Status and Distribution of Four Endemic Vascular Plants in the Gobi Altai
Karsten Wesche1, Eckehart J. Jäger1, Henrik von Wehrden1, Ragchaa Undrakh2
1Institute of Geobotany and Botanical Garden, Martin-Luther-University Halle-Wittenberg, 06099 Halle, Germany
2Faculty of Biology, National University of Mongolia, P.O. Box 377, Ulaanbaatar, Mongolia
The paper presents distributional data on the four vascular plants Papaver saichanense, Saussurea saichanensis, Potentilla ikonnikovii and Galitzkya macrocarpa, all of which are restricted to Mongolian mountains. Updated biogeographical data demonstrate that all four are Mongolian endemics. In terms of their taxonomic relationships, S. saichanensis and P. saichanense belong to a group of species occurring mainly on continental Asian mountains. Potentilla ikonnikovii has relatives with a mainly East-Asian distribution, and the genus Galitzkya is a predominantly Mid-Asian element. New maps of the local distribution in the Gobi Altai and adjacent mountains indicate that all species are highly fragmented and are so far only known to occur in less than a dozen localities. We have since discovered new sites and subsequently have little reason to regard the species as threatened, although the overall rarity suggests that some form of rough monitoring is advisable.
Endemism is a central issue in nature conservation and governments are often held responsible for the protection of those species that have the entire or at least major parts of their geographical range on the country’s territory (Bonn et al., 2002). Knowledge on the overall distribution of any given species is of straightforward importance in this respect (Jдger & Hoffmann, 1997). Responsibility measures have been implemented in schemes for setting priorities in species’ conservation; prominent examples are the designation of so-called biodiversity hotspots (Myers et al., 2000), or the Flora and Fauna Habitat Directive of the European Union, which assigns high priority to the conservation of European endemics and subendemics. However, numbers of restricted-range species are relatively low for most European countries (e.g. Germany; Welk, 2002) apart from the Mediterranean region. The situation in Mongolia is comparable in this respect. Although the country covers a vast territory, the majority of the occurring species have even larger distributional ranges and are shared with neighbouring countries. The flora of Mongolia is divided along two principal gradients. The most obvious differences are between the northern parts - which belong to the central Siberian-Daurian floristic region, and the dry Gobi - which is part of the Central Asian floristic region (Meusel & Jдger, 1992). A somewhat less obvious, but equally important, distinction occurs between the western and eastern parts of Mongolia. The latter are characterized by predominant summer rains, lack of snow cover in winter, and relatively fast temperature changes in spring and autumn. The division line runs along a longitudinal line of 100°E (Hilbig et al., 2004); east of this limit species mainly have eastern and Central Asian distributions, whereas west of this line many species comprise Mid-Asian or even Mediterranean elements. Most of Mongolia’s endemic and subendemic species are Central Asian elements (Grubov, 1989). Numbers of endemics listed differ between sources: a conservative estimate indicates 4% truly endemic, i.e. exclusively Mongolian vascular plants; if subendemics are included the number increases to some 10% of the flora (Grubov, 1989). Here, “subendemic” refers to Mongolian species with restricted ranges but which are also found in the neighbouring countries. A survey by Ulziikhutag (1989) lists 5.9 % of the overall flora of Mongolia as endemic, and a further 8.1% as subendemic. Here, several species known to occur in the neighbouring parts of China are also listed as endemic, explaining at least part of the difference. Examples of subendemic species with a Central Asian distribution and with occurrences in the Gobi parts of Mongolia and China include Psammochloa villosa and Hedysarum fruticosum (Hilbig et al., 1999). Central Asian species with their entire distribution range in Mongolia, i.e. country endemics in a strict sense, include Galitzkya macrocarpa