The Role of Base Saturation and Altitude in Habitat Differentiation Within Philonotis in Springs and Mires of Three Different European Regions

The role of base saturation and altitude in habitat differentiation within Philonotis in springs and mires of three different European regions

PETRA HA´ JKOVA´ Institute of Botany and Zoology, Faculty of Science, Masaryk University, Kotla´rˇska´ 2, CZ 611 37 Brno; Department of Ecology, Institute of Botany, Academy of Science of the Czech Republic, Porˇı´cˇı´ 3b, Brno, CZ-603 00, Czech Republic e-mail: [email protected]

BLANKA SHAW Department of Biology, Box 90338, Duke University, Durham, NC 27708-0338, U.S.A. e-mail: [email protected]

MICHAL HA´ JEK Institute of Botany and Zoology, Faculty of Science, Masaryk University, Kotla´rˇska´ 2, CZ 611 37 Brno; Department of Ecology, Institute of Botany, Academy of Science of the Czech Republic, Porˇı´cˇı´ 3b, Brno, CZ-603 00, Czech Republic e-mail: [email protected]

DIRK HINTERLANG Institut fu¨r Landschaftso¨kologie, Westfa¨lische Wilhelms-Universita¨t, Robert-Koch- Str. 26, Mu¨nster, D-48149, Germany e-mail: [email protected]

VI´TEˇZSLAV PLA´ SˇEK Department of Biology and Ecology, University of Ostrava, Chittussiho 10, CZ-710 00 Ostrava, Czech Republic e-mail: [email protected]

ABSTRACT. Philonotis species are frequent and often dominant components of many wetland habitats. We modeled ecological optima of the four commonest Philonotis species in three regions of Europe, using measurements of pH, water conductivity, altitude and vegetation gradients. We did not observe obvious differences in ecological optima among three climatically and geographically different regions. Philonotis calcarea was well separated from other species along the water pH and conductivity gradient and occupied the most mineral- rich and alkaline springs in all cases. Philonotis seriata was at the other end of the mineral- richness gradient. The morphologically similar species, P. fontana and P. caespitosa, had very little differentiated niches with respect to water pH and conductivity, which differed slightly

THE BRYOLOGIST 110(4), pp. 776–787 0007-2745/07/$1.35/0 Copyright Ó2007 by the American Bryological and Lichenological Society, Inc. Ha´jkova´ et al.: Philonotis habitats in Europe 777

among regions. Whereas P. fontana occupied more alkaline and mineral-rich habitats than the rarely occurring P. caespitosa in the West Carpathians, the opposite situation was recorded in Mid-West Europe. In Bulgaria, these two species exhibited practically the same niches with respect to mineral richness. The ecological niche of P. fontana was wider in all regions when compared to P. caespitosa. Altitude represents a discriminating factor only for P. seriata, whereas other species exhibit very wide intervals of response to altitude. We concluded that species with clearly differentiated niches like P. seriata, P. fontana and P. calcarea are accurate indicators of the chemical parameters of spring waters. Alternatively, known water pH and conductivity can support or call into question the reliability of Philonotis determinations. The total species composition of vegetation can also have a certain predictive value for species occurrence.

KEYWORDS. Alps, Bulgaria, bryophytes, ecological niche, Europe, HOF response model, pH, Philonotis, water conductivity, West Carpathians.

