International Association of Vegetation Science

Floristical and Ecological Characterization of the Polar Desert Zone of Author(s): Christian Bay Source: Journal of Vegetation Science, Vol. 8, No. 5 (Nov., 1997), pp. 685-696 Published by: Blackwell Publishing Stable URL: http://www.jstor.org/stable/3237373 . Accessed: 18/02/2011 16:22

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=black. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Blackwell Publishing and International Association of Vegetation Science are collaborating with JSTOR to digitize, preserve and extend access to Journal of Vegetation Science.

http://www.jstor.org Journal of VegetationScience 8: 685-696, 1997 ? IAVS;Opulus Press Uppsala. Printed in Sweden 685

Floristical and ecological characterization of the polar desert zone of Greenland

Bay, Christian

Botanical Museum,University of Copenhagen,Gothersgade 130, DK-1123 CopenhagenK, Denmark; Tel. +45 35 322209; Fax +45 35 322210; E-mail [email protected]

Abstract. compositionand biomass of fourplant com- Most authors assign the northernmost areas of the munitieswere investigatedin two coastal polar desertareas in Holarctic with an impoverished flora and vegetation to eastern North Greenland,bordering the North East Water the polar desert zone (e.g. Aleksandrova 1988; - an ice-free sea area - and Polynya kept open by upwelling Matveyeva 1994; Bliss 1988; Bliss & Svoboda 1984; comparedwith inlandareas in North Greenland.Herb barren, Bliss et al. 1984; Edlund 1980). The concept of arctic the pooresttype, has a species richnessof 6 species/m2,a cover deserts has been discussed Aleksandrova (1988) of 0.7 %, and an above-groundbiomass of 0.6 g/m2 (vascular polar by ).The richesttype, Saxifragaoppositifolia snowbed, has and others (e.g. Bliss et al. 1984; Funder & Abrahamsen 10 species/m2,5.0 % cover, and 11.2 g/m2biomass. 1988). Landscapes of the northernmost coastal parts of A floristic and vegetation boundaryexists a few kilome- the High Arctic are considered Arctic polar deserts; they tres from the coast. The coastal areasbordering the NorthEast are without continuous cover, have a low number WaterPolynya had an impoverishedflora and vegetationcom- of species and lack certain plant groups otherwise char- paredto areasnear the ice-coveredsea, possiblycaused by very acteristic of the circumpolar High Arctic. The total low summer and of clouds. temperatures high frequency cover of vascular plants in this patchy vegetation is A new delimitationof the polar deserts of Greenlandis generally < 2 %, only locally exceeding 5 %, whereas proposedon the basis of the numberof vascularplant species, the cover be as as 75 %. Character- the occurrenceof species with a specific inlanddistribution in cryptogam may high NorthGreenland and the dominatinglife forms. At presentthe istic features of the polar desert zone are the absence of polardesert zone includesonly areaswithin a zone up to ca. 15 woody species which are otherwise widely distributed km from the outercoast of high arcticGreenland - northof ca. in the High Arctic, e.g. Salix spp., Diyas spp. and 80? N. Large areas formerly classified as polar deserts in Cassiope tetragona. In addition to vascular cryptogams, eastern North Greenland,as well as in WashingtonLand in aquatic and salt marsh species and representatives of the western North Greenland,are excluded. New floristic data plant families Asteraceae and Cyperaceae occur. confirm that Greenlandis included in the Canadian correctly The primary ecological factors controlling the im- provinceof the arcticpolar deserts, whereas there is no reason poverishment of flora and vegetation of the coasts around for subdividingthe polar deserts of the Canadianprovince. the Arctic Ocean are low temperatures and a short grow- ing season (Young 1971; Edlund 1983; Elvebakk 1985; Keywords: Arctic; Floristic delimitation; Phytogeography; Funder & Abrahamsen 1988). Aleksandrova (1988) esti- Plant community;Vegetation type. mates the annual radiation balance to be < 10 kcal/cm2. Bliss (1988) suggested that the basic control of polar Nomenclature: Bocher et al. (1978). deserts is the combination of the geological substrate, low nutrientlevels and cryoturbationcausing a poor soil devel- opment and a low moisture retaining capacity during the Introduction short summer. However, Bliss & Matveyeva (1992) later modified this view by stating that a low summer tempera- Botanists have divided the northern circumpolar High ture and a short growing season, with surface soils that dry Arctic into zones based on floristic and vegetation char- out in mid-season, are also important. acteristics. Young (1971), Aleksandrova (1988), Bliss The present study was part of a project focusing on (1993) and Yurtsev (1994) have proposed floristic and the terrestrial ecology of land bordering the open water phytogeographical zones for the entire Arctic. Regional area off the coast of easternmost North Greenland. For studies of Canada (Edlund 1990; Edlund & Alt 1989), comparison an inland area was included in the study. (Komarkova & Webber 1978), Greenland The present paper aims to describe the flora and vegeta- (Bay 1992), Svalbard (Elvebakk 1989) and Russia tion of the polar desert zone of Greenland, present a new (Aleksandrova 1988; Matveyeva 1994) have outlined re- delimitation of the zone and update the delimitation of gional delimitations of these zones. the polar desert zone in the circumpolar Arctic. 686 Bay, C.

vegetation,and climaticdata from a coastal areaborder- ing a mostly ice-covered sea with a coastal area under influence of the polynya. These areas,occurring within the floristic province of North Greenlandand including a coastal zone up to 50 km inland, were classified as polardesert (Bay 1992). Datacollected on HenrikKr0yer Holme, Eskimonasset, Fimdalen, and from the semi- nunatakKilen are also included (Fig. 1). For comparison,available floristic data from a dis- tant inland area within the interiorfloristic district of North Greenlandby CentrumS0, KronprinsChristian Land (80? 10' N) are compiled from Bay (1992), in additionto materialcollected by others.

