High-throughput amplicon sequencing reveals the spatiotemporal effects, abiotic and biotic shaping factors for the microbial communities in tropical mangrove sediments in Sanya, China


 Background

Microbial roles in element cycling and nutrient providing are crucial for mangrove ecosystems and serve as important regulators for climate change in Earth ecosystem. However, some key information about the spatiotemporal influences and abiotic and biotic shaping factors for the microbial communities in mangrove sediments remains lacking.
Methods

In this work, 22 sediment samples were collected from multiple spatiotemporal dimensions, including three locations, two depths, and four seasons, and the bacterial, archaeal, and fungal community structures in these samples were studied using amplicon sequencing.
Results

The microbial community structures were varied in the samples from different depths and locations based on the results of LDA effect size analysis, principal coordinate analysis, the analysis of similarities, and permutational multivariate ANOVA. However, these microbial community structures were stable among the seasonal samples. Linear fitting models and Mantel test showed that among the 13 environmental factors measured in this study, the sediment particle size (PS) was the key abiotic shaping factor for the bacterial, archaeal, or fungal community structure. Besides PS, salinity and humidity were also significant impact factors according to the canonical correlation analysis (p ≤ 0.05). Co-occurrence networks demonstrated that the bacteria assigned into phyla Ignavibacteriae, Proteobacteria, Bacteroidetes, Chloroflexi, and Acidobacteria were the key biotic factors for shaping the bacterial community in mangrove sediments.
Conclusions

This work showed the variability on spatial dimensions and the stability on temporal dimension for the bacterial, archaeal, or fungal microbial community structure, indicating that the tropical mangrove sediments are versatile but stable environments. PS served as the key abiotic factor could indirectly participate in material circulation in mangroves by influencing microbial community structures, along with salinity and humidity. The bacteria as key biotic factors were found with the abilities of photosynthesis, polysaccharide degradation, or nitrogen fixation, which were potential indicators for monitoring mangrove health, as well as crucial participants in the storage of mangrove blue carbons and mitigation of climate warming. This study expanded the knowledge of mangroves for the spatiotemporal variation, distribution, and regulation of the microbial community structures, thus further elucidating the microbial roles in mangrove management and climate regulation.


Introduction
Mangroves are productive wetland ecosystems in the world and occur in the intertidal zones on tropical and subtropical coast [1,2]. Mangroves have attracted much research interest because of the various ecological functions of mangrove ecosystems, including runoff and ood prevention, storage and recycling of nutrients and wastes, cultivation, energy conversion, and their great signi cance on science and education [3]. Mangroves act as the interface of land, ocean, and atmosphere and are the centers for the ow of energy and matter among these ecosystems. Mangroves, as typical blue-carbon ecosystems, store considerable amounts of carbon (C) into the sediments, thus making them important regulators for climate change [4]. Therefore, mangroves are essential in the Earth ecosystem.
Microorganisms in mangroves are important. Universal studies on the structure and diversity patterns of microorganisms in mangrove sediments and leaves, including the communities of bacteria [5][6][7], archaea [7,8], and fungi [9][10][11], have shown that the microorganisms in mangroves are important element cyclers and nutrient providers, thus elucidating the microbial roles for mangrove ecological functions. As we know, comprehensive information for the biotic and abiotic processes across space and time needs to be established to fully understand ocean ecosystems [12]. However, this key knowledge of mangrove microbial communities remains lacking. Speci cally, the existing unilateral investigations on limited spatial and temporal dimensions cannot reveal the comprehensive effects of spatiotemporal variations on the microbial communities in mangroves [13,14]. The key abiotic and biotic factors that shape the mangrove microbial communities also need to be explored. These questions should be addressed to further understand the assembly of mangrove microbial communities, the roles of mangrove microorganisms in climate regulation, and the management of mangrove forests [15,16].
In this study, we investigated three kinds of microbial communities, namely, bacteria, archaea, and fungi, in the sediments from tropic mangroves in Sanya City, along with 13 abiotic environmental factors from different spatial (two locations and two depths) and temporal (four seasons) dimensions. We aimed to determine the laws of spatiotemporal effects on microbial variations and reveal the key abiotic and biotic factors that shape the microbial communities, including bacteria, archaea, and fungi, in mangrove sediments.