^^^

Ten species of Philonotis are reported from conti- standing their ecology and taxonomy. Many papers nental Europe (Hill et al. 2006), six of which are have described habitat differentiation of related widespread. Here we attempt to assess ecological species within the genus Sphagnum (Andrus et al. demands and niche differentiation in four of the most 1983; Bragazza 1997; Du¨nhofen & Zechmeister 2000; frequent: Philonotis calcarea, P. caespitosa, P. fontana Gignac 1992; Gignac et al. 1991; Ha´jkova´ &Ha´jek and P. seriata. The genus Philonotis is generally 2004; Vitt & Slack 1984), which are dominant considered taxonomically difficult because of high components of mire ecosystems. Relationships be- levels of phenotypic variation (Buryova´ & Shaw 2005) tween phylogeny and habitat adaptations were studied and also intergradation between species. Difficulties within wetland moss genera of the Amblystegiaceae distinguishing some Philonotis species and occasional (Hedena¨s & Kooijman 1996; Kooijman & Hedena¨s species coo¨ccurrence can complicate the usefulness of 1991) that dominate in calcareous fens. No previous Philonotis taxa as ecological indicators in vegetation study has dealt with differentiation of habitat and ecological studies. The pattern of genetic requirements within Philonotis in detail. As a part of variation in two Philonotis species pairs was described our long-term European wetlands survey, we gathered (Buryova´ 2004b; Buryova´ & Hradı´lek 2006), but data that allowed us to ask the following questions: 1. phylogenetic relationships among Philonotis species Do Philonotis species differ in their ecological niches have not been published. Genetic differences between and how much do their niches overlap? and 2. Can we the two morphologically similar species, P. fontana detect regional variation in habitat differentiation and P. caespitosa, were confirmed, but their mutual within Philonotis on a European scale? hybridization, producing morphologically intermedi- Ecological niche modelling for plant species has ate populations, was also suggested (Buryova´ 2004b). received much attention in recent years in order to Species of the genus Philonotis (Bartramiaceae) quantify the response of species to abiotic factors are frequent components of wetland habitats. They (Coudun & Ge´gout 2005; Guisan et al. 2002; often dominate and form extensive mats in many Wamelink et al. 2005). The establishment of response types of (sub)alpine springs or submontane fens. This curves allows finding ecological optima and ampli- fact is reflected in vegetation classification and tudes, which facilitate the comparison of reactions to nomenclature (Hinterlang 1992; Valachovicˇ 2001). ecological factors among species. This method allows Precise knowledge of the habitat preferences of us to investigate quantitatively the ecological response wetland and other bryophytes is essential for under- of Philonotis species to measured water properties in 778 the bryologist 110(4): 2007 three climatically different regions, where shifts in the West Carpathians and in Bulgaria during the last ecological behavior of Sphagnum species and some six years. The sample areas were 16 m2 with the vascular plant species has been observed (Ha´jkova´ & exceptions of extremely small sites. Vegetation data Ha´jek 2006; Ha´jkova´ et al. 2006). If no geographical from Mid-West Europe were taken from the pub- variation in habitat differentiation within Philonotis is lished paper by Hinterlang (1992). The sample area revealed, then the species response model may be size used was somewhat smaller than in the West valid on a European scale. Carpathian and Bulgarian ones. Cover of all species was estimated using the nine-grade Braun-Blanquet MATERIALS AND METHODS scale (van der Maarel 1979). In total, we used 1266 Study area. We investigated high-mountain and vegetation plots of mires and springs in this study, submontane mires and springs in three European which contained 466 occurrences of Philonotis species, regions which differ in characteristics of climate, accompanied by measured ecological data (ground- bedrock and flora. (1) Bulgaria, where natural water pH and conductivity, altitude). Water conduc- conditions support mire and spring occurrence tivity and pH, both standardized at 208C, were mainly in mountains with peaks reaching 2900 m, has measured in situ using portable instruments (CM 101 a rather continental climate with relatively high and PH 119, Snails Instruments). Water conductivity evaporation in lower altitudes and a cold and humid well reflects the mineral richness (CaþMg) of the climate in high-mountains (Lieth et al. 1999). The groundwater, especially in spring-fed mires (Ha´jek & Bulgarian flora is one of the richest in Europe Hekera 2004; Malmer 1986; Sjo¨rs & Gunnarson 2002). (Meshinev & Apostolova 1998) combining central Conductivity caused by Hþ ions was subtracted in European (Alpine, Carpathian) elements with Medi- acidic waters with pH , 5.5 (Sjo¨rs 1952). The terranean, Continental (including Boreal-Continen- measurements were conducted with an aim to tal), Euxinian-Caucasian and Balkan elements. characterize the parameters of springwater entering Mineral-poor rocks prevail in many regions including the ecosystem. Water chemistry was measured directly mire refuge areas. (2) The West Carpathians (Czech in free spring water. When the water level was several and Slovak republics, Poland) have a colder and more centimeters below the surface (in wet meadows), a humid climate than does Bulgaria. The West Carpa- small shallow pit was dug and spring water was thian flora involves mostly central European elements allowed to clarify before measurement. Whenever combining Alpine and Carpathian species; the Boreal- spatial variation of water pH or conductivity was Continental elements are also represented in mires. observed, several replications within vegetation plots Many mires develop on flysch bedrock, which is were conducted and arithmetic means were calculat- characterized by variable mineral contents from very ed. Altitude and coordinates were measured by GPS low to extremely high. The Inner West