1. Introduction

Due to the uneven distribution of organisms, species diversity varies greatly from region to region across the earth. Studies on geographic pattern of biodiversity are an important content of macroecology and biogeography (Ricklefs, 2004). Some well-known patterns have been revealed by researchers, for instance, diversity generally decreases from the tropics to the poles (Pianka, 1966; Gaston, 2000; Hillebrand, 2004). Compared to a large number of studies on the variation of species diversity at local (i.e. α-diversity) and regional (i.e. γ-diversity) scales, research on β-diversity (i.e. the variation in community composition among sites) has significantly increased in recent years (Si et al., 2017). β-diversity is a scalar between α- and γ-diversity (Whittaker, 1960; Qian et al., 2020, 2021a), and studies on the spatial patterns and environmental correlates of β-diversity can help to understand the variation of α-diversity between local sites, and the origin and the maintenance of γ-diversity in the region (Qian et al., 2021a).

Current composition of species in a region reflects the interplay between ecological and evolutionary processes (Ricklefs, 2004; Qian et al., 2019), thus, in addition to taxonomic diversity, attention should be paid to the variation of phylogenetic diversity along environmental gradients. Taxonomic β-diversity is quantified with taxonomic dissimilarity metrics (e.g. Jaccard and Sørensen dissimilarity indices; Baselga, 2010), and can provide information about the degree of overlap and distinctness of species between different communities (Qian et al., 2021a). However, this information does not take into account the phylogenetic relationship between species in these communities. Compared with taxonomic β-diversity, phylogenetic β-diversity is quantified with phylogeny-based dissimilarity metrics (e.g. phylogenetic Sørensen dissimilarity index; Bryant et al., 2008) and can more effectively detect community similarity in the context of evolutionary history, particularly when focal communities share no species (i.e. with the highest taxonomic dissimilarity) but share the same genera or families (Ives and Helmus, 2010; Qian et al., 2021a). Thus, a comprehensive understanding of the geographic patterns of β-diversity and its relationship with environmental variables requires that taxonomic and phylogenetic β-diversity of communities are compared (Graham and Fine, 2008; Peixoto et al., 2017; Qian et al., 2020, 2021a).

Geographic patterns of taxonomic or phylogenetic β-diversity and their environmental correlates for terrestrial organisms have been well investigated at global, continental and regional scales (McKnight et al., 2007; Wang et al., 2012; Weinstein et al., 2014; Peixoto et al., 2017; Pinto-Ledezma et al., 2018; Qian et al., 2020, 2021a). Geographic patterns of taxonomic β-diversity vary greatly among different biogeographic regions. Some studies show that the geographic range and niche width of a species increase with latitude (Stevens, 1989), and taxonomic β-diversity follows a latitudinal gradient, with β-diversity high in tropical regions and low in temperate regions (Stevens and Willig, 2002; Dyer et al., 2007; Lewinsohn and Roslin, 2008). Other studies show that compared with their surrounding regions, some mountains (such as Andes, Rocky Mountains, and Himalayas) have higher taxonomic β-diversity (McKnight et al., 2007; Gaston et al., 2007; Melo et al., 2009). Geographic patterns of phylogenetic β-diversity have been shown to be consistent with the pattern of taxonomic β-diversity, but the values of phylogenetic β-diversity are consistently lower than those of taxonomic β-diversity across the whole research scale (Terlizzi et al., 2009; Qian et al., 2013, 2020, 2021a). For instance, Qian (2009) compared the β-diversity of major terrestrial vertebrates in the world and showed that the β-diversity at the species level (taxonomic β-diversity) was 1.24 and 1.85 times that at the genus and family levels (phylogenetic β-diversity), respectively.

In general, β-diversity reflects the change of species composition on the environmental gradient across large geographic scales (Whittaker, 1960). The dispersal ability and environmental adaptability of species limits their distribution range; thus, it could be predicted that β-diversity is positively correlated with the geographic distance and environmental difference between areas (Nekola and White, 1999; Qian and Ricklefs, 2007; Qian et al., 2020). Previous studies have found that environmental differences play a more important role than geographic distances in shaping patterns of β-diversity for plants in North America (Qian and Ricklefs, 2007; Qian et al., 2013), whereas geographic distances are more important for β-diversity patterns of freshwater fish assemblages in the same continent (Qian et al., 2021b). Geographic and environmental differences may shape β-diversity in different types of organisms at different degrees, depending on habit, living environment and distribution mode of organisms. There is a consistent and strong collinearity between the environmental differences and geographic distances, thus, the greatest possibility is that the two factors jointly affect the geographic patterns of β-diversity (Qian and Ricklefs, 2007; Qian et al., 2009; Chen et al., 2010).

