Seagrasses of West Africa: New Discoveries, Distribution Limits and Prospects for Management
, Salomão Bandeira 2,3,*, Seydouba Soumah 4, Gnilane Diouf 4, Elisabeth Mayé Diouf 5, Omar Sanneh 6, Noelo Cardoso 7, Abubacarr Kujabie 8, Melissa Ndure 4,9, Lynette John 10, Lisdália Moreira 11, Zofia Radwan 11, Iderlindo Santos 2,12, Adam Ceesay 12, Marco Vinaccia 2 and Maria Potouroglou 13
Karine Gagnon
Vasillis Papathanasiou
Round 1
Reviewer 1 Report (Previous Reviewer 4)
Review of revision. Manuscript diversity-2050585 "Seagrasses of West Africa: new discoveries, distribution limits and prospects for management"
Thank you for the very thorough revisions to this manuscript! There have been a lot of great improvements. I still have a few (relatively minor) comments and suggestions to help improve the readability and clarity of the text. Please also carefully proofread prior to resubmission, I noticed some errors in the species names, missing spaces between genus and species, etc. Thank you for your hard work and I look forward to seeing this manuscript published!
Abstract
Line 30: remove the decimal points in the seagrass hectares – that level of precision is not needed in an abstract, especially since you start the sentence with “about”. You could even rewrite it to “at least 62000 hectares…” which would acknowledge the fact that there are probably some unmapped areas.
Introduction
I think the introduction needs a few short sentences introducing the species you are looking for (or at least which species have previously been noted in the area). (Either in the introduction or in the methods section 2.2 – see below).
Table 1
There is an extra horizontal line before the second last entry (Karfaya-Kartong) that needs to be removed.
Methods, Section 2.2.
As I noted above, either here or in the intro, please mention which seagrass species you are mapping. And importantly, which one you are not (Ruppia maritima). See my comments in the results section below as well regarding Ruppia and how it is treating a bit confusingly within the manuscript.
Methods, Section 2.4.
You note that systematic observations reveal pressures on the seagrass. Please elaborate on this – did you notice seagrass loss, or map specific pressures (if so, which and how), etc.
Results, Section 3.1.
As I noted above, I am still a bit confused about Ruppia in this study. In the first sentence of this paragraph it is listed with the other seagrass species. However, further down below in the same paragraph, you note that it is “not documented as part of this study”. But then it is included in the Cabo Verde map as if it were documented. Please clarify!
Results, Section 3.2
In the last sentence you write that the southernmost Z. noltei meadow appears under stress. Can you elaborate on why more slender plants would imply stress (a reference to other studies is likely needed).
Results, Section 3.3
You are missing a unit (ha) after the total seagrass area in the third sentence.
Discussion, Section 4.1.
In the last paragraph, again some confusion about Ruppia. If you did not document it in your study, please start the paragraph with “Though the species was not included in our study, ….”
Discussion, Section 4.2.
In the second paragraph, please standardize the large numbers with spaces between the thousands, instead of commas. This avoids confusion with languages that use commas as a decimal separator.
Also, I assume the 100 000 000 MG C for the whole Banc d'Arguin national park? It could be interesting to also include the figure in C per hectare, to enable comparison to other ecosystems (if you have that data, of course).
In the last sentence, there is a “(1)” in parentheses. Is this an error in reference formatting?
In the third paragraph, the explanation of seagrasses trapping sediment is slightly oversimplified. Sediment and particle trapping from the water column to the sediment is due to the complex ABOVEGROUND structure of the plants (the blades/leaves), while binding the sediment and preventing erosion is due to the complex BELOWGROUND structure (roots and rhizomes). Please clarify this in the text.
In the last paragraph, please elaborate and clarify the stressors. You start with discussing both dust events and temperature, but then only discuss the dust events. For the dust events, please clarify whether these stressors are positive or negative (the first sounds negative, the second is unclear). And discuss how temperature varies and effects seagrass since you introduce it at the beginning.
Discussion, Section 4.3.
In the last section, you discuss the limitations of satellite imagery in regards to limited water clarity. It might be worth elaborating on this slightly and also noting that using satellite imagery is especially limiting for deeper waters where there can be also seagrass meadows that are difficult to detect.
Conclusions.
I think I suggested adding a conclusion in the first review. Thank you for adding it! But I think it is a bit too long. The purpose of a conclusions section should be the highlight the most important results, rather than summarizing everything. I would suggest condensing it to one paragraph only.
Figures
In the figures 3-10 (except Figure 6), could you make the text in the white box slightly larger (i.e. the same as in Figure 6, which is good!).
Also in Figure 10, could you make the dots representing sites slightly larger. It is difficult to see them unless zooming in.
Author Response
Dear Madame/Sir,
Seagrasses of West Africa: new discoveries, distribution limits and prospects for management
By Sidi Cheikh, Bandeira…Potouroglou
Following the reviewer comments to the manuscript diversity-2050585, we hereby
- re-submit an updated version of this manuscript the above mentioned manuscript.
- Attached also a file responding to the reviewer #1 comments.
- Attached a last file responding to the reviewer #2 comments
On behalf of the co-authors,
Sidi Cheikh,
Salomão Bandeira
Maria Potouroglou
Author Response File: 
Author Response.docx
Reviewer 2 Report (New Reviewer)
Seagrasses of West Africa: new discoveries, distribution limits 2 and prospects for management
The authors have gone through a large-scale effort to map seagrass meadows in the West Africa, an area will little existing information. This paper is very interesting and provides original data that need to be published. However, major corrections are needed in the manuscript before publishing, not so much in the methodological approach, but instead to the way this work is presented. It is apparent that the authors have done extensive research in the field, but seems that they are confused as to what to present here. For example, reading the manuscript one understands that mapping through point presence absence with GPS is followed. However, at different points in the manuscript satellite analysis is mentioned. It isn’t clear what exactly they did for the morphological study and how often. Finally, an effort to present their results in a better way should further improve the paper. The authors could also consider within this publication to upload their shapefile to any one of the free repositories, so that their research can have highest impact.
My comments in detail are the following.
