Morphology, Phytochemistry, and Genetics-Based Analysis of Endemic Species Belonging to Allium sect. Schoenoprasum (Amaryllidaceae) from the Kazakhstan Altai

1. Introduction

The genus Allium L. (Amaryllidaceae) is among the most species-rich and ecologically diverse groups of monocots, comprising more than 1000 taxa distributed across temperate and subtropical regions of the Northern Hemisphere [1]. This genus includes globally important crops such as onion (A. cepa), garlic (A. sativum), and leek (A. porrum), valued not only as vegetables and spices but also as sources of bioactive compounds with wide-ranging pharmacological properties [2,3]. Wild relatives of cultivated Allium species also play key ecological roles and represent an important genetic reservoir for crop improvement and adaptation to environmental stress [4,5].

Among Central Asian ecoregions, the Kazakhstan Altai stands out as a biodiversity hotspot, harboring a high concentration of endemic and relict plant species [6,7]. This mountainous system forms part of the Altai–Sayan floristic region, characterized by high levels of endemism and a mosaic of habitats shaped by complex topography and climatic gradients. Within this system, the genus Allium demonstrates exceptional adaptive radiation, with over 100 species documented across Kazakhstan, several of which are endemic or narrowly distributed [8,9,10]. Recent studies in the neighboring Kyrgyz Alatau documented 36 Allium species representing 13 subgenera, including 13 narrow endemics, some of which are listed in national Red Data Books and the IUCN Red List [11]. These findings highlight the exceptional concentration of diversity and rarity in Central Asian mountain systems, reinforcing the likelihood that the Kazakhstan Altai harbors similarly high levels of underexplored endemic diversity [12,13]. However, despite this richness, population-level, ecological, and molecular data for Altai endemics remain scarce, limiting conservation and taxonomic resolution.

Within the genus, section Schoenoprasum Dumort. Is of particular interest due to its ecological specificity and economic potential. This section includes the well-known chive (Allium schoenoprasum L.), a perennial herb widely cultivated for culinary use and noted for its antioxidant and anti-inflammatory activities [14,15]. Recent phytochemical studies have highlighted the richness of phenolic compounds in A. schoenoprasum, linking them to strong antioxidant capacity and potential applications in nutraceuticals [16]. However, its wild relatives in Central Asia remain poorly characterized in terms of morphology, ecology, and genetics, despite their likely role as reservoirs of adaptive and phytochemical diversity.

The Kazakhstan Altai harbors several endemic species of section Schoenoprasum, yet their taxonomic status, morphological differentiation, and ecological roles remain inadequately explored. While preliminary floristic surveys and checklists confirm their presence, comprehensive population-level and integrative studies are scarce [10,17]. Recent work has provided valuable insights into seed morphology of section Schoenoprasum endemics in Eastern Kazakhstan [18], but no study to date has combined ecological, morphometric, phytochemical, and molecular datasets to evaluate species limits and adaptation within this group.

Research on Kazakhstan’s Allium flora has expanded considerably in recent decades, particularly in taxonomy, phylogenetics, and population genetics. Several new and cryptic Allium taxa have been described from Central Asia [19,20,21], underscoring the genus’s underestimated diversity and taxonomic complexity. However, most molecular studies have focused on economically important groups such as A. cepa or A. altaicum, leaving section Schoenoprasum underrepresented in comparative or phylogenomic analyses [22,23,24,25,26].

Phytochemically, Allium species are rich in sulfur-containing compounds, flavonoids, and phenolics, which underlie their antimicrobial, antioxidant, and anticancer properties [27,28,29]. Studies on Kazakhstan endemics have revealed unique phenolic profiles with cosmeceutical and pharmacological potential [30], but the extent of intraspecific variation and environmental modulation of these compounds remains unknown.

Ecologically, Allium species often serve as indicators of alpine and subalpine habitat conditions [31]. Population studies from neighboring regions highlight the vulnerability of endemic Allium taxa to habitat degradation and overharvesting [32,33]. In the Kazakhstan Altai, however, integrated data connecting demography, ecology, and molecular diversity are still lacking, constraining effective conservation planning.

Despite growing research, major knowledge gaps persist for endemic section Schoenoprasum species of the Kazakhstan Altai. Specifically, their population dynamics, morphological plasticity, bioactive potential, and genetic differentiation remain unquantified. Addressing these aspects is essential for resolving species boundaries, assessing adaptive evolution, and informing conservation priorities.

Several endemic Allium species of the Kazakhstan Altai are listed in the Red Data Book of Kazakhstan, underscoring their vulnerability and the urgent need for conservation-oriented research [11]. Beyond their ecological significance, wild relatives of section Schoenoprasum represent valuable genetic resources for breeding programs, particularly for traits such as drought and cold tolerance that are increasingly relevant under climate change [5,34]. The Altai region itself holds exceptional biogeographic importance, functioning as a floristic bridge between Central Asia, Siberia, and Mongolia, and serving as a center of speciation for many taxa [7]. Despite this, section Schoenoprasum remains globally underexplored compared with crop-related sections such as Cepa [35,36] and Porrum [37,38]. Recent advances in phylogenomics and DNA barcoding [19,20,21] now provide powerful tools for resolving complex Allium lineages, highlighting the timeliness and novelty of our integrative approach.

