Contrasting archaeal and bacterial community assembly processes and the importance of rare taxa along a depth gradient in shallow coastal sediments

32 Marine microbial communities assemble along a sediment depth gradient and are 33 responsible for processing organic matter. The high-resolution mechanisms of the 34 vertical assembly processes in marine sediments remain poorly described. We analyzed 35 31 depth layers of 3 sediment cores from the shallow sediment zone at 3 stations in the 36 Bohai Sea, and obtained high-resolution vertical profiles (2 cm per sample) of microbial 37 communities. We analyzed 78 archaeal and 76 bacterial communities based on 16S 38 rRNA gene amplicon sequencing, together with 14 selected metagenomes. We grouped 39 these samples into three layers (Top, 0-18 cm; Middle, 18-38 cm; and Bottom, below 40 38 cm) to analyze trends in diversity and assembly processes along the depth gradient. 41 We found that alpha diversity increased for the Thaumarchaeota -dominated archaeal 42 community but decreased for the Proteobacteria -dominated bacterial community as 43 depth increased. The mechanisms determining archaeal community assembly were 44 mostly deterministic, while bacterial community assembly was mostly stochastic. Co- 45 occurrence networks among different taxa and key functional genes revealed a tight 46 community with low modularity in the bottom sediment, and disproportionately more 47 interactions among low abundance ASVs. This suggests a significant contribution to 48 community stabilization by rare taxa, and suggests that the bottom layer, rather than 49 surface sediments may represent a hotspot for benthic microbial interactions. 50 Importance: Marine sediments harbors a giant microbial diversity. To date, our 52 understanding of the mechanisms determines their distribution remain limited. We 53 identified different mechanism dominated the distribution of archaeal and bacterial 54 assemblies mostly may due to the metabolic plasticity in bacteria. We demonstrated an 55 increased microbial co-occurrence network with depth. We further stressed the 56 importance of rare taxa in the microbial community based on the co-occurrence 57 network of the relative abundance of different taxa and different key genes in marine 58 biogeochemical cycling. This work reveals the complex mechanisms determines the 59 distribution of microbial assemblies in coastal marine sediments.


61
Over 70% of the Earth's surface is covered by the ocean, and microbes in marine  The Bohai Sea is a semi-enclosed marginal sea of the Northwest Pacific. The 73 biogeochemistry in the Bohai Sea is complex. First, it is jointly influenced by terrestrial 74 inputs, e.g., the discharge from rivers (7), and the neighboring North Yellow Sea (8). 75 Second, the Bohai Sea has a high rate of sedimentation (over 10 mm•yr -1 ) (9). Third, 76 the Bohai Sea is characterized by clear seasonal variations in environmental conditions, 77 e.g., temperature, chlorophyll-a, and primary production (10). Moreover, recently 78 observed seasonal hypoxia events in the bottom waters of the Bohai Sea in summer (11,79 12) may increase buried organic matter due to insufficient degradation. Furthermore, 80 the redox gradients in the sediment change rapidly along the depths.

81
Extensive studies of microbial communities in the Bohai Sea have commonly 82 revealed a strong effect of anthropogenic activities on microbial distribution (13)(14)(15)(16). 83 The microbial composition in marine sediments is largely affected by both the abiotic 84 or environmental factors (17)(18)(19), such as the nutrient availability and redox gradients, 85 and the biotic factors, e.g., competition and predation. However, due to the complexity 86 of the biogeochemical conditions, the mechanism of prokaryotic distribution has not 87 been well studied in the Bohai Sea. Energy-limitation is usually a key factor in processes that determine their distribution. 100 We hypothesized that the vertical distributions of benthic prokaryotic 101 communities, which include both bacteria and archaea, are primarily affected by the 102 biogeochemical conditions rather than stochastic processes. In this study, we examined 103 the spatial distribution (from the surface to 62 cm below the surface maximum) of both 104 the bacterial and archaeal communities with high spatial resolution (2 cm for each 105 sample), together with samples selected for metagenome sequencing at three different 106 stations in the shallow sediment zones of the Bohai Sea. We assessed different 107 ecological processes governing the microbial community assembly. The co-occurrence 108 of different taxa and key functional genes also stressed the importance of rare taxa in 109 the microbial interactions driving element cycling in marine sediments.

111
We collected 3 sediment cores from 3 stations (M3, M8, and BHB10) in the Bohai Sea. to each other than those in the surface sediments at three stations (Fig. S1).