Line 30: were documented for the studied… change for to in
Line 32: Zostera noltei in italics
Line 41: you don’t need both coastal and marine!
Line 72: add… early-warning indicators or signals or…!
Line 73: what do you mean bring back? You had management measures before; they were gone and you’re trying to bring them back?
Line 77: I would argue that while we have a lot of published work about ecosystem services and the benefits of seagrass globally, based on model and regression analysis, we are missing the much needed regional and local data. I would say that your work contributes to that end, to providing knowledge about all these ecological mechanisms in a smaller scale!
Line 79: what is the difference between seagrass distribution limits and seagrass distribution?
Line 84: what do you mean near-shallow areas? Give a measurable metric, i.e. 0-25m or how you set it to be!
Line 87: Please provide some more detail here. What do you mean known occurrence? From your previous studies or from interviews with i.e. fishermen?
Line 125: you’re referring to the 10 sites you present in table 1?
Line 146: datapoint were collected not calculated!
Line 163: Do you mean one point for every meadow? Just to get an idea of your areas, how large do you define a meadow? Why not take more points per meadow? With a hand-held GPS you could have taken 10s to 100s easily increasing the accuracy of your mapping!
Line 167: How were the boundaries digitised? With one point collected from the field, when you then go to QGIS, how did you draw limits?
Line 170: Why only in Mauritania? Here you suddenly present the possibility to use a remote sensing approach! Is this something that you did within the work of this paper? If so, you need to present clearly all the steps taken and the exact methodology. Also, why only in Mauritania. If this is some work done in a parallel project or package of your program, mention here that you took these results from another work and reference it.
Line 176- 178: These lines are discussion or introduction and don’t belong in a methodology section!
Line 179-185: This paragraph is poor. I would suggest enriching it. You say for example you measured cover and density…in how many replicates? What depth? You did this once? Further on in the manuscript you imply that you did it monthly! If that is the case describe here your sampling scheme.
Line 192: It is more accurate to say that table 2 shows species occurrence and not distribution!
Line 192: In this paragraph you’re providing the reader with some information about the species themselves! This is not needed and especially in the results section! You could refer to these in the introduction if you like, but it isn’t necessary!
Line 225-231: Sometimes it becomes a bit difficult for the reader to follow through! For example, here, you introduce the concept of a temporal analysis. This is new to the manuscript and confusing! You said previously that the NIT’s revisited the areas and mapped more meadows, however you’re not saying anything about a temporal sampling of the same meadows. If this happened, please provide details in the methodology. If this happened only in some case studies also say it. The way this part is written, it implies that you have data for the period of 2020-2022 (??). Moreover, you can’t talk in an abstract way for things like meadow decline. If you have indeed through mapping measured a decline, give here the numbers i.e. the meadows in Banc d’Arguin lost x ha/m2/Km2… Also, you’re referring to all species found there? Finally, a slight decline in a seagrass meadow (which when not accompanied with an exact number) can be anything from anthropogenically to naturally caused. For example, C. nodosa meadows can have huge coverage shifts for a year only to come back in full the following years. My suggestion is that this slight changes that you say here should be mentioned as slight but not significant changes, and of course…make sure you explain the temporal vector of your analysis…in the next lines you even imply a seasonal (monthly??) sampling scheme!
Line 230: To which species are you referring to?
Line 242: How do you mean that? In which way do they appear less impacted? Did you measure something specific? Any indicators?
Line 244-245: it would be nice to have some numbers here and in general about the extent.
Line 245: Again, the temporal aspect needs to be discussed. 2020-2022 monthly, or when did you measure?
Line 247: Not apparently, since no experimental or measurable data exist, but rather use the word probably!
Line 251: It feels that a lot of weight has been given to the previous countries and these last 3 are being overlooked with just three lines each!
Line 274: this sentence needs re-writing.
Line 275: You should really give more effort to presenting these results in detail. I believe that this is important information. Maybe you could keep it from this small reference and publish it in a different paper, focusing on morphometrical traits of west Africa seagrass species.
Line 277: m2
Line 278: unit in the brackets.
Line 279: leaf length needs units. Also since there is no standard deviation or error next to it, this is the highest leaf length measurement you took? Also, in this paragraph you’re talking about the southernmost Z. noltei meadow, which was found in Senegal. You don’t need to give as here information about Banc d’Arguin just because you have these data!
Line 280: “Potentially under stress” isn’t very scientific and definitely not an expression for results! Give numbers and measurable metrics here. You can then discuss in the discussion the activities in the area that imply in turn a level of pressure for the meadow.
Figure 2. You could if you choose exclude this figure, as it doesn’t add anything to the paper, and the species are known.
Line 301: There is no table 3! Do you mean the one further down the manuscript after table 4?
Line 301: What initial satellite assessment? There is no such thing mentioned in the methodology.
Line 303: 62108 ha!
Line 315-321: This paragraph needs to move to methodology to the description of the remote sensing method after it is written!
Figure 3. The map resolution is poor and also this isn’t a distribution map. Distribution implies polygons! This is the ground truthing map with its results. The same goes for the rest of the images. Even so, I don’t understand since the authors have satellite data (it’s still not clear to me weather or how these satellite data keep popping up and then disappearing!) why not give the polygons? From where did they measure area coverage?
Figure 10. In general, the presentation of the results could be better. This map also has really low resolution making it hard to read. I’d suggest a unified map of the area of high resolution with the polygons that were produced by this study. You could accompany the mixed map with maps of each area in a combined figure.
Line 366: Sea turtles disturb seagrasses?
Line 376: Why is climate change something that troubles only Banc d’Arguin?
Line 394: Table 3 needs to be Table 4 needs to be formatted. Also there is no Figure 1 higher up. The one in page 16 showing different densities is redundant and you can delete it.
Line 407-: What seagrass extent assessment???when did you do that? It’s not mentioned before! Especially from 2015! Anything that you borrow for your work that belongs to another project, put it in the discussion and cite it there! Also these morphometrics that you give here. They are for 2022? When? From previously we get a feeling that you have taken these measurements in different months. So when do these refer to?
