The intensively discussed taxonomic complexity of the Dactylorhiza genus is probably correlated with its migration history during glaciations and interglacial periods. Previous studies on past processes affecting the current distribution of Dactylorhiza species as well as the history of the polyploid complex formation were based only on molecular data. In the present study the ecological niche modeling (ENM) technique was applied in order to describe the distribution of potential refugia for the selected Dactylorhiza representatives during the Last Glacial Maximum. Additionally, future changes in their potential habitat coverage were measured with regard to three various climatic change scenarios. The maximum entropy method was used to create models of suitable niche distribution. A database of Dactylorhiza localities was prepared on the grounds of information collected from literature and data gathered during field works. Our research indicated that the habitats of majority of the studied taxa will decrease by 2080, except for D. incarnata var. incarnata, for which suitable habitats will increase almost two-fold in the global scale. Moreover, the potential habitats of some taxa are located outside their currently known geographical ranges, e.g. the Aleutian Islands, the western slopes of the Rocky Mountains, Newfoundland, southern Greenland and Iceland. ENM analysis did not confirm that the Balkans, central Europe or central Russia served as the most important refugia for individual representatives of the Dactylorhiza incarnata/maculata complex. Our study rather indicated that the Black Sea coast, southern Apennines and Corsica were the main areas characterized by habitats suitable for most of the taxa.
Citation: Naczk AM, Kolanowska M (2015) Glacial Refugia and Future Habitat Coverage of Selected Dactylorhiza Representatives (Orchidaceae). PLoS ONE 10(11): e0143478. https://doi.org/10.1371/journal.pone.0143478
Editor: Tzen-Yuh Chiang, National Cheng-Kung University, TAIWAN
Received: October 11, 2014; Accepted: November 5, 2015; Published: November 23, 2015
Copyright: © 2015 Naczk, Kolanowska. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The present study was financially supported by the Polish Ministry of Science and Higher Education (grant no. NN 303 5491) and The Foundation For Polish Science (Fundacja na rzecz Nauki Polskiej, FNP). The publication is financed from European Social Fund as a part of the project "Educators for the elite—integrated training program for PhD students, post-docs and professors as academic teachers at University of Gdansk" within the framework of Human Capital Operational Programme, Action IV. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The Last Glacial Maximum (LGM) refers to the period between 26,500 and 20,000 years ago  that greatly affected the distributions and population sizes of many temperate plant species. Migration routes and the history of colonization after the LGM have been studied for various taxa, e.g. Viola rupestris , Lathyrus vernus , Silene dioica , Calluna vulgaris , and Betula pendula . For a long time, it has been commonly assumed that during the LGM a lot of temperate plant species survived within refuge areas in the Balkan, Apennine and Iberian Peninsulas and in the Caspian and Caucasian regions (“the southern refugia hypothesis” [7–8]. It has also been established that the general view of high genetic diversity and haplotype richness in refugial areas in the south is the result of refugial persistence and accumulation of genetic variation during ice ages, in comparison with low diversity in glaciated areas in the north. Populations in previously glaciated areas are genetically depleted as a consequence of rapid postglacial colonization and the repeated bottleneck effect during stepwise migration [9–10]. The hypothesis is just a general concept and the individualistic nature of species' responses to climate change implies that the location of refugia varies according to the climatic conditions preferences as well as to the way individual species or populations adapt . The incoming evidence suggests that the southern refugia for the temperate species were complemented by more northern refugia during the LGM. "The northern refugia hypothesis" assumes more complex patterns for the distribution of genetic diversity, where suitable niches were also distributed much more widely in Europe during the LGM, not only across Southern Europe, but also in Central Europe close to the line of the ice sheet [12–13]. This proposal has also been indicated in phylogeographic studies of selected Dactylorhiza species .
Dactylorhiza Neck. ex Nevski is a temperate orchid genus which includes taxa of various ploidy levels [15–18]. They are either diploids (2n = 40) or tetraploids (2n = 80). Most currently recognized Dactylorhiza species belong to the Dactylorhiza incarnata/maculata polyploid complex. The most problematic taxa within this complex belong to D. majalis s.l., which evolved by multiple and independent hybridization events between two broadly defined parental lineages: D. incarnata s.l.—recognized as the paternal lineage and D. maculata s.l.—considered to be the maternal lineage [18–29]. The taxonomic complexity of this genus is probably due to its migration history during glaciations and interglacial periods, as well as polyploidization episodes, which took place several times e.g. [18,24,28–30].
As assumed by Hedrén et al. , this complex must have originated before the Weichselian glaciation and its representatives are now distributed across Europe and Asia Minor [32–33]. Within this range, the allotetraploids often occupy limited occupancy areas [17,34] and many of them are restricted to those regions in more northern or western Europe that were completely covered by the ice sheet during the Weichselian glaciation. It has been postulated that numerous allotetraploid species evolved after the ice age on several, independent occasions by repeated local polyploidization events in areas where they are currently found (e.g. [24,31]). This hypothesis has been supported by molecular data, including allozyme variation [18–19,20,35] and AFLPs . However, analysis of plastid DNA [26,28,36] has disclosed that some variants within the allotetraploids have not been encountered in the extant parental lineages, indicating that the allotetraploid complex may also include older taxa which currently remain unknown.
The Dactylorhiza incarnata/maculata polyploid complex constitutes an extremely dynamic model of polyploid speciation and extinction, in which polyploid species evolve continuously from the same set of broadly defined parental lineages. The pattern of colonization inferred for the complex representatives seems to be unusual compared with most other temperate taxa, where polyploids have proven to be strong colonizers of Arctic regions , whereas their diploid progenitors have remained much further south .
As is the case with other orchids, Dactylorhiza is not present in fossil material. Previous studies on past processes affecting the present distribution of the genus representatives as well as the history of the formation of polyploid complex have been based only on molecular data.
Environmental niche models, which are generated by combining species occurrence records (and/or absence data) with environmental GIS data layers have become increasingly important tools to address various issues in biogeography, ecology, evolution and conservation biology research [38–39]. Many methods have been used for modeling of species distribution (eg. BIOCLIM [40–41], DOMAIN , GLM , MaxEnt ). MaxEnt is considered as the most reliable machine learning programme in computing presence-only data (e.g. [45–48]). This application is particularly useful in the course of determining locations of glacial refugia of plants and animals, especially when the fossil material is poor (e.g. [49–51]. Nowadays, when the rate of climate change accelerates, it is increasingly important to understand the consequences  and here species distribution models based on current ecological niche constraints are used to project future species distributions (e.g. [53–55]).
While the frequent niche shifts in the polyploid complexes could be expected, the evidences for alternative patterns were reported. Findings published in the recent studies suggest that the niche conservatism may be more common between different cytotypes than previously recognized. The tendency for the niche of a taxon to be little changed over time was confirmed for several polyploid species complexes, i.e. Claytonia perfoliata (Portulacaceae) , Larrea tridentata (Zygophyllaceae) , Houstonia (Rubiaceae)  and Heuchera cylindrica (Saxifragaceae) . Based on the frequent spatial segregation of the diploid and the polyploid cytotypes, and considering the fact that the polyploidization may drive the ecological divergence, species distribution modeling seems to be adequate approach for studies on biogeography of polyploid complexes.
Because no study revealed any ecological shifts within Dactylorhiza species so far, we assume that their niches have remained unchanged since the LGM and they will not transform in the predictable future. The quantified niche can be therefore projected across a geographic area for the purposes of mapping applicable climatic conditions for studied taxa and predicting its potential distribution . In our research, the ecological niche modeling (ENM) technique has been applied in order to estimate distribution of suitable niches for three interesting Dactylorhiza species groups (D. incarnata, D. maculata and D. majalis) during the LGM and in the predictable future. Noteworthy, the studied tetraploid taxa represent fixed hybrids between known parental lineages. They should be therefore considered as separated entities with their own evolutionary history and characterized by specific habitat requirements. The aim of the study was to confront the phylogeographic insights into distribution of glacial refugia with the outcomes of the climate envelope models as well as to evaluate the future changes in the potential habitat coverage of the studied orchids.