Carpathians Garmin Etrex Summit (WGS 84 system) with the are formed of a siliceous core with extended carbonate altimeter calibrated by current atmospheric pressure. lithofacies. (3) Mid-West Europe includes localities in Data analysis. Tukey post-hoc tests following the Alps (Austria, Germany), Hercynian mountains one-way ANOVA can be conducted to test for (Germany), Vosges (France) and Belgium. Contrary differences in abiotic ecological factors among to other studied regions, the flora contains a higher regions. If these factors significantly differ, it may not rate of (sub)Atlantic elements. Geological bedrock be meaningful to test for significant differences in varies between siliceous and calcareous. The Mid- occurrence of species among regions. Under this West European dataset also includes forest springs, circumstance, significant differences in species oc- but Philonotis species occurred only rarely in such currence can simply reflect differences in ecological habitats. This last dataset was published by Hinterlang factors among regions, even when the species are in (1992). fact ecologically undifferentiated among regions. This Field data sampling. Vegetation plots were taken consideration supports the use of response curves, from all available springs and mires, which the first which illustrate the probability of species occurrence three authors found during their extensive research in under a given ecological factor value. Response curves Ha´jkova´ et al.: Philonotis habitats in Europe 779 are therefore not so dependent on over-representation Bryophyte nomenclature follows Frey et al. (1995) of data from certain parts of the study gradients. The except for Scorpidium (Kooijman & Hedena¨s 1991); Tukey post-hoc test revealed that means of measured vascular plant nomenclature follows Ehrendorfer environmental variables do differ significantly (p, (1973) with the exception of species occurring only in 0.05) among all study areas, with only one exception the Bulgarian data set, which followed Kozhuharov (pH between the West Carpathians and Mid-West (1992). Europe). Hence, we constructed species response curves. Response curves to selected environmental RESULTS gradients (pH, conductivity and altitude) were Altogether six Philonotis species were recorded in derived from logistic regression using the Huisman, the three studied regions (Table 1). The Bulgarian Olff and Fresco (HOF) type 5 model (Huisman et al. dataset included all six species; the Mid-West 1993; Oksanen & Minchin 2002). Ecological optima European and West Carpathian five of them. Only and response widths of Philonotis species were derived four species occurred frequently and were abundant. from HOF response curves. Optima were defined as Philonotis seriata and P. fontana were the most the highest point of the response curve in the case of a common species in all regions. Philonotis calcarea unimodal relationship, or as the maximum or occurred less frequently, although it was the most minimum value of the data set whenever monotonic frequent among Philonotis species in the West response was obtained. The response width (i.e., Carpathians. On the other hand, West Carpathian species tolerance) was defined as the distance between mires were poor in P. caespitosa. Only five sites in two those parts of the gradient where the predicted localities for this species were found during our probability of occurrence reached more than half of intensive research. Philonotis tomentella was recorded the maximum predicted probability of occurrence only once in Bulgaria and it was not observed in the (see also Schro¨der et al. 2005). Particular Philonotis West Carpathians, whereas it was more common in species were compared with respect to environmental the Mid-West European region (n¼10). Philonotis factors separately for three studied regions and, marchica was the rarest Philonotis species in all three finally, for the entire study area after merging all data study areas. Niche differentiation was studied within sets. Tukey post-hoc tests were applied only for a the four commonest species (P. calcarea, P. caespitosa, comparison of species occurrences within particular P. fontana and P. seriata). study areas. Box and whisker plots were constructed Ecological demands of Philonotis species in for measured environmental variables. particular regions. Niche differentiation within Phil- Each species response curve shows a realized onotis was similar in all three regions. Using HOF niche with respect to a single factor, which, however, models, we detected slightly shifted ecological optima may not be the most important factor affecting between the western and eastern parts of our study species distribution. Moreover, construction of spe- area (Mid-West European vs. Bulgarian datasets; see cies response curves is impossible for species with a Table 1). Philonotis caespitosa and P. seriata both had low number of occurrences. Hence, we subjected our higher optima with respect to pH in Mid-West vegetation samples to correspondence analysis (CA), Europe than in Bulgaria. The differences were 0.8 and with downweighting of rare species, in order to 0.6 units, respectively. On the contrary, pH optimum further test habitat differentiation within Philonotis for P. fontana was highest in Bulgaria and lowest in using total species composition of the vegetation. Mid-West Europe. Philonotis calcarea occupied the Philonotis species were used as supplementary (inac- mineral-richest and the most alkaline springs in all tive) species in this analysis to avoid circular cases (Fig. 1a, b). Philonotis seriata occurred at the argumentation. The Canoco 4.5 package (ter Braak & other end of the mineral-richness gradient; it S˘milauer 2002) was used for ordination analyses. SPSS occupied springs with significantly lower water software was used for ANOVA analyses and for conductivity than the other Philonotis species (Fig. constructing scatter plots. Response curves were 1b). Water pH of springs populated by P. seriata was modeled using R and JUICE (Tichy´ 2002) software. not as low relative to other species as it was in terms 780 the bryologist 110(4): 2007

