Can Rock-Rubble Groynes Support Similar Intertidal Ecological Communities to Natural Rocky Shores?
Round 1
Reviewer 1 Report
This is a nice descriptive study that is of potentially broad practical use for people seeking to design shoreline alterations that are as beneficial as possible for marine biodiversity and abundance. I have put a few 'sticky notes' on the text where more detail is needed - especially descriptions of the types of pools studied, and how quadrat location were chosen -without this information it is not possible to critically evaluate or explain the results. The text seems wordy, especially the Introduction, although not drastically so.
I think it would be useful for the authors to include a paragraph in the Discussion or Conclusions that clearly describes what kinds of studies would be helpful as 'next steps' for this descriptive work. What is the next logical direction to go in answering their initial questions? Is there some experimental work that would shed light on these issues?
Comments for author File: 
Comments.pdf
Author Response
*** Author responses
This is a nice descriptive study that is of potentially broad practical use for people seeking to design shoreline alterations that are as beneficial as possible for marine biodiversity and abundance. I have put a few 'sticky notes' on the text where more detail is needed - especially descriptions of the types of pools studied, and how quadrat location were chosen -without this information it is not possible to critically evaluate or explain the results. The text seems wordy, especially the Introduction, although not drastically so.
***All sticky notes have been addressed. We have also aimed to reduce the text in the introduction to make less wordy.
I think it would be useful for the authors to include a paragraph in the Discussion or Conclusions that clearly describes what kinds of studies would be helpful as 'next steps' for this descriptive work. What is the next logical direction to go in answering their initial questions? Is there some experimental work that would shed light on these issues?
*** A good idea. Along with R2’s comments that suggest highlighting the importance of certain confounding variables, at the end of various paragraphs in the discussion (rather than all in one paragraph), we have included suggestions for future studies that we feel could be of benefit to the community. L384-392, L406-413, L424-427, L470.
*** Thank you very much for your constructive and helpful comments
Reviewer 2 Report
In this manuscript the authors compare species richness, abundances and invertebrate communities in intertidal rock-rubble groyne structures and natural rocky shores, in different regions and intertidal levels. The field work is old (2008) and there are some confounding factors that could be influencing the results, but nevertheless I consider this study well written and broadly interesting to those interested in ecology, biodiversity, and coastal conservation. Therefore, this manuscript is worthy of publication in LAND, with some changes.
General comments:
The main problem of this study is that there are many confounding factors that could be influencing the results: the rock type, pool sizes and depth, tidal regime and wave exposure, presence of ecosystem engineers. Some of them have already been discussed in the text (rock type, tidal regime). I consider rock type the most important, since shape, texture, moist, hardness and friability of the rocks can influence their communities. Physical and chemical substrate properties (such as roughness, hardness, free energy, and polarity) are known to affect organismal adhesion and persistence (see for example: Fletcher RL, Callow ME. The settlement, attachment, and establishment of marine algal spores. Br Phycol J. 1992;27:303–29). You state in the discussion: “We paired sample sites based on proximity rather than rock type, in order to address questions pertaining to the spread of geographic distributions and compare nearby biological assemblages. Therefore, we are unable to separate possible effects of structure type and rock type, ..”. However, I consider rock type is a factor that should have been taken into account, both for groynes and natural rocks. By looking at Supplemental Information 1 I notice not only groynes are all different between sites (limestone, syenite, granite), but also the natural rocks are different (glacial materials, devonian sedimentary, mudstones). In addition, in some beaches rocks appear to be boulders while others are composed of a flat, continuous platform. Differences will be found here, e.g. as boulders create refuges between rocks. In my experience, communities are completely different on a gravel beach and a boulder beach. To avoid rock type as a confounding error, groynes of the same material (and age) would need to be selected among sites; and next to them, rock platforms of similar composition and topography. I understand that this is not always possible, but you should have tried to minimize differences at the most, maybe by choosing a smaller scale (a small region with many similar groynes and similar natural rocky shores). Although it is true that by reducing the scale the question about the spread of geographic distributions would be difficult to answer.
See my comments about the pools under materials and methods section.
Tidal regime and wave exposure: In Sandbanks Peninsula the tidal regime is complicated and different to the other sites. Then why did you choose this site? Ideally, the tidal regime should by similar in all sites. Also the wave exposure, as exposed and protected sites are not the same and many studies show different communities are related to one or the other. You should add a couple of sentences about wave exposure. Are all of your sites exposed? Is it the north more exposed than the South?
