Author response:
The following is the authors’ response to the original reviews.
eLife Assessment:
This study provides a valuable genome-centric characterization of microbial communities across deep sediment cores from a Spartina patens salt marsh. The study provides claims on the metabolic capabilities of the deep sediment microbiome as well as on a burial microbial assembly process and functional complementarity at depth. However, some of these claims remain incomplete and would benefit from further supporting evidence. Overall, this work will be of interest to microbial ecologists working on wetlands.
We appreciate all of the efforts of the reviewers and editors. Additional analysis and extensive edits were made to the original manuscript to address the comments of both reviewers and editor. We believe this effort has significantly improved the manuscript.
Public Reviews:
Reviewer #1 (Public review):
Summary:
In this manuscript, Vineis et al. examined the structure and functional potential of microbial communities along a vertical sediment profile of a salt marsh, using a genome-centric metagenomic approach. They attempted to test whether (1) the microbial communities within dynamic upper layers contain genomes with diverse functional potential, (2) the energy limited deeper sediments contain microbial consortia assembled to metabolise complex carbon, and (3) microbial compositional changes in the low energy sediments mirror the burial processes observed in marine environments with similar energetic limitations. Results revealed a core microbial consortia that contains a collective metabolic potential for complex carbon and aromatics degradation, suggesting putative syntrophic interactions. Besides, the recovery of MAGs assembled independently from multiple depths in the same core and the consistent relative abundance structure of MAGs within co-occurrence network modules together suggest burial process as a likely mechanism for microbial assembly.
Strengths:
(1) Two long sediment cores (down to 240 cm deep) were collected in this study, allowing investigation of the less well characterised subsurface microbiome in salt marsh.
(2) A genome-centric metagenomic approach was employed here, which provides information on both the structure and functional potential of the salt marsh sediment microbiome, which is not possible in commonly performed 16S rRNA-based surveys.
Weaknesses:
(1) In both the abstract and conclusion, the authors claimed that results from this study provide a "mechanistic understanding" of the assembly and distribution of the microbial communities in salt marsh sediment (P2, L31 and P35, L645-649). However, both claims are speculative and not supported by solid evidence. Firstly, the genomic data presented in this study and supplementary physical properties of sediments in the broader area are not enough to make a solid claim (that appears in the title) on microbial assembly being governed by a burial process. Alternative explanations include residual bioturbation, slow porewater advection, etc. Therefore, this remains an interesting hypothesis unless additional evidence is provided to rule out the alternative explanations. Similarly, the claim on the detailed syntrophic interactions among members within a co-occurrence network module (e.g. P36, L649-652) is purely speculative and warrants functional validation experiments to prove.
The reviewer makes two points in section 1 of their review, and we have addressed each point as follows: 1) We have removed the term “mechanistic understanding” from the manuscript and instead focused on the co-occurring group of microbes and the evidence for overall community and population structure with depth. Statistical analyses were added to the manuscript including a test for the influence of depth on the overall community structure (Fig. 3) and sample-specific single nucleotide variants (SNVs) within a collection of MAGs that were abundant and prevalent throughout the subsurface (Fig. 7). The text now contains additional text in the discussion (lines 702-720 and 843-845) that addresses alternative explanations of our observations including the potential for advection and residual bioturbation. 2) We agree that to prove syntrophy, functional validation through experimentation is required and we have added a statement to this effect on line 858-864.
(2) A major aim of this work was to study complex carbon degradation. However, neither CAZymes, the first-line carbon degradation enzymes, nor peptidases, which can be important contributors to carbon degradation at depth, was examined here. METABOLIC, which the authors used for functional annotation of MAGs, by default generates peptidases outputs and can be easily integrated here.
We expanded our analysis to include CAZymes and peptidases including an analysis of the number of peptidase pathways among our network modules (Table S3, Fig. S6), and a supplemental table of glycoside hydrolase and polysaccharide lyase (Table S2). Methods regarding the CAZyme and peptidase analysis are included on lines 282-296, results on 490-509, and 581-588.
(3) No geochemical data is available to provide context for the genomic analysis here. Without such information, readers cannot even tell whether the surface sediment samples were oxic or anoxic. A reference to a PhD thesis is provided (P6, L126) but it would be most helpful to extract relevant data from there and provide as a supplementary table.
Oxygen concentrations were not measured as part of that study. Although oxygen concentrations are heterogeneous, due to bioturbation, and radial oxygen loss from roots in marsh systems similar to ours they are generally below detection at the shallowest depth examined in our study. We have added additional citations relevant to this point on lines 747-753.
