Background

Saline lakes are found all over the world and play important roles in global biogeochemical cycling [1, 22] was used to excise bases with a mass value below 20 at the tail of the reads, and a window of 10 bp was set if the mean mass value within the window was below 20. The back-end bases were cut off from the window, sequences containing N and short sequences were filtered after quality control, and sequences with low complexity were filtered out. The FASTA files were de-replicated, abundance sorted, and singleton sequences were removed. The OTUs (using a 97% similarity cut-off) of archaea were clustered de novo using USEARCH (v.11.0.667) [23]. Mothur (1.43.0) [24] was used to determine the alpha diversity index. The Redundancy, Venn, Mantel-test, and network analysis of different samples were performed using R (v.4.3.0) software. Archaeal metabolism functional prediction was conducted using PICRUSt (v.1.1.4) [25, 26]. The ecological functional prediction of the archaeal communities was conducted using FAPROTAX (v1.2.1) [27].

Results

Differences in archaeal diversity, taxonomic compositions, and potential ecological functions

The overall number of archaeal sequences in Qinghai saline lake and Chaka hypersaline lake was 233,097 and 179,359, which were clustered into 791 and 428 operational taxonomic units (OTUs, at a 97% cut-off), respectively. As Table 1 shown, the richness (Chao index) of archaea in Qinghai saline lake (in an average of 532) was significantly higher (Wilcox.test, p = 0.004 < 0.05) than that in Chaka hypersaline lake (in an average of 352), suggesting more archaeal species were adaptable to the saline lake, but sensitive to the hypersaline environments. However, the alpha diversity (both Shannon and Simpson Index) in Chaka hypersaline lake showed no significant difference (Wilcox.test, p = 0.823 and 0.123 > 0.05) with that of Qinghai saline lake, demonstrating a stable and highly adapted archaea community existed in the hypersaline lake. Using Principal coordinate analysis (PCoA) with Bray-Curtis distance, we compared the beta diversity of archaeal communities in different lakes. Based on our results, the first two principal components explained 53.9% and 11.5% of the total variation, respectively. Furthermore, the PERMANOVA (p = 0.002 < 0.005) results and PCoA ordination plots indicated that there were significant differences in the composition of archaea between the two lakes (Fig. S1).

Table 1 Statistical analysis of archaea alpha diversity in different habitats
Full size table

At the phyla level, the relative abundance of archaeal sequences in Qinghai saline lake was dominated by Woesearchaeota (43.94%), followed by Euryarchaeota (29.16%) and Thaumarchaeota (4.8%) (Fig. 1a). By contrast, nearly all the sequences (95.74%) in Chaka hypersaline lake were annotated on Euryarchaeota. Except for Euryarchaeota, a few archaeal sequences in lake Chaka were also annotated on Nanohaloarchaeota (2.51%), Woesearchaeota (0.4%), and Thaumarchaeota (0.06%) (Fig. 1a). Notable, although phylum Euryarchaeota was dominant in both two lakes, at the genus level, the Euryarchaeota sequences in lake Qinghai and Chaka were mainly annotated on Methanothrix/ Methanocalculus/ Methanomassiliicoccus and Halorubrum/ Halohasta/ Halonotius/ Natronomonas, respectively (Fig. 1b). Thus, these evidences revealed archaea with markedly different taxonomic compositions in Qinghai saline lake and Chaka hypersaline lake.

Fig. 1
figure 1

Relative abundance of archaea composition at phyla (a) and genus level (b) in various samples collected from the two lakes

Full size image

Using the PICRUSt software, we predicted the metabolic functions of archaea based on the COG database. In the Chaka hypersaline lake, archaeal metabolism was advantaged by ‘replication, recombination and repair’ and ‘inorganic ion transport metabolism’, which were some functions to tackle the extremely high-salinity environment. Comparatively, archaea in the Qinghai saline lake highly in ‘energy production and conversion’ and ‘translation, ribosomal structure and biogenesis’ (Fig. 2a). Furthermore, we utilized FAPROTAX to predict the ecological functions of the archaea communities in different habitats. Archaea in Qinghai saline lake showed a higher abundance of ‘methanogenesis’, ‘hydrogenotrophic methanogenesis’, and ‘methanogenesis by CO2 reduction with H2’ functions. In contrast, the archaea predominantly displayed ‘nitrate reduction’, ‘chemoheterotrophy’, and ‘aerobic chemoheterotrophy’ functionality in Chaka hypersaline lake (Fig. 2b).