Finally, the presence of ecosystem engineers (see Jones et al.1994, and related papers. Jones, C.G.; Lawton, J.H.; Shachak, M. Organisms as ecosystem engineers. Oikos 1994, 69, 373. ) can have an impact on biodiversity. Biodiversity is frequently enhanced by species that structurally modify the environment (ecosystem engineers) and, in so doing, create microhabitats where the impact of abiotic conditions becomes reduced. Oftentimes, many species occur in strict association with the microhabitats engineered by one or a few species. Physically engineered microhabitats are known to be critical for biodiversity in rocky intertidal shores where wave exposure, desiccation, and temperature often exceed species tolerance limits. Therefore, many of the species inhabiting this environment rely on the creation of protected microhabitat by a few tolerant and often spatially dominant sessile species (e.g., mussels, tunicates, seaweeds). They can increase species diversity or abundances enormously. I consider the presence of ecosystem engineers could have a strong influence on your results and should be at least mention. You mention the polychaete Eulalia was always found related to mussels (ecosystem engineer). I imagine many of the shrimps, amphipods and isopods could be related to a specific seaweed, that provides habitat in a harsh environment. Ecosystem engineers can create differences in species richness or abundances at a small scale, in sites were they are present, and mask/lower other factors.
Many interactions between factors seen in the ANOVA analysis could be explained by all these confounding factors.
VERY IMPORTANT: I could not find the references list. This list must be at the end of the manuscript but is not there. I consider this a serious fault, since now the reviewers cannot check the cited information. I have no idea what papers are you talking about and if this information even exists. Especially in this journal, where references are only numbers. Please double check for next time that everything is uploaded (tables, figures, complete manuscript and REFERENCES!), otherwise reviewers can get mad very easy.
Below are comments specific to each section:
TITLE: Personally, I don’t like the actual title. You are emphasizing the fact that you found some species of concern on groynes, but also the number of species in groynes is lower than in natural rocks and you are not saying this. A more neutral title would be better. Perhaps a question.
INTRODUCTION:
The introduction is fine but I found it very long. Try to shorten it a bit. For example, paragraphs 2 and 3 can be merged into one, and there is no need to explain one by one every study conducted at the sites. Try to resume information.
Line 78. “Codium fragile ssp. Tomentosoides” is tomentosoides (all lower case). This is a subspecies name, not a genus.
MATERIALS AND METHODS:
Line 114: Add the distance between natural rocky shores and groynes: … in reasonably close proximity to each other (XX to XX m, Fig 1). Also, Sidmouth and Stokes Bay seem very far away (100 km) from each other compared to the other sites (less than 20 km?). Couldn´t you find a natural rocky shore more closed to Sidmouth? I´m not sure that with this great distance (100 km) these two sites are comparable.
Line 116: Summer 2018? You should specify the sampling date or at least the season; because it is not the same to sample during summer or winter (seasonal species could be present). To minimize variability ideally you should sample all sites within the same month, or at least the same season.
Line 126: “In each sampling section, we also selected three rock pools (of similar sizes across
structures) in the same way, or all rock pools in the section if fewer than four were present.”
I truly doubt that rock pools were of similar sizes, especially when there are few. Did you measure diameter and depth of the pool? Depth is important, because a shallow and deep pool can be very different and host completely different communities. Also, how did you count the animals in the pool? Did you only count the swimming animals, or also the ones (sessile or mobile) on the bottom of the pool? What about the pool walls, did you count animals in there too? Please explain this further. I said this because there could be an error here, since rock pools and flat rock surfaces normally differ in surface area. For example, for sessile animals a rock quadrat and a pool quadrat could differ in richness and abundances not because of the water on the pool, but simply because the pool has more rock surface area. If you measured the size and depth of the pool, you can make corrections in the counts and calculate the densities for the same surface area.
Line 134: ..,”we recorded abundance as percentage cover.” Did you do this by using a grid or simply by sight? Please explain.
Figure 1: Geographic coordinates should be added to the sides of the map.
Data Analysis:
I’m not sure about the DCA analysis. DCA is more used when applied to gradient data (for example, to study succession in plants colonizing a new habit or in paleontological studies reconstructing communities in time-series). Also, DCA corrects the arch effect problem of the CA but some authors say this is not an enough solution.
