Karst landscapes are developed on soluble rocks (primarily limestone), covering about 10–15% of Earth's ice-free land surface (Ford and Williams, 2007). Their unique geological properties have given rise to complex, heterogeneous, and "island-like" habitats that, combined with periodic drought, high calcium, and low-nutrient stresses, foster high species diversity and endemism (Clements et al., 2006; Hao et al., 2015; Oliver et al., 2017; Monro et al., 2018). These conditions make karst ecosystems ideal "natural laboratories" for studying speciation and adaptive evolution (Clements et al., 2006; Oliver et al., 2017). However, such high specialization and fragmentation also render karst plants—particularly rare and endangered taxa, as well as plant species with extremely small populations (PSESP)—highly vulnerable to threats such as habitat destruction, climate change, and human disturbance (Chen and Xu, 2023; Chen et al., 2024; Asatulloev et al., 2026; Xie et al., 2026). Rapid advances in sequencing technologies and bioinformatic tools now offer unprecedented opportunities for conservation genomics research on karst plants, and allow us to systematically decipher the genetic basis of their adaptation to extreme habitats, reconstruct their evolutionary history, and scientifically assess their endangered status and conservation potential (Paez et al., 2022; Loegler et al., 2026). Recently, several studies on karst plant genomics, in which chromosome-level genomes of these plants were assembled, have comprehensively analyzed and evaluated their genetic adaptation, endangerment mechanisms, and conservation needs (Chen et al., 2024; Peng et al., 2025; Asatulloev et al., 2026; Xie et al., 2026; Yan et al., 2026). These findings provide a scientific basis for formulating effective rescue and protection strategies for rare and endangered species and PSESP.

These studies reveal the genetic vulnerability of endemic plants in karst regions. Multiple investigations demonstrate that karst plant species have generally undergone long-term and continuous population declines (Chen et al., 2024; Peng et al., 2025; Asatulloev et al., 2026; Xie et al., 2026; Yan et al., 2026). For instance, two Urophysa species experienced significant population reductions since the Miocene (Xie et al., 2026). Oreocharis mileensis underwent two severe population bottlenecks during the Mid–Pleistocene Transition, Late Pleistocene, and Last Glacial Maximum (LGM), followed by a continued decline (Asatulloev et al., 2026). Camellia rubituberculata populations contracted sharply during the Naynayxungla Glaciation and continued to decline from the Penultimate Glaciation to Last Glaciation (Yan et al., 2026). O. esquirolii populations experienced bottlenecks across multiple glacial periods, with a particularly dramatic reduction during the LGM (Peng et al., 2025). All four studied Begonia species underwent historical bottlenecks during the Pleistocene (Chen et al., 2024). Moreover, human activities have drastically exacerbated karst habitat fragmentation, leading to severe genetic isolation among plant populations. In Oreocharis mileensis, for example, interpopulation genetic differentiation index (F_ST) values reached as high as 0.77, with nearly severed gene flow; core populations exhibited high inbreeding coefficients (F_IS), long runs of homozygosity (ROH), and high burdens of deleterious mutations (Asatulloev et al., 2026). Similarly, U. rockii is endangered due to restricted gene flow, severe inbreeding, and high habitat specialization (Xie et al., 2026). Additionally, the endangered species O. esquirolii (Peng et al., 2025) and B. masoniana (Chen et al., 2024) also showed high interpopulation genetic differentiation, reduced genetic diversity, inbreeding, and accumulation of deleterious mutations. These studies indicate that the endangered status of these species results from the combined effects of long-term historical pressures and recent anthropogenic impacts.