Climate

Only a few meteorological data sets are available from North Greenlandnorth of 80? N. Of the stations recording daily temperatures,only Station Nord and Kap Moltke have been recordingfor more than 15 yr (Table 1). Very limited informationon precipitationis available and only from the mannedstations with long recordingperiods. Mean summertemperature and mean annualtemperature are given in additionto the precipi- tation.Summer is defined as July andAugust, which are the only months with a mean temperatureabove zero. The meanannual temperature is relatedto latitude,the 1. areas (black dots) in eastern North Greenland Fig. Study exception being Henrik Kr0yer Holme, situated in the with localities mentioned.The new delimitationof the polar polynya, which has a markedly annual desert zone of eastern North Greenland is indicated by the higher tempera- dashed line. ture correspondingto the temperaturerecorded at Dan- markshavn450 km to the Southon the East coast. How- ever, the summertemperature is not correlatedwith lati- Study areas tudebut withthe distanceto the sea; inlandlocalities have a mean temperatureof 3 - 4 ?C, whereas in the coastal Two areasin the coastalpart of easternNorth Green- areasthe valuesare just 0 - 2 ?C (Bay 1992). land (ca. 81? N) south of the ice cap Flade Isblink (Fig. Precipitationis low, showing a significantdifference 1) were investigatedin July and August 1993 duringthe between the outer coast and the inlandlocality, ranging NEWLandproject, which focused on the terrestrialecol- annuallybetween ca. 200 mm in coastalareas and ca. 25 ogy of the land areas borderingthe North East Water mm in inlandareas. The snow cover in coastalareas lasts Polynyajust off the coast between 79? N and 82? N. for a longerperiod, postponing the onset of plantgrowth SophusMiillers Naes is situatedat the coastof eastern- and reducingthe length of the growing season. mostAmdrup Land (80? 57' N, 14?40' W); the studyarea the flat foreland to 10 km from the coast comprises up Geology (Fig. 2). The othermain locality is situatedin the central partof westernAmdrup Land (80? 49' N, 17? 35' W) ca. The coastal areas at Sophus MtillersNaes consist of 25 km from the outercoast and 8 km from Ingolf Fjord, UpperPalaeozoic and Mesozoic limestone, locally cov- which is ice-coveredall year round(Fig. 5). ered by Quaternary glacial and marine sediments Grazing pressure from mammalian herbivores at (Hakansson et al. 1993). The bedrock at the Amdrup AmdrupLand is negligibleas the densityof herbivoresis Land locality consists of carboniferouslimestone and extremelylow and non-existentat the coastal localities. sandstone, which only exceptionally penetratesthe till In addition, investigationswere carriedout in July (Fredskild1995). In the studyarea at PrinsesseIngeborg and August 1986 at the peninsula Prinsesse Ingeborg Halv0 the carboniferousbedrock is covered by raised Halv0 west of Flade Isblink(81? 36' N, 16? 39' W), and marine silt, beach ridges and glacial deposits. the results are includedfor the comparisonof the flora, - The polar desert zone of Greenland:floristical and ecological characterization 687

Material and Methods Table 1. Climatedata for the summer(July and August) in NorthGreenland north of 80? N computerizedby the Mete- Plant communities orologicalInstitute of Denmarkand compiled from Bay (1992) andKristensen & Kristensen HKH= Henrik were carried out the period from (1993). Kr0yer Analyses during PIH= Prinsesse KM= to when the Holme; IngeborgHalv0; KapMoltke; mid-July mid-August, vegetationoptimally KMJ= Morris HL = Hall Land.M.a. = Mean and the maximal. Plant com- Kap Jesup; temp. developed standingcrop annualtemperature; M.s. temp. = Meansummer temperature; munitieswere identified on the basis of ground survey M.a.prec. = Meanannual precipitation. and floristical observations during one week of field work in each area.Plant communities were then charac- Locality HKH PIH KM KMJ HL terizedby species composition,dominant species andlife Location 80?13'N 81?36'N 82?09'N 83?38'N 81?44'N forms, and physical parameterssuch as topography,ex- 13?43'W 16?40'W 29?57'W 33?22'W 59?00'W posure,soil watercontent and texture. Recordingperiods 1985-87 1961-72 1974-82 1981-87 1983-87 In each three 1990-92 1975-87 1984-87 plantcommunity distinguished, analyses 1990-92 were performed randomlyby establishing 15 plots of 50 cm x 50 cm and cover and M.a. temp. (?C) -12.9 17.0 -15.7 19.0 -20.0 recording frequency.Only M.s. temp. (?C) 1.5 2.5 4.2 0.2 3.3 one stratumwas recognized,and cover was estimatedto Degree days > 0 ?C 203 221 305 70 276 thenearest 1 %for vascularplant species andfor the total M.a. prec. (mm) 227 25* of each:bryophytes, organic crust including lichens, bare * Precipitationat Kap Moltke recorded at the nearby locality Br0nlundhus soil andstones. Species representedby very few plantsor during 1948-1950. leaves, cover < 1 %,received an arbitrarycover value of 0.5 %. Species contributingwith very small values are not includedin the tables,but their numbers and cover are Results given in additionto the numberof species in the types. Above-groundlive biomass of vascular plants was Plant communities harvestedin every thirdplot. Vascularplant material was sorted by species, and pre-driedin the field to avoid Four types of plant communitieswere identified at carbonloss due to respiration.In the laboratorysamples SophusMiullers Nes; they only cover small areas(Table were driedfor two days at 50 ?Cand the drybiomass was 2). None had a continuousplant cover, and no woody recorded.Since biomassfigures are relatedto cover val- species were present.Amdrup Land has six types with a ues they are not included(but are availableon request). higher species diversity, higher percentage cover and with woody species occasionallydominating (Table 3). Soils Three plant communities were identified at Prinsesse Soil pits were dug in one of the standsof each plant IngeborgHalv0; one is characterizedby a woody species communityin AmdrupLand and Sophus Mullers Nes. (Salix arctica) (Table 2). Sampleswere collectedat differentdepths to studyverti- cal differences in the chemical propertiesof the soils, Sophus Miillers Naes althoughno profiledevelopment was observed.Samples A rich Saxfifaga oppositifolia-dominated plant com- were collected at 0 - 1 cm and for each 10 cm down to a munity is closely associated with the few large snow- depthof 30 - 40 cm, and air-driedin the field. pH (H20), driftsin the areaand is consequentlya raretype (Fig. 3). total carbon,nitrogen, phosphorus, organic carbon,and The soil is silty,occasionally with gravel and wet through- carbonatewere analysedby StatensHusdyrbrugsfors0g, out the growing season. Except for monocots, the spe- Ministryof Agriculture.Organic carbon was determined cies composition is closely related to the Saxifiaga as total carbonburt in a Leco instrument.The samples oppositifolia-Papaver radicatum plant community, but were checked for carbonateby adding acid. If signs of the cover and biomass of all plant groups are larger. CaCO3were traced, the content of CaCO3was determined. On fine textured soils a Saxifiaga oppositifolia- Total nitrogencontent was determinedby the Kjeldahl Papaver radicatum-dominated plant community occurs, method; total phosphoruswas determinedby digestion being the most frequenttype in the study area.It occurs with perchloricand sulphuricacid. on moist polygon soils as well as on non-sortedsoils. The third type with Saxifraga oppositifolia and Phytogeography Papaver radicatum, 'herb barren', is the poorest plant Plantcollections kept at the GreenlandHerbarium in communityin NorthGreenland, regarding cover, diver- Copenhagen (see Bay 1992) and later collections by sity, species frequencyand biomass. It covers vast areas Daniels & Alstrup, Berg & Kapel and Schwarzenbach on mesic silty or sandy soils. were used to delimit the polardesert zone in Greenland. No truefens dominatedby sedges, i.e. Carex spp. or 688 Bay, C.

