1. INTRODUCTION

Freshwater mollusks, a major taxonomic group in freshwater ecosystems, play crucial ecological roles in water purification, nutrient cycling, and habitat structuring (Vaughn and Hoellein 2018). Among them, freshwater bivalves (Bivalvia) act as efficient filter feeders, removing suspended organic matter from the water column and contributing to water quality improvement (Spooner and Vaughn 2012). Their biodeposition activity further stabilizes sediment environments and enhances benthic structure. Such activities promote the abundance and diversity of aquatic insects, creating biodiversity hotspots within river ecosystems. Freshwater gastropods (Gastropoda) function as primary consumers, feeding on periphyton and organic detritus attached to substrates and macrophytes, thereby regulating primary production (Wang and Liu 2024). They also constitute an essential food source for fishes and aquatic birds, maintaining energy flow and trophic connectivity within freshwater food webs. Furthermore, several mollusk species serve as reliable bioindicators of habitat quality and ecological integrity due to their sessile lifestyles and sensitivity to environmental change (Atkinson et al. 2023).

Although fine-scale microhabitat conditions (e.g., water quality, substrate composition) strongly influence their survival and settlement (Urban-Malinga et al. 2021), the large-scale distribution and persistence of freshwater mollusks are primarily shaped by macro- environmental gradients such as climate, topography, and land use that determine the regional structure of freshwater assemblages (Arbuckle and Downing 2002). Therefore, quantitatively identifying the spatial distribution and habitat preferences of freshwater mollusks across these gradients provides essential baseline information for conservation planning and assessing ecosystem vulnerability to climate change (O’Brien et al. 2025).

Despite their ecological significance, freshwater mollusks are among the most imperiled faunal groups worldwide (Bohm et al. 2021). Their limited mobility, habitat specialization, and high sensitivity to anthropogenic disturbance make them particularly vulnerable to extinction (Sayer et al. 2025). In the United States, approximately 10% of freshwater mollusk species are already extinct, and 65-70% are currently listed as threatened or endangered (Johnson et al. 2018). Similar patterns are reported in Europe, where 44- 59% of species face extinction risk, ranking freshwater mollusks as the most threatened taxonomic group on the continent (Cuttelod et al. 2011). Nevertheless, quantitative research addressing how large-scale environmental gradients influence freshwater mollusk distributions remains scarce in Korea.

Most previous research has focused on specific rivers or basins such as the Geumgang and Seomjingang river systems examining local community composition or conducting laboratory-based feeding experiments in Korea (Hwang et al. 2002;IFRI 2010). Such localized or experimental approaches have provided valuable ecological insights but lack a nationwide framework integrating spatial distribution data with environmental gradients. Consequently, significant knowledge gaps remain regarding which species occupy particular habitats and how environmental factors structure freshwater mollusk assemblages across Korea. To establish effective management and conservation strategies, a comprehensive national-scale assessment of species occurrences and habitat associations is urgently needed (Jeong et al. 2024).

Therefore, analyzing species’ ecological niche breadths offer a powerful quantitative framework for understanding how macro-environmental gradients shape species distributions (Slatyer et al. 2013). Ecological niche breadth represents the range of environmental conditions that a species can tolerate or utilize for survival and reproduction (Carscadden et al. 2020). This approach is particularly valuable for freshwater mollusks, whose limited dispersal, strong habitat dependence, and high sensitivity to environmental disturbance make them ideal taxa for linking environmental gradients with species vulnerability and conservation needs (Vaughn and Hoellein 2018;Sayer et al. 2025). Species with broad niche breadths (generalists) can persist across diverse environmental settings, whereas those with narrow niches (specialists) are restricted to limited, specific habitats (Chichorro et al. 2019). These differences strongly influence extinction vulnerability: species with narrower niches are generally more susceptible to environmental change and habitat disturbance (Lee and Jetz 2011).

Accordingly, this study integrates nationwide occurrence data of freshwater mollusks with climatic, topographic, and land-cover variables to quantify the ecological niche breadths of 27 representative species in Korea. By comparing species-specific habitat preferences and tolerance ranges, we aim to identify key environmental drivers shaping freshwater mollusk distributions. This work addresses critical knowledge gaps in freshwater biodiversity research in Korea and provides a foundational framework for conservation prioritization under changing environmental conditions.