Therefore, this study integrates population ecology, morphology, phytochemistry, and genetics to establish a comprehensive baseline for endemic section Schoenoprasum species in the Kazakhstan Altai and to elucidate the ecological and genetic mechanisms driving their differentiation.

2. Materials and Methods

2.1. Field Expeditions

The objects of the study were 6 populations of 4 species of the genus Allium section Schoenoprasum collected in the territory of the Kazakhstan Altai from June to August in 2023 and 2024 (Figure 1).

A. ivasczenkoae and A. ubinicum are narrow endemics of the Kazakhstan Altai, A. ledebourianum is a regional endemic of the Altai, and A. schoenoprasum is a widespread species in the region.

Field surveys in the Kazakhstan Altai were conducted using route–reconnaissance methods appropriate for heterogeneous terrain and large spatial scales [39]. Populations were located in representative habitats according to species-specific ecological preferences [40]. Site conditions were evaluated following Ellenberg’s ecological indicator scales (EIV) [41], with assessments of light (L), temperature (T), moisture (M), soil reaction (R), nutrient availability (N), and salinity (S). L was measured using a luxmeter UNI-T UT383S (UNI-T, China) under standardized daytime conditions (between 10:00 and 14:00 h) on cloud-free days to minimize temporal variation. Soil samples were collected from the root zone of representative plants and analyzed for R, S, and N in the laboratory. These clarifications ensure that environmental data reflect consistent site conditions. M and T were recorded with a thermohygrometer RGK-TH 30 (RGK, Russia). R (pH) was determined using a pH meter Yieryi HH-1000 (Yieryi, China). N was estimated based on soil fertility characteristics, including organic matter content and visible vegetation cover, while S was evaluated according to soil type and the presence of halophytic plant species.

Individuals were assigned to ontogenetic stages following the criteria of Jones C.G. [42], based on the structure of the shoot system, bulb morphology, and presence of reproductive organs.

2.2. Morphometric Analysis

In the field, morphological and morphometric traits of 30 adult plants per population were studied using eight parameters: plant height (cm), number of stems per clump (count), number of leaves per clump (count), stem length (cm), stem diameter at base (mm), flower stalk length (mm), leaf length (cm), and leaf width (mm). To describe floral morphology, a random sample of 30 flowers was collected from each population [43].

Inflorescences were collected in June 2024. Flowers were preserved in a mixture of ethanol, glycerol, and distilled water (3:2:1) at 16 °C. Only fully developed and healthy specimens were used for analysis. Morphological descriptions focused on diagnostic characters common to flowers of each species within section Schoenoprasum. Floral morphometric measurements were taken using a stereomicroscope (Magus A18T, Russia).

The main morphometric parameters of the flower were studied: petal length (mm), petal width (mm), stamen length (mm), pistil column length (mm), ovary length (mm), ovary width (mm), anther length (mm), and anther width (mm) (Figure 2).

2.3. Analyses of Cytotoxicity and Antioxidant Activity

Cytotoxic activity of Allium extracts was assessed using the brine shrimp (Artemia salina) lethality assay [44]. Extracts were prepared using Soxhlet extraction with ethanol and concentrated under reduced pressure at ≤ 45 °C. Stock solutions were prepared at 1 mg/mL in ethanol (1000 μg/mL). Working concentrations were obtained by dilution in artificial seawater to final concentrations of 1, 5, 10, 25, and 50 μg/mL.

Each concentration was tested in triplicate, with 20 nauplii per replicate, and the total volume per test tube was 2 mL. The solvent control consisted of artificial seawater containing 0.5% (v/v) ethanol, matching the highest solvent fraction used in test wells. A separate distilled-water control was also included for reference.

After 24 h of exposure, the numbers of live and dead nauplii were recorded. Percent mortality was calculated using the formula:

$$Mortality(\%)=\left({\displaystyle \frac{numberofdeadnauplii}{initialnumberoflivenauplii}}\right)\times 100$$

LC₅₀ values (concentration causing 50% mortality) and their 95% confidence intervals were calculated using probit regression in GraphPad Prism (v. 9.0). χ², p and slope parameters were recorded.