Fig. 2
figure 2

Functional annotation results of archaea in Qinghai saline lake and Chaka hypersaline lake. (a) metabolic pathways annotation (b) ecological function annotation

Full size image

Impact of the environmental factors on archaeal communities

Following the exclusion with high inflation factors (> 10) (such as pH, NO3− and DO), redundancy analysis (RDA) was used to understand the difference in the archaeal community structure of the two habitats and identify the significant environmental factors regulating their community structure. The RDA plot clearly demonstrated that the archaeal communities in the saline lake and hypersaline lake were substantially distinct (Fig. 3a). Salinity, Chlorophyll a (Chl_a), total nitrogen (TN), and temperature were identified as the most critical elements influencing the development of archaeal communities. Further variance partitioning analyses (VPA) confirmed that salinity largely impacted the LP community structure variation (Fig. 3b). But the contributions of Chl_a, temperature, and nutrient (TP + TN + PO43−) to the archaeal communities’ variation were also unneglectable. In fact, archaeal variation explained by salinity combined temperature (representing 34% of the variation), Chl_a (representing 26% of the variation), or nutrient (representing 21% of the variation) was largely higher than the variation single explained by the salinity (represent 13% of the variation).

Fig. 3
figure 3

Archaeal communities’ variations across different habitats. (a) RDA analysis of archaea communities (b) VPA analysis of archaea communities

Full size image

The Mantel Test was used to examine the relationships between environmental conditions and the archaeal phyla. Consistent with prior findings, although salinity was found to significantly affect a lot of phyla, such as Euryarchaeota and Crenarchaeota, but it was not the only important factor affecting these phyla. Especially for the Euryarchaeota, which was one of the most important phyla to contribute to archaeal communities in both Chaka hypersaline and Qinghai saline lakes, DO, pH, and PO4− were also found to have a large effect on them (Fig. 4).

Fig. 4
figure 4

Mantel test for the effects of environmental variables on different phyla

Full size image

Archaeal interactions in different habitats and their response to environmental factors

Co-occurrence network analysis of the dominant archaeal OTUs (Top 50) was used to infer their interactions in different habitats. In Qinghai saline lake, the total nodes and correlation number of the network were 51 and 118, respectively (Fig. 5a). Surprisingly, the correlation between archaeal dominant OTUs in Qinghai saline lake was entirely positive, indicating the wide co-presence of the dominant archaea species. The number of the total nodes and correlation number of Chaka hypersaline lake were 53 and 120, respectively, which was not statistically different from those of Qinghai saline lake (Fig. 5b). However, some negative correlations between different dominant archaeal OTUs were found in Chaka hypersaline lake (∼ 8% of the total dominant archaeal OTUs’ correlations), representing the mutual exclusion existing in dominant archaeal OTUs under the extremely high salt environment.

Fig. 5
figure 5

Dominant archaeal co-occurrence networks in Qinghai saline (a) and Chaka hypersaline (b) lakes. Positive and negative correlations were colored red and blue, respectively

Full size image

Among the correlation relationships identified between dominant archaeal OTUs and environmental factors, there were also significant differences between the two habitats (Fig. 5). In the Qinghai saline lake, dominant archaeal OTUs were mainly correlated with salinity, total phosphorus (TP), and Chlorophyll a, but in the Chaka hypersaline lake, dominant archaeal OTUs were associated with DO, TN, pH, and PO43−. Notable, although salinity was observed with correlation with dominant archaeal OTUs in both two habitats, their correlations’ positive/negative relationship was different. Specifically, the relationships between salinity and dominant archaeal OTUs were all positive in the network of Qinghai saline lake, but all negative in the network of Chaka hyper-saline lake. Moreover, in each habitat, the importance of some nutrient factors was higher than that of salinity. Especially, the number of correlations between dominant archaeal OTUs and phosphorus (TP/PO43−) was equal to and exceeded that of salinity in the Qinghai and Chaka lakes, respectively.