131
Alpha and beta diversity of bacterial and archaeal communities 132 Generally, the species richness of the archaeal community was lower than that of the 133 bacterial community (Fig. 1). Archaeal alpha diversity increased from top to bottom 134 sediments ( Fig. 1a and S2), especially for the top layer, while bacterial alpha diversity 135 decreased ( Fig. 1b  which mainly consisted of Deltaproteobacteria (26 ± 6 %), Gammaproteobacteria (18 160 ± 10 %), and Alphaproteobacteria (4 ± 1 %) (Fig. 2b). taxon, which contributed a total relative abundance over 1% in each domain, were only 177 present in one sample, except for Crenarchaeota and Pacearchaeota (Fig. 3c).

178
However, ASVs that occurred in only one sample had low relative abundance.

179
The dissimilarity of microbial communities between samples increased with 180 depth for both archaeal and bacterial assemblages in all three sites ( Fig. 3d and 3e), and 181 this trend was stronger in the archaeal community than in the bacterial community. To

215
Microbial co-occurrence networks 216 We retained 547, 470, and 620 archaeal and bacterial ASVs for the co-occurrence  Both archaeal and bacterial communities in the sediments were influenced by 308 different environmental factors, such as nitrite, ammonium, sulfate, and CO (Fig. S6).

309
However, there was no significant correlation between depth and the entire archaeal or  The cutoff of shown connections were P < 0.01 and |rho| > 0.7. 695 Figure 1 The archaeal and bacterial diversity in the Bohai Sea sediments at three stations (BHB10, M3, and M8) at different depths. (a) Shannon index of each sample for the archaeal community, (b) Shannon index of each sample for the bacterial community, (c) Shannon index for the archaeal community in different layers (top: 0-18 cm; middle: 18-38 cm; and bottom: below 38 cm), (d) Shannon index for the bacterial community in different layers. Signi cant differences between layers are marked by stars (***P < 0.001; *P < 0.05). Principal coordinate analysis (PCoA) of amplicon sequence variants (ASVs) abundance based on the Bray-Curtis dissimilarity for the archaeal (e) and bacterial (f) communities colored by depth in three stations (BHB10, M3, and M8) with 95% con dence.

Figure 2
Stacked bar chart showing the relative abundance of the archaeal taxa (a) and bacterial taxa (b) at different depths in the three stations in the Bohai Sea sediments.

Figure 3
The occurrence pattern of ASVs in different samples and the pattern of dissimilarity along with depth. The relationship, indicated by Spearman's rank correlation, between the relative abundance of each archaeal (a) and bacterial (b) ASVs and the occurrence across all the samples. (c) Occurrence patterns of ASVs belonging to the taxa in phylum level with relative abundance over 1% for archaeal or bacterial community. Numbers in the parentheses above the phylum name indicate the number of ASVs shown in the gure for each phylum. Numbers within the gure represent the total number of ASVs for each phylum, the number of ASVs occurring only in one 34 sample (not included in the gure for better visualization), and their corresponding relative abundance (percentage in parentheses). Relationships between the Bray-Curtis dissimilarities and the depth between each sample for archaeal (d) and bacterial (e) communities.

Figure 4
Ecological processes determining the archaeal and bacterial assemblies based on null model analysis with iCAMP in marine sediments. (a) The relative importance of different mechanisms governing the community in different layers. (b) Relative abundance of the top 15 abundant phylogenetic groups (bins), which account for a total relative abundance of 63.99% among the total 162 archaeal bins and 28.45% among the 599 bacterial bins. (c) Relative abundance of the classi ed genus with the greatest relative abundance in each bin. The relative importance of ecological processes across different phylogenetic groups in the top (d), middle (e), and bottom (f) layers.

Figure 5
Co-occurrence networks and Zi-Pi plot of the archaeal and bacterial ASVs in three different layers: (a, b, and c) top layer, (d, e, and f) middle layer, and (g, h, and i) bottom layer. The cutoff of shown connections were P < 0.01 and |rho| > 0.7. The size of each node is proportional to the relative abundance of each ASV (a, d, and g), and the number of connections (b, e, and h). The nodes were classi ed into four categories: (I) peripherals with Zi ≤ 2.5 and Pi ≤ 0.62; (II) connectors with Zi ≤ 2.5 and Pi > 0.62; (III) module hubs with Zi > 2.5 and Pi ≤ 0.62; and (IV) network hubs with Zi > 2.5 and Pi > 0.62. Numbers denote the ASV counts in each category (c, f, and i).

Figure 6
Correlation of the relative abundance of each archaeal and bacterial phylum and the relative abundance of functional genes involved in oxygen, nitrogen, and sulfur redox reactions. The cutoff of shown connections were P < 0.01 and |rho| > 0.7.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.