High uncertainty in the effects of data characteristics on the performance of species distribution models

Species distribution models (SDM) are widely used as indicators of different aspects geographical ranges for many purposes, from conservation to biogeographical and evolutionary analyses. However, these techniques susceptible various sources uncertainty. Data coverage, species’ ecology, the characteristics their geographic distributions can affect SDM results, often generating critical errors in predicted maps. We assess influence data quality, species distributions, ecological traits on performance. predict dung beetle Madrid region (central Spain) using six validate them an independent dataset. relate variations model performance with environmental completeness, characteristics, through a partial least squares analysis. In this analysis, body size, nesting behaviour, marginality, rarity, prevalence, Relative Occurrence Area (ROA), range niche breadth, completeness predictors assessment metrics (sensitivity, specificity, kappa, TSS, CCR, AUC). Marginality prevalence were variables that most influenced performance, followed by ROA, breadth: presenting higher marginality smaller ROA breadth associated better models. Nesting size had minor importance Our results highlight taking into account when modelling comparing large groups SDM. This implies estimates richness composition based stacked SDMs show high levels error if they constructed diverse types distributions. suggest holding lead poor should not be included constructing composite biodiversity variables. Further effort is needed develop methodologies protocols such source