Polychaete Invasion May Lead to Biogeochemical Change in Host Marine Environment

Round 1

Reviewer 1 Report

The manuscript of Quintana & Kristensen "Polychaete invasion may lead to biogeochemical change in host marine environment" is a study on the shifts of biogeochemistry of the sediment caused in the Baltic by the invasive polychaete Marenzelleria viridis. The authors are using exeperiments in aquaria to study the SO₄ reduction, the AVS and CRS profiles. Although the subject of the study is very interesting, it suffers from lack of data. The overall work is based on small number of experimental replicates (only two) and poor amount of data. Therefore I recommend the rejection of the article until adequate amount of data is obtained to support the conclusions of the research. More specifically:

• Only two aquaria of M. viridis were used for the experiments. At least three replicates are needed to draw safe conclusions. 
• The species H. diversicolor was used to compare the in situ H₂S profile but not in the experimental part. 
• At the legend of Figure 1 is impossible to read the values. 
• Figure 3: the graph of M. viridis for Spring and H. diversicolor for Summer has great standard errors. 
• Figure 3: in the legend N. diversicolor should change to H. diversicolor. 
• Table 1: the Spring and Summer values of H₂S for H. diversicolor have great standard errors e.g. the error is almost 7 for a value of 12. 
• Lines 165-166 "while CRS was deposited even further away": the progile in Aq. 2 shows a different profile than the one described here. 
• Lines 181-182 "However, chemoautotrophic bacteria are probably not a major food source for M. viridis in shallow coastal areas where abundant microphytobenthos is available": A reference is needed. 
• Lines 187-188 "Otherwise, the depth-integrated in situ H₂S levels in M. viridis sediment was consistently higher than in H. diversicolor sediment during all other seasons": The profiles don't show consistently higher levels of H₂S. Only the mean values are such. But the mean values could be driven by great values that appear occasionally in each season. 
• Discussion lacks comparison to other studies. Conclusion is driven by limited amount of data. 

Author Response

Our response is provided below the reviewer's comments in italics:

The manuscript of Quintana & Kristensen "Polychaete invasion may lead to biogeochemical change in host marine environment" is a study on the shifts of biogeochemistry of the sediment caused in the Baltic by the invasive polychaete Marenzelleria viridis. The authors are using exeperiments in aquaria to study the SO₄ reduction, the AVS and CRS profiles. Although the subject of the study is very interesting, it suffers from lack of data. The overall work is based on small number of experimental replicates (only two) and poor amount of data. Therefore I recommend the rejection of the article until adequate amount of data is obtained to support the conclusions of the research. More specifically:

• Only two aquaria of M. viridis were used for the experiments. At least three replicates are needed to draw safe conclusions. 

We are aware of the need of more replicates for the thin-aquaria part. The initial set-up had 4 replicates with M. viridis, but unfortunately, we lost 2 during the acclimation phase (lines 74-76). However, instead of applying repeated experimental set-ups to force optimal replication, our approach was to focus on the successful aquaria that we obtained with visible burrows and with comparable sediment conditions (lines 76-79). Additionally, the study is supported by field observations, which provide an important link between the findings in the lab and responses in natural systems. In the conclusions, we clearly mention that the results of this study should be interpreted with care due to the small number of replicates (lines 241-245).  

• The species H. diversicolor was used to compare the in situ H₂S profile but not in the experimental part. 

Several studies point out the increase in aerobic metabolism stimulated by H. diversicolor, therefore the focus of the experimental part was on mapping with high resolution the effects of M. viridis on the sulfate reduction

We chose to prioritize the sulfate reduction assays in M. viridis aquaria, since the SR assays are laborious, time-constrained and they must be done simultaneously in all treatments to allow valid comparisons. Furthermore, it is well known that H. diversicolor inhibit sulfate reduction in sandy sediments (Banta et al. 1999, Kristensen 2000, Kristensen et al. 2011). To facilitate comparisons, we added the previous data from Banta et al. (1999) with similar sandy sediment conditions to Figure 2. We found a mistake in the depth-integrated SR data, which considered before 1 cm depth for all sediment depths. Therefore, we corrected the depth-integrated SR using 2 cm depth (except the first sediment section) and the corresponding text in the discussion (lines 142-143).