and Potentilla ikonnikovii (Grubov, 1989). Mongolia clearly has a high level of responsibility for conservation of these species, as well as for those subendemics that have major parts of their distribution range on Mongolian territory. Local abundance is a second important aspect of a given species’ distributional status. There is a general tendency for restricted-range species to show low levels of abundances even in their distributional centre (e.g. Bonn et al., 2002). Thus, rarity is a second important factor in conservation assessments (cf. IUCN, 2001), as small populations are widely held to face increasing risks of extinction (“small populations paradigm”, see e.g. Channell & Lomolino, 2000). The mountain ranges of the south-eastern Gobi Altai form archipelago-like islands of relatively moist conditions in an otherwise dry matrix of steppe and semi-desert environments (Gunin & Vostokova, 1995). Plant species of the upper mountain ranges have highly isolated populations and, not surprisingly, several endemic species have evolved in the dry mountains of Mongolia. The following is an assessment of the distribution status of the four vascular plant species that have their centre of distribution in southern Mongolian high mountains and are considered endemic by Ulziikhutag (1989): Papaver saichanense Grub.; Saussurea saichanensis Kom. ex Lipsch.; Potentilla ikonnikovii Juz.; Galitzkya macrocarpa (Ik.-Gal.) V. Bocz. First of all, we present an analysis of the overall geographical distribution range based on an updated compilation of available literature, thereby assessing Mongolia’s level of responsibility. Secondly, we present new data on the species’ local distribution in southern Mongolia. In combination, these data allow us to give some recommendations on their conservation statuses.
Material and Methods
The number of flora references in the neighbouring territories has increased considerably in the last few years, so we had to consult 32 different sources to assemble the biogeographical data. A full reference list thereof has not been included here, but the information is available from us on request. Floristic data were used to compile the maps in Figs. 1-3. The maps also give the ranges of the most closely related species in order to demonstrate biogeographical affinities. The distribution of species in Mongolia can be inferred from the recent account given by Gubanov (1996), and detailed information provided by e.g. Grubov (2001). The authors have divided the country into 16 sub-regions, which are used to give a general impression of the local distribution in Mongolia (see inserts in Fig. 4). Based on the available information on species distribution, fieldwork was restricted to southern Mongolia, and covered parts of the Mongolian Altai, the Gobi Altai and most mountain ranges south of the Altai system. The focus lay on the four major protected areas in the region. We visited all mountains >2000 m asl. in the Gobi Gurvan Saikhan National Park, but also sampled adjacent parts of the Gobi Altai (Arts Bogd, Ikh Bogd, Noyon Uul). Additionally, we checked several mountains in the Great Gobi A (Atas Bogd, Tsagaan Bogd, Endrengiin Nuruu) and Great Gobi B Special Protected Areas (Khavtagiin Nuruu and adjacent hills). The Little Gobi Special Protected Areas A & B were also surveyed but they lack high mountains, and thus suitable habitats. All species are restricted to the upper parts of mountain ranges (Table 1) with their relatively high rainfall. Species occur in shallow soil on steep slopes and on rock outcrops where they face limited competition from other plants and are often protected from intensive livestock grazing. Access to sites is difficult and sometimes dangerous with respect to the often highly weathered rock. Thus, we were unable to provide census data on population sizes. Detailed maps were compiled based on SRTM topographical data (“shuttle radar topographical mission”, downloaded from the Global Landcover Facility at http://glcf.umiacs.umd.edu/index.shtml) in a WGS 84 projection. Two sets of maps were compiled; for each species we provide a large-scale map of the overall geographic distribution and a small-scale map giving the local distribution within our sampling area. Maps were compiled using ArcGIS 8.2.