2. MATERIALS AND METHODS

2.1. Compilation and processing of biological data

The occurrence data of freshwater mollusks used in this study were compiled by integrating national biodiversity monitoring programs with internationally available datasets. The primary data sources were as follows: (1) the National Aquatic Ecological Monitoring Program (rivers and estuaries) of the National Institute of Environmental Research (NIER: https:// water.nier.go.kr/); (2) the National Park Biodiversity Survey of the Korea National Park Research Institute (KNPRI); (3) the National Endangered Species Survey and National Environmental Survey of the National Institute of Ecology (NIE); (4) freshwater specimen records from the Integrated Platform for Freshwater Biodiversity of the Nakdonggang National Institute of Biological Resources (NNIBR; https://freshwaterbiodiversity. go.kr); and (5) occurrence records registered in the Global Biodiversity Information Facility (GBIF; https://www.gbif.org).

Following data integration, spatial quality control and filtering procedures were performed. Duplicate coordinates and erroneous locations (e.g., marine areas, outside administrative boundaries) were removed. To minimize sampling bias during environmental analyses, all occurrence points were standardized to a 1 km×1 km spatial grid. If multiple records of the same species occurred within a single grid cell, they were counted as a single presence. For robust statistical analysis of habitat responses, only species with more than 100 occurrence records were retained. As a result, a total of 27 species and 11,406 valid occurrence records were included, covering all major river basins in Korea, including the Hangang, Geumgang, Nakdonggang, Seomjingang, and Yeongsangang Rivers, as well as Jeju Island (Figs. 1 and 2).

2.2. Environmental variables

Environmental variables expected to influence freshwater mollusk distributions and available nationwide were selected and categorized into three groups: climate, topography, and land cover.

Meteorological variables (annual mean, maximum, and minimum temperatures) were obtained from the Korea Meteorological Administration (KMA) Climate Information Portal (https://climate.go.kr). These data are part of the national standardized climate grid, derived through dynamical downscaling of a global numerical prediction model over East Asia and the Korean Peninsula, followed by statistical bias correction reflecting domestic topographic characteristics. The final dataset provides values at approximately 1 km spatial resolution, and long-term averages for the period 2001-2020 were used in this study. Topographic information was extracted from the Digital Elevation Model (DEM) provided by the National Geographic Information Institute (NGII), with a spatial resolution of 90 m. Land-cover data were obtained from the Environmental Geographic Information Service (EGIS) of Ministry of Climate, Energy and Environment, Korea. The national land-cover map follows the official classification system based on multisource imagery, including Landsat and multipurpose satellite data, aerial orthophotos, and topographic maps.

Three representative categories-urban, agricultural, and forest-were used to capture the gradient of anthropogenic disturbance and natural habitat conditions. For each occurrence point, the proportion (%) of each land-cover type was calculated within a 500 m radius buffer (Bae and Kim 2025).

All environmental variables were subsequently used as input for calculating Levins’ indices (B: niche breadth; BA: evenness) and for optimum-tolerance (median and t95-t05 range) analyses.

2.3. Data analysis

This study quantitatively analyzed species-specific habitat preferences and environmental tolerance ranges of freshwater mollusks in Korea based on nationwide occurrence data. The analysis included 27 species that were recorded more than 100 times across major river systems. Three analytical approaches were employed as follows:

First, basic biological and distributional information was summarized for each of the 27 species meeting the frequency criterion (≥100 occurrences). For each species, we identified its taxonomic group (Bivalvia or Gastropoda), national protection status (e.g., endangered species, Korean endemic), and spatial distribution across major river basins.

Second, species’ environmental tolerance breadth was quantified using Levins’ B (niche breadth) and Levins’ BA (standardized evenness index). Levins’ index measures how broadly and evenly a species utilizes environmental resources along a given gradient (Slatyer et al. 2013).

The Levins’ B index for niche breadth was calculated as:

where R is the total number of environmental intervals, and P_j is the proportion of occurrences of a species within the j-th environmental interval relative to its total occurrences. Higher B values indicate broader environmental tolerance and a generalist habitat strategy, whereas lower values indicate specialization within a narrower range of conditions.

The Levins’ BA (standardized niche evenness index) was calculated as:

The BA index ranges from 0 to 1, with values close to 1 indicating an even distribution of occurrences across the full range of an environmental gradient (i.e., uniform resource use). Conversely, values approaching 0 indicate that occurrences are concentrated within specific portions of the gradient. This index allows cross-variable comparison of relative niche breadth and was used to assess how evenly each species occupies available habitat conditions.