Table 1. Ecological optima and response widths for Philonotis species derived from HOF response curves. Asterisks indicate optima inferred from minima or maxima of curves, when response curves were not unimodal and occurrence probabilities increased towards one gradient end (see text). Only measured values were used instead of optima when species occurred at only one locality. Such species are listed below the table, under the line. Abbreviations: n: number of samples, con: water conductivity, pH: water pH, alt: altitude.

Bulgarian data set West Carpathian data set Mid-West Europe data set

optimum response optimum response optimum response (HOF) width (HOF) width (HOF) width

n pH con alt pH con alt n pH con alt pH con alt n pH con alt pH con alt

P. calcarea 26 7.6 365 971 0.7 214 1181 60 7.6 406 174* 1.3 624 456 18 7.5 782* 1545 0.5 473 257 P. caespitosa 20 6.1 110 1190 1.0 83 645 5 5.6 56 923 0.3 26 78 14 6.9 217 240* 0.4 66 2160 P. fontana 68 6.2 118 1268 1.7 160 1302 38 5.7 278 493 0.8 197 1046 28 4.3* 11* 1333 1.3 140 808 P. seriata 121 6.0 18* 2529* 1.6 38 846 31 6.3 13* 2075* 0.8 17 586 25 6.6 23* 2400* 0.4 72 389 P. marchica 1 7.3 464 725 - - - 1 7.8 636 355 - - - 0 ------

of conductivity. The running and oxidized water was other species along the pH gradient (Fig. 2a). Its comparable to that characteristic of P. fontana and P. occurrence probability steeply increased towards pH caespitosa. Realized-niche differentiation between P. above 7. Other species exhibited similar responses to fontana and P. caespitosa with respect to pH and the pH gradient. Similarly, P. calcarea occupied conductivity was very small and not significant, and in extremely mineral-rich springs with an optimum addition, differed among regions. Whereas P. fontana conductivity of about 400 lScm1 as compared to occupied more alkaline and mineral-rich habitats others species (Fig. 2b). To the contrary, P. seriata than P. caespitosa in the West Carpathians, the was found in very mineral-poor springs and the opposite situation was recorded in Mid-West Europe. probability of its occurrence steeply decreased with In Bulgaria, these two species exhibited practically the increasing conductivity. Philonotis caespitosa and P. same niches with respect to mineral richness and pH fontana did not exhibit such clearly differentiated (Fig. 1a, b). The ecological niche of P. fontana was niches as the two previous species. The optimum of P. wider in all regions when compared to P. caespitosa fontana was toward more mineral-rich mires and its (Table 1). Philonotis species distributions also re- response curve was wider than that of P. caespitosa flected the altitudinal gradient (Fig. 1c). Generally, all (Fig. 2b). Altitude represented a discriminating factor Philonotis species grew in higher altitudes in Bulgaria only for P. seriata, whereas other species exhibited and Mid-West Europe than in the West Carpathians, very wide intervals of responses to altitude (Fig. 2c). which is caused by the different latitude and partly by Our response models suggest only weakly differenti- the different absolute altitude among regions. Phil- ated niches between P. caespitosa and P. fontana, onotis seriata occupied springs in the (sub)alpine belt, contrary to the differentiated niches of P. calcarea and whereas P. calcarea and P. caespitosa occurred mostly P. seriata (Fig. 2a–c). in lower altitudes with the exception of Mid-West Indicator value of studied species and ecological Europe, where calcareous springs also support the predictions. Species with clearly differentiated and occurrence of P. calcarea in higher altitudes. rather narrow responses to pH and conductivity The general response model valid for entire gradients like P. seriata and P. calcarea are good study area. We modeled species responses to indicators of the chemical parameters of spring measured environmental variables together for the waters. On the other hand, known pH and conduc- entire study area because of similar trends in species tivity can be used to evaluate the reliability of distribution along ecological gradients in different Philonotis determination. For example, if we mea- regions. Philonotis calcarea was well separated from sured pH as about 7.5 in the field, and we found a Ha´jkova´ et al.: Philonotis habitats in Europe 781

Figure 2. HOF response curves to water pH (A), water conductivity (B) and altitude (C) for Philonotis species. The complete dataset including samples from all three studied regions and also samples without Philonotis species were used in the analyses to determine real probabilities of Philonotis species occurrence relative to environmental factors (water pH, conductivity and altitude).

Philonotis species, this species would most probably be P. calcarea (Fig. 3a). On the contrary, Philonotis Figure 1. Box and whisker plots of water pH (A), water species growing with pH under 6.9 can be, with conductivity (B) and altitude (C) for Philonotis species. Box probabilities of about 30%, P. fontana, P. calcarea or vertical length is the interquartile range. The line across the box indicates the median, the stars indicate outliers. Abbreviations of Philonotis species names: calc (P. calcarea), caesp (P. caespitosa), font (P. fontana), seri (P. seriata). The different small letters parameters within the regions. Differences were tested by Tukey identify signiﬁcant differences between species in ecological Post-Hoc tests following one-way ANOVAs. 782 the bryologist 110(4): 2007

