Glacier retreat induces contrasting shifts in bacterial biodiversity patterns in glacial lake water and sediments

doi:10.21203/rs.3.rs-4683287/v1

Download PDF

Research Article

Glacier retreat induces contrasting shifts in bacterial biodiversity patterns in glacial lake water and sediments

https://doi.org/10.21203/rs.3.rs-4683287/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Glacial lake ecosystems are experiencing rapid changes due to accelerated glacier retreat. As glaciers recede, their influence on downstream lakes diminishes, potentially affecting the biodiversity of glacial lake microbial communities. However, there remains a knowledge gap regarding how bacterial biodiversity patterns in glacial lakes are altered by diminishing glacial influence. Here, we investigated shifts in bacterial communities from paired water and sediment samples collected from seven glacial lakes on the Tibetan Plateau, using a space-for-time substitution approach to understand the consequences of glacier retreat. Our findings reveal that bacterial diversity in lake water increases significantly with higher glacial influence index (GI), whereas sediment bacterial diversity exhibits a negative correlation with GI. Both water and sediment bacterial communities display significant structural shifts along the GI gradient. Notably, reduced glacial influence decreases the complexity of bacterial co-occurrence networks in lake water but enhances network complexity in sediments. This divergence in diversity and co-occurrence patterns highlights that water and sediment bacterial communities respond differently to changes in glacial influence over these lake ecosystems. This study provides insights into how diminishing glacial influence impacts the diversity and community structure of bacterial communities in glacial lake water and sediments, revealing contrasting patterns between the two environments. The findings emphasize the need for comprehensive monitoring to understand the implications of glacier retreat on these fragile ecosystems.

Glacier retreat

Glacial lakes

Water and sediment

Bacterial biodiversity

Space-for-time approach

Glacial lakes are directly affected by glacier retreat (Wang et al., 2016), with their developments influenced by long-term climate changes (Zhang et al., 2020; Pandey et al., 2021). In recent decades, rapid global warming has led to substantial mass loss in glaciers worldwide (Yao et al., 2012; Zemp et al., 2019; Immerzeel et al., 2020; Nie et al., 2021). These changes can prompt shifts in glacial lake ecosystems from their initial states to alternative ones (Tiberti et al., 2020). On one hand, new glacial lakes are formed in previously ice-covered areas as the results of glacier retreat, which offer suitable habitats for various biota (Slemmons et al., 2013). On the other hand, as glaciers continue to melt, the distance between glacial lakes and the glacier increases, reducing the glacial influence on lake ecosystems. Consequently, when glacial lakes lose their hydrological connection to the glacier, they transition from a cold and turbid state to a warmer and clearer state (Sommaruga, 2015). These significant changes in glacial lakes are expected to have profound effects on the structure and biodiversity of microorganisms inhabiting these environments. Previous observations have highlighted the importance of the distance of glacial lakes from the ice cap as a driver of microbial diversity in West Greenland lakes (Somers et al., 2020). Recent evidence suggests that bacteria originating from glaciers can contribute significantly—ranging from 40–96%—to downstream glacial lake communities (Liu et al., 2021). Moreover, it has been observed that the contribution of glacier-originated bacteria to downstream lake communities decreases with increasing distance from the glacier terminus. In addition, previous studies indicated that glacier retreat can affect microbial biodiversity of downstream glacial lakes by altering their physicochemical conditions (Milner et al., 2017; Liu et al., 2019). For example, glacier meltwater can transport a large amount of fine inorganic particles into glacier-fed ecosystems, leading to a high turbidity in these ecosystems (Egholm et al., 2012). The high turbidity limits the penetration of light available for primary production (Peter et al., 2018; Liu et al., 2021), further affecting the microbial communities.

Within lake ecosystems, water and sediment represent two distinct lacustrine habitats, each characterized by unique environmental properties (Jiang et al., 2006). Water bacteria originate from sources such as glacier meltwater, rainfall, groundwater, and soil (Nelson, 2009; Sun et al., 2021), whereas sediment bacterial communities develop over time through processes of accumulation, deposition, and erosion (Du et al., 2020). Nevertheless, surficial sediments remain in continual contact with the overlying water, receiving microorganisms and nutrients from the latter (Zeng et al., 2019). Previous studies have pointed out that bacterial communities in water or sediments response differently to environmental change (Xia et al., 2014; Zeng et al., 2019; Nevers et al., 2020; Gao et al., 2021b; He et al., 2021). For example, planktonic bacterial communities in freshwater lakes are more susceptible to eutrophication than sediment bacterial communities (Zeng et al., 2019). In addition, a study conducted in glacier-fed rivers demonstrated that water bacterial diversity was more sensitive to mean annual precipitation, while sediment bacterial diversity was more influenced by nitrogen concentration (Zhang et al., 2021a). However, the extent to which bacterial communities in water and sediments change during the ontogeny of glacial lakes, and whether there are differences in response to environmental changes induced by glacier retreat, remains unclear.

To address this knowledge gap, we explored the biodiversity and co-occurrence patterns of bacterial communities in water and surface sediment samples from seven lakes formed by the retreat of the Kuoqionggangri glacier. Although the formation times of these lakes are unknown, the distance of the lakes from the glacier also reflects, to some extent, a gradient of the ontogenetic age of the glacial lakes, with the nearest lakes being most recently formed. Therefore, these glacial lakes offer a unique opportunity to examine changes in bacterial communities during lake ontogeny as consequence of the glacier retreat. Given the significant role of glaciers in the ontogeny of glacial lakes, we hypothesized that the biodiversity and co-occurrence patterns of bacterial communities in the water and sediments of downstream glacial lakes would be strongly influenced by glacial influence. Furthermore, we hypothesized that downstream glacial lake water and sediment bacterial communities would exhibit differential responses to glacial influence.

Study area and sample collection

Kuoqionggangri glaciers (KQGR, 29°51′48″N, 90°12′39″E) are located in southern regions of the Tibetan Plateau (Figure S1). As one of the largest glaciers in Nyainqen Tangglha range, KQGR are mainly controlled by the South Asian monsoon (Zhang et al., 2021c). The annual mean air temperature in this region is -1.28℃ and the annual mean precipitation is ~ 550 mm, with approximately 90% of the precipitation occurring in the summer season from June to September (Gao et al., 2021a). KQGR consist of seven small glaciers, and at least nine glacier lakes are expanding as the glacier calving. Among them, seven lakes are originated from a rapidly retreating KQGR Ⅰ glacier (Figure S1). Of the seven lakes, three lakes are in close proximity to the glacier but are located in different directions of the glacier. The fourth lake is connected to lake 3, which lake 5 is connected to the glacier by a stream. Lake 6 and lake 7 are situated behind a ridge, and has lost connect to the glacier (Figure S1).