Eriophorumspp., are present,but along melting streams to mesic soils. The soil surfaceis underthe influence of below snowdriftsnarrow stripes of graminoidvegetation solifluction and only few herbaceousplants can estab- occur. lish. On mesic, silty soil a Saxifragaoppositifolia-domi- nated herbaceous community occurs with an organic AmdrupLand crust. On shelteredsouth and west-facing slopes in the The hummocky Salix arctica-dominatedsnowbed non-coastalareas of the peninsula, i.e. at least ca. 5 km community has the highest standing crop value and from the seashore, a relatively rich Salix arctica-domi- number of species; it covers small areas on east and nated plant communitywith an organic crust occurs. south-facingslopes (Fig. 4). On slopes with a long-lasting snow cover a mossy Soils snowbedis found.Bryophytes are prominent with a cover of 56 %;the total numberof vascularspecies is 25. Soils are alkaline;pH varies between 8.8 and 9.3 at Fens occur along streamsand at the edge of ponds. SophusMiillers Noes and between 7.3 and9.3 at Amdrup Theyare species-rich but mainly composed of almostpure Land (Table4). Organicmatter content from the soils at standsof Arctagrostislatifolium and Carex stans (Fig. 5). Amdrup Land mostly shows a decrease from the soil The open Salix arctica community is the most widely surfacedownwards in all plant communities.This is not distributedcommunity at AmdrupLand. It differs from the case for the two Saxifragaoppositifolia communities the herb barren by the presence of Salix arctica and Carex nardina in addition to a larger cover, biomass and a higher number of species. Table 3. Plantcover and diversityof plantcommunities at The algida-dominated community on wet AmdrupLand. 1 = Salix arctica snowbed;2 = Mossy snowbed; solifluction soils occurs below snowdrifts and on slopes 3 = Fen comm.; 4 = Open Salix arctica comm.; 5 = Phippsia 6 = Herb barren. with silty soils that are wet throughoutthe growing season. algida comm.; areas are and snow-free or with a Large wind-exposed Plant community 1 2 3 4 5 6 thin snow cover during winter; they are bare or have a Saxlifraga oppositifolia 0.5 0.5 0.1 0.1 0.3 herb barren vegetation. The soil is dry, well-drained, and Papaver radicatum 0.3 0.4 0.2 0.2 have a vascular cover Stellaria crassipes 0.3 0.2 0.1 mostly silty. Herb barrens species Draha hellii 0.2 0.1 of only 1.5 %, whereas bryophytes are almost absent. Saxifraga cernua 0.2 0.5 0.4 0.3 Dra-hb suhcapitata 0. 1 Dahba adamsii 0.2 0.4 0.2 Prinsesse Ingeborg Halvo Cerastium arcticurn 0.2 0.1 0. I 0. Phippsia algida' 0.3 0.6 The most common plant community on Prinsesse Salix arctica 6.5 0.1 1.0 Halv0 is the herb barren, which occurs on wet Alopecurus al/pinius 0.2 0.2 Ingeborg hyperhorea 0.4 pratensi.s 0(.1 (. 1 Carex nardina 0.1 0.3 J.lnculs hilunmis 0.2 0.1 0.4 0.1 Table 2. Plant cover and diversity of plant communities at Potentilla ruhricaulis 0.2 0.1 Millers Naes(SM) and PrinsesseIngeborg Halv0 (PI). Minuartia rubella 0.3 0.1 0.1 Sophus Cerastiumn regelii . 1 0.2 0.2 1 = Saxifraga oppositifolia snowbed; 2 = Saxifraga oppo- Potentilla hyparctica 0. sitifolia-Papavercomm.; 3 = Herbbarren; 4 = Phippsiaalgida Taraxacum pumilum 0.1 Braya purpurascens 0.1 0. 1 comm.; 5 = Herbbarren; 6 = Saxifiragaoppositifolia comm.; 7 Luzulaarctica 0.1 0.1 = Salix arctica comm. angustata 0.2 0.1 Sagina intermedia 0.1 0.6 SM SM SM SM PI PI PI Carex stans Locality scheuchzeri 0.1 1 2 3 4 5 6 7 Ersiophorum Eriophorum triste 0.1 0.1 2.5 0.2 Julncuustriglumis Saxifiragaoppositifolia 3.5 0.8 0.2 0.6 0.3 0.3 0.1 latifolia PaIpver radicatum 0.5 0.3 0.1 0.1 0.1 Pleuropogon sabinei Stellaria crassipes 0. I 0.1 0.1 0.1 Ranunculus sulphureus Draha bellii 0.3 0.1 0.1 0.4 Saxifi-agafioloosa Saxif-aga cernua 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.4 Cardamine bellidifolia Draba subcapitata arvense 0.1 Draha adamsii 0.1 Equisetum Poa abbhheviata 0.1 0.4 Cerastiumarcticum 0. 1 Potentilla pulchella 0.2 algida algidiftrmis 0.1 0.2 0.5 0.1 Phippsia ssp. Melandriumtriflorum 0.1 Colpodiumvahlianumn 0.5 0.1 Taraxacumphymatocarpum 0.3 Sax-ifragaplatysepala 0.1 0.1 I 2 Salix arctica 8.7 ssp. algidiformis; var. colpodea Numberof other vascularspp. 4 3 1 0 10 10 1 Numberof other vascularspp. 16 12 5 11 9 8 Total numberof vascularspp. 10 10 6 3 14 13 9 Total numberof vascularspp. 36 25 21 20 17 15 Total vascularplant cover (%) 5.0 1.8 0.7 0.8 1.7 4.4 9.6 Total vascularplant cover (%) 10.5 4.4 3.6 2.7 2.1 1.5 Bryophytecover (%) 6.6 1.6 0.1 0.1 0.0 1.5 5.6 Bryophytecover (%) 36.6 56.0 99.0 0.1 10.5 0 Organiccrust cover (%) 14.6 2.5 0.6 50.3 0.1 94.1 85.4 Organiccrust cover (%) 45.2 33.0 0 36.0 77.6 6.2 Total plantcover (%) 26.2 5.9 1.4 51.2 1.8 97.2 100.5 Total plant cover (%) 92.3 93.4 102.6 38.8 90.2 7.7 The polar desert zone of Greenland:floristical and ecological characterization- 689 at SophusMiillers Naes. Generally, the organicmaterial A comparisonof the plant communityherb barren, constitutesonly 1 - 3 %.A maximumof 5.2 %was found which covers the largestarea and is characteristicof both in the mossy snowbedcommunity at AmdrupLand. the Canadian,Greenlandic, and Russian polar deserts, Nitrogenlevels at all sites are low: 0.03 to 0.04 % at shows thatthe totalplant cover is 2 - 3 %.The biomassin Sophus Miillers Naes and 0.02 to 0.25 % at Amdrup all polar desert communitiesis very low, rangingfrom Land. The nitrogen content decreases from the soil 0.6 g/m2 in herb barrens to 11.2 g/m2 in Saxifraga surfaceto a depth of 40 cm at AmdrupLand. Phospho- oppositifoliasnowbeds. The above-groundvascular plant rus levels vary from 226 to 304 ppm at Sophus Millers biomassin Canadianpolar deserts was 15.5 + 18.1 g/m2, Naesand 275 to 519 ppm at AmdrupLand. 82.9 ? 35.2 g/m2and 169.7 ? 143.8 g/m2in herbbarrens, The mean values of phosphorusand nitrogen are cushionplant communities and snow-flush communities, higher at Amdrup Land comparedto Sophus Miillers respectively(Bliss et al. 1984). Nes, 391 vs. 269 ppm, and0.13 vs. 0.04 %,respectively. The numberof vascularspecies in the plantcommuni- ties in the polar desert site near the polynya varies be- tween 3 and 10, whereasit is between9 and 14 at the polar Discussion desert site at the ice-covered sea at Prinsesse Ingeborg Halv0. Saxifraga oppositifolia and Papaver radicatum Plant communities are the most frequent species followed by Saxifraga cernua, Stellaria crassipes and Draba subcapitata.The The low summertemperatures prevailing at the coast diversityin theplant communities in the semi-polardesert and the presence of the open water polynya giving a areaat AmdrupLand is between 15 and 36, Salix arctica high frequencyof cloudy and foggy weatherduring the being the most frequentspecies in mesic and dry plant growing season are the majorphysical factorsresponsi- communities, whereas Carex stans and Arctagrostis ble for the impoverishmentof the vegetation, in respect latifolia dominatein the wet habitats. to species diversity, plant cover and biomass. In large areas the soils originatingfrom easily weatheredlime- stone give rise to a coarse textured, well-drained soil which inhibits plant growth. On silty soils with water Table 4. Soil chemistry of samples from different depths at available partly or during the entire growing season, two study sites. Org. m. = organic matter. four types of plant communitiesare identified. The vascularand cover in Community Depth pH Ct,, Org. m. CO3-C Nto, Pt,, average cryptogam veg- (cm) (%) (%) (%) (%) (ppm) etatedareas at SophusMiillers Naes are 2.1 % and 3.8 %, Sophus Miillers Naes respectively.The plantcover of polardeserts in high arctic Saxifraga 0-1 9.0 10.30 2.35 7.95 0.4 298 Canadais two to threetimes larger:5.3 % and 9.3 % for oppositifolia- 10-20 8.8 10.59 2.05 8.54 0.4 304 comm. 30 8.8 10.39 2.54 7.85 0.3 293 vascular and cryptogam cover, respectively (Bliss et Papaver al. Aleksandrova statedthat the Saxiftaga 0-1 8.8 8.03 0.80 7.23 0.4 260 1984). (1988) generalplant oppositifolia 10-20 8.8 7.53 1.04 6.48 0.4 261 cover is much largerin polardeserts of the Barentsprov- snowbed 30 8.9 7.25 1.11 6.14 0.4 266 ince thanin the easternpart of high arcticRussia. These Herb barren 0-1 9.0 7.26 1.52 5.74 0.4 226 rich and the have a 10-20 9.2 7.58 1.27 6.31 0.4 247 'nanocomplexes' open aggregations 30-40 9.3 7.04 1.00 6.04 0.3 266 cover varying between 85 - 90 % and 1 - 70 %, respec- Mean 8.9 0.4 269 tively,with a contributionof vascularplants of only 1 - 9 % and 1 - 7 %, respectively.The climatic differencesbe- Amdrup Land Open Salix arctica 0-1 8.4 4.15 2.69 1.46 0.23 426 tween the polardeserts of the Canadianprovince and the vegetation 10-20 8.6 1.55 0.73 0.82 0.07 364 Barentsprovince might explain the difference in cryptogam 30-40 8.8 2.47 0.36 2.12 0.04 322 cover. A branch of the Gulf Streamreaches Svalbard Salix arctica 0-1 8.0 4.61 3.67 0.94 0.21 406 snowbed 10-20 8.5 4.23 2.20 2.04 0.19 404 resultingin open water conditionsduring summer with 30-40 8.2 3.20 1.74 1.47 0.14 422 andhumid weather with The cloudy frequentprecipitation. Mossy snowbed 0-1 7.3 5.15 5.15 - 0.25 485 annualmean temperature, - 5 ?Cto - 10 ?C,is lowerthan 10-20 8.3 2.32 2.32 - 0.19 480 - the valueof ca. - 18 ?Cmentioned for comparableareas 30-40 8.3 2.27 2.27 0.18 519 in Canada(Bliss & Svoboda 1984) andNorth Greenland Herb barren 0-1 8.9 2.45 1.20 1.25 0.10 367 10-20 9.2 4.54 0.14 4.40 0.02 275 (Bay 1992). Aleksandrova(1988) states that the lichen 30-40 9.3 2.58 0.24 2.34 0.03 310 crust is of much in the formationof greaterimportance Phippsia algida 0-1 8.5 1.69 1.33 0.36 0.11 346 the plant aggregationsin the Barents province than in snowbed 10-20 8.5 1.10 1.01 0.08 0.09 361 8.5 1.25 1.12 0.14 0.09 378 the Siberian province, where the climate is distinctly 30-40 continental. Mean 8.5 0.13 391 690 Bay, C.