caespitosa occurrence. In such habitats, P. fontana occurred with 60% probability and P. caespitosa with a probability of about 30%, whereas the two other species have probabilities of occurrence lower than 10% (Fig. 3b). Position of Philonotis species along the main vegetation gradients. The main vegetation gradient was predominantly connected with altitude in all studied regions (Fig. 4a–c), which was indicated by the occurrence of other vascular plant and bryophyte species. Philonotis seriata was accompanied by Epi- lobium alsinifolium, Cratoneuron decipiens, Saxifraga stellaris, Deschampsia caespitosa, Viola biflora and other species. All these species indicated (sub)alpine springs. In Mid-West Europe, a different position of Philonotis calcarea along the main vegetation gradient was evident. Philonotis calcarea was placed in the upper part of the elevation gradient, similar to the (sub)alpine species Philonotis seriata (see also the position of other (sub)alpine species Saxifraga aizoides and Polygonum viviparum). This pattern was caused by a different altitudinal distribution of calcareous habitats in Mid-West Europe compared to other study areas. Whereas calcareous fens had a distributional center in low altitudes in the West Carpathians and in Bulgaria, they occurred at high altitudes in Mid-West Europe. In this dataset, the first ordination axis was partially connected also to light conditions. Semi-open forest springs were an important component of the left part of the scatter. The second vegetation gradient was connected with pH Figure 3. HOF response curves to water pH (A), water and mineral-richness. Philonotis calcarea occupied the conductivity (B) and altitude (C) for Philonotis species. The same part of the gradient as calcicole species such as probabilities of occurrence were calculated only from samples Cratoneuron commutatum, Blysmus compressus and with at least one Philonotis species, to show the probability of Carex flava, whereas P. fontana and P. caespitosa occurrence of particular Philonotis species at sites where at least one Philonotis occurred. optima corresponded to (sub)neutrophilous, acid- ophilous and meadow species (Fig. 4b, c). Moreover, P. seriata. In regard to altitude, P. seriata reached the in the Mid-West European dataset, these two species highest probability of occurrence at about 2000 m had similar positions along main gradients as forest- (Fig. 3c). Water conductivity values of about 130–140 spring species like Chrysosplenium oppositifolium and lScm1 can be a good predictor of P. fontana and P. Plagiomnium undulatum and the open-spring species

! Figure 4. Correspondence analyses (CA) scatter diagram for (A) the Mid-West European dataset (Eigenvalues: 1st axis 0.746, 2nd axis 0.291), (B) the Bulgarian dataset (Eigenvalues: 1st axis 0.559, 2nd axis 0.318), (C) the West Carpathian dataset (Eigenvalues: 1st axis 0.455 and 2nd axis 0.350). Philonotis species were used as supplementary species in analyses. Only Philonotis species and species with high ﬁt and high weight in the analysis are shown. Philonotis species with an absolute frequency of ten samples or less are written with small letters. Ha´jkova´ et al.: Philonotis habitats in Europe 783 784 the bryologist 110(4): 2007

Veronica beccabunga and Stellaria alsine (Fig. 4a). The common Philonotis species in the neighboring areas total species composition of vegetation can therefore of the Bohemian Massif (Buryova´ 1996). A detailed also have a certain predictive value for Philonotis specimen-based distribution for this species is species occurrence. unfortunately not available from the West Carpa- thians. The distributional pattern of P. caespitosa in DISCUSSION the West Carpathians, where this species is rare, We calculated model-described ecological opti- having fen localities especially at the northwestern ma of four Philonotis species in three regions of margin of the West Carpatians (Pla´sˇek & Stebel Europe, using measurements of pH, water conduc- 2002), has an analogy in the distribution pattern of tivity, altitude and vegetation gradients. We did not some vascular plant and bryophyte species with an observe obvious differences in ecological optima oceanic distribution tendency. The species Hydro- among the three climatically and geographically cotyle vulgaris, Juncus bulbosus, Montia fontana, Lotus different regions. uliginosus, Carex demissa and Sphagnum denticulatum Mixed populations of species pairs P. caespitosa- are much more common in the Bohemian Massif P. fontana, P. calcarea-P. fontana and P. fontana-P. than in the neighboring West Carpathians (Ha´jek & seriata have been observed in natural habitats (Bury- Ha´jkova´ 2002). Isozyme analysis of P. fontana and P. ova´ 2004a). Our study revealed well-differentiated caespitosa in Mid-Europe did not show any geo- ecological niches for P. seriata and P. calcarea. The graphic structure within the two species (Buryova´ two morphologically similar species, P. fontana and P. 2004b). caespitosa, had only slightly differentiated ecological We detected no marked geographical variation in niches. Our results showed only small ecological ecological optima of four of the most common differences between P. caespitosa and P. fontana, and Philonotis species on a European scale. Only slightly only in the Mid-West European dataset, where P. different pH and conductivity optima were found in caespitosa grew under somewhat higher pH. Mor- some cases, but it is impossible to generalize such phological similarity between and variability within P. small differences in terms of generally higher value in fontana and P. caespitosa have led to taxonomic one region as compared to others. This result may be disagreement about whether these species should be caused rather by a different structure of the datasets evaluated as varieties (e.g., Crum & Anderson 1981) than by a real influence of different climate, or separate species. It is difficult to assess whether competition pressure or evolution history (Guisan et similar morphology or close phylogenetic relationship al. 2002; Thuiller et al. 2004). Relatively stable generally coincides with similar ecology in bryo- ecological behavior of Philonotis species throughout phytes. A lack of ecological differentiation seems to the whole of Europe is interesting as compared with exist between some pairs of morphologically similar some Sphagnum (Ha´jkova´ &Ha´jek 2007) and vascular taxa which are recently accepted at the species level plant species (Ha´jkova´ et al. 2006), where significant (Hill et al. 2006), for example between Campylium differences in ecological demands among different stellatum and C. protensum or between Cratoneuron regions have been detected. filicinum and C. curvicaule (Hedena¨s 2003). On the Papers dealing with niche differentiation within other hand, Kooijman and Hedena¨s (1991) docu- one genus are rather scarce, with the exception of mented habitat differentiation of the morphologically Sphagnum, where niche differentiation along major similar species Scorpidium revolvens and S. cossonii ecological gradients has been studied in many parts of and Ha´jkova´ and Ha´jek (2004) showed a partial the Northern Hemisphere (Bragazza 1997; Gignac differentiation of habitats of Sphagnum fallax and S. 1992; Rydin 1986; Vitt & Slack 1984). A large number flexuosum. of Sphagnum studies are probably motivated by the The distributional frequency of P. caespitosa in high number of species in that genus and its particular regions has an uneven pattern. Whereas P. dominant role in many mire ecosystems. Niche caespitosa seems to be rare in most parts of the West structure within genera other than Sphagnum has Carpathians, it is documented as the second most been studied less often (e.g., Ha´jkova´ 2005; Kooijman Ha´jkova´ et al.: Philonotis habitats in Europe 785

