The influence of habitat conditions on abundance and selected individual traits of Arum alpinum in the communities Populetum albae and Tilio cordatae–Carpinetum betuli (Western Carpathians)

Abstract

Studies of selected habitat conditions, as well as spatiotemporal variability of the number and selected traits of individuals of the species Arum alpinum were carried out in 2020–2021 in the foothills of the Western Carpathians. The investigations were conducted in permanent patches located in the Golesz nature reserve (Patch I), near the village of Markowce (Patch II), in the Kozigarb nature reserve (Patch III), and near the village of Żółków (Patch IV). Patches I and III were established in a Tilio cordatae–Carpinetum betuli oak–hornbeam forest with undergrowth dominated by low-growth vegetation with narrow leaves, whereas Patches II and IV were established in a Populetum albae riparian forest with undergrowth dominated by broad-leaved species. The most abundant population of A. alpinum was noted in Patch I, but substantial numbers of both vegetative and reproductive individuals were also present in Patches II and IV. The occurrence of temporal variability of individual traits increased from its lowest level in Patch IV, through Patches I and II, to its highest level in Patch III. The significant positive correlation in all populations between length of petioles and blade dimensions, as well as between length of generative stems and infructescence traits confirmed previous findings. Moreover, we showed that A. alpinum was not closely affiliated with a specific forest community. Significant shading and moist nutrient-rich soils are suitable for this species, while dry soils and excessive insolation may limit the flowering and fruiting of individuals.

摘要

栖息地条件对白杨与枫树林群落中高山天南星的丰富度和选定个体特征的影响

本研究于2020至2021年在西喀尔巴阡山脉的山麓地区进行，研究了高山天南星(Arum alpinum)的物种数量和选定特征的空间-时间变异。调查分别在4个永久样地中进行(Patch I、 II、 III、 IV)。Patch I和Patch III位于以低矮植被为主的枫树-鹅耳枥林(Tilio cordatae-Carpinetum betuli)林地中，而Patch II和Patch IV位于以阔叶物种为主的白杨(Populetum albae)河滩林中。调查结果表明，高山天南星最丰富种群出现在Patch I，但Patch II和Patch IV中也存在大量的营养体和繁殖个体。个体性状的时间变异性在Patch IV最低，其次是Patch I和Patch II，在Patch III最高。所有种群中均发现叶柄长度与叶片尺寸以及有性茎长度与果序特征之间的关系显著呈正相关，这一发现证实了先前的研究。这些研究结果表明，高山天南星与特定的森林群落没有密切关联。充足的遮荫和湿润富养分的土壤对该物种有利，而干燥的土壤和过多的日照可能会限制其个体的开花与结果。

INTRODUCION

The genus Arum includes 29 species, the majority of which are highly concentrated in the Mediterranean, the Balkans, and the Middle East, where the genus has its centre of diversity (Boyce 1993, 2006; Lobin et al. 2007; Croat and Ortiz 2020). Arum inflorescences are known for their fetid odour, which has a fascinating relationship with their fraudulent pollination mechanism and is a remarkable example of evolution (Gibernau et al. 2004; Linz et al. 2010; Gfrerer et al. 2022; Szenteczki et al. 2022). Arum species are found in a variety of habitat types, from open, hot and dry areas in the Mediterranean to forested areas in northern Europe with a wetter and cooler climate (Croat and Ortiz 2020). Generally, Arum species grow on nutrient-rich soils with a neutral or slightly alkaline pH (Bedalov et al. 2006; Kozuharova et al. 2014a).

The most problematic taxa in the taxonomy of the genus Arum include Arum cylindraceum Gasp. and Arum alpinum Schott & Kotschy. The former was long considered a local and poorly understood species. The taxonomic history of the second has been more frequently debated. Arum alpinum was first described from the Transylvanian Alps (Bedalov and Küpfer 2005). Many authors classified it as Arum maculatum L., A. maculatum subsp. danicum and later recombined as A. orientale subsp. danicum. It has also often been identified as A. maculatum var. immaculatum or included in A. italicum (Bedalov and Fischer 1995; Bedalov and Drenkovski 1997). Based on detailed morphological, taxonomic, and phytogeographical studies, it was restored to the status of a separate species (Terpό 1973; Bedalov and Küpfer 2005). In the following years it was treated differently, as a subspecies of A. orientale, a synonym of A. orientale subsp. orientale, and a microspecies within A. maculatum agg. (Terpό 1973; Bedalov and Fischer 1995; Bedalov and Drenkovski 1997; Bedalov et al. 2006). It was subsequently restored to A. alpinum (Boyce 1993). The above overview of the taxon’s history shows that it has often been misidentified. Until the turn of the 20th and 21st centuries, the status of this plant was the subject of many systematic controversies; however, according to Boyce (2006), A. alpinum should be called A. cylindraceum because this name appeared first. Despite these findings, in Poland this taxon is still known as A. alpinum (Mirek et al. 2020).

