DOI: 10.2478/hacq-2014-0013 HACQUETIA 13/1 • 2014, 171-189 DRY GRASSLAND TYPES IN THE PRESPA NATIONAL park (nw greece), including the southernmost occurrence of the priority habitat type "pannonic sand steppes" (code 6260) Georgios FOTIADIS1'*, Michael VRAHNAKIS2, Yannis KAZOGLOU3 & Ioannis TSIRIPIDIS4 Abstract A recently completed project on the Natura 2000 sites of Prespa National Park revealed that the area hosts 49 habitat types according to the EU classification, of which eight have a narrowly restricted distribution in Greece. The priority habitat type "*6260 Pannonic sand steppes" is reported here for the first time for Greece. The new locality represents its southernmost occurrence on the Balkan Peninsula. The aim of this paper therefore is to describe the Greek stands of this habitat type, investigate the factors affecting its occurrence and discuss the associated conservation issues. This is based on a total of 87 relevés sampled for this study and 8 additional relevés from the literature. The relevés were classified by applying TWINSPAN and ordinated using Detrended Correspondence Analysis (DCA). Six vegetation units could be distinguished. One of these corresponds to the habitat type *6260 and was found exclusively on inland sand dunes. Syntaxonomically, we assigned this unit to the alliance Sileno conicae-Cerastion semidecandri s.l. of the class Koelerio-Corynephoretea. Its occurrence in the study area seems to be determined by climatic factors, as well as by the sandy substrate, while anthropogenic disturbances such as grazing also appear to be beneficial to some degree. Apart from the Koelerio-Corynephoretea, another five grassland classes were distinguished in the national park, namely the Thero-Brachypodietea, Stellarietea mediae, Festuco-Brometea, Daphno-Festucetea and Juncetea trifidi. Key words: Habitats Directive, Koelerio-Corynephoretea, phytosociology, Sileno conicae-Cerastion semidecandri, syntaxonomy. Izvleček Nedavno končani projekt o območjih Natura 2000 v Narodnem parku Prespa je pokazal, da je na območju po EU klasifikaciji 49 habitatnih tipov, od katerih jih je osem z ozko razširjenostjo v Grčiji. Pojavljanje prednostnega habitata "*6260 Panonske stepe na peščenih tleh" v Grčiji objavljamo prvič. Nova lokacija predstavlja najjužnejše pojavljanje na Balkanskem polotoku. Namen članka je opisati sestoje tega habitatnega tipa v Grčiji, preučiti dejavnike, ki vplivajo na njegovo pojavljanje in razpravljati o vprašanjih, povezanih z njegovim ohranjanjem. Raziskava temelji na 87 vegetacijskih popisih in 8 dodatnih popisih iz literature. Popise smo numerično obdelali s pomočjo TWINSPAN klasifikacije in z uporabo korespondenčne analize z odstranjenim trendom (DCA). Ločili smo šest vegetacijskih enot. Ena ustreza habitatnemu tipu in smo jo našli le na celinskih peščenih sipinah. Sintaksonomsko smo jo uvrstili v zvezo Sileno conicae-Cerastion semidecandri s.l. in razred Koelerio-Corynephoretea. Njeno pojavljanje v preučevanem območju je pogojeno s klimatskimi dejavniki in peščeno podlago, antropogene motnje, kot na primer paša, pa do neke mere nanjo vplivajo pozitivno. Poleg travišč razreda Koelerio-Corynephoretea smo v parku našli še pet travišč iz razredov Thero-Brachypodietea, Stellarietea mediae, Festuco-Brometea, Daphno-Festucetea in Juncetea trifidi. Ključne besede: Habitatna direktiva, Koelerio-Corynephoretea, fitosociologija, Sileno conicae-Cerastion semide-candri, sintaksonomija. 1 Department of Forestry and M. N. E., TEI of Sterea Ellada, Mpakogianni str., Karpenissi, 36100, Greece, gfotiad95@ gmail.com 2 Department of Forestry and M. N. E., TEI of Thessaly, Terma Mavromihali str., Karditsa, 43100, Greece, mvrahnak@ teilar.gr 8 Society for the Protection of Prespa (current address: Municipality of Prespa), 53077, Greece, ykazoglou@gmail.com, 4 School of Biology, Aristotle University of Thessaloniki, P. O. Box 104, 54124, Greece, tsiripid@bio.auth.gr 1. INTRODUCTION The recently accomplished project to record, assess and map the rangeland and forest habitat types of the Natura 2000 sites of the National Park of Prespa ("Ethnikos Drymos Prespon-GR1340001" and "Ori Varnounta-GR1340003") revealed an exceptional diversity of plant species and habitats. The national park hosts 49 habitat types (19 more than those recorded in 2000), 70 vegetation types and more than 2,100 vascular plant species (almost 30% of the Greek flora) with 194 considered as important according to Annex II of the Habitats Directive (Vrahnakis et al. 2011). Seven habitat types are of conservation priority according to Annex I of the Directive 92/43/EEC (Habitats Directive); they occupy 26% (11,000 ha) of the total area of the national park. Eight of the recorded habitat types have quite restricted distribution in Greece. In the previous report on habitat types of the national park in 2000 (Dafis et al. 2001), three grassland priority habitat types (indicated with "*") from a total of 31 types had been documented. The new mapping in 2011 revealed four dry grassland priority habitat types (*6210, *6220, *6230 and *6260) occupying an area of 8,754 ha. The present study expands the work carried out during the recent mapping of the national park (Vrahnakis et al. 2011) by means of additional phytosociological sampling. Most importantly it reports for the first time the occurrence of the phytosociological order Festuco-Sedetalia acris Tx. 1951 (sensu Dengler 2001 et seq.) and the priority habitat type *6260 (Pannonian sand steppes) in Greece. Therefore, this study specifically aims to describe the floristic composition and ecology of the Greek stands of the Festuco-Sedetalia acris, as well as investigate its differentiation from other dry grassland vegetation types occurring in the same national park. Furthermore, the factors determining the occurrence of this regionally rare habitat type are explored and conservation recommendations are made. 2. STUDY AREA The study area comprised the whole of the Prespa National Park, which is located in northwestern Greece on the frontiers with Albania and the Former Yugoslav Republic of Macedonia. It is characterized by the presence of two lakes, Megali Prespa and Mikri Prespa at approx. 850 m a.s.l., while the surrounding mountains of Varnountas (2334 m), Devas (1372 m), Triklario (1750 m) and Vrondero (1351 m) contribute to the high landscape diversity of the area. The eastern sector of the national park is influenced by the perennial stream of Agios Germanos. The stream was diverted in the 1930s, so as to flow into Lake Megali Prespa instead of Lake Mikri Prespa (Hollis & Stevenson 1997) (Figure 1). The bedrock is siliceous (mostly granitic) in the eastern part of the national park and calcareous in the western part (I. G. M. E. 1983). Alluvial deposits dominate the plain of the basin, whereas sandy soils are found in the southeastern part of Megali Prespa (Figure 2) and on the Slogi islet (Figure 1). The climate of the area experiences both Mediterranean and continental influences and could be characterized as sub-continental Central European (Ristevski 2000). The mean annual temperature is 10.8 °C (Meteorological Station of Pretor; period 1961-1991), with the warmest month being July (19.2 °C) and the coldest January (0.2 °C). The mean annual precipitation for that station is 730 mm; in the lower parts of the national park, precipitation ranges from 600 to 700 mm, while in the mountainous zone it increases to 800900 mm, and in the higher mountainous areas up to 1400 mm (Ristevski 2000, Fotiadis et al. 2012). 