Interactions between temperature and energy supply drive microbial communities in hydrothermal sediment

Temperature and bioavailable energy control the distribution of life on Earth, and interact with each other due to the dependency of biological energy requirements on temperature. Here we analyze how temperature-energy interactions structure sediment microbial communities in two hydrothermally active areas of Guaymas Basin. Sites from one area experience advective input of thermogenically produced electron donors by seepage from deeper layers, whereas sites from the other area are diffusion-dominated and electron donor-depleted. In both locations, Archaea dominate at temperatures >45 °C and Bacteria at temperatures <10 °C. Yet, at the phylum level and below, there are clear differences. Hot seep sites have high proportions of typical hydrothermal vent and hot spring taxa. By contrast, high-temperature sites without seepage harbor mainly novel taxa belonging to phyla that are widespread in cold subseafloor sediment. Our results suggest that in hydrothermal sediments temperature determines domain-level dominance, whereas temperature-energy interactions structure microbial communities at the phylum-level and below.

In general, the interpretations of these results are a bit too strong and would benefit from a bit more nuance. For example, although the comparison to the non-seep area (NSA) is an important and valuable component of this study, it is still only an n=2 for making the case for a domain-level trend. One group of Archaea (Bathyarchaeota) dominates the non-seep areas while a different group of Archaea (Crenarchaeota) dominates the seep area. This is an exciting and intriguing finding, but the extrapolation from these two data points to a universal domain-level principle should be more careful. Understanding why these groups are dominant at high temperature requires investigating the biology of these specific taxa, and I'm not sure if ascribing these trends to an ancient, domain-level distinction is warranted with the present data or necessarily helpful for understanding this system.
With its extensive description of microbial taxa identified by 16S rRNA amplicon sequencing and its massive phylogenetic trees (in the supplementary materials), the paper reminds me, in a good way, of the classic exploratory studies of novel environments that characterized environmental microbiology in the early days of environmental DNA sequencing. The text is long and descriptive, and much of the discussion is speculative, although the speculations are interesting and wellinformed.
The supplementary materials, especially the Excel spreadsheets listing all OTUs and taxonomic classifications, are excellent and very useful companions to the paper. They will be very much appreciated by practitioners in this field.
Additional coments by line number: 52: "according to ecotype" seems to be referring to ecological context, which is not what an "ecotype" is. An ecotype is a species-like unit with a very specific definition based on one theory of speciation. 61: "ecotype" again. I think the authors mean "niche" or "habitat". 64: I'm confused why this sentence is constructed as "while...., less is known...". It seems to me that both halves of the sentence refer to slightly different aspects of the same thing, and they are not in any way opposed to each other.
151 Please clarify that sulfate concentrations decrease rapidly at some sites but not others 167 Please define SCOA here.
265 Please define ZOTU here. 431 Surely there are other potential explanations in addition to membrane stability? It seems too simplistic to discuss this deep biological mystery as though it must be explained by one of the two explanations presented here.
442 The phrase "controlled at the phylum level" seems to slip a little too far down the slope toward implying that selection is acting at the phylum level. All selection is at the species level (or below), of course, so domain-level patterns must be a legacy of ancient evolutionary events or an accident of taxonomic filtering. Either way, nothing is being currently "controlled" at the phylum level today, even if we can observe phylum-level trends today.
554 The grand conclusion "interactions between temperature and energy supply drive microbial community assembly at the phylum level and below" seems too strongly worded because I don't see any reason to focus on the phylum level in particular, as this could be just as easily a genus level pattern that propagates up to the phylum level when plotted that way. In addition, temperature and energy influence the distribution of all life on Earth (as stated in the first sentence of the abstract). The results of this study are consistent with that principle, which is great to see, but the writing style gives the reader the impression that these results have never been seen before.
Reviewer #2 (Remarks to the Author): The manuscript evaluates the correlation between various geochemical parameters (with a focus on temperature and potential substrate concentrations) and 16S rRNA gene diversity in advection vs diffusion dominated hot sediments. They find temperature correlates best with domain level abundance patterns and that available energy sources may explain community structure at lower taxonomic levels.
The manuscript is well written, clear, and is commended for including some potential caveats in the study and interpretations.
Major comments: 1) I could not find any representation of error or standard deviations on any of the plots. It is hard to interpret what is a significant difference between concentrations versus replicate or measurement error. If error is smaller than the symbol size, please indicate this in figure captions.
