Distinct Soil Microbial Communities in habitats of differing soil water balance on the Tibetan Plateau

Abstract

Global change may be a severe threat to natural and agricultural systems, partly through its effects in altering soil biota and processes, due to changes in water balance. We studied the potential influence of changing soil water balance on soil biota by comparing existing sites along a natural water balance gradient in the Qinghai-Tibetan Plateau. In this study, the community structure of bacteria, archaea and eukaryotes differed between the different soil water conditions. Soil moisture was the strongest predictor of bacterial and eukaryotic community structure, whereas C/N ratio was the key factor predicting variation in the archaeal community. Bacterial and eukaryotic diversity was quite stable among different soil water availability, but archaeal diversity was dramatically different between the habitats. The auxotype of methanogens also varied significantly among different habitats. The co-varying soil properties among habitats shaped the community structure of soil microbes, with archaea being particularly sensitive in terms of community composition, diversity and functional groups. Bacterial and archaeal phylogenetic community turnover was mainly driven by deterministic processes while stochastic processes had stronger effects on eukaryotic phylogenetic community turnover. Our work provides insight into microbial community, functional group and phylogenetic turnover under different soil conditions in low-latitude alpine ecosystem.

Figure 1

Relative abundances of the dominant microbial phyla in all samples at Haibei Station (a) bacteria; (b) archaea; (c) eukaryotes). Relative abundances are based on the proportion of sequences that could be classified at the phylum level. Samples were clustered using UPGMA method based on OTU table using bray-curtis dissimilarity.

Figure 2

Microbial community (a) bacteria; (b) archaea; (c) eukaryotes) compositional structure in different habitats at Haibei Station calculated by non-metric multi-dimensional scaling (NMDS) using Bray-Curtis dissimilarity (▲ Alpine meadow; ● Marsh meadow; ■ Marsh). Environmental factors that had the most explanation to microbial community variation in DistLM analysis were fitted to the plots and color gradient indicate the value of the factor.

Figure 3

Microbial alpha diversity variation (a) phylogenetic diversity; (b) observed species; (c) Simpson evenness) across vegetation types. PD and OTUs variation were determined by dividing diversity value of each sample by that of the first sample in Alpine meadow. All Diversity indices were calculated using a random subset of 2800, 1110 and 640 sequences per sample, respectively. Error bar stands for standard error.

Figure 4

Regression of environmental distance and microbial dissimilarity (a) bacteria; (b) archaea; (c) eukaryotes). Environmental distance was calculated using euclidean dissimilarity and only factors that were significantly correlated to microbial community variation were chosen to perform the calculation (that is, SM and TN for bacteria; C/N, TC, DOC and TN for archaea; SM and TC for eukaryotes). Microbial dissimilarity was calculated using bray-curtis dissimilarity. Between: plots pair belonging to different habitats; within: plots pair in same habitats.

Figure 5

Microbial phylogenetic community turnover pattern tested by beta NTI. −2 and +2 were the ecological processes threshold for deterministic and stochastic force. Between: comparison of different habitats; withi: comparison in same habitats.