Fig. 2. Herb barrenlandscape at SophusMiillers Naes bordering the NorthEast Water Polynya.Black stripes in the center are Phippsia algida communitiesalong melting water channels. vahlia- num and Saxifraga cernua are the onlyother vascular plants occurring.

Delimitationand distributionof polar desert zone in precipitationand high summertemperatures (Bay 1992; Greenland Fredskild1996), e.g. Carex maritima,Erigeron compo- situs, Braya thorild-wulffii,Lesquerella arctica, Poa Holmen (1957) divided the vegetation of eastern hartziiand Calamagrostis purpurascens. The othertype, NorthGreenland into two districtsaccording to physical 'the fell-field district',occurs both in inlandand coastal conditions, distributionof selected species and charac- regions. It is of a more humid character,organic matter teristicvegetation types. The 'high arcticdesert district' is accumulatingand Cassiope tetragona is a key spe- includes mostly inland areas characterizedby its ex- cies. Surprisingly,both districts comprise inland areas tremelycontinental climate and the presenceof mineral as well as coastal areas. This classification is in disa- soils withoutaccumulations of organicmaterial, and the greement with the findings from other parts of Green- vegetation is characterizedby certain species: Dryas land, wherethere are distinctdifferences in the flora and integrifolia and Kobresia myosuroides in dry habitats the composition of the vegetation between coastal and and Carex stans in wet sites. The species list given by inland areas. Holmen includes species which, in otherparts of Green- Funder& Abrahamsen(1988) reviewed the discus- land, are distinctly related to inland areas with low sion of the high arctic desert concept in Greenlandand

- .00*00 :SW _~~No

Fig. 3. The rich Saxifiragaoppositi- foliacommunities atSophus Miillers Nas occuron wet soil associated withpermanent snow banks. Saxi- fraga oppositifoliaforms a network in the rimsof the polygons.Other importantspecies are Papaver radi- catum,Stellaria crassipes and Draba bellii. The framecovers 0.25 m2. The polar desert zone of Greenland:floristical and ecological characterization- 691