Line 428: italics
Line 473: According 63 and comma, not stop!
Line 484-496: This paragraph is for the introduction.
Line 502-509: The methodology of mapping prediction is irrelevant and should be deleted from here.
Line 549: Now you mention 2016! It is very confusing and one can’t understand what it is you exactly did!
Comments for author File: Comments.docx
Author Response
Dear Madame/Sir,
Seagrasses of West Africa: new discoveries, distribution limits and prospects for management
By Sidi Cheikh, Bandeira…Potouroglou
Following the reviewer comments to the manuscript diversity-2050585, we hereby
- re-submit an updated version of this manuscript the above mentioned manuscript.
- Attached also a file responding to the reviewer #1 comments.
- Attached a last file responding to the reviewer #2 comments
On behalf of the co-authors,
Sidi Cheikh,
Salomão Bandeira
Maria Potouroglou
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report (New Reviewer)
As previously discussed, this is a very important project and worth publishing. The revised manuscript is significantly better and a great effort to work around any issues has been made. Two issues remain before it can be published, that I raised in the previous review that haven’t however been addressed, and while they are simple, they form major issues for publication.
The authors have made use of remote sensing techniques. If so, they absolutely need to explain within the manuscript the whole process of the satellite mapping, as done in similar publications. One cannot just say I used remote sensing and just present results. If this is presented within a different publication and the authors just use the results here, they also need to cite that work. If not, then the authors need to add a paragraph in the methodology, explaining how from the 185 points that the NIT’s gathered they went to draw polygons in QGIS and present actual cover values in hectares. This is a major issue with this work. The results are great and important but more effort is needed in order to explain to the public how the results were acquired!
Secondly, like said in the previous cycle, within the monitoring methodology a series of measured parameters are numbered (line 300). However, only a small bit of results is presented for these. I would expect to have either a table (like table 3 that is deleted!) or a combined figure showing these results. Else, there is no need to mention them in the methodology and the authors can refrain to saying that further measurements per meadow were conducted for use in future research! Or nothing at all! Also, it isn’t clear if these measurements were seasonal. Further down the manuscript the authors conduct seasonal analysis. Please provide further details!
Line 276: These three lines here are a completely new and different research paper! If remote sensing techniques were followed, then the authors need to give within the manuscript the complete methodology used. Which products were analysed? What kind of process, classification methodology etc. There is a reference 97 here but I can’t understand how this reference is associated with this paragraph. If the remote sensing mapping has been published within a different paper, then of course the correct reference here would suffice.
Line 306: Check the sentence.
Line 379: You haven’t mentioned anything about seasonal sampling before… Add some text in methodology explaining this. So far, the reader understands that you went in one effort to 185 points to collect your data, then a second to get supplementary data and then you had your monitoring program. Where in this scheme was seasonality, so that you can then present seasonal dynamics!
Line 489: Again, the mention of remote sensing techniques. As I stressed in the previous review cycle, you can’t just use it as something trivial! You need to strengthen the manuscript and explain how you did a remote sensing mapping!
Author Response
To MDPI Diversity,
Dear Madame/Sir,
Seagrasses of West Africa: new discoveries, distribution limits and prospects for management
By Sidi Cheikh, Bandeira…Potouroglou
Following the reviewer #2 comments (ROUND 2) to the manuscript diversity-2050585, we hereby
- re-submit an updated version of this manuscript the above mentioned manuscript;
- Attached a last file responding to the reviewer #2 comments, Round #2;
- Added a letter, addressing managing editor comments;
On behalf of the co-authors,
Sidi Cheikh,
Salomão Bandeira
Maria Potouroglou
Author Response File: Author Response.docx
Round 3
Reviewer 2 Report (New Reviewer)
The authors have adequately addressed all my concerns and comments. As such I believe that this work can now be accepted for publication.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
This is an important contribution to our knowledge of seagrases distribution and diversity in Western Africa – a region with very little information. My only major criticism of the manuscript is that it is just too long (19 figures) and a little disintegrated – it seems like maybe different authors wrote different parts. I would suggest consolidating and reducing the text and figures and moving some figures to supplementary material (see below).
Another issue I have is with the treatment of Ruppia maritima. I missed it being listed right at the beginning in the abstract and then in the results but then it appears briefly in the discussion. Personally I consider it to be a seagrass, suffering all the vulnerabilities and providing all the ecosystem services of the other species, so should be included in this study, but if the authors do not consider it a seagrasss then I suggest this be clearly stated in the introduction or methods and any mention removed from the discussion.
I think it should be stated somewhere that H. wrightii the most widespread species is a fast growing pioneer which can actually benefit from some disturbance (for example if later successional seagrasses or algae are impeded by the disturbance that H. wrightii can withstand) so it is quite robust.
There are quite a few typos and the English although totally understandable could benefit from a revision.
Other points:
Line 34 Define important.
LIne 38-9 Better: Extensive meadows of Zostera noltei were recorded for the first time at Saloum Delta, Senegal, the southernmost distribution limito f the species.
Line 77 correct “avaialable”
Results – again why no Ruppia maritima?
Lines 334-343 revise for symbols and format of the latitudes and longitudes
Line 334 correct “raching”
The Results is far too long and far too many figures – I would place Figs. 5-11 as supplementary material. Figs. 13 and 14 could be made into a single plate (a and b); Fig. 17 is unnecessary – the data in it could be described in the text e.g. “....lack of knowledge (34%), prioritization (19%), and unregulated or destructive fishing (16%)...”
Fig. 16 Sargassum should be capitalized and italic
Lines 484-485 – it would be better to state the name of the type of fishing gear used
Line 504 “While seagrass occurrence in Guinea Bissau is new” this needs rewriting - the occurrence is not new, the scientific record is.
Lines 624-630 Repetitive- This paragraph could be reduced or removed as C was not measured in the study (nor is seagrass biomass presented).
Author Response
We are pleased to submit our responses addressing the editorial comments to the manuscript on “Seagrasses of West Africa: new discoveries, distribution limits, ecosystem services and prospects for management”, which needed a Major Revision.
