K) Check for updates Received: 18 August 2022 Revised: 29 December 2022 Accepted: 8 January 2023 DOI: 10.1111/avsc.l2711 RESEARCH ARTICLE Applied Vegetation Science Subalpine vegetation changes in the Eastern Sudetes (1973-2021): Effects of abandonment, conservation management and avalanches Klára Klinkovská1 | Anna Kučerová1 | Štěpánka Pustková1 | Jaroslav Rohel1 | Karolína Šlachová1 | Vojtěch Sobotka1 | Daniel Szokala1 | Jiří Danihelka1 '2 | Martin Kočí1 | Eva Šmerdová1 © | Milan Chytrý1 Abstract Aims: The summit grasslands of many European mountain ranges were historically used for summer grazing, which ceased in the 20th century. These grasslands are changing, partly through succession after abandonment and partly owing to environmental changes. Subalpine vegetation is also affected by artificially reduced avalanche frequency. Recent conservation efforts have attempted to reverse the negative trends of change. We ask: (1) How has subalpine vegetation changed following the abandonment and avalanche control? (2) Was conservation management able to reverse the post-abandonment trend of vegetation change? (3) Did avalanche disturbance have a positive effect on plant species diversity? Location: Summit area of the Hrubý Jeseník Mountains (1,491m a.s.l.), Eastern Sudetes, Czech Republic. Methods: Vegetation plots sampled in the 1970s were resurveyed in the 2000s and again in 2021. Subalpine vegetation was classified into six types, and transitions between these types over time were quantified. Vascular plant species richness and the proportion of threatened species were compared between periods, between areas with and without conservation management, and between areas affected vs unaffected by a large avalanche from 2019. Species composition was analysed using principal coordinate analysis and distance-based redundancy analysis. Results: Vegetation types remained relatively stable except for species-rich grasslands, some of which changed to heathlands or tall-forb vegetation. Some competitive species have increased, and species specialized to threatened habitats declined. Conservation management systematically implemented after 2010 slowed the decline of habitat-specialized species but did not reverse it. Disturbance by an avalanche positively affected species richness but not the number of threatened species. Conclusions: Subalpine vegetation is slowly losing its plant diversity owing to grazing cessation and possibly acidification from past atmospheric deposition. Recently Klára Klinkovská, Anna Kučerová, Štěpánka Pustková, Jaroslav Rohel, Karolína Šlachová, Vojtěch Sobotka and Daniel Szokala are joint first authors. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2023 The Authors. Applied Vegetation Science published by John Wiley & Sons Ltd on behalf of International Association for Vegetation Science. Department of Botany and Zoology, Faculty of Science, Masaryk University, Brno, Czech Republic 2 Czech Academy of Sciences, Institute of Botany, Průhonice, Czech Republic Correspondence Milan Chytrý, Department of Botany and Zoology, Faculty of Science, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic. Email: chytry@sci.muni.cz Funding information Grantová Agentura České Republiky Co-ordinating Editor: Jorg Ewald Appi VegSci. 2023;26:el2711. https://doi.org/10.llll/avsc.12711 wileyonlinelibrary.com/journal/avsc | l o f l 9 implemented conservation management and restoration of avalanche activity are essential to stop this trend, but future monitoring is needed to evaluate the success of management actions. K E Y W O R D S alpine grassland, avalanche, conservation management, Hrubý Jeseník Mountains, subalpine scrub, succession, vegetation change, vegetation resurvey 1 | INTRODUCTION Summer grazing by livestock was common in the past in the summit areas of many European mountains (Herzog et al., 2005). Treeless areas above the timberline were expanded, and summits of lower ranges that did not reach the climatic timberline were cleared to provide more pasture land (Pini et al., 2017). Extensive grazing contributed to the high species richness of mountain grasslands (Hoiss et al., 2013). However, mountain grazing became increasingly unprofitable during the 20th century, leading to the abandonment of many former pastures (MacDonald et al., 2000). The most obvious consequence of abandonment is the spread of forest and the upward shift in the previously artificially lowered alpine timberline (Didier, 2001; Gehrig-Fasel et al., 2007; Bracchetti et al., 2012; Treml et al., 2016; Mietkiewicz et al., 2017), and shrub encroachment above the timberline (Dullinger et al., 2003; Buhlmann et al., 2014; Palaj & Kollar, 2021; De Toma et al., 2022). In addition, there may be a significant loss of species diversity within grassland plant communities after management ceases (Dullinger et al., 2003; Schwaiger et al., 2022). Abandonment of grazing and changes in land use are not the only causes of change in alpine and subalpine grasslands (Kulakowski et al., 2011). These grasslands are also affected by global warming (Klanderud & Birks, 2003; Walther et al., 2005; Pauli et al., 2007; Varricchione et al., 2022), increasing atmospheric nitrogen deposition (Britton et al., 2009) and acidification (McGovern et al., 2011; Ross et al., 2012). Because these processes are parallel, their effects are likely combined, and it is challenging to disentangle the main drivers of biodiversity change in mountain grasslands (Kaufmann etal.,2021). Another factor affecting the diversity of alpine and subalpine grasslands is the alteration of the disturbance regime owing to anthropogenic suppression of avalanches. By disturbing subalpine forests and shrublands, avalanches maintain natural grasslands that can harbour high diversity, combining alpine belt species with forest belt species (Erschbamer, 1989; Rixen et al., 2007; Bebi et al., 2009). However, many artificial barriers have been created in mountain areas to protect against avalanches, resulting in lower avalanche frequency, which in turn triggers successional changes in vegetation (Kulakowski et al., 2006; Rixen et al., 2007). In protected areas, conservation authorities are trying to halt the decline of mountain grasslands and their biodiversity by reintroducing extensive grazing or mowing (Krahulec et al., 2001; Lochon et al., 2018). In places without potential risk to humans, attempts are also being made to restore the natural avalanche regime by removing protective barriers. However, in some areas, these measures were introduced several decades after avalanche regulation was abandoned. In the meantime, mountain grasslands have changed through succession, which may have led to biodiversity change that is difficult to reverse in the short term (Pechackova et al., 2010). The question is how much of the former diversity has been lost and whether the mere reintroduction of historical or alternative management practices can restore it. An example of a subalpine area that has gone through the journey from grazing through abandonment and avalanche control to conservation management is the summit of the Eastern Sudetes (Hruby Jesenik) in the Czech Republic (Mackovdn et al., 2021). This area was used for livestock grazing from the 17th century but was abandoned in the 1940s. Beginning in the late 19th century, the non-native dwarf pine (Pinus mugo) was planted there to reduce avalanche activity. The lack of disturbance from grazing and avalanches led to successional changes in vegetation in the second half of the 20th century. Beginning in the late 1980s, conservationists have been gradually removing Pinus mugo, and in the 1990s, they also began mowing some areas of subalpine grasslands and heathlands. After several decades of no significant avalanche activity, a large avalanche was released in January 2019, destroying an extensive area of birch woodland. Vegetation in this area has been surveyed repeatedly in the past, providing a unique opportunity to examine changes in subalpine plant communities following abandonment and avalanche regulation, and subsequent implementation of conservation management and restoration of natural avalanche dynamics. In 2021, we resurveyed vegetation plots from the 1970s and 2000s and asked the following questions: (1) How has subalpine vegetation changed following abandonment and avalanche control? (2) Was conservation management able to reverse the post-abandonment trend of vegetation change? (3) Did avalanche disturbance have a positive effect on plant species diversity? 