Fig. 4. Salixarctica snowbeds occur on west andsouth-facing slopes. This communityis the most lush andspe- cies-richat AmdrupLand. Other im- portantspecies are Festuca hyperborea, Minuartiarubella, Stellaria crassipes and Papaver radicatum. define polardeserts in agreementwith Edlund(1983) as Epilobium arcticum, Pedicularis flammea and Tofieldia areas where low temperaturesin the growing season coccinea). restrictand nearly inhibitthe growthof vascularplants. Several authorshave proposeda delimitationof the Consequently,vast areasof interiorNorth Greenland do polar desert zone in Greenland in connection with a not belong to the polar desert zone according to this general classification and zonation of the flora and definition,as a relativelylush vegetationoccurs locally, vegetation of the Arctic (Young 1971; Komarkova& wherethe waterregime is favourableduring summer, and Webber 1978; Aleksandrova 1988; Elvebakk 1985; the species diversity is high, i.e. 60 - 90 species of Yurtsev 1994). As no botanicalevidence has been pub- vascularplants (Aastrup et al. 1986;Bay 1992;Daniels & lished in the period from 1957 (Holmen 1957) to the Alstrup1996). The summertemperature inland is mark- mid-1980s (Aastrupet al. 1986; Funder& Abrahamsen edly higherand the physicalfactor mainly controlling the 1988), none of the delimitationsof the polardesert zone low plantcover is the low annualand summerprecipita- in these papersare based on recent botanicaldocumen- tion. In addition,several middle arcticspecies and even tation. Aleksandrova(1988) and Elvebakk (1985) in- low arcticspecies areoccasionally found (e.g. Cardamine clude only coastal areasof westernNorth Greenland and pratensis, Empetrum nigrum ssp. hermaphroditum, Peary Land northof ca. 82? N in the polar desert zone,

Fig.5. Relativelylush fen vegetation occurringon moist soil in depres- sions in the landscapeat Amdrup Land,25 km fromthe outercoast. Carexstans andArctagrostis latifolia arethe dominatingspecies followed by Eriophorumscheuchzeri, E. triste andJuncus biglumis. 692 Bay, C. whereasYoung (1971) considersonly the northernmost PearyLand (ca. 83 ? N) as belonging to the polar desert zone. Funder & Abrahamsen (1988) show in their palynological study that areas in eastern North Green- land southof 82? N also belong to the polardesert zone. On the other hand, they exclude all areas west of Peary Land from this zone. Yurtsev (1994) does not include any partof Greenlandin his High-Arctictundra subzone which otherwise coincides in other parts of the Arctic with the polar desert zone proposedby others. Based on the distributionpatterns of those vascular plantspecies (124 spp.) in NorthGreenland, only occur- ring in the inland,where they reachtheir northern distri- bution limit, Bay (1992) divided the floristic province North Greenlandinto two districts.The coastal district delimitsthe polardesert zone of Greenland.The present Fig.6. Distributionof the Arctic polar desert zone in Greenland. investigations show, however, that this delimitation should be revised: the polar desert zone of Greenland includes only the most coastal areas,i.e. a zone up to ca. eastern North Greenland is at ca. 79? 30' N as coastal 15 km wide (Fig. 6). The Amdrup Land locality was localities furtherto the south have a richerflora includ- originally classified as belonging to the polar desert ing woody species, sedges and vascular cryptogams zone of North Greenland(Bay 1992). The high species (Bay & Fredskild1991). The southernborder in western diversity, the occurrence of 18 species classified as North Greenlandis at ca. 82? N in Hall Land (Fig. 6). distinctly related to the warm continental inland of This new delimitationof the arctic polar desert zone of North Greenland,the presence of woody species (Salix Greenlandexcludes WashingtonLand from the previ- arctica, Dryas integrifolia,D. octopetala, and Cassiope ous delimitationby Bay (1992). The western delimita- tetragona) and aquatic species (Pleuropogon sabinei tion joins the borderline of the Canadianpolar desert and Ranunculushyperboreus), in additionto fens with zone proposedby Edlund(1990). Carex stans and two species of Eriophorum,strongly imply that this area does not belong to the polar desert Soils zone. Likewise, Prinsesse Ingeborg Halv0 is divided into a coastal and an inland zone based on the distribu- pH values correspond to values reportedfrom the tion of key taxa. The following species are only found in Canadian polar deserts, where they vary from 7.9 to sheltered inland habitats: Salix arctica, Dryas inte- 8.5; the same holds for the nitrogenlevel. However, the grifolia, Braya purpurascens,B. thorild-wulffii,Carex phosphoruslevel is much higher in the Greenlandsites misandra, C. nardina, Oxyria digyna and Ranunculus (Bliss & Svoboda 1984). This may be explainedby the sulphureus. different methods used. A comparison of a coastal locality in Peary Land Comparingthe chemicalproperties of the soils of the (82?30' N) including areas up to 5 km from the coast two study sites representinga polar desert and a polar (Fredskildet al. 1992) with a luxuriantlocality only 10 semi-desert shows that there are large similarities and km furtherinland, confirms the narrowcoastal delimita- only minordifferences. This disagrees with the conclu- tion of the polar desert zone. The delimitation of the sion by Bliss et al. (1984) that soil conditions are the polar desert zone at Holm Land (Fig. 1) is based on a major factor influencing the impoverishment of the comparisonof the totalnumber of species andthe number plant cover in polar deserts. of inland species at two localities at approximatelythe The mean pH of the areas differs by 0.5. There is a same latitude. 51 species, of which 23 are inland spe- gradualdecrease in the content of organic matterfrom cies, were foundat Firdalen duringa shortstay, whereas the top to the bottom in all the plant communities at the number of species at the outer coast locality, Amdrup Land and in the herb barren community at Eskimonaesset,has only 17 species of which none are Sophus Mullers Naes. classified as inland species. M0lgaardin Hakanssonet There is a mean difference of 122 ppm of phospho- al. (1993) divides the semi-nunatakKilen into floristic rus between the coastal and inlandpart of AmdrupLand regions based on floristics. His ca. 5 km broad coastal and the mean total nitrogen content is 0.04- 0.13 in zone belongs to the polar desert zone. coastal and inlandparts of AmdrupLand, respectively. The southern border of the polar desert zone in The polar desert zone of Greenland:floristical and ecological characterization- 693

