Influence of "cryoconcentration" on the composition of bacterial communities in semi-enclosed shallow water lakes

Abstract

Bacteria serve as vital indicators of the lake ecosystem health. Although substantial progress has been made in investigating the structural features of lake bacterial communities, limited attention has been paid to the dynamic assembly processes and driving factors affecting bacterial communities in ice and water environments during the freeze-up period. In this study, we investigated aggregation and compositional changes in bacterial communities in the internal ice-covered state of Lake Hulun. We examined the effects of cryoconcentration under low-temperature conditions on community assembly and systematically analyzed the physicochemical parameters as well as α- and β-diversity of bacterial communities in bottom ice (BI) and surface water (SW) media. Bacterial diversity was significantly higher in SW than in BI. Among the dominant taxa, eight phyla were shared between both environments. Firmicutes and Patescibacteria were dominant in the BI, whereas Gemmatimonadota and Bdellovibrionota were dominant in the SW. Nutrient transport driven by cryoconcentration emerged as a key factor influencing bacterial community assembly. Specifically, total nitrogen and salinity regulated the balance between stochastic and deterministic processes in BI and SW, respectively. Overall, the distinct environmental conditions of BI and SW weakened the diffusion capacity of bacterial communities, resulting in diffusion-limited and drift-dominated assembly processes. These findings offer new insights into the mechanisms underlying bacterial interactions and community assembly in ice-covered lake habitats and provide a scientific foundation for the management and preservation of lake ecosystems under ice-covered conditions.

Keywords: Assembly process; bacterial community structure; co-occurrence network; freezing process; ice-water interface.

FIGURE 1

(a) Location of the study area in China; (b) geographic location of Lake Hulun in Inner Mongolia, China; (c) 3D topographic map of the study area; (d) spatial distribution of sampling sites in Lake Hulun.

FIGURE 2

Contents of environmental indicators in BI and SW samples. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

FIGURE 3

(a) α-Diversity indices of bacterial communities in BI and SW; (b) principal coordinate analysis (PCoA) based on Bray–Curtis dissimilarity showing intergroup differences in bacterial communities, with ellipses indicating 95% confidence intervals.

FIGURE 4

(a,b) Relative abundances of bacterial communities at the phylum level in BI (a) and SW (b) samples; (c) significance testing of intergroup differences in bacterial phyla between BI and SW samples; (d,e) relative abundances of bacterial communities at the genus level in BI (d) and SW (e) samples; (f) significance testing of intergroup differences in bacterial genera between BI and SW samples. Only dominant phyla and genera are displayed; taxa with relative abundances < 1% are grouped into the category “others.” For (a,b) and (d,e), the left half-circle indicates the taxonomic composition at each sampling points, with different colors representing different sampling points and taxa in outer and inner rings, respectively; the right half-circle indicates the overall taxonomic composition in all BI or SW samples as well as the distribution of each taxon across different sampling points, with different colors representing different taxa and sampling points in outer and inner rings, respectively. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

FIGURE 5

(a) Linear discriminant analysis (LDA) illustrating differences in bacterial community composition between BI and SW samples. (b) LDA scores (threshold = 4). Higher LDA scores indicate a greater contribution of specific taxa to observed group differences.

FIGURE 6

(a,b) Co-occurrence networks of bacterial communities in BI (a) and SW (b). Each node represents a bacterial genus that is colored according to phylum affiliation. The node size is proportional to the node degree. Edges represent significant interactions between genera (p < 0.05), with green indicating positive correlations and purple indicating negative correlations. (c–e) Zi-Pi scatter plots showing intra-module connectivity (Zi) versus inter-module connectivity (Pi) of key taxa in the network. Highlighted taxa include Crenothrix, Elizabethkingia, BIyi10, Roseomonas, Lautropia, Nitrosomonadaceae, SWB02 (Proteobacteria), Planomicrobium (Firmicutes), and Brachybacterium (Actinobacteriota).

FIGURE 7

(a,b) Neutral community model results for bacterial assemblages in BI (a) and SW (b). The solid blue line indicates the model fit, and the dashed line represents the 95% confidence interval. Orange and green points denote the taxa frequencies above and below the model expectations, respectively. (c) Scatter plot of Raup–Crick (RC) and β-nearest taxon index (βNTI) values for BI and SW samples. (d,e) Relationships between βNTI values of bacterial communities and significant environmental factors in BI (d) and SW (e) samples. Only the environmental factors with statistically significant correlations are shown. (f) Relative contributions (%) of five ecological processes: homogeneous selection (HoS), heterogeneous selection (HeS), homogeneous dispersal (HD), dispersal limitation (DL), and drift (DR).