In this study, we collected surface snow (0-15cm) on KQGR Ⅰ glacier, paired water and surface sediment samples were collected synchronously from seven nearby pro-glacier lakes on 11 July 2020 (Figure S1). In each lake, at least three replicate water and sediment samples were collected. In total, 16 snow samples, 24 water and 24 sediment samples. For snow sampling, we used a sterilized Teflon shovel to transfer surface snow into sterile 700 ml Whirl-Pak bags. The samples were stored in cooling containers and transported to the laboratory and slowly thawed over 24–48 h. After melting, approximately 500 ml snowmelt water was filtered onto a 0.22 µm polycarbonate membrane (47 mm, Millipore, USA) for DNA extraction. For water sampling, 1 L of water was collected at depths of 10–20 cm with a 1000 mL Ruttner water sampler (Hydro-Bios, Altenholz, Germany) (Liu et al., 2021) into a sterile plastic bottle. In the laboratory, approximately 500 ml water first filtered through 20 µm mesh (Millipore, USA) to remove large particles, and then filtered through a 0.22 µm polycarbonate membrane. Another portion was used for physicochemical analysis. Surface sediment samples were collected with stainless-steel shovel and placed into the sterile Whirlpak bags. Subsequently, samples were brought to the laboratory and stored at -80 ℃ until further analysis.

Estimation of glacial influence

To assess the influence of glaciers on bacterioplankton biodiversity, glacier index (GI) was calculated according to (Jacobsen and Dangles, 2012):

$$\text{G}\text{I}=\frac{\sqrt{\text{S}\text{i}\text{z}\text{e}}}{\text{D}\text{i}\text{s}\text{t}+\sqrt{\text{S}\text{i}\text{z}\text{e}}}$$

where GI is the glacier index, Size is the glacier size (km²), and Dist is the distance of the sampling site from the glacier terminus (km). GI value of 1 indicates that the distance between the sampling site and the glacier terminus is zero, where the glacier has the greatest influence, while GI value of zero means that the glacier has the least influence.

Physicochemical analysis

Sampling site coordinates and altitude were collected using a Global Positioning System (GPS). Lake area and straight-line distances from sampling site to glacier terminus were calculated from remote sensing image based on GPS coordinates (Kohler et al., 2022).

The water temperature (Temp), pH, conductivity (Con), total dissolved solids (TDS) were measured in the field using a multiprobe Water Quality Sonde (Hydrolab DS5, Hach, USA). In the laboratory, water samples were filtered through 0.45 µm hydrophilic polyethersulfone syringe filter (25 mm, Anpel) to determine the dissolved organic carbon (DOC), total nitrogen (TN), nitrate (NO₃⁻) and sulfate (SO₄²⁻). DOC and TN were measured by a TOC-Lcph (Shimadzu Corp., Japan) following the standard methods (APHA 2005). The NO₃⁻ and SO₄²⁻ concentration was determined with Dionex ion chromatography (IC) system 600 (Thermo Fisher Scientific Instrument Co., LTD). Sediment pH and conductivity was determined in a 1:5 sediment/water suspension using a PHS-3C pH and DDS-307A conductivity meter, respectively. The sediment total carbon (TC) was determined using a TOC-Lcph analyzer. The sediment NH₄⁺-N and NO₃⁻ -N were extracted using a ratio of 5 g sediment to 25 mL 2 mol/liter KCl. After shaking for 1 h, the extracts were filtered using a filter paper and the concentrations were measured using a continuous flow Injection analyzer (Smartchem200) (Hou et al., 2019). Total nitrogen (TN) and total phosphorus (TP) contents was determined by digestion H₂SO₄ and H₂SO₄-HClO₄, respectively, and ultimately determined via flow analyzer (Smartchem200). Details about the samples and physicochemical measurements are in the table S1.

DNA extraction and sequencing

The total bacterial DNA was extracted from the filter membrane or 0.5 g sediment samples using the MoBio Power Soil DNA isolation kit (MoBio Laboratories, Carlsbad, CA, USA) according to the manufacturer’s instructions. The extracted DNA was resuspended into 100 µL of nuclease-free water and quantified using a Qubit 4.0 fluorometer with the dsDNA Broad Range assay kit (Invitrogen, Singapore).

The V4-V5 region of the bacterial 16S ribosomal RNA (rRNA) gene was ampli-fied using primer set 515F (5′-GTGCCAGCMGCCGCGGTAA-3′) and 907R (5′-CCGTCAATTCMTTTRAGTTT-3′) (Chen et al., 2020a). The 50 µL polymerase chain reaction (PCR) reaction systems were prepared in triplicate and each contained 10 ng DNA template, 5 µL 2x Premix Taq DNA polymerase (Takara Biotechnology, Dalian, China), 1 µL of each primer (10 µM) and 20 µL of nuclease-free water. The PCR reaction was performed under the condition: 94℃ for 3 min, 30 cycles of 94℃ for 30 s, 60℃ for 30 s, 72℃ for 1 min and a final cycle of 5 min at 72℃. Afterwards, PCR products were purified with Agencourt AMpure XP beads, pooled at equal volume, and sequenced on an Illumina MiSeq machine (Illumina, San Diego, CA) with a paired-end strategy. All raw sequence obtained in this study have been deposited into the National Center for Biotechnology Information (NCBI) Sequence Reads Archive (SRA) database under the BioProject number PRJNA873432.

Sequence analysis

Raw reads of 16S rRNA gene amplicons were quality-controlled and analyzed using QIIME2 (2020.04) pipeline. Basic criteria were respected following the official suggestions. DADA2 was used for the detection and correction of amplicon sequence data, as well as the generation of the phylogenetic tree. Taxonomic assignment of sequences was assigned using the Naive Bayes classifier trained by the Silva 138 16S rRNA database (Quast et al., 2012). With the pipeline provided by QIIME2, Amplicon Sequence Variants (ASVs) were produced in place of Operational Taxonomic Units (OTUs) for a higher resolution of taxonomic annotations (Glassman and Martiny, 2018). ASVs that were annotated as chloroplasts, archaeal and unassigned sequences (not affiliated with Bacteria) were excluded from subsequent analysis. Overall, we obtained a total of 1605568 high-quality sequences from 64 samples, which were clustered into 10820 ASVs at the cutoff of 97% sequence identity.