Material and Methods
List of localities
The potential niche modeling was performed using the maximum entropy method implemented in MaxEnt version 3.3.2 [44,61–62] and based on species presence-only observations. The list of locations in which the Dactylorhiza populations were found was prepared based on available literature data [14,26,29,30,35,63–68] and information gathered during the field works . Only these localities which could be precisely placed on the map (Fig 1) were used in the analysis and the duplicate presence records were removed. The complete database is provided in S1 Annex. Taxonomic classification within the Dactylorhiza incarnata/maculata complex followed Hedrén’s concept .
Ecological niche modeling
The maximum entropy method, as implemented in Maxent version 3.3.2 software, was used to create models of the suitable niche distribution [44,62,70]. Nineteen climatic variables in 2.5 arc minutes (± 21.62 km2 at the equator) developed by Hijmans & al.  as well as the altitudinal data were used as input data (S1 Table). The maximum iterations were set to 10,000 and the convergence threshold to 0.00001, thereby forcing the program not to finish before the threshold was reached. For each run, 20% of the data were used to be set aside as test points . The "random seed" option was applied and it provided a random test partition and background subset for each run. The run was performed as a bootstrap with 100 replicates and the output was set to logistic. All operations on GIS data were carried out on ArcGis v. 9.3 (ESRI). Because only current location data for studied taxa were available the records were the base for all models. The contemporary species-climate relationships were projected for the LGM and three future layers using the settings described above.
The bioclimatic data for the LGM were developed and mapped by Paleoclimate Modelling Intercomparison Project Phase II  (PMIP2, CCSM) and downloaded from www.worldclim.org. In the projections the climatic data are extrapolated onto a modern world map, so it is easier to compare the distribution of glacial refugia with the current range of the studied taxa. For the purposes of showing the actual location of the refugial areas, we also ran the ENM using CCSM4 bioclims maps for the LGM developed by Coupled Model Intercomparison Project Phase 5 (CMIP5 ).
The future climatic projections related to a hypothetical climate change in 2080 with A1b (CCCMA-CGCM3 simulation), A2a (CCCMA-CGCM2) and B2a (CCCMA-CGCM2) scenarios obtained from the CIAS website (http://ccafs-climate.org; ). Those models were used in numerous recent studies focused on climate change impact on the distribution of various organisms (e.g. [76–78]). The A1b scenario is characterized by a balance across all energy sources (where balanced is defined as not relying too heavily on one particular energy source, provided that similar improvement rates apply to all energy supply and end-use technologies). The A2 storyline describes a highly heterogeneous future world with regionally oriented economies. The main driving forces involve a high rate of population growth, increased energy consumption, land-use changes and slow technological change. The B2 is a scenario with a general evolution towards environmental protection and social equity. The modeling for the future was performed with the same settings as for the present time. The global coverage area of the most suitable niches was calculated in order to evaluate general effects of the global warming on the potential distribution of Dactylorhiza. To precisely determine the possible negative effect of climate changes on the studied orchids, the same computation was made for the region of their known geographical range (within the longitude of 15.37–180.30 and the latitude of 29.21–86.98). Because most of the known populations of the species studies were found in the zones characterized by high habitat fitness of 0.7, the number was applied as the threshold value of the most suitable niches.
All the models created received high AUC (area under ROC curve) scores of 0.987–0.998 (S2 Table). The results are consistent with the outcomes of previous studies which indicated the reliable performance of the method for the purposes of developing ecological niche models based exclusively on presence-only data .
The models of the suitable niches of the selected Dactylorhiza representatives' distribution in the LGM in Europe are presented in Figs 2–4 (PMIP2 based models). The global models are presented in S1–S6 Figs (PMIP2 based models) and S4–S6 Figs (CMIP5 based models).
Within the known geographic range of the studied taxa, the most probable refugia of D. incarnata var. cruenta Hyl. were located on the western and northern coasts of the Caspian Sea as well as on the coast of the Sea of Azov and the western coast of the Black Sea (Danube delta). Some less suitable niches were distributed on the eastern slopes of the northern Apennines, the western slopes of the western Alps and along the Danube. The main refugia of D. incarnata var. incarnata were located in Corsica and the southern Apennines. Populations of D. incarnata var. ochroleuca (Wüstnei ex Boll) P.F. Hunt & Summerh could survive on the north-eastern coast of the Caspian Sea, the southern part of the Caspian Depression, the eastern slopes of the Caucasus and the southern coast of the Sea of Azov, including western Crimea. Some less suitable habitats of this taxon were located in Corsica, the eastern slopes of the northern Apennines and the western Alps.
The models created indicated the eastern part of the Black Sea, the southern Apennines and Corsica as the most probable refugia for D. maculata ssp. fuchsii Hyl. Additional habitats could be found in Crete and Sicily. The potential glacial range of D. maculata ssp. maculata included the eastern coast of the Black Sea, Crete, the Cyclades, Rhodes as well as the southern Apennines, Sicily, Corsica, Sardinia and Menorca. Some less suitable niches could be located in Andalusia.
During the LGM, populations of D. majalis ssp. lapponica H. Sund. could survive in Corsica, the southern Apennines, northern Sicily, and the southern coast of the Sea of Azov. Additional, fragmented refugia were located in Crete. Populations of D. majalis ssp. majalis in the LGM could be distributed in the southern Apennines, Corsica and Sicily. Unlike those taxa, refugia of D. majalis ssp. traunsteineri (Saut. ex Rchb. f.) H. Sund. were probably distributed on the Pannonian Plain, Dinaric Alps as well as along the Danube and at the lower altitudes along the range of the Alps. Numerous less suitable niches were also located in the eastern Pyrenees and the slopes of the Caucasus.
When considering the global distribution of the potential glacial refugia of the studied taxa, several areas were indicated by ENM analysis to be well outside their current geographical ranges (Figs 2–4). The most important regions that could be characterized by climatic conditions suitable for the studied Dactylorhiza representatives are the Aleutian Islands and the Alaska Peninsula in North America, Patagonia in South America and the Kamchatka shore.
Additionally, the south-eastern slopes of the Himalayas were indicated by MaxEnt as suitable for D. majalis ssp. lapponica, D. majalis ssp. majalis and D. incarnata var. incarnata. In the LGM period, the Zambezi River Basin could have served as a refugium for D. majalis ssp. majalis, D. maculata ssp. maculata and D. incarnata var. incarnata. Suitable niches of D. majalis ssp. traunsteineri were also located on the north shore of the Gulf of Mexico and the east of the southern Appalachian Mountains. In the glacial period, suitable habitats of D. majalis ssp. traunsteineri, D. incarnata var. cruenta and both representatives of D. maculata complex were also located in southern Japan.
Current potential distribution
The distribution of the suitable niches for allotetraploid representatives of D. majalis in Europe and Asia Minor indicated by ENM analysis corresponds to the known geographical range of those taxa [17,79]; however, several regions where none of these orchids have been found so far, have also been specified as potentially suitable. For D. majalis ssp. lapponica, the suitable niches outside its geographical range are located in the southern parts of Svalbard, the Kamchatka Peninsula, the Aleutian Islands, the Alaska Peninsula, the eastern part of Newfoundland and south-western Greenland. Some less suitable habitats are also found in Iceland (Fig 5A). Some habitats of lower suitability potentially suitable for D. majalis ssp. majalis were indicated in the models in the Aleutian Islands, the Alaska Peninsula and small areas in eastern Newfoundland (Fig 5B). The Kuril Islands, Kamchatka Peninsula, Aleutian Islands, Alaska Peninsula, northern Coast Mountain and northern Rocky Mountains were indicated as potentially suitable for D. majalis ssp. traunsteineri. It is noteworthy that the models did not indicate numerous areas in central Asia as suitable for this taxon, despite known occurrence reports of D. majalis ssp. traunsteineri in these regions (Fig 5C).
In all the three models created for the diploid representatives of D. incarnata, the Aleutian Islands and Alaska Peninsula were indicated as potentially suitable for those taxa; however, none of these have been found in the regions (Fig 6). Additionally, habitats suitable for D. incarnata var. cruenta are also located along the shores of the Black and Caspian Seas and the eastern margins of the Kamchatka Peninsula. Some less appropriate habitats are also located in the Rocky Mountains and Patagonia (Fig 6A). Among the areas indicated by the ENM analysis as suitable for D. incarnata var. ochroleuca, it was only the northern Caspian region that this taxon was not observed in. Additional habitats were indicated in the Aleutian Islands and the Alaska Peninsula (Fig 6C). Less suitable niches are located in the northern Rocky Mountains and northern Iceland.