Antioxidant activity of Allium samples was determined using the standard 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging assay [45,46]. A 0.1 mM ethanolic solution of DPPH was freshly prepared and kept in the dark to prevent photodegradation. Test extracts were prepared as 1000 μg/mL stock solutions in ethanol, and aliquots of 10, 20, 30, 40, and 50 μL were added to 2 mL of the DPPH solution. Antioxidant activity of Allium samples was determined using the standard 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging assay [46]. A 0.1 mM ethanolic solution of DPPH was freshly prepared and kept in the dark to prevent photodegradation. Test extracts were prepared as 1000 μg/mL /mL stock solutions in ethanol, and aliquots corresponding to final concentrations of 5, 10, 15, 20, and 25 μg/mL were added to 2 mL of the DPPH solution. The mixtures were thoroughly vortexed and incubated at room temperature in the dark. The decrease in absorbance was monitored at 517 nm using a UV-5100 spectrophotometer (Metash, China) equipped with a 3 mL quartz cuvette and a 1 cm path length. Measurements were taken at 5 min intervals up to 30 min to capture the kinetics of radical scavenging. Antioxidant activity was expressed as the percentage of DPPH inhibition relative to the control (ethanol + DPPH without extract) using the formula:

$$Inhibition(\%)=\left({\displaystyle \frac{A_{control}-A_{sample}}{A_{control}}}\right)\times 100$$

IC₅₀ values (concentration required to inhibit 50% of the DPPH radicals) were calculated to compare the antioxidant efficiency of the samples.

2.4. Genetic Analysis

Fresh leaves were collected from each plant, immediately dried to prevent degradation, and stored at –80 °C until further use. Samples were maintained under these conditions to preserve DNA integrity and prevent enzymatic or microbial activity prior to extraction.

Genomic DNA was extracted from leaf tissue using a modified CTAB protocol [47], followed by chloroform–isoamyl alcohol (24:1) purification and precipitation with cold isopropanol. DNA integrity was verified by electrophoresis on 1% agarose gels, and concentration and purity were determined with a NanoDrop 2000 spectrophotometer (Thermo Scientific, Singapore).

Three chloroplast markers were selected for genetic analysis: MatK and RbcL (coding genes commonly used for plant barcoding) and the PsbA–TrnH intergenic spacer (highly variable). PCR amplification was performed on a T100 thermocycler (Bio-Rad, USA) in 25 µL reactions containing 1× buffer, 2.0 mM MgCl₂, 0.2 mM each dNTP, 0.4 µM of each primer, 1 U Taq DNA polymerase (Thermo Scientific, USA), and 20–50 ng of template DNA. Thermal cycling conditions were as follows: MatK – 94 °C for 3 min; 35 cycles of 94 °C for 30 s, 50 °C for 30 s, 72 °C for 1 min; final extension 72 °C for 5 min. PsbA–TrnH – same profile with annealing at 54 °C. RbcL – same profile with annealing at 52 °C. PCR products were checked on 1.5% agarose gels with ethidium bromide. Amplicons were purified with the QIAquick PCR Purification Kit (Qiagen, Germany) and sequenced bidirectionally on an ABI 3730xl DNA Analyzer (Applied Biosystems, USA). Sequence editing and assembly were performed in Geneious Prime v.2023 (www.geneious.com/features/prime). Alignments were generated with MUSCLE and manually refined. For each marker, nucleotide composition (A, T, G, C), alignment length, and mean values with standard deviations were calculated in MEGA X (www.megasoftware.net).

Phylogenetic trees were constructed using the Neighbor-Joining (NJ) method with the Kimura 2-parameter (K2P) evolutionary model. Node reliability was evaluated through bootstrap analysis with 1000 pseudoreplicates. Analyses were conducted independently for each marker and for the combined dataset. Dendrogram visualization and design were performed using FigTree v.1.4.4 (www.tree.bio.ed.ac.uk/software/figtree).

2.5. Statistics

All data analyses were performed using standard statistical and bioinformatic software. Morphometric parameters were summarized as means ± standard deviations and coefficient of variation (CV). Principal component analysis (PCA), correlation analysis, and hierarchical clustering (sectional dendrograms) were conducted in RStudio v.1.3.1093 (www.posit.co/download/rstudio-desktop) using the packages factoextra, corrplot, dendextend, cluster, and ggplot. Diagrams of the general floral structure and morphometric parameters were prepared in AutoCAD LT 2024 (Autodesk Inc., USA), which provides precise control over geometric shapes and proportions, ensuring accuracy in the illustration of diagnostic morphological features.

3. Results

3.1. Populational and Ecological Features

An assessment of plant communities containing A. ledebourianum, A.ivasczenkoae, A. ubinicum, and A. schoenoprasum in the Kazakhstan Altai showed that the studied populations are restricted to humid habitats with constant light exposure and nutrient-rich soils. The soils are predominantly mountain chernozems, and the terrain is generally flat, with slopes rarely exceeding 10°. No signs of diseases or pest damage were observed in the examined plants. The populations were in good condition, exhibiting dispersal tendencies. Age structure analysis indicated that the populations are demographically diverse, comprising individuals from all developmental stages (Figure 3).

In all four species studied, young individuals represented the most abundant ontogenetic stage, accounting for the largest proportion of each population. In A. ledebourianum, young plants predominated, followed by juveniles and immatures, while virginile and generative individuals were rare. A. ivascenkoae exhibited a similar pattern, with young individuals dominating, moderate numbers of juveniles and immatures, and few virginile and generative individuals. In A. ubinicum, young and juvenile individuals were most common, whereas immature, virginile, and generative plants were infrequent. A. schoenoprasum showed the highest number of young individuals overall, along with substantial proportions of immatures and juveniles, and smaller cohorts of virginile and generative plants. Collectively, these results indicate that populations of all species are dominated by early developmental stages, while later stages occur less frequently, reflecting species-specific demographic structures within the sampled habitats.