Species diversityof the arctic polar desert zone A comparisonof the numberof vascular species at well-investigatedlocalities within a latitudinalrange of Most species, i.e. 60 - 70 %,have an arctic-alpineor ca. 150 km along a climatic gradientfrom the inland to arcticcircumpolar distribution (App. 1) andspecies with the coast, a distanceof 50 km, shows a close relationship a low arcticdistribution are absentfrom the arcticpolar between the number of vascular species and the tem- desertzone. Differentnumbers have been given for the peraturesduring summer (Table 5). The total numberof total numberof vascularplants within the polar desert vascularspecies and the numberof species classified as zone. Young (1971) mentions50 vascularplant species, inland species judging from their distributionpattern in and the numberfrom the Russian polar desert zone is the North Greenlandicfloristic districts(Bay 1992) are between 17 and 75, giving a total of 98 species for all compared with the recorded and estimated mean July Russianpolar deserts (Aleksandrova1988). temperature.In additionto the localities studiedin 1986 The coastal floristic districtof North Greenlandhas and 1993, the only other well-investigatedlocalities in 55 species (Bay 1992). As the present investigations the region, CentrumS0 andKilen, aretaken into consid- have alteredthe delimitationof this district, some spe- eration.The numberof species varies between 81 in the cies previously considered as occurring both in the continentalinland and six species on the islands in the coastal and continentaldistrict of North Greenlandare polynya. CentrumS0 and AmdrupLand differ signifi- excluded. 33 vascularspecies occur in the Greenlandic cantly from the other localities in respect to both the polar desert zone (App. 1). The bioclimatic zone 1 by total number of vascular species and the number of Edlund (1983), including areas north of the northern inland species. None of the other localities have any limit of prostrateshrubs and sedges, has less than 35 inland, woody or aquaticspecies, and sedges are absent vascular species. Elvebakk (1989) and Aleksandrova or rare (App. 2). The number of species in the outer (1988) listed44 speciesfrom the polar deserts of Svalbard coast localities ranges between 40 and 6. The localities and 50 from the Barentsprovince of Russia. borderingthe open waterat the NorthEast Water polynya Rannie (1986) dealt with the relationshipbetween have a smallernumber of species comparedto the coastal summerair temperatureand number of vascularspecies locality at PrinsesseIngeborg Halv0 at a partlyice cov- in arcticCanada and showed a strong correlationand a eredsea (21 and23 vs. 40. The most impoverishedarea is vascularspecies diversitygradient of 24 - 26 species per found in the polynya with only six species of vascular degree. From her regression,the July temperatureof a plants.The same numberof vascularspecies was found locality can be estimatedfrom its vascularspecies rich- on a small islandat the coast of northernmostGreenland ness to a standarderror of 0.5 ?C. For the three study (Anderssonet al. 1980). sites the mean July temperatureis calculatedas 2.7 ?C, Using Elvebakk's (1990) index of thermophilyfor 3.3 ?C, and 4.5 ?C for Sophus Miillers Naes, Prinsesse defining the biogeographical zone confirms that the IngeborgHalv0 and AmdrupLand, respectively. Mete- areas in Greenlandwith the species composition given orological data are only available from Prinsesse in App. I should be classified as arctic polar deserts, as IngeborgHalv0. The mean July temperaturefor a 27-yr only two species classified as weakly thermophilousare recordingperiod was 3.2 ?C, which correspondswell to found only rarelywithin the zone. the calculatedmean July temperature(3.3 ?C). Judging The investigations of the polar deserts at Sophus from this calculationonly the two first mentioned sites MuillersNaes, support the hypothesis by Young (1971), are within the polar desert zone according to the tem- Edlund (1983) and Funder& Abrahamsen(1988) that peratureregime used by Edlund(1983). She defined the the primaryecological factor controlling the impover- 'mini-treeline'by a July mean temperatureof less than ishmentof the flora and vegetation at the coasts around 3.0 - 3.5 ?C and designatedareas north of this to be in zone 1, which merges with the zone defined as the polar desertzone others Aleksandrova Table 5. Northernlatitude (N lat.), Distancefrom the outer by (Young 1971; 1988). coast Numberof vascular inland The evaluationof Landas classified (D-coast), species: (S-inland) Amdrup being high andtotal (S-total), and July temperature (July temp.; based on arcticsemi-desert from the matches temperatureregime Rannie1986). CS = CentrumS0; AL = AmdrupLand; PIH the biological facts that three woody species occur, PrinsesseIngeborg Halv0: SMN = SophusMiillers Naes; KCP among which Salix arctica is dominantin some of the = Kilen, coastal part;HKH = HenrikKr0yer Holme. plant communities,the vascularspecies diversity of 73 Locality CS AL PIH SMN KCP HKH is more than double the numberfor polar deserts given by Edlund(1983), andboth sedges (five Carex spp., two N lat. 80?10' 80045' 81?36' 80?55' 81?10' 80?40' D-coast (km) 150 25 0-5 0-5 0-5 0 Eriophorumspp. and Kobresia myosuroides) and aquatic S-inland 36 17 0 0 0 0 species (Ranunculus hyperboreus and Pleuropogon S-total 81 73 40 23 21 6 sabinei) are present. July temp. (?C) 4.8 4.5 3.3 2.7 2.6 2.1 694 Bay, C.

Thermophily and now includes only coastal parts of Nordaustlandet, easternmost Spitsbergen, eastern Barents0ya and Edge0ya. Consequently, the polar desert zone of Svalbard is extended as far to the south as 76? 30' N. The delimitation to the west roughly coincides with the 2 ?C July isotherm (Elvebakk 1989). The polar desert zone of Greenland includes the main parts of the coastal floristic district of North Greenland (Bay 1992) except for a few areas in eastern North Green- land, now comprising a much larger area than that pro- posed by Aleksandrova (1988) and Elvebakk (1985). Edlund & Alt (1989) showed a congruence of the Canadian polar desert zone and the 2 ?C isotherm. There is no evidence, according to Edlund (1990), that the Canadian high arctic archipelago should be divided into two polar desert subzones as proposed by Aleksandrova (1988). The Canadian polar deserts are all included in the southern belt of the arctic polar deserts.

Acknowledgements. The 1993 projectwas carriedout as part of the NEWLandproject supportedby the Danish Research Council for the Humanities, Ole M. Windsted's memorial Fig. 7. Distributionand subdivisionof the polar deserts in the foundationand the Aage V. Jensen Foundation.Their finan- Arcticbased on Aleksandrova(1988), Elvebakk(1989), Edlund cial support is gratefully acknowledged. The logistics were (1990), Bay (1992) and the presentstudy. Only the northern- arrangedby T.I. Hauge Andersson,Danish Polar Center.The most islandsin high-ArcticRussia are included in the northern U. S. Coast Guardassisted with helicoptertransport to Henrik belt of the polar desert zone of the circumpolarArctic. Kr0yer Holme. The Centrallaboratoriet,Foulum, carried out soil analyses. I thankT.I. HaugeAndersson, F.J.A. Daniels, V. Alstrup, T.B. Berg, C. Kapel and F.H. Schwarzenbachfor puttingtheir plant collections at my disposal. The the Arctic Ocean is the low temperatures during the The 1986 field work was financially supportedby Commission for Scientific Research in Greenland,where the short growing season, because a very low plant cover were J.P.S. the Danish and were also recorded in areas with logistics organized by Clemmensen, species diversity Air Forces. I also thank the staff at Station Nord who have soils and adequate water supply during the growing silty been very helpful during both field seasons, and R.M. season. The low summer are to temperatures supposed Kristensen for putting meteorological data at my disposal. factor for establishment and be the main limiting plant Finally, I am indebtedto B. Fredskildfor his criticalreading of growth in the Greenlandic polar deserts. The conditions the manuscriptand to N. Auerbachfor her design of a map recorded in the continental polar deserts of Queen (Fig. 7). Constructivecomments from the reviewers greatly Elisabeth Island, Canada, where the surface soil is dry in improvedthe manuscript. the mid-season and is believed to be a major controlling factor responsible for the impoverished flora and plant References cover (Bliss & Matveyeva 1992), do not occur in the coastal polar deserts of Greenland. Aastrup, P., Bay, C. & Christensen, B. 1986. Biologiske milj0unders0gelseri Nordgr0nland,1984-85. Gr0nlands Fiskeri- og Milj0unders0gelser1986: 11-58. Conclusions Aleksandrova,V.D. 1988.Vegetation of theSovietpolar deserts. CambridgeUniversity Press, Cambridge. The circumpolar distribution of the arctic polar desert Andersson,T.I.H., Funder,S. & Hjort, C. 1980. PA verdens zone as presented by Aleksandrova (1988) has to be nordligstelandjord. Naturens Verden 10: 314-322. changed for three areas: central Svalbard, eastern North Bay, C. 1992. A phytogeographicalstudy of the vascularplants - ? Greenland, and the western Canadian high-Arctic archi- of northernGreenland north of 74 northernlatitude. Medd. Gr0nlandBiosci. 36: 1-102. pelago (Fig. 7). The Papaver dahlianum zone comprising Bliss, L.C. 1988. Arctic tundraand polar desert biome. In: the polar deserts in Svalbard (Brattbakk 1986) was slightly Barbour,M.G. & Billings, W.D. (eds.) North American altered by Elvebakk (1989) by using his Index of - The polar desert zone of Greenland: floristical and ecological characterization - 695