All four species belong, at least with respect to the distribution of their taxonomical relatives, to completely different biogeographical elements (Figs. 1-3). The flora of the study region is characterised by species with typically central to eastern Asian, continental distribution ranges. A similar general pattern is also found in Papaver saichanense. Unfortunately, its taxonomical status is not yet understood. Papaver species in the Mongolian mountains belong to a group of taxa around Papaver nudicaule (sect. Meconella Spach., syn. Scapiflorae Rchb. ex Elkan), but the exact relationships within that group remain unknown. Morphological differences among taxa are not conclusive (cf. Grubov, 1982), and chromosome numbers in this polyploidic complex have not yet been extensively studied. The specialist P. Hanelt (Gatersleben, Germany) tentatively named most of our specimens from the Gobi Gurvan Saikhan as P. saichanense Grub., which is an endemic taxon of the south-eastern Gobi Altai. However, P. pseudocanescens M. Pop., P. nudicaule L. and P. pseudotenellum Grub. also occur in the area (P. Hanelt, pers. comm.), as does Papaver rubroaurantiacum (DC) Fisch. ex Steud. (Grubov, 1982). All these taxa are closely related and species’ limits are unclear. Thus, we refrained from drawing a distribution map. The Papaver nudicaule group has a Turkestanian – Himalayan – western Chinese – Siberian distribution. The westernmost outposts are found in the eastern Caucasus and in western Afghanistan. There are also remote outposts in continental south-eastern Alaska, which must have migrated there when land bridges were formed during the Quaternary. The distributional centre of the genus Saussurea (>20 species) shows a rather similar spatial extent, including the presence of outposts in SE-Alaska (four species). Saussurea glanduligera is the species most closely related to S. saichanensis (Lipshits, 1979); both form the series Glanduligera (Fig. 1).
Saussurea glanduligera is widespread in Central Asian highlands. In contrast, S. saichanensis is endemic to Mongolia with somewhat scattered occurrences from the far western to northern parts of the country; in the south, the species occurs from the western Mongolian Altai to the Gobi Gurvan Saikhan. So far, it has not been found in any neighbouring country. The distributional range of Potentilla ikonnikovii and related species is much more extensive compared to the previously described (east-) central Asian mountain distributions. The section Tanacetifoliae covers some 18, mainly typical eastern Asian, species (Fig. 2). The group of species is widely distributed in continental central and eastern Asia; outposts in regions with summer precipitation occur in eastern Afghanistan, Armenia, westwards to the Caucasus, with some glacial relics surviving in the calcareous outcrops around Voronesh and Kursk, in the western part of the Balkan Peninsula and northwards to Jakutsk and the Kolyma region. The closest relative to P. ikonnikovii is P. kryloviana (Juzepchuk, 1955). The latter occurs in the south-eastern Altai with an outpost in the southern Khangai. Potentilla ikonnikovii is an endemic of the Gobi Altai, but the distribution of its relatives implies that it has reached the Gobi Gurvan Saikhan from the northwest rather than from the south-west (Tien Shan) or the east. Thus, in this respect biogeographical affinities are similar to those described for P. saichanense and S. saichanensis. In contrast, Galitzkya macrocarpa is a western element and belongs to a biogeographical group of species that is very rare in the flora of Mongolia east of 100°E. The genus Galitzkya is concentrated in the continental regions of Mid-Asia (Fig. 3), which receive at least part of their precipitation in winter. This is also implied by the species’ morphology, as Galitzkya spp. have a suffruticose growth form and are semi-evergreen, while the other three species are herbs that bear leaves only in summer and have tap roots or a rhizome (Table 1).
Galitzkya spathulata is among the most widespread and common species in the petrophytic steppes of mountains and hills in central Kazakhstan. Eastwards, the species is replaced by G. potaninii with occurrences in eastern Xinjiang, western Gansu and south-western Mongolia. Galitzkya macrocarpa occupies the easternmost outposts in this continental western Asian genus and is an endemic of the Gobi Altai and adjacent mountain regions in Mongolia.
The taxonomic difficulties render an assessment of the distribution of Papaver saichanense almost impossible. We collected samples that were tentatively named Papaver saichanense (P. Hanelt, pers. comm.) from all three main ranges of the Gobi Gurvan Saikhan and from the Ikh Bogd (2200-2800 asl., Fig. 4).