Previous studies have focused on not only total β-diversity but also its two distinct components, turnover and nestedness (Baselga, 2010; Fu et al., 2019). Turnover reflects the replacement of species between sites, whereas nestedness reflects the degree to which species in a species-poor site represent a subset of species in the species-rich site (Baselga, 2010). Areas with a warm climate are expected to have higher species turnover for β-diversity, whereas the contribution of species nestedness to β-diversity is expected to be higher in colder and harsher environments (Baselga, 2010; Dobrovolski et al., 2012). The relative importance of turnover and nestedness components for total β-diversity among species communities vary with spatial scale and local environmental factors (Dobrovolski et al., 2012; Fu et al., 2019). The mechanisms generating variation in species diversity (including taxonomic and phylogenetic diversity) across a region can be better understood by exploring the relationship between geospatial and environmental factors with β-diversity and its components.

Compared with terrestrial ecosystems, freshwater ecosystems are less impacted by climate (Cook, 1985; Santamaría, 2002; Grosberg et al., 2012; Alahuhta et al., 2017), and the dispersal of freshwater organisms is somewhat affected by water systems (Qian et al., 2021b; Larsen et al., 2021). Geographic distance is expected to play a more important role than climate distance in determining the β-diversity of freshwater organisms, as observed in previous studies on freshwater fishes (Griffiths, 2017; Qian et al., 2021b). Aquatic and marshy plants are an important part of freshwater ecosystems (Vymazal and Kröpfelová, 2008). A large number of previous studies have focused on the geographic patterns and environmental correlates of α- and γ-diversity of aquatic plants (Alahuhta, 2015; Alahuhta et al., 2018, 2020; Murphy et al., 2019, 2020; Zhou et al., 2022), but there are few studies focusing on β-diversity and its components of aquatic plants across a large geographic scale, especially for the phylogenetic β-diversity (Alahuhta et al., 2021).

China has a great diversity of geographic and environmental features, and has a rich flora with about 29,000 angiosperm species (Qian et al., 2019), including over 800 aquatic plant species (Chen et al., 2012; Wang et al., 2021). Thus, China is an ideal place for studying the β-diversity patterns of aquatic plants in relation to geographic and climatic distances. Here, we use a plant distribution database derived from multiple sources to explore the geographic patterns and climatic determinants of taxonomic and phylogenetic β-diversity of aquatic angiosperms in China. In particular, we intend to answer the following questions: 1) Are the geographic patterns of taxonomic and phylogenetic β-diversity congruent for aquatic angiosperms in China, and is the taxonomic β-diversity consistently higher than phylogenetic β-diversity? 2) Is the ratio between the nestedness component and total β-diversity (βratio) high in northwestern China and low in southeastern China? 3) Does the geographic distance play a more important role than climatic distance in shaping the patterns of β-diversity for aquatic angiosperms across China?

2. Materials and methods 2.1. Study area and environmental variables

China was divided into grid cells of 100 km × 100 km, and cells with less than 75% of land and non-ocean water areas located within China were excluded. As a result, 903 grid cells were used in this study, as in Qian et al. (2021a). Latitude and longitude of the middle point of each grid cell were extracted using ArcGIS (v.10.2; ESRI, 2016). Nineteen bioclimatic variables at the resolution of 30 arc-second were obtained from the WorldClim database (http://www.worldclim.org), and the average value of each climate variable for each grid cell was calculated using ArcGIS. Environmental variables refer to water environmental conditions in aquatic angiosperm habitats, including salinity, dissolved oxygen (DO), pH, total nitrogen (TN), total phosphorus (TP), and ammonia nitrogen (NH3N). Raw data of these variables were obtained from China National Environmental Monitoring Centre (http://www.cnemc.cn/). We used spatial interpolation methods to convert these point data to national scale surface data in ArcGIS, and extracted and calculated the average value of each variable in each grid cell. We also used ArcGIS to extract and calculate elevation, slope and relief degree of land surface (RDLS) from the SRTM 250 m DEM, which was obtained from the CIAT-CSI SRTM website (http://srtm.csi.cgiar.org). ArcGIS was also used to extract water area (WA) and shoreline length (SL) data in China from 1:250,000 national basic geographic database, which was downloaded from the National Geomatics Center of China (NGCC, http://www.ngcc.cn/). The ratio of water surface to land area (RWL) of each grid cell was calculated. The shoreline development index (SDI) indicates the irregularity of the shoreline controlling for water area effects, and is defined as , where SL is shoreline length and WA is water area (Eadie and Keast, 1983). We used elevation, slope, RDLS, WA, SL, RWL, and SDI to represent topographic variables.