Previous research on A. alpinum (syn. A. cylindraceum) focussed on its distribution (Bedalov and Fischer 1995; Bedalov and Drenkovski 1997; Draper and Rosselló-Graell 1997; Dajdok and Kącki 2001; Bedalov et al. 2006; Croat and Ortiz 2020; Jeanmonod and Schlussel 2012; Kozuharova et al. 2014b; Trigas et al. 2021) and taxonomic complexities (Terpό 1973; Bedalov and Küpfer 2005; Boyce 2006; Lendel et al. 2007), while in recent years more attention has been paid to phylogenetics and biogeography (Lendel et al. 2006; Mansion et al. 2008; Espíndola et al. 2010; Linz et al. 2010; Joudi et al. 2016). In some works, the issue of habitat preferences of the species has been discussed, but usually this presented general information, not supported by detailed analyses of the abiotic conditions of the habitat (Barć et al. 2004; Bedalov et al. 2006; Lendel et al. 2007; Kozuharova et al. 2014a; Wójcik and Ziaja 2015). In turn, biometric studies covered basically only the southern part of the range of the species. A few studies in this field were carried out in Corsica (Fridlender 1999) and Croatia (Lendel et al. 2007), while from the Carpathians comparative research analyses prepared by Terpό (1973) are known, based on herbarium specimens. Only recently, the results of studies on the variability of selected features of A. alpinum in the fertile Carpathian beech forest in the foothills of the Carpathians have been published (Wójcik and Kostrakiewicz-Gierałt 2020). However, the state of our understanding of the habitat requirements, population size, and diversity of features of this rare species is still insufficient. This is especially at of the northern end of its range, where the species occurs in beech forests, oak–hornbeam forests and riparian forests. With this in mind, research addressing the variability of selected features of individuals and populations of A. alpinum in riparian forests and hornbeam forests located in the Carpathian foothills was carried out.

Considering the current state of knowledge, we hypothesize that (i) the habitat conditions differ between the localities of the species; (ii) the habitat conditions exert an influence on the abundance of A. alpinum individuals; (iii) the individual traits of A. alpinum show spatial and temporal variability; (iv) the length of leaf petioles is positively correlated with blade dimensions, and the length of the stem in generative individuals is positively correlated with the infructescence length and the number of fruits.

MATERIALS AND METHODS

Study species

Individuals of A. alpinum (syn. A. cylindraceum) create a vertical rhizome in the shape of a flattened tuber. The leaves are long and usually slightly arrow shaped. The inflorescence consists of two parts: the spadix and the spathe. The spadix has unisexual flowers and adaptations to generate heat, while the spathe is a modified bract surrounding the spadix (Boyce 1993; Bedalov et al. 2006). The structure of the inflorescence attracts pollinating insects by producing heat and volatile compounds. The combination of scent emission and the presence of hairs in the upper part of the flower chamber (acting as a barrier) is a key feature of successfully catching arthropods during the pollen production period. After pollination, the spathe wilts and releases the insects. The fruit is a red berry with 1–3 seeds (Gibernau et al. 2004; Bröderbauer et al. 2013; Gfrerer et al. 2022).