However, only samples from the Khalgan and from the Zuun Saikhan are clearly identified based on morphological characteristics given in Grubov (1982). In any case, populations of poppies are commonly found in moist mountain steppes of the upper slopes and are often quite large (numbering dozens, to hundreds of flowering individuals on any given slope). Thus, we might conclude that the south-eastern Gobi Altai hosts a distinct taxon within the P. nudicaule group, which occurs on the moister mountain ranges of the Gobi Altai and develops reasonably large populations there. Saussurea saichanensis can also be found in mountain steppes (2200-2800 m asl.), where populations can amount to several dozen to hundreds of specimens. The largest sites were in the Zuun Saikhan and the Dund Saikhan, whereas populations in the dry Nemegt Uul were smaller and less dense. The species is the most widely distributed of the four discussed here, and it is also known to occur in Tolbo Kungay in the Khovd region, Tsengel, Khan Khokhii, Mankhan, Aj Bogd, Taishiryn Uul, Ikh Khavtag Uul in the Dsungarian Gobi, Khasagt Khairkhan (Dod Nur) in the Mongolian Altai and Otgontenger from Khangai. It has also been found near Khovsgol and in parts of Khentii (Grubov, 1982; Gubanov, 1996; Herbarium of the Botanical Institute of the Mongolian Academy of Science, Ulaanbaatar). Potentilla ikonnikovii is easily identified in the field because of its pinnate leaves and its habitat requirements, as the species is restricted to rock fissures and small crevices between 2050 and 2500 m asl. Thus, it occupies similar sites as G. macrocarpa. Potentilla occurs on almost all ranges in the Gobi Gurvan Saikhan, except for the Dund Saikhan where it is apparently absent (Fig. 4). However, a transplantation experiment revealed that the species can grow in the upper Dund
Saikhan, so dispersal limitations may cause its limited distribution. Counting of specimens was not feasible due to the inaccessibility of the habitat, but where present several specimens were usually found in a radius of <10 m. Nonetheless, populations appear to be somewhat smaller and the species is, on the whole, less widely distributed than G. macrocarpa. We saw a single herbarium specimen that apparently originated from the Khadjin Uul in the Mongolian Altai. Galitzkya macrocarpa was commonly found at mountain sites in the region at altitudes ranging between 1950 and 2570 m asl. It has colonised all ranges in the Gobi Gurvan Saikhan NP, and also occurs on several sites in the Arts Bogd. We visited parts of the neighbouring Noyon Uul without finding specimens of the species, but previous records from there seem reasonable, and the Tost Uul may also be colonised. We were unable to find G. macrocarpa in the Endrengiyn Nuruu, and found only G. potaninii in the Trans-Altai Gobi (Atas Bogd). Though we lack quantitative data, we had little difficulty in finding between 5 and 15 individuals of G. macrocarpa within a radius of 10 m at colonized sites, so populations are at least not extremely small. All four species were found to be flowering. Flowers were even produced in the zud (drought) year of 2001 and in dry 2002, but numbers appeared higher in the moist years of 2003 and 2004. Flower production in 2005 was similar to 2002. Thus, seed production is presumably reduced in drought years but seeds were nonetheless produced during the whole observation period. In summary, P. ikonnikovii is known from one district (Gobi-Altai, see inserts Fig. 4), but there may be an outpost in a second (Mongolian Altai). G. macrocarpa occurs in two districts (Gobi-Altai and Trans-Altai), while P. saichanense appears to be restricted to the Gobi-Altai district. S. saichanensis is the most widespread species, with occurrences in seven districts (Khovsgol, Khentii, Khangai, Khovd, Mongolian Altai, Gobi-Altai and Dzungarian Gobi). Recommendations for species’ conservation The distributional data clearly show that all four species are true endemics of Mongolia. Moreover, Galitzkya is a small genus with only three species (Boczantzeva, 1979), and Mongolia takes part of the range of G. potaninii. The survival of this species may be also questionable if Mongolian populations became extinct. Thus, from a biogeographical point of view, Mongolia has full responsibility for their survival, and should be highly concerned about all four species. None of the species are listed in the Mongolian Red Data book (Shiirevdamba et al., 1997), and they are similarly absent from IUCN’s International Red List. In fact, the latter does