Global change seriously threatens biodiversity at all organisational levels ranging from genetic diversity, performance and occurrence of single species, taxonomic, phylogenetic and functional diversity of communities and species assemblages to properties of whole ecosystems including the provision of ecosystem services for human well-being (Lavergne et al. 2010; Parmesan 2006; Potts et al. 2010; Schröter et al. 2005). But species are not equally at risk when facing global change (e.g. Settele et al. 2008). In the context of climate change, several species-specific ecological characteristics have been identified to determine vulnerability, including diets, habitat requirements, ecological specialisation and plasticity and the ecological characteristics of interacting species (Heikkinen et al. 2010; Pöyry et al. 2009; Schweiger et al. 2012; Visser 2008; Warren et al. 2001). Thus, good knowledge of the ecological characteristics relevant for the reaction of species and communities to particular drivers of global change is needed, which can then be utilised as powerful indicators for conservation planning and action.

One of the most important ecological characteristics to assess how species react to climate change obviously is the climatic niche. While knowledge on particular species characteristics such as habitat requirements is already available for some species groups, crucial publicly available information about climatic requirements is still lacking for the majority of the species. Here we present a unique dataset on climatic niche characteristics of 397 (91%) butterfly species in Europe, which have been shown to be particularly sensitive to changing climates (Hill et al. 2002; Settele et al. 2008; Warren et al. 2001). Based on projections of future suitable climatic conditions, Settele et al. (2008) showed that under the assumption of unlimited dispersal 7% of the European butterflies are at an extremely high or very high risk (i.e. a loss of more than 95% and 85%, respectively of their current range size until 2080), 6% are at high risk (>70% loss) and 18% are at risk (>50% loss; Fig. 1). However, the more realistic assumption of no dispersal (in the given amount of time) projected 33% of the butterflies to be at an extremely high or very high risk, 26% to be at high risk and 19% to be at risk (Fig. 1).

Based on detailed data on the distribution of European butterflies, which also led to the ‘Distribution Atlas of European Butterflies’ (Kudrna 2002), the ‘Climatic Risk Atlas of European Butterflies (Settele et al. 2008) and the ‘Distribution Atlas of Butterflies in Europe’ (Kudrna et al. 2011), we extracted measures of climatic conditions (indicating niche breadth and position) within the distributional range of each species. As a consequence of this approach, users of this dataset should be aware that the provided measures refer to the realised climatic niche and not to the fundamental niche (sensu Hutchinson 1957; but see discussion in Araújo and Guisan 2006). The extracted measures reflect two primary properties of climate, energy and water, which are known to affect butterfly species performance and distributions as a consequence of physiological limitations (Buckley et al. 2011; Roy et al. 2001). Most of these measures are quite independent from each other and cover different aspects of the climatic niche (Fig. 2).

By combining a comprehensive database on the distribution of European butterflies with publicly available climatic data in combination with a constantly high level of quality control at crucial steps of the data generation, CLIMBER represents a unique and ready-to-use dataset for a broad variety of potential applications. Analysis of phylogenetic signals in the climatic niche characteristics, for instance, can be used to assess past niche evolution which can lead to projections of potential future risks in the face of rapid climate change (for a comparable analysis for birds see Lavergne et al. 2013). Also, analyses relating climatic niche properties to other species traits can be helpful to assess interdependencies of different ecological characteristics, as has been done recently for birds and their temperature and habitat preferences (Barnagaud et al. 2012). So far the most powerful application of climatic niche characteristics provided in this dataset comes from the ‘species temperature index’ (STI). The STI is simply the mean temperature value per species across its range. Based on the STI, the ‘community temperature index’ (CTI) has been suggested as a powerful and robust tool to measure the response of local communities to temperature change (Devictor et al. 2008; Devictor et al. 2012a; Devictor et al. 2012b). The CTI is calculated as the average STI value across the species or specimens in a community and has been used to analyse the temporal response to climate warming of local bird and butterfly communities across Europe. One striking result of this study was the detection of time lag effects in the community response to climate warming and that these lag effects differed between the two species groups (Devictor et al. 2012a).