The range of this species covers Europe and Asia Minor. The western geographical limit of A. alpinum occurs in Spain (Draper and Rosselló-Graell 1997), the eastern reaches the regions of north-eastern Turkey (Bedalov and Fischer 1995), while the southernmost sites are located in Italy (Sicily), the Balkan Peninsula, and on the Greek islands (Evvia, Crete, Zakinthos) (Bedalov and Fischer 1995; Bedalov and Drenkovski 1997; Trigas et al. 2021). The northern limit of its range runs through Denmark, Germany, and Poland (Terpό 1973; Dajdok and Kącki 2001; Boyce 2006). In Poland, A. alpinum grows in the southern part of the country, with most of its sites located in lower mountain locations in the Carpathians and Sudetes and in the adjacent areas (Zając and Zając 2001). In the northernmost sites in Denmark, the species grows in the community of Carpinus betulus–Anemone nemorosa (Lawesson 2000), and in the upland part of Poland it occurs in the oak–hornbeam forest Tilio cordatae–Carpinetum betuli (Medwecka-Kornaś and Kornaś 1963). In the Carpathians, it is recorded in communities of the order Fagetalia sylvaticae. It most often occurs in various forms of oak–hornbeam forests from the Tilio cordatae–Carpinetum betuli association (Święs 1982; Hadač and Terray 1989; Wróbel 2007; Kollár et al. 2009; Towpasz and Stachurska-Swakoń 2010), in beech forests from the Dentario glandulosae–Fagetum (Święs 1982; Barć et al. 2004; Wójcik and Ziaja 2015; Wójcik and Kostrakiewicz-Gierałt 2020) and Phyllitidi–Fagetum (Nicolin and Imbrea 2009) associations, as well as in riparian forests and riverside thickets (Święs 1982; Hadač and Terray 1989; Kollár et al. 2009; Petrášová and Jarolímek 2012). It also grows sporadically in linden-sycamore forests of the association Tilio platyphyllis–Acerion pseudoplatani (Mijal 2015) and in shrub thickets (Jarolímek and Mucina 1979; Hadač and Terray 1989). In the areas south of the Carpathians, it occurs in the forest-steppe zone: oak–ash groves from the Fraxino pannonicae–Ulmetum association (Kevey 2020) and oak–hornbeam forests from the Convallario–Carpinetum association (Kevey et al. 2017) and Carpino betuli–Quercetum roboris (Lendel et al. 2007). In addition, it is recorded in more thermophilic communities on the border of forest and scrub (Bedalov et al. 2006).

Arum alpinum is partially protected in Poland (ISAP 2014). As already mentioned, the localities of A. alpinum are mainly found in the southern part of the country with a mountainous character. Although it has been noticed in many sites, it is considered as a rare species. Therefore, it has been included into many local red lists, among others with the category EN (Fabiszewski and Kwiatkowski 2002; Parusel and Urbisz 2012) and VU (Nowak et al. 2008). In addition, it has been included in several national red lists of endangered species (Gärdenfors 2005; Petrova and Vladimirov 2009; Grulich 2017).

Study area

The research was conducted in the foothills of the Western Carpathians (Solon et al. 2018) at four study sites (Fig. 1). Sites located in natural forests, i.e. two sites in oak–hornbeam forest and two in riparian forest, were selected for the study. Locality 1 was situated in the Strzyżów Foothills in Krajowice (49.778631° N, 21.439977° E), within the Golesz nature reserve and the Natura 2000 area Golesz PLH180031. The Golesz Nature Reserve protects sandstone outcrops with the surrounding oak–hornbeam forest, with stands of rare and protected plant species (including A. alpinum, Cephalanthera damasonium (Mill.) Druce, Cephalanthera longifolia (L.) Fritsch), as well as the ruins of the medieval Golesz Castle (Towpasz and Stachurska-Swakoń 2010). The research was carried out in the vicinity of the ruins of the castle on a steep slope (15°–20°) with a southern (S) exposure (320 m asl.), where Patch I was designated.

Locality 2 was situated in the Bukowiec Foothills in Markowce (49.533219° N, 22.111701° E) in the Sanoczek River valley. The research was carried out within the floodplain (330 m asl.) on the established Patch II, surrounded by arable fields, fallow land and built-up areas. Natural and semi-natural vegetation (disused meadows, herbaceous plants, willow thickets, and fragmentarily preserved riparian forests) occurs in the form of narrow strips along the riverbed. However, there are several protected species here: Colchicum autumnale L., Gentiana cruciata L., Listera ovata (L.) R. Br., and Ophioglossum vulgatum L.

Locality 3 is situated in the Dynów Foothills in the town of Bachórzec (49.800473° N, 22.342133° E) within an area with a high protection regime (the Kozigarb nature reserve, the Przemyskie Foothills Landscape Park, and the Natura 2000 area of the Przemyskie Foothills PLB180001). The Kozigarb reserve protects a forested peak (321 m asl.), which is characterised by a rich microrelief. The diversified orography of the area and habitat conditions has influenced the development of rich veg:tation. There are several plant communities here (Luzulo luzuloidis–Fagetum, Dentario glandulosae–Fagetum, and Tilio cordatae–Carpinetum betuli) and numerous protected species (including A. alpinum, Daphne mezereum L., Galanthus nivalis L., Lilium martagon L., Neottia nidus-avis (L.) Rich., Primula elatior (L.) Hill, and Scilla bifolia L.). Patch III was established in the San river valley (253 m asl.) on the river bank (5°–15°) with a western (W) exposure.

