The subjects of this study were pteridophyte species found along observation trails in the Gandul Mount area, Wonogiri District, Wonogiri Regency, Central Java, Indonesia. Field sampling was conducted from July to October 2025 at altitudes between 195 and 355 m above sea level. The total length of the survey trail was approximately 1.21 km (Figure 1), covering representative habitats within the study area, including terrestrial, epiphytic, and lithophytic environments.

The materials and instruments used in this study included writing instruments (notebook and pencil), a digital Global Positioning System (GPS) device to record sampling locations, a digital camera for specimen documentation, newspaper sheets for temporary specimen preservation, label paper for sample coding, and a standard fern identification guidebook. Additional taxonomic references were obtained from online databases, including Plants of the World Online (POWO), Plantamor, and the Global Biodiversity Information Facility (GBIF).

This study employed a descriptive exploratory design using the cruise method for field inventory survey. The cruise method in field survey was chosen to allow comprehensive exploration of heterogeneous habitats across the karst landscape of Gandul Mount and to effectively document fern diversity. Field surveys are widely applied in botanical surveys due to their effectiveness in heterogeneous landscapes. This approach involves the systematic exploration of representative habitat types and allows detailed observation of morphological characteristics and habitat preferences (U’un et al., 2021). Gandul Mount, Wonogiri, is dominated by karst limestone formations, resulting in shallow soils, low rainfall, and uneven vegetation distribution. These conditions limit the effectiveness of plot-based sampling techniques and make field surveys particularly suitable for documenting pteridophyte diversity, including species occupying specific microhabitats such as epiphytic and lithophytic niches (U’un et al., 2021).

The research procedure consisted of four main stages such as preparation, sampling, identification, and inventory. In the preparation stage, the observation route was determined based on topography, accessibility, and habitat representation, followed by preliminary field observations and a literature review. During the sampling stage, pteridophyte specimens found along the observation route were selectively collected, ensuring the presence of key identification characteristics such as spores and other parts such as leaves, stems, and roots, that are complete and undamaged. The specimens were then placed on newspaper sheets and labeled with the sampling code and location data for further examination.

The identification and inventory stages were conducted based on morphological characteristics, including stem structure, leaf shape and arrangement, root type, and sori position and shape. Species identification followed the reference journals "Identification of Ferns Based on Sorus Position and Altitude Differences in Bondowoso Regency" (Pradipta et al., 2023), "Field Guide for Ferns (Pteridophyta) at Ragunan Zoo" (Agatha et al., 2019), and verified using online taxonomic databases (POWO, Plantamor, and GBIF). Habitat classification is done by recording the type of substrate for each species, namely terrestrial (soil), epiphytic (tree surface), and lithophytic (rock surface).

Data analysis was conducted using a qualitative descriptive approach. Each collected specimen was examined to document morphological characteristics, including leaf structure, spore characteristics, stem and root system characteristics, and habitat substrate type. Species composition and distribution patterns along the observation route were summarized to provide an overview of pteridophyte diversity in the study area. The resulting inventory data served as the basis for subsequent in silico phylogenetic analyses using molecular sequence data obtained from public genetic databases.

Species identified from the field inventory were then analyzed using molecular data obtained from the National Center for Biotechnology Information (NCBI) database. The DNA sequence of the chloroplast matK gene was chosen as a molecular marker due to its high discriminatory power and widespread use in plant DNA sequencing and phylogenetic studies. For each identified species, available matK sequences were retrieved in FASTA format from the NCBI GenBank database.

Sequence alignment was performed using the alignment function implemented in MEGA (Molecular Evolutionary Genetics Analysis) software to ensure homologous nucleotide positions among all taxa. The aligned sequences were then evaluated to determine the most appropriate nucleotide substitution model using MEGA's built-in model selection tool based on statistical criteria such as the lowest Bayesian Information Criterion (BIC) or Akaike Information Criterion (AIC) values.

Phylogenetic tree reconstruction was performed using the Maximum Likelihood (ML) method under the best-fitting evolutionary model obtained from model testing. Bootstrap analysis with multiple replications was applied to assess the reliability and robustness of the inferred clades. The resulting phylogenetic tree was visualized and interpreted to elucidate the evolutionary relationships among the recorded pteridophyte species. Haplotype data were obtained from DnaSP v.6.