Models created for the D. maculata-complex indicated several regions suitable for the studied taxa. The Black Sea, south-western Himalayas, Kuril Islands, Aleutian Islands and Alaska Peninsula are potentially suitable for D. maculata ssp. fuchsii (Fig 7A) and the Kuril Islands, Kamchatka Peninsula, Aleutian Islands, Alaska Peninsula, and the eastern part of Newfoundland are suitable for D. maculata ssp. maculata (Fig 7B). Noteworthy is the fact that the models did not indicate numerous Asian areas as suitable for the latter taxon, the occurrence of which had been reported in previous publications (e.g. ).
Future habitat coverage
Firstly, the evaluation of the global coverage of their suitable habitats was calculated (Table 1). The habitats of almost all taxa will decrease in scenario A1b. In scenario A2a a loss of habitats will be observed for D. incarnata var. ochroleuca, D. maculata ssp. fuchsii, D. maculata ssp. maculata and D. majalis ssp. majalis and negative changes will affect D. maculata ssp. maculata and D. majalis ssp. majalis in scenario B2a. The most interesting change may be observed with regard to the coverage area of suitable niches for D. incarnata var. incaranta in A2a and B2a scenarios. According to the models, the area of suitable habitats of this orchid will increase almost two-fold. However, there are no substantial changes concerning the areas of occurrence of D. incarnata var. cruenta, D. majalis ssp. lapponica and D. majalis ssp. traunsteineri in any scenario.
It is noteworthy that the suitable habitats of some taxa are located outside their known geographical ranges, e.g. the Aleutian Islands (D. incarnata var. cruenta, D. maculata ssp. maculata), the western slopes of the Rocky Mountains (D. incarnata var. cruenta, D. maculata ssp. maculata), Newfoundland (D. maculata ssp. maculata, D. majalis ssp. lapponica), southern Greenland and Iceland (D. majalis ssp. lapponica), Tierra del Fuego (D. incarnata var. cruenta, D. incarnata var. ochroleuca), and New Zealand (D. maculata ssp. fuchsii, D. maculata ssp. maculata).
To determine the precise possible negative effect of climate changes within the known geographical range of each taxon, the coverage of the most suitable niches was calculated for Eurasia (Table 2). Within the region in A1b scenario only the coverage of D. incarnata var. incarnata habitats will slightly increase (by about 2%). The available niches of all other taxa will decrease. The changes predicted in the scenario will be the most destructive for niches of the studied orchids. In A2a scenario a reduction of suitable habitats will not be observed only in D. maculata ssp. fuchsii (0.4% increase). In B2a scenario the coverage of suitable niches of D. maculata ssp. fuchsii and D. maculata ssp. maculata will not be negatively affected.
Potential glacial refugia within the known geographical range
The almost complete distinction between diploid lineages of D. incarnata s.l. and D. maculata s.l. at allozyme loci suggests that the lineages characterized with a long period of separated history and that gene flow between them did not exist for a long time [18,23]. However, according to our study, suitable niches of the taxa overlapped during the LGM in Europe.
In contrary to the genetic research results [14,31] the ENM analysis did not confirm that the Balkans, central Europe or parts of central Russia located between the Fennoscandian ice sheet and the Urals served as the most important refugia for most representatives of the Dactylorhiza incarnata/maculata complex during the LGM. Also our analysis did not point out any areas of sheltered topography in central Europe to be refugial for D. maculata ssp. fuchsii which was postulated by Ståhlberg and Hedrén . Our study indicated the Black Sea coast, southern Apennines and Corsica as the main areas characterized by habitat suitable for all taxa, except D. majalis ssp. traunsteineri.
Below, we present a more detailed discussion of the differences and similarities between the results obtained in the course of our study and data collected by other authors, mainly based on the distribution of plastid DNA haplotypes.
D. majalis complex (ssp. lapponica, ssp. majalis, ssp. traunsteineri).
Numerous allotetraploids have originated by hybridization between parental lineages related to present-day D. incarnata s.l. and D. maculata s.l. Within the D. majalis core complex (e.g. D. majalis ssp. majalis, D. majalis ssp. lapponica, and D. majalis ssp. traunsteineri sensu Nordström and Hedrén ) presence of markers which are no longer found in the parental lineages was notified [28,36]. This suggests that some allotetraploids may be relatively old and those most probably originated long before the last glaciation. It is generally accepted that allotetraploids in D. majalis s.l. are often narrow endemics, where populations were established by a single founder event or immigrated by long-distance dispersal. Subsequently, the initial population of Dactylorhiza expanded to form the present metapopulation, where genetic drift and natural selection driven by ecological conditions caused the evolution of distinct subspecies within the group . The analysis of plastid DNA, the results of which were presented in previous studies [30–31,80–81], indicated several migration routes for the allotetraploid representatives of Dactylorhiza.
Given the distribution of the H1 haplotype, Scandinavian populations of D. majalis ssp. traunsteineri/lapponica originated in continental Europe and entered the Swedish and Norwegian mainland probably via Denmark, while the analysis of the H4 haplotype indicated that other routes of migration might have occurred via Estonia and Finland to northern Sweden and Norway [30,80–81]. The high frequencies of the two haplotypes may be explained by the rapid expansion of populations carrying the haplotypes after the last glaciations.
Our study confirmed the location of glacial refugia of D. majalis ssp. traunsteineri around the Alps and the area also maintained most of the recognized plastid haplotypes and the largest number of unique haplotypes for this taxa . However, the suitable niches of other Dactylorhiza representatives were located further to the south in the ENM analysis (e.g. Corsica, the southern Apennines, the southern coast of the Sea of Azov), but we do not have molecular data for the remaining subspecies of D. majalis studied, in order to be able to compare them. Several of the plastid haplotypes that were found in Greece have not been detected in material outside this country. Other findings suggest that also northern parts of the Balkans are characterized by high overall diversity . Hereby parts of the Balkans could also served as an important refugium from which northern European allotetraploids were recruited. Dactylorhiza populations could have also survived in other areas and it is suggested that allotetraploid populations in eastern Europe and Russia should be collected for extended studies .
D. incarnata complex (var. cruenta, var. incarnata, var. ochroleuca).
Allozyme data clearly indicated that Turkish populations of D. incarnata s.l. were far more genetically variable than populations from northern Europe, because the species appears to have lost genetic variation during its recolonization from southern refugia after the Weichselian glaciation [18,22–23,64,83]. Pillon et al.  revealed a higher level of genetic diversity in the Caucasus and the Mediterranean Basin than in western Europe using plastid and nuclear DNA sequence data. Numerous studies have presented a similar picture of differentiation, based on the plastid polymerase chain reaction-restriction fragment length polymorphisms (PCR-RFLPs; [24,26]), nrDNA internal transcribed spacer/external transcribed spacer (ITS/ETS) sequence data [27,84–85] and length-variable ITS fragments [28,86]. Generally, the common occurrence of several haplotypes in southern Scandinavia and also in adjacent areas to the south and the east of the Baltic Sea suggests that D. incarnata s.l. has been dispersed across the Baltic on repeated occasions. Furthermore, all Scandinavian haplotypes were present in the material analyzed from Turkey further to the southeast. The pattern obtained suggests that the principal direction of migration into Scandinavia was from the south or the southeast, although it is possible that additional migrants may have come from other directions .
Our study confirmed previously published hypothesis on the location of glacial refugia of D. incarnata s.l. [22,28,84,87]. The suitable niches of D. incarnata s.l. during the LGM were located in Corsica, the southern Apennines, the southern Balkans and the northeastern coast of the Black Sea. In addition to the available literature data  the location of habitats preferred by D. incarnata var. cruenta was indicated in our test around the Alps.