2 | METHODS 2.1 | Study area The study area is located in the Hruby Jesenik Mountains (the highest mountain group of the Eastern Sudetes) in the northeastern Czech Republic. It includes the highest part of this mountain KLINKOVSKA ETAL. Applied Vegetation Science 3 of 19 range: Mount Praděd (1,491m a.s.l.), the Vysokoholský hřbet range (Mount V y s o k á hole, 1,464m a.s.l.) and the Velká kotlina glacial cirque on the southeastern slope of the Vysokoholský hřbet ( 5 0 ° 2 ' 3 1 " - 5 0 ° 5 ' 6 " N , 1 7 ° 1 2 , 4 2 " - 1 7 ° 1 4 ' 4 6 " E ; Figures 1 and 2). Our sampling was conducted at elevations between 1,111 and 1,464m a.s.l. The summit areas have flat topography with plateaus and rounded mountaintops, covered by monotonous vegetation of species-poor grasslands and heathlands. By contrast, the Velká kotlina cirque harbours a wide range of plant communities, many of which are species-rich (Jeník, 1961; Jeník et al., 1980; Hédl & Kočí, 2015; Bureš, 2022). The bedrock consists mainly of Palaeozoic (Variscan) metamorphic rocks, mostly phyllite, but also gneiss, mica schist, schist with graphite, greenschist and metadolerite (https://mapy.geology.cz/ geocr50/). The Velká kotlina cirque also contains patches of calcareous and alkaline igneous rocks (Novotný & Bureš, 2018). The most common soil type is acidic podzol, but there are also rankers on steep upper slopes and shallow peat accumulations on flat summit areas (https://mapy.geology.cz/pudy/). The soil pH values we measured in the vegetation plots ranged from 2.9 to 6.0. The mean annual temperature at the former climate station on Mount Praděd (1,490m a.s.l.) in 1970-1997 was 1.2 ° C (-6.5 ° C in January, 9.7 ° C in July), and the total annual precipitation was 1,094mm (with a maximum in July) (Czech Hydrometeorological Institute, www.chmi.cz). There was no statistically significant change in mean annual temperature and total annual precipitation at Mount Praděd between 1970 and 1997. At Mount Šerák (1,328 m a.s.l.) station, located 17 km from the study area, there was also no change in these variables in the period 2004-2021 (Appendix SI). Continuous snow cover lasts on average until the second half of April, but small patches of snow often persist until June. Small avalanches fall in Velká kotlina at intervals of several years. The timberline in the Hrubý Jeseník is, on average, at about 1,300m a.s.l. (Treml & Migoii, 2015) but is pushed down by avalanches in the Velká kotlina cirque (Figure 2c). The treeline ecotone is dominated by Picea abies, occasionally with an admixture of Fagus F I G U R E 1 The treeless summit area of the Hrubý Jeseník Mountains with the locations of vegetation plots sampled three times (red, concentrated in the Velká kotlina cirque) and plots sampled twice (blue). Areas where conservation management (mainly mowing) was carried out after 2010 are shown in light green. An interactive map with a zoom option, showing the exact position of plots, small shifts between the 2000s and 2021 plot coordinates and photographs of plot sites in 2021, is available at https://arcg. is/0uSrz9 PL DE \ CZ SK AT plots sampled three times (1970s. 2000s. 2021) • plots sampled two times (2000s, 2021) I I management over the last decade F I G U R E 2 Study area and major disturbance regimes. The area includes (a) the Velká kotlina glacial cirque with steep slopes and rock outcrops and (b) treeless summit plateaus with gentle slopes, (c) Disturbed birch woodlands in a lower part of the 2019 avalanche track, (d) Formerly grazed but later abandoned grassland overgrown by extensive heathlands dominated by Vaccinium myrtillus. Some of these heathlands were recently mown as part of conservation management sylvatica. The forest transitions directly into alpine grasslands, because there is no natural krummholz belt as in other Central European mountains. However, woodland or scrub of Betula pubescens subsp. carpatica and Salix silesiaca are present on avalanche tracks and sites with long-lasting, deep snow cover (Jeník et al., 1980). It is likely that the summit areas were largely covered by forest in the mid-Holocene (Treml et al., 2008; Novák et al., 2010), whereas open areas existed only on avalanche tracks and rocky slopes in the cirques. The continuous presence of natural grassland patches in this area is confirmed by the isolated occurrence of numerous subalpine grassland species, including some neo-endemics (Jeník, 1961; Jeník et al., 1983; Kaplan, 2017; Bureš, 2022). Large-scale deforestation by humans began in the 12th-13th centuries (Novák et al., 2010). The summit area was used for sheep and cattle grazing and hay-making from the 17th century (Hošek, 1972, 1973). As a result, the timberline dropped to its lowest level in the 18th and 19th centuries. In the late 19th and early 20th centuries, belts of non-native dwarf pine (Pinus mugo) were planted above the timberline to prevent avalanches (Hošek, 1972; Bureš et al., 2009). Livestock grazing began to decline already in the second half of the 19th century, and finally ended during World War II. Natural regeneration of spruce on abandoned grasslands and artificial afforestation led to a rise in the timberline and increasing forest density in the treeline ecotone (Treml et al., 2016; Mackovčin et al., 2021). Birch woodland developed on the inactive lower part of the avalanche track in the Velká kotlina cirque. Velká kotlina was designated a nature reserve in 1946, and protection was later extended to the entire study area and beyond. Nature conservation authorities have reintroduced mowing and locally (especially near the huts) also grazing. In small areas (ca. 0.1 ha), mowing was already being used experimentally in the 1990s. However, systematic mowing started only around 2010 and mowing of large areas in 2017. Areas managed after 2010, in most cases by mowing of expanding tall herbs and the dwarf shrub Vaccinium myrtillus, are shown in Figure 1. The frequency of mowing varied, and many places were mown only once in the past decade. In the late 1980s, nature conservation also began removing planted shrubs of Pinus mugo (Bureš et al., 2009). This likely contributed to the departure of an avalanche that fell in January 2019 and severely disturbed the birch woodland. This avalanche was the largest in at least 60 years and formed a 900-m track that ended at an elevation of 1,115m a.s.l., whereas previous avalanches that fell during the past 50years were shorter than 600m. 2.2 I Vegetation sampling In this study, vegetation change is analysed by comparing species composition in vegetation plots surveyed in three periods. The first survey (1970s) was conducted by Leo Bureš and Zuzana Burešová in the Velká kotlina cirque and the adjacent slopes of Mount Vysoká hole in the summers of 1973-1978. This survey was conducted simultaneously with detailed vegetation mapping, and the location of each plot was assigned to a specific patch on the map, which contained 935 patches with 37 phytosociological associations and sub-associations. The second survey (2000s) was conducted by Martin Kočí and Leo Bureš in the summers of 2004-2010 (Kočí, 2005; Bureš & Kočí, 2010). During this survey, plots in Velká kotlina were located in the same patch of the vegetation map as the corresponding plot in the 1970s, often with a spatial uncertainty of <10 m, based on the memory of Leo Bureš. In addition, several plots were sampled at new locations, particularly on summit plateaus. Plot coordinates were measured using GPS with an uncertainty of <10 m. The third survey (2021) was conducted by the authors of this article, including Martin Kočí, in the second half of June 2021. Each plot position was determined based on the GPS coordinates from the 2000s. In a few cases, because of GPS measurement inaccuracies, the coordinates were found in a habitat different from the one sampled in the 2000s. In such cases, the location of the new plot was moved a few metres from the coordinates to sample the same habitat as sampled in the 2000s. If such a habitat was not nearby, the plot was rejected. In 2021, we measured the coordinates of most plots using a GPS with an accuracy of ca. 5 cm (Topcon HiPer SR with Getac PS336 Data Collector) in each corner of the plot. In such a way, we established a network of permanent vegetation monitoring plots. The size and shape of the plots varied by vegetation type. Repeated sampling always used the same plot size as the previous sampling. Most plots were squares of 1 0 0 m 2 in woodlands and 16 m 2 in grasslands or rectangles of 10 m 2 in springs. All species of vascular plants were recorded in each plot, and their cover was estimated using the nine-grade Braun-Blanquet scale (Westhoff & van der Maarel, 1978). Bryophytes and lichens were recorded in only some plots, focusing on the dominant species; they were not included in the analysis owing to inconsistent sampling. The final data set included data from 148 vegetation plots, of which 66 were sampled three times (1970s, 2000s, 2021) and 82 were sampled twice (2000s, 2021), i.e. 362 vegetation-plot records in total. In 2021, soil samples were collected from the mineral soil horizon at a depth of 5-10 cm below the raw humus layer. Within each vegetation plot, soil samples were collected from four locations approximately in the middle of each plot quarter, mixed, dried at room temperature, sieved (mesh size 2mm), and used to measure soil pH and electrical conductivity. A soil-water suspension (weight ratio 1:2.5) was shaken in the Biosan PSU-lOi orbital shaker for 5 min at 280rpm, and pH and conductivity were measured the next day using the H Q 4 0 D digital multimeter. GIS layers indicating the areas managed (in most cases mown) for conservation purposes in the past lOyears were provided by the Administration of the Protected Landscape Area Jeseníky. Plot coordinates measured in the field were overlaid with these layers using ArcGIS 10.7 (ESRI). Based on this overlay (Figure 1), the plots were divided into managed and unmanaged. In addition, the plots were divided into affected and unaffected by the 2019 avalanche based on the locations of damaged trees observed in the field. 