The E ﬀ ect of Visitors on the Properties of Vegetation of Calcareous Grasslands in the Context of Width and Distances from Tourist Trails

: Over the last decades, valuable natural areas considered as zones of silence and rest have been increasingly struggling with the problem of mass tourism. In this study, an investigation of the e ﬀ ect of visitors on the properties of vegetation of calcareous grasslands in the context of width and distances from tourist trails is performed. The study was conducted in seven localities in Cracow (southern Poland) involving calcareous grasslands impacted by tourist trails. The results show that the lower height of plants, the greater number of species and the greater percentage of plant cover damaged by trampling in plots located close to the edge of tourist trails, as well as lower total plant cover and greater mean cover-abundance degree per species along narrow pathways. The dominance of meadow and grassland species, as well as the prevalence of native species, suggests that the composition of the examined vegetation has not been drastically changed. In the majority of the study plots, the dominance of hemicryptophytes and chamaephytes, inconsiderable share of phanerophytes and therophytes, as well as the low share of geophytes, were observed. The infrequent occurrence of species presenting Bidens dispersal type along narrow pathways, as well as in plots located close to the edge of tourist trails, suggests low external transport of epizoochorous seeds by passing people, while the prevalence of species presenting Cornus type in plots located away from the edge of tourist trails might be the e ﬀ ect of dung deposition by animals.


Introduction
The phenomenon of overtourism, which is understood as the opposite of sustainable tourism, has been known and described so far mainly in the urban areas (e.g., [1][2][3][4]). However, it is also beginning to appear in regions considered as zones of silence and rest [5][6][7]. According to numerous authors [8][9][10][11][12][13][14][15][16], valuable natural areas are increasingly struggling with the problem of mass tourism. The negative consequences of the excessive tourist traffic represented, among others, by excessive water intake and sewage production, waste generation, noise, increased probability of fire initiation, synanthropisation of flora and fauna, scaring away of animals, as well as changes in the structure of biocoenoses were repeatedly noted in protected areas. Another consequence of excessive tourist traffic is trampling, which leads to the creation of informal trails. The trampling might contribute to changes in vegetation through mechanical damage of plants and species loss [17,18], as well as the influence on seed germination, seedling establishment, growth functions after establishment, vigour and biomass production, as well as flowering and fruiting [19]. Moreover, the trampling might improve occur mainly in the warm half-year as single days or two-day periods. The atmospheric precipitation achieves ca. 690 mm and the peak of precipitations occurs in July.
Due to their location within the border of Cracow and easy access by public transport, all the study sites are exposed to the recreational activities of citizens and tourists. "The Cracow City Forest Trail" is a marked walking and cycling route leading, among others, through Bogucianka, Górka Pychowicka, Uroczysko Kowadza and Uroczysko Wielkanoc. Moreover, Perzanowska [47] pointed out that Fort Bodzów and Uroczysko Kowadza are perfect places suitable for outdoor recreation (Table 1).