At the legend of Figure 1 is impossible to read the values. 

We increased the font size and now the values are clearer.

• Figure 3: the graph of M. viridis for Spring and H. diversicolor for Summer has great standard errors. 

We double checked the standard errors and there are no mistakes. Great standard errors are expected when we analyze intact sediment cores collected in the field due to heterogeneity. Therefore, no changes were made.

• Figure 3: in the legend N. diversicolor should change to H. diversicolor.

Legend changed to H. diversicolor.

• Table 1: the Spring and Summer values of H₂S for H. diversicolor have great standard errors e.g. the error is almost 7 for a value of 12. 

There was a typo mistake in H. diversicolor standard error, it is corrected now for 3.9 while summer values are correct.

• Lines 165-166 "while CRS was deposited even further away": the progile in Aq. 2 shows a different profile than the one described here. 

We reformulated the sentence specifying that this was observed in Aq. 1 (line 186).

• Lines 181-182 "However, chemoautotrophic bacteria are probably not a major food source for M. viridis in shallow coastal areas where abundant microphytobenthos is available": A reference is needed. 

The reference “Dauer (1997)” was added (line 207).

• Lines 187-188 "Otherwise, the depth-integrated in situ H₂S levels in M. viridis sediment was consistently higher than in H. diversicolor sediment during all other seasons": The profiles don't show consistently higher levels of H₂S. Only the mean values are such. But the mean values could be driven by great values that appear occasionally in each season. 

We reformulated this sentence for clarification (lines 213-215).

• Discussion lacks comparison to other studies. Conclusion is driven by limited amount of data. 

We added more references to the discussion following revisions as well required by the other reviewers. However, there are few examples of sulfate reduction rates measured in 2D planar aquaria. The conclusion was also reformulated as mentioned previously.

Reviewer 2 Report

I think the paper is well written and based on a complex experimental design, however, the title must be changed because it seems that all the alien species could have a detrimental effect. In my opinion, the authors have to refer to this specific case related to the studied species.

What it sounds strange for me is that the experiment was compared to a previous experiment with Hediste, but are the method comparable? However, data are in accordance with field observations that considers both the taxa. 

The most relevant thing that I cannot judge, however, is that it seems strange to me that the bioturbation by a polychaete creates an anoxic condition, more than in deufanated control, but I am not an expert.

Author Response

Reviewer 2:

I think the paper is well written and based on a complex experimental design, however, the title must be changed because it seems that all the alien species could have a detrimental effect. In my opinion, the authors have to refer to this specific case related to the studied species.

The species name is described both in the abstract and in keywords, therefore searching platforms can quickly find these options. So, we decided to not change the title.

What it sounds strange for me is that the experiment was compared to a previous experiment with Hediste, but are the method comparable? However, data are in accordance with field observations that considers both the taxa. 

Yes, the methods are fully comparable, since it was made using the same sediment type and incubation conditions. This is now stated in L97-L99.

The most relevant thing that I cannot judge, however, is that it seems strange to me that the bioturbation by a polychaete creates an anoxic condition, more than in deufanated control, but I am not an expert.

The bioturbation by Marenzelleria viridis did not create more anoxic conditions than defaunated control, but ventilation is too slow for oxygen to penetrate from the burrow into the sediment. It stimulates sulfate reduction in determined zones of the sediment, as shown in this study, by the transport of sulfate to sediment zones where it is depleted, producing H2S. Sulfate can be depleted in non-bioturbated sediments without replenishment of this solute by bioturbation and the formation and percolation of H2S can alter the redox conditions. This is explained along the discussion session.