3. METHODS 3.1 Vegetation data The study covers all grassland types within the area of the national park. Field data were sampled during 2009 and 2010 according to the Braun-Blanquet method (Braun-Blanquet 1964, Dier-schke 1994). The plots were selected subjectively, with the aim of having all different grassland types represented in the data set by an adequate number of relevés, i.e. corresponding roughly to the area covered by each vegetation type. Plots were selected so as to be floristically and ecologically homogeneous, as well as representative of the vegetation type in which they were sampled. Plots were square in shape and had an area of 16-25 m2 each, which is recommended for grasslands (Knapp 1984, Kent & Coker 1992, Chytry & Otypkovâ 2003). Only vascular plants were recorded. The modified 9-point Braun-Blanquet scale was applied for species cover estimation Figure 1: Distribution map of Koelerio-Corynephoretea (red colour) vegetation represented by its relevés (black points) in various localities including the Slogi islet (points 8-12). Slika 1: Karta razširjenosti vegetacije razreda Koelerio-Corynephoretea (rdeče območje) s predstavljenimi različnimi lokacijami popisov (črne točke), vključno z otočkom Slogi (točke 8-12). (Barkman et al. 1964, van der Maarel 1979, Parol-ly 2003). Besides the floristic data, in each relevé the following parameters were recorded: geological substrate, altitude (in m a.s.l.), exposition (in degrees), slope inclination (as percentage), total vegetation cover as well as cover of shrub and herb layers (the latter three as an estimate in %). In total, 87 relevés of dry grasslands were sampled in the national park, and eight were used from a previous study from the region (Quézel 1969). All relevés are stored in the Balkan Dry Grassland Database (Vassilev et al. 2012), with code EU-00-013 in the Global Index of Vegetation-Plot Databases (GIVD; Dengler et al. 2011). Nomenclature of taxa follows Dimopoulos et al. (2013). The cited syntaxonomy follows Mucina et al. (1993), Mucina (1997), Schaminée et al. (1996), Dengler (2001), Rodwell et al. (2002), Dengler et al. (2003), Berg et al. (2004) and Chytry (2007). Figure 2: Stand of the Silene frivaldszkyana-Erysimum microstylum (Sileno-Cerastion) community (Koelerio-Corynephoretea) at the shores of Lake Megali Prespa. Slika 2: Sestoj združbe Silene frivaldszkyana-Erysimum microstylum (Sileno-Cerastion, Koelerio-Corynephoretea) na obali jezera Megali Prespa. 3.2 Numerical analyses The relevés were classified using TWINSPAN (Hill 1979). Three pseudospecies cut-levels (0%, 5%, 25%) were used, and four levels of divisions were applied. Detrended Correspondence Analysis (DCA) was also applied to facilitate the ecological interpretation of vegetation differentiation, as well as to check the discrimination of the groups produced by the classification. For each relevé, the "relative abundances" of the diagnostic species of six phytosociological classes (Koelerio-Corynephoretea Klika in Klika et Novak 1941, Thero-Brachypodietea Br.-Bl. ex A. de Bolos y Vayreda 1950, Stellarietea mediae R. Tx. et al. ex von Rochow 1951, Festuco-Brometea Br.-Bl. & Tx. in Br.-Bl. 1949, Daphno-Festucetea Quézel 1964 and Juncetea trifidi Hadac 1946) were calculated according to the assignment in Mucina (1997). As "relative abundances", we used the summed percentage values corresponding to the mid-points of the respective Braun-Blanquet cover-abundance category raised to the power 0.2. These six classes were used as passive variables in the DCA because they were represented by a significant number of taxa in the data set. This projection al- so aimed at exploring the relationships between the distinguished vegetation units and certain syntaxa (classes) (see Bergmeier et al. 2009). Furthermore, an additional DCA was applied, using a partial data set in order to better explore the floristic and ecological affinities of the vegetation unit under study with the most closely related vegetation units. In both numerical analyses (classification and ordinations), species cover abundances were square-root transformed. Diagnostic taxa were determined using the algorithm of Tsiripidis et al. (2009). This algorithm uses a fidelity threshold based on relative constancy differences of taxa between groups. Its advantage in comparison with other numerical means of diagnostic species determination is that it conducts multiple comparisons between different combinations of vegetation groups. Specifically, the algorithm searches in a data set for a group or a combination of groups that is differentiated by a taxon against another group or combination of groups. In this way it can reveal differentiating structures in a data set (Tsiripidis et al. 2009). As the algorithm is based on differences of relative constancy values, the diagnostic taxa resulting from this algorithm were tested Table 1: Eigenvalues, length of gradient and total inertia for the DCA axes of the two data sets. Data set A includes all the 95 relevés, while data set B includes the relevés of Koelerio-Corynephoretea, Thero-Brachypodietea and Stel-larietea mediae (26 relevés). Tabela 1: Lastne vrednosti, dolžina gradienta in variabilnost vseh ordinacijskih osi DCA v obeh podatkovnih nizih. Podatkovni niz A vsebuje 95 popisov, niz B pa popise razredov Koelerio-Corynephoretea, Thero-Brachypodietea and Stellarietea mediae (26 popisov). EIGENVALUES GRADIENT LENGTH DATA SET TOTAL INERTIA AXIS 1 AXIS 2 AXIS 1 AXIS 2 A 0.726 0.477 5.847 4.153 12.242 B 0.470 0.351 3.296 3.112 4.718 with Fisher's exact test. This test compares the absolute constancy of taxa in the group or the groups it differentiates positively, with the one in the groups it differentiates negatively or not differentiating at all. The level of significance chosen in Fisher's exact tests was 0.05. The classification was made using the TWINSPAN algorithm within JUICE ver. 7 (Tichy 2001), and an ordination was carried out with CANOCO ver. 4.5 (ter Braak & Smilauer 2002). 4. RESULT AND DISCUSSION 4.1 Classification and syntaxonomy Six vegetation units of grasslands could be distinguished based on the TWINSPAN classification (Table 2). Specifically, from the eight groups produced at the third level of divisions in TWIN-SPAN, two pairs of groups were fused to one group each, on the basis of their floristic similarity, as well as the DCA ordination diagrams. In Figure 3 the divisions made by TWINSPAN are presented schematically. The first unit belongs to Koelerio-Corynepho-retea, as easily concluded from the dominance of Koelerio-Corynephoretea diagnostic taxa (see Table 3), and it is well differentiated from the other vegetation units of dry grasslands in the study area. The most frequent species in this vegetation unit are Silene conica, Bromus rubens, Cruciata pedemon-tana, Trifolium arvense, Vulpia myurus, Trifolium arvense and Linaria genistifolia. Most of these species are annuals and diagnostics (according to Mucina 1997, Dengler 2000, Rusina 2005, Kuzemko 2009) for the subclass Koelerio-Corynephorenea (Klika in Klika & Novak 1941) Dengler in Dengler et al. 2003 and/or the order Festuco-Sedetalia acris sensu Dengler (2003) of the class Koelerio-Corynephoretea. Due to the occurrence of Silene conica