Introduction ACTA BIOLOGICA SLOVENICA LJUBLJANA 2004 Vol. 47, Št. 1: 3-11 Sprejeto (accepted): 2003-11-21 Coenological and synphysiological investigations on loess grassland vegetation (Salvio-Festucetum rupicolae) close to Godollo Hills (Hungary) Tibor SZERDAHELYJI, Szilvia FOTI1, Szilard CZOBEL', Janos BALOGH2, Zoltan NAGY1 & Zoltan TUBAJ,2 'Department of Botany and Plant Physiology, Szent Istvan University, H-2103 Godollo 2Departmental Research Group of Hungarian Academy of Sciences, Szent Istvan University, H-2103 Godollo Abstract. Parallel coenological and synphysiological examinations were car- ried out on three typical xerophilous loess grassland stands of Salvio-Festucetum rupicolae community in the Godollo Hills, near Isaszeg village. Three stand types, a Carex humilis, a Chamaecytisus austriacus and a Stipa dasiphylla dominated ones were investigated at the same spatial scale (mesoscale). In our sample area apparent dominance by three species (Festuca rupicola, Stipa dasiphylla, and Carex humilis) suggested that these stands of this subassociation have been emerged from three types of differing species composition. According to these analyses three groups dominated by Carex humilis, Chamaecytisus austriacus and Stipa dasiphylla oj festucetosum rupicolae subassociation of a stand of Salvio- Festucetum rupicolae association were separated. Results of the synphysiological measurements show, that water shortage and senescence in autumn are responsible for the different physiological performances of the three types. In the well-watered summer period we found significant difference only between the physiological activity of the Carex- and Chamaecytisus- dominated types, however this differ- ence is due to the significant LAI-difference and after relating photosynthesis to 1 m2 leaf area, the difference is non-significant. Keywords: coefficient of variation, coenology, grassland, loess stands, multi- variate analysis, Salvio-Festucetum rupicolae Z6lyomi 1958, synphysiology. Temperate grasslands cover large areas of the Earth's vegetation (COUPLAND 1992), and they are locate in one of the regions where the impact of globa! climate change is predicted to be high (MITCHELL et al. 1990). The temperate grassland surface has large and increasing areas with arid 4 Acta Biologica Slovenica, 47 (1), 2004 climate. Even in the middle ofEurope, Hungary has ·areas where the relatively low and unevenly dis- tributed yearly precipitation results in semi-arid grasslands like loess grasslands. Loess grassland stands in the Carpathian Basin are among the richest in species considering the plant communities of the Pannonian vegetation. These stands are at the westem edge of the continental-Eurasian forest steppe-steppe zone. Unfortunately, intact stands exist only as small patches today. The Salvio- Festucetum rupicolae association has severa! subassociations differing in their physiognomy. These stands were dominated by Festuca rupicola, but differed in physiognomy, texture, spatial pattem, vegetation dynamical traits and physiological activity from each other. (VIRAGH, BARTHA 1998). This may partly be caused by the constantly changing long term dynamism of spots of species (therophytes, hemitherophytes, stoloniferous species). The long term processes are frequently mod- ified in rates or in directions of s mali scale disturbances. These rapid changes are expressed by degra- dation, decreased diversity of stands or invasion by advancing weeds or disturbance tolerant species. Coenological and synphysiological examinations were carried out on three typical xerophilous loess grassland stands ( Carex humilis, Chamaecytisus austriacus and Stipa dasiphylla dominated) of Salvio-Festucetum rupicolae Z6lyomi 1958 community in the Godoll6 Hills, near Godollo (Isaszeg village). The synphysiological measurements were started in three apparently different and arbitrar- ily chosen types. We were interested whether these types could be detected also by multivariate analysis from many coenological samples of the area. We also tried to reveal what kind of synphysiological traits are characterising the physiognomically distinct, adjacent, mosaic-like types of the Jestucetosum sul- catae subassociation. Materials and methods Study area, coenological