The Inverted Microbial Loop Stimulates Mineralisation of Sedimentary Organic Detritus
Abstract. Respiration is a key process in the organic carbon cycle of marine sediments. The microbial community is considered the dominant actor in the overall sedimentary respiration, but knowledge is lacking about interactions with other components, particularly the macrofauna. The ‘inverted microbial loop’ hypothesis suggests that macrofaunal activity stimulates the microbial respiration of organic carbon through the mixing of fresh organic carbon to depth, and subsequent priming (i.e. activation of refractory detritus by co-respiration with fresh detritus). We conducted experimental incubations to partition respiration amongst the microbial and macrofaunal components of the community and investigate interactions between them. We prepared sediment cores with native benthic communities, macrofauna only and microbial communities only. We added 13C labelled fresh organic matter to these cores and measured respiration over 7 days, quantifying both O2 consumption (reflecting remineralisation of all sedimentary organic C) and production of 13C dissolved inorganic C (DIC, reflecting remineralisation of labile organic C).
Macrofaunal and microbial communities showed an approximately equal contribution to the total community respiration, while the fate of the added fresh organic C in different treatments suggested competition for this resource between macrofauna and microbes. Consumption of O2, which reflected remineralisation of ambient as well as added fresh organic C, showed greater rates when macrofaunal and microbial communities were present together than the sum of their separate rates. This provides direct experimental evidence that the inverted microbial loop mechanism stimulates mineralisation of less reactive, ambient organic C. The inverted microbial loop effect is likely to be enhanced following deposition of fresh organic C onto the seafloor, as occurs after a spring bloom.
The paper by Woulds et al. presents the results of two stable-isotope labelling experiments, conducted in estuarine sediments to investigate the relationship between macrofauna and microbial processing of organic matter. Historically, the microbial community has been considered the dominant driver of sediment community respiration, and the paper correctly identifies the interactions between the microbial community and faunal communities as key to understanding the specific controls upon the benthic portion of the marine carbon cycle. The manuscript is generally well written and the results are potentially worthy of publication. I believe the results of this paper are of broad interest across the disciplines of marine ecology, biogeochemistry and the wider environmental sciences. The paper, therefore, clearly falls within the scope of Biogeosciences as an academic journal. However, the paper requires significant revision before it could be accepted. Below I provide a detailed outline of my concerns, which I hope the authors will be able to address in a revised manuscript.
Major Revisions Required
1. The level of replication within the study is low, with only three replicates of each treatment. I am well aware of the challenges that running stable-isotope pulse-chase experiments and sediment incubations can bring, both in terms of the logistics and costs. However, there are issues in making generalizations based on so few cores with relatively small internal diameters. Are these cores representative of the wider estuarine environment, specifically in terms of macrobenthic densities and species diversity? Whilst I don't expect the authors to be able to resolve these issues. I believe that this needs to be discussed in the paper and the associated limitations this places on the papers conclusions to be identified.
2. Given the low levels of replication in the study, I am not sure what value the use of inferential statistics brings to the study. I am not convinced that presenting and discussing statisitical significance in terms of p-values is of any real value. Whilst you can estimate a mean and variance based on three replicates, I'm not convinced that those values are really meaningful. Also, there isn't really any way to assess the distribution of a dataset based on experimental treatments of n=3. The use of nonparametric statistical tests won't solve this problem, which is an inherent feature of low replication. The use of Kruskal-Wallis and Mann-Whitney tests require the ranking of datasets, to allow comparison of relative magnitude. This will not resolve issues with regard to the frequency-distribution or variance of a small dataset. In essence using these approaches applies a rather drastic transformation that will reduce any information on the spread of the data from dataset. As such, the author's are likely inflating the potential for a type 2 error, where they fail to detect a 'real' treatment effect. I would recommend the authors remove any statistical testing from the manuscript and discuss the results in terms of relative differences between treatments, and the limited spread of the data. If the authors strongly believe that p-values are needed, then the use of ANOVA with an appropriate post-hoc test (Tukey HSD, Student-Newman-Keuls) would be a better option. ANOVA is reasonably robust to departures in normality and heteroscedacity, but the authors need to be clear in outlining the limitations of the study.