Locality 4 was situated in the Jasło-Krosno Basin in Żółków (49.704038° N, 21.468534° E). Patch IV was located in the Natura 2000 area of the Wisłoka with its tributaries PLH180052 on the floodplain (226 m asl.). In this area, water habitats and communities of forest and meadow vegetation located in the Wisłoka river valley are protected. The most valuable natural habitats include forests and alluvial thickets (Salicetum triandro-viminalis, Salicetum albo-fragilis, Populetum albae, Carici remotae-Fraxinetum, Alnetum incanae, Caltho-Alnetum, and Ficario-Ulmetum campestris), which in many places are preserved in a natural state or only slightly transformed.

Field measurements

In the year 2020, the abundance of generative and vegetative individuals of A. alpinum in populations growing within particular localities was surveyed. Then, in the central part of each locality one square-shaped, permanent representative study patch with the dimensions 10 m × 10 m was established and marked. These were Patch I in Locality 1 (Golesz), Patch II in Locality 2 (Markowce), Patch III in Locality 3 (Kozigarb), and Patch IV in Locality 4 (Żółków). Further investigations were carried out in 2020–2021 solely within permanent patches. Each year, 30 vegetative and 30 generative individuals were selected and marked with a plastic peg with a label. In the case of a lower number of individuals within a study patch, all individuals were marked. Additionally, the distribution of particular individuals was illustrated in working cartograms. Detailed analyses focused on selected characteristics of leaf rosettes (number of leaves in a rosette, length of the longest petiole, length and width of the leaf blade) and selected characteristics of flowering stems (length from the ground level to the base of the infructescence, length of infructescence and number of fruits per infructescence). Measurements of the characteristics of leaf rosettes were made each year in May, and the characteristics of generative shoots in August. Field research was carried out in accordance with the methodology adopted in our previous publication (Wójcik and Kostrakiewicz-Gierałt 2020).

Within each permanent patch in 2020, floristic lists were made and the percentage coverage of all plant layers was determined. Species covering at least 10% of the surface of a given layer were considered dominant. The nomenclature of vascular plants is given after Mirek et al. (2020), and mosses after Ochyra et al. (2003). The investigations of habitat conditions included measurements of the height of herbaceous vegetation, soil pH, soil moisture, and light intensity at ground level (measured in sunny weather). The measurements were taken on 28–31 May 2020, from 12:00 to 15:00. The sites of measurements were designated in the central part of a metal ring measuring 30 cm in diameter, which was thrown within each study patch. In each study site five tosses were taken from the centre of the patch towards each of its borders (in total 20). The pH of the soil was measured with a soil acid meter using Hellig’s solution. Soil moisture was measured with an Extech MO750 hygrometer, while light intensity was investigated with a TES 1335 luxmeter.

Statistical analysis

The arithmetic mean (x) and standard deviation (SD) of the soil moisture, soil pH, height of standing vegetation, and light intensity at ground level in each study site in the first year of observation were calculated. Moreover, the arithmetic mean (x) and SD of the following individual traits were calculated:

• - number of rosette leaves, length of the petioles of the longest leaves, length of the blades of the longest leaves, width of the blades of the longest leaves in all vegetative and generative individuals occurring within the study plots;

• - length of stem to base of infructescences, length of the infructescences, number of fruits.

The normal distribution of untransformed data considering the aforementioned habitat parameters and individual traits were checked using the Kolmogorov–Smirnov test, while homogeneity of variance was tested using the Levene test at the significance level of P < 0.05. As the values in some groups were not consistent with normal distribution, and the variance was not homogeneous, the analysis was based on non-parametric tests. The Kruskal–Wallis H test was applied to test the statistical significance of differences among study sites in: (i) the studied habitat parameters, (ii) the studied traits of vegetative individuals, as well as (iii) the studied traits of generative individuals. Furthermore, the Mann–Whitney U test was applied to check the temporal differences in particular individual traits between the first and second year of observations. The correlation between the length of the longest leaf petiole and blade dimensions, as well as the height of flowering stem and number of fruits and infructescence length was tested using the Spearman coefficient. The aforementioned tests were performed using STATISTICA 13 software. Moreover, the interactive chi-square test (Preacher 2001) was used to check if the proportion of generative and vegetative individuals of A. alpinum differed between study patches and years. All analyses were performed at the significance level of P < 0.05.