STI values for European butterflies can be of great value for governmental and non-governmental conservation organisations (Van Swaay et al. 2010; Van Swaay et al. 2008). Based on the STI, the CTI was recently adopted as an indicator of climate change impact on biodiversity by the pan-European framework supporting the Convention on Biological Diversity (Streamlining European Biodiversity Indicators 2010; http://ec.europa.eu/environment/nature/knowledge/eu2010_indicators). Thus, STI and corresponding CTI values can perfectly complement and enrich the analysis of all kind of butterfly monitoring schemes. To address the fact that temperature is not the only changing climatic factor or aspect of the climatic niche, we think that the additionally provided climatic niche characteristics concerning water availability and annual climatic variability can help to enrich the landscape of target-specific analyses and indicators (Fig. 2). By providing public access to this dataset, we hope to contribute to improvements of the scientific understanding of how climate change affects species and communities and to improve monitoring and conservation actions for climate change mitigation.

For the description of the metadata we followed the standards suggested by Michener et al. (1997) in a slightly modified way.

CLIMBER: Climatic niche characteristics of the butterflies in Europe

The taxonomic coverage spans all butterfly species within the selected geographic window (397 species) and represents 91% of all European species (Fig. 5). Thirty-eight species from less well sampled Eastern European countries, Atlantic and small Mediterranean islands have not been considered (Fig. 5a). The taxonomy of European butterfly species follows Kudrna et al. (2011). Erroneous use of brackets around authors’ names was corrected in 15 cases (cf. Tshikolovets 2011; Table 2).

Aricia artaxerxes (Fabricius, 1793) and Aricia montensis Verity, 1928 are treated in CLIMBER as distinct species with parapatric distributions (see Sanudo-Restrepo et al. 2013). The latter species is confined to the Iberian Peninsula and North Africa.

For the local Macedonian endemic Pseudochazara amymone Brown, 1976 no data were available for the considered time period. After its first discovery in Greece in 1975, the species was not reliably recorded again until its recent rediscovery in Southern Albania (Eckweiler 2012). According to Eckweiler (2012), Pseudochazara amymone should be considered a subspecies of Pseudochazara mamurra (Herrich-Schäffer, [1846]), which is widespread in the Middle East.

The following species in our database actually comprise records of more than one species, most of which were recognized only recently, and are difficult or impossible to distinguish without genitalia examination or molecular methods.

•Carcharodus alceae (Esper, 1780) probably contains data of the sibling species Carcharodus tripolinus (Verity, 1925) from the Southern Iberian Peninsula, differing only in genitalia characters.

•Leptidea sinapis (Linnaeus, 1758) is a complex of three sibling species, and includes data of Leptidea juvernica Williams, 1946, and Leptidea reali Reissinger, 1990 (Dincă et al. 2011b; Dincă et al. 2013). Whereas Leptidea sinapis can be separated by their genitalia, the other two taxa can only be separated from each other by molecular characters. Leptidea reali seems to replace Leptidea juvernica in SW Europe, and both occur largely in sympatry with Leptidea sinapis.

•Lycaena tityrus (Poda, 1761) includes data of Lycaena bleusei Oberthür, 1884 from Central Spain and Central Portugal, which appears to be a distinct species according to unpublished molecular data.

•Melitaea athalia (Rottemburg, 1775) includes the Southwest European Melitaea nevadensis Oberthür, 1904 (syn. celadussa Fruhstorfer, 1910) which might only be a subspecies of the former. Molecular data are inconclusive regarding the taxonomic status of these parapatric taxa.

•Melitaea phoebe (Goeze, 1779) recently turned out to be a complex of at least two largely sympatric species with distinctive larval colouration, and our data probably include records of Melitaea ornata Christoph, 1893 (syn. telona Fruhstorfer, 1908 and emipunica Verity, 1919) (see Toth et al. 2013; Toth and Varga 2011; Tshikolovets 2011).

•Polyommatus icarus (Rottemburg, 1775) includes data of Polyommatus celina (Austaut, 1879), which was recognized as a distinct species from North Africa and the Canary Islands by molecular methods (Wiemers et al. 2010), but also occurs in Southern Spain, and appears to replace Polyommatus icarus in the Balearic Islands, Sardinia, and Sicily (Dincă et al. 2011a).

•Pontia daplidice (Linnaeus, 1758) includes the data of the sibling species Pontia edusa (Fabricius, 1777), a parapatric taxon, which can only be distinguished by molecular methods (Geiger and Scholl 1982; John et al. 2013; Wiemers unpubl.).