D. maculata complex (ssp. fuchsii and ssp. maculata).
A large number of regionally focused studies based on morphological and cytological data [29,88–94], as well as on plastid DNA and nuclear ribosomal DNA markers [24,26–28,36,86,95], have indicated several potential sources of tetraploid populations of D. maculata s.l. According to the results, central Europe could serve as the source of the southern/western lineage of D. maculata ssp. maculata, as well as for D. maculata ssp. fuchsii. This genealogical lineage is currently distributed from Portugal in the south, through the British Isles, western and central Continental Europe, to the Scandinavian mountain range in the north, and also from Iceland in the west to Romania in the east and is characterized by the Group II of haplotypes. However, the northern/eastern lineage of D. maculata ssp. maculata may have survived the LGM in central Russia and also west of the Urals and at present, it is spread from northern Norway to the Urals in the east . It is also expected that the pattern is additionally complicated by the areas of sheltered topography in central Europe which may have provided suitable habitats for the more thermophilous D. maculata ssp. fuchsii during the LGM [13–14]. The Mediterranean region and the Caucasus have not contributed to the northward migration of either D. maculata ssp. fuchsii or D. maculata ssp. maculata . However, the distribution of genetic diversity suggests that northern Europe has been colonized from the south to the north while, there is no support for a north-eastern recolonization route for D. maculata ssp. fuchsii [14,29]. For Scandinavia, the average gene diversity over loci gradually decreases from the south to the north, in accordance with a stepping-stone model of gene dispersal . Nonetheless, our study did not indicate the presence of suitable niches in central Europe or central Russia either for diploid D. maculata ssp. fuchsii or for autotetraploid D. maculata ssp. maculata. The suitable niches obtained for D. maculata representatives were located more to the south in the ENM analysis (e.g. Corsica, the southern Apennines, the eastern coast of the Black Sea). However, Ståhlberg & Hedrén  received a low frequency of private haplotypes for these areas which is reflected in low values in terms of haplotype richness and gene diversity . This limited diversity indicates a history of small population size and a consequent loss of variation through genetic drift . It should be stressed that in the previous molecular studies  only two populations were sampled in southern Apennines and no data from the Corsica were included in the analysis. Extended sampling of D. maculata s.l. in broader areas of the Balkans, as well as in central Europe and central Russia would be needed to resolve the differences at a further stage.
Glacial refugia outside known geographical range
The ENM analysis indicated several additional potential refugial areas for Dactylorhiza representatives, all of which were located near the ice sheet borders. The Aleutian Islands, the edges of the Alaska Peninsula and Kamchatka shore indicated in the models, correspond to the ice-free part of Beringia which served as an important refugium for numerous plants during the LGM . The other region of North America where Dactylorhiza populations could hypothetically survive was located between the Cordilleran Ice Sheet and the Laurentide Ice Sheet . The refugia in Patagonia were located both east and west of the glacier . While the climatic conditions of those areas were apparently appropriate for the studied orchids, the absence of their populations near the mentioned potential refugia bring into question their actual usefulness for those plants. Considering the current geographical ranges of the studied taxa, we would argue that those regions actually contribute to the current patterns of Dactylorhiza diversity and distribution.
Ecological niche modeling vs molecular analysis
The three major southern Mediterranean peninsulas (Iberian, Apennine, Balkan) have for a long time been considered as main Pleistocene refugia for European flora and fauna . The recent studies indicated that the paradigm is too simplistic to explain observed diversity patterns [101–102]. Some authors have postulated that presence of refugia within refugia  or multiple northern refugia [13,104] should be considered as additional factors affecting current distribution of European taxa. Traditionally, paleoecological data [105–106] and analysis of endemic taxa concentration  were used to localize glacial refugia for various organisms.
Advances in molecular techniques allowed to reveal historical patterns of population divergence, including identification of potential refugial sites . Many of these studies have been devoted to the Pleistocene refugia. The molecular analyses are useful in pointing out regions characterized by the presence of multiple lineages or high genetic diversity which are indicators of putative glacial refugial areas . However, extinction of genetic variants, incomplete sampling and large-scale postgalcial range shifts can interrupt phylogenetic patterns and affect conclusions on the past distribution of studied organisms . Moreover, the phylogeographic approaches have one inherent weakness: candidate refugia outside the present species range can rarely be identified . This limitation of phylogeographic approach can be overcome through the use of species distribution models. Such projections assume that a species is in equilibrium with its environmental requirements—its distribution is determined mainly by the environment, not other factors such as competition or dispersal limitation. The main advantage of the ENM approach is providing linkage to bioclimatic and geographical factors affecting the distribution of glacial refugia. This technique gives first approximation of the spatial distribution and extent of potential Pleistocene refugia. The inadequacy of ENM models may be the result of insufficient samples (location records) included in the analysis or biased samples of occurrence data , however we believe that we can avoid these troubles by removing duplicated records from the database. It is inappropriate to ask the question which of the two approaches (ENMs or phylogeography) is more reliable in describing distribution of glacial refugia. The rigorous population genetic nature of the phylogeographic approaches is made more explicit spatially by the ENM approaches resulting in more quantitative product .
Obviously the niche analysis would be more reliable with incorporation of the biotic factors, such as pollinators or mycorrhizal fungi; however those components would not be useful in our study. Unlike numerous other orchids  Dactylorhiza representatives do not show high pollinator specificity [114–116]. The published data on the symbiotic relationship indicated that the roots of Dactylorhiza representatives contain several fungi belonging to both asco- and basidiomycetes [117–118]. It would be difficult to obtain occurrence data on all possible pollinators and/or mycorrhizal fungi to include those information into any analysis. For Dactylorhiza representatives the climatic factors seem to be the most reliable input data for the niche distribution modeling.
Future habitat coverage
Our research indicated that in the next 65 years the coverage of suitable niches of most of the studied orchids will decrease and that the most damaging niche modifications will be observed in the A1b scenario of climate change. Clearly, the ENM analysis includes only climatic factors and some important ecological relationships, such as pollinator and mycorrhizal fungi occurrence, are omitted.
Orchidaceae is a group within the flora of Europe which is particularly vulnerable. It is caused, inter alia, by their specific biology, which makes them very sensitive to changes in their habitat, but also to the direct destruction of the plants, and of the localities occupied by them. At present, the most frequent reasons for the extinction of orchid populations are cessation of mowing, intensive fertilization, and fertilizer run-off from nearby fields and slopes. In addition, excessive fertilization of meadows and areas adjacent to the locality can lead to an imbalance in the mycoflora and to the extinction of pollinators [119–120]. Currently, the number of Dactylorhiza populations is decreasing due to habitat loss related mainly to the drainage of wetlands and the excessive fertilization of wet meadows. However, several taxa have shown the ability to colonize disturbed and anthropogenic sites, e.g. D. maculata ssp. fuchsii and D. majalis spp. praetremissa (Druce) D.M. Moore & Soó .
It has been hypothesized that each allotetraploid taxon occupies a narrow niche, has its own range and is a good indicator of specific habitats . The higher expansion potential of allopolyploids may be explained by their recent origin. It has been suggested that diploids are characterized by lower genetic variation in the north part of their ranges due to the loss of alleles during their recolonization from southern refugia . Allopolyploids are variable in their genetic aspect as they arose from species with different allele compositions. Allotetraploids may express the highly heterozygous genome of divergent parents with unique properties. This would explain their ability to grow in a broad spectrum of habitats of intermediate conditions preferred by their diploid progenitors [18,23,122]. External factors such as the preadaptation of the floral morphology to available pollinators, a suitable phenology, competition with existing allotetraploids and the ability to colonize previously unexploited habitats may determine which allotetraploid derivatives can multiply and spread [18,123].
The ENM analysis indicated, however, that the future available habitat coverage will not be related to the ploidy level of the studied taxa. An increase in the potential global range will be observed in D. incarnata var. incarnata, D. majalis ssp. lapponica and D. majalis ssp. traunsteineri. The most surprising results were obtained for D. incarnata var. incaranta, which demonstrated a two-fold increase of coverage area of suitable niches for this taxon; however, only on a global scale. In its currently known geographic area of occurrence, the increase will be not substantial. This orchid is characterized by a relatively narrow ecological amplitude and prefers a habitat with neutral or slightly alkaline pH, but these habitats like others have been—and indeed still are—subject to human pressure . The available research results suggest that, as a result of intensive drainage work, a reduction in the number of individuals of D. incarnata in the locality can occur, and this can be observed a year or two after conducting the work, and if the unfavorable conditions remain, the total disappearance of the population occurs within 4–6 years . A lowering of the groundwater level leads to a number of processes changing, e.g., the structure, thermals, and pH of the soil, which usually becomes acidified. We should keep in mind the fact that the groundwater level is a limiting factor for the occurrence of this taxon but such hydrological factors are not taken into account in the ENM analysis. As indicated in the analysis, the potential habitats of some Dactylorhiza representatives will also be distributed far outside their currently recognized geographical range; however, the chance of their appearance in the locations is unlikely. While a single representative of the genus, D. aristata (Fisch. ex Lindl.) Soó occurs in the Aleutian Islands, the migration of the European taxa to this region is difficult to imagine. Also, expansion to other areas indicated by the models as suitable for some of the studied taxa, e.g. southern Greenland, Tierra del Fuego and New Zealand appears to be improbable. These regions may not be inhabited by representatives of the genus, e. g. due to unfit soils, the lack of mycorrhizal fungi or preferred pollinators, and other ecological factors, which were not included in the selected model. It should be emphasized that our ENM analysis allowed us to identify only the climatic niches suitable for the studied orchids, not their realized niches, which in our opinion are restricted in distribution to Europe and temperate Asia.