Geographic, climatic, environmental and topographic distances between each pair of grid cells were measured as Euclidean distances (Qian et al., 2020). Geographic distance was calculated based on the middle points of paired cells, and was log₁₀-transformed before subsequent analysis (Qian et al., 2021b). For the calculation of climatic, environmental and topographic distances, we used principal components analysis (PCA) to generate synthetic variables. PC1 and PC2 of climate variables (accounting for 60.53% and 20.76% of the variance, respectively) were used to represent the 19 original climatic variables, and PC1, PC2 and PC3 of environmental variables (accounting for 34.83%, 23.92% and 17.56% of the variance, respectively) were used to represent six original environmental variables, and PC1 and PC2 of topographic variables (accounting for 43.81% and 28.25% of the variance, respectively) were used to represent seven original topographic variables.

2.2. Aquatic angiosperm data set and phylogeny construction

A checklist of 889 aquatic angiosperms was compiled from several monographs and online sources on aquatic plants in China (Zhou et al., 2023). These plants were recorded as “aquatic plant” at least one time in the literature. Woody plants and plants growing in the sea were removed from our list. Species distribution data reported by Lu et al. (2018) were used as a primary data source for the distribution information of each aquatic angiosperm. Because the data set of Lu et al. (2018) did not include some of the aquatic angiosperms used in our study, we supplemented the data with floras, monographs, research papers, and online sources, including the National Specimen Information Infrastructure (NSII, www.nsii.org.cn) and the Global Biodiversity Information Facility (GBIF, 2021), and plant checklists for local floras (e.g. counties, nature reserves and national forest parks) in the literature (Zhao et al., 2006; Qian and Chen, 2016; Qian et al., 2018).

Species names were standardized mainly according to “The Plant List” version 1.1 (http://www.theplantlist.org), using the package U.Taxonstand (Zhang and Qian, 2023). The names on Species 2000 China (http://www.sp2000.org.cn/) and World Flora Online (http://www.worldfloraonline.org) were also checked for some new species or new treatments that did not appear on The Plant List. The phylogenetic tree of aquatic angiosperms was generated from the largest dated mega-tree using the R package “V.PhyloMaker2” based on GBOTB.extended.TPL.tre and build.nodes.1 (Jin and Qian, 2019, 2022). For species absent from the mega-tree, we used the Scenario S3 approach to add them to their families and genera in the mega-tree. Phylogenetic trees generated by V.PhyloMaker2 or its predecessors (i.e. S.PhyloMaker and V.PhyloMaker; Qian and Jin, 2016; Jin and Qian, 2019) have been commonly used in studies on phylogenetic diversity and structure of plants (e.g. Qian et al., 2019, 2020; 2021a; Yue and Li, 2021; Zhang et al., 2021; Huang et al., 2023).

2.3. Metrics of taxonomic and phylogenetic β-diversity

Sørensen dissimilarity index (βsor) was used to measure both taxonomic and phylogenetic β-diversity (Tβsor and Pβsor). Following Baselga (2010), βsor was partitioned into two components, βsim and βnes. βsim (including Tβsim and Pβsim), which is the Simpson dissimilarity index, quantifies β-diversity due to turnover of species between sites. βnes (including Tβnes and Pβnes) quantifies β-diversity resulting from nestedness of species between sites. βsor and its two components are defined as follows: βsor = (b+c)/(2a+b+c), βsim = min(b,c)/[a+min(b,c)], and βnes = βsor-βsim, where a is the number of species shared by the two sites, b is the number of species unique to one site and c is the number of species unique to the other site (Baselga, 2010). When applied to phylogenetic β-diversity, shared and unique species are replaced with shared and unique branch lengths, respectively (Leprieur et al., 2012; Qian et al., 2020). Both taxonomic and phylogenetic β-diversity were calculated using the R package “betapart” (Baselga and Orme, 2012).