& Hedena¨s 1991) despite the importance of some LITERATURE CITED non-peatmoss species in mire ecology and classifica- Andrus, R. E., D. J. Wagner & J. E. Titus. 1983. Vertical zonation tion (Gorham & Janssens 1992; Vitt 2000). In general, of Sphagnum mosses along hummock-hollow gradients. detailed ecological knowledge about congeneric spe- Canadian Journal of Botany 61: 3128–3139 Bell, P. R. & E. Lodge. 1963. The reliability of Cratoneuron cies, which may often have speciated in response commutatum (Hedw.) Roth as ‘‘indicator moss.’’ Journal of microhabitat variation, provides important informa- Ecology 51: 113–122. tion of use in vegetation classification and the Bragazza, L. 1997. Sphagnum niche diversification in two indicator value of those communities for abiotic oligotrophic mires in the southern Alps of Italy. The environmental factors (Dakskobler et al. 1999; Loidi Bryologist 100: 507–515. et al. 1999; Mu¨tter et al. 1998). ———, H. Rydin & R. Gerdol. 2005. Multiple gradients in mire vegetation: a comparison of Swedish and an Italian bog. Nevertheless, closely related species can be Plant Ecology 177: 223–236. reliable indicators of environmental parameters such Buryova´, B. 1996. Rozsˇı´rˇenı´ druhu˚ rodu Philonotis vCˇ eske´ as mineral concentrations (Bell & Lodge 1963) or Republice. [Distribution of Philonotis species in the Czech other soil characteristics (Shaw 1981), but only if Republic]. M.Sc. thesis [in Czech], Department of Botany, morphologically similar, taxonomically difficult spe- Charles University, Prague, Czech Republic. cies are correctly determined. This point highlights ———. 2004a. Genetic variation in two closely related species of the value of collaboration between taxonomic spe- Philonotis based on isozymes. The Bryologist 107: 316–327. ———. 2004b. Morphological and Genetic Variation in cialists and ecologists, such as in the present study. Selected Species of the Genus Philonotis (Bartramiaceae, Our results seem to be important especially in the Bryopsida). Ph.D. thesis, Department of Botany, Charles light of the fact that Philonotis species are often University, Prague, Czech Republic. confused with one another because of morphological ——— & Z. Hradı´lek. 2006. Clonal structure, habitat age, and similarities. Since base-saturation and altitude are conservation value of the moss Philonotis marchica in crucial factors determining variation in mire vegeta- Kotoucˇ quarry (Czech Republic). Cryptogamie Bryologie 27: tion (Bragazza et al. 2005; Ha´jkova´ et al. 2006; Nekola 375–382. ——— & A. J. Shaw. 2005. Phenotypic plasticity in Philonotis 2004; Tahvanainen 2004), the incorrect determination fontana (Bryopsida: Bartramiaceae). Journal of Bryology 27: of Philonotis calcarea, P. fontana and P. seriata may 13–22. substantially alter the results of any vegetation- Coudun, C. & J. C. Ge´gout. 2005. Ecological behaviour of environment relationships. Our analyses show that herbaceous forest species along a pH gradient: a comparison these closely related species can be useful for between oceanic and semicontinental regions in northern ecologists and vegetation scientists working in wet- France. Global Ecology and Biogeography 14: 263–270. Crum, H. & L. E. Anderson. 1981. Mosses of Eastern North land ecosystems. America, Vol. 1. Columbia University Press, NY. Dakskobler, I., A. Selisˇkar & B. Vresˇ. 1999. Stellaria nemorum L. ACKNOWLEDGMENTS and S. montana Pierrat (Caryophyllaceae) in the forest The authors express their thanks to the Grant Agency of the communities of Slovenia. Folia Geobotanica 34: 115–126. Czech Academy of Sciences (project no. B6163302). The Du¨nhofen, A. M. & H. G. Zechmeister. 2000. Sphagnum- research was performed within the long-term research plans of Zonation entlang von Wasserstands- und Wasserchemie- the Botanical Institute of the Czech Academy of Sciences (no. gradienten in zwei o¨stereichischen Moorgebieten. Herzogia AVZ0Z6005908 and KSK6005114) and of Masaryk University, 14: 157–169. Brno (no. MSM0021622416). We are also very grateful to Iva Ehrendorfer, F. 1973. Liste der Gefa¨ßpflanzen Mitteleuropas, 2., Apostolova and Tenyo Meshinev for crucial help with erweitere Auflage. Gustav Fischer Verlag, Stuttgart-Jena-New organization of our common field investigations in Bulgaria, York. and to Jon Shaw for constructive comments on the manuscript Frey, W., J.-P. Frahm, E. Fischer & W. Lobin. 1995. Die Moos- and language correction of the text. Special thanks belong also und Farnpflanzen Europas. In Kleine Kryptogamenflora. to Daniel Dı´teˇ, who guided us through many well-preserved Vol. IV. 6th ed. Gustav Fischer Verlag, Stuttgart-Jena-New West-Carpathian mires and provided a great number of his York. own vegetation samples. David Zeleny´ is especially acknowl- Gignac, L. D. 1992. Niche structure, resource partitioning, and edged for crucial help with statistical analyses. Urban species interactions of mire bryophytes relative to climatic Gunnarsson and Denis Gignac provided useful comments as and ecological gradients in western Canada. The Bryologist reviewers. 95: 406–418. 786 the bryologist 110(4): 2007