This cover letter explains, point by point, the details of the revisions carried to the manuscript and our responses one-by-one to the four referees’ comments.
Below we list all reviewer’s comments in bold and our response in normal font. Track Changes were used throughout the attached manuscript. We also checked the relevance of all references and one (1) referee was deleted: Lebrun, J.-P., Stork, A.L. Catalogue des plantes vasculaires de Mauritanie et du Sahara occidental. 1997.
but five more added, some of them suggested by the reviewers.
The authors would like to thank the Editor and the Reviewers for their detailed and insightful comments, and we hope that these improvements have adequately addressed all the concerns.
We look forward to hearing from you at your earliest convenience.
On behalf of all the authors, we are grateful for your excellent input once again.
With our sincere thanks and best regards,
Sidi Sheick, Salomão Bandeira, Maria Potouroglou
Author Response File: Author Response.docx
Reviewer 2 Report
There is an urgent need for an inventory of seagrass species and a map of seagrass meadow extent for the West African region and I applaud the authors for raising capacity and highlighting the importance of seagrass ecosystems to the region. Although the paper presents some important findings and raises some interesting points, I cannot recommend it for publication.
Overall, the paper needs to be more concise and focused. The methods section is exceeding poor, the results are unsupported by the methodology applied. The discussion has some interesting points, but could be presented more succinctly. I found the mapping approach poorly described and results lacking sufficient evidence to support the massively extrapolated distribution maps – these do not provide any confidence in the maps produced.
The significant issues with the present article include:
2. By the authors own admission, actually 4 seagrass species occur in the region – however the authors have chosen to omit records of Ruppia reported in the literature/herbarium collections.
2. The methodology for collecting and describing the perceptions needs to be explained.
3. The discussion is interesting, but very long and could be helped by reducing text and focusing more directly on discussing the results
4. Of biggest concern is the “predictive” seagrass mapping, of which there are a number of serious issues which raise red flags:
· Using composite images from Sentinel-2 to conduct the classification. Models are sensitive to the specific image the classification is tied to, as it learns directly from pixel values. Additionally, composites like this could end up pooling imagery acquired months apart, making it virtually impossible to match it to the field validation data.
· The lack of sufficient field validation data. 101 field data points for a 290625 km2 area is not appropriate and should negate conducting such a mapping exercise. Additionally, there is a huge data imbalance with 86 points being seagrass present to 15 for absent. As a consequence, the output is very unreliable; and the accuracy value cannot be trusted as it is based on the unbalanced and low number of points. By increasing the number of points by simply guessing based on the imagery, plus “very high certainty knowledge”, is not good enough. As a result, the 90/10% training the authors used to validate reveals that their accuracy assessment is based on less than 10 real field validation points and 140 “guessed” ones.
· The post-processing seems very manual and user-driven. Although the authors detail the ArcGIS tools they used, this is not really the issue. What is not transparent is the decision making process they used for removing or reclassifying what they judge as incorrect. The authors describe “iteratively by zooming into every region to make sure all incorrectly identified seagrass areas had been erased”, which seems incredibly laborious. I don’t understand the point of developing a model to create the map and then manually drawing the map. Why not just manually draw the map in the first place?
· There is no mention of seagrass percent cover ranges for their field validation data or the lowest level of seagrass cover detectable from the imagery. For example, were the authors able to detect seagrass less than 25% cover?
· After acknowledging the predictive mapping is very rough and probably inaccurate, the authors return to the field to conduct surveys and replace the original predicted map output with a manually digitized map. Why did the authors not attempt to refine the model further with better imagery for those specific areas?
· The distribution map figures and accuracy assessment results demonstrate the output is not acceptable. On closer examination, the maps/polygons are incredibly crude (could have been more accurately drawn by hand) and the predicted hectares of seagrass reported do not include the potential range or error based on the accuracy assessment or a map confidence value.
All these issues aside, I would suggest the authors constrain their results and findings to the seagrass sites/points examined and not extrapolate beyond. I recommend removal of the figures of seagrass meadows and replace with figures showing the sites and points assessed (using symbology to identify presence and absence of seagrass). I would also recommend the point data be made available either in Supplementary Materials or through a data publisher, e.g. Pangaea.de
The whole document should be carefully checked for spelling errors – there are quite a few. Also, please check for consistency of terminology, e.g. seagrass meadows vs seagrass beds.
I would encourage the authors to re-evaluate their approach and consider publishing the information as two separate articles,
1. A thorough review of seagrass knowledge across the region with improved species distribution and meadow extents, but restrict finding to only the “reference sites” (exclude the predictive mapping), including the assessment of threats and discussion of how management can be improved to conserve seagrasses in West Africa;
2. A more thorough analysis and presentation of seagrass goods and benefits and how they contribute to the livelihoods and wellbeing of the peoples of West Africa.
More specific comments are listed below:
Ln 49. The sentence is a poor start to the paper as it is a bit scrambled and needs re-writing.
Ln 54. What was the reason for the initiative? I think this will help link with the previous sentences.
Ln 58. What is meant by "value"? This term is commonly used throughout and would benefit by a clear definition very early on in the text. It is unclear if the authors are referring to a measured/quantified and assigned economic value as a measure of to how important ecosystem goods and services are to people?
Ln 60 change “protection of” to “food for”
Ln 65-68. The sentence is a bit scrambled and needs re-writing.
Ln 68 significant references are missing, e.g. McKenzie et al 2021. This reference provides a clear evidence base for why seagrass maps are needed for the region.
Ln 78-87. These examples are not relevant in the introduction and should be removed or relocated to discussion
Ln 90 & 93 change “benefits” to “goods and benefits”
Ln 90 Consider replacing “ecosystems supply to humans” with “ecosystems contribute to human livelihoods and wellbeing”
Ln 94 as a “coupled socio-ecological system (SES) framework”
Ln 110 Rather than starting on hydrodynamic processes, consider possibly moving the paragraph from Lines 135 to 140 to the start of section 2.1
Ln 110-127. suggest reducing the hydrodynamic processes as it overwhelms the description of the study area
Ln 128. as the sentence is about MPA, change “sheltered” to “protected”
Ln 138. What is the definition of a “reference site”? I suggest the use of the term "reference site" is not appropriate, as the authors cannot demonstrate the sites represent the desired future characteristics and condition to be achieved, or are representative of the region. Consider replacing with “case study sites” or similar.