The value of the total taxonomic and phylogenetic β-diversity and their components of turnover and nestedness of each grid cell for aquatic angiosperms in China was calculated using the neighborhood (moving window) approach, which represented the average of β-diversity values resulting from all pairwise comparisons between the focal cell and each of its first-order neighboring cells (Qian et al., 2020). For both taxonomic and phylogenetic β-diversity, we calculated the ratio of βnes to βsor, using the formula: βratio = βnes/βsor (Qian et al., 2020), and denoted Tβratio and Pβratio for βratio of taxonomic and phylogenetic β-diversity, respectively. A value of βratio less than 0.5 indicates that β-diversity is determined mainly by turnover whereas a value of βratio greater than 0.5 indicates that nestedness is the more important component than turnover in driving β-diversity (Dobrovolski et al., 2012; Qian et al., 2020). In addition, the deviation (βdev) between total taxonomic and phylogenetic β-diversity for each grid cell was calculated using the formula: βdev = (Tβsor-Pβsor)/Tβsor (Qian et al., 2020, 2021a). The value of βdev reflects the degree of phylogenetic lineage exchange between communities, i.e. high values of βdev reflect that the communities may share no species but do share genera or families, whereas low values of βdev reflect that the communities share species from different lineages (Peixoto et al., 2017; Qian et al., 2021a).

2.4. Data analysis

Mantel test (Smouse et al., 1986) was used to determine relationships between taxonomic and phylogenetic β-diversity, as well as the relationships of β-diversity with geographic, climatic, environmental and topographic distances. The values of correlation coefficients (Spearman’s rank correlation for Mantel test, r_s) were considered to be strong for |r_s| > 0.66, moderate for 0.66 ≥ |r_s| > 0.33, and weak for |r_s| ≤ 0.33 (Qian et al., 2019, 2020). Multiple regression on distance matrices (MRM; Legendre et al., 1994) was conducted to determine the relative contribution of dispersal limitation (geographic distance) and environmental filtering (climatic distance) on taxonomic and phylogenetic β-diversity of aquatic angiosperms in China. We used a variance partitioning approach based on MRM to separate the total explained variance (R²) into three parts (Legendre and Legendre, 2012): explained uniquely by geographic distance, explained uniquely by climatic distance, and explained jointly by geographic and climatic distances. Mantel correlation analysis and MRM were performed using the R packages 'vegan' (Oksanen et al., 2020) and 'ecodist' (Goslee and Urban, 2022), respectively. All the analyses were carried out using the R 3.3.3 software (R Core Team, 2017).

3. Results 3.1. Geographic patterns of β-diversity, βratio and βdev

Taxonomic β-diversity (Tβsor, Tβsim, and Tβnes) showed strong correlations with phylogenetic β-diversity (Pβsor, Pβsim, and Pβnes) (r_s = 0.96 for Tβsor versus Pβsor, r_s = 0.92 for Tβsim versus Pβsim, and r_s = 0.78 for Tβnes versus Pβnes; P < 0.001 in each case based on 999 permutations of Mantel test). Geographic patterns of total taxonomic β-diversity and their components of turnover and nestedness (i.e. Tβsor, Tβsim, and Tβnes, respectively) were highly consistent with those of their counterparts of phylogenetic β-diversity (i.e. Pβsor, Pβsim, and Pβnes, respectively) for aquatic angiosperms across China (Fig. 1). The turnover component of β-diversity tended to be greater in southeastern than in northwestern China; in contrast, the reverse pattern was observed for the nestedness component of β-diversity (Fig. 1). The deviation (βdev) of taxonomic and phylogenetic β-diversity was positive in all grid cells across China (Fig. 2A), but showed no relationship with latitude or climate variables. The high values of βdev were found in western China, middle and lower reaches of the Yangtze River, and northeastern China, however, a broad region in central northern China tended to have low values of βdev (Fig. 2A). Geographic patterns of Tβratio and Pβratio were also highly congruent across China, with the ratio being higher in northwestern China, and lower in southeastern China (Fig. 2B and C).

3.2. Relationships of β-diversity with geographic, climatic, environmental and topographic distances

Total taxonomic and phylogenetic β-diversity (Tβsor and Pβsor) and their components of turnover (Tβsim and Pβsim) were moderately associated with geographic and climatic distances (r_s ranging from 0.404 to 0.654), and weakly associated with environmental and topographic distances (r_s ranging from 0.102 to 0.325) (Table 1). The nestedness component of taxonomic β-diversity (Tβnes) has weak or no correlation with all the four distances, whereas the nestedness component of phylogenetic β-diversity (Pβnes) was moderately associated with geographic and climatic distances, and weakly associated with environmental and topographic distances (Table 1).