To study the communities associated to groynes/natural rocks, intertidal levels and the other factors, and taken into account that your dataset includes null values, I would advise you to use Bray-Curtis similarity matrix and nMDS ordination. Then you can perform a PERMANOVA analysis to test for significant differences between groups. For me this is an advantage over DCA, as to my knowledge, no significant tests are available with DCA.
However, the choice of ordination methods depends on 1) the type of data you have, 2) the similarity distance matrix you want/can use, and 3) what you want to say (see debate about this in Researchgate, https://www.researchgate.net/post/How_to_choose_ordination_method_such_as_PCA_CA_PCoA_and_NMDS). With this being said, if you can properly justify why you chose DCA, it is fine for me.
RESULTS
Figure 2 and 3: In the “average abundance” (y axis) plots, you are drawing here together bars for algae, lichen and sessile animals (measured as percent cover) and mobile animals (measured as densities). How can you plot all together? The densities have been converted to percentages some way or are they still densities? Please explain in the captions.
Figure 3: Although this figure has a lot of information, some bars are hardly visible and would need another scale (more numbers in the y axis). If you want to leave it as it is I suggest you to move this figure to supplemental information and enlarge it in more pages.
Tables 1 and 2: To reduce the size of the tables I would suggest removing the P values (all P rows) and adding asterisks next to the significant F values (P<0.5*, P<0.01**).
Line 262: “Overall, this analysis suggested only a slight differentiation between communities on groynes versus natural rocky shores,…” This is only an inference. With PERMANOVA you could statistically test if the centers (centroids) of distance matrices differ.
Sup. Information 4: Labels are wrong. Scientific name goes in the second column and common name in the third. Common and scientific name are by no means synonyms. The scientific name (or Latin name) is composed by the Genus and specific epithet of the species. The common name refrains to how the animal is called locally (in a specific region) by the people.
Instead of DECORANA analysis you should write DCA analysis.
“DCA is sometimes erroneously referred to as DECORANA; however, DCA is the underlying algorithm, while DECORANA is a tool implementing it. (https://en.wikipedia.org/wiki/Detrended_correspondence_analysis).
DISCUSSION
Lines 353-358. You only think of trophic interactions here, but algae such as the sea oak and Corallina officinalis can be considered ecosystem engineers (sensu Jones 1994), providing refuge and habitat for fishes.
Lines 359-366. The rock type could be the answer here, as barnacles normally prefer hard rocks.
Lines 379-393. You say the age of an artificial structure is important for the development of an ecological community. With this in mind, and taking into account that the sampling date in this study is quite old (2008), it would be nice to repeat the sampling in recent years and check whether with older structures the communities are still the same or they have changed.
Author Response
*** Author Responses
General comments:
The main problem of this study is that there are many confounding factors that could be influencing the results: the rock type, pool sizes and depth, tidal regime and wave exposure, presence of ecosystem engineers. Some of them have already been discussed in the text (rock type, tidal regime). I consider rock type the most important, since shape, texture, moist, hardness and friability of the rocks can influence their communities. Physical and chemical substrate properties (such as roughness, hardness, free energy, and polarity) are known to affect organismal adhesion and persistence (see for example: Fletcher RL, Callow ME. The settlement, attachment, and establishment of marine algal spores. Br Phycol J. 1992;27:303–29). You state in the discussion: “We paired sample sites based on proximity rather than rock type, in order to address questions pertaining to the spread of geographic distributions and compare nearby biological assemblages. Therefore, we are unable to separate possible effects of structure type and rock type, ..”. However, I consider rock type is a factor that should have been taken into account, both for groynes and natural rocks. By looking at Supplemental Information 1 I notice not only groynes are all different between sites (limestone, syenite, granite), but also the natural rocks are different (glacial materials, devonian sedimentary, mudstones). In addition, in some beaches rocks appear to be boulders while others are composed of a flat, continuous platform. Differences will be found here, e.g. as boulders create refuges between rocks. In my experience, communities are completely different on a gravel beach and a boulder beach. To avoid rock type as a confounding error, groynes of the same material (and age) would need to be selected among sites; and next to them, rock platforms of similar composition and topography. I understand that this is not always possible, but you should have tried to minimize differences at the most, maybe by choosing a smaller scale (a small region with many similar groynes and similar natural rocky shores). Although it is true that by reducing the scale the question about the spread of geographic distributions would be difficult to answer.
