Ecological indicator values reveal missing predictors of species distributions

The questions of how much abiotic environment contributes to explain species distributions, and which abiotic factors are the most influential, are key when projecting species realized niches in space and time. Here, we show that answers to these questions can be obtained by using species' ecological indicator values (EIVs). By calculating community averages of plant EIVs (397 plant species and 3988 vegetation plots), we found that substituting mapped environmental predictors with site EIVs led to a doubling of explained variation (22.5% to 44%). EIVs representing light and soil showed the highest model improvement, while EIVs representing temperature did not explain additional variance, suggesting that current temperature maps are already fairly accurate. Therefore, although temperature is frequently reported as having a dominant effect on species distributions over other factors, our results suggest that this might primarily result from limitations in our capacity to map other key environmental factors, such as light and soil properties, over large areas.