The associated flora of the studied Allium plant communities comprised 124 species belonging to 76 genera and 34 families (Table S1). Characteristic species list included Juncus compressus Jacq., Calamagrostis epigejos (L.) Roth, Deschampsia cespitosa (L.) P. Beauv., Trollius altaicus C.A. Mey., Filipendula ulmaria (L.) Maxim., Geum rivale L., Salix viminalis L., Veronica longifolia L., and Angelica decurrens (Ledeb.) B. Fedtsch. The most species-rich families were Poaceae, Asteraceae, and Cyperaceae.

Comparison with regional floristic lists indicated that these communities are most similar in composition and dominant families to the flora of the West Altai Nature Reserve (Table 1).

In Schoenoprasum-associated communities in the Kazakhstan Altai, Poaceae (14.5%), Asteraceae (9.6%), and Cyperaceae (8.8%) dominated. This mirrors patterns in the Koktau Mountains, West Altai Reserve, and broader Kazakhstan Altai flora, where Poaceae (7.7–12.6%), Asteraceae (11.9–15.5%), and Cyperaceae (3.8–9.6%) prevailed. The top 10 families comprised 71.7–81.0% of species across datasets, reflecting similar dominant family structures.

As for the ecological conditions, four Allium species from the section Schoenoprasum in the Kazakhstan Altai revealed distinct ecological preferences across key environmental gradients, as quantified by Ellenberg Indicator Values (EIVs) (Table S2). EIVs revealed that all studied species of section Schoenoprasum are predominantly heliophilous to semi-shade tolerant, with light preferences ranging from L = 7 to 8.5; the highest light demand was recorded for A. ubinicum (L = 8.5), whereas A. ivasczenkoae exhibited the lowest value (L = 7). Temperature preferences indicated adaptation to montane-temperate conditions (T = 4–5.5), with A. ledebourianum showing the strongest cool indication (T = 4) and A. ubinicum together with A. schoenoprasum reflecting moderate heat requirements (T = 5.5). Moisture preferences corresponded to fresh to well-moistened soils (M = 7–9), with A. schoenoprasum displaying the highest wetness value (M = 9) characteristic of flood meadows, while A. ledebourianum (M = 7.5) and A. ubinicum (M = 7) indicated moderately moist habitats. Soil reaction values (R = 5–5.5) suggested tolerance of slightly acidic to near-neutral conditions, with A. ivasczenkoae and A. schoenoprasum showing the broadest range. Nutrient preferences ranged from moderately to strongly nutrient-rich sites (N = 4.5–8), with A. ledebourianum exhibiting the highest demand (N 8) and A. ubinicum the lowest (N = 4.5). Salinity values (S = 0–4) confirmed that all taxa are glycophytes, with A. ubinicum (S = 3) and A. schoenoprasum (S = 3.5) showing the lowest tolerance, indicating strict confinement to non-saline habitats.

3.2. Morphometric Characteristics of Plants and Flowers

The second step was the assessment of the morphometric and quantitative traits of reproductive individuals of the four studied species based on eight parameters (Table 2).

Plant height ranged 30–110 cm, highest in A. ledebourianum and lowest in A. ubinicum. Stem number per clump and leaf number were highest in A. ledebourianum. Leaf length averaged 83–95 cm in A. ledebourianum vs. 32–44 cm in A. ubinicum. Stem base diameter and flower stalk length varied moderately. Leaf width ranged 2–12 mm, with greatest variability in A. ledebourianum. CV showed moderate-high variation, especially in leaf number and height.

The morphometric characteristics of inflorescences were analyzed across four Allium species and presented in Table 3.

Inflorescence length varied significantly among species, with A. ledebourianum exhibiting the widest range and A. ivasczenkoae the shortest. Inflorescence diameter showed a similar trend, with the largest in A. ledebourianum and the smallest in A. ivasczenkoae. The number of flowers per inflorescence was also highest in A. ledebourianum (74.16 ± 11.87) and lowest in A. ivasczenkoae (20.5 ± 1.37), indicating pronounced interspecific differences in floral density. The highest variability among these morphometric traits was observed for A. schoenoprasum.

Of particular importance in the morphometric analysis of adult reproductive individuals is the flower, the principal reproductive organ of the plant. Assessment of its structure revealed consistent and distinctive differences in morphology (Figure 4, Table 4).

Raw floral morphometry data is provided in Table S3. Floral morphology of the studied species showed that, although A. ledebourianum possesses the largest inflorescences, A. ubinicum and A. ivasczenkoae produce the largest individual flowers. Populations of A. ledebourianum were characterized by comparatively long stamens (5.99–6.76 mm), whereas A. ubinicum exhibited a relatively robust ovary (3.69 mm) (Table 4).

