3.1. Archaeal community composition and diversity
Archaeal α-diversity (Shannon and Chao1 indices) varied among different soil types (Fig. S1, p < 0.05). Generally, soils with medium pH (6.0 < pH < 7.0) showed higher archaeal α-diversity than soils with pH over 7.5. Archaeal community composition also differed between the different sites (Fig. S1).Crenarchaeota was the most abundant in most samples at the phylum level, accounting for 20% to 70% of total archaea, followed byHalobacterota (1% to 70%) as the second most abundant phylum (Fig. S2). The relative abundance of Euryarchaeota andThermoplasmatota was 2%–33% and 5%–20%, respectively.Asgardarchaeota and Micrarchaeota were also detected in some samples with lower relative abundance. Bathyarchaiea andNitrososphaeria were the two most dominant classes in most samples at the class level (Fig. 2a). However, the relative abundance ofBathyarchaeia differed drastically among different sites, ranging from 3% to 68% of total archaea. Methanosarcinia andMethanobacteria were also detected at relatively high abundance in most samples, which belong to Halobacterota andEuryarchaeota , respectively.
3.2. Distribution and diversity ofBathyarchaeia in paddy soils
The α-diversity of Bathyarchaeia in various paddy soils exhibited significant differences among distinct paddy soils, with the Shannon index ranging from 2.16 to 5.63 (refer to Fig. S3, p < 0.05). Notably, soils characterized by a pH > 7.5 displayed a reduced Bathyarchaeial α-diversity when compared to soils with a pH < 7.5. Furthermore, the NMDS analysis indicated thatBathyarchaeia within paddy soils clustered distinctly based on the sampling sites and pH (see Fig. S3).
In the aggregate, five subgroups (Bathy-6 , Bathy-11 ,Bathy-15 , Bathy-17, and Bathy-5bb ) were identified within all the paddy soils (Fig. 2b). Bathy-6 constituted the predominant subgroup across most samples, accounting for 70% to 80% of the total Bathyarchaeia (as depicted in Fig. 2c). Conversely,Bathy-11 , Bathy-17, and Bathy-5bb were detected as less prevalent groups in numerous samples. It is noteworthy that paddy soils in Southeast China exhibited a higher relative abundance ofBathy-6 in comparison to their counterparts in Northwest China (Fig. S4).
The heatmap, based on the 40 most abundant Bathyarchaeial ASVs, was generated with a focus on soil pH (as demonstrated in Fig. 2d). Out of the 40 most abundant Bathyarchaeial ASVs, 36 ASVs were affiliated with Bathy-6 , while three ASVs were associated with Bathy-15 , and one ASV was linked toBathy-17 . Furthermore, soil pH had a substantial impact on the distribution of these dominant ASVs (see Fig. 2d). Notably, two ASVs (ASV1970 and ASV1998) affiliated with Bathy-6 were present at elevated levels in all the samples. Nevertheless, most other ASVs were solely detected in samples with acidic soil conditions (pH < 7.0).
3. 3. Bathyarchaeial community assembly in paddy soils
The NCM results at ASV levels indicated that stochastic processes were dominant in archaea community assembly, with an R2value of 0.613 (Fig. 3a). We found that 86.2% of archaeal ASVs and 87.0% of Bathyarchaeial ASVs fitted to the neutral model, indicating that a majority of archaeal ASVs assembled in paddy soils following stochastic processes. However, the neutral model fitted ASVs accounted for only 40.9% and 47.0% relative abundance of total archaeal and Bathyarchaeial community, respectively. Approximately 12.6% of Bathyarchaeial ASVs deviated from the neutral expectation, accounting for more than 45.0% relative abundance. These results indicate that most low-abundance Bathyarchaeia assembled following stochastic processes, whereas high-abundanceBathyarchaeia mainly assembled following deterministic processes. This phenomenon can be further supported by the assembly process of the highest abundant Bathy-6 (Fig. 2), exhibiting a higher percentage of the deterministic process (about 38.5%) than total archaea (13.8%) and total Bathyarchaeia (13%, Fig. 3b).
Among the 13% of Bathyarchaeial ASVs that diverged from neutral expectations, 12.6% were more frequently observed than anticipated by the NCM prediction, indicating that they exceeded the neutral prediction. ASVs exceeding the prediction are believed to possess greater migratory capabilities and a heightened ability to relocate to new habitats. Notably, Bathy-6 , as previously mentioned, was the most prevalent Bathyarchaeial subgroup and exhibited a higher proportion of ASVs surpassing the NCM prediction compared to other archaea andBathyarchaeia . This suggests that Bathy-6 may have more robust migratory abilities and experience less constraint on dispersal than other subgroups. The breadth of ecological niches at the class level, as determined by the null model, was computed to assess the diversity of resource utilization among Bathyarchaeia . The outcomes demonstrated that the niche width of Bathyarchaeia in paddy soils surpassed that of other archaea, indicating a greater diversity in resource utilization among Bathyarchaeia in paddy soils (as visualized in Fig. 3c).