Descriptions of sediment samples and abiotic environmental factors
Bulk sediment samples were collected from the tropic mangroves located in Yalong Bay (YLW), Sanya River (SYH), and Tielu Bay (TLG) in Sanya City, China (Fig. 1A). Totally, 22 sediment samples were obtained from two depths (surface: 0-2 cm and subsurface: 12-14 cm), three locations (YLW, SYH, and TLG), and four seasons (March, June, September, and December). However, two samples from YLW in December were missing, because the sampling path was blocked by the municipal engineering unit. The sample name was formatted as "location_season_depth". For example, the name "YLW_6_surf" indicates that the sample was collected from the surface sediments of YLW mangroves in June, while "TLG_9_sub" indicates that the sample was collected from the subsurface sediments of TLG mangroves in September.

Compositions of the microbial communities
Proteobacteria was the dominant phylum for bacterial community in all samples, and the phyla Bacteroidetes, Chloro exi, Cyanobacteria, Planctomycetes, Acidobacteria, Firmicutes, Verrucomicrobia, Nitrospirae, Latescibacteria, Spirochaetae, Actinobacteria, and Gemmatimonadetes were detected in all sediment samples (Additional le 1: Fig. S2). Classes that belonged to phylum Proteobacteria were separately analyzed. Except for the YLW samples in June and September, delta-and gammaproteobacteria classes were dominant in other samples (Fig. 1B).

Spatiotemporal Effects On The Alpha Diversities Of Microbial Communities
The alpha diversities of bacterial, archaeal, and fungal communities were analyzed. For the bacterial community (Additional le 1: Fig. S5), the depth signi cantly affected (analysis of variance [ANOVA], p ≤ 0.05) the diversity (based on Shannon and Simpson indices) and abundance (based on ACE and Chao indices). Signi cant difference was not found in the samples from different locations. However, for archaeal community (Additional le 1: Fig. S6), signi cant difference was observed in the alpha diversity of the samples obtained from different locations (ANOVA, p ≤ 0.05) but in different depths. For fungal community (Additional le 1: Fig. S7), location and depth could signi cantly affect the abundance indices (ACE and Chao indices, ANOVA, p ≤ 0.05) but not the diversity indices (Shannon and Simpson indices). Signi cant difference was uniformly not detected in seasonal samples in the bacterial, archaeal, or fungal community.

Spatiotemporal Effects On The Structures Of Microbial Communities
According to the principal coordinate analysis (PCoA) and the 95% con dence ellipses in the results, the location and depth have signi cant effect on microbial communities, including bacteria, archaea, and fungi ( Fig. 2A). The structural patterns of bacterial and fungal communities from surface and subsurface sediments were obviously different. However, depth effect on archaeal community was not found (Fig. 2B). Seasonal impact was not detected in the bacterial, archaeal, or fungal community (Fig. 2C).
LDA effect size (LEfSe) analysis yielded similar results ( Fig. 3 and Additional le 1: Fig. S7). Totally, 11 and 21 bacterial phyla with signi cant differences were detected in the samples from different locations and depths (LDA ≥ 2.5), respectively, and only four bacterial phyla had signi cant difference in the samples from different seasons (LDA ≥ 2.5). Correspondingly, differential phyla belonging to archaea and fungi were found with LDA values ≥ 2.5 from location and depth samples. However, no differential species on phylum level was found in archaeal and fungal communities from seasonal samples.
Results from the analysis of similarities (ANOSIM) and permutational multivariate ANOVA (Adonis) supported the signi cant difference among the samples from different locations (p ≤ 0.05, Additional le 2: Table S2) in the bacterial, archaeal, or fungal community. ANOSIM and Adonis also supported that the depth could signi cantly affect bacterial and fungal communities (p ≤ 0.05, Additional le 2: Table S2). Similar with the PCoA result, no signi cant difference was found in the archaeal community from the depth samples based on the results of ANOSIM and Adonis (Additional le 2: Table S2). Moreover, no seasonal difference was found in bacterial, archaeal, and fungal communities (Additional le 2: Table  S2).