While revealing the genetic vulnerability, these studies further explore potential adaptive mechanisms. Whole-genome duplication (WGD) events can provide abundant genetic material for adaptation (Feng et al., 2024). Oreocharis mileensis experienced three lineage-specific WGD events during its evolution (Asatulloev et al., 2026). Camellia rubituberculata underwent an additional WGD event shared with closely related species (Yan et al., 2026). Emerging evidence suggests that such WGD events may contribute to adaptive evolution in karst plants, though the generality of this pattern requires broader taxonomic testing. For example, a study on the karst cave plant Primulina huaijiensis also identified a lineage-specific WGD event, and gene family expansions (e.g., WRKYs) resulting from gene retentions after this event are thought to contribute to adaptation to high-salinity and drought stresses in karst habitats (Feng et al., 2020). Notably, a recent genus-wide pan-genome study of Primulina provides broader evidence for these patterns, revealing that karst-adapted species exhibit lineage-specific WGD duplicate-retention biases, especially for transcription factors, potentially facilitating adaptive evolution (Feng et al., 2025). Such WGD events and the biased retention of duplicates may constitute a potential genetic foundation for adaptive evolution in karst plants. It should be noted that current insights are derived from a limited number of lineages, which could potentially lead to ascertainment bias, as genome projects often target unusual genomes or charismatic endemics, and selection scans can be sensitive to demography and sampling. Molecular adaptation to specific stresses is another critical aspect. Xie et al. (2026) identified multiple candidate Ca²⁺ regulatory genes (e.g., TPC1 and CAX3) under positive selection in Urophysa plants. Similarly, pan-genomic analyses in Primulina identified adaptive evolution hotspots in ion transport and ABC transporter pathways linked to edaphic specialization in karst habitats, with ABC transporters and ion channels emerging as candidates for drought and salt stress tolerance (Feng et al., 2025). Asatulloev et al. (2026) identified genes associated with drought resistance and rapid recovery after rehydration in the "resurrection plant" O. mileensis. C. rubituberculata exhibited specific expansions and contractions of gene families—such as the expansion of genes related to plant hormone signaling and photosynthesis, and the contraction of some calcium signaling-related genes—reflecting adaptive trade-offs in resource allocation and stress response (Yan et al., 2026). These findings identify candidate components and generate testable hypotheses regarding the core gene regulatory networks underlying karst plant adaptation, though the adaptive effects of these genomic patterns remain to be functionally validated.

These genomic studies provide a theoretical basis for formulating precise and forward-looking conservation strategies for endangered karst plants. For example, based on high F_ST values and genetic clustering results, the ten populations of Oreocharis mileensis were divided into eight genetic management units requiring differentiated conservation measures; for core populations with severe inbreeding, assisted gene flow must be carefully planned; for highly isolated populations, ex situ conservation should be prioritized (Asatulloev et al., 2026). For the highly habitat-specialized Urophysa rockii, populations with low genetic diversity warrant priority protection (Xie et al., 2026). For O. esquirolii, with its severely contracted populations and high deleterious mutation burden, ex situ conservation is as important as habitat protection (Peng et al., 2025). For Begonia masoniana populations with high deleterious mutation loads, the highest conservation priority and urgent intervention are needed (Chen et al., 2024). Together, these cases demonstrate that tailoring conservation actions according to population-specific genetic characteristics—such as delineating genetic management units, planning assisted gene flow, prioritizing ex situ conservation for isolated or genetically depauperate populations, integrating habitat protection with ex situ measures, and prioritizing urgent intervention for populations with high deleterious mutation burdens—offers a scientifically rigorous and operationally practical framework for protecting endangered karst plants.

Collectively, current karst plant genomics studies provide deep insights into their unique endangerment mechanisms and adaptive evolution, while effectively integrating theoretical findings into conservation strategy formulation. This drives a paradigm shift from traditional approaches toward precision conservation guided by genomic information.

Despite significant progress, given the high species richness and habitat heterogeneity of karst regions, there remains considerable room for further exploration in the following areas:

1. Constructing pan-genome maps and integrating machine learning to decipher the genetic basis of adaptation. Pan-genomes can capture variations missed by reference genome-based approaches, particularly structural variations (SVs), which may be key sources of adaptive potential (Fang and Edwards, 2025; Loegler et al., 2026). The recent Primulina pan-genome exemplifies how such resources can reveal adaptive SV hotspots in ion transport and ABC transporter pathways linked to edaphic specialization (Feng et al., 2025). Constructing pan-genome and SV maps for key karst plant taxa will provide a more comprehensive understanding of genetic adaptation mechanisms. However, the resulting massive, high-dimensional, and complex data challenge traditional population genetics methods. Machine learning approaches, such as deep learning, have demonstrated strong potential in population genetics for efficiently detecting selection signals and predicting phenotypic effects of adaptive variants (Huang et al., 2024). For example, in Setaria italica, the integration of pan-genome, genome-wide association studies (GWAS), and interpretable machine learning identified a key SV that regulates leaf sheath and pulvinus color, illustrating how interpretable machine learning can decode complex trait architecture by integrating pan-genomic data with transcriptomic and phenotypic information (Wang et al., 2025). Deep integration of pan-genome resources with machine learning algorithms—for example, using graph neural networks (GNNs) to integrate multi-omics data including pan-genomic, transcriptomic, epigenomic, and environmental factors—can systematically mine the complex genetic regulatory networks and key gene modules driving adaptation.