The Overview of the Study Design and Characteristics of the Study Plots
In each location, two visitor-created (informal) pathways were selected: a narrow one (up to 50 cm in width) and a wide one (at least 115 cm in width) based on the assumption that the width of the path is positively correlated with the intensity of tourist traffic ( Table 1). The narrow trail can be used by one person, while the wide trail allows at least two people to pass in one direction or to pass each other. The pairs of 1 × 1 m research plots were established along each pathway. The pairs were systematically distributed every 2 m (alternately on both sides of the pathway). Each pair consisted of a plot labelled CL situated close to the edge of the pathway at a distance of 10 cm, and a plot labelled FU located much further at a distance of 150 cm from the plot CL ( Figure 1). The distance from the edge of the trail was chosen arbitrarily on the basis of the behavior of tourists. Plots CL were established in places often trampled by tourists to avoid the trail after rainfalls, when the surface is muddy and slippery. Plots FU were established at a greater distance from the trail where descending from the pathway is due to willingness of taking photographs, curiosity, repose, or other causes. Sustainability 2020, 12,   The measurements of the abiotic habitat conditions, which were tested in particular plots by using the handheld device BIOWIN, evidenced that light intensity ranged from 740.0 Lx to 2000.0 Lx, soil moisture ranged from 1.0 to 6.7, whereas soil pH was from 7.1 to 7.8 ( Table 2). The ANOVA analysis of main effects evidenced that the values of light intensity were similar in both the narrow and wide patches, the soil moisture was significantly greater in the plots CL than in the plots FU (F = 9.85, p < 0.01), and the soil reaction achieved greater values in the plots located along the wide pathways than in the narrow ones (F = 10.30, p < 0.01). The measurements of the abiotic habitat conditions, which were tested in particular plots by using the handheld device BIOWIN, evidenced that light intensity ranged from 740.0 Lx to 2000.0 Lx, soil moisture ranged from 1.0 to 6.7, whereas soil pH was from 7.1 to 7.8 ( Table 2). The ANOVA analysis of main effects evidenced that the values of light intensity were similar in both the narrow and wide patches, the soil moisture was significantly greater in the plots CL than in the plots FU (F = 9.85, p < 0.01), and the soil reaction achieved greater values in the plots located along the wide pathways than in the narrow ones (F = 10.30, p < 0.01). Table 2. The mean light intensity (Lx) (± SD), soil moisture (± SD) and soil pH (± SD) noted in closer (CL) and further (FU) plots located along the narrow (width ≤ 50 cm) and wide (width ≥ 115 cm) pathways situated within the investigated study site.

The Field Trial
Field study was conducted in July 2019. The height of the tallest plant shoot from the ground level to the top of the stem was measured in each study plot using a folder tape. Within each plot, the total percentage of plant cover and the percentage of plant cover damaged by trampling were visually estimated using a cover-abundance scale with an interval of 5%. Furthermore, the vascular plant species growing in the herbaceous layer within each plot were inventoried. The seedlings and saplings were removed and determined according to Csapodý [48] and Muller [49]. The degree of cover-abundance of each species was visually estimated according to a scale of Braun-Blanquet [50]. The explication of particular points of scale is as follows: • -"+"-species covers less than 1% of the studied area, • -"1"-species covers 1-5% of the studied area, • -"2"-species covers 6-25% of the studied area, • -"3"-species covers 26-50% of the studied area, • -"4"-species covers 51-75% of the studied area, • -"5"-species covers 76-100% of the studied area.

The Species Groups
To assess the species response to tourist activities, we selected four traits (i.e., habitat affiliation, life form, native or alien status in the flora and mode of seed dispersal) that were "ecologically meaningful" in accordance with the ability to persist in the stressful conditions caused by man (trampling) and accompanying animals (ground browning, wallowing). The traits of particular vascular plant species evidenced in the plots are presented in Table A1.
The life form of species proposed by Raunkiaer [52] was assigned according to the database "Ecological Flora of the British Isles" [53]. The following life forms were distinguished: phanerophytes, chamaephytes, hemicryptophytes, geophytes and therophytes. In the case of missing data the publication of Ellenberg et al. [54] was included.
The origin of species was assigned based on the database "Alien species in Poland" [57]. The alien species was understood as a species or lower taxon, introduced outside its natural past or present range that might survive and subsequently reproduce. The native species to a given area is a species that has been observed in the form of a naturally occurring and self-sustaining population from historical times.
Data concerning habitat affiliation and life form of Erigeron acris ssp. serotinus (Weihe) Greuter, which are lacking in the aforementioned sources, were taken from the publication of Pliszko [58]. Plants identified solely to the rank of a genus (e.g., Carex sp.) were excluded from the analyses. Moreover, the cultivated plants such as Cerasus vulgaris Mill. and Malus domestica Borkh. were excluded from the analysis of habitat affiliation.