not include a single vascular plant from the Mongolian flora (http:// www.redlist.org). In order to be listed as “endangered”, restricted-range species must have an overall geographical range smaller than 5000 km², or a so-called “area of occupancy” (i.e. actual habitat) of <500 km² (IUCN, 2001). To be listed as “vulnerable”, the species’ geographical range must be <20,000 km², or the species must not occupy >2000 km². GIS-based analysis of potential habitats suggests that P. saichanense, P. ikonnikovii and G. macrocarpa definitely qualify as vulnerable by the latter criterion. In addition, S. saichanensis, whose overall range is much greater than that required for the “vulnerable” categorisation, has a scattered distribution which suggests that area of occupancy could be smaller than 2000 km². Further research would be required to confirm as such. A second criterion is to be met for inclusion in the Red List. In context of our study, evidence of shrinking populations or loss of habitats would be the most obvious second problem, but our survey in the Gobi Altai region gives no evidence of either. We have newly confirmed most of the populations previously described in the literature and have also added new records. Most of the suitable sites identified are apparently colonised, and populations are at least not extremely small. Moreover, more populations may be found in future studies. The previously published floristic data cover the entire vast country and are comparatively detailed. However, our new records hint at some of the still existing gaps. Our survey covered most southern Mongolian mountains over the last 5 years, but we have still certainly not seen every spot of potential habitat. Because all our new records were within or close to the already known distribution ranges of the four endemics, we nonetheless doubt that the general pattern will change much with more data becoming available. Another criterion for listing as vulnerable is described as a species being “severely fragmented or known to exist in no more then 10 locations” (IUCN, 2001). All species are severely fragmented and P. ikonnikovii and P. saichanense are known to occur in fewer than ten populations (Fig. 2, 4). The number of occurrences of S. saichanensis and G. macrocarpa is around ten, so all four species can be considered rare. Knowledge on the local distribution is still insufficient, but all four species could easily become “vulnerable” if a few populations disappear, or if numbers of individuals start to decline. G. macrocarpa and P. ikonnikovii grow in rock fissures and are unlikely to be affected by livestock grazing. S. saichanensis grows on rocks and also in heavily grazed mountain steppes, but as a rosette plant appears to be equally grazing-tolerant as the mountain steppe species P. saichanense. Thus, there is little reason to expect detrimental impact from human land use. Sites in the Gobi Altai are dry and sexual regeneration is apparently limited in some mountain steppe species (e.g. Wesche et al., 2005). Climate change could pose a problem in this respect, although at present there is no evidence that southern Mongolia is getting drier (Jacoby et al., 2000). Temperatures are currently rising, but all species colonise a wide altitudinal and therefore thermal range. Rising temperatures could lead to increased evapotranspiration, but coinciding rising levels of atmospheric CO2 should increase water use efficiency (Christensen et al., 2004), so overall effects of climate change on these species should be negligible. In summary, our data indicated that all four species are Mongolian endemics and add to the biodiversity value of the country. None of them is significantly threatened at the moment, although some monitoring of their overall distribution is advisable with respect to the limited number of available records. However, no immediate dangers are discernible and we subsequently have enough reason to be confident that future generations will be able to take delight in these gemstones of the Mongolian flora.
We thank all members of the Gobi Gurvan Saikhan Research Project for their help during various stages of research and manuscript preparation, most notably Ts. Jamsran and R. Samiya, N. Soninhishig and Ch. Dulamsuren helped with literature reference. P. Hanelt kindly checked our Papaver specimens, and provided a preliminary identification. H. Zech drew the maps in figures 1-3. Daniel McCluskey had the patience to check our English. Suggestions of two anonymous reviewers helped to clarify the ideas presented here.