and Cerastium semidecandrum (diagnostic species of the alliance Sileno conicae-Cerastion semidecandri Korneck 1974 according to Ellenberg (1988), Dengler et al. (2003) and Faust et al. (201l)), as well as the eastern Balkan endemic species Silene frivald-szkyana and Erysimum microstylum, all relevés of the first vegetation unit were classified within the alliance Sileno conicae-Cerastion semidecandri s.l. and named the Silene frivaldszkyana-Eysimum microstylum community. We use here the wide concept of the Sileno conicae-Cerastion semidecan-dri, which in the delimitation of Dengler (2001, 2003) includes all the annual-dominated pioneer grasslands on dry calcareous sands in temperate Europe. Dengler (2001, 2003) also includes the Bassio laniflorae-Bromion tectorum (Soo 1957) 95 plots Group 1,2,3,4,5 Group 6 Group 1,2,3,4 Group 5 Group1,2 Group 3, 4 Group 1 Group 2 Group 3 Group 4 Figure 3: Schematic representation of the divisions made by TWINSPAN. Slika 3: Shematski prikaz členitve, narejene s programom TWINSPAN. Borhidi 1996 in the Sileno-Cerastion s.l., while other authors (e.g. Rodwell et al. 2002) accept it as a separate alliance. According to its authors, the Bassio-Bromion comprises annual-dominated pioneer grasslands of calcareous sands in the Pan-nonian Basin, thus would be a vicariant alliance to the Sileno-Cerastion s. str. distributed further north and west, and also shows some floristic and ecological similarities to our Silene frivaldszkyana-Erysimum microstylum community. Until the justification of two separate vicariant alliances of sub-continental sand pioneer grasslands has been proven and their delimitation established, we prefer to assign our community provisionally to the Sileno conicae-Cerastion semidecandri in the wide sense, which is also supported by the fact that its two eponymous taxa Silene conica and Cerastium semidecandrum are widespread in the stands examined in Prespa National Park. The alliance Sileno conicae-Cerastion semidecan-dri s.l. is classified within the order Festuco-Sede-talia acris and Koelerio-Corynephorenea subclass of the Koelerio-Corynephoretea class (Dengler 2003, Dengler et al. 2003, Berg et al. 2004). Another syntaxonomical issue is that Pedashenko et al. (2013) report Silene frivaldszkyana from Bulgaria as typical for several alliances of the Festuco-Bro-metea, but also for the Koelerio-Corynephoretea (in the latter with higher abundance). In Prespa National Park, as already mentioned, Silene frivald-szkyana is found exclusively and with high abundance in the Koelerio-Corynephorenea stands. We therefore decided to classify this vegetation type to a new community, as it is differentiated floris-tically from all of the associations of the alliance described so far, such as Erodio-Senecionetum ver-nalis Luhrs 1993 and Sileno conicae-Cerastietum semidecandri Korneck 1974 mentioned by Dengler (2000) for Germany, mainly, because it is rich in Balkan endemic taxa. The Sileno conicae-Cerastion semidecandri s.l. and the Festuco-Sedetalia acris according to the Manual of the European Union Habitats (European Commission 2013) largely correspond to two habitat types, namely *6120 (Xeric sand calcareous grasslands) and *6260 (Pannonic sand steppes). Although our first thoughts were to include the described community in the former habitat type, we have chosen finally to classify it, at least provisionally, to the *6260 habitat type. The reason for this choice is that *6120 has been identified mainly in central and northern Europe (e.g. Germany and Poland (Odman et al. 2011), Norway, Sweden (European Commission 2013), Denmark (Silva et al. 2008), Belgium (Demold-er et al. 2008)), while *6260 is given for eastern Central Europe (Hungary, Slovakia, Austria and Czech Republic (Sefferova Stanova et al. 2008 and references therein)), but also for Southeast Europe (e.g. Romania (European Commission 2013), Bulgaria (Bulgarian Academy of Sciences 2011) and Serbia (Sefferova Stanova et al. 2008)). According to Sundseth (2009), complex climatic patterns, formed due to Mediterranean, Carpathian and Alpic influences, cause the Pannonian region to "exhibit a mosaic vegetation structure instead of the more classic zonal arrangements that one sees in other biogeographical regions". Moreover, Metzger et al. (2012) in the environmental stratification of the European continent described the stratum Pannonian 2 (PAN2) of semi-arid plains as having three areas of distribution: one in the middle Danube plain, one in the Black Sea lowlands and one in the Valley of the Struma in Bulgaria and Greece. In this sense, the area of Prespa potentially exhibits a Panonian-like character, at least concerning the vegetation of this specific location (study area). The second column of the TWINSPAN analysis (Table 2) represents the Thero-Brachypodietea class, since most of the taxa or plant species are diagnostic of the class (e.g. Arenaria leptoclados, Plantago holosteum). It was found in very dry areas of the study area, mostly nearby villages. The class is distinguished from the other classes by annual species, mostly grasses (e.g. Taeniatherum caput-medusae, Aegilops triuncialis, Catapodium rigidum, Lagurus ovatus). The third column represents the Stellarietea mediae class, which is distinguished from the other classes by synanthropic species such as Medicago sativa and M. rigidula, and is dominated by diagnostic taxa of the class, such as Convolvulus arvensis, Cerastium glomera-tum, Echium italicum and Geranium molle. It seems that in the study area both classes, Thero-Brach-ypodietea and Stellarietea mediae, are floristically related to the Koelerio-Corynephoretea, since they mainly represent annual communities affected by human activities. Columns 4, 5, 6 represent the Festuco-Brometea, Daphno-Festucetea and Juncetea trijidi classes, respectively. These three classes are characterized by perennial herbaceous species. The Festuco-Brometea is distinguished from other classes by the presence of perennial grasses such as Phleum phleoides and P. pratense. In the national park, the Festuco-Brometea is represented by several communities, probably of the order Astragalo-Potentilleta-lia Micevski 1971. These communities are found at intermediate altitudes, on sites with moderate grazing. The Daphno-Festucetea is characterized and well differentiated from the other classes by chamaephytes and spiny species (e.g. Astragalus angustifolius, Morina persica, Eryngium amethysti-num, Prunus prostrata). The Juncetea trifidi is well differentiated from the other classes by acido-philic grasses, such as Nardus stricta, Deschampsia flexuosa and Alopecurus gerardii (Table 2). 4.2 Ordination The first DCA diagram (Figure 4; Table 1) shows that grasslands of the Koelerio-Corynephoretea in the national park are clearly discriminated from the Festuco-Brometea, Daphno-Festucetea and Juncetea trifidi classes. On the other hand, the DCA diagram reveals the floristic affinity of Koelerio-Corynephoretea with the vegetation units of Stellarietea mediae and Thero-Brachypodietea. The first axis in the DCA diagram may represent a moisture gradient, with the driest ecological conditions occurring on the left side of the axis. The second DCA axis mainly represents the geological substrate, with the relevés on siliceous substrates occurring in the lower part of the axis and those on calcareous substrates in the upper part of the axis. The second DCA diagram (Figure 5) shows a clear floristic differentiation of the grasslands of Koelerio-Corynephoretea from the vegetation of Thero-Brachypodietea and Stellarietea mediae. The first DCA axis in this diagram is related to Si T-B + 1 □ 2 < 3 V n Incl Ca □S3 |Alt J? \ 1 s' 4 K-C .................ii..........