The results of MRM showed that the explanatory power of geographic and climatic distances together accounted for more than 96.43% of the total explanatory power of the four distances for both taxonomic and phylogenetic β-diversity and their components (Table 2). That is, the addition of environmental and topographic distances does not increase the interpretation of β-diversity. When only geographic and climatic distances were considered for variance partitioning analysis, these two distances together explained 62.1% and 53.9% of the variance in total taxonomic and phylogenetic β-diversity (Tβsor and Pβsor), respectively, and together explained 53.7% and 26.3% of the variance in turnover component of taxonomic and phylogenetic β-diversity (Tβsim and Pβsim), respectively. Surprisingly, geographic and climatic distances together only explained 1.0% of the variance in the nestedness component of taxonomic β-diversity (Tβnes), but explained 20.1% of the variance in the nestedness component of phylogenetic β-diversity (Pβnes) (Table 2; Fig. 3).

4. Discussion

China has a vast territory, with many rivers, lakes and other water systems, harboring a large number of aquatic angiosperms (Chen et al., 2012; Wang et al., 2021). The diversity of aquatic angiosperms is somewhat restricted by water environments, and thus, the geographic patterns of aquatic angiosperm diversity may differ from those of terrestrial plants (Hillebrand, 2004). Although there are many studies on the geographic patterns and environmental determinants of β-diversity of total, herbaceous or woody plants in China (Wang et al., 2012; Ye et al., 2019; Qian et al., 2020, 2021a), there are few studies focusing on β-diversity patterns of aquatic angiosperms in China and elsewhere. Here, we analyzed a comprehensive data set of aquatic angiosperms in China, and explored the patterns of taxonomic and phylogenetic β-diversity of aquatic angiosperms across China. We also related β-diversity to geographic, climatic, environmental and topographic distances between assemblages of aquatic angiosperms to understand the relative contribution of dispersal limitation and environmental filtering on taxonomic and phylogenetic β-diversity.

Our results show that total taxonomic β-diversity (Tβsor) and its two components (Tβsim and Tβnes) are strongly correlated with total phylogenetic β-diversity (Pβsor) and its two components (Pβsim and Pβnes), respectively. This result is consistent with previous studies on total angiosperms in China (Ye et al., 2019; Qian et al., 2020) and freshwater fishes in North America (Qian et al., 2021b). The geographic patterns of β-diversity of aquatic angiosperms in our study are similar to those of β-diversity of total angiosperms in China (Qian et al., 2020). As found for regional assemblages of angiosperms in North America (Qian et al., 2013) and in China (Ye et al., 2019; Qian et al., 2020) and of freshwater fishes in North America (Qian et al., 2021b), as well as of major terrestrial vertebrates in the world (Qian et al., 2009), taxonomic β-diversity is higher than phylogenetic β-diversity for each of the aquatic angiosperms across China (βdev > 0 in most grid cells, Fig. 2). These findings suggest that compared with taxonomic β-diversity, β-diversity based on the phylogenetic relationships between species can more effectively detect similarity among communities (Ives and Helmus, 2010; Qian et al., 2021a).

Qian et al. (2020) found that the ratio between the nestedness component and total β-diversity (βratio) of angiosperms significantly increased with latitude in China. Our results of Tβratio and Pβratio for aquatic angiosperms are consistent with those for the overall angiosperms of China, which show higher values in northwestern China and lower values in southeastern China (Fig. 2; Qian et al., 2020). Generally, in warm and humid regions, the speciation rate is high and the distribution range of species is narrow, whereas in cold and arid regions, the speciation rate is low and the distribution range of species is wide (Stevens, 1989; Guo et al., 2022). Northwestern China, which has more large-ranged species, showed high values of βratio, indicating the greater importance of nestedness, compared to turnover, in β-diversity (Tomašových et al., 2016; Qian et al., 2020). However, the βdev of aquatic angiosperms does not show an obvious geographic gradient, which is not consistent with that of angiosperms overall in China (Fig. 2; Qian et al., 2020). In particular, we observed the highest values of βdev in western China, especially the northern part of the Qinghai-Tibet Plateau. This indicates that phylogenetic β-diversity is lower in these regions when taxonomic β-diversity is accounted for. The water systems of Qinghai-Tibet Plateau are well developed; for instance, there are more than 1000 lakes larger than 1 km² and 346 lakes larger than 10 km² (Wang and Dou, 1998). The Qinghai-Tibet Plateau is densely covered with high mountains, resulting in insufficient connectivity between water systems (Liu et al., 2020), which has led to relatively few of the same species among communities. Meanwhile, the harsh climate of this region may lead to species among communities to be more closely related (from the same genera or families) because of phylogenetic niche conservatism (Qian et al., 2019; Zhou et al., 2022). Thus, we suspect that the developed water systems and harsh climate in these regions may have resulted in high taxonomic β-diversity but low phylogenetic β-diversity.