2.3 | Data analysis We entered vegetation-plot records from all three sampling periods into a database using the Turboveg 2 program (Hennekens & Schaminee, 2001). W e then standardized the taxonomic concepts and nomenclature according to Kaplan et al. (2019) in the Juice 7.1 program (Tichy, 2002). W e removed from the analyses the vernal species Anemone nemorosa and Cardamine pratensis, which were recorded frequently in 2021 but infrequently in the previous surveys because the 2021 sampling was conducted at a slightly earlier phenological stage. To account for the different plot sizes in the comparisons of species numbers, we fitted the species-area curves in log-log space separately for birch woodlands and open habitats. Because the slopes of the curves were not significant, we compared species numbers without accounting for differences in plot sizes. W e also analysed changes in the proportion of threatened species (IUCN categories critically endangered [CR], endangered [EN] and vulnerable [VU]) of all species between sampling periods. Species were classified according to the national Red List (Grulich, 2017); hybrids and species that could not be accurately identified were excluded from this analysis. Changes in species numbers and proportion of threatened species per plot over time were compared within groups of woodland plots and open-habitat plots and then separately for individual open-habitat types. Undisturbed and disturbed plots were also analysed separately to compare vegetation in the 2000s and 2021. For birch woodlands, disturbed plots were those affected by the 2019 avalanche, and for open habitats, disturbed plots were those cut as part of conservation management. The significance of changes in species numbers and proportions was tested using the Wilcoxon test for paired samples. Because changes in alpine vegetation can differ among plant communities (Gritsch et al., 2016), we classified the entire data set of 362 vegetation-plot records from the three different periods into vegetation types. First, t w o types were distinguished based on the dominant species (birch woodlands and heathlands). All plot records with a total tree cover larger than 1 5 % were assigned to birch woodland, and all non-woodland records with a total cover of heath species (Calluna vulgaris, Vaccinium myrtillus, V. vitis-idaea and Empetrum hermaphroditum) larger than 2 5 % and a total cover of bog or spring species smaller than 10% were assigned to heathlands. W e excluded two plots in beech forests and four plots on cliffs because these habitats were represented by only a few plots. For the remaining 267 plots, we used the modified T W I N S P A N classification (Rolecek et al., 2009) in the Juice 7.1 program (Tichy, 2002) with total inertia as a dissimilarity measure and three pseudospecies cut levels corresponding to 0%, 5%, and 25% species covers. The minimum group size for division was set at 2. W e divided the data set into six clusters and then merged two of them based on similar species composition, resulting in five vegetation types: bogs, species-poor grasslands, species-rich grasslands (created by merging two clusters), tall-forb vegetation and spring vegetation. The plot records of bogs (n = 8), beech forests and cliffs were included in the overall analyses, but were not analysed separately owing to the small number of plot records in these types. The types were interpreted as associations and alliances of the national vegetation classification (Chytrý, 2007). Transitions between vegetation types on the same plots through time were visualized in alluvial plots created using the ggalluvial package in R (version 0.12.3; R Core Team, R Foundation for Statistical Computing, Vienna, AT). Data were visualized for three periods, using only plots surveyed three times (1970s, 2000s and 2021), and for two consecutive periods separately, including plots surveyed only in the 2000s and 2021. For the latter two periods, alluvial plots were also created separately for undisturbed and disturbed plots. The gridExtra package in R (version 2.3; R Core Team, R Foundation for Statistical Computing, Vienna, AT) was used to combine two alluvial plots into one display panel. Because different numbers of plots were surveyed in the three periods (1970s, 2000s and 2021), we analysed changes in species richness and composition in two steps: between the 1970s and 2000s (66 plots) and then between the 2000s and 2021 (148 plots, including those first surveyed in the 2000s). We analysed overall changes across the data set and within each vegetation type (birch woodlands, heathlands, species-poor grasslands, species-rich grasslands, tall-forb and spring vegetation). Plots were assigned to these groups based on their vegetation classification at the time of initial sampling (either the 1970s or the 2000s). To allow comparison between the 2000s and 2021, we also conducted separate analyses of disturbed and undisturbed plots within each vegetation type. To visualize changes in species composition, we calculated unconstrained ordination (principal coordinate analysis; PCoA) with square-root transformation of percentage species covers (derived from the Braun-Blanquet scale values) and Bray-Curtis dissimilarity. To analyse changes in species composition, we used a constrained ordination (distance-based redundancy analysis; db-RDA) with square-root transformation of species percentage covers and Bray-Curtis dissimilarity calculated using the vegan package in R (version 2.5-7; R Core Team, R Foundation for Statistical Computing, Vienna, AT). Time was used as a constraining variable. To test the significance of the change in species composition, we permuted observations from different periods within plots 999 times. We then determined which species significantly increased or decreased over time. To do this, we used the "envfit" function and tested the significance of the relationship between species and the constrained axis of db-RDA using a permutation test with 999 permutations within plots. Changes in species occurrence or abundance were considered significant if the p-value was <0.05. For each comparison, unconstrained and constrained ordinations were plotted in R using the packages ggplot2 version 3.3.5 (Wickham, 2016) and gridExtra version 2.3 (R Core Team, R Foundation for Statistical Computing, Vienna, AT). When comparing overall changes in vegetation types over the entire period (1970s-2021), centroids were displayed for each vegetation type. The plots in this ordination were divided into vegetation types based on the 1970s classification. Environmental changes in the resurveyed vegetation plots were assessed using Ellenberg-type indicator values for Czech flora (Chytrý et al., 2018). Differences in unweighted indicator values between pairs of observations from the same plots were tested using repeated measures A N O V A ("aov" function in the stats package in R). Row-based (randomization of sampling time) and column-based (randomization of indicator values for species before calculating plot means) tests were calculated with 999 permutations. The higher pvalue of the two tests was considered (max test). The unweighted means of indicator values and permutations were calculated using the weimea package in R (Zelený, 2018). 3 I RESULTS 3.1 I Vegetation types The following six vegetation types were distinguished and analysed (see the synoptic table of species composition in Appendix S2): Birch woodlands (Figure 3a) are found mainly in the lower parts of the Velká kotlina cirque, especially near streams and at the edges of avalanche tracks. The dominant species in the tree layer is Betula pubescens subsp. carpatica, often accompanied by Picea abies and Salix silesiaca. In more disturbed areas, the dominant woody species grow as shrubs rather than trees. Typical herb-layer species include Calamagrostis arundinacea, C. villosa, Chaerophyllum hirsutum, Cirsium oleraceum, Equisetum sylvaticum, Lunaria rediviva, Luzula sylvatica, Molinia caerulea, Rubus idaeus, Stellaria nemorum and Vaccinium myrtillus. The average soil pH for this group is 4.4 (range 2.9-5.8). This vegetation belongs to the association Salici silesiacae-Betuletum carpaticae (alliance Salicion silesiacae). Heathlands (Figure 3b) are dominated by ericoid plants, such as Calluna vulgaris, Empetrum hermaphroditum, Vaccinium myrtillus and V. vitis-idaea accompanied by acidophilous graminoids, e.g. Avenella flexuosa, Calamagrostis villosa, Festuca supina and Luzula sylvatica. Heathlands are found mainly on the upper slopes and summit plateaus. They occur on acid soils with an average soil pH of 3.9 (3.1-5.3). They mostly belong to the association Festuco supinae-Vaccinietum myrtiili (alliance Genisto pilosae-Vaccinion), and some of them to Avenello flexuosae-Callunetum vulgaris (Loiseleurio procumbentis-Vaccinion). Species-poor grasslands (Figure 3c, referred to as "Poor grasslands" in figures and tables) are found primarily on wind-affected summit plateaus on acidic soils. They are dominated by the grasses Avenella flexuosa, Festuca supina and Nardus stricta, which are accompanied by Bistorta officinalis, Calamagrostis villosa and Carex bigelowii subsp. dacica. The average soil pH is 3.8 (3.5-4.2). They belong to F I G U R E 3 The main vegetation types in the study area: (a) birch woodland, (b) heathland, (c) species-poor grassland, (d) species-rich grassland, (e) tall-forb vegetation and (f) spring vegetation the associations Carici bigelowii-Nardetum strictae (alliance Nardo strictae-Caricion bigelowii), Cetrario-Festucetum supinae (Juncion trifidi), Festuco supinae-Nardetum strictae (Nardion strictae) and Crepido conyzifoliae-Calamagrostietum villosae (Calamagrostion villosae). Species-rich grasslands (Figure 3d, "Rich grasslands" in figures and tables) occur on gentle slopes, especially in the Velká kotlina cirque. This vegetation is characterized by the presence of numerous forbs and graminoids that require deep and nutrient-rich soil. The dominant species include Anthoxanthum odoratum agg., Avenella flexuosa, Calamagrostis villosa, Carex montana, Deschampsia cespitosa, Geranium sylvaticum, Hypericum maculatum, Luzula syivatica, Molinia caerulea, Nardus stricta, Poa chaixii, Potentilla aurea, Rumex arifolius and Trollius altissimus. The average soil pH is 4.5 (3.7-5.8). These grasslands belong to the associations Thesio alpini-Nardetum strictae (alliance Nardion strictae), Crepido