Ln 141. Table 1 needs to include details such as length of coastline in each region, intertidal area, population along the coast, lats/longs of each site, etc.
Ln 143. Please include country boundaries
Ln 145-149. Please provide detail on the sites and field expeditions, e.g. what are they (unclear how sites and field expeditions differ) and what were the data collection methods used at each?
Ln 150. How did you determine that there were on two existing records? Did you conduct a literature review – if so, can you please describe how thorough it was and the approach used.
Ln 153-168. The methods are very poorly explained.
• why do the authors consider the methods applied for mapping Posidonia (a structurally large seagrass species) in the Mediterranean (clear waters) are appropriate for West Africa, where structurally smaller seagrass species occur in optically complex waters? The study by Traganos et al (2018) demonstrated that although dense Posidonia meadows could be successfully mapped in clear waters using Copernicus Sentinel-2 and GEE, the less dense Cymodocea nodosa which occurred across the study region could not? Therefore, it follows that this approach is not demonstrated to be appropriate for West Africa?
• Did the authors use Copernicus Sentinel-2 imagery for West Africs, and were the capture dates close to the field validation dates?
• There is such a big focus on GEE in the methods. I understand it is a great tool for dealing with large datasets, but the model is only as good as the input data, which in this case appears exceedingly poor. Wouldn’t that overall give very low confidence in the final map product?
• How were the field surveys conducted? Were field surveys at sites or as part of the field expeditions (see earlier comment). The opportunistic seagrass communities reported along the West Africa are likely to be highly dynamic as they are similar to species in other tropical regions. So the need to couple imagery as close as possible to the field validation is critical.
• What was the accuracy of the field positioning?
• Only 101 field data points, with 86 seagrass present is an exceedingly poor field validation. I cannot see a figure showing the position/spread of field data points in the article??
• As the data is binary (seagrass present and seagrass absent), the number of field data points needs to be approximately equal in each category to be used for training and testing. Such an unbalance has a higher probability of resulting in false positives.
• One data point for every 2877km2 is not sufficient for such a mapping exercise. I am not convinced that the authors can adjust for this by creating seagrass presence/absence points based on visual interpretation of satellite imagery – this is not acceptable! The “predictive map” should be removed, unless the authors can use other covariants such as biophysical maps of sediment type, bathymetry, water clarity, benthic shear, etc.
Ln 227-242. This section if confusing. What is the “ResilienSEA project”? Unclear if there were two field mapping events overall? Why wasn’t this field data used for the remote sensing exercise earlier?
Ln 243. What was the duration and frequency of the monitoring? Table 1 does not identify which sites monitoring occurred? How were the sites chosen and what was the purpose/reason of the monitoring?
Ln 250. This section has a number of issues and needs re-writing.
• Is it possible to please include a copy of the questionnaire (in supplementary material)?
• the methodology for collecting and describing the perceptions needs to be clearly explained.
• How did the authors define an expert?
• What expert-elicitation approach did the authors use?
• Did the authors follow a standardised framework for the services? See assessments by Nordlund et al. (2016) and McKenzie et al. (2021)
• How did the authors define “perceived value”?
• The authors need to provide a clear definition of a threat, e.g. a threat is the harmful effect from a human activity.
• How did the authors determine the perceived level of a threat, e.g. was a vulnerability assessment used?
Ln 280-282. No need to present general species identification characteristics in the results section. For species habitat, can you please include examples of where in the study area the species was found in eulittoral to sublittoral sandy or muddy substrates in both sheltered and exposed locations.
Ln 294. No need to present general identification characteristics for species in results.
Ln 303-305. It would be really helpful in the results if the authors could please include a map/figure for each country, showing location of sites, seagrass points assessed (seagrass present/absent) and with all places listed in the main text labelled.
Ln 306-313. What is the monitoring that started in 2020 and why are no results presented? Please clarify or move paragraph to discussion with citation.
Ln 315-318. Needs a figure with sites and points examined, otherwise lacks evidence base.
Ln 320-22: this is discussion
Ln 330: How was this assessed?
Ln 334-343. Do you mean points? You have only defined “sites” as “reference sites”. This should be shown on a map, rather than lats/longs (see earlier comment about inclusion within maps/figures)
Ln 351-356: the results should only be confined to the specific reference sites, however, they are being presented as though the entire coastal areas was examined and these were the locations of the most extensive meadows. Please re-write to clarify.
Ln 357-360. This paragraph is difficult to understand. A figure which includes the sites and points examined would be really helpful. It is unclear if Seh was a "reference site"? If so, why is it not included in Table 1? Also, unless the other locations examined are points, they are really discussion notes (needing citations) and should not be included in the results.
Ln 362-364: this should be presented in section 3.4 or the discussion, as not a result in section 3.1
Ln 365: Table 3. Why is it only the two locations that are presented. A comprehensive table including similar data for all reference sites would be beneficial.
Ln 366: Figure 2. This is an excellent figure should the differences in density. Can the authors please explain how these low density meadows can be detected by Sentinel-2?
Ln 369-371. How was this change in extent measured? This was not mentioned in the methods section.
Ln 371-372: This would be better presented in a table of results from all reference sites.
Ln 375: How can you be certain this is the southernmost limit of Zostera, as this was only a reference site. Please present in Figure 3 the other points examined as evidence of southernmost limit.
Ln 380-383: This is interesting data and would be better presented in a table with other reference sites. Also, how often was the site examined to make summer and winter observations?
Ln 390: the “predictive mapping” and this whole section should be removed. I also find the error matrix (Table 5) does not truly represent the accuracy of the overall exercise (the model should be restricted to around the field validation points only). The assumption of the error matrix is that it covers all habitats and seagrass species mapped, i.e. is representative. But unfortunately this cannot be possible with only one data point for every 2877km2, i.e. the prediction area is 1000s of times larger than the validation area. The representativeness of the field validation is incredibly low suggesting the extrapolation is inappropriate and the error matrix possibly misleading. To state an overall accuracy of 85% accuracy based on 10% of acquired field point is concerning.