A question arising in the context of potential biological invasions is: could the studied taxa be able to survive and spread in the areas indicated by the ENM analysis when introduced or accidentally transferred? Temperate orchids seems to have relatively small invasive potential in comparison to the tropical and subtropical taxa. So far, the only successful Orchidaceae representative native for Europe, north Africa and south-west Asia that has spread outside its range is Epipactis helleborine. In our opinion, the invasive potential of Dactylorhiza is poorer than in E. helleborine mainly due to its habitat requirements and pollinator limits. While E. helleborine grows in forest litter where the competition for pollinator service is relatively low, Dactylorhiza representatives occur in various open or semi-open vegetation areas with numerous other plants which are often in flower at the same time. In a newly inhabited region, these orchids would probably vanish due to the high level of pollinator competition with native plants [126–127].
Within the known geographical range of the studied taxa a general habitat loss will be observed as a result of the predicted climatic changes. While the ENM analysis was used in the conservation planning and assessment of possible changes in the suitable niche coverage of numerous taxa (i.a. [128–130]), it is worth noting that the process investigated in our research may be intensified by local anthropogenic pressures [131–133].
S1 Annex. List of localities used in the ecological niche modeling.
S1 Table. Variables used in the ecological niche modeling.
S2 Table. The average training AUC for the replicate runs measured for Last Glacial Maximum (LGM), future climate change scenarios (A1b, A2a, B2a) and present time models.
Standard deviation values are given in parenthesis.
S1 Fig. Global distribution of the suitable niches of D. majalis ssp. lapponica (A), D. majalis ssp. majalis (B), and D. majalis ssp. traunsteineri (C) in the LGM (PMIP2 based models).
S2 Fig. Global distribution of the suitable niches of D. incarnata var. cruenta (A), D. incarnata var. incarnata (B), and D. incarnata var. ochroleuca (C) in the LGM (PMIP2 based models).
S3 Fig. Global distribution of the suitable niches of D. maculata ssp. fuchsii (A) and D. maculata ssp. maculata (B) in the LGM (PMIP2 based models).
S4 Fig. Global distribution of suitable niches of D. majalis ssp. lapponica (A), D. majalis ssp. majalis (B), and D. majalis ssp. traunsteineri (C) in the Last Glacial Maximum (CMIP5 based models).
The current coastal line is indicated with a continuous line.
S5 Fig. Global distribution of suitable niches of D. incarnata var. cruenta (A), D. incarnata var. incarnata (B), and D. incarnata var. ochroleuca (C) in the Last Glacial Maximum (CMIP5 based models).
The current coastal line is indicated with a continuous line.
S6 Fig. Global distribution of suitable niches of D. maculata ssp. fuchsii (A) and D. maculata ssp. maculata (B) in the Last Glacial Maximum (CMIP5 based models).
The current coastal line is indicated with a continuous line.
S7 Fig. Distribution of suitable niches of D. majalis ssp. lapponica (A), D. majalis ssp. majalis (B), and D. majalis ssp. traunsteineri (C) in 2080 modeled for three various climate change scenarios.
S8 Fig. Distribution of suitable niches of D. incarnata var. cruenta (A), D. incarnata var. incarnata (B), and D. incarnata var. ochroleuca (C) in 2080 modeled for three various climate change scenarios.
We wish to express our gratitude to Professors Dariusz L. Szlachetko and Marek S. Ziętara for their valuable comments on the manuscript. This paper was prepared thanks to the support of The Foundation For Polish Science (Fundacja na rzecz Nauki Polskiej, FNP) and The Polish Ministry of Science and Higher Education (grant no. NN 303 5491).We acknowledge the World Climate Research Programme's Working Group on Coupled Modelling, which is responsible for CMIP, and we thank National Center for Atmospheric Research for producing and making available their model output. For CMIP the U.S. Department of Energy's Program for Climate Model Diagnosis and Intercomparison provides coordinating support and led development of software infrastructure in partnership with the Global Organization for Earth System Science Portals.
Conceived and designed the experiments: AN MK. Performed the experiments: MK. Analyzed the data: MK. Contributed reagents/materials/analysis tools: AN MK. Wrote the paper: AN MK.
- 1. Clark PU, Dyke AS, Shakun JD, Carlson AE, Clark J, Wohlfarth B, et al. The Last Glacial Maximum. Science. 2009;25: 710–714.
- 2. Nordal I, Jonsell BA. phylogeographic analysis of Viola rupestris: three post-glacial immigration routes into the Nordic area? Bot J Linn Soc. 1998;128: 105–122.
- 3. Schiemann K, Tyler T, Widén B. Allozyme diversity in relation to geographic distribution and population size in Lathyrus vernus (L.) Bernh. (Fabaceae). Plant Syst Evol. 2000;225: 119–132.
- 4. Malm JU, Prentice HC. Immigration history and gene dispersal: allozyme variation in Nordic populations of the red campion, Silene dioica (Caryophyllaceae). Biol J Linn Soc. 2002;77: 23–34.
- 5. Rendell S, Ennos RA. Chloroplast DNA diversity in Calluna vulgaris (heather) populations in Europe. Mol Ecol. 2002;11: 69–78. pmid:11903905
- 6. Palmé AE, Su Q, Rautenberg A, Manni F, Lascoux M. Postglacial recolonization and cpDNA variation of silver birch, Betula pendula. Mol Ecol. 2003;12: 201–212. pmid:12492888
- 7. Bennett KD, Tzedakis PC, Willis KJ. Quaternary refugia of north European trees. J Biogeogr. 1991;18: 103–115.
- 8. Taberlet P, Cheddadi R. Quaternary refugia and persistence of biodiversity. Science. 2002;297: 2009–2010. pmid:12242431
- 9. Hewitt GM. Some genetic consequences of ice ages, and their role in divergence and speciation. Bot J Linn Soc. 1996;58: 247–276.
- 10. Hewitt GM. Speciation, hybrid zones and phylogeography—or seeing genes in space and time. Mol Ecol. 2001;10: 537–549. pmid:11298967
- 11. Stewart JR, Lister AM, Barnes I, Dalén L. Refugia revisited: individualistic responses of species in space and time. Proc Biol Sci. 2010;277: 661–671. pmid:19864280
- 12. Willis KJ, Rudner E, Sümegi P. The full-glacial forests of central and southeastern Europe. Quatern Res. 2000;53: 203–213.
- 13. Stewart JR, Lister AM. Cryptic northern refugia and the origins of the modern biota. Trends Ecol Evol. 2001;16: 608–613.
- 14. Ståhlberg D, Hedrén M. Evolutionary history of the Dactylorhiza maculata polyploid complex (Orchidaceae). Biol J Linn Soc. 2010;101: 503–525.
- 15. Heslop-Harrison J. A synopsis of the Dactylorochids of the British Isles. Ber Geobot Forschlnst Rübel. 1954;1953: 53–82.
- 16. Averyanov LV. A review of the genus Dactylorhiza. In: Arditti J, editor. Orchid Biology: reviews and perspectives, vol. V. Portland, Oregon: Timber Press; 1990. pp. 159–206.
- 17. Delforge P. Guide des Orchidées d’Europe, d’Afrique du Nord et du Proche-Orient, 3rd ed. Neuchâtel-Paris: Delachaux et Niestlé; 2005.