(4) A single metagenomic binning tool, CONCOCT, was used in this study, which very likely has resulted in a limited number of MAGs recovered. More (high-quality) MAGs are expected with the use of additional binners and a bin consolidation procedure.
We agree that multiple binning tools can potentially identify more bins, contigs can be mistakenly binned when using purely automated processes. To sidestep the chance of including erroneous contigs in our collection, we chose to invest a large amount of time and effort required to bin manually. CONCOCT was used as an initial guide to identify some of the most easily reconstructed MAGs but all contigs within MAGs were subjected to visual inspection of the sequencing coverage profile over several samples, sequence composition congruency, and real-time completion and contamination estimation as contigs were manually added and/or subtracted from the collection in each bin. This approach is not more widely applied because of the expertise and time required to manually reconstruct MAGs from each sample individually. Many researchers find that human involvement is often required to improve the accuracy of automated binning tools, which has given rise to several tools in addition to Anvi’o, BinaRena, ICoVeR, and ggKBase. To elaborate on this procedure, we made a video tutorial of our approach to binning manual-binning-approach. An example of the manual bin refining process can be found here (merenlab-MAG-refinement)
(5) Several terminologies are misleading here. Firstly, the term "co-occurring" or "co-located" microbes or MAGs (e.g. P1, L19 and P31, L537) can be misleading as it could imply a close spatial relationship. However, co-occurrence networks rely on correlations of (relative) abundance and show statistical associations instead of direct spatial or physical relationships. I would suggest alternative names such as co-abundant or statistically associated microbes.
We have added text to address the important point raised by the reviewer regarding the term “cooccurrence”. We added additional language in the methods section to define co-occurrence and to make clear that the term should not be interpreted to infer a direct spatial or physical relationship (lines 121-123). However, the term “cooccurrence” is commonly used to describe the significant correlations identified molecular ecological networks and we think the introduction of additional terminology such as “statistically associated” could lead to additional confusion.
Secondly, the term "persistent conversion of soil organic carbon" (P36, L654) in the conclusion is also misleading as it implies an active process, which cannot be tested without metatranscriptomics or metaproteomics data.
We agree and thus we added metatranscriptomic data to the manuscript to assess the activity of the microbes in the subnetwork of commonly co-occurring MAGs within the subsurface. Our results indicate that the pathways for complex carbon are both present and active at the time of sampling. The methods section describing this additional analysis is included on lines 155-168 and 314-324 and we created a new figure to communicate the present and active MAGs in the Bathyarchaeia BA1 subnetwork (Fig. 7). Even with the additional metatranscriptome data, we agree that the term “persistent conversion of soil organic carbon” is still not supported by our data because we sampled at a single timepoint and we have removed this text from the manuscript.
(6) Based on a NMDS plot of KEGG IDs (Figure 4B), the authors claimed that the functional potential among MAGs in modules 1, 2 and 7 was very similar (P18, L346). However, the dispersions of modules 1 and 2 were just too large. A proper statistical test, such as PERMANOVA, should be used to support the claim.
We replaced the analysis of the KEGG modules with a more detailed characterization of the functions identified in the MAGs within each of the modules (Fig. 5) (including the CAZy and peptidase analysis described above) (Table S2 and S4). This revised approach provides more resolution on the higher-order functions that are differentially detected among the modules. This allows for greater clarity on the functions that are specific to the MAGs within each module.
(7) Genome-scale metabolic networks was analysed using Metag2Metabo (M2M) and results were discussed in detail (P26, L453-466). However, the source data should be provided in a supplementary table to show what metabolites are producible by which MAGs.
The M2M analysis generates a map of thousands of genes and reactions for each MAG which is stored in an XML-formatted file. Files for each MAG are available upon request so the metabolic pathways can be explored in Fluxer (https://fluxer.umbc.edu/) or Escher (https://escher.github.io/). The metabolites produced through complimentary reactions within the Bathyarchaeia BA1 subnetwork are included in Table S6 and the complimentary gene content for the Benzoyl-CoA pathway is shown in Table S7.
Reviewer #2 (Public review):
This work provides a detailed metabolic reconstruction of sediment microbiomes along a depth profile in a Spartina patens salt marsh in Massachusetts, USA. Using a combination of genome reconstruction, co-occurrence network analysis, and metabolic profiling, the authors describe the metabolic potential of co-occurring microbial consortia in understudied deep sediments.