Estimation of the stochastic ratio in community assembly

The community assembly processes of water and sediment bacterial communities were evaluated by using neutral and null models. The neutral community model (NCM) was applied to evaluate the potential importance of stochastic processes on community assembly (Sloan et al., 2006). The null model-based stochastic ratio was calculated using the modified method with the Bray-Curtis distance to determine the relative importance of stochastic and deterministic processes in shaping community assembly (Guo et al., 2018; Liang et al., 2020). The normalized stochasticity ratio (NST) is performed in R using the “NST” package to reflect the contribution of stochastic processes based on relative differences between the observed situation and the null expectation (Ning et al., 2019). The NST index defines 0.5 as the boundary point to determine whether the community assembly is more deterministic (< 0.5) or more stochastic (> 0.5). Levins’ niche breadth index was calculated for both water and sediment bacterial communities using the “niche.width” function within R package “spaa” (Zhang and Zhang, 2013). In general, a bacterial community with a wide niche breadth is expected to be more metabolically flexible at the community level than one with a narrow niche breadth.

Network analysis

To estimate how species coexistence changes in different habitats as glaciers retreat, we first divided the lakes into two groups based on GI values: lakes close to the glacier terminus with high GI values (HGL, lake 1 to lake 3), and lakes away from the glacier terminus with low GI values (LGL, lake 4 to lake 7). Then, constructed separate correlation networks for 24 water and 24 sediment samples. Finally, subnetworks of HGL and LGL were extracted from the water and sediment network using the induced_subgraph function in “igraph” package in R (Csardi and Nepusz, 2006) and analyzed individually. All Robust correlations with Spearman’s correlation coefficients (ρ) > 0.8 and false discovery rate [FDR]-adjusted P -value < 0.01 were used to construct networks. Network visualization was generated with Gephi version 0.9.2 (Bastian et al., 2009). Each node represents one ASVs, and each edge represents a strong and significant correlation between two nodes. For each network, we calculated a set of topological features: average path length, network diameter, average degree, clustering coefficient, and graph density. Higher average degree, clustering coefficient, and graph density suggest a more connected network. In addition, lower average path lengths and diameters indicate closer associations in the network (Barberán et al., 2012; Ma et al., 2016). Furthermore, the 1000 Erdös-Réyni random networks, which exhibit the same number of nodes and edges as the real networks, were calculated in the “igraph” R package.

Community cohesion was calculated to quantify the degree of connectivity of the water and sediment bacterial communities (Herren and McMahon, 2017). In this analysis, positive and negative cohesions were calculated for each sample as the sum of the significant positive or negative correlations between taxa weighted by OTU abundances. We use the absolute value of negative:positive cohesion to test whether changes in network stability are caused by changes in GI values.

Statistical analyses

The Bayesian-based SourceTracker method was performed to quantify the extent of contribution of potential source to the sink (Knights et al., 2011). In the present study, SourceTracker analysis were performed twice: (i) lake water samples were defined as sink, and the snow samples as potential source, and (ii) both lake water and snow samples were considered as potential sources and sediment samples as sinks. Spearman correlation analysis was used to determine the bacterial genera in water and sediment samples that were significantly correlated with the distance from the glacier. Bacterial alpha diversity indices, including the species richness and Shannon index were calculated for each sample in the R package “vegan”. The Bray–Curtis dissimilarities of bacterial communities between water and sediment were examined using Wilcoxon rank-sum tests, performed using the wilcox.test function in “stats” package in R (Field et al., 2012). Water and sediment bacterial community compositions were visualized using non-metric multidimensional scaling (NMDS) based on Bray-Curtis dissimilarities. Vegan’s functions metaMDS was used for ordinations, envfit was used to fit environmental vectors and ordisurf, which uses generalized additive models, was used to visualize the sampling site distance from the glacier in the ordinations (Oksanen et al., 2013). Analysis of similarity (ANOSIM) was used to evaluate differences in bacterial communities between habitats. All statistical analyses were performed in R3.6.1 (http://www.r-project.org). The distance-based multivariate linear model (DistLM) based on the community Bray-Curtis dissimilarities was performed in the DISTLM_forward3 program to determine the combination of environmental variables that significantly explain the observed dissimilarity among the samples (McArdle and Anderson, 2001).

Bacterial taxonomy and alpha diversity patterns

For water samples, GI was significantly positively correlated with pH, but negatively correlated with water temperature (Figure S1a). No significant relationships were observed between GI and TN or DOC. For sediment samples, TN and TC were significantly and negatively correlated with GI (Figure S1b), reflecting the high nutrient concentrations in the sediment of early-formed glacial lakes.

The dominant bacterial phyla (Proteobacteria in Class) present in the water samples, ranked in decreasing order by mean relative abundance, were Alphaproteobacteria (51.5% ± 18.8%), Gammaproteobacteria (22.5% ± 14.5%), and Bacteroidota (13.7% ± 7.1%) (Figure S2a). In sediment samples, Gammaproteobacteria, Bacteroidota, Acidobacteriota and Alphaproteobacteria were the dominant bacterial linkeage, accounting for 28.4%, 20.9%, 13.5% and 11.3% of the total sequence, respectively (Figure S2b).

At the genus level, we found a total of 18 bacterial genera in lake water and sediments were significantly related to the distance from the glacier (P < 0.05) (Table S2). Among them, the genera Longivirga (affiliated to Actinobacteriota), Tetrasphaera (Actinobacteriota), Edaphobaculum (Bacteroidota), RBG-13-54-9 (Chloroflexi), Syntrophorhabdus (Desulfobacterota), Schlesneria (Planctomycetota), Roseococcus (Proteobacteria), UKL13-1 (Proteobacteria), Pseudolabrys (Proteobacteria) and Terrimicrobium (Verrucomicrobiota) both in water and sediment increased with the distance from the glacier. However, the relative abundance of the genera Arcicella (Bacteroidota) and Roseomonas (Proteobacteria) showed decreasing distance trend. In addition, Gemmatimonas, Phenylobacterium, Rhizobacter, Ellin6067 and Undibacterium genera in lake water exhibited a negative correlation with the distance from the glacier, while showed a positive correlation in the sediment.

