Fire and Rhizosphere Effects on Bacterial Co-Occurrence Patterns

Round 1

Reviewer 1 Report

The authors have submitted a manuscript in which they re-analyse sequencing  data obtained and published by Aponte et al. in 2022 and available at the Sequence Read Archive (NCBI). These data concern the structure of the microbiota of soils (bulk and rhizosphere) that were affected or not by a forest fire 33 months earlier.

Major remarks

The main contribution of this re-analysis of previously published data is the analysis of interaction networks of bacterial populations at genus or species level (OTUs). It must be acknowledged that the authors have made an effort to explain the twenty or so parameters used to describe the co-occurrence networks.
The results are interesting but the conclusion cannot be summarized as a difference observed between the network of the rhizosphere soil not affected by fire (RU) and that of the bulk soil affected by fire (BB).

The analysis of the main parameters (nb of nodes, nb of ties, ANS, % deg centralization,  density, fragmentation, modularity, nb of modules, compactness, small wordlness, robustness and effectiveness) needs to be analysed statistically to determine whether the effect of fire and the effect of plant cover (bulk vs rhizosphere) is significant or not.
The same applies to the data presented in Table 3, as only a statistical analysis can draw conclusions from this type of results.

Minors remarks:

Line 133: the correct reference is [27] not [33]

Lines 270 and 273: Paenibacillus instead of Paenicibacillus

Line 459: 'Heulin' instead of 'Heuli'

Line 489: 'Allende R.' instead of 'Rosario Al-lende'

 

 

Author Response

Reviewer 1

Dear Sir/Madam

Thank you for your comments.

The authors have submitted a manuscript in which they re-analyse sequencing  data obtained and published by Aponte et al. in 2022 and available at the Sequence Read Archive (NCBI). These data concern the structure of the microbiota of soils (bulk and rhizosphere) that were affected or not by a forest fire 33 months earlier.

Major remarks

The main contribution of this re-analysis of previously published data is the analysis of interaction networks of bacterial populations at genus or species level (OTUs). It must be acknowledged that the authors have made an effort to explain the twenty or so parameters used to describe the co-occurrence networks.

The results are interesting but the conclusion cannot be summarized as a difference observed between the network of the rhizosphere soil not affected by fire (RU) and that of the bulk soil affected by fire (BB).

This study analyzed the fire effects on soil bacteria on a small spatial scale (rhizosphere vs. bulk soil). So, the general trend that soil microbial communities are distinguished in terms of composition and/or diversity between burnt and unburnt areas was not probably expected to be so clear in this study. Also, the study examined the network of relations among the bacterial OTUs at the level of genus and species and not the composition or the diversity of the community at the level of family or phyla as it was usually happened. Under these circumstances the biggest differences in co-occurrence patterns were revealed between the rhizosphere-unburnt and the bulk-burnt soil. The first appeared as the most fragmented and the second as the most compact.

Further there were some network characteristics that all networks exhibited like modularity and small worldness. We attributed these to the high heterogeneity of the Mediterranean soils that are under continuous anthropogenic pressures such as fire and overgrazing. When Bonanomi et al. (2022) compared the microbial networks in burnt and mowed areas with those in abandoned areas, they found that the first two networks consisted of a higher number of dispersed subcommunities compared to the latter.

As a general conclusion we could say that some networks’ characteristics were unaffected by fire, some other were affected by fire no matter the spatial scale, while the rest differed in relation to fire and spatial scale.

The analysis of the main parameters (nb of nodes, nb of ties, ANS, % deg centralization,  density, fragmentation, modularity, nb of modules, compactness, small wordlness, robustness and effectiveness) needs to be analysed statistically to determine whether the effect of fire and the effect of plant cover (bulk vs rhizosphere) is significant or not.
The same applies to the data presented in Table 3, as only a statistical analysis can draw conclusions from this type of results.

To construct a network describing the network of ties between organisms in a specific treatment you use the organisms’ abundance or their presence/absence recorded in all replicates of this treatment. The analysis based on correlation coefficients or similarity/dissimilarity indices per pair of organisms as this recorded in all replicates of the same treatment. So, the produced network estimates single values for the number of nodes, number of ties, clustering coefficient etc. Thereby, there are no replicates for these parameters, so you cannot apply any statistical analysis in network parameters, namely analysis of variance as it was mentioned by the reviewer.

Note please, that appropriate statistical analysis was applied when constructing the network with the Cytoscape software (App CoNet). This analysis aims to keep only the ties that are statistically significant and to avoid the possibility to get a tie because of chance. This happens by evaluating the ties by using different similarity/dissimilarity indices (Bray–Curtis, Kullback–Leibler, Mutual Information) or correlation coefficients (Spearman, Pearson) or both and then applying permutation and bootstrap methods to keep only those ties that are statistically significant.
Based on the above, we cannot adopt the reviewer’s suggestion
and apply statistical test in the values of the network parameters or in the data presented in Table 3. The data on Table 3 was based on network analysis since they referred only to the most influential species/genera meaning the nodes that exhibit high degree centrality. However, in case that the reviewer insist we could subtract Table 3. 

For the reliability of the information presented above, I attach some relevant literature concerning network analysis:

Stamou, G.P., Monokrousos, N., Papapostolou, A. et al. Recurring heavy rainfall resulting in degraded-upgraded phases in soil microbial networks that are reflected in soil functioning. Soil Ecol. Lett. 5, 220161 (2023). https://doi.org/10.1007/s42832-022-0161-3.