Third, to complement Levins’ indices and to characterize specific habitat preferences, the optimum (preferred condition) and tolerance range (environmental limits) were estimated for each environmental variable. The optimum value of a species was defined as the median of all environmental variable values at its occurrence points. The median was used instead of the mean to minimize the influence of extreme values (outliers) and to better represent the central tendency of environmental preferences (Hartwig et al. 2020). The tolerance range (t95-t05) quantified the environmental range encompassing 90% of all occurrence records, corresponding to the interval between the 5th and 95th percentiles of observed values. A narrow tolerance range indicates higher environmental sensitivity and specialization, whereas a wider range reflects greater adaptability to varying conditions.

All indices and calculations were performed using R software (ver. 4.3.0; R Core Team 2024). Levins’ B (niche breadth) and BA (standardized evenness) indices were computed using custom R scripts with the dplyr (Wickham et al. 2020) and tidyr (Wickham et al. 2023) packages for data wrangling. Quantile-based functions in base R were applied to derive the optimum (median) and tolerance ranges (t95-t05) of each environmental variable. Visualizations such as heatmaps and distribution plots were generated using ggplot2 (Wickham 2016) and ggridges (Wilke 2021).

3. RESULTS

3.1. Distribution patterns and regional occurrence of freshwater mollusks

Based on nationwide occurrence data collected from major river systems in Korea, a total of 27 freshwater mollusk species with more than 100 occurrence records were selected for analysis (Fig. 1, Table 1). The dataset included two nationally endangered species (Clithon retropictum and Koreoleptoxis nodifila), four Korean endemic species (K. nodifila, Semisulcospira coreana, S. forticosta, and S. tegulata), sixteen species regulated under international export approval, and two alien species (Physa acuta and Pomacea canaliculata). Among all species, S. libertina, Radix auricularia, and P. acuta were the most frequently recorded and widely distributed taxa across river systems. Several species, including Cipangopaludina chinensis malleata, Gyraulus convexiusculus, R. auricularia, and Unio douglasiae, occurred in more than 10% of all surveyed river sites, indicating relatively even distribution across major river systems compared to other taxa. In terms of regional distribution, the Nakdonggang River Basin accounted for the highest proportion of occurrences (30.4%, 9,771 sites), followed by the Hangang (21.8%, 7,004 sites), Geumgang (21.0%, 6,745 sites), Seomjingang (15.2%, 4,903 sites), and Yeongsangang River Basins (11.2%, 3,596 sites), with Jeju Island contributing 0.5% (146 sites).

3.2. Ecological niche breadth and environmental evenness based on Levins’ indices

To quantify environmental adaptability, Levins’ B (niche breadth) and BA (standardized evenness) indices were calculated for the 27 freshwater mollusk species (Fig. 2). High B values were observed in S. libertina, R. auricularia, and P. acuta, indicating broad environmental tolerance and a generalist habitat strategy. In contrast, K. nodifila and C. retropictum (both endangered species), as well as Anodonta woodiana, Nodularia breviconcha, and Limnoperna fortunei, showed low B values, suggesting restricted environmental ranges and a specialist strategy. The B index was positively associated with overall occurrence frequency, meaning that more frequently observed species tended to exhibit broader environmental tolerance. However, the BA index, which represents the evenness of habitat use within the occupied environmental range, did not always follow the same pattern as B. For instance, K. nodifila and L. fortunei displayed low B but relatively high BA values, suggesting even and stable distributions within their limited environmental ranges.

Conversely, S. libertina, R. auricularia, and P. acuta had high B but low BA values, implying that although these species can occupy a wide variety of environments, their occurrences were concentrated under specific conditions-particularly in less urbanized areas. Among environmental factors, most species exhibited low BA values with respect to urban cover, indicating strong habitat constraints in urbanized catchments. In contrast, both agricultural and forest cover showed relatively higher BA values across most taxa, suggesting that freshwater mollusks tend to utilize these environments more evenly, regardless of variation in land-cover proportion. This pattern indicates that while urbanization serves as a major limiting factor for freshwater mollusk habitats in Korea, agricultural and forested landscapes provide relatively stable and evenly distributed habitat opportunities for many species.

3.3. Optimum and tolerance ranges along macro-environmental gradients

Complementary analyses of species’ optimum (median) values and 90% tolerance ranges (5th-95th percentiles) across environmental gradients (Fig. 3) revealed a distinct dichotomy in ecological strategies among Korean freshwater mollusks. Species inhabiting upper river reaches generally exhibited adaptation to cold, less disturbed environments, whereas those distributed in lower reaches preferred warmer and more anthropogenically influenced habitats.