Field inventory conducted in the Gandul Mount area recorded a total of 17 pteridophyte species belonging to eight families (Table 1). Among these families, the Pteridaceae family showed the highest species richness with seven species, followed by the Polypodiaceae with three species, the Dryopteridaceae with two species, and the other families each represented by one species.

Habitat analysis showed that lithophytic species dominated the study area, with 13 species growing on damp moss-covered rocks, rock crevices, and vertical rock surfaces near drainage channels. The other four species were terrestrial, growing on flat ground surfaces in open areas. This distribution pattern is closely related to the karst-dominated geology of Gandul Mount, where limestone outcrops and rocky substrates predominate. Such environmental conditions favor lithophytic growth forms and shape the structure of fern communities. These results demonstrate that habitat-based inventories complement morphological analyses in providing a comprehensive understanding of species diversity and ecological adaptations in heterogeneous landscapes.

The dominance of Pteridaceae indicates high ecological adaptability to diverse environmental conditions, including habitats with low moisture availability and rocky substrates. Members of this family have small, segmented leaves that reduce evaporation and enable survival in both shaded and open environments. This finding is consistent with previous research by Khairunisa et al. (2023), which reported that Pteridaceae species are widespread due to their ability to grow in moist to dry habitats under varying light conditions. Globally, this family is recognized as one of the largest within the Pteridophyta, comprising approximately 950 species, further explaining its frequent occurrence in diverse ecosystems.

Based on conservation assessments using the IUCN Red List, two species, Adiantum capillus-veneris and Pteris vittata, are categorized as Least Concern, indicating stable populations and a low risk of extinction. The remaining 15 species have not yet been evaluated (status Unevaluated). Although most species lack formal conservation assessments, their presence highlights the ecological importance of Gandul Mount as a habitat for fern diversity.

Morphological analysis of the 17 recorded species revealed substantial variation in leaf, stem, root, and reproductive structures (Table 2), reflecting adaptation to heterogeneous environmental conditions.

Leaf morphology exhibits the highest degree of variation in surface texture, arrangement, margin type, shape, and complexity. Most species exhibit smooth, hairless anterior surfaces; however, Cheilanthes tenuifolia, Tectaria aurita, and Deparia japonica have rough, hairy surfaces, which likely enhance water retention and protection against environmental stress. Alternating phyllotaxy predominates among species, while Selaginella aristata uniquely displays a tetraphyllous leaf arrangement. The anterior margin varies from serrated to lobed and entire. Pinnate compound leaves are common, as observed in Pteris vittata, while bipinnate and tripinnate forms are found in Tectaria aurita and Dryopteris nodosa. In contrast, Selaginella aristata has microphyllous leaves arranged in four rows. This observation supports the findings of Rizky et al. (2018), who described Selaginella species as having small, scale-like microphyllous leaves without petioles.

Stem morphology also varies significantly between species. Eight species have creeping rhizomes, including Dryopteris nodosa; four species have erect rhizomes, such as Pteris vittata; three species exhibit erect stems, such as Pityrogramma calomelanos; and two species have creeping stems (Selaginella aristata and Pteris biaurita). Interestingly, Drynaria quercifolia exhibits distinctive rhizome characteristics: thick, fleshy, creeping, and densely covered with brown scales. These characteristics align with the description provided by (Lindasari et al., 2015), who reported that this species has thick, rounded rhizomes covered with fine brown scales.Root morphology is relatively uniform, with 16 species having fibrous roots emerging from the rhizome. Only Selaginella aristata exhibits adventitious roots emerging from each stem node, a characteristic also reported by (Sartika et al., 2021). Root color is consistently brown across all species.

Reproductive structures further differentiate the species. Sori vary in shape, position, presence of an indus, and coloration. Rounded abaxial sori are observed in Dryopteris nodosa, while linear marginal sori occur in Pteris ensiformis. Most species exhibit white sori when immature, changing to yellowish-brown as they mature, while Lygodium flexuosum uniquely retains green sori. Indus types range from open (Pteris vittata), closed (Dryopteris nodosa), to absent (Elaphoglossum sp.). Selaginella aristata does not form sori but produces terminal strobili with a tetragenous arrangement, in which sporangia are arranged in four rows. This reproductive strategy aligns with the findings of (Elsifa et al., 2019), who reported that Selaginella species form strobili rather than sori.