———, D. H. Vitt & S. E. Bayley. 1991. Bryophyte response between S. revolvens and S. cossonii. Journal of Bryology 6: surfaces along ecological and climatic gradients. Vegetatio 619–627. 93: 29–45. Kozhuharov, S. (ed.). 1992. Field Guide to the Vascular Plants in Gorham, E. & J. A. Janssens. 1992. Concepts of fen and bog re- Bulgaria. Nauka i Izkustvo, Sofia [in Bulgarian]. examined in relation to bryophyte cover and the acidity of Lieth, H., J. Berlekamp, S. Fuest & S. Riediger, (eds.). 1999. surface waters. Acta Societatis Botanicorum Poloniae 61: 7– Climate Diagram World Atlas. CD-ROM, Backhuys 20. Publishers, Leiden. Guisan, A., T. C. Edwards Jr. & T. J. Hastie. 2002. Generalized Loidi, J., M. Herrera, I. Biurrun & I. Garcıá-Mijangos. 1999. linear and generalized additive models in studies of species Relationships between syntaxonomy of Thero-Salicornietea distributions: setting the scene. Ecological Modelling 157: and taxonomy of the genera Salicornia and Sueda in the 89–100. Iberian Peninsula. Folia Geobotanica 34: 97–114. Ha´jek, M. & P. Ha´jkova´. 2002. Vegetation composition, main Malmer, N. 1986. Vegetational gradients in relation to gradient and subatlantic elements in spring fens of the environmental conditions in northwestern European mires. northwestern Carpathian borders. Thaiszia 12: 1–24. Canadian Journal of Botany 64: 375–383. ——— & P. Hekera. 2004. Can seasonal variation in fen water Meshinev, T. & I. Apostolova. 1998. Distribution of higher chemistry influence the reliability of vegetation-environment plants biodiversity and the problem of habitat diversity in analyses? Preslia 76: 1–14. Bulgaria. Phytologia Balcanica 4: 151–159. ´ ´ ´ˇ ´ Hajkova, P. 2005. Bryophytes. Pages 154–173. In A. Poulıckova, Mu¨tter, H., H. J. B. Birks & A. Odland. 1998. The comparative M. Ha´jek & K. Rybnıćˇek (eds.), Ecology and Palaeoecology ecology of Polystichum aculeatum, P. braunii, and P. lonchitis of Spring Fens of the West Carpathians. Olomouc, Palacky´ in Hordaland, western Norway. Nordic Journal of Botany Univerzity. 18: 267–288. ——— & M. Ha´jek. 2004. Bryophyte and vascular plant Nekola, J. C. 2004. Vascular plant compositional gradients responses to acidity and water level gradients in Western within and between Iowa fens. Journal of Vegetation Science Carpathian Sphagnum-rich mires. Folia Geobotanica 39: 15: 771–780. 335–351. Oksanen, J. & P. R. Minchin. 2002. Continuum theory revisited: ——— & ———. 2007. Sphagnum distribution pattern along what shape are species responses along ecological gradients? environmental gradients in refugial region of Bulgaria. Ecological Modelling 157: 119–129. Journal of Bryology 29: 18–26. Pla´sˇek, V. & A. Stebel. 2002. Bryophytes of the Cˇ antoryjsky´ hrˇbet ———, ——— & I. Apostolova. 2006. Diversity of wetland range /Czantoria range/ and its foothills (Western Carpa- vegetation in the Bulgarian high mountains, main gradients thians - Czech Republic, Poland). Cˇ asopis Slezske´ho Muzea and context-dependence of the pH role. Plant Ecology 184: Opava (A) 51: 1–87. 111–130. Rydin, H. 1986. Competition and niche separation in Sphag- Hedena¨s, L. 2003. The European species of the Calliergon- num. Canadian Journal of Botany 64: 1817–1824. Scorpidium-Drepanocladus complex, including some related Schro¨der, H. K., H. E. Andersen & K. Kiehl. 