RESULTS

Habitat conditions

In Patch I (Golesz), A. alpinum occurred in the Tilio cordatae–Carpinetum betuli oak–hornbeam forest (Table 1, Fig. 2) developed in a moderately humid habitat, on brown soil with a large proportion of rock rubble. The herbaceous layer was floristically diverse and typical of oak–hornbeam forests. It was dominated by low-growing species, often with small and narrow leaves (e.g. Galeobdolon luteum Huds. emend. Holub, Galium odoratum (L.) Scop., Stellaria holostea L.). In Patch II (Markowce), A. alpinum occurred in the Populetum albae association, which developed on a moist alluvial site. In total, 35 species of vascular plants and 3 species of mosses were recorded in the community. The undergrowth was lush and reached full coverage. The largest share was represented by species with a high habit, often with large and spreading leaves (e.g. Aegopodium podagraria L., Anthriscus nitida (Wahlenb.) Hazsl., Lamium maculatum L., Stachys sylvatica L.). The moss layer was also well developed, with Plagiomnium undulatum (Hedw.) T.J. Kop., Plagiomnium affine (Blandow ex Funck) T.J.Kop., and Oxyrrhynchium hians (Hedw.) Loeske. In Patch III (Kozigarb), A. alpinum occurred in the Tilio cordatae–Carpinetum betuli association in a slightly desiccated habitat on brown soil. It was the most species-rich community, as 52 species of vascular plants and 6 species of mosses were recorded there. The layer of herbaceous plants was very well preserved and floristically diverse. It was dominated by species characteristic of oak–hornbeam forests, with narrow leaved (Carex pilosa Scop., S. holostea) and low-growing species belonging to the order Fagetalia sylvaticae (e.g. Asarum europaeum L., Pulmonaria obscura Dumort.). In Patch IV (Żółków), A. alpinum occurred in the Populetum albae association on a moist alluvial site. In total, 33 species of vascular plants and 1 species of moss were recorded there. The herbaceous vegetation was lush and differentiated into two layers: A. podagraria and Salvia glutinosa L. dominated in the upper layer, while low-growing species (A. europaeum, Ficaria verna Huds., Anemone nemorosa L.) dominated in the lower layer. Mosses represented by O. hians had little coverage.

The statistical analysis showed that the soil moisture, soil pH, and height of standing vegetation were significantly lower in Patches I and III than in Patches II and IV. The light intensity at the ground level was much greater in Patch III than in the other study sites (Table 2).

Abundance of population and number of individuals in study patches

The greatest abundance of the A. alpinum population was recorded in Locality 1 (Golesz), where 513 vegetative and 826 generative individuals were noted. The remaining populations were much smaller. In Locality 3 (Kozigarb), 327 vegetative and 65 generative individuals were recorded, in Locality 2 (Markowce) 72 vegetative and 124 generative individuals occurred, whilst in Locality 4 (Żółków) 125 vegetative and 81 generative individuals were noted.

In both study years, the greatest number of vegetative and generative individuals was recorded in Patch I (Golesz). The lowest number of vegetative individuals was noted in Patch II (Markowce), while the lowest number of generative individuals was observed in Patch III (Kozigarb) (Fig. 3). Similarly, the greatest number of generative individuals with infructescences was recorded in Patch I, while the lowest number was found in Patch III (Fig. 4).

During both study years, the share of generative individuals was significantly greater than vegetative individuals in all the study sites excluding Patch III. However, only in the year 2020 were the differences statistically significant (Table 3).

Traits of vegetative and generative individuals

In the vegetative individuals, the number of rosette leaves as well as the length and width of the longest leaf blades did not differ among the populations. The greatest length of the leaf petioles was noticed in Patch II, whereas the lowest value of this parameter was recorded in Patch III (Table 4). The Mann–Whitney U test showed temporal variability in the number of leaf rosettes in Patch II (U = 135.0, P ≤ 0.05) and in Patch III (U = 280.0, P < 0.01). Furthermore, the temporal variability was displayed by the length of petioles in Patch I (U = 277.0, P < 0.01), Patch II (U = 93.5, P ≤ 0.05), and Patch III (U = 170.0, P ≤ 0.05) and by the length of blades in Patch III (U = 216.0, P < 0.01). The width of blades was similar in both years in all populations.

In the generative individuals, the number of rosette leaves was similar in all patches, while the length of leaf petioles as well as the lowest length and width of the longest leaf blades were noted in Patch III (Table 4). The statistical analysis showed the temporal variability of individual traits solely in the case of the length of the longest leaf petioles in Patch II (U = 182.0, P < 0.001), the length of the longest leaf blades in Patch IV (U = 297.0, P < 0.001), and the width of the longest leaf blades in Patch II (U = 274.0, P < 0.001).