Abiotic Factor That Shape The Microbial Communities
Four abiotic factors with variance in ation factors of more than 10, including TN, NN, TC, and MMT, were removed because of their high collinearity with the other factors (Additional le 1: Table S3). Subsequently, canonical correlation analysis (CCA) was performed to assess the effects of abiotic environmental factors on the microbial communities (Fig. 4A). SAL, HUM, and PS were the abiotic factors that signi cantly affected the bacterial, archaeal, and fungal community (p ≤ 0.05, Additional le 3: Table  S4). Additionally, AN, TP, OC, and pH were the signi cant abiotic factors for bacterial community, and TEMP could signi cantly affect the archaeal community. For fungal community, TP, OC, and pH were signi cant factors (Additional le 3: Table S4).
To investigate the key abiotic factors that shaped the microbial communities, we individually tted the signi cant factors from CCA results (Additional le 3: Table S4) with the PC1 values from principal component analysis (PCA) by using the linear models. The factors possessed the highest R 2 value in the tting models were identi ed as the key biotic factors for the microbial communities. Uniformly, PS had the highest R 2 value in the bacterial, archaeal, or fungal community ( Fig. 4B and Additional le 1: Fig. S8).
Mantel test was performed to con rm this result, and this analysis supported that PS served as the key abiotic factor for the microbial communities. Results showed that among the abiotic environmental factors collected, PS possessed the highest R 2 value and a signi cant p value (p ≤ 0.05) in the bacterial, archaeal, or fungal community (Table 1).