Major strengths of this study include the detailed metagenomic characterization of the understudied deep marsh sediments. The authors recovered genomes representing a substantial portion of the deep sediment microbiome (up to ~60%) and provided an initial explanation of pathways related to the potential for organic carbon decomposition in this environment. Of particular interest is the capability of the deep sediment microbiome to process aromatic organic compounds, highlighting the need for a collaborative consortium to carry out their decomposition. Improved understanding of the microbial transformation of deep sediment organic carbon in blue carbon ecosystems is vital to better understand the fate of this large carbon pool in the face of climate change.
However, I have a few concerns in the interpretation of the results, and in the case of the surface sediments there is a lack of strong evidence in my opinion.
(1) A stronger ecological interpretation is needed regarding the meaning of the co-occurrence network analysis. The authors correctly note that their analysis identifies groups of co-occurring genomes, which may indicate shared niche space, not necessarily interspecific ecological interactions (as the authors imply for instance in lines 423-425). When performing network analysis using samples from the entire sediment profile (0-240 cm), they identified consortia that co-vary in relative abundance along the depth gradient most likely because of shared environmental filtering forces, such as changes in redox potential and sediment chemistry. Supplementary Figure S4 showing that different modules have distinct abundance distributions along the sediment profile supports this idea. Being that the case, I would like the authors to define the ecological significance of the "connector hub". Is it merely taxa that is prevalent in the whole sediment profile? Since the modules are physically separated (in different sediment depth layers), they are not really interacting between each other. As it stands, it is not clear why the authors decide to study connector hubs in greater detail, along with their subnetworks.
These are all very good points raised by the reviewer and we have taken the following steps to clarify the role of connector nodes. Additional text regarding the environmental role of a connector node is included on lines 273-278, 440-442. We added additional text regarding our decision to study the Bathyarchaeia BA1 MAGs that were classified as connectors in greater detail (lines 514-534).
(2) I question if the lack of network modules in the surface sediment is really a consequence of non-significant interspecific ecological interactions and not the result of methodological biases. The low MAG recovery and thus short read recruitment in surface-level metagenomes may hinder the ability of the authors to identify co-varying microorganisms in the surface sediment. The high diversity of the surface sediment prevents proper assembly of the surface microbiome. I would also argue that as redox potential declines sharply in salt marsh sediments just below the root surface, the microbial community in the first few centimeter's changes rapidly and is significantly different from the more stable deep sediment microbiome. Due to the sampling design, the study has less representation of the surface layer (only 0-30 cm, while the cores extend down to 240 cm). Grouping sediment microbiomes by depth based on similarity in their sequence space (e.g., Mash) or functional profile (e.g., KEGG annotation) before performing network analysis could help to better infer ecological relationships within the distinct ecological niches of the marsh sediment profile, rather than performing a single network analysis of all samples combined.
We agree with rationale that the low MAG recovery from surface sediments inhibits our ability to build a reliable network that would capture interactions within this region of the sediment profile. While the reviewer’s idea of grouping samples based on functional similarity or depth is and interesting idea, this decreases the number of samples used for estimating co-occurrence. Because we had few MAGs derived from surface sediments, it is unlikely that this recommended approach would yield additional connections. Additional analysis of surface sediments is certainly warranted in order to tease apart the interactions and their relevance to the biogeochemistry of the sediment. The reviewer comments may be useful to others attempting to identifying niche spaces and connections in the surface. See lines 775-783.
(3) Normalizing the relative abundance of MAGs by dividing by the total reads mapping to a particular sample can be misleading due to differences in recruitment levels across samples (and depths). A better approach would be to normalize by metagenome library size, or preferably by genome equivalents (e.g., using MicrobeCensus) or a similar approach.
We decided to normalize to the number of reads mapped to our non-redundant collection of MAGs instead of to the total number of sequences in each sample for several reasons. 1.) We do not know what the proportion of unmapped reads represents and could include virus and eukaryotic life. Comparing the amount of microbial DNA to the eukaryotic fraction can be misleading because changes in the relative abundance from these sources would skew the microbial relative abundance. In human microbiome studies, subtraction of human reads is often carried out prior to calculating relative abundance, but we lack any such reference in our system. We added text to explain this in the methods (lines 219-232). 2.) We are not comparing the total abundance of MAGs across samples or comparing individual genes across samples, so tools such as MicrobeCensus would not offer an improved analysis. Additionally, many of these tools rely on reference genomes within the workflow which is challenging in these sediments due to the many novel taxa. There are challenges with any relative abundance calculation and clear explanation of our read mapping strategy and Fig. S2 allows the reader to see the potential bias.
Recommendations for the authors:
Reviewing Editor Comments:
It appears that the main issue is that some of the claims in the paper are not adequately justified. You can find the detailed assessment of the two reviewers in their reviews, and their recommendations are listed below. We encourage you to examine these comments and incorporate the recommendations in your manuscript. In particular, after consulting the reviewers, we encourage you to prioritize the following:
(1) Addressing alternative hypotheses.