conyzifoliae-Calamagrostietum villosae, Violo sudeticae-Deschampsietum cespitosae (both Caiamagrostion villosae) and Bupleuro longifoliae-Calamagrostietum arundinaceae [Caiamagrostion arundinaceae). Tall-forb vegetation (Figure 3e) is found on moist and nutrientrich soils near springs and streams and in seepage areas. The vegetation is dominated by Adenostyles aiiiariae, Alchemilla spp., Calamagrostis arunďmacea, C. villosa, Carduus personata, Cirsium heterophyllum, Dactylis glomerata, Deschampsia cespitosa, Filipendula ulmaria, Geranium sylvaticum, Hypericum maculatum, Laserpitium archangelica, Molinia caerulea, Poa chaixii, Rubus idaeus, Senecio nemorensis agg. and Trollius altissimus. Shrubs of Salix hastata also dominate at some sites. The average soil pH is 5.3 (4.7-5.8). They belong to the associations Ranunculo platanifolii-Adenostyletum aiiiariae, Trollio altissimi-Geranietum syivatici, Laserpitio archangelicaeDactylidetum glomeratae (all in the alliance Adenostylion aiiiariae) and Violo sudeticae-Deschampsietum cespitosae (Caiamagrostion villosae). Spring vegetation (Figure 3f) occurs on slopes where groundwater comes to the surface, especially in the Velká kotlina cirque. The dominant component of this vegetation is hygrophilous mosses and liverworts, e.g. Bryum pseudotriquetrum, Hygrohypnum ochraceum, Palustriella decipiens, Philonotis caespitosa, Phiionotis fontána and Philonotis seriata. Typical herbaceous plants include Adenostyles aiiiariae, Allium schoenoprasum, Caltha palustris, Cardamine pratensis, Carex flava, Chaerophyllum hirsutum, Cirsium oleraceum, Dactylis glomerata, Deschampsia cespitosa, Filipendula ulmaria, Molinia caerulea, Phalaris arundinacea, Phragmites australis and Senecio nemorensis agg. The average soil pH is 5.4 (4.9-6.0). This vegetation belongs to the associations Crepido paludosae-Philonotidetum seriatae, Swertietum perennis and Cardaminetum opicii (alliance Swertio perennis-Dichodontion palustris). 3.2 | Transitions between vegetation types Very few plots in birch woodlands and species-poor grasslands changed their vegetation type over time (Figure 4). Most of the changes occurred in the species-rich grasslands between the 1970s and 2000s, which often changed to heathlands and tall-forb vegetation. However, at some sites managed after 2010, the original vegetation type returned. In a few cases, primarily at disturbed sites, transitions occurred between spring and tall-forb vegetation. 3.3 | Changes in species richness and composition In birch woodlands, there were no significant changes in the number of species or proportion of threatened species over time (Figure 5). However, in areas disturbed by the 2019 avalanche, species numbers increased between the 2000s and 2021 (Figure 6b). In open (a)l9705-2000s(n = 128) Birch Birch Heathland Heathland Poor grassland •tato*! Heathland Poor grassland Poor grassland Rich grassland Poor grassland Rich grassland Rich grasslandRich grassland Tall-forb Tall-forb Tall-forb Tall-forb Spring Spring Spring 1970S 2000S (b) 20005-2021 [n = 284.) Birch Birch Heathland Heathland Poorgrassland Poor grassland Rich grassland Rich grassland Tall-forb Tall-forb Spring Spring 20C0S 2021 (C) 2000s-2021, disturbance effects (n = 284) Birch Rirr.h Heathland Heathland Poor grassland wPoor grassland Rich grassland Rich grassland Tall-forb Tall-forb Spring M- Spring 2000S 2021 Undisturbed Disturbed F I G U R E 4 Alluvial diagrams showing transitions between vegetation types between two successive surveys (a, b) and the effects of disturbance regimes between the 2000s and 2021 (c) KLINKOVSKA ETAL. (a) Birch woodlands, number of species per plot 1970s-2000s; n = 10; p-value = 0.443, n.s. • 60Z -:$ 20Applied Vegetation Science ^gt (b) Birch woodlands, number of species per plot 2000s-2021; n = 17; p-value = 0.013, * 9 of 19 Z 40 O | 20 \ J > < (c) Birch woodlands, % of threatened species per plot 1970s-20QQs; n = 10; p-value = 0.432, n.s. < o ľ o* O O C ^ Q o • Q C] í o • o • CD I—1 fl> •<> a * o 0 • • c o o o • O® 1 1 O o o o 0 • u o • • ( J • 3 • O -1.0 ^05 OČÍ ČľŠ Axis 1 O 1970s O 2021 • Birch O Poor grassland O Tall-forb • 2000s O Heathland O Rich grassland O Spring F I G U R E 7 Changes in species composition of vegetation types throughout the whole study period (1970s-2021) analysed using unconstrained ordination (principal coordinate analysis; PCoA). Only the plots surveyed three times were used. Plots with different colours belong to different vegetation types according to the modified TWINSPAN classification. Centroids for each vegetation type (larger symbols) were calculated based on the 1970s classification of plots. The arrows run from the centroid of plots in the 1970s to their centroids in the 2000s and 2021. 4 | DISCUSSION Our comparison of vegetation-plot records from the 1970s, 2000s and 2021 shows that most types of subalpine vegetation in the Hruby Jesenfk Mountains have remained relatively stable over the past half-century. However, species-rich grasslands partially changed into tall-forb vegetation or species-poor heathlands. Vascular plant species richness in these grasslands declined, mainly due to the decline of habitat-specialist species, many of which are on the national Red List. Declines in habitat specialists continued over the past decade but have not been as severe as before. Conservation management and the 2019 avalanche have reduced some of the expanding dominant species, but so far, these factors have not restored the diversity of habitat-specialist and Red List species. 4.1 | Limited evidence of changes in the abiotic environment Mean Ellenberg-type indicator values were relatively stable over time, suggesting a limited change in abiotic environmental factors. Although other studies of alpine grasslands (Evangelista et al., 2016; Gritsch et al., 2016) showed an increase in temperature indicator values, we found such an increase only for certain vegetation types, namely species-rich grasslands between the 1970s and 2000s and heathlands between the 2000s and 2021. This weak response is consistent with the non-significant increase in temperature revealed by local climate station data (Appendix SI). The observed increase in temperature indicator values may be caused by the decline of specialized subalpine species, which have low temperature values, but their decline may also be caused by factors other than climate warming. A t the same time, light indicator values decreased between the 2000s and 2021 in all managed open habitats combined, and specifically in species-rich grasslands. This suggests that light-demanding species are declining while grasslands are becoming denser. It is likely that the changes in grassland plant communities were not primarily caused by abiotic environmental changes such as temperature increase, eutrophication or acidification, although the area was subject to acidification from sulphur deposition in the 1970s-1980s (Rodhe et al., 1995; Hédl et al., 2011) and still receives high deposition of atmospheric nitrogen (Krúpová et al., 2018). This is in contrast to some studies that have found impacts of temperature rise on alpine grasslands, especially on high mountain summits (Walther et al., 2005; Pauli et al., 2007; Steinbauer et al., 2018), or suggested impacts of atmospheric nitrogen deposition (Britton et al., 2009; McGovern et al., 2011; Ross et al., 2012). However, it is consistent with other studies that consider land abandonment in subalpine or alpine grasslands previously used for summer grazing as TA B L E 1 Increasing and decreasing species in the entire data set in the periods 1970s-2000s and 2000s-2021, determined by partial distance-based redundancy analysis (db-RDA) with time as the only constraining variable 1970s-2000s (n = 132, p < 0.001) Increasing Hypericum maculatum*** Aconitum plicatum** Digitalis grand/flora** Crepis paludosa* Vaccinium myrtillus* Trollius altissimus (VU)* Decreasing Crepis mollis (NT)*" Festuca supina (VU)*** Potentilla erecta*** Rhinanthus riphaeus (EN)*** Dianthus superbus (EN)** Festuca rubra** Crepis conyzifolia (VU)** Hieracium prenanthoides (EN)** Hypochaeris uniflora (NT)** Polygonatum verticillatum** Viola lutea subsp. sudetica (EN)** Ranunculus acris** Leontodon hispidus** Ligusticum mutellina (NT)** Cymnadenia conopsea (EN)** Potentilla aurea (NT)* Angelica sylvestris* Ranunculus platanifolius* Silene vulgaris* Primula elatior* Paris quadrifolia* Campanula barbata (VU)* Euphrasia officinalis* Phyteuma spicatum* Geranium sylvaticum* Asarum europaeum* Aconitum variegatum* Hieracium lachenalii* Thalictrum aquilegiifolium* Ranunculus nemorosus* 2000S-2021 (n = 296, p < 0.001) Increasing Acer pseudoplatanus*** Anthoxanthum odoratum agg.*** Dactylorhiza fuchsii*** Festuca supina (VU)*** Luzula sylvatica*** Poa chaixii*** Trientalis europaea*** Veratrum album subsp. lobelianum*** Taraxacum sect. Taraxacum** Galium saxatile** Paris quadrifolia** Chrysosplenium alternifolium** Carex nigra** Ficaria verna** Lilium martagon** Vaccinium myrtillus* Alchemilla sp.* Juncus effusus* Primula elatior* Luzula sudetica* Carex acuta* Poa trivialis* Sorbus aucuparia* Vaccinium vitis-idaea* Carex vaginata (EN)* Ranunculus acris* Ajuga reptans* Helictochloa planiculmis* Fagus sylvatica* Veronica chamaedrys* Decreasing Hypericum maculatum*** Geranium sylvaticum** Betula pubescens subsp. carpatica** Crepis conyzifolia (VU)** Delphinium elatum (EN)** Calamagrostis arundinacea** Hypochaeris uniflora (NT)** Senecio nemorensis agg.