Biotic Factors That Shape The Bacterial Communities
Co-occurrence networks from microbial communities were used to assess the keystone taxa that shaped the microbial communities in mangrove sediments. For bacterial community, networks were built for three locations ( Fig. 5A for SYH, 5B for TLG, and 5C for YLW). Based on these networks, the values of amongmodule connectivities (Pi) and within-module connectivities (Zi) were calculated to assess the keystone taxa (Zi > 2.5 or/and Pi > 0.62) for each location.
Results showed that ve keystone taxa (operational taxonomic unit [OTU]5273, OTU5037, OTU5104, OTU5013, and OTU12246) were found in SYH samples, 12 keystone taxa (OTU11764, OTU11887, OTU12355, OTU11459, OTU11559, OTU6475, OTU11996, OTU7966, OTU10701, OTU16075, OTU12101, and OTU11086) were found in TLG samples, and only one OTU was assessed as keystone taxon (OTU2963) in YLW samples (Fig. 5D). These keystone taxa were assigned into phyla Ignavibacteriae, Proteobacteria, Bacteroidetes, Chloro exi, and Acidobacteria (Table 2). However, we did not obtain any keystone taxon information for archaeal and fungal communities from the data. Inconsistency over the comparison of in uence degree on the bacterial community in mangrove sediments between marine and terrestrial environments Phyla Cyanobacteria and Chloro exi [17] and class gamma-Proteobacteria [18] are typical bacterial communities in marine environments, and phyla Actinobacteria, Acidobacteria, and Verrucomicrobia [19][20][21] and classes alpha-and beta-Proteobacteria [22] are mainly found in terrestrial environments. The relative abundances of these communities could assess the in uence degree on the bacterial community in mangrove sediments from marine and terrestrial environments [23].
Mangrove eco-systems, as typical intertidal wetlands, could be affected by the land and ocean. Marine environment has strong effect on the bacterial and viral communities in mangrove sediments, indicating the higher in uence from oceans than that from land [7,24]. For 16 out of 22 samples in this study, the relative abundances of marine bacteria were greater than those from land, thus supporting former studies ( Fig. 1B and Additional le 1: S2). However, for the six remaining samples, the relative abundances of terrestrial communities were greater than those from oceans (Figs. 1B and Additional le 1: S2). This result is inconsistent with previous research, indicating that the comparison of in uence degree between oceans and land could be more variable than previously observed. For example, anthropogenic factors can complicate this variation [23].
Microbial community structures in mangrove sediments showed spatial variability and temporal stability Location affected the alpha diversities of archaeal (Additional le 1: Fig. S5) and fungal communities (Additional le 1: Fig. S6), and ANOSIM, Adonis (Additional le 2: Table S2), LEfSe (Fig. 3A), and PCoA  Table  S2). To avoid the effects of rhizosphere in mangroves [25], we only collected the bulk sediments away from the plant roots. However, the location and depth were still the signi cant spatial dimensions that shaped the microbial communities in mangrove sediments, suggesting that the distribution of microbial communities in mangrove sediments was highly territorialized in a fairly close sampling range (e.g., in the same city for sampling).
Our work failed to detect signi cant differences among the seasonal samples. The in uence of spatial dimensions supports previous studies [7,26,27]. However, the weak effect from temporal dimension, namely, seasonal change, is unusual, especially for fungi with dynamic community structure in mangrove sediments [28]. We attribute this difference to the stable abiotic factors, except SAL and PS (Fig. S1), among the seasonal samples from the tropics, as previously reported [7].
Generally, our work demonstrated the variability in spatial dimensions and stability in temporal dimension of the microbial communities in tropical mangroves in Sanya City. This characteristic not only ensures the diversity of microbial functions in the different areas of mangrove sediments, but also guarantees the stability of the microbial functions against time variation, indicating a versatile but stable environment in tropical mangrove sediments.
PS served as the key abiotic factor for shaping the microbial communities in mangrove sediments Limited studies focus on the key abiotic factors that shape the microbial communities in mangrove sediments. Among at least 13 abiotic factors that we collected in this work, PS was the key abiotic shaping factor for the bacterial, archaeal, or fungal community according to the results of linear tting models (Fig. 4B, Additional le 1: Figs S8 and S10) and Mantel test ( Table 2).
To the best of the authors' knowledge, this study rst demonstrated PS as a key shaping factor for the microbial communities in mangrove sediments. The in uence of PS on mangrove microbial diversity has been reported before, which showed that PS could affect the diversity and abundance of laccase-like bacteria, and their abundance decrease in the order of sand > clay > silt in Mai Po mangrove sediments of China [29]. Our work enhanced the value of PS for the assembly of not only the laccase-like bacteria but also three typical microbial communities in mangrove sediments.
The possible reasons of the PS in uence on microbial structures are as follows: (i) High nutritional content is present in the soils or sediments with speci c PS [30,31]. The difference in nutritional contents resulted in the variation in the structures of microorganisms. (ii) Soils or sediments with small PS could provide microbes a protective habit from predators [32,33]. Therefore, high abundance of microbes is often detected in samples with small PS. (iii) The crucial enzymes, such as urease, invertase, alkaline phosphatase, and xylanase, for bacterial survival are rich in sediments with speci c PS [34][35][36][37] and could cause the abundance and diversity of microbes in the present samples. (iv) Permeability in uenced microbial structures as con rmed by Probandt et al. [38], and PS is the sole variable that affects the permeability calculation of soils or sediments [39]. Thus, PS could affect the microbial structure by controlling the permeability of sediment samples.
In addition to PS, SAL and HUM were also uniform signi cant abiotic factors for bacterial, archaeal, and fungal communities (Fig. 4A and Additional le 3: Table S4). Considering the element cycling functions of these microorganisms in mangrove sediments, the above factors might indirectly in uence carbon, nitrogen, and sulfur cycles in mangrove sediments by affecting the microbial community structures.
Providers of bioavailable carbohydrates and ammonia nitrogen served as the key biotic factor for shaping the bacterial communities in mangrove sediments Keystone taxa are known as "the engineers" for shaping the community structures in nature because of their high connectivity with other species in the co-occurrence structures. Therefore, the keystone taxa are crucial biotic factors shaping the microbial communities in mangrove sediments. We failed to obtain any information about archaeal and fungal communities because of the possible weak dependency among these microorganisms and the possible interactions of archaeal and fungal communities with other living organisms rather than the creatures of their own kinds. Unlike archaea and fungi, keystone taxa were found in bacterial community. Results showed that the OTUs belonging to phyla Ignavibacteriae, Proteobacteria, Bacteroidetes, Chloro exi, and Acidobacteria served as the bacterial keystone taxa in Sanya mangrove sediments (Table 2).
Polysaccharides, including cellulose and seaweed polysaccharide, are important storage forms of organic C in mangrove sediments [57,58]. However, recalcitrant polysaccharides are di cult to utilize for many microorganisms. Mangroves can also store considerable dissolved inorganic carbons that cannot be utilized by many microorganisms [57]. Therefore, photo-autotrophic and polysaccharide-degrading bacteria can server as the keystone abiotic factors by providing bioavailable carbohydrates to other bacteria that interacted with them, as shown in the results of this work. Mangrove ecosystems also have a typical N-de cient environment [59]. Therefore, the keystone OTUs with N-xation ability in the present study can provide bioavailable ammonia nitrogen for the survival of other microorganisms and living organisms in mangrove sediments. Hence, these keystone species could serve as the centers for bacterial interaction in mangrove sediments because of the nutrition that they provide.
These keystone species drive the microbial structure and function [60] and predict changes in microbial community [61]; hence, their status, especially for bifunctional species that can x both C and N, is important not only for themselves but also for many other bacteria that depend on them. Therefore, the changes of community structure of these keystone species provided in this work were potential important indicators for supervising the health of mangrove ecosystems.
Important contributions of the bacterial modules centered on photoautotrophic keystone taxa in mangrove sediments to the global climate The abundances of keystone species were not dominant in natural environments and in our work, but these taxa are the centers of co-occurrence structures in bacterial community, suggesting that they could organize many bacterial species together and form considerable interaction modules [60]. These bacterial modules might contribute to the ecological functions of mangroves, which are typical blue carbon ecosystems [4].
These bacterial co-occurrence modules centered on photoautotrophic keystone taxa could import inorganic CO 2 into the sediments and support the survival of heterotrophic bacteria in the modules to form and store large amounts of organic carbons in microorganisms. Furthermore, the carbon xed in mangrove sediments could be exported into the oceans [57], while some could be sealed up as recalcitrant dissolved organic matters in oceans for millions of years by the marine microbial carbon pump [62]. N xation is another key function for some of these keystone taxa. Ammonia nitrogen, the product of N xation, is an essential nutrient for living organisms, including primary producers, and nitrate, the product of nitri cation of ammonia nitrogen, is crucial for the photosynthesis of phytoplankton [63]. The keystone taxa with N xation ability could further increase the amount of C xation by providing the limited nitrogenous nutrients for photosynthesis. In brief, these bacterial modules centered on photoautotrophic keystone taxa could be the additional C xers and storage sites besides the large phytobenthos in mangroves and serve as potential vital participants in the storage of blue carbons and mitigation of climate warming.
Our work aimed to reveal the spatiotemporal in uence and the key abiotic and biotic shaping factors for the microbial communities in mangrove sediments through amplicon sequencing and to re ect the structural information of mangrove microorganisms. However, we did not research and discuss the functional information about these communities. Hence, metagenomic and meta-transcriptomic sequencing should be conducted to describe the microbial communities for future works in mangrove ecosystems.

