Palaeoecologists use sequences of fossils within deposits from continents and oceans all over the world in order to produce time-series of past environmental dynamics over decades to millennia or longer. Such information can place current and future environmental change into context, for example by showing how climate, environments, ecosystems and humans interacted during past events, and by enabling verification of climate models through ‘hind-casting’ of such events. Through a meta-analysis and focused literature review of currently used statistical approaches in palaeoecological research, we highlight potential pitfalls and suggest ways forward to a fuller statistical understanding of the possibilities and limitations of palaeoecological studies. Statisticians or at least statistical reasoning should be involved in order to quantify uncertainties across the full analytical pipeline of obtaining, analysing and interpreting fossil time-series and could help optimizing the analytical decisions taken at all these steps.
|Journal||Journal of Agricultural, Biological, and Environmental Statistics|
|Publication status||Published - 03 Sep 2019|