<................. Sm\ Al + 1 □ 2 ^ D-F Ca w \\ • \ V • Nhr < 3 X 4 • 5 ■ 6 lncl\ x >« Alt ^ —»-Jt Sm"*^ \ ^^Si T-B / S-B K-C Al Figure 4: Diagram of all relevés along the first two DCA axes (1: Koelerio-Corynephoretea, 2: Thero-Brachypodietea, 3: Stellarietea mediae, 4: Festuco-Brometea, 5: Daphno-Festucetea, 6: Juncetea trifidi) with explanatory variables passively projected onto the ordination space. The explanatory variables concern the sum per relevé of the relative abundances of the species considered as diagnostic of the classes Koelerio-Corynepho-retea (K-C), Thero-Brachypodietea (T-B), Stellarietea mediae (Sm), Festuco-Brometea (F-B), Daphno-Festucetea (D-F), Juncetea trifidi (Jt), as well as the environmental variables inclination (Incl), altitude (Alt) and geological substrate (Al: Alluvial, Si: Siliceous, Ca: Calcareous). Slika 4: Diagram vseh popisov vzdolž prvih dveh DCA osi (1: Koelerio-Corynephoretea, 2: Thero-Brachypodietea, 3: Stellarietea mediae, 4: Festuco-Brometea, 5: Daphno-Festucetea, 6: Juncetea trifidi) z nedovisnimi spremenljivkami, pasivno pro-jeciranimi na ordinacijski prostor. Neodvisne spremenljivke prikazujejo vsoto relativnih abundanc na popis diagnostičnih vrst razredov Koelerio-Corynephoretea (K-C), Thero-Brach-ypodietea (T-B), Stellarietea mediae (Sm), Festuco-Brome-tea (F-B), Daphno-Festucetea (D-F), Juncetea trifidi (Jt) in okoljskih spremenljivk: naklon (Incl), višina (Alt) in geološka podlaga (Al: aluvij, Si: silikat, Ca: karbonat). Figure 5: Diagram of the relevés classified within the classes Koelerio-Corynephoretea (1), Thero-Brachypodietea (2) and Stellarietea mediae (3) with explanatory variables passively projected onto the ordination space. For the abbreviations of explanatory variables see Figure 4. Slika 5: Diagram popisov, uvrščenih v razrede Koelerio-Corynephoretea (1), Thero-Brachypodietea (2) in Stellarietea mediae (3), z neodvisnimi spremenljivkami, pasivno projiciranimi na ordinacijski prostor. Oznake spremenljivk so kot na Sliki 4. the calcareous substrate and might additionally reflect a disturbance gradient, as the class Stella-rietea mediae contains anthropogenic vegetation units. The second DCA axis is related to the type of substrate (alluvial sands vs. bedrock), as well as with altitude (the relevés of Koelerio-Corynepho-retea occur at lower altitudes where the substrate is alluvial). The floristic affinities of the Koelerio-Corynephoretea grasslands with those of the Stella-rietea mediae have also been documented in other studies. Stroh et al. (2002) recorded that ruderal pioneer communities of Stellarietea mediae have been established after spontaneous succession in sand grasslands belonging to the Koelerio-Corynephoretea, while the important role of thero-phytes in Koelerio-Corynephoretea communities is also supported by Zwaenepoel et al. (2002) and Faust et al. (2011). 4.3 Ecological characteristics of the habitat type *6260 in Greece and conservation issues The occurrence of the *6260 habitat type has been documented mainly from Hungary, but its distribution extends into Lower Austria, Slovakia, Romania and Bulgaria (Sefferova Stanova et al. 2008). The habitat type holds several vegetation types classified in the order Festuco-Sedetalia acris, which represents the subcontinental and continental sand-swards. It represents dry open grasslands developed on mobile or fixed dunes (alluvial sands, subfossil dune systems) of the Pannonic steppes (European Commission 2013). The sub-continental climate type of the study site, and the sandy shores formed by the active hydrological network in the area offer the two main ecological conditions required for the existence of the *6260 habitat type. Specifically, the Silene frivaldszkyana-Erysimum microstylum [Sileno-Cerastion] community was found on sandy soils on the south-southeastern shores of Lake Megali Prespa (Figure 2) and on the Slogi islet in Lake Mikri Prespa (Figure 1). The altitude of the localities varies from 849 m a.s.l. (the water level of Megali Prespa was ap-prox. 847 m a.s.l. during the sampling period of this study) to 855 m a.s.l. (water level of Mikri Prespa was approx. 854 m a.s.l.). The sand dunes on the shores of Lake Megali Prespa were formed by the gradual decrease of the water level of the lake over the last 30 years, and partially by me- chanical excavation works for sand extraction at specific localities in recent decades. The sandy islet of Slogi is located 1-2 m above the water level of Lake Mikri Prespa and probably resulted from historical alluvial deposits of the Agios Germanos River before its anthropogenic diversion to the larger lake. Grasslands of the Koelerio-Corynephoretea are distinguished from those of the Stellarietea mediae and Thero-Brachypodietea, because they are found in areas with almost flat and deep alluvial sandy soils (Figure 5). The stands of the Silene frivaldszkyana-Erysi-mum microstylum community are highly influenced by the fluctuations of the water levels of the lakes. The high abundance of Scirpoides cf. holoschoenus in this community reveals the influence of underground water on this vegetation unit. Moreover, many ephemeral ponds that are occasionally flooded during winter and spring, especially at the isthmus separating the two lakes, favour the dominance of Azolla filiculoides. Given on the one hand the high susceptibility of the *6260 to degradation due to competition and, on the other hand, the uniqueness of the habitat for Greece, special conservation measures are needed to ensure its continued survival in the area. In the study area, the major threats to the habitat type are water level fluctuations caused by anthropogenic influences (e.g. excavations, embankments), invasion of non-native species and the absence of grazing. We assume that any further increase in the water level of the lakes will negatively affect the *6260 habitat type in the study area. However, detailed studies are needed to define a sustainable water level to maintain a balance between safeguarding the habitat type and meeting the irrigation needs of local farmers. The existence of short, shrub-like trees of Sa-lix alba as well as of trees like Robinia pseudoacacia and Morus spp., planted at the localities of the habitat type, may be a threat because of the possible invasion of these species in the habitat. The invasion of Robinia pseudoacacia has been recorded as a serious threat for *6260 in Slovakia (Sefferova Stanova et al. 2008). This species, due to its intensive vegetative spread and its high germination rates after fire, becomes a superior competitor in secondary succession processes. Systematic eradication of such invasive trees is suggested for the study area by means of mechanical cleaning (cutting), subsequent grazing of surviving shoots and monitoring of its regeneration (Kelemen & Warner 1996). The soil surface of the stands of the community is often covered by a conspicuous cryptogam layer (e.g. lichens of the genus Cladonia and xero-phytic mosses of the genus Tortula s.l.), which is typical for the habitat type in