Alpha diversity indexes of bacterial communities in both water and sediment samples are presented in Table S3. In water samples, the average of richness was estimated to range from 220 to 429, with the lowest value observed in lake 6. The Shannon index ranged from 3.67 to 6.00, and the highest value occurred in lake 1, which was close to the glacier terminus. Sediment samples exhibited higher bacterial richness and Shannon diversity index compared with the corresponding water samples. The bacterial richness and Shannon index in sediment samples ranged from 575 to 1144, and 6.88 to 9.06, respectively. Both indexes increased from lake 1 to lake 5 and decreased towards lake 7, which was far away from the glacier terminus. In addition, the bacterial richness and Shannon indexes of water samples increased gradually with increasing GI values (linear regression: R² = 0.097, P = 0.139 and R² = 0.420, P < 0.001, respectively) (Fig. 2a). In comparison, the bacterial richness and Shannon indexes of sediment samples were significantly negative related to GI values (linear regression: R² = 0.591, P < 0.001 and R² = 0.647, P < 0.001, respectively) (Fig. 2b).

Bacterial β-diversity patterns

NMDS plot showed that bacterial communities clustered by habitat types (Fig. 3a). The results were further confirmed by ANOSIM, which demonstrated significant differences between water and sediment bacterial communities (R = 0.894, P < 0.001). Further, bacterial β-diversity (average Bray-Curtis dissimilarity) in the water communities was significantly lower than that in the sediment communities (Kruskal-Wallis test: P < 0.001, Figure S3a). Bacterial community composition showed significant heterogeneity along the GI gradient regardless of whether the samples were from water (environmental factor fit: R² = 0.952, P < 0.001, Fig. 3b) or sediment (environmental factor fit: R² = 0.943, P < 0.001, Fig. 3c). The β-diversity of the water bacterial community increased significantly with increasing GI (R² = 0.170, P = 0.024), whereas the β-diversity of the sediment community decreased significantly with increasing GI (R² = 0.255, P = 0.004) (Figure S3b).

The DistLM analysis revealed the set of environmental variables that explain significant and independent proportions of the variation in water and sediment bacterial community composition (Table 1). For water communities, GI, water temperature, conductivity, lake area, and TN were significantly correlated with bacterial community composition (P < 0.05) and explained 54.4% of the total variation, with GI having the greatest explanatory power (18.7% of the total variation). Three variables, including GI (P = 0.001), NH₄⁺-N (P = 0.001) and TP (P = 0.003) significantly explained 32.3% of the variation in sediment bacterial community composition, with GI explaining the largest percentage, 14.4% of the total variation.

Source of bacteria in KQGR lake ecosystems

Using Bayesian SourceTracker algorithm, we evaluated the importance of glacier surface snow as a source of diversity for downstream glacial lake water and sediment bacterial communities (Fig. 4). When the sink was set to the bacterial communities of the waters, the contribution of ASVs derived from glacier surface snow to the sequences of downstream lake communities averaged 49.3%, ranging from 79.8% ± 8.3% in Lake 1 to 11.3% ± 1.1% in Lake 7 (Fig. 4a). However, when the sink was set to sediment bacterial communities, the contribution of ASVs derived from glacier surface snow was accounting for only 0.7% − 4.1% of sediment sequences (Fig. 4b). Although the contribution of glacial surface snow to glacial lake water and sediment communities varied, the proportion of its contribution to both lake water (R² = 0.474, P < 0.001, Figure S4a) and sediment communities (R² = 0.372, P = 0.002, Figure S4b) decreased consistently with decreasing GI.

The assembly of water and sediment communities

The neutral community model successfully estimated a large fraction of the relationship between the frequency of occurrence of ASVs and their relative abundance variation in the water communities, explaining 73.1% of the community variance for the water bacterial communities (Fig. 5a). However, the dispersal-related stochasticity explained only 48.0% of the sediment bacterial community variance (Fig. 5b). The m value of the water bacterial communities (0.016) was higher than that of the sediment bacterial communities, indicating that species in water habitats were more dispersed than species in sediment habitats.

The assembly of the water and sediment communities was also explored using a null model approach (Figure S5a). The average normalized stochasticity ratio (NST) of the water communities was 65.1%, which was higher than that in the sediment communities (24.3%), suggesting that stochastic processes had more contribution on the water community assemblage, while deterministic processes had more influence on the sediment community assemblage. The water communities exhibited significantly wider niche breadths than the sediment communities, further indicating that the sediment communities were more influenced by deterministic processes (Figure S5b).

Co-occurrence network of water and sediment communities

To determine the co-occurrence of bacterial communities in waters and sediments of KQGR glacial lakes, samples were grouped according to the degree of glacial influence on the lakes and four sub-networks were constructed (Fig. 6). We found that all network degrees exhibited scale-free characteristics for the water (Power law: R² > 0.87) and sediment (Power law: R² > 0.90) networks, indicating that the network structure in both waters and sediments was non-random. The topological properties in the empirical networks were higher than those in the respective Erdös–Réyni random networks, which suggested that the networks had “small-world” properties and modular structure (Table S4). There were 166 nodes and 1052 edges in the water HGL subnetwork, and the number of nodes and edges in the water LGL subnetwork were 84 and 180 respectively. Additionally, the water HGL subnetwork had a higher average degree, graph density, centrality, but lower modularity than the water LGL subnetwork (Table S4). This suggests that the greater the glacial influence, the more frequent the associations between bacterial taxa in water habitats. However, in sediment habitats, the LGL subnetwork showed a more complex network structure compared to the HGL network, characterized by higher average degree and centrality, shorter average path length and lower modularity (Table S4).

The water HGL subnetwork had larger proportion of positive edges (99.5%) than that of the water LGL subnetwork (97.8%), indicating that the water LGL subnetwork was more stable (Fig. 6b). The sediment subnetwork showed the opposite trend, with a smaller proportion of positive edges in the sediment HGL subnetwork (98.0%) and a larger proportion of positive edges in the sediment LGL subnetwork (99.2%). The ratio of negative:positive cohesion was higher in the water LGL and the sediment HGL bacterial communities than in the water HGL and the sediment LGL communities, respectively (Fig. 6c). This further suggests that the stability of the water bacterial co-occurrence network increases with less glacial influence and the stability of the sediment bacterial co-occurrence network decreases with less glacial influence.