** Crepis mollis (NT)** Digitalis grandiflora** Juncus filiformis** Leontodon hispidus* Phyteuma orbiculare (EN)* Parnassia palustris (EN)* Carex echinata* Ranunculus nemorosus* Note: Numbers of plots (n) and p-values of partial db-RDA are indicated. Species with the same trends in both periods are shaded; species with opposite trends are bold. Red List categories critically endangered (CR), endangered (EN), vulnerable (VU) and near threatened (NT) are indicated in parentheses after species names. Species are listed in order of greatest increase or decrease. ***p <0.001, **p <0.01, *p <0.05 the main cause of vegetation change (Dullinger et al., 2003; Vittoz et al., 2009; Czortek et al., 2018). By contrast, the decline in moisture and reaction indicator values in springs may indicate the effects of the dry years 2015-2018 (Buntgen et al., 2021) and the spread of acidophilous species from grasslands to drying springs. 4.2 | Vegetation changes in abandoned subalpine grasslands The main cause of changes in subalpine grasslands in the study area seems to be succession due to the abandonment of summer grazing by sheep and cattle, accompanied by hay-making, which had already been declining since the second half of the 19th century and finally ceased during World War II (Klimeš & Klimešová, 1991). Zeidler et al. (2014) came to the same conclusion based on the assessment of vegetation changes between the 1950s and 2009 in a small part of our study area outside the glacial cirque. However, post-abandonment changes in vegetation must have occurred in the decades prior to our initial survey in the 1970s. Klimeš and Klimešová (1991) assessed changes in the dominant plant species on the summit plateaus in the study area based on botanical literature published between 1910 and 1987. Despite the inconsistent quality of the data, they observed a significant spread of Avenella flexuosa KLINKOVSKA ETAL. 13 of 19 Applied Vegetation Science T A B L E 2 Increasing and decreasing species in individual vegetation types in the periods 1970-2000s and 2000s-2021, determined by partial distance-based redundancy analysis (db-RDA) with time as the only constraining variable Birch 1970s-2000s 2000s-2021 disturbed 2000s-2021 undisturbed Birch n = 20, p = 0.014 n = 14, p = 0.031 n = 20, p = 0.067 Increasing Picea abies* None Veratrum album subsp. lobelianum* Solidago virgaurea* Acer pseudoplatanus* Decreasing None Betula pubescens subsp. carpatica* Betula pubescens subsp. carpatica* Rumex arifolius* Calamagrostis arundinacea* Athyrium distentifolium* Heath n = 14, p = 0.016 n= 10, p = 0.563 n = 32, p = 0.007 Increasing None None Vaccinium vitis-idaea* Trientalis europaea* Decreasing None None None Species-poor grassland n = 14, p = 0.063 n = 8, p = 0.250 n = 48, p<0.001 Increasing Avenella flexuosa* None Festuca supina (VU)*** Decreasing Calamagrostis villosa* None Solidago virgaurea* Festuca sup/no (VU)* Species-rich grassland n = 44, p < 0.001 n = 36, p < 0.001 n = 18, p < 0.398 Increasing Hypericum maculatum*** Trientalis europaea** None Senecio nemorensis agg.* Aconitum plicatum* Decreasing Crepis mollis (NT)*** Geranium sylvaticum*** None Rhinanthus riphaeus (EN)*** Hypericum maculatum*** Crepis conyzifolia (VU)** Potentilla erecta*** Dianthus superbus (EN)** Phleum alpinum (NT)** Viola lutea subsp. sudetica (EN)** Trollius altissimus (VU)** Hypochaeris uniflora (NT)** Crepis mollis (NT)* Ranunculus nemorosus** Ligusticum mutellina (NT)* Festuca rubra** Potentilla aurea (NT)* Gymnadenia conopsea (EN)** Crepis conyzifolia (VU)* Ranunculus acris" Ranunculus nemorosus* Hieracium prenanthoides (EN)** Deschampsia cespitosa* Potentilla aurea (NT)** Leontodon hispidus* Potentilla erecta* Campanula barbata (VU)* Euphrasia officinalis* Angelica sylvestris* Rumex arifolius* Festuca supina (VU)* Tall-forb n = 22, p = 0.002 n = 10, p = 0.063 n = 44, p < 0.001 Increasing Hypericum maculatum* None Dactylorhiza fuchsii* Aconitum plicatum* Decreasing Geranium sylvaticum* None Geranium sylvaticum*** Primula elatior* Epilobium alpestre (NT)** 7 = I T 7 = (Continues) T A B L E 2 (Continued) 1970s-2000s 2000s-2021 disturbed 2000s-2021 undisturbed Birch n = 20, p = 0.014 n = 14, p = 0.031 n = 20, p = 0.067 Aconitum variegatum* Calamagrostis arundinacea** Laserpitium archangelica (EN)* Senecio nemorensis agg.* Digitalis grandiflora* Hypericum maculatum* Delphinium elatum (EN)* Spring n = 14, p = 0.031 n= 10, p = 0.063 n = 34, p< 0.001 Increasing None None Dactylorhiza fuchsii*** Carex nigra* Rumex arifolius* Decreasing None None None Note: The numbers of plots (n) and p-values of the partial db-RDA are indicated. Red List categories critically endangered (CR), endangered (EN), vulnerable (VU) and near threatened (NT) are indicated in parentheses after species names. Species are listed in order of greatest increase or decrease. *"p <0.001, **p <0.01, *p <0.05 in grasslands previously dominated mainly by Nardus stricta, which is considered an indicator of grazing (Klimešová, 1992; Hejcman et al., 2006). In recent decades, Avenella flexuosa also increased in other mountain groups of the Sudetes, namely in the Krkonoše Mountains (1,602m a.s.l.; M . Fabšičová, personal communication 2022) and on Mount Králický Sněžník (1,423m a.s.l.; Husek, 2020). These authors suggested that the abandonment of grazing was the main cause of vegetation change in the summit grasslands. Consequently, the changes observed during our study period should be considered as a continuation of a long successional process that began much earlier. This succession may have also been influenced by grazing by red deer (Cervus elaphus), which are widespread in the study area, and introduced chamois (Rupicapra rupicapra), which find suitable habitat on the rocky slopes of the Velká kotlina cirque. However, both species differ from domestic livestock in their grazing preferences and impact on grassland vegetation. As in some other studies of abandoned mountain pastures (Britton et al., 2009; Mayer et al., 2009; Schwaiger et al., 2022), we found no consistent trend of decline in species richness but noted changes in dominant species. The increase in Avenella flexuosa observed by Klimeš and Klimešová (1991) in species-poor grasslands on the summit plateaus continued into the 2000s. Interesting is the pattern of change in Festuca supina, another Red List species, a tussocky narrow-leaved grass that co-dominates species-poor grasslands on summit plateaus. It declined between the 1970s and 2000s and was gradually replaced by Avenella flexuosa, but it increased again in unmanaged areas between the 2000s and 2021. Its recent increase may be due to the summer droughts of 2015-2018, which resulted in the dieback of some patches of species-poor grasslands. In the subsequent wetter years of 2019-2021, Festuca may have regenerated more quickly than Avenella and increased its abundance. Similar effects of summer drought on narrow-leaved Festuca species have been observed in Central European lowlands (Kroel-Dulay & Garadnai, 2008; Fischer et al., 2020), and negative effects of drought on the dominant species have also been found in experiments in Central European mountain grasslands (Stanik et al., 2021). The most obvious vegetation change, however, was the transition from species-rich grasslands to heathlands dominated by Vaccinium myrtillus (on mesic, oligotrophic soils) or tall-forb vegetation (on moist, nutrient-rich soils). This change occurred mainly between the 1970s and 2000s, and was also observed in the other mountain groups with summit grasslands in the Sudetes (Krkonoše: M . Fabšičová, personal communication 2022; Králický Sněžník: Husek, 2020). It is interesting to note that Klimeš and Klimešová (1991), in their rough assessment of the change in dominant species on summit plateaus between the 1900s and 1980s, considered Vaccinium myrtillus to be a non-increasing, possibly even slightly decreasing species. This suggests that the expansion of Vaccinium myrtillus (and, to a lesser extent, an increase in Vaccinium vitis-idaea) found in our analysis across the study area may have occurred only since the 1980s. It is unclear whether this change was caused by a delayed response to grazing abandonment or by acid rains in the 1970s-1980s, which may have promoted these acidophilous dwarf shrubs in competition with herbaceous species. Another group of expanding species were some tall, nutrient-demanding plants, which spread mainly in the species-rich grasslands in the Velká kotlina cirque. Klimeš and Klimešová (1991) observed the spread of Rubus idaeus, which did not continue in our study period; however, we noticed the spread of Senecio nemorensis agg. (mainly Senecio hercynicus). Both of these species require nutrient-rich, ungrazed habitats. Their spread may be due in part to grazing abandonment and in part to enrichment by atmospheric nitrogen deposition. Besides changes in dominant species, the most alarming trend in vegetation change since the 1970s has been the decline of several habitat-specialist species, including threatened species according to the national Red List (Grulich, 2017), such as Campanula barbata, Crepis conyzifolia, Dianthus superbus, Gymnadenia conopsea, Hieracium prenanthoides, Rhinanthus riphaeus and Viola lutea subsp. sudetica. At the same time, total species richness per plot decreased between the 1970s and 2000s and increased again in the following period. The inconsistent pattern of total species richness could be due to methodological issues: most of the plots in the 2000s were sampled by a single researcher, whereas most of the 2021 plots were sampled by two to three researchers who had the 2000s plot record available. However, the trend of decline in habitat specialists occurred in both periods, although it was less severe between the 2000s and 2021. Declines in rare and habitat-specialist species have been observed in several plot-based studies of vegetation change in mountain