Reviewer 3 Report

The manuscript "Polychaete invasion may lead to biogeochemical change in host marine environment " by Quintana and Kristensen is combining the results of: (1) a laboratory experiment (1DV and 2D cartographies) and (2) field measurements to assess the impact of the invasive species Marenzelleria viridis on sulfate reduction and resulting H2S concentrations in porewater (comparison between M. viridis and a H. diversicolor dominated areas).

The study is original (2D sampling adopted during the laboratory experiments) suffering from low replication as acknowledged by the authors themselves. I do not believe that this weakness precludes publication since the adopted design generates considerable work that provides relevant information on the tackled question.

However, I would stronly recommend the authors to be more cautious with their use of statistics.They presently use 1W ANOVA to compare the results of the two experimental aquaria with M. Viridis with those of a single control aquarium. Several problems here: (1) the use of parametric ANOVA is inapropriate for such a small number of replicates and in this case the power of non parametric ANOVA is unsufficient as well, (2) to overcome this difficulty, the authors apparently mixed the variability within and between aquaria which is not correct from a statistical standpoint, and (3) they did not check for variance homogeneity before running the analysis (apparently much less variability in the control than in the presence of worms). The same flaws apply to the analysis of field results. In addition, the significance of the interaction is the first point that should be checked in a 2W ANOVA. This was not done although interaction is likely significant here due to the difference between the summer and the other 3 seasons experiments. The mix between different sources of variances in the profile figures is also potentially misleading. I would strongly recommend the authors to acknowledge the fact that they cannot run hypothesis testing on their experimental data and to redraw Figure 3 by showing the averave vertical profiles and their within aquarium variability in the 3 aquaria. In my opinion, this should be justified by the originality of the new provided insights.

 The field component of the manuscript clearly suffers from the confounding effect of temperature. It would also have been interested to include a control area although this is already justified by the authors.

Other suggestions:

(1) In the introduction, the gap in the litterature (l52-54) is not really in line with the content of the study. This could be improved to better show the interest of the study and its contribution to the tackled scientific question.

(2) Figure 2 upper panes, SR scale not readable.

(3) Figure 2 Shaded area barely visible and not clear what it corresponds to (thin aquaria area?). sediment or what?

(4) Table 1 Not necessary clear why winter temperature was higher at the H. diversicolor (although shallower) than at the M. viridis station

Author Response

Our response is indicated in the text below in italics:

The manuscript "Polychaete invasion may lead to biogeochemical change in host marine environment " by Quintana and Kristensen is combining the results of: (1) a laboratory experiment (1DV and 2D cartographies) and (2) field measurements to assess the impact of the invasive species Marenzelleria viridis on sulfate reduction and resulting H2S concentrations in porewater (comparison between M. viridis and a H. diversicolor dominated areas).

The study is original (2D sampling adopted during the laboratory experiments) suffering from low replication as acknowledged by the authors themselves. I do not believe that this weakness precludes publication since the adopted design generates considerable work that provides relevant information on the tackled question.

However, I would stronly recommend the authors to be more cautious with their use of statistics.They presently use 1W ANOVA to compare the results of the two experimental aquaria with M. Viridis with those of a single control aquarium. Several problems here: (1) the use of parametric ANOVA is inapropriate for such a small number of replicates and in this case the power of non parametric ANOVA is unsufficient as well, (2) to overcome this difficulty, the authors apparently mixed the variability within and between aquaria which is not correct from a statistical standpoint, and (3) they did not check for variance homogeneity before running the analysis (apparently much less variability in the control than in the presence of worms). The same flaws apply to the analysis of field results. In addition, the significance of the interaction is the first point that should be checked in a 2W ANOVA. This was not done although interaction is likely significant here due to the difference between the summer and the other 3 seasons experiments. The mix between different sources of variances in the profile figures is also potentially misleading. I would strongly recommend the authors to acknowledge the fact that they cannot run hypothesis testing on their experimental data and to redraw Figure 3 by showing the averave vertical profiles and their within aquarium variability in the 3 aquaria. In my opinion, this should be justified by the originality of the new provided insights.