general (Sefferova Stanova et al. 2008). Such relation of inland dunes and dense moss layers has already been indicated for ecologically similar habitat types (e.g. for *6120 by Odman et al. (2011)). The study area is only occasionally grazed by cattle. Abandonment of grazing may lead to closed grasslands and increased moss cover (Molnar 2003). Animal trampling can prevent the formation of dense moss carpets. However overgrazing must be avoided, as this will enrich the sand with nutrients, facilitating the occurrence of some diagnostic species of Stellarietea mediae or Molinio-Arrhenatheretea, such as Plantago lanceolata. On the other hand, the cessation of grazing on sand dunes is expected to cause increased biomass production and subsequently higher litter and humus accumulation. Such changes will further favour generalists and may cause the local extinction of some open sand grassland species (Onodi et al. 2006). It has been reported that grasslands suitable for open Pannonic sand steppes in Hungary turn to weedy fields due to grassland abandonment (Torok et al. 2000). Consequently, a proper stocking rate must be estimated and maintained in the study area for conservation purposes. 5. CONCLUSIONS This paper reports the southernmost occurrence of the priority habitat type *6260 "Pannonic sand steppes" in the Balkan Peninsula, from sandy soils at the south-southeastern shores of Lake Megali Prespa and on the Slogi islet in Lake Mikri Prespa, in the Prespa National Park of Greece. The prevailing sub-continental climate, the gradual drop of the water level of Lake Meg-ali Prespa over the past few decades, the alluvial deposits of the river Agios Germanos in Lake Mikri Prespa and, to some extent, mechanical excavation works on the shores of Lake Megali Prespa in recent decades are considered as factors that positively influenced this habitat type. Syntaxonomically, the stands are classified as Silene frivaldszkyana-Erysimum microstylum community within the alliance Sileno conicae-Cerastion semidecandri s.l. (including Bassio laniflorae-Bro-mion tectorum), the order Festuco-Sedatalia acris (= continental grasslands of base-rich, nutrient-poor sands) and the class Koelerio-Corynephoretea. One of the most important factors for the sustainabil-ity of the habitat type is grazing; undergrazing will probably lead to the dominance of moss carpets, which is a typical feature of the habitat type under no vegetation management, while overgrazing will affect vegetation by increasing nutrient content of the soils, which will probably favour the invasion of diagnostic species of Stellarietea mediae or Molinio-Arrhenatheretea. Finally, the control of the water level against rising in Lake Mikri Prespa is considered crucial since this will probably favour the invasion of species such as Phragmites australis and Azolla filiculoides that would eventually alter the vegetation type. In Lake Megali Prespa the water level cannot be artificially controlled, but the areas occupied by the habitat type do not appear threatened by a potential water level rise as they lie 3-5 m above the present water level. Finally, the spread of "aggressive" trees, like Robinia pseudoacacia, from plantations must be controlled. ACKNOWLEDGMENTS The research was partly financed by the project "Record, Assessment and Geographical Presentation of the Range and Forest Habitat Types of the Natura 2000 Sites Prespa National Park (Ethnikos Drymos Prespon-GR 1340001), Mt. Varnountas (Ori Varnountas-GR 1340003) and Adjacent Areas", managed by the Society for the Protection of Prespa. We thank Dr. Jürgen Den-gler and three anonymous reviewers for their constructive comments in improving the manuscript. Finally, we thank Laura Sutcliffe (Göttingen) for linguistic editing and EDGG for covering the costs of this with an IAVS grant. REFERENCES Barkman, J., Doing, H. & Segal, S. 1964: Kritische Bemerkungen und Vorschläge zur quantitativen Vegetationsanalyse. Acta Botanica Neer-landica 13: 394-419. Berg, C., Dengler, J., Abdank, A. & Isermann, M. (eds.) 2004: Die Pflanzengesellschaften Mecklenburg-Vorpommerns und ihre Gefährdung. Textband. Weissdorn, Jena, 606 pp. Bergmeier, E., Konstantinou, M., Tsiripidis, I. & Sykora, K. V. 2009: Plant communities on metalliferous soils in northern Greece. Phyto-coenologia 39: 411-438. Boch, S. & Dengler, J. 2006: Floristische und ökologische Charakterisierung sowie Phytodi-versität der Trockenrasen auf der Insel Saare-maa (Estland). Arbeiten aus dem Institut für Landschaftsökologie Münster 15: 55-71. Borhidi, A., Kevey, B. & Varga, Z. 2001: Checklist of the higher syntaxa of Hungary. Annals of Botany (N. S.) 1: 159-166. Braun-Blanquet, J. 1964: Pflanzensoziologie. Grundzüge der Vegetationskunde. 3rd. ed. Springer, Wien, 865 pp. Bulgarian Academy of Sciences. 2011: Red Data Book of the Republic of Bulgaria-Digital edition. Published on the Internet: http://e-ecodb. bas.bg/rdb/en/vol3 [accessed 31/10/2013]. Chytry, M. & Otypkova, Z. 2003: Plot sizes used for phytosociological sampling of European vegetation. Journal of Vegetation Science 14: 563-570. Chytry, M. (ed.) 2007: Vegetation of the Czech Republic-1. Grassland and heathland vegetation [in Czech, with English summaries]. Academia, Praha, 526 pp. Dafis, S., Papastergiadou, E., Lazaridou, T. & Tsia-fouli, M. 2001: Technikos Odigos Anagnorisis, Perigrafis kai Chartografisis Tipon Oikotopon tis Elladas (Technical guide for identification, description and mapping of habitat types of Greece). Greek Wetland and Biotope Centre (EKBY), Thessaloniki, 908 pp. Demolder, H., Delescaille, L. M., Van Landuyt, W., Wouters J., Van Looy, K. & Paelinckx, D. 2008: Conservation status of the Natura 2000 habitat 6120 (Xeric sand calcareous grasslands) for the Belgian Atlantic region. In: Pa-elinckx D., Van Landuyt W. & De Bruyn L. (eds.): Conservation status of the Natura 2000 habitats and species. Report of the Research Institute for Nature and Forest, IN-BO. R. 2008. 15, Brussels, pp. 1-4. Dengler, J. 2000: Synsystematische Stellung und Gliederung deruckermärkischen Sandtrockenrasen. Berichte der Naturwissenschaftlichen Gesellschaft Bayreuth 24: 302-306. Dengler, J. 2001: Erstellung und Interpretation synchorologischer Karten am Beispiel der Klasse Koelerio-Corynephoretea. Berichte der Reinhold-Tüxen-Gesellschaft 13: 223-228. Dengler, J. 2003: Entwicklung und Bewertung neuer Ansätze in der Pflanzensoziologie unter besonderer Berücksichtigung der Vegetationsklassifikation. Archiv Naturwissenschaftlicher Dissertationen 14: 1-297. Dengler, J., Berg, C., Eisenberg, M., Isermann, M., Jansen, F., Koska, I., Löbel, S., Manthey, M., Päzolt, J., Spangenberg, A., Timmermann, T. & Wollert, H. 2003: New descriptions and typifications of syntaxa within the project "Plant communities of MecklenburgVorpommern and their vulnerability". Part I. Feddes Repertorium 114: 587-631. Dengler, J., Jansen, F., Glöckler, F., Peet, R. K., De Cáceres, M., Chytry, M., Ewald, J., Olde-land, J., Lopez-Gonzalez, G., Finckh, M., Mu-cina, L., Rodwell, J. S., Schaminée, J. H. J. & Spencer, N. 2011: The Global Index of Vegetation-Plot Databases (GIVD): a new resource for vegetation science. Journal of Vegetation Science 22: 582-597. Dierschke H. 1994: Pflanzensoziologie-Grundlagen und Methoden. Ulmer, Stuttgart, 683 pp. Dimopoulos, P., Raus, T., Bergmeier E., Constan-tinidis, T., Iatrou G., Kokkini, S., Strid, A. & Tzanoudakis, D. 2013: Vascular Plants of Greece. An annotated checklist. Botanic Garden and Botanical Museum Berlin-Dahlem & Hellenic Botanical Society, Berlin & Athens, 372 pp. Donitä, N., Popescu, A., Paucä-Comänescu, M., Mihäilescu, S. & Biri§, I. A. 2005: Habitele din Romania. Edit. Tehnicä Silvicä, Bucure§ti, 500 pp. Ellenberg, H. 1988: Vegetation Ecology of Central Europe, 4th ed. University of Cambridge, Cambridge, 705 pp. European Commission. 