survey Sample plots were chosen between villages Isaszeg and Kerepes (230 m a. s. l.) in typical and continuous stands ofthe Salvio-Festucetum rupicolae association (ZOLYOMI 1958). In this area the steppe grassland stands occupy marginal positions in the vicinity of Godollo Hills. The annual pre- cipitation: 601 mm, the average temperature: 9,1 °C, the medium daily maximum: 14,1°C, the medi- um daily minimum: 4, l 0 C. (KAKAS 1969). This flora spectrum is poor in characteristic species and contains several steppe species typical of the _Central Range. Hungarian loess vegetation strictly related to eastem European steppe vegetation. This flora composition of westem steppe fragments similar to e. g. Ukrainian steppe vegetation (ZOLYOMI & FEKETE 1994), so these sample plots of this area could represent adequately the East European loess vegetation. We tried to set up plots representative to this association, rich in species, diverse in physiogno- my, far from plantations (Populus forest, hedges, etc.). The synphysiological study was carried out in June (summer aspect) and in September (autumn aspect) 2000. During June we made physiological measurements both under stressed and under non- stressed conditions in the vegetation's fully developed phase, while in September measurements were made under wet conditions, but in the vegetation's senescent phase. The coenological survey was car- ried out on 3rd July 2001. Seventy five quadrats of 2 x 2 m were chosen. The species list and cover values are detected. The species composition consists of 62 species. Dominant species with decreas- ing cover values are the following: Festuca rupicola, Carex humilis, Dorycnium germanicum, Cytisus austriacus, Chrysopogon gryllus, Stipa dasiphylla. There are severa! subdominant species, T. Szerdahelvi, S. F6ti, S. Cz6bel, J. Balogh, Z. Nagy & Z. Tuba: Coenological and synphysiological... 5 e. g.: Phleum phleoides, Seseli osseum, Teucrium chamaedrys, Galium verum,Helictotrichon pubes- cens, Filipendula vulgaris, Euphorbia pannonica, Asperula cynanchica. Characteristic and accesso- rial species the next: Salvia nemoralis, lsatis tinctoria and Agropyron repens, Agropyron caninum, Centaurea sadleriana, Hypericum perforatum, Falcaria vulgaris, Dactylis glomerata, Anthericum ramosum, Adonis vemalis. Por statistical evaluation several ecological characters of species were taken from Hungarian Database 1.2 (HORV ATH et al. 1995). The ecological indicator values of species (BORHIDI 1995) are given in percentage pro rata. These characters are: the relative temper- ature requirement of species (TB), soil acidity (RB), hurnidity (WB), nitrate supply (NB), relative light intensity (LB), continentality (KB) and the types of social behaviour (SOC). The nature con- servation ranks (NCR) (Simon 1988) are given in 10 categories separated in two groups. Proportion of the first group (U-unique, KV-strictly protected, V-protected, E-native species, K-accessorial species, TP-nature pioneers) reflects natura! conditions, proportion of the second group indicates the degree of degradation (TZ-disturbance tolerant native species, A-adventives, G-cultivated species, GY-weeds) compared to the Hungarian average. The categories of simplified flora element spectrum of Hungary (FL) contain the main groups completing with relationship of Pannonic viewpoint. This kind of simplification is suitable for the analysis. The coenosystematical groups (COENOS) are established for categories of the hierarchical system. Multivariate analysis on cover values of species was carried out by using the SYN-TAX program (PODANI 1993, 1997). Por the cluster analysis the Czekanowsi-index was used (PODANI 1993, 1997). Synphysiological methods COz-exchange, transpiration, air-temperature, relative hurnidity and vapour pressure, and stom- atal conductance were measured by using a portable closed-loop IRGA gas exchange system (LI- COR 6200) sampling the air in a plexi chamber of 60 cm diameter (ground area of the chamber is 2826 cm2) and 70 cm height with three replicate measurements in each plot (BALOGH et al. 2002). PPFD values were recorded and LAI was estimated using sunfleck ceptometers (Decagon). Canopy- surface temperature was measured with an infrared thermometer (MX4, Raytek). Soil water