Discussion

In this study, the results of high-throughput sequencing revealed that the predominant phylum within the archaeal communities in Chaka hypersaline lake was Euryarchaeota (Fig. 1a). This observation was consistent with those of previous studies on the dominant archaeal phylum in the Great Salt Lake [28] and the Dead Sea [29]. Genus-level annotation further revealed Euryarchaeota in Chaka hypersaline lake was dominated by Halorubrum, Halohasta, Halonotius and Natronomonas, which was consistent with the archaeal community composition in high-altitude Andean lakes [30]. Based on culturable and genomic analysis, previous studies uncovered many Natronomonas and Halorubrum strains (e.g. Natronomonas pharaonis DSM 2160, and Halorubrum lacusprofundi ATCC 49,239) with the ability to reduce NO3− and NO2− assimilatorily [31, 32]. Interestingly, nitrate-reduction function was also found abundant in the hypersaline lake (Fig. 2b). Thus, these lines of evidence suggested archaea played important role in the nitrogen cycling of Chaka hypersaline lake. In the Qinghai saline lake, a previous culturable-dependent study supposed Euryarchaeota and Woesearchaeota were the first and second dominant phyla [33], respectively, which was different from the results of this study. Clearly, the most likely explanation for that variance was the difference in study methods. As a relatively new member of the superphylum DPANN (Diapherotrites, Parvarchaeota, Aenigmarchaeota, Nanoarchaeota, and Nanohaloarchaea), Woesearchaeota was surprisingly diverse and abundant in a wide range of extreme environments, such as deep oil reservoir, oligo-trophic lakes and indicating a high diversity of their roles in global biogeochemical cycles [34, 35]. Recently, multivariate regression analysis further revealed that Woesearchaeota might function in consortium with methanogens in the cycling of carbon [36]. Interestingly, our results also detected the wide co-presence of Woesearchaeota and methanogenesis-functional Euryarchaeota (Fig. 5a), which supported the previous observation. Furthermore, in the Qinghai saline lake, methanogenesis-related functions and taxa were distinctive characteristics of the archaeal annotation. Especially, the dominant genera in Qinghai saline lake were Methanothrix (average relative abundance > 8.30%) (Fig. 1b), which was widely distributed in both natural and artificial anoxic environments and played a major role in global methane production [37, 38]. Previous research has shown that some methanogenic archaea are halotolerant and can function with a salinity up to 20 g/L [39]. Thus, the appropriate salinity level (in an average of 16.5) and other environmental factors of Qinghai saline lake may suitable for the growth of these methanogenic archaea and allow them to gain some advantage in the communities. Considering the importance of methane in the greenhouse effect [40], the contribution of Qinghai saline lake to the greenhouse effect is worth further evaluating.

In this study, both network, RDA and Mantel-test results suggested the single influence of salinity on the structuring of archaeal communities was limited, which distinguished from prior study results that focused on the variation of bacterioplankton communities in different saline lakes [12]. This could be owing to the fact that many archaea have the ability to ‘live with salt’ [9]. Instead, the composite effect of salinity with diverse environmental parameters (e.g., temperature, chlorophyll a, total nitrogen, and total phosphorus) dominated the explanation of the variations in archaeal community structure. In oceanic habitats, several previous studies have suggested the structure of archaeal communities was significantly influenced by temperature, nitrogen, and chlorophyll-a concentrations [41, 42]. While in some freshwater lakes, a growing body of evidence suggested that phosphate or total phosphorus can essentially alter the structure of archaeal communities [43, 44]. The influence of environmental factors on archaea in saline lakes seems to be a combination of the two habitats mentioned above, but the specific regulatory effects of these factors and their mechanisms still need further analysis, especially based on culture-dependent studies and in situ investigations.

Although a lot of environmental factors have been considered, nearly 40% of the constrained variance of archaeal communities in different habitats was still unexplained, highlighting the limitations of employing environmental factors to explain community shifts. Currently, it has been reported that community structure predictions considering biotic information are more accurate than those based on only environmental factors [15, 12]. According to our findings, the above theory may also apply to archaea communities.

Conclusions

Using high-throughput sequencing, this study revealed the significantly different communities and potential ecological functions of archaea between Qinghai saline lake and Chaka hypersaline lake. Based on RDA, VPA, Mantel-test, and network analysis, we further suggested the differences of archaeal communities between saline and hypersaline lakes were driven by the combination of multiple environmental and non-environmental factors (archaeal interactions). Overall, this research improved our understanding of the structure and ecological role of archaea in saline lakes and provided an update on the mechanisms that shape their communities.