2) Similarly, a minimum quantification limit for the qPCR is mentioned, but I could not find this value depicted on the figures or mentioned in the text.
3) Since 16S is always a not a single copy gene, how are differences in copy number between taxa taken into consideration in these comparisons? 4) Since the geochemistry text is mostly descriptive, bringing more of the microbio supplemental figures to the main text would help keep the focus on microbial communities (which the title suggests is the main focus). For example, the description of new taxonomic groups is more significant/impactful than downcore plots of geochemistry and should be brought into the main text. Table 1 and geochemisty-only figures -> supplement Figure S1 -> main text, with the addition of panels with seafloor photos that exemplify each site type. Figure S7 and S8 (or key portions of?) -> main text Figure S12 -> main text 5) The manuscript could benefit from an overview schematic of dominant processes and taxa across the compared systems to tie the discussion together.
Minor comments: 1) Figure 1 -What is the significance of the line across the bottom, right panel?
2) The reader is asked to remember a lot of acronyms specific to this paper/site. It may be worth the extra words in some cases in order to reduce the number of acronyms.
3) Figure S9 -color data points by source as in Figure S10.
Reviewer #3 (Remarks to the Author): The authors compared microbial communities in shallow advection-and diffusion-dominated sediments in Guaymas Basin, Gulf of California. The communities were compared based on 16S rRNA copy number (bacteria vs archaea) and sequence identify (zero-radius operational taxonomic units, ZOTU). The authors also reported geochemical data describing the sediment cores (temperature, temperature gradients, concentrations of inorganic carbon, total organic carbon, total nitrogen, short chain organic acids, sulfate and sulfide). They found that archaea dominate above 45 C and bacteria below 10; at the phylum level high-temperature sites dominated by diffusion have taxa common to cold sediments and hot sites with advection dominating the transport regime have hydrothermal-vent and -spring taxa. The authors conclude that the microbial community structure in Guaymas basin sediments are determined by a combination of temperature and energy availability.
This manuscript is well-written and will be of interest to a broad community of biogeochemists and microbial ecologists. To ensure that these communities, and more, benefit from this manuscript, the authors should consider moving some of the material in the Supplemental Materials section into the main text (see below). My main critical points concern how the authors frame energy and transport and how the discussion is presented. In particular, the authors conclude that energy availability is key to understanding how advective systems can host microorganisms at higher temperatures than diffusion-controlled systems. Yet, they use a rather narrow conceptualization of energy availability being the presence of organic electron donors (and short chain organic acids at that). Unless the microorganisms are all engaged in fermentation (and I see from their data that there are likely fermenters present), then it is the presence of electron acceptors AND donors that will determine the energy availability -if the electrons have nowhere to go, then no energy can be gained. Furthermore, since transport is key to this energy availability, the authors should consider discussing this in terms of the rate of energy availability, or power.
Also, the authors broadly distinguish between Seep Areas and Non-Seep Areas, yet I don't see how these determinations were made. Typically, the Péclet number is used to characterize the relative contributions of advection and diffusion. The authors might not have the data to determine Pe values, but as it stands, it isn't clear how this designation is being made. Also, though some of the sites might be dominated by advection vs diffusion, it's the magnitude of this difference that could be responsible for observed geochemical and microbiological trends. This is also noteworthy with regard to temperature -hot sediments that are dominated by advection will also have faster diffusion rates.
Finally, there is the issue of time. The GC sites have been sampled down to 5 meters whereas some of the other NSAs and all of the SA sites have only been sampled down to between 20-45 cm. Given similar sedimentation rates, this means that the organisms at the bottom of the GC sites have been down there roughly 10 times longer than those at the bottom of the sites where shorter cores were taken. This means that these organisms have endured their entombment in sediments far longer and have likely been subjected to the stresses of low-energy and nutrient deprivation far longer than their shallow counterparts. The authors should consider at least noting this fact and even look for temporal trends in their microbiological data.
Smaller, specific comments Why is sulfide just considered a respiration end-product and not also an electron donor? Presumably methane is considered to be both.
Is total organic carbon actually just particular organic carbon (POC), or does it include dissolved OM? This should be clarified.