A. ubinicum and A. ivasczenkoae exhibited larger petal sizes compared to the smaller overall dimensions of A. ledebourianum. A. schoenoprasum showed intermediate characteristics, with notable variability in ovary and anther sizes. Variability and error metrics across species suggest reliable measurements with natural dispersion, highlighting distinct morphological profiles among the Allium species studied.

Despite the observed diversity in floral morphology among species of section Schoenoprasum, the flowers of individuals from the Kazakhstan Altai can be standardized for systematic and taxonomic purposes. Morphometric analysis enabled the development of diagrams representing the mean dimensions of floral structures (Figure 5).

To determine the underlying relationships among floral architecture traits and their potential response to environmental pressures, Pearson’s correlation analysis was performed on the morphometric parameters for each species, as well as on the mean values of the four Allium species against the relevant ecological indicator values (EIVs) (Figure 6).

In A. ledebourianum, strong positive correlations were observed between PL and PW, and between SL and SW (Figure 6A). Petal width also showed a strong positive correlation with OL. Conversely, significant negative correlations were noted between AL and both OW and AW. A moderate negative correlation was also found between AL and OL. In A. ivasczenkoae, PL exhibited a moderate positive correlation with PW (Figure 6B). SL showed positive correlations with SW and OL. Petal width also correlated positively with OL. AL displayed a notable positive correlation with AW. Similar to A. ledebourianum, negative correlations were present between AL and both OW and AW. A. ubinicum displayed fewer significant correlations among the morphometric parameters (Figure 6C). A strong positive correlation was observed between PL and PW. A very strong positive correlation was found between OL and OW, and also between SL and PCL. For A. schoenoprasum, a moderate positive correlation was found between PL and PW (Figure 6D). SL also showed a positive correlation with SW. A strong positive correlation was observed between OL and OW. Distinctly, AL demonstrated moderate negative correlations with OW and AW.

As for the environmental factors, PW showed a strong positive correlation with L. PL exhibited strong positive correlations with EIVs for M and N (Figure 6E). Conversely, PL was negatively correlated with R and S. AL showed strong negative correlations with L, M, and N, while correlating positively with S and R. EIVs for T were moderately correlated with SL and negatively correlated with PCL.

To assess the multivariate patterns and relationships among the four Allium species based on their morphometric parameters, PCA and hierarchical clustering were performed (Figure 7).

The PCA biplot (Figure 7A) illustrates the distribution of the four Allium species in a two-dimensional space defined by the first two principal components, which together accounted for 94.18% of the total variance in the dataset. The plot reveals a clear separation among the species, with A. ledebourianum and A. ivasczenkoae clustering closely together, indicating a high degree of similarity in their floral morphology. A. schoenoprasum occupies an intermediate position but is positioned closer to the A. ledebourianum / A. ivasczenkoae cluster than to A. ubinicum, which is distinctly separated from the others, suggesting unique morphometric characteristics. The dendrogram (Figure 7B), on the opposite, grouping A. ubinicum and A. ivasczenkoae into a single, tightly knit cluster, with A. schoenoprasum joining this cluster at a moderate distance. In contrast, A. ledebourianum remains an outlier, joining the main cluster only at the highest linkage distance, thus confirming its morphological distinctiveness.

The ANOVA results (Table S4) revealed that temperature was the predominant environmental driver influencing floral morphology in Allium species, showing the strongest effects across several traits, particularly petal length (PL) and petal width (PW) (P < 0.001). Light intensity and soil reaction also exerted significant but more selective influences, especially on ovary and anther dimensions. Notably, salinity had a pronounced effect on anther traits – both anther length (AL) and anther width (AW) – indicating high environmental sensitivity of reproductive structures to soil salinity levels. In contrast, factors such as nutrient availability and moisture showed trait-specific effects, contributing moderately to the variation in petal, style, and ovary traits. Overall, these results highlight that temperature and salinity are key determinants shaping interpopulation differences in floral morphology.

3.3. Cytotoxic and Antioxidant Activity of Ethanolic Extracts

The study demonstrated that extracts from four Allium species collected in the Kazakhstan Altai exhibited different levels of cytotoxic activity against Artemia salina (Table 5).

The brine shrimp lethality assay revealed that ethanolic extracts of the four Allium species exhibited dose-dependent cytotoxic effects (Table 5). Mortality remained at 0% in the solvent control (0.5% EtOH), confirming that the ethanol fraction (≤ 0.5% v/v) did not contribute to toxicity. Among the tested species, A. ubinicum showed the highest cytotoxic activity, with an LC₅₀ of 5.9 μg/mL, followed by A. ivasczenkoae (7.8 μg/mL) and A. schoenoprasum (9.6 μg/mL). A. ledebourianum displayed the lowest toxicity, with an LC₅₀ of 10.9 μg/mL. No neurotoxic symptoms (spasmodic movements or paralysis) were observed in nauplii for any extract. The order of cytotoxic potency was therefore: A. ubinicum > A. ivasczenkoae > A. schoenoprasum > A. ledebourianum. These results indicate that ethanolic extracts of A. ubinicum and A. ivasczenkoae possess relatively higher cytotoxic potential, warranting further bioactivity-guided fractionation and compound identification studies.