3. The bar charts in Figures 2, 5, 6, 7 and 8 would benefit from having the raw data overlaid as dot-plots. This would be more honest and would allow the reader to see how the mean and error bars displayed relate to the actual distribution of the data.
4. There is a fundamental difficulty with the design of experiments to partition faunal and microbial respiration in sediment ecosystems. In a perfect world, the experiment would have included sediments that were de-faunated by sieving
5. It is interesting to note that there is effectively no microbial uptake of labile organic matter in the Fauna treatment (Figure 5). This needs to be given further explanation / consideration in the discussion. Whilst I would expect to see a treatment effect, this seems very unusual. Whilst the author's discuss his in relation to competition, this feels like an oversimplification. The uptake values are based on single end-points and I think this needs to be considered given the cascading pathways for nutrient transfer through the sediment ecosystem. Consideration of the temporal changes in macrofaunal and microbial uptake need to be considered and discussed in more detail. This is critical to understanding the partitioning of respiration between the microbial and faunal groups. Furthermore, I think the data within the experiment would be amenable to a Linear Inverse Modelling approach to better parse out the microbial and faunal contributions to sediment community respiration, which might allow the author's to consider the responses to labile organic matter inputs into a system where there is a larger pool of autochthenous refractory organic matter, and incorporate additional data on microbial and faunal respiration rates.
6. The two experiments were conducted in different years, and they investigate the responses to different doses of added carbon. Furthermore, the core diameters and thus the volumes of sediments sampled differ between the two experiments. The author's need to make a stronger case in the discussion as to why these two experiments should be considered together. Are they comparable? The current presentation suggests that they are, but I think perhaps there needs to be clearer distinction of the two experiments.
7. The title needs to be revised to specify that the conclusions of the study are specific to an estuarine system. The study would need to be repeated in some format in subtidal sediments to allow broader generalizations to be made about marine sedimentary ecosystems. In addition, given that the fate of organic in marine sediments is strongly influenced by grain size, I would expect different relationships between the microbial and macro-benthic communities of sandy and muddy sediments. The current title suggests that the papers conclusions are much broader.
Minor Revisions Required.
Page 2, Line 35: Delete 'Hence'.
Page 2, Lines 37 and 38: Replace long dashes with commas.
Page 2, Line 39: Replace 'and as a consequence' with 'consequently'.
Page 2, Line 41: Replace 'high' with 'higher'.
Page 2, Lines 46-48: 'In contrast to our knowledge...' Revise for clarity.
Page 3, Line 54: Delete 'For this reason'.
Page 4: Lines 85-95: Please clarify the aims and clearly outline the hypothesis of the study.
Page 4, Line 98: 'incubated over time.' Please specify the incubation time.
Page 9 Line 201 - Page 10 Line 215: Any statistical tests used need to be described and explained in the materials and methods.
Page 9 Line 201: Can the author's publish the dataset, so that this can be viewed and possibly interrogated by the reader.
Page 23, Line 348-350: 'In summary, for the majority of our timepoints, the observed O2 consumption rates, which reflect degradation of total sedimentary organic matter, supported the occurrence of the inverted microbial loop.' I think at this point a recap of what is mean by the 'inverted microbial loop' would help. I realize this is outlined in the introduction, but a single sentence would suffice.
Page 25. Line 368-369: "The 13C that could not be accounted for presumably remained in the sediment, although data are not available to confirm this. " This statement is partially true. Another major pool of 13C that is not measured is the Dissolved Organic Carbon pool. Whilst the authors are not seeking a mass balance, it is worth referencing the potential shunt of organic matter into the DOC pool, which may be important in terms of the biogeochemistry of the system.
Page 30 Lines 456 - 464: The conclusions need to clearly state that the results of this experiment are focused on an estuarine system. I don't believe this diminishes the findings, but highlights the wider knowledge gap around microbial-faunal interactions in marine sediments.