E.M. Papatheodorou, N. Monokrousos, E. Angelina, G.P. Stamou, 2021.Robustness of rhizosphere microbial communities of L. sativa originated from soils of different legacy after inoculation with Plant Growth Promoting Rhizobacteria. Appl. Soil Ecol. Volume 167, 104028, https://doi.org/10.1016/j.apsoil.2021.104028

Stamou, G.P.; Argyropoulou, M.D.; Rodriguez-Polo, I.; Boutsis, G.; Kapagianni, P.; Papatheodorou, E.M. A Case Study of Nematode Communities’ Dynamics along Successional Paths in the Reclaimed Landfill. Diversity 2020, 12, 274. https://doi.org/10.3390/d12070274

Go Y., Yu L., Zhao L. 2022. Ecological Networks In Agroecosytems: Approaches and applications. Front. Agr. Sci. Eng. 9: 523–535. https://doi.org/10.15302/J-FASE-2022466

Pérez-Valera E., Goberna M., Faust K., Raes J., García C., Verdú, M. 2017. Fire modifies the phylogenetic structure of soil bacterial cooccurrence networks. Environmental Microbiology, 19, 317–327. https://doi.org/10.1111/1462-2920.13609

Lan G., Yang C, Wu Z., Sun R., Chen B., Zhang X. 2022 Network complexity of rubber plantations is lower than tropical forests for soil bacteria but not for fungi. SOIL, 8: 149–161. https://doi.org/10.5194/soil-8-149-2022

Deng, Y., Jiang, Y-H., Yang, Y., He, Z., Luo, F. and Zhou, J., “Molecular ecological network analyses”, BMC Bioinformatics, 13, 113, 2012. https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-13-113

Minors remarks:

Line 133: the correct reference is [27] not [33] DONE

Lines 270 and 273: Paenibacillus instead of Paenicibacillus DONE

Line 459: 'Heulin' instead of 'Heuli' DONE

Line 489: 'Allende R.' instead of 'Rosario Al-lende' DONE

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript is well-written, clear, and on point. However, I have a few technical matters that require to be adjusted before final acceptance, the figure 1 should be well ordered in a finer square, as we can see the bulk soil burnt was hidden due to mal-organization, you should make a square for each treatment and squares should be equal for all four. Moreover, authors should justify the choice of bacteria instead of fungi in their analysis and also give a brief description of the way of sampling should be given, I know it was already done in a previous paper but a scientific paper should be independent in itself, therefore, it is crucial to describe sampling method, number of replicates, timing, brief DNA extraction method, PCR primers, pipeline used for bioinformatics. Finally, as an update of the literature, I recommend the authors to read this paper to improve their discussion section, it's a long-term experiment on the effect of burning/fire on the microbiota of a Mediterranean grassland (Impact of prescribed burning, mowing and abandonment on a Mediterranean grassland: A 5-year multi-kingdom comparison. Science of The Total Environment 834, 155442).

Author Response

Reviewer 2

Dear Sir/Madam

Thank you for your comments.

The manuscript is well-written, clear, and on point. However, I have a few technical matters that require to be adjusted before final acceptance, the figure 1 should be well ordered in a finer square, as we can see the bulk soil burnt was hidden due to mal-organization, you should make a square for each treatment and squares should be equal for all four.

We did it. Please see Figure 1.

Moreover, authors should justify the choice of bacteria instead of fungi in their analysis and also give a brief description of the way of sampling should be given, I know it was already done in a previous paper but a scientific paper should be independent in itself, therefore, it is crucial to describe sampling method, number of replicates, timing, brief DNA extraction method, PCR primers, pipeline used for bioinformatics.

In the present study we investigated the soil microbial interactions in small-scale environment (rhizosphere vs. bulk soil) in relation to fire effects. The high diversity and functional redundancy that characterize the soil bacterial communities gives the opportunity to check the effect of environmental filtering since different genera/species or strains could be selected under slightly different environmental conditions. This makes prokaryotes more suitable for studying changes in interactions or community composition in points nearby to each other. As Bonanomi et al. (2022) mentioned the three soils (abandoned, mowed and burnt) share 138 common fungal OTUs and only 79 bacterial OTUs indicating that prokaryotes had higher discriminatory power.

Some details for the experimental design, DNA extraction and pipelines used are presented in lines 133-161.

 

Finally, as an update of the literature, I recommend the authors to read this paper to improve their discussion section, it's a long-term experiment on the effect of burning/fire on the microbiota of a Mediterranean grassland (Impact of prescribed burning, mowing and abandonment on a Mediterranean grassland: A 5-year multi-kingdom comparison. Science of The Total Environment 834, 155442).

We followed your suggestion, and we make some changes in the discussion based on this article. Please see lines 340-343 and 422-427.

 

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors' response to the remarks about the lack of statistical treatment of the data are well argued given the design of the experiment, but I still think that the impossibility of distinguishing the effect of fire from the rhizosphere effect prevents clear and sufficiently argued conclusions from being given. At least in the conclusions, this impossibility of distinguishing the two effects should be mentioned.

Author Response

We thank the reviewer for taking the time to provide valuable feedback. We appreciate the concern regarding the inability to distinguish the effect of fire from the rhizosphere effect and how it may have impacted our conclusions. We acknowledge that this is a complex issue and, although we have presented a well-designed experiment, we agree that it is difficult to attribute the observed effects solely to fire or the rhizosphere. We have included additional language in the conclusions to emphasize this point and to acknowledge that this limitation may have affected our ability to draw definitive conclusions.

Please see lines 454-458. “Overall, our study provides valuable insights into the complexity and robustness of bacterial networks in burnt Mediterranean soils; however to distinguish the fire effect from the rhizosphere effect is difficult and highlights the need for further research to disentangle the effects of these two agents on bacterial community structure”.

 

Author Response File: Author Response.docx