*** A set of valid points, and something that with the benefit of hindsight we might have tried harder to address. However, the number of rock-rubble groynes around the coast of the UK is limited, meaning that we could probably not have succeeded in controlling rock type anyway, and so we had to focus on the geographic variation rather than substrate. Further, the reality of the situation, around the English coast at least, is that the rock types of the groynes and nearby natural shores do not tend to match, so our sampling is appropriate to the applied questions we address, even if it limits our ability to ascribe differences to the specific distinction between groynes and natural rocky shores. Thus part of the groyne effect may be rock type. It may be worth extending the reasoning to absurdity: if we were able to find groynes that are identical to natural rocky shores in ALL respects (including age) then there would be no distinction and we would find no differences in the biological assemblages living on them. That would be pointless. What we have done is found a groyne effect, and that is an important step. Further work is needed to understand this effect fully, and that would include investigating the influence of rock type. Unfortunately, we cannot make progress on that point in this manuscript because including rock type as a variable in any analysis of our data does not provide meaningful information (no replication, and confounded with other variables). Instead, we have now emphasised the role of rock type more explicitly in the discussion (L338-341) and tempered some of our statements accordingly. We have also added the suggested reference.
Tidal regime and wave exposure: In Sandbanks Peninsula the tidal regime is complicated and different to the other sites. Then why did you choose this site? Ideally, the tidal regime should by similar in all sites. Also the wave exposure, as exposed and protected sites are not the same and many studies show different communities are related to one or the other. You should add a couple of sentences about wave exposure. Are all of your sites exposed? Is it the north more exposed than the South?
*** We have now added a few sentences and references on exposure and aspect in the methods (L129-144) and discussion (L441-454). These were considered in the study design and our method and our discussion now reflects this.
*** With regard to the selection of Sandbanks, we chose this site as it has been the focus of the only two previous studies exploring the ecological communities of groynes, and the fact that these studies found contrasting results in the same study area (this was mentioned in the introduction). Consequently, we felt it warranted further research. We have re-emphasised this in our discussion (L364-368).
Finally, the presence of ecosystem engineers (see Jones et al.1994, and related papers. Jones, C.G.; Lawton, J.H.; Shachak, M. Organisms as ecosystem engineers. Oikos 1994, 69, 373. ) can have an impact on biodiversity. Biodiversity is frequently enhanced by species that structurally modify the environment (ecosystem engineers) and, in so doing, create microhabitats where the impact of abiotic conditions becomes reduced. Oftentimes, many species occur in strict association with the microhabitats engineered by one or a few species. Physically engineered microhabitats are known to be critical for biodiversity in rocky intertidal shores where wave exposure, desiccation, and temperature often exceed species tolerance limits. Therefore, many of the species inhabiting this environment rely on the creation of protected microhabitat by a few tolerant and often spatially dominant sessile species (e.g., mussels, tunicates, seaweeds). They can increase species diversity or abundances enormously. I consider the presence of ecosystem engineers could have a strong influence on your results and should be at least mention. You mention the polychaete Eulalia was always found related to mussels (ecosystem engineer). I imagine many of the shrimps, amphipods and isopods could be related to a specific seaweed, that provides habitat in a harsh environment. Ecosystem engineers can create differences in species richness or abundances at a small scale, in sites were they are present, and mask/lower other factors.
*** Another good point – thanks! We have now explicitly identified ecosystem engineers in our discussion and added references. We aimed to explore these interactions through the community analysis rather than confound the statistical models further by adding proxies for biotic interactions into the parameterisation. We have cited some additional references to support this decision, and make explicit reference to the points you raise, in the results and discussion sections (L406-413).
Many interactions between factors seen in the ANOVA analysis could be explained by all these confounding factors.
*** Through all the above text, we hope we have now more fully addressed the possibility of confounding factors on the statistical analysis
VERY IMPORTANT: I could not find the references list. This list must be at the end of the manuscript but is not there. I consider this a serious fault, since now the reviewers cannot check the cited information. I have no idea what papers are you talking about and if this information even exists. Especially in this journal, where references are only numbers. Please double check for next time that everything is uploaded (tables, figures, complete manuscript and REFERENCES!), otherwise reviewers can get mad very easy.