The Data Analysis
The mean height of the tallest plant shoot, number of species, percentage of aboveground biomass damage by trampling, total plant cover percentage, as well as degree of cover-abundance of a particular species (±SD) were calculated in the research plots CL and FU, as well as in the plots located along narrow and wide pathways in each study site.
The normal distribution of the untransformed data was tested using the Kołmogorov-Smirnov test, whereas the homogeneity of variance was verified using the Levene test at the significance level of p < 0.05. The ANOVA analysis of main effects followed by the post-hoc Tukey HSD test was applied to check the statistical significance of the effect of pathway width and plot distance from the pathway on (i) the height of the tallest plant shoot, (ii) the number of species, (iii) total plant cover percentage, (iv) percentage of plant cover damaged by trampling, as well as (v) degree of cover-abundance of a particular species within the study plots. The aforementioned analyses were computed using a STATISTICA software (version 13). The chi-square test with Yates correction for continuity was applied to check whether there were significant differences among the plots located along the narrow and wide pathways, as well as in the plots situated at a diverse distance from the border of trails in cover-abundance degree of species showing various habitat affiliation, life form, seed dispersal mode and species origin. The chi-square test was conducted using the interactive calculation tool [59].

The Plant Cover and Richness
The mean height of the tallest plant ranged from 66.2 cm to 123.7 cm ( Figure A1) and it was similar in the plots situated along the narrow and wide pathways (F = 0.08, p = 0.77), at the same time it was significantly greater in the plots FU than in the plots CL (F = 31.64, p < 0.001). The mean number of species per plot amounted from 11.1 to 17.6 ( Figure A1) and it did not differ in the plots situated along the narrow and wide pathways (F = 0.17, p = 0.67) but it was greater in the plots CL than FU (F = 5.39, p ≤ 0.05).The mean percentage of total plant cover achieved from 43.0 to 85.0 ( Figure A2). The statistical analysis evidenced, that the values noted in the plots located along the narrow pathways were significantly lower than along the wide ones (F = 5.19, p ≤ 0.05) and they were similar in the plots CL and FU (F = 0.22, p = 0.63). The mean percentage of plant cover damaged by trampling achieved from 0.0 to 82.5 ( Figure A2). It was similar in plots located along narrow and wide pathways (F = 0.01, p = 0.90) and remarkably greater in the plots CL than in the plots FU (F = 145.93, p < 0.001). The mean cover-abundance degree of particular species per plot according to the Braun-Blanquet scale ranged from 0.1 to 0.7 ( Figure A3). The values recorded along the narrow pathways were greater than along the wide ones (F = 12.00, p < 0.001), whereas values noted in the plots CL and FU were similar (F = 0.15, p = 0.69).

The Species Groups Characteristics
Along the narrow and wide pathways, meadow and grassland species prevailed over ruderal plants, while forest taxa occurred sporadically. The statistical analysis showed a lack of differences in most study areas ( Figure 2). The similar spectra of habitat affiliations were observed in the plots CL and FU in the majority of the study sites. Only in one study area, in closer plots, was the considerable dominance of ruderal taxa evidenced ( Figure 3). The statistical analysis proved significant differences in life form spectra among the narrow and wide pathways (Figure 4), as well as among the plots CL and FU ( Figure 5) regardless of the dominance of hemicryptophytes and chamaephytes, slight presence of phanerophytes and therophytes, as well as the lowest cover-abundance degree of geophytes in the majority of the study areas. The cover-abundance degrees of species with particular dispersal mode occurring in plots situated along pathways with different width varied significantly ( Figure 6). In the majority of plots situated along the narrow pathways, the lowest cover-abundance degree showed species with dispersal type Bidens, while taxa with Allium, Cornus or Epilobium dispersal type prevailed in at least one study site. In the plots located along the wide pathways, different patterns of dispersal mode spectra were noticed. The statistical analysis showed that also the cover-abundance degrees of species with particular dispersal mode occurring in the plots CL and FU differed significantly ( Figure 7). In most plots CL, species with Bidens type occurred sporadically, whereas in plots FU species with type Cornus dominated in the majority of places. Despite the prevalence of native species over alien species, the statistical analysis showed significant differences among the narrow and wide pathways (Figure 8), as well as among the plots CL and FU ( Figure 9).  Figure 2. The mean cover-abundance degree of a species (± SD) affiliated to the forest (F), meadow (M), grassland (G), and ruderal habitats (R) per plot located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥ 115 cm) pathways situated within the investigated study sites. The statistical significance level of χ 2 test (df = 3): ns -not significant, * p ≤ 0.05, ** p < 0.01, *** p < 0.001.