To quantify the antioxidant activity of the samples, the IC₅₀ value was determined, representing the concentration of the extract required to reduce free radical levels by 50% (Table 6).

The results (Figure 8) show that A. ledebourianum exhibited no detectable antioxidant activity, as its optical density remained unchanged over time across all tested volumes, indicating the absence of radical scavenging.

Among the studied species, A. ubinicum exhibited the strongest antioxidant activity, as indicated by its lowest IC50 value, followed by A. schoenoprasum and A. ivasczenkoae. In contrast, A. ledebourianum did not show detectable antioxidant activity. The decline in optical density over time was most pronounced at higher extract volumes, particularly for A. ubinicum, confirming its superior radical scavenging potential. Thus, the antioxidant activity of the studied Allium species decreases in the order A. ubinicum > A. schoenoprasum > A. ivasczenkoae.

3.4. Genetic Analysis

In this study, a comparative analysis of three chloroplast markers – matK, psbA–trnH, and rbcL – was conducted for four Allium species. Raw sequences are provided in Table S5. For each marker, the total aligned sequence length and nucleotide composition (A, T(U), G, C) were determined, followed by phylogenetic assessment using the Neighbor-Joining (NJ) method. The sequence statistics are summarized in Table 7.

The matK region (684 bp) showed an AT-rich composition (A 39.4%, T 32.1%) typical of chloroplast coding sequences, with minimal interspecific variation (±0.3%), confirming its high conservation. The psbA–trnH spacer (648 bp) exhibited higher T content (34.1%) and slightly greater variability due to its noncoding nature, while rbcL (538 bp) was the most compositionally balanced (A 28.0%, T 28.6%, G 22.0%, C 21.5%), consistent with strong functional constraints on this gene.

Based on the aligned sequences, dendrograms were constructed for each marker using the NJ method (Figure 9).

The matK and psbA–trnH phylograms revealed a consistent topology, showing A. ledebourianum and A. ubinicum as the most closely related species (pairwise distance 0.000–0.001). A. ivasczenkoae formed a sister clade to this pair, while A. schoenoprasum was the most genetically distinct, branching basally to the others. In contrast, the rbcL tree displayed a different topology: A. ledebourianum clustered with A. ivasczenkoae (distance 0.0019), whereas A. ubinicum appeared as the most divergent taxon. Thus, while matK and psbA–trnH support a close relationship between A. ledebourianum and A. ubinicum, rbcL indicates affinity between A. ledebourianum and A. ivasczenkoae. Such topological incongruence among chloroplast markers is not uncommon in Allium and may reflect heterogeneous evolutionary histories—for instance, incomplete lineage sorting, hybridization, or chloroplast capture among sympatric taxa of the section Schoenoprasum. Consequently, the present data should be interpreted as showing two alternative but related hypotheses: (1) A. ledebourianum is central to both clusters, sharing plastid similarity with A. ubinicum (matK, psbA–trnH) and with A. ivasczenkoae (rbcL); and (2) A. schoenoprasum remains consistently divergent across all markers. Further multilocus or genomic analyses (e.g., complete plastome sequencing or nuclear ITS/GBS datasets) are required to resolve these interspecific relationships more conclusively.

4. Discussion

4.1. Demographic Structure, Ecological Niche and Morphometric Differentiation of the Allium Section Schoenoprasum in Kazakhstan Altai

The ontogenetic spectra of A. ledebourianum, A. ivasczenkoae, A. ubinicum, and A. schoenoprasum are dominated by early stages (young, juvenile, immature) with few virginile and generative individuals (Figure 3). Such regeneration-biased distributions are typical for perennial herbs in montane environments and likely reflect high recruitment with slow or uncertain transition to reproductive maturity rather than decline from biotic damage – an interpretation supported by the absence of pest or disease impacts [48].

Floristically, Allium-bearing communities represent a broad meadow–steppe assemblage (124 associated vascular species) dominated by Poaceae, Asteraceae, and Cyperaceae (Table 1), consistent with regional West and Kazakhstan Altai vegetation patterns. This indicates that Allium populations occur within common regional matrices rather than isolated, taxonomically unique assemblages. Ecological indicator values (EIVs) further define their niches: all four taxa are heliophilous to semi-shade tolerant (L = 7–8.5), inhabit fresh to moist soils (M = 7–9), prefer neutral to slightly acidic substrates (R ≈ 5–5.5), require moderate–high nutrients (N = 4.5–8), and exhibit low salinity tolerance (S = 0–4) (Table S2). Within-section differences – A. ledebourianum’s higher nutrient demand (N = 8) and A. schoenoprasum’s preference for wetter microsites (M = 9) – reflect microtopographic specialization. The use of EIVs as ecological proxies is well established in regional floristics and provides a reliable framework for interpreting species distributions where detailed environmental data are lacking [49,50]. Comparable ecological patterns have been reported in other Central Asian endemics such as Allium karataviense, A. oreoprasoides, and A. fedtschenkoanum, which also occupy narrow alpine–subalpine niches and display strong specialization to light and moisture gradients in the Tien Shan and Pamir–Alai regions [51,52]. This similarity suggests that adaptive responses to steep ecological gradients are a common feature driving diversification of endemic Allium taxa throughout Central Asia