Ln 439: Can figures 13, 14 and 15 be combined into a single figure to make is easier for the reader?
Ln 450 (Figure 16): The threats are a mix of threatening processes and human activities. Although using process terms may be useful in a biological sense, it make the results difficult to analyse as they are not clearly linked to the source of the threatening activity. Terms such as pollution are very general and should be avoided as it often means different things to different people - the authors need to clarify if pollution is: Air pollution (urban, climatic, radioactive); Water pollution (thermal, silt from soil erosion, chemical fertilizers, pesticides, runoff from septic tanks and livestock, stormwater, sewage and other urban wastes); Land pollution (solid wastes —e.g., used cars, cans, bottles, plastics, paper); Noise pollution; Chemical pollution (chlorinated hydrocarbon pesticides, polychlorinated biphenyls, heavy metals). The main threats are so vague that the results provide little used for coastal and environmental managers who want to improve the condition/state of seagrasses. Also, "lack of knowledge" is not a threat -it is a problem, possibly a driver, that results in threatening activities. Similarly, "Prioritization and policy" is not a threat, again it is a problem if it is poor, but can be a solution to removing threatening activities if it is good. The authors would be better to have structured their questionnaire in consideration of a framework such as DPSIR.
Ln 459: I am still not clear what the ResilienSEA project actually is? Much of the following text should be moved to the discussion.
Ln 494: This section should focus on results only and discussion moved to appropriate section.
Ln 547: This is actually more than 13 different services – but the authors have not followed a standardised framework (See Nordlund et al. (2016) and McKenzie et al. (2021)) and many of the services/contributions are grouped as a consequence.
Ln 590 (Figure 19): why are the services listed different from the 13 presented in Figure 18?
Ln 572: The review “The global distribution of seagrass meadows” by McKenzie et al (2020) is noticeably missing, which slightly dates this section.
Ln 589: This is an important finding. Why was Ruppia not included in this study? Ruppia has been classified as a seagrass for many year, and Kees den Hartog was working to ensure its inclusion in the seagrass literature in the later period of his life. I recommend it be included in all future articles for the region.
Ln 608-656: The discussion focusses too much on Banc d’Arguinre, and needs to be more balanced between countries.
Ln 683: Agree that the acquisition of more field data points is critical, but the current lack is not confined to just deep-water or turbid areas.
Ln 705: I cannot see in the results section any data on seagrass monitoring comparing February and July??
Ln 729-733: What was the time of year for each sampling visit? Is the change just due to seasonality or variability within the bay? Was this study conducted at the same sampling position (lat/long)?
Ln 742-745: What are the existing legislative instruments in each of the countries examined?
Ln 748: This section presents information not raised within the main text, e.g. What is NITs.
Author Response
Dear reviewer,
We are pleased to submit our responses addressing the editorial comments to the manuscript on “Seagrasses of West Africa: new discoveries, distribution limits, ecosystem services and prospects for management”, which needed a Major Revision.
This cover letter explains, point by point, the details of the revisions carried to the manuscript and our responses one-by-one to the four referees’ comments.
Below we list all reviewer’s comments in bold and our responses in normal font. Track Changes were used throughout the attached manuscript. We also checked the relevance of all references and one (1) referee was deleted: Lebrun, J.-P., Stork, A.L. Catalogue des plantes vasculaires de Mauritanie et du Sahara occidental. 1997.
but five more were added, some of them suggested by the reviewers.
The authors would like to thank the Editor and the Reviewers for their detailed and insightful comments, and we hope that these improvements have adequately addressed all the concerns.
We look forward to hearing from you at your earliest convenience.
On behalf of all the authors, we are grateful for your excellent input once again.
With our sincere thanks and best regards,
Sidi Sheick, Salomão Bandeira, Maria Potouroglou
Author Response File: 
Author Response.pdf
Reviewer 3 Report
Manuscript ID: diversity-1753413
Title: Seagrasses of West Africa: new discoveries, distribution limits, ecosystem services and prospects for management
The manuscript describes new seagrass beds in the West Africa, and area that has not been very well described compared to other coastal areas. The information is very valuable since it provide the basis for future seagrass conservation and management. Also adds global information of seagrass distribution. The study should make a more clear distinction of what was really mapped and what is predicted to be seagrass. The ecosysem evaluation is usefult to determine the status and threats of seagrass in the region.
Specific comments.
L. 110. The paragraphs starts describing circulation of air masses but then shifts back and forth with seawater flow movement and ocean currents. Please, revise the paragraph to make clear distiction between both air and water circulation patterns.
L. 169. Could you describe with more detail how the ways points were collected? What type of GPS, accuracy, how seagrass was observed (low tide, snorkel, from boat), etc
L262. It would be beneficial for future studies to have access to the full questionnaire used. The questionnaire could be added as Supplementary material. Also could you define what are “Experts”?
Table 4. The figure shows the prediction of seagrass in the area. It would be useful to have another column with the actual mapped area for each region to highlight the amount of seagrass remaining to be fully mapped.
Figure 5 to 11. Please differentiate between what was classified as seagrass after the field surveys and what was predicted as seagrass. Including some groundtruth points in the figure could also be beneficial to see the extent of the surveys and to guide future surveys.
Figures 8, 9, 10, 11, 12. It would be useful for the reader to have in the figure captions the name of the species at each location instead of searching back and for the species present.
Discussion. Would be good to include a few sentences about using drones combined with satellite images in the future as a method in quick expansion for seagrass mapping.
*Huber S, Hansen L, Nielsen L, Rasmussen M, Sølvsteen J, Berglund J, von Friesen CP, Danbolt M, Infantes E, Envall M, Moksnes P-O (2021) Novel approach for large-scale monitoring of submerged aquatic vegetation – a nationwide example from Sweden. Integrated Environmental Assessment and Management, DOI: 10.1002/ieam.4493
Conclusion. Apart for the relevance of the study, would be beneficial to state the main findings of the study too.