(2) Improving MAG binning and assembly.
(3) Revising statistical analyses.
(4) Including CAZymes and peptidases in the analysis.
(5) Revising the terminology and interpretation (e.g. regarding co-occurrence).
Reviewer #1 (Recommendations for the authors):
(1) Title: consider shortening it.
We agree and shortened the title
(2) P3, L49-52: incomplete sentence.
The corrected sentence in on lines 45-47.
(3) P3, L54: consider adding a citation for "blue carbon stocks" for a broad readership.
We added a citation. Line 54.
(4) P7, panel A: consider adding coordinate axes; panel B: would be helpful to add more details on the tools (e.g. CONCOCT for binning), key parameters (e.g. dereplication at 95% ANI), and/or key statistics used (e.g. the number of high-quality and medium-quality MAGs).
We added additional details to figure 1.
(5) P8, L131 (and L148): define MAG at its first use. Are they bins with >50% completeness and <10 contamination MAGs?
MAGs are first defined on line 24. Bins are a minimum of 50% complete and maximum 10% redundant (lines 200-205).
(6) P9, L168: the default setting for dRep uses -comp 75, which won't result in MAGs with a completeness <75%; however, some MAGs in Table S1 have a completeness <75%.
The methods on this point are clarified on lines 191-206.
(7) P10, L173: by the "75% threshold", do you mean for completeness/completion?
We have clarified that sentence to specify “completion” on line 203.
(8) P10, L175: provide the version number for GTDB used.
The version is now provided in the text on line 205.
(9) P10, L185: information in the GitHub repository is incomplete; e.g. the code for running CONCOCT binning is missing.
The GitHub is now up to date including a link to a video explaining how binning and MAG refining were conducted for one of the samples.
(10) P14, L268: how many MAGs are novel at each taxonomic ranks? Table S1 does not include genus level data.
We have included additional text regarding the novelty of MAGs on line 370-375. Table S1 now has genus-level classification.
(11) P14, L271, 274, and 276: please provide the exact numbers and/or percentages.
The percentages are now reported on lines 372-375.
(12) P14, L272: however, Figure 2 only shows phylum data.
We have reworked this paragraph and corrected this error (lines 384-406)
(13) P14, L281: deep instead of shallow?
The paragraph was rewritten and this point now appears on lines 392-394.
(14) P16, L298-302: would be helpful to provide a supplementary table showing these values.
These values were added to table S1.
(15) P16, L314: aren't they 30% and 50%, respectively?
This error was corrected. See lines 378-379.
(16) P18, L339: five Dehalococcoidales and no GIF9 MAGs?
We revised this section and the details are now included on lines 438-440.
(17) P21, L385: module 5 instead of 4?
This section was revised and the corrected details can be found on lines 460-462.
(18) P21, L386: Figure S4 is irrelevant here.
This section was revised and the corrected details can be found on lines 460-461.
(19) P22, L394: Figure 5 instead?
This section was removed during revision
(20) P24, L413: to me, these MAGs were present along the entire core instead.
The section regarding “bimodal” MAGs was removed during revision.
(21) P24, L411: there are multiple typos in the scientific names in Figure S6.
We have removed this supplemental.
(22) P25, L430: Fig. 6A instead?
The correction of this error is on line 525 and references figure S7A.
(23) P25, L440: module 1?
The GIF9 classification was replaced with the family-level classification AB-539-J10 and we have corrected any error assigning these MAGs to the incorrect module. The original sentence containing this error was revised, but the section including these results can be found on lines 514-534.
(24) P28, L501: these pathways were present in not even half of the MAGs.
The details of selenate and arsenate are now presented in figure 5.
(25) P30, L508: also provide a key for the size of nodes.
The original figure 6 is now figure S7 and was revised to include a key for genome size in the network.
(26) P31, L553: please provide a supplementary table showing the metabolic potential of each MAG to support this claim.
We created figure 5 to address this comment in addition to table S9. The associated text can be found on lines 739-745.
(27) P32, L564: I would suggest changing the word "consistency" to "correlated pattern" for clarity.
We are trying to describe the correlated pattern as being consistent across multiple samples. See lines 730-732 for the revised use of this term.
(28) P32, L567: also Bacteroidales.
We have clarified the genomic content of MAGs containing benzoyl-CoA reductase subunits and production of benzoyl-CoA from phenol on lines 630-635
(29) P32, L575: please tone down this statement; some MAGs are capable of both (Figure 6C).