The greatest length of the stem from the ground level to the base of infructescences in 2020 was recorded in Patch II and its lowest value was noted in Patch II. In turn, the value of this parameter in 2021 was significantly higher in Patches I and II than in Patches III and IV. During both study years, the length of the infructescences was greater in Patches I and II than in Patches III and IV, while the lowest number of fruits was recorded in Patch III (Table 5). The statistical analysis confirmed the temporal variability of the length of the stem in Patch I (U = 207.0, P < 0.001) and Patch III (U = 118.0, P < 0.01), the length of the infructescences in Patch II (U = 167.0, P < 0.01) and in Patch III (U = 117.0, P < 0.01), and the number of fruits in Patch I (U = 285.0, P ≤ 0.05), Patch III (U = 127.0, P < 0.001), and Patch IV (U = 266.0, P < 0.01).

Moreover, the vegetative and generative individuals exhibited a positive correlation between the length of the petioles of the longest leaves and the dimensions of the leaf blades. Such a tendency was observed in all populations throughout the study period (Table 6). Furthermore, a positive correlation between the length of the flowering stems and the infructescences was noted in the majority of the populations. Also, a positive correlation between the length of the stem and the number of fruits was observed in all patches (Table 7).

DISCUSSION

Habitat conditions

The studied A. alpinum populations occurred in oak–hornbeam forests from the Tilio cordatae–Carpinetum betuli association and riparian forests from the Populetum albae association. However, syntaxonomically, A. alpinum shows a wide phytocoenotic tolerance (see Study species), which may be significantly influenced by the vertical range of the species. Individuals of A. alpinum were recorded in the transition zone of the foothills and the lower montane zone and in the foothill areas, where numerous plant communities occur. Nevertheless, other investigations (Lawesson et al. 1998; Wójcik and Ziaja 2015) have showed that this taxon is attached to old, well-preserved deciduous forests.

According to other authors (Zarzycki et al. 2002; Chytrý et al. 2018), A. alpinum prefers a moist substrate and moderate light intensity. However, our observations showed that populations of the abovementioned species occur in sites with significant soil moisture and low-light intensity caused by the presence of a large dense layer of shrubs (Patches II and IV), in a site with fresh or slightly dried soil and low-light intensity (Patch I), as well as in a site with fresh or slightly dried soil and substantial light intensity (Patch III). The obtained results partly correspond with the observations carried out in a beech forest (Wójcik and Kostrakiewicz-Gierałt 2020) showing the occurrence of A. alpinum populations in shaded sites, one of which was located in the vicinity of a small watercourse, while the other was located on a steep slope covered by soil with low humidity. At the same time, it is worth emphasizing that the observations of other authors indicate the occurrence of A. alpinum populations mainly in damp and shaded places. Barć et al. (2004) found the presence of a population of the aforementioned species in a beech forest, on a steep slope (40°), NE exposure, in the spring area of a mountain stream in the Beskid Mały (Western Carpathians). Moreover, in Southern Europe, populations of A. alpinum have been observed in shady ravines with significant humidity (Nicolin and Imbrea 2009), as well as in places with varying inclinations (0°–40°) and northern or similar exposures (NW, NE, and E), in relatively close proximity (30–100 m) to watercourses (Kozuharova et al. 2014a).

The conducted research confirmed previous observations indicating that A. alpinum prefers neutral or slightly alkaline soils with a high content of organic matter (Zarzycki et al. 2002; Chytrý et al. 2018). In the observed patches of oak–hornbeam forest (Patches I and III), there was soil with a neutral pH (6.58 and 6.53), while in riparian forests (Patches II and IV) soils with slightly alkaline pH values (7.05 and 7.28) were found. Also, according to Lawesson (2003), the optimal soil pH for A. alpinum is 6.5, and the value of the soil reaction indicator (R) reaches 7.0. On the other hand, in the Carpathians, A. alpinum has also been recorded in places with an acidic substrate (Barć et al. 2004; Wójcik and Kostrakiewicz-Gierałt 2020). Furthermore, the performed observations showing the occurrence of populations of A. alpinum in alluvial soils and brown soils supports the findings of Zarzycki et al. (2002) and Chytrý et al. (2018) arguing that individuals of the aforementioned taxon prefer eutrophic soils. Moreover, our findings are consistent with previous observations (Barć et al. 2004; Wójcik and Kostrakiewicz-Gierałt 2020). At the same time, it should be added that in the southern part of the species’ range, the occurrence of A. alpinum populations on brown forest soils with a loose structure was also recorded (Kozuharova et al. 2014a). In addition, populations have been observed on deep, humus soils, rich in rough material (Bedalov et al. 2006; Lendel et al. 2007), on typical rendzinas, rich in humus with a high content of calcium carbonate (Nicolin and Imbrea 2009), as well as on sandy and stony soils (Kozuharova et al. 2014a). In sum, the results were consistent with the hypothesis that A. alpinum occurs in a variety of habitats.