Conclusions
This study focused on determining the spatiotemporal in uence and the abiotic and biotic shaping factors for the microbial communities in Sanya tropical mangrove sediments. Results showed variability on spatial dimensions (e.g., locations and depths) and stability on temporal dimension (seasons) of bacterial, archaeal, or fungal communities, which ensured the diversity of microbial functions in the different areas of mangrove sediments and guaranteed the stability of the microbial functions against time variation, indicating that the tropical mangrove sediments are versatile but stable environments.
PS was the key abiotic shaping factor for the bacterial, archaeal, or fungal communities, suggesting that PS along with other signi cant in uence factors, such as SAL and HUM, could indirectly participate in the material circulation in mangroves by in uencing microbial community structures.
Furthermore, our work found that photoautotrophic, N-xing, and polysaccharide-degrading bacterial keystone taxa served as the key biotic factors for shaping the bacterial community in mangrove sediments. Hence, the providers of bioavailable carbohydrates and ammonia nitrogen are essential for shaping the bacterial communities in mangrove sediments. These keystone taxa are the potential indicators for monitoring the health of mangroves, and the bacterial modules centered on these photoautotrophic taxa are also crucial participants in the storage of mangrove blue carbons and the mitigation of climate warming.
Our study expanded the knowledge of mangrove ecosystems for the change, distribution, and regulation of the microbial community structure in sediments, thus further elucidating the microbial roles in mangrove management and climate change mitigation.