content was measured by a TDR reflectometer (ML2, Delta-T Devices) in three replications at 5, 10, 20 cm soil depths. Forty plots were measured in summer and 30 in autumn in each stand types. Results and discussion Coenological pattern The dendrogram of the sample plots can be seen on Figure l. Using the Czekanowski-index there are 7 well organized groups at 0.58 similarity value. The first group (1-37. sample plots) is charac- terized by the dominance of Carex humilis, Chrysopogon gryllus, Chamaecytisus austriacus, Dorycnium germanicum, Teucrium chamaedrys and the presence of Phleum phleoides, Filipendula vulgaris (Carex dominated type). This fragment has three levels: the tall grasses (50-70 cm), the hemikryptophyte and chamaephyte species (30-50 cm) and th therophytes (5-15 cm). The levels have compact stand. The Chrysopogon gryllus is almost absent from the second group (2-50. samples), and there is less Chamaecytisus austriacus, Phleum phleoides, Filipendula vulgaris, Dorycnium germanicum, Teucrium chamaedrys too, however the dominance of Festuca rupicola is more pronounced, and the 6 Acta Biolo!tica Slovenica, 47 (!), 2004 O.B 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Figure 1. Dendrogram of sample plots of Festucetum rupicolae stand. denominative species of this association, Salvia nemorosa is also present (Cytisus dominated type). This fragment is compact too, but smaller (50 cm) than the Carex type. The third group (3-68.) (Stipa dominated type) is characterized by lower cover values of Stipa dasiphylla, Chrysopogon gryllus, Chamaecytisus austriacus, Dactylis glomerata, Festuca rupicola, Filipendula vulgaris, Phleum phleoides, Teucrium chamaedrys, Helictotrichon pubescens. The four other groups are different from each other in species composition and cover values, but are coupled by the presence of Stipa dasiphylla. This fragment is not so compact, it comprised loose and tall grass tufts and spots (50-70 cm) and smaller other hemikryptophyte and charnaephyte species. There are sevaral gaps between the grass tufts (30-40 cm). Result of principal coordinates analysis (PCoa) shows 3 loose groups of sample plots (data are not shown), identical with the three types found by cluster analysis. The rest of sample plots contains rare species making them diverse and mosaic-like, not typical to festucetosum rupicolae subassosi- ation. Analysing severa! standard coenological attributes the investigated area and the three grassland types can be characterized as follows. The distribution of temperature indicator values (TB) shows two peaks, similarly to the continentality diagrarn's two maximum values. The proportion of meso- phyl and xerophyte species is more than 55 %. The high frequency of these species is characteristic for both open or closed grassland communities. The distribution of relative hurnidity indicator values (WB) shows a normal bell-shaped form (Table l.). Its maximum value is due to the high frequency of xerotolerant plants. The diagram is asymmetric in the case of xeromorph habitats. The distribution of soil acidity values (RB) sign the high frequency of basiphylous species in the area's three types. 10 % of them are lime indicator species and do not occur on acidic soil (Carex humilis, Chamaecytisus austriacus, Thalictrum pseudominus). T. Szerdahelyi, S. F6ti, S. Cz6bel, J. Balogh, Z. Nagy & Z. Tuba: Coenological and synphysiological... 7 Table l. Percentage (%) distribution of relative humidity indicator values (WB) of the species in the investigated temperate loess grassland. categories 1 2 3 4 s 6 7 WB values of Carex-type 0,15 3,75 20,6 5,86 1,2 1,65 0,15 WB values of Cytisus-type 0,18 4,42 19,3 6,63 1,29 1,47 WB values of Stipa-type 0,18 4,59 19,6 6,53 0,53 1,94 Distribution of relative nitrogen indicator values (NB) can be seen in Table 2. Half of the species are characteristic of habitats very poor in nitrogen, 10 % of them live in places extremely pooF in nitrogen or in dry places (Seseli osseum, Minuartia _verna, Phleum phleoides). Table 2. Percentage (%) distribution ofrelative nitrogen indicator values (NB) of the species in the investigated temperate loess grassland. categories 1 2 3 4 s 6 7 NB values of Carex-type 8,11 15,6 4,35 1,8 0,6 1,65 1,35 NB values of Cytisus-type 8,29 15,3 