and other grasslands (Britton et al., 2009; Prince et al., 2012; Ross et al., 2012; Husek, 2020). Currently, these threatened species are more frequently found at slightly disturbed sites along roads and trails, suggesting that their decline in the study area is primarily due to grassland abandonment and competition from species that have increased in cover. Nevertheless, other causes may also be involved. Acid rains in the 1970s-1980s (Klimeš & Klimešová, 1991) and current atmospheric nutrient deposition may have affected individual plant species, although indicator values for the whole community did not suggest any significant changes. Indeed, Hédl (2004) revealed a dramatic change in beech forest species diversity in a nearby mountain range as a result of acidification. Other possible causes could be pathogens or selective grazing of some rare species by introduced chamois (e.g. Anemone narcissiflora, Campanula barbata, Crepis conyzifolia, Hypochaeris uniflora and the endemic Plantago atrata subsp. sudetica; Klimeš & Klimešová, 1991). Monitoring and maintaining populations of rare plant species, possibly through the restoration of grazing in selected areas, should be among the priorities of nature conservation in the study area. The area is not significantly affected by plant invasions, which is consistent with the generally low number of alien species in mountain areas of the Czech Republic (Chytrý et al., 2021). However, we have detected an increase in Galium saxatile, a species native to northwestern and western central Europe (including the Western Sudetes). It has been spreading in the mountains of the eastern part of the Czech Republic (including the Eastern Sudetes) since the first half of the 20th century, possibly supported by the transport of spruce seedlings for forestry plantations (Štěpánková & Kaplan, 2000). 4.3 I Effects of conservation management Our results show contradictory effects of conservation management practised in some parts of the area during the past decade. Management reduced the cover of expanding Vaccinium myrtillus, which is cut in large areas. It also appears to have slowed or stopped the declines of some habitat-specialist and threatened species in tall-forb and spring vegetation. However, our data show that some habitat-specialist and Red List species of species-rich grasslands continue to decline in managed areas but not in unmanaged areas. This unexpected pattern can be caused by the fact that management was introduced primarily in areas with the greatest successional change in vegetation, where these species had previously been largely lost. Large areas that have been managed since 2017 have actually become established primarily where Vaccinium myrtillus is spreading. Future monitoring is needed to show whether management is restoring species-rich grasslands and populations of their specialized species, as is the case of mountain grasslands in other protected areas (Schwaiger et al., 2022). 4.4 | Effects of avalanches on subalpine birch woodlands In general, plant species diversity tends to be higher on avalanche tracks than in adjacent undisturbed mountain forests (Rixen et al., 2007). Our study took the unique opportunity to analyse vegetation change in a birch woodland on an avalanche track that had been inactive for several decades due to anthropogenic avalanche suppression and then disturbed by a large avalanche in January 2019. Successional changes were observed in the tree layer, namely an increase in spruce (Picea abies) between the 1970s and 2000s, followed by an increase in sycamore maple (Acer pseudoplatanus) associated with a decline in birch in undisturbed areas between the 2000s and 2021. These changes indicate that a latesuccessional forest community was replacing the pioneer birch woodland. However, the composition of the herb layer, species richness and the number of Red List species did not change significantly. After the tree canopy was disturbed by the avalanche, the total species richness increased considerably within two years. All new species that appeared were typical of natural subalpine grasslands, indicating that avalanches periodically increase species richness. This confirms the importance of the natural avalanche regime in maintaining high local plant species diversity, as shown in studies from the Alps (Rixen et al., 2007). The proportion of threatened species did not increase, nor did we observe a significant establishment of alpine plants from higher elevations on the avalanche track. Perhaps these species need more time to become established than just twoyears between the avalanche fall and our resurvey. However, if avalanches become more frequent in the future, aided by the removal of Pinus mugo planted as avalanche barriers, the transport of seeds and propagules of alpine and subalpine plants to lower elevations will likely increase (Erschbamer, 1989; Rixen et al., 2007). In such a case, threatened and alpine species could find suitable habitats for long-term occurrence on avalanche tracks. 5 | CONCLUSIONS Over the past five decades, subalpine vegetation in the Eastern Sudetes has slowly changed, mainly due to the cessation of grazing, although other factors, such as acidification, may have played a role. The most notable change is the decline of species-rich subalpine grasslands and several threatened plant species typical of these grasslands. Because palaeoecological and historical studies support the view that most grasslands in the study area are secondary, developed on formerly forested land due to long-term continuity of grazing, conservation plans should focus on management rather than a non-intervention approach. Conservation management applied over the past decade, replacing former grazing with mowing of stands with spreading competitive herbs and dwarf shrubs, is slowing negative trends in vegetation change. However, to date, there is limited evidence that these trends are reversing and that previous species diversity is being restored. It is likely that the time since 2010, when management became systematic, and especially from 2017, when it was introduced over large areas, has been too short for positive changes in biodiversity to appear. Therefore, we recommend continued mowing, possibly reintroducing extensive grazing in some areas and monitoring changes in plant species diversity. W e also demonstrate that periodic avalanche releases have a positive effect on the maintenance of the subalpine birch woodland and its plant species richness, suggesting that the removal of avalanche barriers was a good decision for biodiversity in this mountain area. Plant community succession on the avalanche track also requires further monitoring. Our network of permanent vegetation plots provides a solid basis for such monitoring. A U T H O R C O N T R I B U T I O N S M.C. and M.K. conceived the idea. M.C. and J.D. supervised the project. M.K. conducted field sampling in the 2000s, and all authors conducted field sampling in 2021. All authors processed the collected materials and prepared the data sets. K.K., A.K., Š.P., V.S., D.S. and M.C. analysed the data. K.S. prepared the maps, including the online version. A.K., J.R. and K.S. conducted the literature review. M . C , K.K., A.K., Š.P. and V.S. wrote the text and all authors revised the text. A C K N O W L E D G M E N T S This article results from research carried out in the course Geobotanical Project at Masaryk University. We thank the Nature Conservation Agency of the Czech Republic and the Forests of the Czech Republic for access and research permits; Leo Bureš for vegetation-plot records from the 1970s; Radek Stencl for shapefiles indicating the extent of conservation management in the study area and for logistic support; Ondrej Hájek and Martin Večera for technical assistance with spatial data and mapping; Ľubomír Tichý and David Zelený for discussion of statistical analyses and help with R scripts; Martina Fabšičová for information on vegetation changes in other mountain groups in the Sudetes. F U N D I N G I N F O R M A T I O N J.D. and M.C. were supported by the Czech Science Foundation (project no. 19-28491X). J.D. was also partly supported by longterm research development project no. RVO 67985939 of the Czech Academy of Sciences. DATA AVAILABILITY S T A T E M E N T The data used in this study are stored in the Zenodo repository (https://doi.org/10.5281/zenodo.7338814). ORCID Klára Klinkovská S> https://orcid.org/0000-0002-1644-2140 Anna Kučerová ) https://orcid.org/0000-0003-1636-0281 Štěpánka Pustková https://orcid.org/0000-0002-9384-4872 Jaroslav Rohel https://orcid.org/0000-0002-4684-4669 Karolína Šlachová https://orcid.org/0000-0002-5984-2654 Vojtěch Sobotka t> https://orcid.org/0000-0002-0668-3258 Daniel Szokala https://orcid.org/0000-0002-3593-1791 JiříDanihelka https://orcid.org/0000-0002-2640-7867 Eva Šmerdová https://orcid.org/0000-0003-4589-6317 Milan Chytrý https://orcid.org/0000-0002-8122-3075 R E F E R E N C E S Bebi, P., Kulakowski, D. & Rixen, C. (2009) Snow avalanche disturbances in forest ecosystems - state of research and implications for management. Forest Ecology and Management, 257(9), 1883-1892. Available from: https://doi.Org/10.1016/j. foreco.2009.01.050 Bracchetti, L., Carotenuto, L. & Catorci, A. (2012) Land-cover changes in a remote area of central Apennines (Italy) and management directions. Landscape and Urban Planning, 104(2), 157-170. Available from: https://doi.Org/10.1016/j.landurbplan.2011.09.005 Britton, A.J., Beale,CM.,Towers, W. & Hewison, R.L. (2009) Biodiversity gains and losses: evidence for homogenisation of Scottish alpine vegetation. Biological Conservation, 142(8), 1728-1739. Available from: https://doi.Org/10.1016/j.biocon.2009.03.010 Bühlmann, T., Hiltbrunner, E. & Körner, C. (2014) Alnus viridis expansion contributes to excess reactive nitrogen release, reduces biodiversity and constrains forest succession in the Alps. Alpine Botany, 124(2), 187-191. Available from: https://doi.org/10.1007/s0003 5-014-0134-y Büntgen, U., Urban, O., Krusic, P.J., Rybníček, M., Kolář, T., Kyncl, T. et al. (2021) Recent European drought extremes beyond common era background variability. Nature Ceoscience, 14, 190-196. Available from: https://doi.org/10.1038/s41561-021-00698-0 Bureš, L. (2022) Fenomén Velká kotlina 1. Flóra, vegetace, fauna [Velká kotlina phenomenon 1. Flora, vegetation, fauna]. Praha: Academia. Bureš, L. & Kočí, M. (2010) Problematika dlouhodobých změn flóry a vegetace subalpínského stupně Hrubého Jeseníku [Long-term changes of flora and vegetation in the subalpine belt of the Hrubý Jeseník Mountains]. Jeseník: Správa CHKO Jeseníky. Bureš, L., Kočvara, R., Kuras, T. & Hradecký, J. (2009) Problematika kleče v Hrubém Jeseníku [Krummholz in the Hrubý Jeseník Mountains], In: Vítek, O. (Ed.) Analýza antropických vlivů v nejcennějších částech CHKO Jeseníky [Analysis ofanthropic effect in the most valuable parts of the Jeseníky protected landscape area]. Praha: Agentura ochrany přírody a krajiny ČR, pp. 9-57. Chytrý, M. (Ed.). (2007) Vegetace České republiky 1. Travinná a keříčková vegetace/vegetation of The Czech Republic 1. Grassland and heathland vegetation. Praha: Academia. Chytrý, M., Danihelka, J., Kaplan, Z., Wild, J., Holubová, D., Novotný, P. et al. (2021) Pladias Database of the Czech Flora and Vegetation. Preslia, 93(1), 1-87. Available from: https://doi.org/10.23855/presl ia.2021.001 Chytrý, M., Tichý, L., Dřevojan, P., Sádlo, J. & Zelený, D. (2018) Ellenberg-type indicator values for the Czech flora. Preslia, KLINKOVSKA ETAL. Applied Vegetation Science 17 of 19 90(2), 83-103. Available from: https://doi.org/10.23855/presl ia.2018.083 Czortek, P., Eycott, A.E., Grytnes, J.-A., Delimat, A., Kapfer, J. & Jaroszewicz, B. (2018) Effects of grazing abandonment and climate change on mountain summits flora: a case study in the Tatra Mts. Plant Ecology, 219(3), 261-276. Available from: https://doi. org/10.1007/sll258-018-0794-6 De Torna, A., Carboni, M., Bazzichetto, M., Malavasi, M. & Cutini, M. (2022) Dynamics of dwarf shrubs in Mediterranean high-mountain ecosystems. Journal of Vegetation Science, 33, el3143. Available from: https://doi.org/10.llll/jvs.13143 Didier, L. (2001) Invasion patterns of European larch and Swiss stone pine in subalpine pastures in the French Alps. Forest Ecology and Management, 145(1-2), 67-77. Available from: https://doi. org/10.1016/S0378-1127(00)00575-2 Dullinger, S., Dirnbóck, T., Greimler, J. & Grabherr, G. (2003) A resampling approach for evaluating effects of pasture abandonment on subalpine plant species diversity. Journal of Vegetation Science, 14(2), 243-252. Available from: https://doi.org/10.llll/ J.1654-1103.2003.tb02149.x Erschbamer, B. (1989) Vegetation on avalanche paths in the Alps. Vegetatio, 80(2), 139-146. Available from: https://doi.org/10.1007/ BF00048037 Evangelista, A., Frate, L, Carranza, M.L., Attorre, R, Pelino, G. & Stanisci, A. (2016) Changes in composition, ecology and structure of high-mountain vegetation: a re-visitation study over 42years. AoB Plants, 8, plw004. Available from: https://doi.org/10.1093/ aobpla/plw004 Fischer, F.M., Chytrý, K., Těšitel, J., Danihelka, J. & Chytrý, M. (2020) Weather fluctuations drive short-term dynamics and long-term stability in plant communities: a 25-year study in a central European dry grassland. Journal of Vegetation Science, 31(5), 711-721. Available from: https://doi.org/10.llll/jvs.12895 Gehrig-Fasel, J., Guisan, A. & Zimmermann, N.E. (2007) Tree line shifts in the Swiss Alps: climate change or land abandonment? Journal of Vegetation Science, 18(4), 571-582. Available from: https://doi. org/10.1111/j'.1654-1103.2007.tb02571.x Gritsch, A., Dirnbóck, T. & Dullinger, S. (2016) Recent changes in alpine vegetation differ among plant communities. Journal of Vegetation Science, 27(6), 1177-1186. Available from: https://doi.org/10.llll/ jvs.12447 Grulich, V. (2017) Červený seznam cévnatých rostlin Č R [The Red List of vascular plants of the Czech Republic]. Příroda, 35, 75-132. Hédl, R. (2004) Vegetation of beech forests in the Rychlebské Mountains, Czech Republic, re-inspected after 60years with assessment of environmental changes. Plant Ecology, 170(2), 243-265. Available from: https://doi.Org/10.1023/B:VEGE.0000021681.83068.53 Hédl, R. & Kočí, M. (2015) Hrubý Jeseník mountains. In: Chytrý, M., Danihelka, J. & Michalcová, D. (Eds) Botanical excursions in Moravia. Brno: Masaryk University, pp. 205-216. Hédl, R., Petřík, P. & Boublík, K. (2011) Long-term patterns in soil acidification due to pollution in forests of the Eastern Sudetes Mountains. Environmental Pollution, 159(10), 2586-2593. Available from: https://doi.Org/10.1016/j.envpol.2011.06.014 Hejcman, M., Pavlů, V , Hejcmanová, P., Gaisler, J., Hakl, J. & Rauch, O. (2006) Farmer decision making and its effect on subalpine grassland succession in the Giant Mts., Czech Republic. Acta Societatis Botanicorum Poloniae, 75(2), 165-174. Available from: https://doi. org/10.5586/asbp.2006.020 Hennekens, S.M. & Schaminée, J.H.J. (2001) TURBOVEG, a comprehensive data base management system for vegetation data. Journal of Vegetation Science, 12(4), 589-591. Available from: https://doi. org/10.2307/3237010 Herzog, R, Buňce, R.G.H., Pérez-Soba, M., Jongman, R.H.G., Sal, A.G. & Austad, I. (2005) Policy options to support transhumance and biodiversity in European mountains: a report on the TRANSHUMOUNT Stakeholder Workshop, Landquart/Zurich, Switzerland, 26-28 May 2004. Mountain Research and Development, 25(1), 82-84. Available from: https://doi.org/10.1659/0276-4741(2005)025[0082:POTST A]2.0.CO;2 Hoiss, B., Gaviria, J., Leingartner, A., Krauss, J. & Steffan-Dewenter, I. (2013) Combined effects of climate and management on plant diversity and pollination type in alpine grasslands. Diversity and Distributions, 19(4), 386-395. Available from: https://doi. org/10.1111/j'.1472-4642.2012.00941.x Hošek, E. (1972) Dosavadní vývoj horní hranice lesa v Jeseníkách [The development of timberline in the Jeseníky Mountains]. Ochrana přírody, 27,101-113. Hošek, E. (1973) Vývoj dosavadního hospodaření v nejvyšších polohách Jeseníků a jeho vliv na horní hranici lesa [The development of management on the summits of the Hrubý Jeseník Mountains and its effect on the timberline]. Campanula, 4, 69-81. Hušek, V. (2020) Dlouhodobé změny subalpínské vegetace Králického Sněžníku [Long-term changes of subalpine vegetation of the Králický Sněžník Mts.]. MSc. thesis. Olomouc: Palacký University. Jeník, J. (1961) Alpinská vegetace Krkonoš, Králického Sněžníku a Hrubého Jeseníku [Alpine vegetation of the Krkonoše, Králický Sněžník and Hrubý Jeseník]. Praha: Nakladatelství Československé akademie věd. Jeník, J., Bureš, L. & Burešová, Z. (1980) Syntaxonomic study of vegetation in Velká Kotlina cirque, the Sudeten Mountains. Folia Ceobotanica et Phytotaxonomica, 15(1), 1-28. Jeník, J., Bureš, L. & Burešová, Z. (1983) Revised flora of Velká Kotlina cirque, the Sudeten Mountains, l-ll. Preslia, 55,25-62,123-142. Kaplan, Z. (2017) Flora and phytogeography of the Czech Republic. In: Chytrý, M., Danihelka, J., Kaplan, Z. & Pyšek, P. (Eds) Flora and vegetation of the Czech Republic. Cham: Springer, pp. 89-163. Kaplan, Z., Danihelka, J., Chrtek, J., Jr., Kirschner, J., Kubát, K., Stech, M. et al. (Eds). (2019) Klíč ke květeně České republiky [Key to the flora of the Czech Republic], 2nd edition. Praha: Academia. Kaufmann, R., Mayer, R., Schallhart, N. & Erschbamer, B. (2021) Effects of climate change vs. grazing exclusion on species diversity over 18years along an elevation gradient in the European Alps. Frontiers in Ecology and Evolution, 9, 640103. Available from: https://doi. org/10.3389/fevo.2021.640103 Klanderud, K. & Birks, H.J.B. (2003) Recent increases in species richness and shifts in altitudinal distributions of Norwegian mountain plants. The Holocene, 13(1), 1-6. Available from: https://doi. org/10.1191/0959683603hl589ft Klimeš, L. & Klimešová, J. (1991) Alpine tundra in the Hrubý Jeseník Mts., the Sudeten, and its tentative development in the 20th century. Preslia, 63(3-4), 245-268. Klimešová, J. (1992) Rostlinná společenstva alpinského stupně se smilkou tuhou (Nardusstricta) v Hrubém Jeseníku I. Charakteristika společenstev ve vztahu k dynamice cenopopulací smilky tuhé [Alpine plant communities with Nardus stricta in the Hrubý Jeseník Mts. (The Sudeten Mts., Czechoslovakia) I. Description of the communities with respect to the dynamics of Nardus stricta coenopopulations]. Preslia, 64(3), 223-239. Krahulec, R, Skálová, H., Herben, T., Hadincová, V , Wildová, R. & Pecháčková, S. (2001) Vegetation changes following sheep grazing in abandoned mountain meadows. Applied Vegetation Science, 4(1), 97-102. Available from: https://doi.org/10.llll/jM654- 109X.2001.tb00239.x Kočí, M. (2005) Inventarizační průzkum národní přírodní rezervace Praděd. Botanika - flóra a vegetace [An inventory of the National Nature Reserve Praděd. Botany - flora and vegetation]. Jeseník: Správa CHKO Jeseníky. Króel-Dulay, G. & Garadnai, J. (2008) The role of disturbances in sand grassland dynamics. In: Kovács-Láng, E., Molnár, E., Króel-Dulay, G. & Barabás, S. (Eds) KISKUN UER: long-term ecological research in the Kiskunság, Hungary. Institute of Ecology and Botany: Vácrátót, pp. 41-44. 