Substituting space for time, the present study provides empirical evidence that glacier retreat exerted significant effects on bacterial communities in glacial lakes, with higher diversity and network complexity in lake water of newly formed lakes near the glacier terminus, and lower diversity and network complexity in sediments of these newly formed lakes. Understanding these contrasting biodiversity patterns is crucial for predicting how glacial lake ecosystems might respond to ongoing glacier retreat driven by climate change.

Distinct patterns of bacterial α-diversity in lake water and sediments under glacial influence

We found that bacterial alpha diversity in lake water increased with higher glacial influence index (GI), suggesting newly-formed lakes closer to the glacier terminus harbor more diverse bacterial communities than early-formed glacial lakes (Fig. 2a). This result aligns with findings that the lake nearest to a retreating alpine glacier had higher diversity than other lakes (Peter and Sommaruga, 2016). The higher bacterial alpha diversity observed in the water of newly-formed, glacier-proximal lakes in this study could be attributed to massive inputs of bacteria originating from glacier surface snow (Fig. 4a). Previous work showed that bacterial communities in the glacier-fed connected hydrological continuum are mainly regulated by stochastic dispersal processes (Liu et al., 2021). Indeed, we observed the dispersal-related stochasticity played a key role in structuring the lake water communities. As glacier melting, those glacier-derived microbes can flow into downstream aquatic ecosystems, leading to higher bacterial alpha diversity in lakes that closer to glacier terminus.

Contrarily, we found that the alpha diversity of sediment bacterial communities decreased with increasing GI (Fig. 2b), indicating higher diversity in the sediment of early-formed glacial lakes further from the glacier terminus. Despite sediments also receiving glacier snow-derived bacteria similarly to lake water, the average proportion was much lower (49.3% versus 82.5% in water) (Fig. 4b). This implies that sediment microbial diversity appears less influenced by inputs of glacial snow-derived species. There are several possible explanations for this result. One possible explanation is that sediments can act as ecological reservoirs, capturing bacteria from varied environments like soils, sands, and plant debris (Crump et al., 2012), enabling higher diversity in newly-formed lakes. Another possibility is that the distinct environmental conditions observed along the GI gradient likely exert a sorting effect on sediment bacterial communities (Figure S1b). Local environmental conditions may favor or inhibit specific bacterial populations, influencing the sediment communities through deterministic ecological processes, such as selection and competitive exclusion (Crump et al., 2007; Besemer et al., 2012; Crump et al., 2012), resulting in lower diversity in newly-formed glacial lakes. This is further supported by our neutral and null model results that a strong influence of deterministic processes on community assembly in sediments.

Shifts in bacterial community composition with glacial influence

Insights into how glacier retreat and environmental drivers shape microbial communities can inform conservation strategies aimed at preserving biodiversity and maintaining ecological balance in these sensitive environments. The glacial index emerges as a pivotal determinant of bacterial community composition in both water and sediment samples (Fig. 3). This suggests a direct link between glacier retreat and microbial community changes. As glaciers recede, modified landscapes and altered flows of nutrients and organic matter likely reshape bacterial communities in adjacent glacial lakes. Besides the glacial influence, local environmental conditions such as conductivity and temperature also played important roles in shaping the distribution of bacterial communities in the water. This finding is consistent with earlier studies highlighting the sensitivity of microbial communities to water temperature and conductivity in freshwater lakes in northern Europe (Lindström et al., 2005) and Tibetan glacial lakes (Liu et al., 2019). Nutrients like ammonium and total phosphorus are also significant factors influencing the shift of sediment bacterial community composition. This points to the critical role of nutrient dynamics in driving the composition of sediment bacterial communities. Studies have shown that the enrichment of C, N, and P from hydrological changes with glacier retreat can reshuffle microbial communities and impact ecosystem functions (Zhang et al., 2021b). Therefore, our results underscore the dynamic nature of microbial ecology in glacial lakes influenced by both natural processes and human activities, contributing to our understanding of ecological responses to environmental change.

In the KQGR glacial lakes, we found 18 genera in water and sediment with relative abundances significantly related to GI (Table S2). For example, the relative abundances of the genera Gemmatimonas, Phenylobacterium, Rhizobacter, Ellin6067 and Undibacterium were positively correlated with GI in lake water, but negatively correlated with GI in sediments. Members of the genus Gemmatimonas are known for their abilities to involve in phosphonate and phosphinate metabolisms, and can also participate in the denitrification process (Peng et al., 2019; Yang et al., 2021). Phenylobacterium, Rhizobacter, Ellin6067 and Undibacterium species can participate in denitrification processes and have been found in a variety of freshwater environments (Baek et al., 2019; Chen et al., 2020b). This indicates that in newly-formed glacial lakes, more lake water microorganisms mediate nitrogen and phosphorus cycling, whereas in early-formed glacial lakes, more sediment microorganisms are involved in nitrogen cycling.

Linkages between bacterial diversity and co-occurrence network complexity

Network analysis can improve our understanding of complex interactions between microorganisms (Yuan et al., 2021). Our results revealed that lake water bacterial networks were more complex in high glacial influence (HGL) lakes near the glacier than low influence (LGL) lakes further away from the glacier. In contrast, sediment networks showed the opposite pattern with greater complexity in LGL lakes (Fig. 6a and Table S4). Network complexity is often positively associated with community stability, as highly connected networks can promote stability (Thébault and Fontaine, 2010). However, we found water LGL and sediment HGL networks exhibited stronger stability compared to water HGL and sediment LGL networks as evidenced by our analyses of proportion of negative edges and cohesion (Fig. 6b and 6c). This finding is in line with previous researches that higher interaction complexity might weaken network stability (de Vries et al., 2018; Fan et al., 2018). This may be because complex species interactions can lead to unstable coupling in microbial interactions, and weakening species interactions usually promotes community stability (Coyte et al., 2015). It is also possible that an increase in the strength of interactions between some key taxa could destabilize trophic level interactions (Kuiper et al., 2015).