2013: Interpretation Manual of European Union Habitats. European Union, Bruxelles, 146 pp. Faust, C., Süss, K., Storm, C. & Schwabe, A. 2011: Threatened inland sand vegetation in the temperate zone under different types of abiotic and biotic disturbances during a ten-year period. Flora 206: 611-621. Fotiadis, G., Angelova, N., Nikolov, N., Melovski, Lj., Karadelev, M., Avukatov, V. & Nikolov, L. 2012: Conservation Action Plan for Grecian Juniper Forests in the Prespa Lakes Watershed (Final Report). UNDP/GEF project "Integrated ecosystem management in the Prespa lakes basin", F. Y. R. of Macedonia, Skopje, 46 pp. Hill, M. O. 1979: TWINSPAN - a FORTRAN program for arranging multivariate data in an ordered two-way table by classification of the individuals and the attributes. Ecology & Systematics, Cornell University, Ithaca, NY, 45 pp. Hollis, G. E. & Stevenson, A. C. 1997: The physical basis of the lake Mikri Prespa systems: geology, climate, hydrology and water quality. Hydrobiologia 351: 1-19. I. G. M. E. [Institute of Geology and Mineralogy Research] 1983: Geologikos Chartis tis Elladas (Geological Map of Greece), scale 1 : 500.000. Athens. Kelemen, J. & Warner, P. 1996: Nature Conservation Management of Grasslands in Hungary. Summary. Conservation Handbook Series of the Hungarian National Authority for Nature Conservation, Budapest, 39 pp. Kent, M. & Coker, P. 1992: Vegetation description and analysis. A practical approach. John Wiley, London, 336 pp. Knapp, R. 1984: Sample (relevé) areas (distribution, homogeneity, size, shape) and plot-less sampling. In: Knapp, R. (ed.), Handbook of Vegetation Science 4. Junk, The Hague, pp. 101-119. Kuzemko, A. 2009: Dry grasslands on sandy soils in the forest and forest-steppe zones of the plains part of Ukraine: present state of syntax-onomy. Tuexenia 29: 369-390. Laime, B. & Tjarve, D. 2009: Grey dune plant communities (Koelerio-Corynephoretea) on the Baltic coast in Latvia. Tuexenia 29: 409-435. Löbel, S. & Dengler, J. 2007: Dry grasslands communities on southern Öland: phytosociology, ecology, and diversity. Acta Phytogeographica Suecica 88: 13-31. Matuszkiewicz, W. 2008: Przewodnik do ozna-czania zbiorowisk roslinnych Polski. Wydaw-nictwo Naukowe PWN, Warszawa, 537 pp. Metzger, M. J., Shkaruba, A. D., Jongman, R. H. G. & Bunce, R. G. H. 2012: Descriptions of the European Environmental Zones and Strata. Alterra Report 2281. Wageningen, 154 pp. Molnâr Z. (eds.) 2003. Dry sand vegetation of the Kiskunsag. Termeszet BUVAR Alapitvany Ki-ado, Budapest, 30 pp. Mucina, L., Grabherr, G. & Ellmauer, T. (eds.) 1993: Die Pflanzengesellschaften Österreichs. Teil I: Anthropogene Vegetation. Fischer, Jena, 578 pp. Mucina, L. 1997: Conspectus of classes of European vegetation. Folia Geobotanica et Phyto-taxonomica 32: 117-172. Odman, A. M., Mártensson, L.-M., Sjoholm, C. & Olsson, P. A. 2011: Immediate responses in soil chemistry, vegetation and ground beetles to soil perturbation when implemented as a restoration measure in decalcified sandy grassland. Biodiversity and Conservation 20: 3039-3058. Onodi, G., Kertesz, M. & Botta-Dukat, Z. 2006: Effects of simulated grazing on open perennial sand grassland. Community Ecology 7: 133-141. Parolly, G. 2003: Towards common standards in phytosociological papers submitted to the Turkish Journal of Botany: A Letter from the Editor. Turkish Journal of Botany 27: 163-165. Pedashenko, H., Apostolova, I., Boch, S., Ganeva, A., Janisová, M., Sopotlieva, D., Todorova, S., Ünal, A., Vassilev, K., Velev, N. & Dengler, J. 2013: Dry grasslands of NW Bulgarian mountains: first insights into diversity, ecology and syntaxonomy. Tuexenia 33: 309-346. Quézel, P. 1969: La vegetation du massif de Bela Voda (Macedoine Nord-Occidentale). Biologia Gallo-Hellenica 2: 91-112. Ristevski, P. 2000: Climatic and Agroclimatic Characteristics in the Prespa Lake Basin. International Symposium-Sustainable Development of Prespa Region, Otesevo, pp. 212-223. Rodwell, J. S., Schaminée, J. H. J., Mucina, L., Pignatti, S., Dring, J. & Moss, D. 2002: The diversity of European vegetation-An overview of phytosociological alliances and their relationships to EUNIS habitats. Rapp. EC-LNV 2002 (054): National Reference Centre for Agriculture, Nature and Fisheries, Wageningen, 168 pp. Rüsina, S. 2005: Diagnostic species of mesophy-lous and xerophylous grassland plant communities in Latvia. Latvijas Universitates Raksti 685: 69-95. Schaminée, J. H. J., Stortelder, A. H. F., Weeda, E. J. (eds.) 1996: De Vegetatie von Nederland-Deel 3. Plantengemeenschappen van graslanden, zomen en droge heiden. Opulus, Uppsala, 360 pp. Sefferova Stanova, V., Vajda, Z. & Janak, M. 2008: Management of Natura 2000 habitats. 6260 *Pannonic sand steppes. European Commission, Bruxelles, 24 pp. Silva, J. P., Toland, J., Jones, W., Eldridge, J., Thorpe, E. & O'Hara, E. 2008: LIFE and Europe's grasslands: Restoring a forgotten habitat. European Commission, 56 pp. Stroh, M., Storm, C., Zehm, A. & Schwabe, A. 2002: Restorative grazing as a tool for directed succession with diaspore inoculation: the model of sand ecosystems. Phytocoenologia 32: 595-625. Sundseth, K. 2009: Natura 2000 in the Pannonian Region. European Commission, 12 pp. ter Braak, C. J. F. & Smilauer, P. 2002: CANOCO Reference Manual and CanoDraw for Windows User's Guide: Software for Canonical Community Ordination (version 4.5). Microcomputer Power, Ithaca, 351 pp. Tichy, L. 2001: JUICE, software for vegetation classification. Journal of Vegetation Science 13: 451-453. Torok, K., Szili-Kovacz, T., Halassy, M., Toth, T., Hayek, Z., Paschke, M. W. & Wardell, L. J. 2000: Immobilization of soil nitrogen as a possible method for the restoration of sandy grassland. Applied Vegetation Science 3: 7-14. Tsiripidis, I., Bergmeier, E., Fotiadis, G. & Dimo-poulos, P. 2009: A new algorithm for the determination of differential taxa. Journal of Vegetation Science 20: 233-240. van der Maarel, E. 1979: Transformation of cover-abundance values in phytosociology and its effect on community similarity. Vegetatio 39: 97-114. Vassilev, K., Stevanovic Dajic, Z., Custerevska, R., Bergmeier, E. & Apostolova, I. 2012: Balkan Dry Grasslands Database. In: Dengler J., Oldeland J., Jansen F., Chytry M., Ewald J., Finckh M., Glockler F. Lopez-Gonzalez G, Peet R. K., Schaminee J. H. J., Vegetation databases for the 21st century. Biodiversity & Ecology 4: 330-330. Vicherek, J. 1972: Die Sandpflanzengesellschaften des unteren und mittleren Dnjeprstromge-bietes (die Ukraine). Folia Geobotanica et Phytotaxonomica 7: 9-46. Vrahnakis, M., Fotiadis, G. & Kazoglou, Y. 2011: Katagrafi, Aksiologisi kai Geografiki Apoti-posi ton Livadikon kai Dasikon Tipon Oiko-topon ton Periochon tou Diktiou Phisi 2000: (Ethnikos Drymos Prespon - GR 1340001), Mt. Varnountas (Ori Varnountas - GR 1340003) kai periochon periks auton (Registration, Assessment and Geographical Representation of the Range and Forest Habitat Types of the Natura 2000 Sites Prespa National Park (Ethnikos Drymos Prespon - GR 1340001), Mt. Varnountas (Ori Varnountas -GR 1340003) and Adjusted Areas) - Final Report. TEI of Larissa, Society for the Protection of Prespa, 107 pp. + annexes. Zwaenepoel, A., T'Jollyn, F., Vandenbussche, V. & Hoffmann, M. 2002: Systematiek van natu-urtypen voor Vlaanderen: 6.2 Graslanden, Droge graslanden. Verslag van het Instituut voor Natuurbehoud, Bruxelles, 21 pp. Received: 27. 