3,13 1,47 1,47 1,66 2,03 NB values of Stipa-type 8,11 18,2 3,17 1,06 0,18 1,94 0,71 In terms of temperature (TB), water (WB) and soil (RB) requirements the distribution and the maximum of values are similar in the three types. The tendency of the distribution of nitrogen requirement (NB) is similar to the others, but the rate of Stipa type is higher at the maximum value (9% ), that at the Chamaecytisus type. Distribution of relative light indicator values (LB) shows high ratio of light plants (45%), full light plants of open habitats (20%) and shadow tolerant ones (32% ). There are only few typical shad- ow plants (Table 3.). Table 3. Percentage (%) distribuiton of relative light indicator values (LB) of the species in the inves- tigated temperate loess grassland. categories s 6 7 8 9 LB values of Carex-type 0,3 0,45 11,6 15 5,86 LB values of Cytisus-type 0,74 0,18 12,3 14 6,08 LB values of Stipa-type 0,35 0,71 10,9 12,5 8,82 The distribution of plants according to degree of continentality shows two peaks (Table 4.). The first maximum corresponds to the suboceanic species, mainly central European but expanding to East (Bromus sterilis, Teucrium chamaedrys, Galium verum, Agrimonia eupatoria), the second max- Table 4. Percentage (%) distribuiton of species according to degree of continentality of the species in the investigated temperate loess grassland. categories 3 4 s 6 7 8 KB values of Carex-type 1,35 9,46 5,56 4,05 9,61 3,15 KB values of Cytisus-type 1,1 10,3 4,79 3,31 12 1,84 KB values of Stipa-type 1,23 8,99 5,64 5,82 8,11 3,53 8 Acta Biologica Slovenica, 47 (!), 2004 imum sbows tbe continental-subcontinental species spread in East Europe (Festuca rupicola, Isatis tinctoria, Si/ene otites) . Tbe difference between the maximum values of ligbt requirement (LB) and nitrogen requirement (NB) of tbe Stipa and Chamaecytisus type is nearly tbe same (8%). The values of continentality (KB) are markedly different being 10 % bigber in tbe Chamaecytisus type, tban in tbe Stipa type. Tbis Stipa dominated group of plots is the most uniformly distributed out of tbe three types. Tbe distribution of categories of nature conservation values (SIMON 1988) sbows native species are tbe most frequent (60%), native accessorial species are less abundant (20%). The proportion of disturbance tolerant native species is low (10 %), as the proportion of weeds, cultivated plants and adventitious species too (less, tban 4%). Tbere is bowever considerable quantity of strictly protect- ed and protected species of Hungary (8 % ), like Stipa dasiphylla, Centaurea sadleriana, Adonis ver- nalis, Thalictrum pseudominus. According to distribution of floral elements (FL) the sbare by tbe Eurasian elements is tbe higb- est (27%), the submediterranean and tbe Pontic-Pannonian elements (e. g. Chamaecytisus austria- cus, /satis tinctoria) are present in considerable quantity (together 28 %) and there is a smaller quan- tity (10%) oftbe Pannonian (Dianthus pontederae) and tbe Pannonian-Balkan elements (Euphorbia pannonica). Tbe proportion of cosmopolitan species is less than 5% (Koeleria cristata), and tbe rate of adventive species is low (1,5%, e. g. Erigeron canadensis, Onobrychis viciifolia). This sbows tbat tbis area is more or less intact. Tbe distribution of tbe flora! elements sbows tbat tbe Stipa type bas tbe bigbest quantity ofEurasian, continental and submediterranean elements, in contrast to tbe Carex type tbat almost bave tbe lowest levels of these. Tbe Chamaecytisus group bas tbe bigbest quantity of European and Pontic-Pannonian elements. Table 5 sbows the distribution of social bebaviour types (SOC). Generalists (G) and tbe competi- tors (C) give balf of the species. Tbe cover by natura! pioneers (NP)(Myosotis stricta) and special- ists (S)(Adonis vemalis, Thalictrum pseudominus) is remarkable, too (togetber 26 % ). Tbere are few (15 %) disturbance tolerant (DT) species (Achillea collina, Agrimonia eupatoria, Galium verum). Tbe quantity of weeds (W), alien competitors (AC) and ruderal competitors (RC) is insignificant (Viola arvensis, Descurainia sophia) , less than 4,5 %. Table 5. Percentage (%) distribution of social bebaviour types (SOC) of tbe species in tbe investigat- ed temperat~ loess grassland. categories C s g np dt w rc ac