These demographic and ecological traits align with broader Allium patterns, as many taxa show strong phenotypic plasticity to light, moisture, and nutrient gradients that structure niches in montane habitats [52]. Juvenile-dominated spectra and narrow habitat preferences likely reflect life-history strategies and environmental filtering. Petal and anther traits vary with ecological factors (Figure 6E), indicating environment-driven allocation shifts – larger petals in moist, shaded sites for pollinator attraction, versus greater investment in male traits in open, nutrient-poor areas. Similar plastic floral and foliar responses are reported in A. prattii and related taxa along elevation and moisture gradients [53].

Multivariate analyses (PCA, clustering) separate A. ledebourianum and A. ubinicum from the remaining taxa, consistent with seed-morphology data showing A. ubinicum’s distinct propagule traits (larger seed mass), suggesting divergent life-history strategies within the section (Figure 7). The congruence of floral, vegetative, and seed traits supports integrated adaptive divergence or strong phenotypic plasticity across environmental gradients in the Kazakhstan Altai [54]. Overall, despite co-occurrence in humid, nutrient-rich microsites, the four species exhibit distinct demographic patterns and coordinated morphological responses to environmental factors. These findings align with broader literature on Allium ecological and morphological variation [50], emphasizing the combined roles of life-history strategy, environmental filtering, and plasticity in shaping species differentiation within montane meadow–steppe systems.

Vegetative and reproductive morphometrics reveal species-specific profiles: A. ledebourianum reaches the greatest stature and leaf number, while A. ubinicum is smallest but bears relatively larger individual flowers – indicating a trade-off between vigor and flower size (Table 2 and Table 4). Inflorescence traits (length, diameter, flower number) vary interspecifically, reflecting divergent allocation strategies (many small vs. few large flowers) (Table 3). Correlation matrices reveal integrated floral modules (petal and ovary dimensions) and, in some taxa, trade-offs between male and female organs (negative anther–ovary correlations in A. ledebourianum) (Figure 6).

The demographic and ecological features observed in the Kazakhstan Altai Allium taxa carry important conservation implications. The dominance of early ontogenetic stages and narrow ecological amplitudes indicate limited reproductive turnover and potential sensitivity to habitat alteration. Endemic species such as A. ivasczenkoae and A. ubinicum, confined to specific alpine and subalpine niches, may be especially vulnerable to anthropogenic pressures and climate-induced shifts in moisture regimes. Conservation measures should therefore prioritize the protection of moist meadow–steppe habitats and maintenance of population connectivity to support natural regeneration. Integrating demographic monitoring with molecular data could help identify genetically distinct or isolated populations that warrant targeted in situ and ex situ conservation efforts.

4.2. Bioactivity in Four Allium Species from the Kazakhstan Altai

The differential cytotoxic and antioxidant responses of the Allium section Schoenoprasum extracts from the Kazakhstan Altai parallel patterns observed in other Allium species, yet with distinctive traits. The A. ubinicum extract showed the strongest toxicity (LC₅₀ of 5.9 μg/mL), exceeding the relatively weak effects seen in A. ledebourianum, which aligns with prior reports of low cytotoxicity in A. cepa and A. sativum extracts [29]. In antioxidant assays, A. ubinicum again performed best (IC₅₀ = 88 µL), whereas A. ledebourianum lacked detectable radical scavenging activity. Comparable IC₅₀ ranges in other Allium taxa (e.g. A. rotundum with IC₅₀ ≈ 260 µg/mL) support the view that antioxidant potency varies widely even within the genus [18]. Together, these results suggest that A. ubinicum may harbor enriched bioactive compounds – likely phenolics or sulfur-containing molecules – that confer both moderate cytotoxicity and strong antioxidant potential, making it a promising candidate for deeper pharmacological investigation.

4.3. Phylogenetic Implications and Evolutionary Insights of Chloroplast Markers

Molecular phylogenetics is essential for clarifying evolutionary relationships within Allium, one of the most taxonomically complex monocot genera due to its diversity, hybridization, and polyploidy [55,56]. Using three chloroplast markers (matK, rbcL, and psbA–trnH), we examined four species of section Schoenoprasum from the Kazakhstan Altai. The matK and psbA–trnH trees each recovered two main clusters: one grouping A. ivasczenkoae with A. ledebourianum, and another uniting A. schoenoprasum and A. ubinicum. In contrast, the rbcL tree showed a slightly different topology, with A. schoenoprasum occupying an intermediate position. Although the topologies varied among markers, all analyses consistently indicated a close relationship between A. ivasczenkoae and A. ledebourianum.