*** How frustrating! We recognise that the reference list is very important, not least in the review process. This appears to be some sort of system error because they don’t appear on Referee 1’s commented PDF either, and yet the references were included in the manuscript we submitted, and are currently showing on the author download version of the manuscript. We also went through the painstaking process of manually ordering them (as we don’t trust Endnote or Zotero), and reordering these with the new additions, so a lot of work did go into them. We will inform the editor of this when we submit the revisions, but if they don’t appear please do contact the journal and they can send over them over as they have been included
TITLE: Personally, I don’t like the actual title. You are emphasizing the fact that you found some species of concern on groynes, but also the number of species in groynes is lower than in natural rocks and you are not saying this. A more neutral title would be better. Perhaps a question.
*** The title has now been changed to be less controversial
INTRODUCTION:
The introduction is fine but I found it very long. Try to shorten it a bit. For example, paragraphs 2 and 3 can be merged into one, and there is no need to explain one by one every study conducted at the sites. Try to resume information.
*** We have edited the introduction to remove unnecessary words. We hope that it is now more concise and focused.
Line 78. “Codium fragile ssp. Tomentosoides” is tomentosoides (all lower case). This is a subspecies name, not a genus.
*** this has now been removed with the editing of the Introduction
MATERIALS AND METHODS:
Line 114: Add the distance between natural rocky shores and groynes: … in reasonably close proximity to each other (XX to XX m, Fig 1). Also, Sidmouth and Stokes Bay seem very far away (100 km) from each other compared to the other sites (less than 20 km?). Couldn´t you find a natural rocky shore more closed to Sidmouth? I´m not sure that with this great distance (100 km) these two sites are comparable.
*** These distances have been added to Table SI1
*** We have now also explicitly addressed the distance between Stokes and Sidmouth in the discussion (L384-392). This beach represented the closest appropriate, safe and publicly accessible beach, and we have now clarified this.
Line 116: Summer 2018? You should specify the sampling date or at least the season; because it is not the same to sample during summer or winter (seasonal species could be present). To minimize variability ideally you should sample all sites within the same month, or at least the same season.
*** Good point. We have now specified the season
Line 126: “In each sampling section, we also selected three rock pools (of similar sizes across
structures) in the same way, or all rock pools in the section if fewer than four were present.”
I truly doubt that rock pools were of similar sizes, especially when there are few. Did you measure diameter and depth of the pool? Depth is important, because a shallow and deep pool can be very different and host completely different communities. Also, how did you count the animals in the pool? Did you only count the swimming animals, or also the ones (sessile or mobile) on the bottom of the pool? What about the pool walls, did you count animals in there too? Please explain this further. I said this because there could be an error here, since rock pools and flat rock surfaces normally differ in surface area. For example, for sessile animals a rock quadrat and a pool quadrat could differ in richness and abundances not because of the water on the pool, but simply because the pool has more rock surface area. If you measured the size and depth of the pool, you can make corrections in the counts and calculate the densities for the same surface area.
*** This has now been addressed with explanation in the methods (L129-144) and more description in the discussion (L348-352).
Line 134: ..,”we recorded abundance as percentage cover.” Did you do this by using a grid or simply by sight? Please explain.
*** We used a grid. This has now been clarified.
Figure 1: Geographic coordinates should be added to the sides of the map.
*** Added
Data Analysis:
I’m not sure about the DCA analysis. DCA is more used when applied to gradient data (for example, to study succession in plants colonizing a new habit or in paleontological studies reconstructing communities in time-series). Also, DCA corrects the arch effect problem of the CA but some authors say this is not an enough solution.
To study the communities associated to groynes/natural rocks, intertidal levels and the other factors, and taken into account that your dataset includes null values, I would advise you to use Bray-Curtis similarity matrix and nMDS ordination. Then you can perform a PERMANOVA analysis to test for significant differences between groups. For me this is an advantage over DCA, as to my knowledge, no significant tests are available with DCA.
However, the choice of ordination methods depends on 1) the type of data you have, 2) the similarity distance matrix you want/can use, and 3) what you want to say (see debate about this in Researchgate, https://www.researchgate.net/post/How_to_choose_ordination_method_such_as_PCA_CA_PCoA_and_NMDS). With this being said, if you can properly justify why you chose DCA, it is fine for me.