soc. behav. typ. Carex-type 6,16 1,95 17,9 0,15 5,41 0,6 0,9 0,15 soc. behav. typ. Cytisus-type 5,16 1,66 18,2 0,37 5,16 1,29 1,47 soc. behav. typ. Stipa-type 7,41 1,06 18,7 5,29 0,18 0,53 0,18 The clistribution according to tbe coenosystematic categories sbows the group of indifferent species is comprised of severa! xeropbilous species (Hypericum perforatum, Melandrium album) and species occurring due to anthropogenic disturbance (Erigeron canadensis), witb a sbare of 14 %. Presence of forest elements sbows tbe busby cbaracter of this area. The rest are the elements of grass- land associations (80 % of ali) . Tbe distributions of tbe social bebaviour type (SOC) and of tbe nature conservation rank categories (NCR) are similar in the tbree types. The rate of monocotyledons/dicotyledons/Fabaceae sbows (Table 6.) that tbe quantity of Fabaceae species is double in the Stipa type as compared to tbe two other groups (Carex, T. Szerdahelyi, S. F6ti, S. Cz6bel, J. Balogh, z. Nagy & Z. Tuba: Coenological and synphysiological... 9 Chamaecytisus). Effects of a warmer and drier climate may favor the spread of species from the Chamaecytisus and the Stipa types. Table 6. The percentage (%) rate of monocotyledons/dicotyledons/Fabaceae of the species in the investigated temperate loess grassland. categories monocotyl dicotyl fabaceae Carex-type 8,69 24,63 2,17 Cytisus-type 10,52 22,8 1,75 Stipa-type 10,71 22,61 4,76 Synphysiological characteristics Synphysiological measurements were started in three apparently, physiognomically different types of the community. The above mentioned coenological considerations prove that there's many differences in the three types' composition, constraints by abiotic conditions, etc., but the physiolog- ical perfonnance is not necessarily different. Considering the whole dataset of summer (120 plots) and autumn (91 plots) aspect physiologi- cal measurements (temporal variability) it is obvious, that the average values and coefficient of vari- ation of photosynthesis are higher in summer (Table 7.). Decreasing values of these variables till autumn is partly caused by senescence. The last two columns show that comparing summer water- stressed (56 plots) and non-stressed (64 plots) periods separately, CV values are much smaller in the latter case. Table 7. Some data of the photosynthetical performance of a temperate loess grassland in summer and autumn 2000. Summer Autumn Non-stressed Water-stressed Carex-type, non- Carex-type, water- aspect aspect summer period summer period stressed summer stressed summer period period Mean ofLAI (m2• m"2) 3,09 3,84 2,91 3,30 1,64 3,66 SDofLAI 1,25 0,97 1,44 0,96 0,49 0,84 CVofLAI (%) 40,35% 25,20% 49,34% 29,13% 30,05% 23% The estimated average value of LAI (Table 8.) was higher in autumn than in summer, but the method used for LAl-estimation does not distinguish photosynthetically active and fully senescent leaves. Nor did we pursue the interannual change of LAI, the measurements in summer may have not been at the highest LAias it can be inferred from increasing average LAI with decreasing CV in June (non-stressed period: 2000. 06.08., 13., 15., 16.; water-stressed period: 2000. 06. 21., 22.). Table 8. Some data of the estimated leaf area index. Summer Autumn Non-stressed Water-stressed Carex-type, non- Carex-type, water- stressed summer stressed surnmer aspect aspect summer period summer period period period Mean of LAI (m'. m·') 3,09 3,84 2,91 3,30 1,64 3,66 SDofLAI 1,25 0,97 1,44 0,96 0,49 0,84 CVofLAI(%) 40,35% 25,20% 49,34% 29,13% 30,05% 23% 10 Acta Biologica Slovenica, 47 (1), 2004 Consideration of the three different types shows, that water shortage and senescence in autumn are responsible for the different physiological performances. In the well-watered summer period (soil water content: 12,38% in the Carex-type, 10 plots; 13,54% in the Chamaecytisus-type, 12 plots and 11,86% in the Stipa-type, 40 plots) we found significant difference only between the physiological activity of the Carex- and Chamaecytisus- dominated types, however this difference is due to the significant LAI- difference and after relating photosynthesis to 1 m2 leaf area, the difference is non-significant. In the case of water-stressed summer period (soil water content: 6,55% in the Carex-type, 28 