Ethnobotanical assessment revealed that 11 of the 17 recorded species have potential uses as medicinal plants, ornamental plants, decorative materials, craft resources, and phytoremediation agents (Table 3). Medicinal use was the most frequently documented application, indicating that Gandul Mount harbors biologically valuable plant resources for traditional healthcare practices.

Several Pteris species have demonstrated phytoremediation potential, particularly for the uptake of heavy metals such as mercury (Hg). These findings suggest a relationship between the fern community and the mineral-rich sedimentary rocks of Gandul Mount, which contribute to its distinctive soil chemistry. Pteridophytes with high tolerance to heavy metals appear to be well-adapted to such environments. Overall, the research results indicate that pteridophytes on Gandul Mount have significant ecological, taxonomic, and ethnobotanical value. However, sustainable use and habitat conservation are crucial to ensure the long-term survival of these species in their natural ecosystems.

A total of 14 matK sequences were included in the analysis after the alignment process. The selected region encompassed 227 nucleotide positions, with no sites excluded due to gaps or missing data. Of these, 189 sites varied, indicating a high level of nucleotide polymorphism among the sampled taxa. This proportion of varying sites reflects the considerable genetic divergence among the analyzed pteridophyte species. Haplotype analysis revealed the presence of 14 distinct haplotypes (h = 14), which corresponds exactly to the number of sequences analyzed, with each species represented by a unique haplotype. Consequently, haplotype diversity was maximal (Hd = 1.0000), indicating complete haplotype differentiation among the taxa. Each haplotype was associated with a single species accession from GenBank. This pattern confirms that no shared haplotypes were detected among the sampled species, supporting a clear genetic separation between the samples.

A Maximum Likelihood phylogenetic tree, generated using General Time Reversible has invariant sites (GTR+I) based on the chloroplast marker matK (1,000 bootstrap replicates) revealed several major clades and was largely consistent with the modern fern classification system proposed by Pteridophyte Phylogeny Group I (PPG I, 2016). Species belonging to the Dryopteridaceae (Dryopteris nodosa and Elaphoglossum callifolium) were more closely related to the Hypodematiaceae (Leucostegia immersa), followed by the Polypodiaceae (Microsorum commutatum and Tectaria aurita), all clustering into a strongly supported eupolypod clade (70–100 bootstraps), reflecting their common evolutionary ancestry despite morphological differences. Similarly, members of the Pteridaceae (Hemionitis tenuifolia, Pityrogramma calomelanos, Pteris vittata, P. biaurita, and P. ensiformis) form a distinct monophyletic group with high bootstrap support (99–100), corroborating their taxonomic placement and shared morphological features such as marginal or linear sori and adaptation to relatively dry habitats. Microlepia speluncae (Dennstaedtiaceae) and Deparia japonica (Aspleniaceae) occupy an intermediate position consistent with their early divergence among leptosporangiate ferns, whereas Lygodium flexuosum (Schizaeaceae) is placed in a basal lineage, in keeping with its status as an early-diverging fern family in the PPG I system.

However, there are some deviations from the expected family-level grouping, particularly in the placement of Adiantum capillus-veneris outside the main Pteridaceae clade. These differences are likely due to methodological constraints rather than genuine evolutionary inconsistencies, including reliance on a single molecular marker (matK), limited availability of sequence data in GenBank (often represented by only one accession per species), and the use of partial coding sequences for Microlepia speluncae, Microsorum commutatum, and Adiantum capillus-veneris. Furthermore, this study also used concatenated rps16–matK intergenic space sequences for Tectaria aurita, while other taxa are represented by only matK, resulting in alignment heterogeneity that affects branch topology. In line with the statement by Gu et al. (2024) that cpDNA analyses of Pteridaceae encountered alignment difficulties in non-coding regions (such as intergenic spacers), similar to the rps16–matK fusion in Tectaria, leading to genomic size variation and potential phylogenetic bias. Overall, the phylogenetic reconstructions show similar results to the PPG I (2016) higher-level taxonomic framework, while highlighting the need for multi-locus analyses and locally generated sequences to improve the resolution and stability of species-level relationships among Pteridophyta from Gandul Mount. In line with the findings of Olsson et al. (2021), PPG I (2016) is generally supported in DNA barcoding analyses of Pteridophyta, however, single markers such as matK/psbA-trnH are insufficient for identifying closely related species, so multi-locus analysis is recommended for greater reliability.