biased and how this can affect important decisions in application. In a third step I will give a brief roadmap on the potential role of statistics in ecology and conservation in the next years. I will also outline how statistical tools can create benefit to foster research and knowledge, and how it can ultimately trigger interdisciplinary research.