2. Deepening the investigation of transposable elements (TEs) in genome evolution and adaptation. TEs, as a class of mobile genetic elements, can mediate speciation and rapid environmental adaptation by influencing genome structure and gene expression (Fan et al., 2025; Zheng et al., 2025). Recent analyses of the genus-wide Primulina pan-genome provide causal evidence linking TE dynamics to karst adaptation: karst-adapted species exhibit significantly reduced LTR retrotransposon content, resulting in genome downsizing that likely mitigates nitrogen limitation in nutrient-poor limestone habitats (Feng et al., 2025). Elucidating TE dynamics and their regulatory roles in key traits will enhance our understanding of the molecular mechanisms behind rapid evolution and stress adaptation in heterogeneous karst habitats.

3. Fine-scale genomic studies from the species to the ecotype level. Following ecotype discrimination criteria proposed by Johannesson et al. (2025), systematically evaluating whether different population types constitute true ecotypes is critical for defining the genetic and ecological boundaries within karst habitats. This process is central to understanding micro-adaptation and speciation.

4. Investigating convergent adaptation across independent karst lineages. Karst habitats impose well-defined selective pressures—including high calcium concentrations, nutrient limitation, and periodic water stress—that may drive independent lineages toward similar phenotypic and genomic solutions. A critical question remains: To what extent do phylogenetically distant karst plants evolve convergent adaptive traits and reuse the same genetic pathways, gene families, or regulatory modules? Future studies should explicitly address the extent and mechanisms of convergence at multiple levels: (i) phenotypic convergence in traits such as calcium tolerance or water-use efficiency; (ii) genomic convergence, including the reuse of specific pathways (e.g., calcium signaling), gene families (e.g., TPC1), or regulatory modules across independent karst lineages. For example, Cao et al. (2023) revealed that TPC1—a gene involved in calcium regulation—underwent convergent evolution in the karst-endemic woody plant Platycarya longipes and karst herb genus Primulina. Expanding such comparisons to diverse karst lineages will help decipher general principles of adaptive convergence and test whether karst environments drive predictable evolutionary outcomes.

5. Precisely assessing the feasibility of genetic rescue based on genomics. When applying assisted gene flow (genetic rescue) to mitigate inbreeding depression in isolated populations, the risk of outbreeding depression must be systematically evaluated (Tengstedt et al., 2026). A robust framework for genetic rescue based on whole-genome data (ROH, chromosome number and structural variations, population divergence time, gene flow, and local adaptation signals) can be used to scientifically evaluate the feasibility of introducing individuals across populations (Tengstedt et al., 2026). This data-driven approach balances enhancing genetic diversity with minimizing outbreeding depression risks, ensuring that genetic rescue efforts are effective and sustainable.

6. Extending karst plant genomic insights to other extreme or special environments. The genomic insights from karst plants offer valuable models for understanding genome evolution in other extreme habitats. Like karst habitats, alpine systems impose severe stresses—including low temperatures, high UV radiation, and hypoxia—that drive genomic responses such as WGD events and contraction of disease-resistance genes (Zhang et al., 2023, 2024). Similarly, desert plants exhibit remarkable genomic adaptations to drought, including the expansion of gene families related to desiccation tolerance (Marks et al., 2024). Mangrove genomes revealed convergent amino acid usage patterns and transposable element reductions as adaptive responses to intertidal salinity fluctuations (He et al., 2020; Xu et al., 2020). Even cave plants have demonstrated lineage-specific WGDs that facilitate adaptation to limestone habitats (Feng et al., 2020). By positioning karst plants within this broader comparative framework, universal mechanisms of environmental adaptation can be distinguished from habitat-specific responses, ultimately allowing karst systems to be established as a model for understanding genome evolution under extreme or special environmental stress.

Acknowledgments

Preparation of this paper was equally funded by the Joint Funds of the National Natural Science Foundation of China (U2571210), the Strategic Priority Research Program of Kunming Institute of Botany, Chinese Academy of Sciences (KIBXD202401), National Natural Science Foundation of China (32471734), the Yuelushan Laboratory Breeding Project, and the Caiyun Postdoctoral Program of Yunnan Province.

CRediT authorship contribution statement

Yongpeng Ma: Conceptualization, Supervision, Funding acquisition, Writing-original draft, Writing-review & editing. Xiongfang Liu: Writing-original draft, Writing-review & editing.

Declaration of competing interest

The author Yongpeng Ma is an Editorial Board Member for Plant Diversity and was not involved in the editorial review or the decision to publish this article. The other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.