Figure 2.
The mean cover-abundance degree of a species (± SD) affiliated to the forest (F), meadow (M), grassland (G), and ruderal habitats (R) per plot located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥ 115 cm) pathways situated within the investigated study sites. The statistical significance level of χ 2 test (df = 3): ns -not significant, * p ≤ 0.05, ** p < 0.01, *** p < 0.001.      Figure 4. The mean cover-abundance degree of a species (± SD) representing phanerophytes (PH), chamaephytes (CH), hemicryptophytes (H), geophytes (G) and therophytes (T) per plot located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥115 cm) pathways within the investigated study sites. The statistical significance level of χ 2 test (df = 4) is given in Figure 2.             Figure 8. The mean cover-abundance degree of alien (A) and native (N) species (± SD) per plot located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥115 cm) pathways within the investigated study sites. The statistical significance level of χ 2 test (df = 1) is given in Figure 2.

The Plant Cover Characteristics
The performed observations evidenced that the values of height of the tallest plant shoots, species richness, as well as the percentage of plant cover damaged by trampling did not differ in the narrow and wide pathways. The height of the tallest plants achieves greater values in the distant plots than in the plots situated closely to the pathways, while species richness and percentage of plant cover damaged by trampling show an inversed trend. Therefore, our hypotheses that (i) the height of plants is lower in plots located near pathways than that in plots located away from pathways, (ii) the percentage of plant cover damaged by trampling is higher in plots located near pathways than that in plots located away from pathways, and (iv) the number of species is higher in plots located near pathways than that in plots located away from pathways can be fully accepted. At the same time, the hypothesis (iii) that the percentage of plant cover damaged by trampling is higher along narrow pathways than that along wide pathways must be rejected.
The lower height of the tallest plant in plots situated close to the tourist trails might be an effect of damage to plant tissue, especially shoot fractures by passers-by. Also, a much greater percentage of damaged plant cover by trampling in the close plots might be linked with the activity of visitors bypassing the pathways or descending from them due to the taking of photographs, curiosity, repose or other causes. Such tourist dispersion was frequently observed in trails by numerous authors e.g., [32,34].
The obtained results proving much greater species richness close to tourist trails correspond with the findings of Root-Bernstein and Svenning [60], while other investigators observed an inversed tendency [61]. A higher number of plant species near the tourist/recreation pathways within the calcareous grasslands can be explained by the fact that plant diaspores are easily transported on shoes, clothes and vehicles. Similarly, it was evidenced by Tikka et al. [62] that road verges serve as dispersal corridors for grassland species. It is also worth mentioning that some plants such as Lolium perenne L. and Trifolium repens L. tolerate trampling and often occur on roadside verges [23,63]. However, their abundance in the examined plots was rather low (except some plots with L. perenne). Moreover, tourist trails are often used for migration by wild animals (e.g., [64]), which may also promote the plant dispersal along the pathways.
The performed observations evidenced that the distance from the trails does not have an influence on total plant cover percentage, as well as the cover-abundance degree of a particular species per plot. The obtained results are not consistent with the studies of Jägerbrand and Alatalo [43], who noted that due to the decrease in understory cover, the abundance of litter, rock and soil increased with the proximity to the trail in alpine heath. The noted in the present studies lower values of total plant cover in the plots situated along the narrow pathways might indicate the occurrence of a greater number of gaps where bare substratum is visible. It is worth mentioning that such openings in continuous turf are considered as safe sites for seedling recruitment sensu Harper [65], regeneration niche sensu Grubb [66] and space "free from competition" [67]. The beneficial role of small-scale gaps enabling spontaneous recruitment and establishment of seedlings in calcareous grasslands was repeatedly proved in naturally originated [68], as well as experimentally made openings [69,70]. In the present study, the recorded greater mean cover-abundance degree of a species along the narrow pathways than along the wide ones might suggest the successful generative propagation and/or undisturbed vegetative spread, leading to the multiplication of individuals and/or ramet number, as well as an area of individuals presumably owing to the non-intensive use of trails by visitors. The increase of pathway width as the result of the augmentation of the intensity of tourist traffic, as well as the frequency of passes, was previously recorded among others by Kiszka [16].

