Co-occurrence pattern and function prediction of bacterial community in Karst cave

Abstract

Background: Karst caves are considered as extreme environments with nutrition deficiency, darkness, and oxygen deprivation, and they are also the sources of biodiversity and metabolic pathways. Microorganisms are usually involved in the formation and maintenance of the cave system through various metabolic activities, and are indicators of changes environment influenced by human. Zhijin cave is a typical Karst cave and attracts tourists in China. However, the bacterial diversity and composition of the Karst cave are still unclear. The present study aims to reveal the bacterial diversity and composition in the cave and the potential impact of tourism activities, and better understand the roles and co-occurrence pattern of the bacterial community in the extreme cave habitats.

Results: The bacterial community consisted of the major Proteobacteria, Actinobacteria, and Firmicutes, with Proteobacteria being the predominant phylum in the rock, soil, and stalactite samples. Compositions and specialized bacterial phyla of the bacterial communities were different among different sample types. The highest diversity index was found in the rock samples with a Shannon index of 4.71. Overall, Zhijin cave has relatively lower diversity than that in natural caves. The prediction of function showed that various enzymes, including ribulose-bisphosphate carboxylase, 4-hydroxybutyryl-CoA dehydratase, nitrogenase NifH, and Nitrite reductase, involved in carbon and nitrogen cycles were detected in Zhijin cave. Additionally, the modularity indices of all co-occurrence network were greater than 0.40 and the species interactions were complex across different sample types. Co-occurring positive interactions in the bacteria groups in different phyla were also observed.

Conclusion: These results uncovered that the oligotrophic Zhijin cave maintains the bacterial communities with the diverse metabolic pathways, interdependent and cooperative co-existence patterns. Moreover, as a hotspot for tourism, the composition and diversity of bacterial community are influenced by tourism activities. These afford new insights for further exploring the adaptation of bacteria to extreme environments and the conservation of cave ecosystem.

Keywords: 16S rRNA gene; Bacterial community; Co-occurrence network; Function prediction; Karst; Oligotrophy; Tourism; Zhijin cave.

Fig. 1

Bacterial community composition. a Relative abundances of the 10 most abundant phyla in each sample. The relative abundance not shown in chart if fewer than 4%. b Non-metric multidimensional scaling (NMDS) of bacterial community in three sample types

Fig. 2

The results of LEfSe analysis. a Cladograms indicating the phylogenetic distribution of bacterial lineages associated with the samples. b Indicator bacterial group significantly differentiated across the three sample types with LDA values higher than 3

Fig. 3

The PICRUSt predicted function in samples. a Predicted function of bacteria among the three sample types. b The second level of KEGG pathway was shown in the heatmap

Fig. 4

The co-occurring network analysis of the bacterial communities across the three sample types. The nodes are colored by phylum level, the size of each node is proportional to the relative abundance of specific genus level. The color of each edge is positive and negative of correlation coefficient, grey represents positive correlation, and red represents negative correlation. The thickness of each edge is proportional to the correlation coefficient (Spearman’s r > ±0.8 and P < 0.01)

Fig. 5

Distribution of sampling sites. A, J, S, T, W, X, Y represent different sampling sites