We followed the suggestion of reviewer 3 and removed the one-way ANOVA test on the aquaria. We checked for interaction between seasons and intertidal zones in the upper 10 cm sediment with two-way ANOVA, but it was not significant (lines 168-170).

 The field component of the manuscript clearly suffers from the confounding effect of temperature. It would also have been interested to include a control area although this is already justified by the authors.

We agree, but the field component is essential to translate and compare results found in the lab.

Other suggestions:

(1) In the introduction, the gap in the litterature (l52-54) is not really in line with the content of the study. This could be improved to better show the interest of the study and its contribution to the tackled scientific question.

We modified the sentence and now it is more in line with the content of the study (lines 54-56).

(2) Figure 2 upper panes, SR scale not readable.

We supposed the reviewer meant Figure 1. The scale was increased as also requested by reviewer 1.

(3) Figure 2 Shaded area barely visible and not clear what it corresponds to (thin aquaria area?). sediment or what?

The shaded area represents the area of the thin aquaria used for Figure 1. This is mentioned in Figure 2 caption (line 151). We have darkened the area to enhance its visibility.

(4) Table 1 Not necessary clear why winter temperature was higher at the H. diversicolor (although shallower) than at the M. viridis station

We sampled in late winter when the shallower and upper H. diversicolor intertidal zone was exposed to winds and may reach similar temperatures as the air. This is corrected in line 103.

Reviewer 4 Report

The authors present a well-written paper with important information on biogeochemical effects from a wide-spread invasive species. My sole critique is the wrong use of statistics which is absolutely necessary to modify before publication. It is not possible to do a one-way anova with 2 replicates in one group and 1 replicate in another. Samples from a single aquaria are to be seen as biased samples and not indipendent samples. Instead of falsely/inappropriately applying statistical tests, only descriptive statistics should be used. There is nothing inherently wrong with presenting qualitative, non-interpreted data in a scientific paper and definitely better than performing statistics wrong. The information should be published and the results are clear enough that descriptive statistics are ok for the aquaria experiment and for the seasonal cores, there are enough replicates to apply statistics in a proper way if a one-way design is used to test the species effect.

The table 1 caption describes n=2-3 and a two-way ANOVA cannot have one group with 2 replicates as it then is not statistically ok to compare variances across groups. From my visual interpretation of the data, the seasonal effect is mainly attributed to differences in temperature (which also methodologically could be improved by measuring sediment temperature rather than bottom water temperature). Thus I would recommend a one-way ANCOVA or PERMANCOVA test, where the species effect is tested statistically and temperature is used as covariate. This further has the advantage that it at least statistically tries to correct the large difference in summer temperatures for the different habitats the species occupy. 

Comments for author File: Comments.pdf

Author Response

Our response is indicated below the comments in italics:

The authors present a well-written paper with important information on biogeochemical effects from a wide-spread invasive species. My sole critique is the wrong use of statistics which is absolutely necessary to modify before publication. It is not possible to do a one-way anova with 2 replicates in one group and 1 replicate in another. Samples from a single aquaria are to be seen as biased samples and not indipendent samples. Instead of falsely/inappropriately applying statistical tests, only descriptive statistics should be used. There is nothing inherently wrong with presenting qualitative, non-interpreted data in a scientific paper and definitely better than performing statistics wrong. The information should be published and the results are clear enough that descriptive statistics are ok for the aquaria experiment and for the seasonal cores, there are enough replicates to apply statistics in a proper way if a one-way design is used to test the species effect.

We thank the reviewer for the valuable suggestions. We decided to drop the statistical analysis of the aquaria and focus on the qualitative interpretation of the experimental part.