Microbial co-occurrence patterns further provide us with a better understanding of the maintenance of bacterial diversity in the water and sediment of glacial lakes during lake ontogeny. The relationship between species interactions and microbial diversity can be described by two models: the Ecological Control model (EC) where high diversity limits further diversification due to niche occupancy, and the Diversity Begets Diversity model (DBD) where interactions create new niches promoting diversification (He et al., 2021; Rabosky and Hurlbert, 2015; Calcagno et al., 2017). Our results supported the latter model, with higher diversity in water of HGL lakes where species associations were more clustered and complex compared to in LGL lake water. Similarly, the relationship between sediment bacterial community diversity and species interactions followed this model as well. This suggests that in glacial lakes, microbial interactions in water and sediment could serve as diversifying selective forces via niche differentiation, influencing diversification and community stability (Calcagno et al., 2017; He et al., 2021).

In summary, by substituting space for time, our results suggest that bacterial diversity, network complexity and stability in lake water and sediments showed opposite trends during glacial lake ontogeny. In lake water, bacterial diversity increased with increasing glacial influence, with higher diversity in newly-formed glacial lakes, which may be related to the increased contribution of glacier-derived bacteria. However, the deposition of exogenous microorganisms and the enhancement of species interactions may be the reasons for the increase of bacterial diversity in sediments in early-formed glacial lake sediments. As glacier retreat progresses with continued anthropogenic climate change, the ontogeny of glacial lakes could lead to the extinction of highly adapted species and the homogenization of aquatic microbial communities. Although glacier retreat has increased bacterial diversity in lake sediments currently, the risk of lake desiccation increases with future climate warming. Moving forward, it would be valuable to explore how these contrasting patterns of microbial diversity impact broader ecosystem functions and processes, and contribute to the resilience or vulnerability of glacial lake ecosystems in the face of climate change. Such insights could further inform conservation strategies and management practices aimed at preserving these unique and rapidly changing habitats.

ACKNOWLEDGEMENTS

This work was supported by the Key Collaborative Research Program of the Alliance of International Science Organizations (Grant No. ANSO-CR-KP-2021-04), National Natural Science Foundation of China (Grant No. 42330410), and the Second Tibetan Plateau Scientific Expedition and Research Program (STEP) (Grant No. 2019QZKK0503).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

AUTHOR CONTRIBUTIONS

Keshao Liu: Writing – original draft, Writing – review & editing, Data curation, Conceptualization. Qi Yan: Investigation, Methodology, Writing – original draft, Formal analysis, Data curation. Xuezi Guo: Methodology, Visualization. Wenqiang Wang: Investigation, Data curation; Zhihao Zhang: Data curation, Visualization. Mukan Ji: Writing – review & editing, Conceptualization. Feng Wang: Methodology, Data curation. Yongqin Liu: Writing – review & editing, Funding acquisition, Conceptualization.