Regards, Eduardo Infantes
Author Response
Dear reviewer,
We are pleased to submit our responses addressing the editorial comments to the manuscript on “Seagrasses of West Africa: new discoveries, distribution limits, ecosystem services and prospects for management”, which needed a Major Revision.
This cover letter explains, point by point, the details of the revisions carried to the manuscript and our responses one-by-one to the four referees’ comments.
Below we list all reviewer’s comments in bold and our responses in normal font. Track Changes were used throughout the attached manuscript. We also checked the relevance of all references and one (1) referee was deleted: Lebrun, J.-P., Stork, A.L. Catalogue des plantes vasculaires de Mauritanie et du Sahara occidental. 1997.
but five more were added, some of them suggested by the reviewers.
The authors would like to thank the Editor and the Reviewers for their detailed and insightful comments, and we hope that these improvements have adequately addressed all the concerns.
We look forward to hearing from you at your earliest convenience.
On behalf of all the authors, we are grateful for your excellent input once again.
With our sincere thanks and best regards,
Sidi Sheick, Salomão Bandeira, Maria Potouroglou
Author Response File: Author Response.docx
Reviewer 4 Report
Review for “Seagrasses of West Africa: new discoveries, distribution limits, ecosystem services, and prospects for management”, by Sidi Cheikh et al.
General comments:
First of all, I just want to say that this is a really important paper, and fills in some important knowledge gaps about seagrass distribution (and also acknowledges that there are remaining questions to answer for future research). It is clear that there is a lot of multidisciplinary work, including field research, mapping, and social science here.
I think this study could be a great addition to the seagrass literature, but there are a lot of improvements that could be done to really emphasize the key results. The organisation of some sections could be improved, the methods are missing a lot of detail in most sections, and there are too many extraneous tables and figures that obscure the important results. The discussion also does not emphasize the key results. Please see my detailed comments below, and I hope that these comments and suggestions will help your improve this manuscript and move it towards an excellent paper!
Introduction
I would suggest to re-arrange the introduction a bit: in particular I would combine the first and third paragraphs as they are overlapping topics, then the second paragraph leads nicely into the fourth paragraph. Emphasize here that there is a big knowledge gap about seagrass species in western Africa, and that you are providing these data for the first time. Also emphasize the multidisciplinary nature of this work – it is quite impressive.
Material and methods
There is a lot of information missing from the material and methods section, and it can be difficult to understand what you did.
Section 2.2 is especially lacking in detail! Which methods did you use? How were new seagrass areas discovered? Was there a systematic search system? Did you use satellite mapping, etc.
Section 2.3: The first paragraph is difficult to understand: if there was a script generated from this project, it should be presented or linked somewhere e.g. in the supplement or as an appendix or an open data site. I also have difficult understanding how you predicted seagrass distribution? Is this based on habitat suitability – if so, what were the environmental variables used. I especially do not understand the “turbidity” rating.
Section 2.4: This is fine, but these results are presented mixed up with the results of section 2.2, so it is difficult to find the information.
Section 2.5: It would be helpful to add more details on the questions and interview topics in the supplement/appendix. Please also provide more information about the “descriptive statistics” used and the information on threats (was that also from interviews or was this a different thing).
Results
The results section is quite long and it can be difficult to find the information and results. In addition, there are just too many figures and tables, some are less relevant and obscure the main results. See my separate comments for the tables and figures below. I note that this section needs to be re-organised, and I provide some ideas and suggestions, but feel free to come up with other ideas of how to do this!
Section 3.1: This section is quite long and descriptive and it seems to combine two parts of the methods: sections 2.2. and 2.4? Please separate these out, and the results of the morphometrics could be instead presented in a table form (replacing some of the other tables). It could also be helpful to provide subsections e.g. 3.1.1. “species found”, 3.1.2. “seagrasses in different countries”, etc. Then section 3.2. could also be combined into the Senegal subsection, as it does not seem necessary to have a separate section just for this single result.
I also wonder if it is possible to add the points where the different species were found to the distribution maps of section 3.3. This would be much more informative than just listing site names
Section 3.3. In the first paragraph, please remove the total areas for each country from the text, as they are provided in the table below.
Section 3.4. Again this section is quite long, especially the long paragraph on Page 19 is also really hard to read –please divide into at least 2 paragraphs. Also, please clarify the sources of these results – are these all from the interviews and surveys? Or is there also included some previously known results, which should be cited. Otherwise, please add to the methods section how you quantified these threats?
Section 3.5. Again references are missing here, particular in regards to known biodiversity using the meadows. As in section 3.4, please clarify that this is all local data reported from the interviews? Because it seems but at least some seems to be previously published data which needs to be cited? e.g. lines 501-503 about gastropods, and lines 520-524, 528-535, and 535-358.
For both sections 3.4 and 3.5, if this is a mix of data reported from this study, and previous results, some of that should instead be in the discussion, as supporting your results.
Tables
Table 1: Either remove this or Figure 1, since the information does not need to be shown twice. I suggest keeping this table, and adding the gps coordinates, as well as the seagrass species known in each site, and some references for previous studies that show seagrass in these sites.
Table 2: This table could also be removed, the information is already in the text, and especially if you instead add seagrass species to the habitat maps (Figures 5-11)
Table 3: A bit of context needed here: why was this comparison highlighted?
Table 4: ok
Table 5: perhaps move to supplement
Table 6: A bit more context, why was this highlighted. E.g. to provide a concrete example of how a knowledge gap was filled.
Figures
Figure 1: As mentioned above, either remove this figure or Table 1, as they are showing the same information. Since this figure is essentially the same as Figure 12, I would suggest removing this one.
Figure 2: See my comment for Table 3. It could instead be added to Figure 4, as examples of high and low density meadows?
Figure 3: Not needed. Highlight on the Senegal map instead.
Figure 4: Its nice to see pictures, but a bit more context would be useful in the figure caption: are these photos presented as examples of seagrass, or are they highlighting particular things.