- 18. Hedrén M. Genetic differentiation, polyploidization and hybridization in northern European Dactylorhiza (Orchidaceae): evidence from allozyme markers. Plant Syst Evol. 1996;201: 31–55.
- 19. Hedrén M. Notes on the esterase variation in Swedish Dactylorhiza incarnata s.l. (Orchidaceae). Nordic J Bot. 1996;16: 253–256.
- 20. Hedrén M. The allotetraploid nature of Dactylorhiza praetermissa (Druce) Soó (Orchidaceae) confirmed. Watsonia. 1996;21: 113–118.
- 21. Hedrén M. Conservation priorities in Dactylorhiza, a taxonomically complex genus. Lindleyana. 2001;16: 17–25.
- 22. Hedrén M. Systematics of the Dactylorhiza euxina/incarnata/maculata polyploid complex (Orchidaceae) in Turkey: evidence from allozyme data. Plant Syst Evol. 2001;229: 23–44.
- 23. Hedrén M. Speciation patterns in the Dactylorhiza incarnata/maculata polyploid complex (Orchidaceae): evidence from molecular markers. Journal Europäischer Orchideen. 2002;34: 707–731.
- 24. Hedrén M. Plastid DNA variation in the Dactylorhiza incarnata/maculata polyploid complex and origin of allotetraploid D. sphagnicola (Orchidaceae). Mol Ecol. 2003;12: 2669–2680. pmid:12969470
- 25. Hedrén M, Fay MF, Chase MW. Amplified fragment length polymorphisms (AFLP) reveal details of polyploid evolution in Dactylorhiza (Orchidaceae). Am J Bot. 2001;88: 1868–1880. pmid:21669620
- 26. Devos N, Raspé O, Oh S-H, Tyteca D, Jacquemart A-L. The evolution of Dactylorhiza (Orchidaceae) allotetraploid complex: insights from nrDNA sequences and cpDNA PCR-RFLP data. Mol Phylogenet Evol. 2006;38: 767–778. pmid:16439164
- 27. Devos N, Tyteca D, Raspé O, Wesselingh RA, Jacquemart A-L. Patterns of chloroplast diversity among western European Dactylorhiza species (Orchidaceae). Plant Syst Evol. 2003;243: 85–97.
- 28. Pillon Y, Fay MF, Hedrén M, Bateman RM, Devey DS, Shipunov AB, et al. Evolution and temporal diversification of western European polyploid species complexes in Dactylorhiza (Orchidaceae). Taxon. 2007;56: 1185–1208.
- 29. Ståhlberg D, Hedrén M. Systematics and phylogeography of the Dactylorhiza maculata complex (Orchidaceae) in Scandinavia: insights from cytological, morphological and molecular data. Plant Syst Evol. 2008;273: 107–132.
- 30. Nordström S, Hedrén M. Genetic differentiation and postglacial migration of the Dactylorhiza majalis ssp. traunsteineri/lapponica complex into Fennoscandia. Plant Syst Evol. 2008;276: 73–87.
- 31. Hedrén M, Nordström S, Persson H, Pedersen HÆ, Hansson S. Patterns of polyploid evolution in Greek Dactylorhiza (Orchidaceae) as revealed by allozymes, AFLPs and plastid DNA data. Am J Bot. 2007;94: 1205–1218. pmid:21636487
- 32. Delforge P. Orchids of Britain and Europe. London: Harper Collins; 1995.
- 33. Pridgeon AM, Cribb PJ, Chase MW, Rasmussen FN. Genera Orchidacearum, vol. 2. Orchidoideae, part 1. Oxford: Oxford University Press; 2001.
- 34. Baumann H, Künkele S, Lorenz R. Orchideen Europas Mit angrenzenden Gebieten. Stuttgart, Germany: Eugen Ulmer KG; 2006.
- 35. Pedersen HÆ. Dactylorhiza majalis s.l. (Orchidaceae) in acid habitats: variation patterns, taxonomy and evolution. Nordic J Bot. 2004;22: 641–658.
- 36. Hedrén M, Nordström S, Ståhlberg D. Polyploid evolution and plastid DNA variation in the Dactylorhiza incarnata/maculata complex (Orchidaceae) in Scandinavia. Mol Ecol. 2008;17: 5075–5091. pmid:19120990
- 37. Abbott RJ, Brochmann C. History and evolution of the arctic flora: in the footsteps of Eric Hultén. Mol Ecol. 2003;12: 299–313. pmid:12535083
- 38. Guisan A, Thuiller W. Predicting species distribution: offering more than simple habitat models. Ecol Lett. 2005;8: 993–1009.
- 39. Warren DL, Glor RE, Turelli M. Environmental niche equivalency versus conservatism: quantitative approaches to niche evolution. Evolution. 2008;62: 2868–2883. pmid:18752605
- 40. Busby JR. A biogeographical analysis of Nothofagus cunninghamii (Hook.) Oerst. in southeastern Australia. Aust J Ecol. 1986;11: 1–7.
- 41. Nix H. A biogeographic analysis of Australian elapid snakes. In: Longmore R, editor, Atlas of Elapid Snakes of Australia. Canberra: Australian Government Publishing Service; 1986. pp. 4–15.
- 42. Carpenter G, Gillison AN, Winter J. DOMAIN: a flexible modeling procedure for mapping potential distributions of plants, animals. Biodivers Conserv. 1993;2: 667–680.
- 43. McCullagh P, Nelder JA. Generalized Linear Models. London: Chapman & Hall; 1989.
- 44. Phillips SJ, Dudík M, Schapire RE. A maximum entropy approach to species distribution modeling. In: ICML ′04 Proceedings of the twenty-first international conference on Machine learning. New York: ACM; 2004. pp. 655–662.
- 45. Hernandez PA, Graham CH, Master LL, Albert DL. The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography. 2006;29: 773–785.
- 46. Guisan A, Graham CH, Elith J, Huettmann F. Sensitivity of predictive species distribution models to change in grain size. Divers Distrib. 2007;13: 332–340.
- 47. Wisz MS, Hijmans RJ, Li J, Peterson AT, Graham CH, Guisan A, et al. Effects of sample size on the performance of species distribution models. Divers Distrib. 2008;14: 763–773.
- 48. Elith J, Graham CH. Do they? How do they? Why do they differ? On finding reasons for differing performances of species distribution models. Ecography. 2009;32: 66–77.
- 49. Varela S, Lobo JM, Hortal J. Using species distribution models in paleobiogeography: a matter of data, predictors and concepts. Palaeogeogr Palaeoclim Palaeoecol. 2011;310: 451–463.
- 50. Tóth JP, Varga K, Végvári Z, Varga Z. Distribution of the Eastern knapweed fritillary (Melitaea ornata Christoph, 1893) (Lepidoptera: Nymphalidae): past, present and future. J Insect Conserv. 2012;17: 245–255.
- 51. Kolanowska M. Determination of potential glacial refugia and possible migration routes of Campylocentrum (Vandeae, Orchidaceae) species through the Darién Gap. Acta Soc Bot Pol. 2015;84(1): 97–102.
- 52. Wiens JA, Stralberg D, Jongsomjit D, Howell CA, Snyder MA. Niches, models, and climate change: assessing the assumptions and uncertainties. Proc Natl Acad Sci USA. 2009;106(Suppl 2): 19729–36. pmid:19822750
- 53. Khanum R, Mumtaz AS, Kumar S. Predicting impacts of climate change on medicinal asclepiads of Pakistan using Maxent modeling. Acta Oecol. 2013;49: 23–31.
- 54. Wang W, Tang X, Zhu Q, Pan K, Hu Q, He M, et al. Predicting the Impacts of Climate Change on the Potential Distribution of Major Native Non-Food Bioenergy Plants in China. PLoS ONE. 2014;9: e111587. pmid:25365425
- 55. Velásquez-Tibatá J, Salaman P, Graham CH. Effects of climate change on species distribution, community structure, and conservation of birds in protected areas in Colombia. Reg Environ Change. 2013;13: 235–248.
- 56. McIntyre PJ. Polyploidy associated with altered and broader ecological niches in the Claytonia perfoliata (Portulacaceae) species complex. Am J Bot. 2012;99: 655–662. pmid:22434773
- 57. Laport RG, Hatem L, Minckley RL, Ramsey J. Ecological niche modeling implicates climatic adaptation, competitive exclusion, and niche conservatism among Larrea tridentata cytotypes in North American deserts. J Torrey Bot Soc. 2013;140(3): 349–363.