2. 2013 Accepted: 8. 3. 2014 Co-ordinating editor: Jürgen Dengler Table 2: Constancy synoptic table of the differential taxa of the six dry grassland vegetation units. Dark grey color indicates positive differentiation, medium grey color indicates positive-negative differentiation, white color and italic letters indicates negative differentiation and light grey color indicates non-differentiation (see Tsiripidis et al. 2009); 1: overall constancy, 2: Koelerio-Corynephoretea, 3: Thero-Brachypodietea, 4: Stellarietea mediae, 5: Festuco-Brometea, 6: Daphno-Festucetea, 7: Juncetea trifidi. Only species with statistically higher constancy (Fisher's exact test; a = 0.05) in the groups that differentiate positively vs. those that differentiate negatively or not differentiating at all are presented as differential. Species with a constancy lower than 20% in any column, and companion species with overall constancy lower than 10%, have been omitted. Tabela 2: Sinoptična tabela s stalnostjo razlikovalnih taksonov šestih vegetacijskih tipov suhih travišč. Temno siva barva prikazuje pozitivno razlikovanje, srednje siva barva nakazuje pozitivno-negativno razlikovanje, bela barva in poševna pisava nakazujeta negativno razlikovanje in svetlo siva barva označuje nobenega razlikovanja (Tsiripidis et al. 2009); 1: celotna stalnost, 2: Koelerio-Corynephoretea, 3: Thero-Brachypodietea, 4: Stellarietea mediae, 5: Festuco-Brometea, 6: Daphno-Festucetea, 7: Juncetea trifidi. Samo vrste s statistično visoko stalnostjo (Fisherjev natančni test; a = 0.05) v skupinah, ki se pozitivno razlikujejo napram tistim, ki se razlikujejo negativno ali sploh ne, so prikazane kot razlikovalne. Vrste s stalnostjo, manjšo od 20% v enem stolpcu in spremljevalke s celotno stalnostjo, manjšo od 10%, niso prikazane. Vegetation unit 5 6 Number of relevés Geology (Sa: sandy, Al: alluvial, Si: siliceous, Ca: calcareous) 95 12 Sa Si, Ca, Al 5 Al 35 Si, Ca, Al 14 Ca 20 Si Mean altitude (m a.s.l.) 853 987 976 1300 1490 1900 Altitudinal range (m a.s.l.) 848-857 850-1110 850-1260 850-1940 1310-1760 1730-2060 Mean number of taxa per relevé 33 43.7 44.0 28.0 34.0 43.5 24.0 Differential species (p < 0.01) Silene frivaldskyana 13 100 Euphorbia baselicis 9 75 20 Scirpoides cf. holoschoenus 9 75 Cruciata pedemontana 24 100 22 40 17 7. Hypericum perforatum 19 83 22 20 14 Torilis arvensis 18 75 11 20 6 29 . Draba muralis 17 67 11 20 17 Delphinium balcanicum 8 58 7. Medicago minima 12 42 11 20 9 7. Papaver rhoeas 6 42 3 Verbascum cf. eriophorum 7 42 11 7. Vicia lathyroides 9 42 11 9 Linaria genistifolia subsp. sofiana 18 92 40 11 Jasione heldreichii 19 67 14 25 Tordylium maximum 11 42 6 21 Silene conica 36 100 100 23 36 Bromus rubens 27 100 100 14 Erysimum microstylum 31 92 67 20 20 29 Poa bulbosa 59 92 100 40 63 64 15 Filago arvensis 27 75 100 20 7. Centaurea cuneifolia 17 75 33 20 9 Herniaria incana 16 50 44 14 Petrorhagia saxifraga 18 50 44 14 14 Aira elegantissima 12 42 22 11 Stachys angustifolia 7 33 22 3 Hypericum olympicum 17 42 33 17 14 Trifolium hirtum 21 42 100 17 Trifolium arvense 41 92 100 20 46 14 . 1 2 3 4 7 9 Vegetation unit 1 2 3 4 5 6 7 Thymus sibthorpii 58 75 67 40 77 21 40 Astragalus onobrychis 29 75 44 40 7 . Cerastium brachypetalum subsp. roeseri 14 25 44 3 36 Avena sterilis 27 92 100 40 11 Cichorium intybus 14 42 22 40 11 Alyssum chalcidicum 9 25 33 20 6 Convolvulus cantabrica 9 25 33 20 6 Anchusa officinalis 18 83 11 40 11 Eryngium campestre Vulpia myurus Potentilla argentea 53 29 15 100 75 25 100 44 60 40 20 66 37 26 21 . -t-t 11 Bromus hordeaceus 44 100 100 40 37 43 Sanguisorba minor 40 83 67 20 49 29 Achillea coarctata 27 58 56 40 26 21 Bromus squarrosus 24 58 33 20 23 29 Trifolium scabrum 35 50 33 20 51 36 Sedum acre 23 42 56 20 20 29 Galium verum subsp. verum 26 67 11 23 57 Melica ciliata 16 42 22 9 36 Anthemis arvensis subsp. arvensis 18 42 20 17 36 Dasypyrum villosum 29 33 100 23 50 Trifolium campestre 29 25 44 37 50 5 Euphorbia myrsinites 46 83 67 40 100 Rumex acetosella 51 100 33 40 40 85 Centaurea alba 13 8 22 40 20 Hordeum murinum 11 56 60 6 Sedum caespitosum 11 44 40 11 Cynosurus echinatus 28 17 89 60 29 21 5 Echium italicum 7 8 22 40 6 Anthemis cretica subsp. carpatica 14 22 29 5 Bellardia latifolia 14 8 33 26 Chamaecytisus eriocarpus 26 22 46 35 Dactylis glomerata 17 33 26 29 Xeranthemum annuum 17 22 23 43 Crepis sancta 20 8 44 29 29 Crupina vulgaris 27 100 20 71 Arenaria leptoclados 32 100 20 29 71 Festuca valesiaca 39 8 33 57 43 35 Taeniatherum caput-medusae 16 33 67 20 6 14 . Convolvulus elegantissimus 11 78 20 6 Plantago holosteum 11 67 9 5 Aegilops triuncialis 14 56 20 17 7. Trifolium glomeratum 4 33 3 Catapodium rigidum 3 33 Lagurus ovatus 3 33 Medicago orbicularis 4 33 3 Convolvulus arvensis 6 17 60 3 Tragopogon porrifolius subsp. 7 17 60 6 eriospermus Medicago rigidula 8 11 60 9 7. Medicago sativa 8 11 60 9 7. Centaurea salonitana 11 11 20 9 36 Vegetation unit 4 5 6 7 20 14 55 20 43 75 20 23 7 25 40 26 60 31 7 20 23 5 26 7 60 29 14 26 5 20 23 29 50 5 23 36 . 6 100 . 6 100 93 20 6 86 79 15 . 11 79 20 23 64 5 . 14 64 . 9 57 57 50 .6 50 50 . 3 50 .6 50 50 . 17 43 43 . 3 43 43 36 36 .9 36 36 36 29 29 .3 29 .6 29 29 29 21 21 21 21 .3 21 95 90 .3 60 Agrostis canina 18 Pilosella leucopsilon 33 Trisetum flavescens 16 Bromus sterilis 14 Plantago lanceolata 20 Verbascum longifolium 12 Hypericum barbatum 23 Phleum phleoides 13 Armeria canescens 12 Phleum pratense 7 Juniperus oxycedrus subsp. oxycedrus 9 Centaurea solstitialis 20 Alyssum montanum subsp. repens 16 Anthyllis vulneraria subsp. rubriflora 18 Eryngium amethystinum 17 Sideritis raeseri subsp. raeseri 14 Stipa pulcherrima 17 Koeleria lobata 15 Hypericum rumeliacum 16 Lomelosia argentea 21 Teucrium capitatum 17 Leontodon crispus 12 Morina persica 8 Cerastium decalvans 7 Erodium absinthoides subsp. guiciardii 9 Genista cf. depressa 7 Minuartia attica subsp. attica 8 Teucrium chamaedrys 12 Achillea holosericea 8 Sideritis montana subsp. montana 13 Astragalus angustifolius 6 Inula oculus-christi 7 Prunus prostrata 6 Onosma visianii 5 Rosa villosa 5 Acinos alpinus subsp. hungaricus 8 Carex kitaibeliana 5 Satureja montana 5 Bupleurum falcatum subsp. cernuum 4 Dianthus cruentus 4 Orlaya grandiflora 6 Thymus boissieri 6 Trifolium fragiferum 5 Draba lasiocarpa 4 Herniaria parnassica subsp. parnassica 3 Pimpinella tragium 3 Sedum ochroleucum 3 Bromus erectus 4 Silene vulgaris subsp. prostrata 4 Nardus stricta 20 Luzula spicata 19 Phleum alpinum 14 17 17 22 11 11 11 11 11 22 22 11 11 11 1 2 3 Vegetation unit 1 2 3 4 5 6 7 Alopecurus gerardii 13 60 Silene roemeri subsp. macrocarpa 13 60 Geranium subcaulescens 16 6 14 55 Avenellaflexuosa 11 50 Bellardiochloa variegata 11 50 Geum montanum 9 45 Veronica chamaedrys subsp. chamaedrys 12 6 45 Dianthus deltoides 16 17 45 Campanula spatulata subsp. spatulata 9 3 40 Scleranthus perennis subsp. marginatus 12 9 40 Thymus stojanovii 8 40 Vaccinium myrtillus 7 35 Viola cf. velutina 11 9 35 Senecio rupestris 