Figures 5-11: These are the main results! Very nice to see the maps, but please standardise the legends and formatting (e.g. is the scale bar in or out of the map, the format of the Mauritania map is different than all the others, sometimes “borders” is shown as a legend. If national parks are shown, please mention in the captions (for Mauritania and Senegal (I assume PNDS is a national park?). As I mentioned above, please consider showing specific where seagrass was found and the different seagrass species (either on the map or in the caption).
Figure 12: This map is fine, but the colours are not great. The green seagrass can be a bit hard to see. Maybe make the blue lighter and use a darker green? Or make the land and sea black/white and keep the seagrass green? Or some other way of adding a bit of contrast?
Figures 13,14,15: Important background info, but move to supplement
Figure 16: Good, but please add in the caption the number of respondents.
Figure 17: From the caption, it is unclear that this represents a ranking. Please clarify. Maybe this could be combined with Figure 16 as two panels (a,b) to cut down the number of figures.
Figure 18: Good, but some ecosystem services seem to be highly overlapping? E.g. fish nursery and breeding habitats, also e.g. sense of place, cultural and spiritual.
Figure 19: See my comment for Figure 17.
Discussion
Right now there is a lot of overlap between the discussion and results, i.e. the discussion is mostly just restating the results. For example, lines 650-656 should be in results, etc.
I think you really need to focus the discussion on (1) the key results and the knowledge gaps that you have now filled, (2) some discussion of threats and key ecosystem services and how that will help management efforts, (management efforts is mentioned in the title, so it should have more emphasis in the discussion (3) how the methods used here can be used for future research, and (4) what knowledge gaps remain to be filled and new questions that have arisen. There is some of the latter already here (on depth limits, Ruppia maritima, further mapping needs along other countries in western Africa, etc., but these need to be emphasized more, perhaps in their own section.
Furthermore, in the conclusions, reemphasize the key results and implications.
Author Response
Dear reviewer,
We are pleased to submit our responses addressing the editorial comments to the manuscript on “Seagrasses of West Africa: new discoveries, distribution limits, ecosystem services and prospects for management”, which needed a Major Revision.
This cover letter explains, point by point, the details of the revisions carried to the manuscript and our responses one-by-one to the four referees’ comments.
Below we list all reviewer’s comments in bold and our responses in normal font. Track Changes were used throughout the attached manuscript. We also checked the relevance of all references and one (1) referee was deleted: Lebrun, J.-P., Stork, A.L. Catalogue des plantes vasculaires de Mauritanie et du Sahara occidental. 1997.
but five more were added, some of them suggested by the reviewers.
The authors would like to thank the Editor and the Reviewers for their detailed and insightful comments, and we hope that these improvements have adequately addressed all the concerns.
We look forward to hearing from you at your earliest convenience.
On behalf of all the authors, we are grateful for your excellent input once again.
With our sincere thanks and best regards,
Sidi Sheick, Salomão Bandeira, Maria Potouroglou
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
I thank the authors for their perseverance to publish these important findings so as to raise awareness of seagrass and the importance of the good and benefits they contribute to a region of the world where information has been sadly lacking. The text has been greatly improved by focussing and presenting the findings more succinctly. Unfortunately, many of the review comments have not been addressed or essentially ignored in the revised manuscript. Unless the authors can adequately address the issues raised regarding the mapping approach used and the results presented, I cannot recommend publication at this time.
The mapping is a significant component of the paper, however, there remain many issues of concern. To report "About 83 093 ha of seagrasses were documented for the studied region comprising seven countries: Mauritania, Senegal, The Gambia, Guinea Bissau, Guinea, Sierra Leone and Cabo Verde." is misleading. I fully appreciate this is a preliminary mapping exercise to fill a critical knowledge gap on which policy and management can be based. However, the authors need to be transparent that the mapping approach and that the results are producing maps with various levels of confidence. I would be grateful if the authors could please address the issues and reconsider how the mapping results are presented.
An alternative approach the authors could consider is presenting the extent/area values in the context of confidence. For example, confidence could be categorised as:
High confidence = polygons for which there is quantitative field validation data, e.g. data collected at case study sites, herbaria collections, etc.
Moderate confidence = polygons where qualitative data confirms seagrass presence from anecdotal reports by experts or individuals with some level of seagrass knowledge (e.g. can confirm seagrass vs macroalgae).
Low confidence = polygons where presence of seagrass is qualitative and interpreted/modelled from imagery (no field validation).
If this approach was implement, a scientifically valid result would be “we have estimated with Moderate to High confidence about 58 808.02 ha of seagrasses, but possibly 83 093 ha with lower confidence, for the study region comprising seven countries: Mauritania, Senegal, The Gambia, Guinea Bissau, Guinea, Sierra Leone and Cabo Verde.”
The authors should also carefully consider addressing several of the terminologies used throughout. For example, they refer to “predictive” map, however, this could be confused with the mapping conducted by Jayathilake and Costello (2018) who’s model predicted 40,373km2 of seagrass across Mauritania, Senegal, The Gambia, Guinea Bissau, Guinea, Sierra Leone and Cabo Verde. I believe these confusing terminologies can be overcome by focusing on the confidence in the mapping product, rather than using terms such as measured area or predicted area. Also, I suggest the authors review the use of terms such as “seagrass habitat” – this has important implications for management, identifying where habitat is suitable for seagrass to grow and persist, but seagrass may be absent. My understanding from this mapping exercise is it mapped seagrass presence, i.e. meadow areas.
My concerns with the mapping relates to how this data will be used as the baseline against which future mapping will be compared. Aerial differences in future mapping will be used to demonstrate losses or gains, which will have great significance for policy and management in both the region and globally. By being open and transparent in regard to limitations and confidence in the maps, will ensure evidence based policy and management.
Please see attached document replying to the responses the authors have given to my original review comments.
Comments for author File: Comments.pdf
Author Response
Dear reviewer,
On behalf of the co-author team of Diversity1753413, below is our detailed response of the Reviewer comments (round 2).
Sincerely,
Sidi Cheick, Salomão Bandeira and Maria Potorouglou
Author Response File: Author Response.pdf