The table 1 caption describes n=2-3 and a two-way ANOVA cannot have one group with 2 replicates as it then is not statistically ok to compare variances across groups. From my visual interpretation of the data, the seasonal effect is mainly attributed to differences in temperature (which also methodologically could be improved by measuring sediment temperature rather than bottom water temperature). Thus I would recommend a one-way ANCOVA or PERMANCOVA test, where the species effect is tested statistically and temperature is used as covariate. This further has the advantage that it at least statistically tries to correct the large difference in summer temperatures for the different habitats the species occupy. 

We thank the reviewer for the suggestions regarding the temperature measures. We agree about the temperature effect, which is mentioned in the discussion (lines 209-214). However, the main goal of the study was not to explain the effects of temperature on H2S, but to investigate levels of SR in the lab experiment and relate them with H₂S in the field. Therefore, we kept the statistical design with the two-way ANOVA and tested for interactions between the factors season and M. viridis and H. diversicolor zones. No significant interaction was found between factors for the H2S in the upper 10 cm of the sediment (lines 168-170).

Reviewer 5 Report

Review of jmse-908498

Polychaete invasion may lead to biogeochemical change in host marine environment

Specific comments to authors

As an overall comment, the amount of replication for experimental part is far too low. Only two aquaria had one worm each (two replicates!) in homogenized sediment and the only control only homogenized sediment. That being said, it is good that the experimental results were expanded with results from the field for more realism.

Overall comment on the results and the design: Using homogenized sediment as a control and homogenized sediment with even only one worm added as a treatment, there will surely be some kind of an effect, but how realistic this effect is, is questionable. Therefore, one needs to be careful about drawing more generalized conclusions about dramatic effects or, and especially, about integrating results from individual studies with only a few replicates into ecological models.

I only have a very few specific comments since the study was well conducted methodologically other than the overarching low sample size, which casts a shadow on the reliability and realism of the results.

Row 64: What were the conditions in the bay at the time of sampling (temperature?)? The acclimatization period was only one day and the temperature in the experiment 15 deg C, a sudden increase in temperature could cause a change in the behavior of the worms.

Row 73: “two optimal aquaria”: what does it mean optimal? Did the worms die in the other two or the burrows were not visible why did you choose these two?

Row 92: n=4-14 and n=2-7: is this the vertical profiles + the horizontal profiles summed per aquaria? How can profiles measured in the same aquaria in different places be replicates? As far as I understood there are two averaged SR profile replicates (one from each treated aquarium) and one from the control.

Row 184 to 187: Indeed the higher temperature would stimulate SR.

Discussion:

I would refrain, based on the limitations concerning your sample size (which you also point out), from talking about dramatic effects on the ecosystem.

Author Response

Our response is shown as italic text below the reviewer's comments:

As an overall comment, the amount of replication for experimental part is far too low. Only two aquaria had one worm each (two replicates!) in homogenized sediment and the only control only homogenized sediment. That being said, it is good that the experimental results were expanded with results from the field for more realism.

Overall comment on the results and the design: Using homogenized sediment as a control and homogenized sediment with even only one worm added as a treatment, there will surely be some kind of an effect, but how realistic this effect is, is questionable. Therefore, one needs to be careful about drawing more generalized conclusions about dramatic effects or, and especially, about integrating results from individual studies with only a few replicates into ecological models.

I only have a very few specific comments since the study was well conducted methodologically other than the overarching low sample size, which casts a shadow on the reliability and realism of the results.

Row 64: What were the conditions in the bay at the time of sampling (temperature?)? The acclimatization period was only one day and the temperature in the experiment 15 deg C, a sudden increase in temperature could cause a change in the behavior of the worms.

We have swapped the information, we meant that “sediment compaction” was allowed for one day while no specific information was given for the acclimatization. We have added now that sediment was stored in temperature-controlled room for acclimatization for three days (lines 68-69).

Row 73: “two optimal aquaria”: what does it mean optimal? Did the worms die in the other two or the burrows were not visible why did you choose these two?