Abundance at the sites and number of individuals in study patches

Previous studies have shown differences in the size of the A. alpinum population. In the southern part of the range, most populations of A. alpinum are rather small and occur in isolated sites which are distant from each other (Fridlender 1999; Kozuharova et al. 2014b; Trigas et al. 2021), while larger populations are rather rare (Fridlender 1999). In Poland, the smallest populations were recorded in the Tilio cordatae–Carpinetum betuli forests (Wróbel 2007), while larger populations of 20–184 individuals were observed in the patches of the Carpathian beech forest Dentario glandulosae–Fagetum (Barć et al. 2004; Wójcik and Kostrakiewicz-Gierałt 2020). The populations observed in these studies were larger and significantly diverse in size. Populations developed in the riparian forests of Populetum albae, with 196 individuals (Locality 2—Markowce) and 206 individuals (Locality 4—Żółków), were significantly smaller than the populations occurring in the forests of Tilio cordatae–Carpinetum betuli, with 392 and 1339 individuals within Locality 3—Kozigarb and Locality 1—Golesz, respectively. Considering this, the first hypothesis about the influence of habitat conditions on the abundance of A. alpinum individuals seems to be supported. The abundance of the abovementioned populations is influenced by the efficiency of vegetative and generative reproduction. Fridlender (1999) emphasizes that in populations inhabiting unstable habitats (steep, crumbling slopes), dismemberment of A. alpinum tubers and vegetative reproduction are frequent. Additionally, Barć et al. (2004) pointed out that the unstable nature of the sliding subsoil favours fragmentation of tubers and colonization of the lower parts of the slope. Considering this, we might speculate that clonal propagation also might have a significant impact on the considerable population abundance in Locality 1, located on a steep slope. On the other hand, generative reproduction, impacted by various factors, might also shape the population abundance. According to Fridlender (1999), in large populations of A. alpinum cross-pollination alone might occur due to self-sterility of individuals, in contrast to small populations (counting less than 10 plants) where individuals are autofertile. The undisturbed seed production and dispersal followed by successful seedling recruitment might contribute to an increase in the population abundance and prolongation of its persistence in the occupied site.

The diminishing of abundance of populations might be caused by mortality of individuals. Our previous observations suggest that excessive substrate moisture is unfavourable, as it leads to rotting and death of A. alpinum individuals (Wójcik and Kostrakiewicz-Gierałt 2020). Considering this, we might speculate that a similar phenomenon might result in the small population sizes in Localities 2 and 4, situated in river valleys prone to frequent flooding periods. Furthermore, the diminishing of population abundance might be caused by damage to tubers as an effect of the activity of large herbivores such as wild boars (Sus scrofa) observed in neighbouring populations (personal observation).

The performed investigations showed the dominance of generative individuals in Locality 1 (Golesz) and Locality 2 (Markowce), as well as in Patches I and II, similar to the aforementioned observed populations in the Carpathian beech forest (Barć et al. 2004; Wójcik and Ziaja 2015; Wójcik and Kostrakiewicz-Gierałt 2020). On the other hand, the predominance of vegetative individuals observed in Locality 3 (Kozigarb) and Patch III could be caused by high insolation and soil drying, limiting flower production and contributing to the rapid drying of entire generative individuals.

Traits of vegetative and generative individuals

To date, studies of individual characteristics of A. alpinum in the field have been conducted on limestone cliffs in Corsica (Fridlender 1999), in a Carpino betuli–Quercetum roboris oak–hornbeam forest in eastern Croatia (Lendel et al. 2007), and in a Carpathian beech forest Dentario glandulosae–Fagetum in southern Poland (Wójcik and Kostrakiewicz-Gierałt 2020). Moreover, investigations of herbarium specimens from the Carpathian region were conducted by Terpό (1973).

