Thermal niche predictors of alpine plant species

Within the context of species distribution models, scrutiny arises from the choice of meaningful environmental predictors. Thermal conditions are not the sole driver, but are the most widely acknowledged abiotic driver of plant life within alpine ecosystems. We linked long‐term measurements of direct, plant‐relevant, near‐surface temperatures to plant species frequency. Across 47 sites located along environmental gradients within the Scandinavian mountain chain, the thermal preferences of 26 focal species of vascular plants, lichens, and bryophytes were explored. Based on partial least‐squares regression, we applied a relative importance analysis to derive inductively the thermal variables that were best related to a species’ frequency. To discover potential seasonal variability of thermal controls, analyses were both differentiated according to meteorological season and integrated across the entire year. The pronounced interspecies and temporal variability of thermal constraints revealed the thermal niches were much more nuanced and variable than they have commonly been represented. This finding challenges us to present, interrogate, and interpret data representing these thermal niches, which seems to be required in order to move beyond purely probabilistic and correlative descriptions of species’ range limits. Thus, this information will help improve predictions of species distributions in complex arctic‐alpine landscapes. In a rapidly changing world, there is a critical need to predict the future geographical ranges of species (Anderson 2013, Yackulic et al. 2015). For this prediction, it is necessary to know the potential range limits of a given species and the factors determining those limits (Wellenreuther et al. 2012). Indeed, a profound understanding of the distribution boundaries of a species is pivotal in regard to projecting potential species distributions, especially under novel environmental conditions (K？rner and Hiltbrunner 2018). The use of species distribution models (SDMs) is a common tool to make inferences on species’ range limits and to project how the distribution of species might change (Morin and Lechowicz 2008, Wellenreuther et al. 2012, Yackulic et al. 2015). Ecological niche models are the cornerstones of such distributional modeling. These models are built upon information on the environmental features that define the current niche (i.e., the environmental requirements) of a species. Future distributions of those features, derived from climate change scenarios, are then used to predict where the species’ niche requirements