The section, "Trends in dissolved SCOAs across locations" could use some reorganization. The first sentence needs to encompass what the paragraph will contain, but its starts out with an observation about SCOAs at two sites, then goes into what compounds were found and then a sentence beginning "In constrast…" that doesn't make sense because it does not contrast the preceding sentence.
Is 'TN' total organic nitrogen or total nitrogen? Please specify.
Line 242 The authors could be a bit more specific and scientific than using the phrase 'Humpshaped' to describe a data trend. Perhaps 'sigmoidal' or 'logistic'?
Define 'ZOTU.' There is a large and growing way to refer the these. And in a few years, who knows which one will be standard.
The authors note that, starting on line 376, "the elevated electron donor input to deeper layers of the hydrothermal seep sites does not compensate for temperature-driven increases in cell-specific energy requirements." However, the authors do not present any values for cell-specific energy requirements at this site, nor do they specify how energy is available from electron donors. Furthermore, they would need to discuss a rate at which this energy is needed since maintenance energies are a rate of energy, not a static amount.
It's great that the authors have a section in the discussion entitled, "Alternative explanations for the observed trends" The authors note that "Energy-depleted high temperature subsurface sediments are dominated by bacterial and archaeal phyla that are also widespread in energy-depleted low-temperature subsurface settings, and likely gain energy from the breakdown of recalcitrant carbohydrates and other organic compounds." If these compounds are 'recalcitrant' then they wouldn't be broken down. Perhaps they just have a low reactivity under the conditions in which they are found. In many systems complex carbohydrates are broken down readily, with the right combination of organisms and geochemistry. On a related point, if the high temperature, high energy systems are relying on labile organics from deeper, why aren't they being consumed deeper down? Perhaps because the ecosystem of origin doesn't make then labile, but recalcitrant. It is only by being transported to another system that they become labile, a system that has microorganism that can consume them at an appreciable rate.
In addition to making biomolecular decay more rapid with higher temperatures, abiotic reactions that compete with biologically catalyzed ones also increase with temperature. Furthermore, the energy yield of different reactions change with temperature. Temperature and energy are more inter-related than the discussion currently allows. The authors should also consider that sulfide is being generated abiotically from hotter sediments via thermochemical sulfate reduction.
The entire paragraph beginning on line 579 should be in the results or discussion or background section! This paragraph answer so many questions that I had about the manuscript up until this point. Why bury this in the Material and Methods section when it is critical background material? Table 1. The temperature gradient for Everest Mound is 490 oC/m. This might be the case over a very short interval, but it suggests extreme heating, supercrital and/or vapor phase water. A bit misleading.
In figure 1, its looks like nothing much is happening with respects to sulfate reduction, methane oxidation or organic carbon oxidation in the non-seep areas. I guess the authors don't have nitrate data? Figure 2. define SCOAs in caption as well as NSA and SA; the caption notes differing scales on the depth (y) axis, but fails to note the same for the concentrations (x axis).
This figure also shows that nothing much is happening in the NSA; the high SCOA concentration deeper at the SA and lower near the SWI makes it look like they are not being utilize at depth, but only near the SWI; Figure 3 define TOC and TN in caption TOC profiles make it look like nothing is happening at NSAs, but something at SAs Figure 4 I'm assuming that the Archaeal 16S rRNA gene copy numbers are the same scale as bacterial, but this should be explicit since the scales change on other plots; upper and lower case letters don't match in caption and plots (a, A)

Figure 5
Spell out NMDS in caption; same with ZOTU Are the seep areas showing the community structure they are because they are more energy rich, or because microbes from elsewhere are being transported into this area?
The 'Supplementary background on study area' in the Supplementary Text is excellent. Consider including this in the main document for context and background.
Perhaps figure S1 should also be in the main body of the text.
13C seems to confirm that nothing is happening in NSA sediments, w.r.t organic carbon Figure S12 (heat map showing correlation between microbial groups of geochemical variables) should also be considered for the main text.
Since I have already reviewed an earlier version of this manuscript, I will skip the normal summary and comments. I think that this manuscript is ready for publication. The only additional comment that I have is that the authors have still not defined all all the acronyms in the supplemental figure captions. I won't hold up anything over this. I suspect that anyone looking at the SI will have read the text and understand the acronyms.
Thank you very much for the positive assessment. We have now spelled out all acronyms in the supplementary figure captions.