The Species Groups Characteristics
Our study, showing that regardless of pathway width and distance from the trail, meadow and grassland species prevailed over ruderal plants, while forest taxa occurred sporadically, suggest that hypothesis (v) about the variability of habitat affiliation spectra must be rejected. Moreover, we evidenced the dominance of native species over alien species irrespective of pathway width and distance from the edge of the trail. Although the area of calcareous grasslands in Cracow has significantly decreased over the last decades [71], their semi-natural value is still high [72]. However, the presence of some alien species such as Erigeron annuus (L.) Desf., E. canadensis L., Robinia pseudoacacia L., Solidago canadensis L. and Vicia grandiflora Scop., which are invasive in Poland [73], suggests the negative effect of human activities on calcareous grasslands in the area of the city. These species can be easily introduced to calcareous grasslands from nearly located roadside verges, abandoned allotment gardens and waste ground. Nevertheless, it might be stated that despite the significant statistical differences regarding the presence of native and alien species along the narrow and wide pathways, as well as among the plots CL and FU, the dominance of native taxa suggests the rejection of hypothesis (v) about the variability of species origin spectra.
Also, in spite of recorded differences in the degree of cover-abundance of species representing particular life forms, depending on trail width and plot location, the similar patterns of life form spectra noticed in the majority of the study areas indicate the rejection of hypothesis (v) about the variability of species life form spectra. The performed investigations evidencing a dominance of hemicryptophytes and chamaephytes supports the findings of Dobay et al. [74], arguing that the species representing the aforementioned life forms are often found in grassland areas. Roovers et al. [38] observed the dominance of hemicryptophytes regardless of level recreational use in meadows, heaths, and forests, whilePescott and Stewart [75] added that vegetation dominated by hemicryptophytes recovers from trampling to a greater extent than vegetation dominated by other life forms. The observed in the present studies scarce number of phanerophytes is not remarkable considering the occurrence of forests in the vicinity of the study sites, whereas the slight abundance of therophytes seems to be very surprising and might be an effect of slight occurrence of diaspores in the soil seed bank and/or unsuitable conditions for seedling recruitment. Additionally, it is worth mentioning, that Skłodowski et al. [39], as well as Zdanowicz and Skłodkowski [40], found the greater number of therophytes along wide pathways than along narrow ones in forests. Apart from this, other researchers recorded a considerably greater share of therophytes in the borders of trails than in more distant sites in forests [37] and meadows [38].
According to Sádlo et al. [56], the dispersal strategies of Allium, Bidens, Cornus, Epilobium and Lycopodium are found within the plants occupying dry grasslands. In our study, we evidenced the presence of species with the strategies of Allium, Bidens, Cornus and Epilobium. The occurrence of different patterns of dispersal mode spectra among the plots located along the narrow and wide pathways, as well as among the plots CL and FU, allows confirming the hypothesis (v) about the variability of species dispersal mode spectra. Simultaneously, the results showing the lowest cover-abundance degree of species presenting Bidens dispersal type (mainly epizoochory and autochory, as well as endozoochory) in plots situated along narrow pathways, as well as in plots located close to the trail edge, might suggest low activity of tourist and visitors passing by pathways in the external transport of diaspores possessing mechanisms to adhere to clothes equipment, vehicles and animals. On the other hand, the prevalence of taxa with Cornus type (endozoochory and autochory) in plots located at a greater distance might be an effect of dung deposition by animals. The considerable recruitment of endozoochorous species seedlings from dung samples was observed in numerous habitats (e.g., [76][77][78]).