*** We spent some time rerunning these analyses using nMDS. However, due to the long computation time, coupled with the lack of model convergence and the compounding impact of rare species made interpreting and exploring these results difficult and made the ordination plot more complicated to interpret. Overall, results were largely similar, with the exception of the extreme outliers of Idotea and Gammarus spp. To overcome this issue we had to remove these species from the analysis, opening our analysis up to a different set of problems. Therefore, we have opted to stick with the DCA. We have provided further justification for this (L177-185), citing the above reasons, but also the underlying unimodal model and the gradient length – the tidal range across which we sampled (in particular) means that our gradient is quite long relative to species’ tolerances. Further, the ordination plot gives no indication of suffering from an arch effect. Finally, we acknowledge the comment about ‘gradients’, and argue in the discussion (and in line with your comments on the compounding factors) that some of our results are perhaps explained by gradients that have been classified, and suggest this as a future avenue of research.
RESULTS
Figure 2 and 3: In the “average abundance” (y axis) plots, you are drawing here together bars for algae, lichen and sessile animals (measured as percent cover) and mobile animals (measured as densities). How can you plot all together? The densities have been converted to percentages some way or are they still densities? Please explain in the captions.
*** The abundance for algae, sessile and lichen are percentage cover and the abundance for mobile animals are raw counts. This is now explained in the caption. Given the values, we don’t see the need to create two Y-axes as this might only complicate the readability of the figure, and the point is not to compare abundances between species types, but between groynes and natural shores (and between tidal levels and counties)
Figure 3: Although this figure has a lot of information, some bars are hardly visible and would need another scale (more numbers in the y axis). If you want to leave it as it is I suggest you to move this figure to supplemental information and enlarge it in more pages
*** We have now split this figure. We have retained species richness in the main text and have moved abundance to supplementary information
Tables 1 and 2: To reduce the size of the tables I would suggest removing the P values (all P rows) and adding asterisks next to the significant F values (P<0.5*, P<0.01**).
*** Changed as advised, except that *indicates P<0.05, not 0.5. Tracked changes have been accepted, here, to allow readability
Line 262: “Overall, this analysis suggested only a slight differentiation between communities on groynes versus natural rocky shores,…” This is only an inference. With PERMANOVA you could statistically test if the centers (centroids) of distance matrices differ.
*** See above comments related to DCA
Sup. Information 4: Labels are wrong. Scientific name goes in the second column and common name in the third. Common and scientific name are by no means synonyms. The scientific name (or Latin name) is composed by the Genus and specific epithet of the species. The common name refrains to how the animal is called locally (in a specific region) by the people.
*** Thanks for spotting this mistake, which we have now corrected
Instead of DECORANA analysis you should write DCA analysis.
“DCA is sometimes erroneously referred to as DECORANA; however, DCA is the underlying algorithm, while DECORANA is a tool implementing it. (https://en.wikipedia.org/wiki/Detrended_correspondence_analysis).
*** True. Changed
DISCUSSION
Lines 353-358. You only think of trophic interactions here, but algae such as the sea oak and Corallina officinalis can be considered ecosystem engineers (sensu Jones 1994), providing refuge and habitat for fishes.
*** Text altered so it now reflects both trophic levels and ecosystem engineers
Lines 359-366. The rock type could be the answer here, as barnacles normally prefer hard rocks.
*** Thanks. We have now provided some suggestions for barnacle distributions
Lines 379-393. You say the age of an artificial structure is important for the development of an ecological community. With this in mind, and taking into account that the sampling date in this study is quite old (2008), it would be nice to repeat the sampling in recent years and check whether with older structures the communities are still the same or they have changed.
*** Yes, that would be a good idea! However, the realities of both authors’ current situations preclude this. We have mentioned this in the final paragraph of the discussion, and provide our data to enable researchers to make comparisons in the future.
*** Thank you very much for your thorough, thoughtful and well-informed comments and suggestions. We hope we have addressed them to a suitable standard, and are grateful for your role in improving our paper
Reviewer 3 Report
I found this to be a well considered study and the paper to be presented clearly and thoroughly. The work is relevant in terms of both non-native invasive species and the conservation of coastal ecologies under threat. The data are effectively described and analysed and the conclusions sound.
Author Response
I found this to be a well considered study and the paper to be presented clearly and thoroughly. The work is relevant in terms of both non-native invasive species and the conservation of coastal ecologies under threat. The data are effectively described and analysed and the conclusions sound.
*** Thank you very much for your comments