Baek, C., Shin, S.-K., and Yi, H. (2019) Phenylobacterium parvum sp. nov., isolated from lake water. International journal of systematic and evolutionary microbiology 69: 1169-1172.
Barberán, A., Bates, S.T., Casamayor, E.O., and Fierer, N. (2012) Using network analysis to explore co-occurrence patterns in soil microbial communities. The ISME journal 6: 343-351.
Bastian, M., Heymann, S., and Jacomy, M. (2009) Gephi: an open source software for exploring and manipulating networks. In Proceedings of the international AAAI conference on web and social media, pp. 361-362.
Besemer, K., Peter, H., Logue, J.B., Langenheder, S., Lindström, E.S., Tranvik, L.J., and Battin, T.J. (2012) Unraveling assembly of stream biofilm communities. The ISME journal 6: 1459-1468.
Calcagno, V., Jarne, P., Loreau, M., Mouquet, N., and David, P. (2017) Diversity spurs diversification in ecological communities. Nature communications 8: 1-9.
Chen, J., Wang, P., Wang, C., Wang, X., Miao, L., Liu, S. et al. (2020a) Fungal community demonstrates stronger dispersal limitation and less network connectivity than bacterial community in sediments along a large river. Environmental Microbiology 22: 832-849.
Chen, Y., Shao, Z., Kong, Z., Gu, L., Fang, J., and Chai, H. (2020b) Study of pyrite based autotrophic denitrification system for low-carbon source stormwater treatment. Journal of Water Process Engineering 37: 101414.
Coyte, K.Z., Schluter, J., and Foster, K.R. (2015) The ecology of the microbiome: networks, competition, and stability. Science 350: 663-666.
Crump, B.C., Amaral-Zettler, L.A., and Kling, G.W. (2012) Microbial diversity in arctic freshwaters is structured by inoculation of microbes from soils. The ISME journal 6: 1629-1639.
Crump, B.C., Adams, H.E., Hobbie, J.E., and Kling, G.W. (2007) Biogeography of bacterioplankton in lakes and streams of an arctic tundra catchment. Ecology 88: 1365-1378.
Csardi, G., and Nepusz, T. (2006) The igraph software package for complex network research. InterJournal, complex systems 1695: 1-9.
de Vries, F.T., Griffiths, R.I., Bailey, M., Craig, H., Girlanda, M., Gweon, H.S. et al. (2018) Soil bacterial networks are less stable under drought than fungal networks. Nature communications 9: 3033.
Du, L., Wang, R., Gao, X., Hu, Y., and Guo, S. (2020) Divergent responses of soil bacterial communities in erosion-deposition plots on the Loess Plateau. Geoderma 358: 113995.
Egholm, D., Pedersen, V., Knudsen, M., and Larsen, N. (2012) Coupling the flow of ice, water, and sediment in a glacial landscape evolution model. Geomorphology 141: 47-66.
Fan, K., Weisenhorn, P., Gilbert, J.A., and Chu, H. (2018) Wheat rhizosphere harbors a less complex and more stable microbial co-occurrence pattern than bulk soil. Soil Biology Biochemistry 125: 251-260.
Field, A., Miles, J., and Field, Z. (2012) Discovering Statistics Using R. In: SAGE Publications Ltd.
Gao, M., Chen, X., Wang, J., Soulsby, C., and Cheng, Q. (2021a) Climate and landscape controls on spatio-temporal patterns of stream water stable isotopes in a large glacierized mountain basin on the Tibetan Plateau. Science of The Total Environment 771: 144799.
Gao, Y., Zhang, W., and Li, Y. (2021b) Microbial community coalescence: does it matter in the Three Gorges Reservoir? Water Research 205: 117638.
Glassman, S.I., and Martiny, J.B. (2018) Broadscale ecological patterns are robust to use of exact sequence variants versus operational taxonomic units. MSphere 3: e00148-00118.
Guo, X., Feng, J., Shi, Z., Zhou, X., Yuan, M., Tao, X. et al. (2018) Climate warming leads to divergent succession of grassland microbial communities. Nature Climate Change 8: 813-818.
He, Q., Wang, S., Hou, W., Feng, K., Li, F., Hai, W. et al. (2021) Temperature and microbial interactions drive the deterministic assembly processes in sediments of hot springs. Science of The Total Environment 772: 145465.
Herren, C.M., and McMahon, K.D. (2017) Cohesion: a method for quantifying the connectivity of microbial communities. The ISME Journal 11: 2426-2438.
Hou, L., Hu, A., Chen, S., Zhang, K., Orlić, S., Rashid, A., and Yu, C.-P. (2019) Deciphering the assembly processes of the key ecological assemblages of microbial communities in thirteen full-scale wastewater treatment plants. Microbes and environments: ME18107.
Immerzeel, W.W., Lutz, A., Andrade, M., Bahl, A., Biemans, H., Bolch, T. et al. (2020) Importance and vulnerability of the world’s water towers. Nature 577: 364-369.
Jacobsen, D., and Dangles, O. (2012) Environmental harshness and global richness patterns in glacier‐fed streams. Global Ecology Biogeography 21: 647-656.
Jiang, H., Dong, H., Zhang, G., Yu, B., Chapman, L.R., and Fields, M.W. (2006) Microbial Diversity in Water and Sediment of Lake Chaka, an Athalassohaline Lake in Northwestern China. Applied and Environmental Microbiology 72: 3832-3845.
Knights, D., Kuczynski, J., Charlson, E.S., Zaneveld, J., Mozer, M.C., Collman, R.G. et al. (2011) Bayesian community-wide culture-independent microbial source tracking. Nature methods 8: 761-763.
Kohler, T.J., Fodelianakis, S., Michoud, G., Ezzat, L., Bourquin, M., Peter, H. et al. (2022) Glacier shrinkage will accelerate downstream decomposition of organic matter and alters microbiome structure and function. Global Change Biology.
Kuiper, J.J., Van Altena, C., De Ruiter, P.C., Van Gerven, L.P., Janse, J.H., and Mooij, W.M. (2015) Food-web stability signals critical transitions in temperate shallow lakes. Nature communications 6: 7727.
Liang, Y., Ning, D., Lu, Z., Zhang, N., Hale, L., Wu, L. et al. (2020) Century long fertilization reduces stochasticity controlling grassland microbial community succession. Soil Biology and Biochemistry 151: 108023.
Lindström, E.S., Kamst-Van Agterveld, M.P., and Zwart, G. (2005) Distribution of typical freshwater bacterial groups is associated with pH, temperature, and lake water retention time. Applied and environmental microbiology 71: 8201-8206.
Liu, K., Liu, Y., Han, B.-P., Xu, B., Zhu, L., Ju, J. et al. (2019) Bacterial community changes in a glacial-fed Tibetan lake are correlated with glacial melting. Science of the Total Environment 651: 2059-2067.
Liu, K., Liu, Y., Hu, A., Wang, F., Zhang, Z., Yan, Q. et al. (2021) Fate of glacier surface snow‐originating bacteria in the glacier‐fed hydrologic continuums. Environmental microbiology 23: 6450-6462.
Ma, B., Wang, H., Dsouza, M., Lou, J., He, Y., Dai, Z. et al. (2016) Geographic patterns of co-occurrence network topological features for soil microbiota at continental scale in eastern China. The ISME journal 10: 1891-1901.
McArdle, B.H., and Anderson, M.J. (2001) Fitting multivariate models to community data: a comment on distance‐based redundancy analysis. Ecology 82: 290-297.
Milner, A.M., Khamis, K., Battin, T.J., Brittain, J.E., Barrand, N.E., Füreder, L. et al. (2017) Glacier shrinkage driving global changes in downstream systems. Proceedings of the National Academy of Sciences 114: 9770-9778.
Nelson, C.E. (2009) Phenology of high-elevation pelagic bacteria: the roles of meteorologic variability, catchment inputs and thermal stratification in structuring communities. The ISME Journal 3: 13-30.
Nevers, M.B., Byappanahalli, M.N., Nakatsu, C.H., Kinzelman, J.L., Phanikumar, M.S., Shively, D.A., and Spoljaric, A.M. (2020) Interaction of bacterial communities and indicators of water quality in shoreline sand, sediment, and water of Lake Michigan. Water Research 178: 115671.
Nie, Y., Pritchard, H.D., Liu, Q., Hennig, T., Wang, W., Wang, X. et al. (2021) Glacial change and hydrological implications in the Himalaya and Karakoram. Nature reviews earth & environment 2: 91-106.
Ning, D., Deng, Y., Tiedje, J.M., and Zhou, J. (2019) A general framework for quantitatively assessing ecological stochasticity. Proceedings of the National Academy of Sciences 116: 16892-16898.
Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’hara, R. et al. (2013) Package ‘vegan’. Community ecology package, version 2: 1-295.
Pandey, P., Ali, S.N., and Champati Ray, P.K. (2021) Glacier-glacial lake interactions and glacial lake development in the central Himalaya, India (1994–2017). Journal of Earth Science 32: 1563-1574.
Peng, C., Gao, Y., Fan, X., Peng, P., Huang, H., Zhang, X., and Ren, H. (2019) Enhanced biofilm formation and denitrification in biofilters for advanced nitrogen removal by rhamnolipid addition. Bioresource technology 287: 121387.
Peter, H., and Sommaruga, R. (2016) Shifts in diversity and function of lake bacterial communities upon glacier retreat. The ISME Journal 10: 1545-1554.
Peter, H., Jeppesen, E., De Meester, L., and Sommaruga, R. (2018) Changes in bacterioplankton community structure during early lake ontogeny resulting from the retreat of the Greenland Ice Sheet. The ISME Journal 12: 544-555.
Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P. et al. (2012) The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic acids research 41: D590-D596.
Rabosky, D.L., and Hurlbert, A.H. (2015) Species richness at continental scales is dominated by ecological limits. The American Naturalist 185: 572-583.
Slemmons, K.E.H., Saros, J.E., and Simon, K. (2013) The influence of glacial meltwater on alpine aquatic ecosystems: a review. Environmental science Processes & impacts 15: 1794-1806.
Sloan, W.T., Lunn, M., Woodcock, S., Head, I.M., Nee, S., and Curtis, T.P. (2006) Quantifying the roles of immigration and chance in shaping prokaryote community structure. Environmental Microbiology Reports 8: 732-740.
Somers, D.J., Strock, K.E., and Saros, J.E. (2020) Environmental controls on microbial diversity in arctic lakes of West Greenland. Microbial Ecology 80: 60-72.
Sommaruga, R. (2015) When glaciers and ice sheets melt: consequences for planktonic organisms. Journal of Plankton Research 37: 509-518.
Sun, C., Zhang, B., Ning, D., Zhang, Y., Dai, T., Wu, L. et al. (2021) Seasonal dynamics of the microbial community in two full-scale wastewater treatment plants: Diversity, composition, phylogenetic group based assembly and co-occurrence pattern. Water Research 200: 117295.
Thébault, E., and Fontaine, C. (2010) Stability of ecological communities and the architecture of mutualistic and trophic networks. Science 329: 853-856.
Tiberti, R., Buscaglia, F., Callieri, C., Rogora, M., Tartari, G., and Sommaruga, R. (2020) Food web complexity of high mountain lakes is largely affected by glacial retreat. Ecosystems 23: 1093-1106.
Wang, X., Liu, Q., Liu, S., Wei, J., and Jiang, Z. (2016) Heterogeneity of glacial lake expansion and its contrasting signals with climate change in Tarim Basin, Central Asia. Environmental Earth Sciences 75: 1-11.
Xia, N., Xia, X., Liu, T., Hu, L., Zhu, B., Zhang, X., and Dong, J. (2014) Characteristics of bacterial community in the water and surface sediment of the Yellow River, China, the largest turbid river in the world. Journal of Soils and Sediments 14: 1894-1904.
Yang, Z., Zhou, Q., Sun, H., Jia, L., Zhao, L., and Wu, W. (2021) Metagenomic analyses of microbial structure and metabolic pathway in solid-phase denitrification systems for advanced nitrogen removal of wastewater treatment plant effluent: A pilot-scale study. Water Research 196: 117067.
Yao, T., Thompson, L., Yang, W., Yu, W., Gao, Y., Guo, X. et al. (2012) Different glacier status with atmospheric circulations in Tibetan Plateau and surroundings. Nature Climate Change 2: 663-667.
Yuan, M.M., Guo, X., Wu, L., Zhang, Y., Xiao, N., Ning, D. et al. (2021) Climate warming enhances microbial network complexity and stability. Nature Climate Change 11: 343-348.
Zemp, M., Huss, M., Thibert, E., Eckert, N., McNabb, R., Huber, J. et al. (2019) Global glacier mass changes and their contributions to sea-level rise from 1961 to 2016. Nature 568: 382-386.
Zeng, J., Jiao, C., Zhao, D., Xu, H., and Wu, Q.L. (2019) Patterns and assembly processes of planktonic and sedimentary bacterial community differ along a trophic gradient in freshwater lakes. Ecological Indicators 106.
Zhang, G., Yao, T., Xie, H., Yang, K., Zhu, L., Shum, C. et al. (2020) Response of Tibetan Plateau lakes to climate change: Trends, patterns, and mechanisms. Earth-Science Reviews 208: 103269.
Zhang, J., and Zhang, M.J. (2013) Package ‘spaa’. R package version 1.
Zhang, L., Delgado-Baquerizo, M., Shi, Y., Liu, X., Yang, Y., and Chu, H. (2021a) Co-existing water and sediment bacteria are driven by contrasting environmental factors across glacier-fed aquatic systems. Water Research 198: 117139.
Zhang, W., Chen, R., Meng, F., Yuan, H., Geng, M., Cheng, L. et al. (2021b) Ecosystem functioning is linked to microbial evenness and community composition along depth gradient in a semiarid lake. Ecological Indicators 132: 108314.
Zhang, Y., Gao, T., Kang, S., Shangguan, D., and Luo, X. (2021c) Albedo reduction as an important driver for glacier melting in Tibetan Plateau and its surrounding areas. Earth-Science Reviews 220: 103735.