12 11 35 Viola tricolor 6 30 Juniperus communis subsp. nana 6 3 25 Ranunculus sartorianus 7 6 25 Genista tinctoria 7 3 7 25 Dianthus myrtinervius 4 20 Pimpinella saxifraga 4 20 Sagina saginoides 4 20 Trifolium parnassii 4 20 Veronica orsiniana subsp. orsiniana 4 20 Thymus praecox subsp. jankae 5 3 20 Other species Potentilla recta 39 58 20 40 57 35 Achillea nobilis 21 8 11 40 34 7 15 Acinos alpinum 21 17 20 26 36 15 Koeleria macrantha 17 56 20 17 20 Trifolium pratense 14 17 40 9 7 25 Lotus corniculatus 19 11 20 20 7 40 Marrubium peregrinum 15 8 33 20 20 14 Linaria peloponnesiaca 17 22 20 7 30 Minuartia recurva subsp. condensata 15 44 23 7 5 Veronica arvensis 15 8 11 20 23 7 10 Dianthus pinifolius subsp. lilacinus 14 8 11 17 36 Lactuca serriola 11 8 22 20 14 7 Cerastium semidecandrum 12 33 22 14 Trifolium repens 12 11 20 11 25 Sherardia arvensis 13 33 17 7 5 Asperula purpurea subsp. apiculata 12 9 21 25 Phleum phleoides 13 22 20 11 Table 3: Percentage frequency of diagnostic taxa of different classes in the six distinguished vegetation units (1: Koelerio-Corynephoretea, 2: Thero-Brachypodietea, 3: Stellarietea mediae, 4: Festuco-Brometea, 5: Daphno-Festucetea, 6: Juncetea trifidi). Tabela 3: Frekvenca v odstotkih diagnostičnih vrst različnih razredov v šestih vegetacijskih tipih (1: Koeler-io-Corynephoretea, 2: Thero-Brachypodietea, 3: Stellari-etea mediae, 4: Festuco-Brometea, 5: Daphno-Festucetea, 6: Juncetea trifidi). Classes Diagnostic taxa of (in %): 1 2 3 4 5 6 Koelerio-Corynephoretea 1 l9 9 20 6 S Thero-Brachypodietea l 26 l S 4 O Stellarietea mediae ll S 2l l S O Festuco-Brometea lS l4 lO 22 2l ll Daphno-Festucetea O O O l 22 O Juncetea trifidi O l O 2 O 29 Artemisietea vulgaris S S S S l l Molinio-Arrhenatheretea l 2 9 S l S Other species 2S 24 SS Sl 40 46 Table 4: Relevés of Koelerio-Corynephoretea in the Prespa area. Diagnostic species are according to Mucina (1997), Rusina (2005), Kuzemko (2009) and Pedashenko et al. (2013). The location of the plots is given in Figure 1. Tabela 4: Popisi razreda Koelerio-Corynephoretea iz območja Prespa. Diagnostične vrste so povzete po Mucina (1997), Rusina (2005), Kuzemko (2009) in Pedashenko et al. (2013). Lokacije popisnih ploskev so podane na Sliki 1. Running number of relevé l 2 S 4 S 6 l S 9 l0 ll l2 Original number of relevé S4 9S 92 2SS 2S6 2S7 2Sl 26S 264 26S 266 267 Plot size (m2) l6 l6 2S 2O 2S l6 2S 2O 2O 2S 2S 2S A b Geology Al Al Al Al Al Al Al Al Al Al Al Al o Altitude (m a.s.l.) SS2 S4S SS2 S49 SS4 sss SS2 SSS SS6 SS6 SSl SSl ut Exposition (o) SlS 29S O O O SlS O S O O 1S7 SlS ft c o n Inclination (%) S S O O O S O 9O O O S S t Cover of vegetation (%) 6O SO 9O 6O 6O 6O lO 9S lO 6O lO 9O anc o Cover of shrubs (%) O O l O O O l O O O O O << Cover of herbs (%) 6O SO 9O 6O 6O 6O l0 9S l0 60 70 90 Diagnostic species of Koelerio-Corynephoretea (incl. Festuco-Sedetalia acris) Cruciata pedemontana + + + + l + + + l l + l l2 Rumex acetosella l l l l l l l + l + l l l2 Poa bulbosa l + l + l + l + + l l ll Trifolium arvense + + l l l + l + l l l ll Linaria genistifolia subsp. sofiana + + l + l l l l 2a + + ll Vulpia myurus + l + + l l l l 2a 9 Filago arvensis + + l l l + r r r 9 Petrorhagia saxifraga r r + r r + 6 Trifolium scabrum + + + r l r 6 Medicago minima r + r + r S Aira elegantissima r r r r r S Sedum acre r r r r r S Vicia lathyroides + + + + l S Trifolium campestre + + + S Sedum annuum r r r S Myosotis ramosissima subsp. ramosissima + r + S Arenaria serpyllifolia + r + S Berteroa incana + + r S Potentilla argentea + l + S Diagnostic species of Sileno conicae-Cerastion semidecandri Silene conica + l + l l l + + l l + l l2 Cerastium semidecandrum r + + + 4 Running number of relevé 1 2 3 4 5 6 7 8 9 10 11 12 Diagnostic species of Silene frivaldszkyana-Erysimum microstylum comm. Erysimum microstylum + 1 1 1 1 + 1 r + + 1 11 Silene frivaldszkyana 1 + 1 + 2a + 1 1 + 9 Diagnostic species of Stellarietea mediae Bromus rubens + + 1 1 + + 1 3 1 2a + 1 12 Avena sterilis 1 + + 1 1 1 r + + 1 3 11 Bromus squarosus + + + r + 1 2 7 Lupinus angustifolius r r 1 + 1 5 Anthemis arvensis subsp. arvensis + + + + + 5 Geranium purpureum + r r 3 Bromus sterilis r r 2 Vicia hirsuta + + 2 Convolvulus arvensis + + 2 Diagnostic species of Artemisietea vulgaris Anchusa officinalis 1 + 1 1 1 2a 1 + 1 1 10 Cichorium intybus r + + + + 5 Verbascum pulverulentum + 2a 2 Crepis setosa + + 2 Cynoglossum officinale + + 2 Diagnostic species of Festuco-Brometea Eryngium campestre 1 + + 1 1 1 1 + + + 2a 1 12 Sanguisorba minor + + 1 + + r + r r r 10 Hypericum perforatum 1 r 1 + 1 + + + 1 1 10 Astragalus onobrychis + + + 1 + r 1 2a 1 9 Galium verum subsp. verum + + + + + 1 1 + 8 Herniaria incana 1 + + 1 + + 6 Melica ciliata + 1 + + + 5 Helianthemum oelandicum subsp. canum r + 1 3 Convolvulus cantabrica + 1 + 3 Lactuca viminea r + + 3 Stipa capillata + r 2 Hieracium bauhini r + 2 Diagnostic species of Thero-Brachypodietea Bromus hordeaceus 3 + + 1 + + + 4 1 + + + 12 Trifolium hirtum r 1 1 1 1 5 Taeniatherum caput-medusae r 1 + + 4 Dasypyrum villosum + 1 + r 4 Alyssum chalcidicum + + + 3 Cerastium brachypetalum subsp. roeseri + r r 3 Cynosurus echinatus r r 2 Petrorhagia dubia + + 2 Diagnostic species of Molinio-Arrhenatheretea Trifolium pratense + + 2 Plantago lanceolata 2a 1 2 Other species Euphorbia myrsinites r + + 1 1 1 1 1 1 + 10 Thymus sibthorpii + + 2a 1 1 + 1 + + 9 Torilis arvensis r r r r r + r r r 9 Centaurea cuneifolia + + + 1 + + + 1 1 9 Draba muralis r r + r + + r r 8 Scirpoides holoschoenus 2a 1 1 3 3 5 5 3 8 Jasione heldreichii + + 1 1 1 + 1 1 8 Running number of relevé 5 6 7 8 9 10 11 12 r r r + 7 + + + 1 1 1 1 7 + + + 1 + + 7 1 + 1 2a 1 6 + r 1 5 r r r 5 1 + + 1 + 5 + 1 1 + 1 5 r + 4 + + 4 + + + + 4 + + 1 1 4 + + 3 2 + + 2 + + 2 Delphinium balcanicum Achillea coarctata Potentilla recta Euphorbia rigida Papaver rhoeas Tordylium maximum Hypericum olympicum Verbascum cf. eriophorum Sherardia arvensis Plantago indica Onosma heterophylla Stachys angustifolia Scabiosa webbiana Tragopogon porrifolius subsp. eriospermus Minuartia verna subsp. collina Acinos alpinus + + r + + r + 1 2 3 4 r r r + r + + Taxa in one relevé: Lathyrus sphaericus 1: r, Matricaria chamomilla 1: +, Phragmites australis 2: +, Petrorhagia prolifera 2: +, Rumex tuberosus subsp. tuberosus 2: +, Alyssum montanum subsp. repens 3: 1, Centaurea alba 3: 1, Astragalus depressus 3: +, Salix alba 3: +, Crepis sancta 3: +, Erodium ciconium 3: +, Potentilla recta subsp. laciniosa 3: +, Veronica verna 3: +, Cynodon dacty-lon 3: +, Echium italicum 3: +, Vicia villosa subsp. villosa 3: +, Anthyllis vulneraria subsp. rubriflora 3: +, Marrubium peregrinum 4: r, Saponaria officinalis 4: +, Achillea nobilis 4: +, Bellardia latifolia 5: r, Ajuga chamaepitys 5: r, Securigera varia 5: r, Galium cf. lucidum 5: +, Centaurea cyanea 5: +, Verbascum nigrum 6: +, Rosa canina 7: 1, Dianthus stenopetalus 7: 1, Dianthus pinifolius subsp. lilacinus 7: +, Lactuca serriola 7: +, Stipa pulcherrima 7: +, Mentha spicata 8: 3, Vicia grandiflora 8: 2a, Galium aparine 8: 1, Cirsium eriophorum 8: 1, Vicia angustifolia 8: +, Vicia lutea 8: +, Potentilla reptans 8: +, Origanum vulgare 9: 2a, Achillea grandifolia 9: 1, Centaurea solstitialis 9: 1, Plantago major subsp. intermedia 9: +, Hylotelephium telephium 10: +, Silene italica 10: +, Veronica arvensis 10: +, Tragopogon dubius 10: +, Festuca cf. varia 11: 1, Trifolium tenuifolium 11: +.