Conclusions
The height of the tallest plant shoots, species richness, as well as the percentage of plant cover damaged by trampling did not differ in the narrow and wide pathways. The significantly lower height of plants in close plots and the greater species number and percentage of plant cover damaged by trampling recorded there is the effect of passers-by contributing to the mechanical fracture of plant organs and the dissemination of diaspores. The distance from trails does not impact the total plant cover percentage, as well the cover-abundance degree of a particular species per plot. The lower value of plant cover percentage along narrow trails creates the opportunities for successful generative propagation and/or vegetative spread, resulting in a greater mean cover-abundance degree of a species.
The dominance of meadow and grassland species over ruderal plants and sporadic occurrence of forest taxa, as well as the prevalence of native species irrespective of pathway width and distance from trail edge, suggests that the composition of the examined patches of grasslands has not been drastically changed by secondary succession and human activity. The dominance of hemicryptophytes and chamaephytes, slight presence of phanerophytes and therophytes and the low cover-abundance degree of geophytes was observed in the majority of study areas regardless of path width and distance from the edge of the trail. The lowest cover-abundance degree of species presenting Bidens dispersal type in plots situated along narrow pathways, as well as in plots located close to the trail edge might suggest the low activity of visitors passing by pathways in the external transport of epizoochorous seeds. The prevalence of taxa with Cornus type in plots located at a greater distance might be an effect of deposition of dung containing endozoochorous seeds by animals.
The investigations performed enlarge the current state of knowledge about the properties of vegetation in the vicinity of visitor-created (informal) tourist trails in calcareous grasslands -areas of high conservation value. Our results can be applied in further studies to evaluate the temporal changes of species composition and plant traits, as well as for comparison with other popular semi-natural areas, where trampling is also an issue.
According to assumptions of plans of protection [79], to preserve the calcareous grasslands in the Natura 2000 areas, it is important to make awareness-raising efforts among the local population and tourists through educational campaigns. Moreover, the monitoring of frequently visited patches, enabling the identification of existing and potential threats caused by visitor activities, is desired.

Conflicts of Interest:
The authors declare no conflict of interest. The funder had no role in the design of the study, in the collection, analyses or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.
Appendix A Table A1. The characteristics of species found in the plant cover of the studied calcareous grasslands regarding habitat affiliations according to Matuszkiewicz [61], life form according to Fitter and Peat [63], dispersal type according to Pladias [55]. Database of the Czech flora and vegetation [65], and origin according to Alien species in Poland [67]. The number of species Figure A1. The mean height (cm) of the tallest plant (± SD) and number of species (± SD) in the closer (CL) and further (FU) plots located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥115 cm) pathways situated within the investigated study sites. Figure A1. The mean height (cm) of the tallest plant (± SD) and number of species (± SD) in the closer (CL) and further (FU) plots located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥ 115 cm) pathways situated within the investigated study sites. Pathway: x x Figure A2. The mean percentage of total plant cover (± SD) and the percentage of plant cover damaged by trampling (± SD) in the closer (CL) and further (FU) plots located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥ 115 cm) pathways situated within the investigated study sites. Figure A2. The mean percentage of total plant cover (± SD) and the percentage of plant cover damaged by trampling (± SD) in the closer (CL) and further (FU) plots located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥ 115 cm) pathways situated within the investigated study sites. Pathway: Figure A3. The mean cover-abundance degree of a particular species according to the Braun-Blanquet scale (± SD) in the closer (CL) and further (FU) plots located along the narrow-N (width ≤ 50 cm) and wide-W (width ≥ 115 cm) pathways situated within the investigated study sites.