Our results showing the mean length of petioles ranging from 18.9 to 32.4 cm in generative individuals and from 14.6 to 28.3 cm in vegetative individuals confirm the observations reported by Fridlender (1999), Wójcik and Kostrakiewicz-Gierałt (2020), and Terpό (1973); however, they are not consistent with the observations described by Lendel et al. (2007), who recorded much higher values in the case of generative individuals. The mean leaf blade length ranging from 9.04 to 11.08 cm in vegetative individuals and from 10.78 to 14.76 cm in generative individuals and the mean leaf blade width ranging from 6.03 to 7.35 cm in vegetative individuals and from 8.38 to 10.56 cm in generative individuals observed in the present investigations correspond with findings reported by Fridlender (1999). However, our earlier research conducted in a Dentario glandulosae–Fagetum beech forest showed slightly higher values (Wójcik and Kostrakiewicz-Gierałt 2020). The average length of the stem from the ground level to the base of the inflorescence ranging from 17.0 to 32.7 cm is consistent with the values presented by Terpό (1973) and Wójcik and Kostrakiewicz-Gierałt (2020), while higher values were observed by Fridlender (1999). Moreover, the observed average infructescence length ranging from 3.5 to 5.8 cm was similar to that previously observed by Wójcik and Kostrakiewicz-Gierałt (2020), while Lendel et al. (2007) reported much higher values. In turn, the average number of fruits ranging from 33.2 to 61.9 cm was higher than that observed in the Dentario glandulosae–Fagetum beech forest (Wójcik and Kostrakiewicz-Gierałt 2020) and the number found in individuals growing on limestone cliffs (Fridlender 1999).

The present investigations partially confirm the hypothesis about the spatial variability of individual traits of A. alpinum, showing the substantially greater length of leaf petioles in Patch II than in the other study sites. This may have been caused by the considerable shading. Interestingly, we observed no differences in the length and width of the leaf blades of the vegetative individuals among the populations and substantial differences in the case of the generative individuals. The small size of the leaf blade in the individuals from Patch III may have been associated with substantial light intensity and the low height of neighbouring plants. It is worth mentioning that the close relative congener A. maculatum is characterized by strong photosynthetic acclimation to low-light conditions reflected in a substantial leaf area index (Popović et al. 2006) and specific area index (Popović et al. 2016). Moreover, the lowest length of stems, infructescences, and number of A. alpinum fruits recorded in Patch III may have been caused by the adverse influence of the considerable light intensity and the occurrence of dry soil.

Our investigations partially confirm the hypothesis of the temporal variability of individual traits. The temporal variability of some of the studied traits, i.e. the length of petioles particularly in the vegetative individuals, is consistent with previous findings (Wójcik and Kostrakiewicz-Gierałt 2020), while the inter-annual differences in other traits (e.g. length of infructescences, number of fruits in infructescences) were not noted previously. Moreover, the temporal variability of the individual traits was found to increase from Patch IV, through Patches I and II, to Patch III. The occurrence of these discrepancies might suggest the need for further long-term observations.

The positive correlation between the length of the leaf petiole and the dimensions of leaf blades in the vegetative and generative individuals and between the length of the stem, the length of infructescences, and the number of fruits confirms our fourth hypothesis and is consistent with previous investigations conducted in A. alpinum populations (Wójcik and Kostrakiewicz-Gierałt 2020) and other related congeners (Joudi et al. 2016).

CONCLUSIONS

The studied populations were diverse in terms of size and analysed characteristics. The largest populations were found in the Tilio cordatae–Carpinetum betuli forests (Patches I and III) on steep slopes, moderately or well lit, in fresh or slightly dry habitats with a neutral pH, where narrow-leaved herbaceous vegetation prevailed. Less numerous were the populations located in the riparian forests of Populetum albae (Patches II and IV) on floodplains, in moist, heavily shaded habitats with a slightly alkaline substrate, where lush herbaceous vegetation with wide and spreading leaves prevailed.

The riparian populations were characterised by a much higher share of generative individuals showing large-dimension infructescences and considerable fruit production. Populations growing in oak–hornbeam forests were more diverse. The individuals found in Patch I were similar in size to riparian populations and were also characterised by significant production of large numbers of infructescences and fruits. On the other hand, individuals from Patch III compared with individuals from other populations were much smaller and produced the least amount of fruit, which was related to significant exposure to sunlight and drying of the habitat.

Our research has shown that A. alpinum can grow in a variety of plant communities, but is not strictly affiliated to a specific plant association. Appropriate habitat requirements are crucial for this species. This species prefers moist habitats rich in nutrients. However, the type of soil and soil pH are less important for it. In particular, populations of A. alpinum develop best in highly shaded positions, as too much sunlight leads to soil drying and premature death of the aboveground parts of plants.

At the same time, it should be added that continued studies of the influence of habitat conditions on (i) the genetic structure of populations, reflecting the effectiveness of vegetative and generative reproduction, and on (ii) the opportunities and constraints of generative reproduction might provide important directions for future research.

Acknowledgements

The authors would like to thank Dr Grzegorz Wolski for the determination of the mosses and Dr Paweł Krąż for assistance in making Fig. 1.

Funding

The field studies were financially supported by Department of Nature Conservation and Landscape Ecology, University of Rzeszów.

Conflict of interest statement. The authors declare that they have no conflict of interest.

REFERENCES