The optimum temperature for all species ranged from 10.65°C to 13.95°C, showing clear differentiation along the thermal gradient. Cold-water taxa such as K. nodifila, S. coreana, and S. gottschei displayed low optimum temperatures and narrow tolerance ranges, indicating high sensitivity to thermal fluctuation. In contrast, A. japonica, L. fortunei, and C. japonica exhibited higher optimum temperatures and broader tolerance ranges, suggesting adaptation to warmer and more variable environments. The narrowest thermal ranges were observed in A. japonica (2.79°C) and A. woodiana (2.87°C), while S. gottschei (4.98°C) and S. libertina (4.86°C) showed the widest ranges, reflecting strong environmental tolerance.

Land-cover gradients showed a similar ecological contrast. Optimum values followed the general pattern of Urban<Agriculture<Forest, with most species showing highest occurrence probabilities in areas with less than 10% urban cover.

K. nodifila exhibited a combination of a low B value (narrow niche breadth) and high BA value (uniform habitat use), with an optimum temperature of 10.65°C, 4.82% urban cover, and 57.34% forest cover, confirming its specialization in cold, forested upstream habitats. This species’ narrow tolerance range and low suitability under urban influence suggest strong vulnerability to habitat disturbance and rising temperatures, identifying it as a representative cold-water indicator species.

Conversely, S. libertina and the invasive L. fortunei exhibited broad environmental tolerance and flexible habitat use, implying that they are likely to persist or even expand under future warming and land-use change. Overall, the thermal and land-cover gradients jointly structure species’ habitat differentiation, highlighting the necessity of considering these interacting environmental factors when predicting biodiversity responses and establishing conservation priorities for freshwater mollusks in Korea.

4. DISCUSSION

4.1. Environmental gradients structuring freshwater mollusk assemblages

This study quantitatively assessed the habitat strategies and environmental responses of 27 freshwater mollusk species (11,406 occurrence records) collected from major river systems across Korea by integrating Levins’ indices (B: niche breadth; BA: evenness) with optimum-tolerance (median and t95-t05) analyses. The results demonstrated that species’ ecological strategies were structured along temperature and land-cover gradients, showing a clear tendency toward differentiation between cold-water species adapted to low-disturbance upstream habitats and warm-water species tolerant of disturbed downstream environments. This bimodal pattern is consistent with global observations of habitat partitioning among freshwater mollusk assemblages (Hayward et al. 2022;Rangaswami et al. 2023) and suggests that ongoing climate warming and land-use intensification could become major drivers of community reorganization in Korean river ecosystems.

4.2. Generalist and specialist strategies revealed by Levins’ indices

Analysis of Levins’ indices further supported this pattern. S. libertina, R. auricularia, and the invasive P. acuta exhibited high B values (broad niche breadth) but low BA values (uneven distribution), indicating that these generalist species can tolerate a wide range of environmental conditions yet remain concentrated in particular habitats, such as less urbanized rivers. Such flexibility corresponds to the ecological traits of disturbance-tolerant taxa capable of persisting under variable flow regimes, pollution stress, or habitat fragmentation (Gates et al. 2015;Vaessen et al. 2024). In contrast, K. nodifila and C. retropictum showed low B and moderate-to-high BA values, indicating a specialist strategy characterized by restricted habitat ranges but relatively uniform occupancy within those limits.

4.3. Optimum-tolerance patterns and key environmental drivers

Consistent results were obtained from the optimum- tolerance analysis. Cold-water taxa such as K. nodifila and Semisulcospira spp. exhibited low optimum temperatures (≤11°C), low urban and agricultural cover, and high forest cover, reflecting adaptation to cold, shaded, and minimally disturbed habitats. Conversely, warm-water species including C. japonica, L. fortunei, and A. japonica showed higher optimum temperatures (≥13°C) and higher tolerance to agricultural or anthropogenic environments. These findings confirm that water temperature and land-cover composition jointly act as dominant environmental drivers shaping freshwater mollusk assemblages in Korea.

Among the land-cover variables, urbanization emerged as the most restrictive factor for species occurrence. Most mollusk species exhibited low BA values along the urban gradient, indicating that their distributions are limited in highly urbanized streams. This likely reflects the combined effects of anthropogenic disturbance, degraded water quality, and habitat discontinuity in urban areas, which reduce habitat suitability for sensitive taxa and favor only a few tolerant species (Gillis et al. 2017). In contrast, agricultural land cover showed relatively high BA values, suggesting that certain tolerant species are capable of maintaining stable populations within agricultural environments. For example, A. woodiana displayed high occurrence frequencies even in catchments with high agricultural cover, demonstrating their adaptability to artificial water bodies such as irrigation reservoirs and drainage canals. These results highlight urbanization as a primary limiting factor for freshwater mollusks in Korea, while agricultural landscapes may still serve as potential habitats for tolerant taxa, albeit at the cost of community homogenization.