2005. Rejecting the or similar species. Meylania 28: 1–116. mean: Estimating the response of fen plant species to ——— & A. Kooijman. 1996. Phylogeny and habitat adaptations within a monophyletic group of wetland moss environmental factors by non-linear quantile regression. genera (Amblystegiaceae). Plant Systematics and Evolution Journal of Vegetation Science 16: 373–382. 199: 33–52. Shaw, A. J. 1981 [1982]. Ecological diversification among nine Hill, M. O., N. Bell, M. A. Bruggeman-Nannenga, M. Brugue´s, species of Pohlia (Musci) in western North America. M. J. Cano, J. Enroth, K. I. Flatberg, J.-P. Frahm, M. T. Canadian Journal of Botany 59: 2359–2378. Gallego, R. Garilleti, J. Guerra, L. Hedena¨s, D. T. Holyoak, J. Sjo¨rs, H. 1952. On the relation between vegetation and Hyvo¨nen, M. S. Ignatov, F. Lara, V. Mazimpaka, J. Munõz & electrolytes in north Swedish mire waters. Oikos 2: 241–258. L. So¨derstro¨m. 2006. An annotated checklist of the mosses of ——— & U. Gunnarsson. 2002. Calcium and pH in north and Europe and Macaronesia. Journal of Bryology 28: 198–267. central Swedish mire waters. Journal of Ecology 90: 650–657. Hinterlang, D. 1992. Vegetationso¨kologie der Weichwasser- Tahvanainen, T. 2004. Water chemistry of mires in relation to quellgeselschaften zentraleuropaïscher Mittelgebirge. Cru- the poor-rich vegetation gradient and contrasting geo- noecia 1: 1–122. Verlag Natur & Wissenschaft Hieronimus chemical zones of the north-eastern Fennoscandian Shield. und Schmidt, Solingen. Folia Geobotanica 39: 353–369. Huisman, J., H. Olff & L. F. M. Fresco. 1993. A hierarchical set ter Braakk, C. J. F. & P. S˘milauer. 2002. CANOCO Reference of models for species response analysis. Journal of Manual and CanoDraw for Windows User’s Guide. Software Vegetation Science 4: 37–46. for Canonical Community Ordination (version 4.5). Bio- Kooijman, A. & L. Hedena¨s. 1991. Differentiation in habitat metris, Wageningen and Cˇ eske´ Budeˇjovice. requirements within the genus Scorpidium, especially Thuiller, W., L. Brotons, M. B. Arau´ jo & S. Lavorel. 2004. Effects Ha´jkova´ et al.: Philonotis habitats in Europe 787

of restricting environmental range of data to project current bryophytes. Pages 312–343. In A. J. Shaw & B. Goffinet and future species distributions. Ecography 27: 165–172. (eds.), Bryophyte Biology. Cambridge University Press. Tichy´, L. 2002. JUICE, software for vegetation classification. ——— & N. G. Slack. 1984. Niche diversification of Sphagnum Journal of Vegetation Science 13: 451–453. relative to environmental factors in northern Minnesota peatlands. Canadian Journal of Botany 62: 1409–1430. Valachovicˇ, M., (ed.). 2001. Plant communities of Slovakia. 3. Wamelink, G. W. W., P. W. Goedhart, H. F. van Dobben & F. Wetland vegetation. Veda, Bratislava. Berendse. 2005. Plant species as predictors of soil pH: van der Maarel, E. 1979. Transformation of cover-abundance in Replacing expert judgement with measurements. Journal of phytosociology and its effects on community similarity. Vegetation Science 16: 461–470. Vegetatio 39: 97–114. Vitt, D. H. 2000. Peatlands: ecosystems dominated by ms. received September 4, 2006; accepted April 2, 2007.