- 58. Glennon KL, Rissler LJ, Church SA. Ecogeographic isolation: A reproductive barrier between species and between cytotypes in Houstonia (Rubiaceae). Evol Ecol. 2012;26: 909–926.
- 59. Godsoe W, Larson MA, Glennon KL, Segraves KA. Polyploidization in Heuchera cylindrical (Saxifragaceae) did not result in a shift in climatic requirements. Am J Bot. 2013;100: 496–508. pmid:23400493
- 60. Peterson AT, Soberón J, Sanchez-Cordero VV. Conservatism of ecological niches in evolutionary time. Science. 1999;285: 1265–1267. pmid:10455053
- 61. Phillips SJ, Anderson R, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Model. 2006;190: 231–259.
- 62. Elith J, Phillips SJ, Hastie T, Dudík M, Chee YE, Yates CJ. A statistical explanation of MaxEnt for ecologists. Divers Distrib. 2011;17: 43–57.
- 63. Pedersen HÆ. Species concept and guidelines for infraspecific taxonomic ranking in Dactylorhiza (Orchidaceae). Nordic J Bot. 1998;18: 289–309.
- 64. Pedersen HÆ. Allozyme variation and genetic integrity of Dactylorhiza incarnata (Orchidaceae). Nordic J Bot. 1998;18: 15–21.
- 65. Pedersen HÆ. Late-flowering dune populations of Dactylorhiza incarnata (Orchidaceae): variation patterns and taxonomic inferences. Nordic J Bot. 2001;21: 177–186.
- 66. Hedrén M, Nordström S. Polymorphic populations of Dactylorhiza incarnata s.l. (Orchidaceae) on the Baltic island on Gotland: morphology, habitat preference and genetic differentiation. Ann Bot. 2009;104: 527–542. pmid:19458026
- 67. Hedrén M, Nordström S, Bateman RM. Plastid and nuclear DNA marker data support the recognition of four tetraploid marsh orchids (Dactylorhiza majalis s.l., Orchidaceae) in Britain and Ireland, but require their recircumscription. Biol J Linn Soc. 2011;104: 107–128.
- 68. Hedrén M, Nordström S, Pedersen HÆ, Ståhlberg D. Systematics and conservation genetics of Dactylorhiza majalis ssp. elatior (Orchidaceae) on Gotland. Nordic J Bot. 2012;30: 257–272.
- 69. Naczk AM, Górniak M, Szlachetko DL, Ziętara MS. Plastid DNA haplotype diversity and morphological variation in the Dactylorhiza incarnata/maculata complex (Orchidaceae) in northern Poland. Bot J Linn Soc. 2015;178: 121–137.
- 70. Elith J, Graham CH, Anderson RP, Dudík M, Ferrier S, Guisan A, et al. Novel methods improve prediction of species’ distributions from occurrence data. Ecography. 2006;29: 129–151.
- 71. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int J Climatol. 2005;25: 1965–1978.
- 72. Urbina-Cardona JN, Loyola RD. Applying niche-based models to predict endangered-hylid potential distributions: are neotropical protected areas effective enough? Trop Conserv Sci. 2008;1: 417–445.
- 73. Braconnot P, Otto-Bliesner B, Harrison S, Joussaume S, Peterschmitt J-Y, Abe-Ouchi A, et al. Results of PMIP2 coupled simulations of the Mid-Holocene and Last Glacial Maximum—Part 1: experiments and large-scale features. Clim Past. 2007;3: 261–277.
- 74. Taylor KE, Stouffer RJ, Meehl GA. An Overview of CMIP5 and the experiment design. Bull Amer Meteor Soc. 2012;93: 485–498.
- 75. Kim SJ, Flato G, Boer G. A coupled climate model simulation of the last glacial maximum, Part 2: Approach to equilibrium. Clim Dynam. 2003;20: 635–661.
- 76. Coops NC, Waring RH. Estimating the vulnerability of fifteen tree species under changing climate in Northwest North America. Ecol Model. 2011;222: 2119–2129.
- 77. Ferro VG, Lemes P, Melo AS, Loyola R. The Reduced Effectiveness of Protected Areas under Climate Change Threatens Atlantic Forest Tiger Moths. PLoS ONE. 2014;9: e107792. pmid:25229422
- 78. Barredo JI, Strona G, de Rigo D, Caudullo G, Stancanelli G, San-Miguel-Ayanz J. Assessing the potential distribution of insect pests: case studies on large pine weevil (Hylobius abietis L.) and horse-chestnut leaf miner (Cameraria ohridella) under present and future climate conditions in European forests. Bull EPPO. 2015;45: 273–281.
- 79. Hultén E, Fries M. Atlas of North European vascular plants, vols 1–3. Königstein: Koeltz Scientific Books; 1986.
- 80. Nordström S, Hedrén M. Genetic diversity and differentiation of allopolyploid Dactylorhiza (Orchidaceae) with particular focus on the Dactylorhiza majalis ssp. traunsteineri/lapponica complex. Biol J Linn Soc. 2009;97: 52–67.
- 81. Nordström S, Hedrén M. Evolution, phylogeography and taxonomy of allopolyploid Dactylorhiza (Orchidaceae) and its implications for conservation. Nordic J Bot. 2009;27: 548–556.
- 82. Schönswetter P, Stehlik I, Holderegger R, Tribsch A. Molecular evidence for glacial refugia of mountain plants in the European Alps. Mol Ecol. 2005;14: 3547–3555. pmid:16156822
- 83. Pedersen HÆ. Apochromic populations of Dactylorhiza incarnata s.l. (Orchidaceae): diversity and systematic significance as revealed by allozyme markers and morphology. Bot J Linn Soc. 2009;159: 396–407.
- 84. Pridgeon AM, Bateman RM, Cox AV, Hapeman JR, Chase MW. Phylogenetics of subtribe Orchidinae (Orchidoideae, Orchidaceae) based on nuclear ITS sequences. 1. Intergeneric relationships and polyphyly of Orchis sensu lato. Lindleyana. 1997;12: 89–109.
- 85. Bateman RM, Hollingsworth PM, Preston J, Luo Y-B, Pridgeon AM, Chase MW. Molecular phylogenetics and evolution of Orchidinae and select Habenariinae (Orchidaceae). Bot J Linn Soc. 2003;142: 1–40.
- 86. Shipunov AB, Fay MF, Pillon Y, Bateman RM, Chase MW. Dactylorhiza (Orchidaceae) in European Russia: combined molecular and morphological analysis. American J Bot. 2004;91: 1419–1426.
- 87. Hedrén M. Plastid DNA haplotype variation in Dactylorhiza incarnata (Orchidaceae): evidence for multiple independent colonization. Nordic J Bot. 2009;27: 69–80.
- 88. Heslop-Harrison J. A comparison of some Swedish and British forms of Orchis maculata L. sens. lat. Svensk Bot Tidskr. 1951;45: 608–635.
- 89. Vöth W. Biometrische Untersuchungen an Dactylorhiza maculata s.l.—Sippen in Niederösterreich. Linz Biol Beitr. 1978;10: 179–215.
- 90. Reinhard HR. Skandinavische und alpine Dactylorhiza-Arten (Orchidaceae). Mitt Bl Arbeitskr Heim Orch Baden-Württ. 1985;17: 321–416.
- 91. Jagiełło M, Lankosz-Mróz M. Cytotaxonomic studies in the Dactylorhiza maculata (L.) Soó group in Poland (Orchidaceae). Fragm Flor Geobot. 1986–1987;31–32: 385–394.
- 92. Bateman RM, Denholm I. A reappraisal of the British and Irish dactylorchids, 3. The spotted-orchids. Watsonia. 1988;17: 319–349.
- 93. Dufrêne M, Gathoye J-L, Tyteca D. Biostatistical studies on western European Dactylorhiza (Orchidaceae)—the D. maculata group. Plant Syst Evol. 1991;175: 55–72.
- 94. Tyteca D, Gathoye J-L. Morphometric analysis of the Dactylorhiza maculata (L.) Soó group in western Europe. Ber Arbeitskr Heim Orchid. 2004;21: 4–35.