Table 1 is available in the Supplementary Files section

No competing interests reported.

Download PDF

Editorial decision: Revision requested
09 Sep, 2024
Reviews received at journal
09 Sep, 2024
Reviews received at journal
04 Sep, 2024
Reviews received at journal
25 Aug, 2024
Reviewers agreed at journal
12 Aug, 2024
Reviewers agreed at journal
10 Aug, 2024
Reviewers agreed at journal
08 Aug, 2024
Reviewers agreed at journal
07 Aug, 2024
Reviewers agreed at journal
05 Aug, 2024
Reviews received at journal
25 Jul, 2024
Reviewers agreed at journal
21 Jul, 2024
Reviewers invited by journal
08 Jul, 2024
Editor assigned by journal
04 Jul, 2024
Submission checks completed at journal
04 Jul, 2024
First submitted to journal
03 Jul, 2024

You are reading this latest preprint version

Glacier retreat induces contrasting shifts in bacterial biodiversity patterns in glacial lake water and sediments

Status:

Version 1

Abstract

Figures

INTRODUCTION

MATERIALS AND METHODS

Study area and sample collection

Estimation of glacial influence

Physicochemical analysis

DNA extraction and sequencing

Sequence analysis

Estimation of the stochastic ratio in community assembly

Network analysis

Statistical analyses

RESULTS

Bacterial taxonomy and alpha diversity patterns

Bacterial β-diversity patterns

Source of bacteria in KQGR lake ecosystems

The assembly of water and sediment communities

Co-occurrence network of water and sediment communities

DISCUSSION

Distinct patterns of bacterial α-diversity in lake water and sediments under glacial influence

Shifts in bacterial community composition with glacial influence

Linkages between bacterial diversity and co-occurrence network complexity

CONCLUSION

Declarations

References

Table

Additional Declarations

Supplementary Files

Status:

Version 1