18 of 19 Applied Vegetation Science KLINKOVSKA ETAL. Krúpová, D., Fadrhonsová, V., Pavlendová, H., Pavlenda, P., Tóthová, S. & Šrámek, V. (2018) Atmospheric deposition of sulphur and nitrogen in forests of the Czech and Slovak Republic. Central European Forestry Journal, 64(3-4), 249-258. Available from: https://doi. org/10.1515/forj-2017-0050 Kulakowski, D., Bebi, P. & Rixen, C. (2011) The interacting effects of land use change, climate change and suppression of natural disturbances on landscape forest structure in the Swiss Alps. Oikos, 120(2), 216-225. Available from: https://doi. org/10.1111/j'.1600-0706.2010.18726.x Kulakowski, D., Rixen, C. & Bebi, P. (2006) Changes in forest structure and in the relative importance of climatic stress as a result of suppression of avalanche disturbances. Forest Ecology and Management, 223(1-3), 66-74. Available from: https://doi.Org/10.1016/j. foreco.2005.10.058 Lochon, I., Colace, M.-P, Devaux, C , Grigulis, K., Rettinger, R. & Lavorel, S. (2018) Taxonomie and functional facets of the resilience to management of mown subalpine grasslands. Applied Vegetation Science, 21(4), 636-646. Available from: https://doi.org/10.llll/ avsc.12392 MacDonald, D., Crabtree, J.R., Wiesinger, G., Dax, T., Stárnou, N., Fleury, P. et al. (2000) Agricultural abandonment in mountain areas of Europe: environmental consequences and policy response. Journal of Environmental Management, 59(1), 47-69. Available from: https:// doi.org/10.1006/jema.1999.0335 Mackovčin, P., Machar, I., Brus, J. & Žáková, B. (2021) Land use changes in the alpine tree line ecotone in the Hrubý Jeseník Mountains (Czech Republic). Journal of Landscape Ecology, 14(3), 65-87. Available from: https://doi.org/10.2478/jlecol-2021-0014 Mayer, R., Kaufmann, R., Vorhauser, K. & Erschbamer, B. (2009) Effects of grazing exclusion on species composition in high-altitude grasslands of the Central Alps. Basic and Applied Ecology, 10(5), 447-455. Available from: https://doi.Org/10.1016/j.baae.2008.10.004 McGovern, S., Evans, CD., Dennis, P., Walmsley, C. & McDonald, M.A. (2011) Identifying drivers of species compositional change in a semi-natural upland grassland over a 40-year period. Journal of Vegetation Science, 22(2), 346-356. Available from: https://doi. org/10.1111/j'.1654-1103.2011.01256.x Mietkiewicz, N., Kulakowski, D., Rogan, J. & Bebi, P. (2017) Long-term change in sub-alpine forest cover, tree line and species composition in the Swiss Alps. Journal of Vegetation Science, 28(5), 951-964. Available from: https://doi.org/10.llll/jvs.12561 Novák, J., Petr, L. & Treml, V. (2010) Late-Holocene human-induced changes to the extent of alpine areas in the East Sudetes, Central Europe. The Holocene, 20(6), 895-905. Available from: https://doi. org/10.1177/0959683610365938 Novotný, P. & Bureš, L. (2018) Fenomén Velká kotlina 2. Geologické a půdní poměry [The phenomenon of Velká kotlina 2. Geology and soils]. Živa, 66(2), 71-75. Palaj, A. & Kollár, J. (2021) Expansion of phanerophytes above the timberline in the Western Carpathians. Biológia, 76(7), 1991-2003. Available from: https://doi.org/10.1007/sll756-021-00782-l Pauli, H., Gottfried, M., Reiter, K., Klettner, C. & Grabherr, G. (2007) Signals of range expansions and contractions of vascular plants in the high Alps: observations (1994-2004) at the GLORIA master site Schrankogel, Tyrol, Austria. Global Change Biology, 13(1), 147-156. Available from: https://doi. org/10.1111/j'.1365-2486.2006.01282.x Pecháčková, S., Hadincová, V., Munzbergová, Z., Herben, T. & Krahulec, F. (2010) Restoration of species-rich, nutrientlimited mountain grassland by mowing and fertilization. Restoration Ecology, 18(sl), 166-174. Available from: https://doi. org/10.1111/j'.1526-100X.2009.00615.x Pini, R., Ravazzi, C , Raiteri, L., Guerreschi, A., Castellano, L. & Comolli, R. (2017) From pristine forests to high-altitude pastures: an ecological approach to prehistoric human impact on vegetation and landscapes in the western Italian Alps. Journal of Ecology, 105(6), 1580-1597. Available from: https://doi. org/10.1111/1365-2745.12767 Prince, H.E., Bunce, R.G.H. & Jongman, R.H.G. (2012) Changes in the vegetation composition of hay meadows between 1993 and 2009 in the Picos de Europa and implications for nature conservation. Journal for Nature Conservation, 20(3), 162-169. Available from: https://doi.Org/10.1016/j.jnc.2012.01.002 Rixen, C , Haag, S., Kulakowski, D. & Bebi, P. (2007) Natural avalanche disturbance shapes plant diversity and species composition in subalpine forest belt. Journal of Vegetation Science, 18(5), 735-742. Available from: https://doi.Org/10.llll/j.1654-1103.2007.tb025 88.x Rodhe, H., Grennfelt, P., Wisniewski, J., Ágren, C , Bengtsson, G., Johansson, K. et al. (1995) Acid reign '95 - conference summary statement. Water, Air, and Soil Pollution, 85, 1-14. Available from: https://doi.org/10.1007/BF00483684 Roleček, J., Tichý, L., Zelený, D. & Chytrý, M. (2009) Modified TWINSPAN classification in which the hierarchy respects cluster heterogeneity. Journal of Vegetation Science, 20(4), 596-602. Available from: https://doi.Org/10.llll/j.1654-1103.2009.01062.x Ross, L.C., Woodin, S.J., Hester, A.J., Thompson, D.B.A. & Birks, H.J.B. (2012) Biotic homogenization of upland vegetation: patterns and drivers at multiple spatial scales over five decades. Journal of Vegetation Science, 23(4), 755-770. Available from: https://doi. org/10.1111/j'.1654-1103.2012.01390.x Schwaiger, H., Lenzer, B.& Essl, F. (2022) No species loss, but pronounced species turnover in grasslands in the Northern Alps over 25years. Applied Vegetation Science, 25, el2700. Available from: https://doi. org/10.1111/avsc.l2700 Stanik, N., Peppler-Lisbach, C. & Rosenthal, G. (2021) Extreme droughts in oligotrophic mountain grasslands cause substantial species abundance changes and amplify community filtering. Applied Vegetation Science, 24(4), el2617. Available from: https://doi.org/10.llll/ avsc.12617 Steinbauer, M.J., Grytnes, J.-A., Jurasinski, G., Kulonen, A., Lenoir, J., Pauli, H. et al. (2018) Accelerated increase in plant species richness on mountain summits is linked to warming. Nature, 556(7700), 231-234. Available from: https://doi.org/10.1038/ S41586-018-0005-6 Štěpánková, J. & Kaplan, Z. (2000) Galium L. - svízel. In: Slavík, B., Chrtek, J., Jr. & Štěpánková, J. (Eds) Květena České republiky [Flora of the Czech Republic], Vol. 6. Praha: Academia, pp. 122-156. Tichý, L. (2002) JUICE, software for vegetation classification. Journal of Vegetation Science, 13, 451-453. Available from: https://doi. org/10.1111/j'.1654-1103.2002.tb02069.x Treml, V., Jankovská, V. & Petr, L. (2008) Holocene dynamics of the alpine timberline in the High Sudetes. Biologia, 63(1), 73-80. Available from: https://doi.org/10.2478/sll756-008-0021-3 Treml, V. & Migoň, P. (2015) Controlling factors limiting timberline position and shifts in the Sudetes: a review. Geographic Polonica, 88, 55-70. Available from: https://doi.org/10.7163/GPol.0015 Treml, V., Šenfeldr, M., Chuman, T., Ponocná, T. & Demková, K. (2016) Twentieth century treeline ecotone advance in the Sudetes Mountains (Central Europe) was induced by agricultural land abandonment rather than climate change. Journal of Vegetation Science, 27(6), 1209-1221. Available from: https://doi.org/10.llll/ jvs.12448 Varricchione, M., Carranza, M.L., Di Cecco, V., Di Martino, L. & Stanisci, A. (2022) Warmer and poorer: the fate of alpine calcareous grasslands in Central Apennines (Italy). Diversity, 14(9), 695. Available from: https://doi.org/10.3390/dl4090695 Vittoz, P., Randin, C , Dutoit, A., Bonnet, F. & Hegg, O. (2009) Low impact of climate change on subalpine grasslands in the Swiss Northern Alps. Global Change Biology, 15(1), 209-220. Available from: https:// doi.org/10.1111/j'.1365-2486.2008.01707.x Walther, G.-R., BeiRner, S. & Burga, C.A. (2005) Trends in the upward shift of alpine plants. Journal of Vegetation Science, 16(5), 541-548. Available from: https://doi.Org/10.llll/j.1654-1103.2005.tb023 94.x Westhoff, V. & van der Maarel, E. (1978) The Braun-Blanquet approach. In: Whittaker, R.H. (Ed.) Classification of plant communities. The Hague: W. Junk, pp. 289-399. Wickham, H. (2016) ggplot2: elegant graphics for data analysis. New York: Springer. Zeidler, M., Chmelinová, B., Banaš, M. & Lesková, M. (2014) Dlouhodobé změny subalpínské vegetace svahu Petrových kamenů v Hrubém Jeseníku [Long-term changes in subalpine vegetation on the hillside of Mt. Petrovy kameny in the Hrubý Jeseník Mts.]. Příroda (Praha), 32, 5-17. Zelený, D. (2018) Which results of the standard test for community-weighted mean approach are too optimistic? Journal of Vegetation Science, 29, 953-966. Available from: https://doi.org/10.llll/jvs.12688 S U P P O R T I N G I N F O R M A T I O N Additional supporting information can be found online in the Supporting Information section at the end of this article. Appendix SI. Time series of climatic data for the study area. Appendix S2. Synoptic table of species composition of the six vegetation types analysed in this study. Appendix S3. Changes in species richness and the proportion of threatened species in individual types of open habitats (boxplots). Appendix S4. PCoA ordination of observations from two periods (1970s-2000s, 2000s-2021) for all vegetation types combined and each vegetation type separately. Appendix S5. PCoA ordination of observations from the 2000s and 2021 for each vegetation type, with separation of disturbed and undisturbed sites. How to cite this article: Klinkovská, K., Kučerová, A., Pustková, Š., Rohel, J., Šlachová, K., Sobotka, V. et al. (2023) Subalpine vegetation changes in the Eastern Sudetes (1973-2021): Effects of abandonment, conservation management and avalanches. Applied Vegetation Science, 26, el2711. Available from: https://doi.org/10.llll/avsc.12711