4.4. Species-level differentiation and conservation implications

Environmental tolerance also varied considerably among congeneric species, revealing substantial within- genus ecological diversification (Bielen et al. 2016;Payton et al. 2016). Distinct ecological differences were also identified among species within the genus Semisulcospira. Although these species share similar phylogenetic and habitat backgrounds, they exhibited contrasting habitat-use strategies across environmental gradients as reflected in their Levins’ indices (B, BA) and optimum-tolerance patterns. The dominant species, S. libertina, showed a high B value and moderate BA value, indicating a broad habitat range (generalist type) with occurrences concentrated under specific conditions-particularly in less urbanized streams. In contrast, S. coreana and S. gottschei exhibited lower B values but moderately high BA values, suggesting narrower environmental ranges but relatively uniform distribution within those ranges. This pattern implies that these species may be more sensitive to environmental change, yet capable of maintaining stable populations under suitable conditions. S. forticosta displayed intermediate characteristics, with higher BA values for forest and agricultural cover, suggesting a tendency toward balanced habitat use across relatively diverse natural environments. These findings demonstrate that clear ecological differentiation exists even among congeneric species, reflecting differences in physiological tolerance and adaptive strategies. In particular, S. libertina exhibited broad environmental tolerance, whereas S. coreana and S. gottschei tended to maintain stable distributions within more restricted environmental ranges. Therefore, treating the genus Semisulcospira as a single management unit may overlook species-specific ecological traits, emphasizing the need for species- level conservation and management approaches tailored to their respective environmental responses (Bolotov et al. 2020;Aldridge et al. 2023).

This study holds particular conservation significance as it quantitatively elucidates the ecological characteristics of K. nodifila, a Korean endemic and nationally endangered species (Endangered Wildlife Class II). The species exhibited a narrow niche breadth (B), relatively high evenness (BA), and limited tolerance range (t95-t05), identifying it as a typical cold-water specialist. Its optimum environmental conditions were characterized by a mean water temperature of approximately 10.6°C, urban cover below 5%, agricultural cover below 20%, and forest cover around 60%, indicating its confinement to cold, forested headwater habitats. These results provide empirical evidence that the distribution of K. nodifila is primarily determined by the combined influence of temperature and land-cover gradients. The narrow tolerance range and low suitability under urbanized conditions suggest high sensitivity to anthropogenic disturbance, implying that stable populations can persist only in regions with high forest cover and minimal human impact. Consequently, K. nodifila should be regarded not only as a target for species-level protection but also as a bioindicator representing the ecological integrity of cold, forested freshwater ecosystems in Korea.

For effective conservation, the protection of upper catchments with intact forest cover, the long-term stabilization of stream temperature, and the maintenance of continuous riparian buffer zones are essential. Particular attention should be paid to mitigating the impacts of deforestation and downstream development, as even moderate temperature increases may lead to population isolation or local extinction. Continuous monitoring and the maintenance of spatial connectivity among habitats are therefore critical for ensuring the persistence of K. nodifila and other cold-water mollusk species.

4.5. Methodological significance and future perspectives

Beyond its species-specific implications, this study provides an academic contribution by integrating nationwide occurrence data with Levins’ indices and optimum- tolerance analyses to quantify multidimensional environmental responses of freshwater mollusks. This approach clarified how macro-environmental gradients-particularly climate, topography, and land use-structure mollusk distribution patterns, and established a quantitative basis for assessing the vulnerability of endangered species using measurable niche parameters.

Nevertheless, this study has certain limitations. While the dataset covers an extensive national scale, fine-scale habitat characteristics-such as flow velocity, substrate composition, and organic matter content were not explicitly incorporated. Future research should integrate such microhabitat variables to perform multiscale analyses, thereby improving the accuracy of realized niche estimation. Moreover, coupling the optimum-tolerance metrics with species distribution models (SDMs) under long-term climate scenarios would enable the prediction of range shifts and population declines in response to warming and urban expansion. Such efforts will provide critical scientific evidence for formulating effective biodiversity conservation policies for freshwater ecosystems in Korea.