- Boczantzeva, V. 1979. The new genus Galitzkya V. Boczantzeva (Cruciferae). Botanicheskiy Jurnal 64: 1440-1442. Bonn, A., Rodrigues, A.S.L. & Gaston, K. 2002. Threatened and endemic species: are they good indicators of patterns in the biodiversity on a national scale? Ecol. Lett. 5: 733-741.
- Channell, R. & Lomolino, M.V. 2000. Trajectories to extinction: spatial dynamics of the contraction of geographical ranges. J. Biogeography 27: 169-179.
- Christensen, L., Coughenour, M.B., Ellis, J.E. & Zuo, Z.C. 2004. Vulnerability of the Asian typical steppe to grazing and climatic change. Climatic Change 63: 351-368.
- Grubov, V.I. 1982. Key to the vascular plants of Mongolia (in Russian). Nauka, Leningrad.
- Grubov, V.I. 1989. Endemismus in der Flora der Mongolei. Erforsch. Biol. Ress. MVR. 6: 87- 90.
- Grubov, V.I. 2001. Key to the vascular plants of Mongolia. Volume I & II. Science Publishers, Plymouth.
- Gubanov, I.A. 1996. Conspectus of the Flora of Outer Mongolia (Vascular Plants). Valang Publishers, Moscow.
- Gunin, P.D. & Vostokova, E.A. 1995. Ecosystems of Mongolia. Moscow.
- Hilbig, W., Bastian, O., Jдger, E. & Bujan-Orsich, C. 1999. Die Vegetation des Uvs Nuur Beckens (Uvs Aimak, Nordwest-Mongolei). Fed. Repert. 110: 569-625.
- Hilbig, W., Jдger, E. & Knapp, H.D. 2004. Die Vegetation des Bogd-uul bei Ulaanbaatar (Mongolei) – Standortsbindung und pflanzengeographische Stellung. Fed. Repert. 115: 265-342.
- IUCN 2001. IUCN Red List Categories and Criteria Version 3.1. IUCN. Gland.
- Jacoby, G., D’Arrigo, R., Pederson, N., Buckley, B., Dugarjav, C. & Mijidorj, R. 2000. Temperature and precipitation in Mongolia based on dendroclimatic investigations. Scientific Journal of the Faculty of Biology, National University of Mongolia 11: 5-9.
- Jдger, E.J. & Hoffmann, M.H. 1997. Schutzwьrdigkeit von GefдЯpflanzen aus der Sicht der Gesamtareale. Zeitschrift fьr Цkologie u. Naturschutz 6: 225-232.
- Juzepchuk, S.V. 1955. Desyat novyh lapchatok. Botanicheskye Materialy 17: 213-238.
- Lipshits, S.Y. 1979. Rod Saussurea. Nauka, Leningrad.
- Meusel, H. & Jдger, E. 1992. Vergleichende Chorologie der zentraleuropдischen Flora. Bd. III Text und Kartenband. G. Fischer, Jena, Stuttgart.
- Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B. & Kent, J. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853-858.
- Shiirevdamba, T., Shardarsuren, O. & Erdenjav, G. 1997. Mongolian Red Book. Ministry for Nature and Environment of Mongolia. Ulaanbaatar.
- Ulziikhutag, N. 1989. Mongol ornii urgamliin aimgiin toim (A summary of the flora of Mongolia). Government Publisher, Ulaanbaatar.
- Welk, E. 2002. Arealkundliche Analyse und Bewertung der Schutzrelevanz seltener und gefдhrdeter GefдЯpflanzen Deutschlands. Bundesamt fьr Naturschutz, Bonn.
- Wesche, K., Ronnenberg, K. & Hensen, I. 2005. Lack of sexual reproduction in dry mountain steppe populations of the clonal shrub Juniperus sabina L. in southern Mongolia. J. Arid. Environ. 63: 390-405.