We have changed the wording from “optimal” to “functional” indicating that these aquaria were appropriate for the experiment. Burrows were not visible in two aquaria, therefore we assumed that the worms have died. We have added this information in the methods (lines 77-79).

Row 92: n=4-14 and n=2-7: is this the vertical profiles + the horizontal profiles summed per aquaria? How can profiles measured in the same aquaria in different places be replicates? As far as I understood there are two averaged SR profile replicates (one from each treated aquarium) and one from the control.

We made a mistake and only one profile was done per M. viridis aquarium, as mentioned previously in the figure caption n = 2-14. We corrected this in the text (line 98) and it is also clearer now that those numbers refer to the vertical and horizontal profiles. We did two profiles in control aquaria with 8 cm apart, since no animals were present. This info is added now to the text (line 88). In addition, we illustrated how horizontal and vertical profiles were sampled with different symbols in Figure S1B for clarification.  

Row 184 to 187: Indeed the higher temperature would stimulate SR.

We reformulated this part as also requested by other reviewers (lines 210-214).

Discussion:

I would refrain, based on the limitations concerning your sample size (which you also point out), from talking about dramatic effects on the ecosystem.

We refrained our discussion by tuning down the arguments about the dramatic effects and limitations of sample size.

Reviewer 6 Report

L8 All benthic organisms modify their environments and BGC, not nvajust isives.

Suggest: The invasive polychaete, M.v., alters the BGC of its surrounding sediments in shallow Baltic Sea areas. We hypothesized that M.v. increases sulfate reduction (SR) in microzones surrounding its burrows….. [drop “in this study”]….

Italics for species and generic names and their abbreviations

L11 and elsewhere H₂S

L13, suggest “rates” rather thea “hotspots” [over used and not descriptive]

L21….e.g.,…..

35, I think that the supplemental figures should be in the text, separately, as Fig 1 and Fig 2. Fig 1 near end of L35 sentence, Fig 2 near at the end of sentence in l67. Then renumber figures 1-3 to 4-6.

L73, drop “have”

L93-94, diversicolor, because previous studies documented increased rates ….

L100, …cores were sliced immediately after return to the laboratory into ….

L102, …intervals 19012 and 14-16 were discarded to simplify analytical procedures…. The reason why the sections were excluded is still not explained.

L109, fact for burrow depth limit for H. diversicolor should have a literature reference

L122 has a significant p value, L138,it is confusing to understand  how the control with only one aquarium for testing has a standard error of the mean and can be used in an ANOVA comparing 2 M. viridis to a single control.

L138, L122 has a significant p value; it is confusing to understand  how the control with only one aquarium for testing has a standard error of the mean and can be used in an ANOVA.

The units for panels A, B and C are not legible. The dark violet is not easily distinguished from the black background. The dark violet color is not distinguishable from the black background in any other figure with that combinations.

L138, Fig 2 it is not clear if these data are in the thin aquaria.

L154, Fig 3 legend, should identify these measurements are made in sediment cores cores

L194, …….at the sediment surface… delete the extra “the”

L196…. i.e.,…..

L 206, suggest: These high SR zones increase porewater …..Could delete “where we observed it, “

L208, agree the data are limited. But you don’t have to say “it must be emphasized that the …  Suggest: We are confident with the data collected that support our conclusions, but the study would have been strengthened by additional thin-aquaria experiments.

L209, I like the idea of recruitment studies, but the area without exposure to H₂S supports a very different benthic community. Could you develop different levels of surface H₂S and follow recruitment?

I did not check inclusion or exclusion of references or format.

Author Response

Our reply is indicated as italic text below the reviewer's comments:

L8 All benthic organisms modify their environments and BGC, not nvajust isives.

Suggest: The invasive polychaete, M.v., alters the BGC of its surrounding sediments in shallow Baltic Sea areas. We hypothesized that M.v. increases sulfate reduction (SR) in microzones surrounding its burrows….. [drop “in this study”]….