plots; 7,54% in the Chamaecytisus-type, 28 plots) photosynthesis were significantly different between the types and the same was found for the autumn period (soil water content: 7,02 and 11,33% in the Carex-type on the two days of investigation, 10,89% in the Chamaecytisus-type) (Table 9.). Table 9. Some data of the photosynthetical performance of different types of the investigated grass- land in different conditions. (, ** means significant difference when referred to 1 m2 leaf area) Non-stressed summer period Stressed /water deficient/ Autumn aspect summer period Carex-type Cytisus-type Stipa-type Carex-type Cytisus-type Carex-type Cytisus-type Stipa-type Mean of A (µmol•m·2•s·1) 17,89 29,15 22,55 2,53('*) 3,29(**) 3,24(*') 4,55(*,**) 2,87(') SDofA 8,00 13,72 10,35 1,53 1,32 1,83 1,60 1,38 CVof A(%) 44,71% 47,07% 45,9% 60,45% 39,97% 56,39% 35,23% 47,9% These results contain both abiotic and biotic variability rates. At higher light intensities (PPFD>lO00 µmol.m-2.s-1) the average CO2 exchange rates were higher and CV s were lower (Table 10.) except for the case of the stressed period measurements. There is a clear trend of increase of the value of photosynthesis and decrease of its CV (number of plots are: summer: 74, autumn: 44, water-stressed period: 24, non-stressed period: 50). Table 10. Data of photosynthetical performance at saturating light conditions. Conclusions Mean of A (µmol·m·2·s·1) SDofA CVofA(%) Summer aspect Autumn aspect s:::::s:::~ 16,63 12,69 76,34% 3,62 1,66 45,72% 23,36 9,80 41,94% Stressed /water deficient/ summer period 2,59 1,60 61,79% In a typical stand we tried to set up plots, which represents this Salvio-Festucetum rupicolae association. Statistical evaluation of several ecological characters of species was carried out togeth- er with a hierarchical cluster analysis. According to these analyses three groups dominated by Carex humilis, Chamaecytisus austriacus and Stipa dasiphylla offestucetosum rupicolae subassociation of a stand of Salvio-Festucetum rupicolae association were separated. Results of the synphysiological measurements show that water shortage and senescence in autumn are responsible for the different physiological performances of the three types. In the well-watered summer period we found signif- icant difference only between the physiological activity of the Carex- and Chamaecytisus- dominat- ed types, however this difference is due to the significant LAI-difference and after relating photosyn- thesis to 1 m2 leaf area, the difference is non-significant. T. Szerdahelyi, S. F6ti, S. Cz6bel, J. Balogh, Z. Nagv & Z. Tuba: CoenologicaJ and svnphysiologicaJ ... 11 Acknowledgements The financial support of the Hungarian Scientific Research Foundation (OTKA-32586 project), the CARBOMONT (EVKZ-CT-2001-00125) and GREENGRASS (EVK2-CT-2001-00105) 5th EU Framework Research Projects are gratefully acknowledged. References BALOGH J. Sz. Cz6BEL, A. JuHAsz, Sz. F6T1, z. NAGY, & Z. TuBA 2002: Seasonal carbon-balance of a semi- desert temperate grassland ecosystem over a year period. - Acta Biologica Szegediensis, pp. 108- 110 BoRHJDI A. 1995: Social behavior types, the naturalness and relative ecological indicator values of the highre plants in the Hungarian Flora. - Acta bot. hung. 39(1-2): 97-181. CouPLAND R. T. 1992: Approach and generalizations. In: CoUPLAND R. T. (ed.): Ecosystems of the World 8A, Natura] Grassland, Introduction and Western Hemisphere. Elsevier, New York, p. 1-6. HORVATH F., Z.K. DOBOLYI, T. MORSCHHAUSER, L. L6K6S, L. KARAS, T. SZERDAHELYI 1995: Fl6ra adatbazis 1.2. Taxon-lista es attributum allomany. [Hungarian Flora Database 1.2. Taxon list and attribut- um stands.]. Vacrat6t, pp. 267. KAKAS J. (ed.) 1969: Magyarorszag eghajlati atlasza. Adattar. [Climate atlas of Hungary. Data]. - Akademiai kiad6, Budapest, pp. 263. MITCHELL J. F. B., s. MANABE v. MELESHKO, T. TOKIOKA 1990: Equilibrium climate change and its impli- cations for the future. In: HOUGHTON J. T., G. J. JENKINS & J. J. EPHRAUMS (eds.): Climate change, The IPCC Scientific Assesment, Chapter 5. Cambridge University Press, New York, p. 131-172. PODANI J. 1993: SYN-TAX 5.0: Computer programs for multivariate dala analysis in ecology and sy-stem- atics . - Abstr. Bot. 17: 289-309. PODANI J. 1994: Multivariate