- 95. Devos N, Oh S-H, Raspé O, Jacquemart A-L, Manos PS. Nuclear ribosomal DNA sequence variation and evolution of spotted marh-orchids (Dactylorhiza maculata group). Mol Phylogenet Evol. 2005;36: 568–580. pmid:15939630
- 96. Petit RJ, Aguinagalde I, de Beaulieu J-L, Bittkau C, Brewer S, Cheddadi R, et al. Glacial refugia: hotspots but not melting pots of genetic diversity. Science. 2003;300: 1563–1565. pmid:12791991
- 97. Hopkins DM, Matthews JV, Schweger CE Jr., Young SB. Paleoecology of Beringia. New York: Academic Press; 1982.
- 98. Clague JJ, James TS. History and isostatic effects of the last ice sheet in southern British Columbia. Quaternary Sci Rev. 2002;21: 71–87.
- 99. Boex J, Fogwill C, Harrison S, Glasser NF, Hein A, Schnabel C, et al. Rapid thinning of the late Pleistocene Patagonian Ice Sheet followed migration of the Southern Westerlies. Nature Sci Rep. 2013;3: 2118.
- 100. Hewitt GM. The genetic legacy of the Quaternary ice ages. Nature. 2000;405: 907–913. pmid:10879524
- 101. Weiss S, Ferrand N. Phylogeography of Southern European Refugia: Evolutionary Perspectives on the Origins and Conservation of European Biodiversity. Dordrecht: Springer; 2007.
- 102. Feliner GN. Southern European glacial refugia: a tale of tales. Taxon. 2011;60: 365–372.
- 103. Gómez A, Lunt DH. Refugia within refugia: patterns of phylogeographic concordance in the Iberian Peninsula. In: Weiss S, Ferrand N, editors. Phylogeography of Southern European Refugia: Evolutionary Perspectives on the Origins and Conservation of European Biodiversity. Dordrecht: Springer; 2007. pp. 155–188.
- 104. Schmitt T, Varga Z. Extra-Mediterranean refugia: the rule and not the exception? Front Zool. 2012;9: 22. pmid:22953783
- 105. Elenga H, Peyron O, Bonnefille R, Jolly D, Cheddadi R, Guiot J, et al. Pollen-based biome reconstruction for southern Europe and Africa 18,000 yr BP. J Biogeogr. 2000;27: 621–634.
- 106. Tzedakis PC, Lawson IT, Frogley MR, Hewitt GM, Preece RC. Buffered tree population changes in a Quaternary refugium: evolutionary implications. Science. 2002;297: 2044–2047. pmid:12242441
- 107. Médail F, Diadema K. Glacial refugia influence plant diversity patterns in the Mediterranean Basin. J Biogeogr. 2009;36: 1333–1345.
- 108. Hewitt GM. Genetic consequences of climatic oscillations in the Quaternary. Philos Trans R Soc Lond B Biol Sci. 2004;359: 183–195. pmid:15101575
- 109. Weisrock DW, Janzen FJ. Comparative molecular phylogeography of North American softshell turtles (Apalone): implications for regional and wide-scale historical evolutionary forces. Mol Phyl Evol. 2000;14: 152–164.
- 110. Willerslev E, Cappellini E, Boomsma W, Nielsen R, Hebsgaard MB, Brand TB, et al. Ancient biomolecules from deep ice cores reveal a forested Southern Greenland. Science. 2007;317: 111–114. pmid:17615355
- 111. Fourcade Y, Engler JO, Rödder D, Secondi J. Mapping Species Distributions with MAXENT Using a Geographically Biased Sample of Presence Data: A Performance Assessment of Methods for Correcting Sampling Bias. PLoS ONE. 2014;9: e97122. pmid:24818607
- 112. Waltari E, Hijmans RJ, Peterson AT, Nyári Á, Perkins SL, Guralnick RP. Locating Pleistocene refugia: comparing phylogeographic and ecological niche model predictions. PLoS ONE. 2007;2: e563. pmid:17622339
- 113. Tremblay RL. Trends in the pollination ecology of the Orchidaceae: evolution and systematics. Can J Botany. 1992;70: 642–650.
- 114. Lammi A, Kuitunen M. Deceptive Pollination of Dactylorhiza incarnata: An Experimental Test of the Magnet Species Hypothesis. Oecologia. 1995;101: 500–503.
- 115. Lammi A, Vallius E, Vauhkonen T, Kuitunen M. Outcrossing, hybridization, pollen quantity, and the evolution of deceptive pollination in Dactylorhiza incarnata. Ann Bot Fenn. 2003;40: 331–338.
- 116. Cozzolino S, Schiestl FP, Müller A, De Castro O, Nardella AM, Widmer A. Evidence for pollinator sharing in Mediterranean nectar-mimic orchids: absence of premating barriers? Proc Biol Sci. 2005;272: 1271–1278. pmid:16024392
- 117. Hadley G. Non-specificity of symbiotic infection in orchid mycorrhiza. New Phytol. 1970;69: 1015–1023.
- 118. Pellegrino G, Bellusci F. 2009. Molecular identification of mycorrhizal fungi in Dactylorhiza sambucina (Orchidaceae). Biologia. 2009;64: 893–897.
- 119. Wotavová K, Balounová Z, Kindlmann P. Factors affecting persistence of terrestrial orchids in wet meadows and implications for their conservation in a changing agricultural landscape. Biol Cons. 2004;118: 271–279.
- 120. Janečková P, Wotavová K, Schödelbauerová I, Jersáková J, Kindlmann P. Relative effects of management and environmental conditions on performance and survival of populations of a terrestrial orchid, Dactylorhiza majalis. Biol Cons. 2006;129: 40–49.
- 121. Pillon Y, Fay MF, Shipunov AB, Chase MW. Species diversity versus phylogenetic diversity: a practical study in the taxonomically difficult genus Dactylorhiza (Orchidaceae). Biol Cons. 2006;129: 4–13.
- 122. Lewis WH. Polyploidy in species populations. In: Lewis WH, editor. Polyploidy. Biological relevance. New York: Plenum Press; 1980. pp. 103–144.
- 123. Grant V. Plant speciation, 2nd ed. New York: Columbia University Press; 1981.
- 124. Szalchetko DL, Skakuj M. Storczyki Polski. Poznań: Wydawnictwo Sorus; 1996.
- 125. Minasiewicz J, Tukałło P, Trzepanowska K. Dactylorhiza incarnata (L.) Soó in the Gdańsk region—morphological and genetic variation of populations and the present state of the localities. Acta Bot Cassub. 2004;4: 139–160.
- 126. Kolanowska M. Niche Conservatism and the Future Potential Range of Epipactis helleborine (Orchidaceae). PLoS ONE. 2013;8: e77352. pmid:24143222
- 127. Adamowski W. Amerykańska kariera europejskiego storczyka. Wiadomości Botaniczne. 1995;39: 105–113.
- 128. Songer M, Delion M, Biggs A, Huang Q. Modeling Impacts of Climate Change on Giant Panda Habitat. International Journal of Ecology. 2012;2012: 108752.
- 129. Freeman LA, Kleypas JA, Miller AJ. Coral Reef Habitat Response to Climate Change Scenarios. PLoS ONE. 2013;8: e82404. pmid:24340025
- 130. Gormley KSG, Porter JS, Bell MC, Hull AD, Sanderson WG. Predictive Habitat Modelling as a Tool to Assess the Change in Distribution and Extent of an OSPAR Priority Habitat under an Increased Ocean Temperature Scenario: Consequences for Marine Protected Area Networks and Management. PLoS ONE. 2013;8: e68263. pmid:23894298
- 131. Cribb PJ, Kell SP, Dixon KW, Barrett RL. Orchid conservation: A global perspective. In: Dixon KW, Kell SP, Barrett RL, Cribb PJ, editors. Orchid Conservation. Kota Kinabalu: Natural History Publications; 2003. pp. 1–24.
- 132. Bernardos S, Tyteca D, Garcia-Barriuso M, Crespi A, Castro A, Amich F. Current status and conservation of the Lusitan Duriensean Orchids (Duero Basin, NE Portugal and CW Spain). Acta Bot Gallica. 2006;153: 273–284.
- 133. Swarts ND, Dixon KW. Terrestrial orchid conservation in the age of extinction. Ann Bot. 2009;104: 543–556. pmid:19218582