We wanted in fact to emphasize the change in biogeochemical conditions when an invasive species arrive in a host environment. All benthic organisms only modify their environments via bioturbation when the system is out of balance or after disturbance (Kristensen et al. 2014). In principle, benthic organisms do not enhance organic matter degradation in a steady state ecosystem (i.e. natural sediments), they maintain the redox conditions in balance. Therefore, no changes were made in this context. We have though removed “in this study” in the following sentence (line 9).

Italics for species and generic names and their abbreviations

We corrected all species and generic names and their abbreviations to italics.

L11 and elsewhere H₂S

We are not sure what exactly the reviewer requested. No changes were made.

L13, suggest “rates” rather thea “hotspots” [over used and not descriptive]

We corrected to “high SR rates”.

L21….e.g.,…..

Comma added after e.g.

35, I think that the supplemental figures should be in the text, separately, as Fig 1 and Fig 2. Fig 1 near end of L35 sentence, Fig 2 near at the end of sentence in l67. Then renumber figures 1-3 to 4-6.

We kept the figure as supplementary material since we want to have the focus on the outcomes of the experiment and field measures. However, we modified the grid aquarium figure for a better understanding of how horizontal and vertical profiles were sampled.

L73, drop “have”

We removed “have”.

L93-94, diversicolor, because previous studies documented increased rates ….

The text was corrected as suggested above.

L100, …cores were sliced immediately after return to the laboratory into ….

The text was corrected as suggested above.

L102, …intervals 19012 and 14-16 were discarded to simplify analytical procedures…. The reason why the sections were excluded is still not explained.

We added “due to limited solute changes in deeper sediment zones” (line 108).

L109, fact for burrow depth limit for H. diversicolor should have a literature reference

The reference Muus (1967) was added to the sentence.

L122 has a significant p value, L138,it is confusing to understand  how the control with only one aquarium for testing has a standard error of the mean and can be used in an ANOVA comparing 2 M. viridis to a single control.

We followed the suggestion of reviewer 2 and removed the one-way ANOVA test from the experimental part.

L138, L122 has a significant p value; it is confusing to understand  how the control with only one aquarium for testing has a standard error of the mean and can be used in an ANOVA.

We followed the suggestion of reviewer 2 and removed the ANOVA test from the experimental part.

The units for panels A, B and C are not legible. The dark violet is not easily distinguished from the black background. The dark violet color is not distinguishable from the black background in any other figure with that combinations.

We changed the dark violet color to light violet color, which represented the lower range of all plots. The plots are now more distinguished from the dark blue colors. We also included the light violet color into the legend.

L138, Fig 2 it is not clear if these data are in the thin aquaria.

We added “measured in the thin-aquaria” in the figure caption.

L154, Fig 3 legend, should identify these measurements are made in sediment cores cores

We added “measured in cores” in the figure caption.

L194, …….at the sediment surface… delete the extra “the”

We deleted the extra “the”.

L196…. i.e.,…..

We added the comma after i.e.

L 206, suggest: These high SR zones increase porewater …..Could delete “where we observed it, “

We corrected to: “These high SR zones increase porewater H2S” and deleted “where we observed it” (line 241).

L208, agree the data are limited. But you don’t have to say “it must be emphasized that the …  Suggest: We are confident with the data collected that support our conclusions, but the study would have been strengthened by additional thin-aquaria experiments.

We followed the suggestion of the reviewer and corrected the text to: “We are confident that the data collected support our conclusions, but the study would have been strengthened if all planned thin-aquaria replicates had been successful”.

L209, I like the idea of recruitment studies, but the area without exposure to H₂S supports a very different benthic community. Could you develop different levels of surface H₂S and follow recruitment?

Yes, we could follow recruitment under different levels of H2S in the field by collecting benthic larvae with nets and benthic fauna in different mesh sizes to check for juveniles. No changes were made but it is an interesting approach suggested by the reviewer.

