Aim: Species' biogeographical patterns are already being altered by climate change.Here, we provide predictions of the impacts of a changing climate on species' geographical ranges within high-latitude mountain flora on a sub-continental scale. We then examined the forecasted changes in relation to species' biogeographic histories.Location: Fennoscandia, Northern Europe (55-72°N). Methods:We examined the sensitivity of 164 high-latitude mountain species to changing climate by modelling their distributions in regard to climate, local topography and geology at a 1 km 2 resolution. Using an ensemble of six statistical modelling techniques and data on current and future (2070-2099) climate based on three Representative Concentration Pathways (RCPs 2.6, 4.5, 8.5), we developed projections of current and future ranges. Results:The average species richness of the mountain flora is predicted to decrease by 15%-47% per 1 km 2 cell, depending on the climate scenario considered. Arctic flora is projected to undergo severe range loss along with non-poleward range contractions, while alpine flora is forecasted to find suitable habitat in a warmer North.A substantial majority (71%-92%) of the studied species are projected to lose more than half of their present range by the year 2100. Species predicted to lose all suitable habitat had ranges centred in the northernmost (>68°N) part of continental Europe. Main conclusions:Climate change is predicted to substantially diminish the extent and richness of Europe's high-latitude mountain flora. Interestingly, species' biogeographic histories affect their vulnerability to climate change. The vulnerability of true Arctic and endemic species marks them as highly important for conservation decisions. K E Y W O R D SAlpine, Arctic, biogeographic history, climate change, range contraction, species distribution models

This study aims to determine the effects of environmental factors on the distribution and species richness of refugia for arctic-alpine vegetation. We will assess the main drivers for the arctic-alpine refugia in our study areas in N Europe, defined as isolated pockets with multiple species occurrences outside their main distribution area, and how well they can be modelled. The study is based on a comprehensive vascular plant distribution data set combined with abiotic environmental data at a resolution of 1 km². Cross-validated Boosted Regression Tree (BRT) modelling was employed to examine the effects of the climatic, topographic and geologic variables on refugia distribution and refugia species richness. Model testing was performed incrementally, i.e. first climate alone, then with additions of topography or geology, and concluding with a model including all predictors. All refugia distribution models (climate-only and different predictor combinations) performed well with mean area under curve (AUC) values higher than 0.85 and true skill statistics (TSS) values higher than 0.57. The inclusion of topography significantly improved model performance for both refugia distribution and refugia species richness. Climate has a central role in controlling the occurrence of refugia. However, topographic variables aid in recognizing the locally heterogeneous environments that sustain refugia. Refugia are thus driven by joint impacts of climatic and topographic factors that determine local thermal and moisture conditions. Our study demonstrates that the spatial patterns of refugia can be successfully modelled but emphasizes a need for high-quality data sampled at resolutions reflecting significant environmental gradients.

Determining the drivers, patterns and hotspots of biodiversity can be of critical importance in supporting regional conservation planning. However, as biodiversity hotspots can be described with several different metrics, it is important to investigate their congruence as well as the spatial overlap of hotspots with protected areas. Here, by using extensive data on climate, topography, soil characteristics and vascular plants combined with boosted regression tree models, we determine the patterns and key drivers of plant diversity hotspots along broad environmental gradients in northernmost Europe spanning from taiga landscapes to treeless tundra. We assess plant diversity with four metrics-species richness, range-rarity richness, threatened species richness, and local contribution to beta diversity-and examine their congruence with each other as well as with contemporary conservation areas. We found that climate plays an important role in governing species diversity, though topoedaphic are highlighted alongside climatic predictors in determining the diversity patterns of many threatened, near-threatened, and range-restricted species. Importantly, the different diversity metrics have contrasting drivers and, overall, their hotspots have low congruence. Furthermore, existing protected areas appear to offer limited coverage for hotspots of vascular plant diversity. Modelling the various facets of diversity and their drivers, such as the topo-edaphic setting, may help conserve diversity in a changing climate. Projected patterns of different aspects of diversity and their congruency can provide insights for the processes underlying biodiversity and be employed to assess the representativeness of protected area networks.

Refugia, sites preserving conditions reminiscent of suitable climates, are projected to be crucial for species in a changing climate, particularly at high latitudes. However, the knowledge of current locations of high-latitude refugia and particularly their ability to retain suitable conditions under future climatic changes is limited. Occurrences of refugia have previously been mainly assessed and modelled based solely on climatic features, with insufficient attention being paid to potentially important landscape-scale factors. Here, climate-only models and 'full' models incorporating topo-edaphic landscape-scale variables (radiation, soil moisture and calcareousness) were developed and compared for 111 arcticalpine plant species in Northern Fennoscandia. This was done for both current and future climates to determine cells with resilient climatic suitability harbouring refugia. Our results show that topographic and edaphic landscape-scale predictors both significantly improve models of arctic-alpine species distributions and alter projections of refugia occurrence. Predictions of species-climate models ignore landscape-scale ecological processes and may thus provide inaccurate estimates of extinction risk and forecasts of refugia where species can persist under a changing climate.