Methods
Sampling and physicochemical analysis Sediment samples were collected from Yalong Bay Qingmei, Sanya River, and Linwang Tielu Mangrove Nature Reserve in Sanya, China (Fig. 1A). Approximately 500 g of bulk sediment sample was collected and placed into sterilized centrifuge tubes at each site and stored in dry ice. Samples were stored at − 80 °C in an ultra-low-temperature freezer.
The SAL of sediments was determined using a handheld SAL meter (ATAGO, Japan). Sample HUM was de ned as the percentage of the weight of water in the sample and that of wet sample. The MMT and MMP of the month in which the samples were obtained was queried at https://www.wunderground.com/history/. Laser particle size analyzer Nicomp 380 (PSS, USA) was used to determine sediment PS. The pH and contents of TN, TP, TC, AN, NN, and OC were measured by Qingdao Hengli Testing Co., Ltd. (China) following the national standards of China. Sediment TEMP was measured using an alcohol thermometer in situ. DNA Extraction, High-throughput Sequencing, And Data Processing DNA extraction was performed with PowerSoil DNA isolation kit (MO BIO, USA). For each sample, 0.5 g of sediments was used for DNA extraction, and the procedures were strictly operated in three replicates following the kit instructions. The repeated DNA from the same sample was pooled to avoid extraction bias. DNA quality was determined by 1% agarose gel electrophoresis and Nanodrop 2000 (Thermo, USA).
Subsequently, 515F and 806R primers [64] with barcodes were used to amplify the 16S rRNA genes of bacteria and archaea, and 817F and 1196R primers [65] with barcodes were used to amplify the 18S Raw fastq les were quality-ltered using Trimmomatic and merged using FLASH based on the following criteria: (i) The reads were truncated at any site that receives an average quality score < 20 over a 50 bp sliding window. (ii) Sequences with overlap of longer than 10 bp were merged according to their overlap with mismatch no more than 2 bp. (iii) The sequences of each sample were separated according to barcodes (exactly matching) and Primers (allowing 2 nucleotide mismatching), and reads containing ambiguous bases were removed.

Statistical analysis
Alpha diversity indices were calculated using the mothur software (version v.1.30.1 http://www.mothur.org/wiki/Schloss_SOP#Alpha_diversity) and were visualized using the ggpubr package of the R software. The bar graphs of bacterial, archaeal, and fungal communities were generated in the R software in accordance with the result of Qiime processing. PCoA was conducted using the "cmdscale()" function in R software based on the Bray-Curtis distance, and the graphs were generated using the ggpubr package of R. CCA was used to evaluate the in uences of environmental factors on the bacterial and archaeal diversity patterns. The "cca()" function in the ade4 package of R was used to execute this analysis. The "vif.cca()" function in R was used to reveal the VIF values of abiotic environmental factors, and the factors with VIFs of more than 10 were removed until all the VIF values of abiotic factors were less than 10. PCA was conducted using the "principal()" function in the psych package of R, and the PC1 values were obtained from the PCA results. Linear equation tting was used to determine the most important shaping factors for the microbial communities by using the "lm()" function with the "singular.ok = FALSE" parameter to void a singular t. The factors with VIFs of less than 10 were tted separately by using the PC1 values from PCA results. Base on the p and R 2 values, the key factors were con rmed. LEfSe analysis was performed in http://huttenhower.sph.harvard.edu/galaxy/.
The keystone taxa were analyzed in http://129.15.40.240/mena/, and the values of Zi and Pi were calculated with the default parameters. The OTUs with Zi > 2.5 or/and Pi > 0.62 were identi ed as the keystone taxa [66]. The bacterial co-occurrent networks were visualized using the Cytoscape software. To display the relationship among the environmental factors, we analyzed and visualized the correlation coe cient and signi cance by using the "corrgram()" function in the corrgram package of R with the Pearson method. ANOSIM, Adonis, and Mantel test were analyzed using the "anosim()", "adonis()", and "mantel.rtest()" functions in R with the Bray-Curtis distance and 999 permutation tests.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
Raw sequencing data in this study were deposited in the NCBI SRA database under accession no.

Competing interests
The authors declare that they have no competing interests.   Sampling locations (A) in Sanya mangroves and the bacterial community composition from Proteobacteria phylum (B). YLW, Yalong Bay; SYH, Sanya River; and TLG, Tielu Bay. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 2
Principal coordinate analysis (PCoA) results of the microbial communities in mangrove sediments from multiple spatiotemporal dimensions. A-C showed the PCoA results in the samples from different locations (A), depths (B), and seasons (C), respectively. According to the 95% con dence ellipses, the spatial dimensions, locations, and depths, could signi cantly affect the bacterial, archaeal, and fungal community structures (p ≤ 0.05). However, seasons as temporal dimensions have no signi cant in uence on any microbial community in mangrove sediments. and seasons (C). Taxa on the phylum level with signi cantly different relative abundances were found in the bacterial, archaeal, and fungal community structures from the spatial samples. However, no phylum with signi cant different relative abundance was found in the temporal samples (see Fig. S7 for detailed legend information).