It was reported that freshwater ecosystems, which are more influenced by water environment than climate, are different from terrestrial ecosystems (Cook, 1985; Santamaría, 2002; Alahuhta et al., 2017). But in our study, we found that β-diversity of aquatic angiosperms in China is weakly associated with environmental and topographic distances, and is restricted by geographic and climatic distances across the large geographical scale (Table 1). This result is consistent with those of studies in Europe (Chappuis et al., 2012) and North America (Alahuhta, 2015; Alahuhta et al., 2020). Feld et al. (2009) also suggested that the general trend in the freshwater realm is that abiotic geo-climatic factors (such as temperature) dominate over other impact factors at broad spatial scales. Taken together, these studies show that, similar to terrestrial plants, on large geographic scales the diversity of aquatic angiosperms is more likely restricted by climate variables, compared to local microenvironments (García-Girón et al., 2020; Zhou et al., 2022).

The β-diversity of aquatic angiosperms has been reported to be mainly caused by turnover instead of nestedness (Viana et al., 2016; Alahuhta et al., 2021). Our results also showed that the distances of variables can be explained to a greater extent by the turnover component, rather than the nestedness, of β-diversity of aquatic angiosperms in China, and this is particularly evident in taxonomic β-diversity (Tables 1 and 2; Fig. 3). However, we did not find that geographic distance explained more variance than climatic distance in β-diversity (Fig. 3). This finding is inconsistent with those of previous studies on freshwater fishes (Griffiths, 2017; Qian et al., 2021b). Qian et al. (2021b) found that geographic distance between watersheds is more important than climate similarity in determining β-diversity between freshwater fish assemblages in North America. Unlike fishes, which live in water through their life cycle, aquatic angiosperms live in water and water–land interfaces, which may also be impacted by climate (Vymazal and Kröpfelová, 2008; Zhou et al., 2022). Furthermore, compared with fishes, which only disperse via water, the seeds of aquatic angiosperms can be dispersed by water flow, wind or animals (Viana et al., 2016; Hilt et al., 2022). Therefore, we believe that the β-diversity of aquatic angiosperms and aquatic animals is affected by different factors. Our study shows that β-diversity of aquatic plants is jointly affected by geographic and climatic distances.

In summary, our study has revealed the geographic patterns of taxonomic and phylogenetic β-diversity of aquatic angiosperms across China, and the correlation between β-diversity and geographic and climatic variables. The geographic patterns of taxonomic and phylogenetic β-diversity are highly congruent for aquatic angiosperms in China, and taxonomic β-diversity is consistently higher than phylogenetic β-diversity. The geographic patterns of taxonomic and phylogenetic β-diversity of aquatic angiosperms in China are jointly affected by geographic and climatic distances. Overall, our work provides insight into understanding the large-scale pattern of aquatic angiosperm β-diversity, and is a critical addition to previous studies on the macroecological pattern of terrestrial organisms.

Acknowledgements

We thank the editor and anonymous reviewers for their useful and professional suggestions. This study was supported by the Second Tibetan Plateau Scientific Expedition and Research (STEP) Program (2019QZKK0502), and the National Science Foundation of China (32260046).

Author contributions

Ya-Dong Zhou: Data curation, Formal analysis, Investigation, Methodology, Wrote the original manuscript draft, Funding acquisition. Hong Qian: Formal analysis, Methodology, Wrote the original manuscript draft, Review & editing. Yi Jin: Formal analysis, Methodology. Ke-Yan Xiao: Investigation. Xue Yan: Investigation, Data curation. Qing-Feng Wang: Funding acquisition, Project administration.

Data availability

This study used published data, which were cited in the article.

Declaration of Competing Interest

The authors declare no conflict of interest.