I did not check inclusion or exclusion of references or format.

We checked and all references are included.

Round 2

Reviewer 1 Report

This is the second round of reviews on the manuscript entitled "Polychaete invasion may lead to biogeochemical change in host marine environment". The authors have made changes to the manuscript, however it is not improved as there are important matters raising mainly by the poor data. More specifically:

• for the experiment two Aquaria are used, the insufficient number of repleicates is obvious in the analysis. In Figure 1 the graphs between the aquaria show very different patterns. This is also shown in the text: "Hotspots of SR ranging from 220 to 539 nmol cm^-3 d^-1 were identified in the center along the burrow walls of the two aquaria with M. viridis down to 16cm depth (Figure 1A and B)." (lines 119-120) and in lines 123-124 "The overall rates of SR in M. viridisaquaria was about 2-fold higher than the control aquarium (Figure 1C).". This means that -since the difference 220 to 539 is more than 2-fold- the differences between the two Aquaria are also substantial and at least a third replicate is necessary to resolve this. The same extends to the rest of the analysis. the patterns of AVS between the two aquaria are different too (lines 125-128). 
• In Figure 2 the values of M. viridis at the depth of 8cm is very low while it is substantially higher at the depth of 10cm. This is not in accordance with the plots of Figure 1 for the same depths. 
• In Table 1 the standard error of H. diversicolor in Spring was not the only problematic one. Also for Summer of the same species the standard error is nearly 1/3 of the mean value. Of course, great standard errors are very common when it comes to in situ data but the uncertainty of the field data are combined with the poor experimental data. Therefore, any conclusion drawn in this study is rather precarious. 

In overall the study is interesting, however it suffers from serious lack of data, therefore I cannot support publication of the manuscript. 

Author Response

This is the second round of reviews on the manuscript entitled "Polychaete invasion may lead to biogeochemical change in host marine environment". The authors have made changes to the manuscript, however it is not improved as there are important matters raising mainly by the poor data. More specifically:

• for the experiment two Aquaria are used, the insufficient number of repleicates is obvious in the analysis. In Figure 1 the graphs between the aquaria show very different patterns. This is also shown in the text: "Hotspots of SR ranging from 220 to 539 nmol cm^-3 d^-1 were identified in the center along the burrow walls of the two aquaria with M. viridis down to 16cm depth (Figure 1A and B)." (lines 119-120) and in lines 123-124 "The overall rates of SR in M. viridisaquaria was about 2-fold higher than the control aquarium (Figure 1C).". This means that -since the difference 220 to 539 is more than 2-fold- the differences between the two Aquaria are also substantial and at least a third replicate is necessary to resolve this. The same extends to the rest of the analysis. the patterns of AVS between the two aquaria are different too (lines 125-128). 

We agree and we have explained the problem with low replicates in the text (L68).

• In Figure 2 the values of M. viridis at the depth of 8cm is very low while it is substantially higher at the depth of 10cm. This is not in accordance with the plots of Figure 1 for the same depths. 

The lower values are due to the location of injection ports used for the vertical profiles, where some of the depths were further away from the burrow. We have modified this figure as requested by the editor to show all replicates from each aquarium.

In Table 1 the standard error of H. diversicolor in Spring was not the only problematic one. Also for Summer of the same species the standard error is nearly 1/3 of the mean value. Of course, great standard errors are very common when it comes to in situ data but the uncertainty of the field data are combined with the poor experimental data. Therefore, any conclusion drawn in this study is rather precarious. 

Three replicate sediment cores taken for porewater analysis is a widely applied approach in biogeochemistry studies, since the cores are sliced producing a high number of samples to be analyzed.

In overall the study is interesting, however it suffers from serious lack of data, therefore I cannot support publication of the manuscript. 

We agree that the study suffers with the lack of data, but we believe these findings are an important contribution to advance research on effects of invasive species and ecology of coastal areas.