terrestrialvegetation, pp. 1-32.Cambridge University Press, Elvebakk, A. 1985. Higher phytosociological syntaxa on Cambridge. Svalbardand their use in subdivisionof the Arctic.Nord. J. Bliss, L.C. 1993. Arcticcoastal ecosystems. In: van derMaarel, Bot. 5: 277-284. E. (ed.) Ecosystemsof the World2 A, pp. 15-22. Elsevier, Elvebakk, A. 1989. Biogeographical zones of Svalbard and Amsterdam. adjacent areas based on botanical criteria. Ph.D. Thesis, Bliss, L.C. & Svoboda, J. 1984. Plant communitiesand plant Universityof Troms0. productionin the westernQueen Elisabeth Islands. Holarct. Elvebakk,A. 1990. A new methodfor definingbiogeographical Ecol. 7: 325-344. zones in the Arctic. In: Kotlyakov,V.M. & Sokolov, V.E. Bliss, L.C., Svoboda,J. & Bliss, D.I. 1984. Polardeserts, their (eds.) Arctic Research:Advances and prospects, pp. 175- plant cover and plant productionin the CanadianHigh 186. Proc. Conf. Arctic. Nord. Count.Coord. Res. Arctic, Arctic.Holarct. Ecol. 7: 305-324. Leningrad1988, Part2. Bliss, L.C. & Matveyeva,N. 1992. Circumpolararctic vegeta- Fredskild,B. 1995. Palynologyand sediment slumping in a high tion. In: Chapin, F.S., Jefferies, R.L., Shaver, G.R. & arcticGreenland lake. Boreas 24: 345-354. Svoboda,J. (eds.)Arctic ecosystems in a changingclimate. Fredskild,B. 1996. A phytogeographicalstudy of the vascular An ecophysiologicalperspective,pp. 79-81. AcademicPress, plantsof West Greenland(62?20'-74?N). Medd. Gronland New York, NY. Biosci. 45: 1-157. Brattbakk,I. 1986. Vegetasjonsregioner- Svalbard og Jan Funder, S. & Abrahamsen,N. 1988. Palynology in a polar Mayen.Nasjonatlas for Norge. Hovedtema4: Vegetation desert,eastern North Greenland. Boreas 17: 195-207. og dyreliv.Statens kartverk. Holmen, K. 1957. The vascularplants of Peary Land, North Bocher,T.W., Fredskild,B., Holmen,K. & Jakobsen,K. 1978. Greenland.Medd. Gronland124, 9: 1-149. GronlandsFlora. 3rded. P. Haase & S0ns Forlag, Copen- Hakansson,E., Birkelund,T., Heinberg,C., M0lgaard,P. & hagen. Pedersen,S.A.S. 1993. The Kilen Expedition 1985. Bull. Daniels,F.J.A. & Alstrup,V. 1996. On the vegetationof eastern Geol. Soc. Denmark40: 9-32. NorthGreenland. Acta Bot. Neerl. 45: 583. Komarkova,V. & WebberP.J. 1978. An alpinevegetation map Edlund,S.A. 1980. Vegetationof LougheedIsland, District of of Niwot ridge, Colorado.Arct. Alpine Res. 10, 1: 1-29. Franklin.Geol. SurveyCanada, Paper 80-1A: 329-333. Kristensen, N.M. & Kristensen, R.M. 1993. Nord0stvands- Edlund,S.A. 1983. Bioclimaticzonation in a high arcticregion: polynya - 0rkeneller oase i havet ud for Nord0stgr0nland. CentralQueen Elisabeth Islands. In: Current Research, part Forskningi Gronlandltusaat1/93: 14-28. A, Geol. SurveyCanada, Paper 83-lA: 381-390. Matveyeva,N.V. 1994. Floristicclassification and ecology of Edlund,S.A. 1990. Bioclimatic zones in the Canadianarctic tundravegetation of theTaymyr Peninsula, northern Sibiria. archipelago.In: Harington,C.R. (ed.) Canada's missing J. Veg. Sci. 5: 813-828. dimension. Science and history in the Canadian Arctic Rannie, W.F. 1986. Summer air temperatureand number of Islands.Vol. 1, pp. 421-441. CanadianMuseum of Nature, vascularspecies in arcticCanada. Arctic 39: 133-137. Ottawa. Young, S.B. 1971. The vascularflora of St. Lawrence Island Edlund,S.A. & Alt, B.T. 1989.Regional congruence of vegeta- with special reference to floristic zonation in the Arctic tion and summerclimate patternsin the Queen Elisabeth regions.Contr. Gray Herb. 201: 11-115. Islands,Northwest Territories, Canada. Arctic 42, 1: 3-23. Yurtsev,B.A. 1994. Floristicdivision of the Arctic.J. Veg. Sci. 5: 765-776.

Received 20 August 1996; Revision received 26 February1997; Accepted 12 June 1997.

App. 1. Comparisonof the vascularplant flora of the arctic App. 2. List of vascularplant species and their occurrencein polardesert zone of the Canadian(C) andBarents (B) provinces the study areas and localities: CS CentrumS0, AL = Amdrup for the areasGreenland (G) basedon Bay (1992), Hakanssonet Land, SM = Sophus MUllers Nes, HKH = Henrik Kr0yer al. (1993) and the presentstudy; Canada (C) based on Edlund Holme, PIH = PrincesseIngeborg Halv6, KCP = Kilen, coastal (1983) and Edlund& Alt (1989); Svalbard,Norway (N) based part.Data fromKilen takenfrom Hakanssonet al. (1993); data on Elvebakk(1989); and Zemlya Frantsalosifa,northwestern from CentrumS0 compiled from Bay (1992) and additional Russia(R) based on Aleksandrova(1988). The Canadianlist is materialcollected in 1993. The frequencyis given as: 5 = very not complete. x = frequent;r = rare. Distributiontypes: A = common, 4 = common, 3 = occasional, 2 = rare, 1 = very Arctic, A-A = Arctic-alpine, HA = High Arctic, Circ = rare. The data from Kilen are given in the categories: circumpolar,Amp-A = amphi-Atlantic,NAm = North Ameri- A = abundant, R = sparse, and L = local, i.e. recorded at one can, Eura= Eurasian.Thermophily indication for some species or two sites. according to Elvebakk (1989): DT, MT, WT = Distinctly, Moderately,Weakly thermophilous. 696 Bay, C.