We have toned down the statement and discussion of the metabolic handoffs and syntrophy. See lines 801-817.
(30) P33, L582-585: aromatics are not products of fermentation so their decomposition doesn't "follow" the initial fermentation.
We revised the statement on lines 815-817.
Reviewer #2 (Recommendations for the authors):
(1) Performing co-assemblies could help improve the recovery of microorganisms from the surface sediment layer.
We have responded to this comment in response to reviewer 1, point #4
We agree that multiple binning tools can potentially identify more bins, contigs can be mistakenly binned when using purely automated processes. To sidestep the chance of including erroneous contigs in our collection, we chose to invest a large amount of time and effort required to bin manually. CONCOCT was used as an initial guide to identify some of the most easily reconstructed MAGs but all contigs within MAGs were subjected to visual inspection of the sequencing coverage profile over several samples, sequence composition congruency, and real-time completion and contamination estimation as contigs were manually added and/or subtracted from the collection in each bin. This approach is not more widely applied because of the expertise and time required to manually reconstruct MAGs from each sample individually. Many researchers find that human involvement is often required to improve the accuracy of automated binning tools, which has given rise to several tools in addition to Anvi’o, BinaRena, ICoVeR, and ggKBase. To elaborate on this procedure, we made a video tutorial of our approach to binning manual-binning-approach. An example of the manual bin refining process can be found here (merenlab-MAG-refinement)
(2) Performing diversity analysis of metabolic pathways in surface vs deep sediments could help prove the idea of more versatility in the surface marsh sediment.
To address this comment, we created a figure detailing the proportion of MAGs containing central microbial functions among the modules with distinct depth distributions (Fig. 5), analyzed the number of complete pathways (Fig. 4, Fig. S5) and differences in CAZyme and peptidase genes (Fig. S6).
(3) To strengthen the idea that burial is key in deep marine sediments, looking at the abundance of mobility structures (i.e., flagella) along the sediment profile could help make a case.
This is an interesting idea, however, many of the same genes used for motility are also used to form biofilms. This analysis could be an interesting aspect of these sediments that warrants an independent study.
(4) Lines 60-63, but oxygen leaked from macrophytes could also exacerbate microbial respiration rates.
This point is clarified on lines 59-64.
(5) Line 49-52: Check this sentence. It does not read properly. "2this" -> "this".
We corrected this error on lines 45-47
(6) Line 173: I assume >75% is only for completion, not contamination, right? I would rephrase to make this clear.
This point is clarified on lines 201-206.
(7) Line 179, what is Bacteria_71? Citation is needed.
A citation has been added for this single-copy gene collection on line 209.
(8) Line 201: When performing log transformations, how are 0s, which are highly abundant in microbiome data, handled?
The details of matrix transformation are clarified on line 252-254. However, unlike amplicon data (ASVs and OTUs), the read recruitment data contained very few zeros in the dataset.
(9) Line 417: The Pseudolabrys MAG most likely utilizes the dsr and apr genes in reverse, as is typical for other Alphaproteobacteria, not for sulfate reduction. Additionally, sat can also function in assimilatory sulfate reduction.
Yes, we agree with the reviewer’s assessment of Pseudolabrys. However, during revision, the section regarding MAGs with bimodal distribution was removed.
(10) Line 414 uppercase mags.
During revision, the section containing this error was removed.
(11) Line 605: "examined".
This error was corrected on line 682.
(12) Line 610: Are independent MAGs from adjacent sediment layers more similar to each other than those from layers that are farther apart? Demonstrating such a pattern could provide stronger evidence that burial acts as a major force shaping the assembly of the deep sediment microbiome.
We conducted a community-level analysis based on MAG relative abundance (Fig. 3) and a population analysis of SNVs for six abundant MAGs to test for an effect of depth (Fig. 8). While we no longer focus our analysis on a mechanistic explanation of the depth-dependent patterns, with the exception of lines 769-772. As requested by the reviewers we discuss the observed patterns in light of several other potential mechanisms (lines 838-845)
(13) While acknowledging that selection pressure following burial is a major driver of microbial community assembly in marine sediments, I would ask whether other biogeomorphic forces could also influence the assembly of sediment microbiomes over millennial timescales, such as bioturbation or sediment redistribution within the salt marsh environment.
Based on our community and population-level analysis, we suggest that bioturbation and sediment redistribution are likely to shape the upper 40 cm. Beyond this depth, we did not find evidence for major restructuring events. This topic is discussed on lines 830-845, but the formation of distinct relative abundance peaks for the modules primarily found in deeper sediments remains an outstanding question (lines 772-774).