The chloroplast markers differed in their resolving capacity. MatK produced well-supported clades, reflecting its relatively high substitution rate and established value as a core plant barcode [57]. psbA–trnH yielded a similar overall topology, though with lower statistical support, which is consistent with its higher variability and potential homoplasy. The congruence of these two markers supports the recognition of two main clusters among the studied taxa. In contrast, rbcL, a locus generally more informative at higher taxonomic levels, showed lower variability and a slightly different topology, in agreement with previous findings of limited resolution in Allium [58,59].

Our phylogenetic results partially align with earlier Allium studies based on ITS and rDNA [35,54], which also recognized Schoenoprasum as a distinct lineage but reported unresolved relationships among Eurasian taxa. The close association between A. ivasczenkoae and A. ledebourianum in matK and psbA–trnH trees corresponds with their morphological similarity in scape and floral traits [60]. The intermediate position of A. schoenoprasum in the rbcL tree likely reflects its wide ecological amplitude and genetic variability across its extensive Eurasian–North American range rather than evidence of recent introgression.

Incongruence among chloroplast markers is a common feature in Allium phylogenetics, often resulting from lineage sorting or historical hybridization events at deeper evolutionary scales [20,61]. Given the maternal inheritance of plastid DNA, our results are interpreted as reflecting maternal lineage relationships rather than direct evidence of hybridization.

The Kazakhstan Altai represents an important center of Allium diversity in Central Asia [60]. The close affinity between the endemic A. ivasczenkoae and the more widespread A. ledebourianum may indicate recent local divergence, possibly driven by ecological specialization of the former in alpine habitats compared with the broader distribution of the latter. Similar patterns of ecological differentiation have been documented in other Allium complexes [58].

Phylogenetic clustering among the Altai taxa likely results from historical isolation, post-glacial recolonization, and habitat-driven divergence. The region’s complex topography and Pleistocene climatic oscillations fragmented montane flora, promoting allopatric differentiation and endemism. The relationship between A. ivasczenkoae and A. ledebourianum may represent a recent diversification event maintained by ecological separation, whereas A. schoenoprasum, with its broad range and heterogeneity, may have acted as a widespread ancestor or genetic bridge facilitating secondary introgression among regional lineages. Similar east–west diversification patterns across the Altai occur in other steppe–alpine taxa, reflecting shared biogeographic processes shaping Central Asian Allium.

Comparable evolutionary trends occur in other Central Asian endemics, such as A. carolinianum, A. oschaninii, and A. oreoprasum from the Tien Shan and Pamir–Alai, which also exhibit strong regional structuring and hybridization between high- and mid-altitude lineages [36]. Like our findings, these taxa show chloroplast haplotype sharing and weak interspecific barriers, indicating that introgression and ecological adaptation shaped diversification across Central Asian Allium. The Kazakhstan Altai thus represents part of a broader evolutionary continuum linking the mountain systems of Central Asia, where repeated glacial–interglacial cycles promoted lineage exchange, divergence, and persistence in microrefugia.

Taxonomically, our results support the distinctiveness of all four species, each forming monophyletic groups in the combined tree. The close clustering of A. ivasczenkoae with A. ledebourianum raises the question of rank, but consistent molecular, morphological, and ecological divergence supports their recognition as separate species [58]. The heterogeneity of A. schoenoprasum echoes ongoing debates over its intraspecific taxonomy.

While plastid markers provided valuable insights, they represent only maternal inheritance and may obscure reticulate evolution [62]. Nuclear loci such as ITS and low-copy genes, and genomic approaches like RAD-seq and genome skimming, are increasingly resolving complex groups. Future studies integrating nuclear and plastid data with morphology and ecology will be essential for clarifying diversification and species boundaries in the Altai Allium

5. Conclusions

All investigated Allium populations occurred in humid, nutrient-rich habitats on gentle slopes (<10°) with mountain chernozems. Age structure analysis showed dominance of young individuals and few mature plants, suggesting stable regeneration. The associated flora comprised 124 vascular species, mainly Poaceae (14.5%), Asteraceae (9.6%), and Cyperaceae (8.8%), typical of Altai meadow–steppe communities. Morphometric assessment revealed clear interspecific variation: A. ledebourianum reached up to 110 cm with large umbels (49 mm, 74 flowers), whereas A. ubinicum was smaller (30–50 cm) but had proportionally larger petals and robust ovaries. Correlation and PCA analyses distinguished species clusters, with A. ledebourianum and A. ivasczenkoae closely related, while A. ubinicum and A. schoenoprasum were distinct, explaining 94.2% of total variance. Floral traits were significantly influenced by temperature, moisture, light, and salinity, indicating strong morphological plasticity and adaptation to local environments. Although this study focused on phenotypic differentiation, future work combining molecular and metabolomic profiling with broader geographic sampling will better resolve evolutionary relationships, adaptive mechanisms, and chemical diversity. Overall, Altai Allium species demonstrate pronounced ecological specialization and adaptive morphological diversity shaped by the region’s environmental heterogeneity.