App. 1 App. 2

Area G C N R Distribution Locality CS AL SM HKH PIH KCP Province C CB B type Alopecurus alpinus x 4 3 3 R Arctagrostislatifolia x 4 Alopecurus alpinus x x x x A-A, Circ Braya purpurascens x 4 1 Arctagrostislatifolia DT x A-A, Circ Braya thorild-wulffii x 3 1 Braya purpurascensWT x HA, Circ Cardaminebellidifolia x 3 3 L Cardaminebellidifolia X X X X A-A, Circ Cardaminepratensis 1 Carex ursina WT x A, Circ Carex misandra x 4 1 Cerastiumalpinuml arcticum X X X X A, Amp-A Carex nardina x 4 2 Cerastiumregelii X X X HA, Circ Cassiope tetragona x 1 Cochlearia groenlandicalarctica X X X X A, Circ Cerastiumarcticum x 4 4 4 R Colpodiumvahlianum X X X HA, Circ Cerastiumregelii x 4 3 4 A Deschampsiaalpina X X A-A, Circ Cochlearia groenlandica 3 2 x 3 A caespitosa ssp. glauca x A-A, Circ Colpodiumvahlianum x 4 2 2 Draba adamsii X X HA, Circ Draba adamsii x 3 4 x 3 A Draba alpina X X A-A, Circ Draba arctogena x 2 3 Draba bellii X X X X HA, Circ Draba bellii x 4 4 4 Draba kjellmanii x HA, Eura Draba lactea 2 Draba lactea X X A, Circ Draba micropetala 2 3 Draba micropetala x HA, Circ Draba subcapitata x 4 4 4 A Draba nivalis x A, Circ Dryas integrifolia x 4 2 Draba oxycaipa x A-A, Amp-A Dryas octopetala x 1 Draba pseudopilosa x A-A, Amp-A Eriophorumtriste x 3 D-aba subcapitata X X X X HA, Circ Festuca hypeirbocrea x 4 Dupontiafisheri s.l. MT X X A, Circ Festuca vivipara 1 Dupontiapsilosantha x A, Circ Juncus higlumis x 4 3 Festuca haffinensis x HA, Circ Luzulaarctica x 4 Festuca brachyphyllas.l. WT x A, Circ Luzula confusa 1 L Festuca vivipara x A, Circ Melandi-iumapetalum x 4 Juncus biglumis X X X X A, Circ Minuartiarossii x 3 2 Luzula arctica X X X X A, Circ Minuai-tiarubella x 4 3 3 Luzulaconfusa (X) x x x A-A, Circ Oxyria digyna x 3 3 Minuaitia rossii WT (X) HA, NAm Papave- radicatum x 5 5 x 5 A Minuartiarubella X X X X A, Circ Pedicularis hirisuta x 4 Oxyria digyna (X) X X X A, Circ Phippsia algida ssp. algidifo-mis x 4 4 x A Papave- radicatunms.l. X X X X A, Circ Poa abbreviata x 4 3 4 A Phippsia algida X X X A, Circ Poa arctica 1 L Phippsia concinna X X A, Eura Poa pratensis var. colpodea x 4 Pleuropcogonsahinbiei x HA, Amp-A Polygonlumviviparum x 4 Poa abbreviata X X X X HA, Circ Potentcillahyparc tic a 2 R Poa allpi/enai l s.l. x AA, Circ Potentillo/pul/chella x 4 Poiaci/linlc x A, Amp-A Puccinellia anguslatatt x 4 3 3 A Pota arcctic cas.l. X X X A, Circ Puccinellia /ruggtemtanni x Poa piratensisvar. col/podea (X) X X HA, Circ Rainunculussulphurelus x 4 1 Pol,/Wcygonlviilv iplarum X X A, Circ Sagina i1ntermnledia 3 PoteltillahI//7larctica (X) X X X HA, Circ Salix arctica x 5 1 4 Potentilla pullclella WT (X) HA, Amp-A Sa.xifragac aespitosa x 2 3 4 A Plc c(inellia angustacta X X X HA, Circ Saxifiaga cerinua x 57 4 x 4 A Picci nellicabruggemanni x HA, NAm Sax.Itrlagahyperborea 1 2 L Plccinelliaph//ic ganodes X X A, Circ Saxifiraganival is 3 2 3 L Ralnuncillls lhypeihorus X X A, Circ Saxifiaga oppositi/olia x 5 5 5 A Ranunculuclissabinei x HA, NAm Satx.\ifragaplat(ysepala x 2 4 Ranlunculussulp/huretts X X A, Circ Sax.ifragatenuiis 3 3 R SRaniclcSagina irc/icmciclicinteirmedia X X A, Circ Stellariaccrassipes x 4 3 x 3 A Salix arctica (X) x A, Circ Tarlaxacnumarcticlurn x 3 Salix polaris X X A-A, Amp-A Taraxacumn alrcltogenllum x Sax.ifaga ccaecspitosas.l. X X X X A, Circ Tarax.acumphymotocarlnlm x 4 Saxifiaga cernua X X X X A, Circ Taraxacumnplnmil/num x 3 Saxifragalfiiolosa X X X A-A, Circ Total numberof inlandspecies 36 18 0 ( 000 Saxlifi-agaliype holrea (X) X X A, Circ Total numberof species 81 73 23 6 40 21 Sacifiaga nicvalis X X X X A, Circ Saxifiaga oppositifolia X X X X A-A, Circ Inland species C''S A\L CS AL Saxifiaga platysepala (X) X X HA, Circ Saxifiaga rivularcis X X A, Circ Almericascabra x Brayachumilis x Saxifi-agatenuis (X) X X A, Circ Calanlagrostispiirpur/ascens x Carex actrofisca x Silenleacaulis x A-A, Amp-A Carex maritima x 3 Carex rupestris Stellaria longipes s.l. X X X X A, Circ Carex stans x 3 CaL-e.supina Chamaetnerionlatifolium x Numberof species 33 28 44 50 Deschanpsia brevcifolia x 4 Dr-abaarctica x 3 Elymushyperarcticus x Epilohiuniarcticum x Equisetumarvense x 3 Equisetlumvariegatum x I Eriigeroni compositus x Erligeron eriocephalus x Eriophor-umcallitrix x Erliophoilrumscheuchzeri x 4 Eutrenmaedwardsii x 3 Festuca haffinensis x Juncus castaneus x luncusstr-iglumis x 2 Kohresia myosur-oides x 2 Kohr-esiasimpliciuscula x Lesquerlellaarctica x I Melandriiumaffine x Melandrium triflor-um x 4 Pedicularis flammea x Pleluropogonlsabinei x 4 x 3 Poa hartzii x Potentilla rubricaulis x 3 Ranuncullushyperboreus x 4 Saxifiaga aizoides x Saxift-agafi/olioosa 3 Silene acaulis x Trisetumspicautum x