Fallacies and fantasies: the theoretical underpinnings of the Coexistence Approach for palaeoclimate reconstruction
- 1University of Vienna, Department of Palaeontology, Vienna, Austria
- 2Nelson Mandela Metropolitan University, Centre of Coastal Palaeoscience, Port Elizabeth, South Africa
Abstract. The Coexistence Approach has been used to infer palaeoclimates for many Eurasian fossil plant assemblages. However, the theory that underpins the method has never been examined in detail. Here we discuss acknowledged and implicit assumptions and assess the statistical nature and pseudo-logic of the method. We also compare the Coexistence Approach theory with the active field of species distribution modelling. We argue that the assumptions will inevitably be violated to some degree and that the method lacks any substantive means to identify or quantify these violations. The absence of a statistical framework makes the method highly vulnerable to the vagaries of statistical outliers and exotic elements. In addition, we find numerous logical inconsistencies, such as how climate shifts are quantified (the use of a "centre value" of a coexistence interval) and the ability to reconstruct "extinct" climates from modern plant distributions. Given the problems that have surfaced in species distribution modelling, accurate and precise quantitative reconstructions of palaeoclimates (or even climate shifts) using the nearest-living-relative principle and rectilinear niches (the basis of the method) will not be possible. The Coexistence Approach can be summarised as an exercise that shoehorns a plant fossil assemblage into coexistence and then assumes that this must be the climate. Given the theoretical issues and methodological issues highlighted elsewhere, we suggest that the method be discontinued and that all past reconstructions be disregarded and revisited using less fallacious methods. We outline six steps for (further) validation of available and future taxon-based methods and advocate developing (semi-quantitative) methods that prioritise robustness over precision.