Tracking the Expansion of Sonneratia apetala and Its Impact on Local Mangroves Using Time-Series Remote Sensing Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Authors present the expansion of Sonnertia apetala in Leizhou Peninsula, China based on time-series remote sensing data. (1) The introduction of the manuscript is too long, the authors should focus more on why study Sonnertia apetala and the meaning of this study.  (2) This manuscript needs polishing, some words are not clearly spelled, for example L189, L425. (3) The descriptions of the classification method and the process is not clear, more details are needed. (4) L232: what software is used here for sampling tool, data management module. (5) Why you use RMSE for classification accuracy verification? Researchers often use Kappa, overall accuracy for accuracy verification. (6) Here you use medium-resolution remote sensing imagery for Sonnertia apetala detection, but is the detection results reliable? since many researchers directly used high-spatial resolution satellite imagery to monitor the mangrove expansion. (7) Why you choose the four metrics for landscape pattern change analysis? 

Comments on the Quality of English Language

 This manuscript needs polishing.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Author,

The manuscript is interesting, well-designed, studied, and well-written. But several points could be improved. Please use italics for the species name.

Abstract:

27-30: Are facts (key findings) numbers 1-3 proven related? Please describe briefly before concluding in lines 33-34: These findings demonstrate that Sonneratia apetala is encroaching upon local mangrove habitats.

Introduction:

80: Sonneratia apetala, an exotic tree species introduced to China. Where was the S. apetala from? Was it from another place in China?

90-109: Since 1998, S. apetala has been widely utilized as an afforestation species, expanding along the coast of South China with a total planting area of 3,800 hm². Studies revealed the negative impact of this planting (Reference 23-27). This study uses an image from 2010-2023; comparing it with the image before 1998 is better. 

Materials and Methods:

175-176: The mangroves on the Leizhou Peninsula are predominantly composed of artificially planted S. apetala communities. Please describe the history of its planting and when it started and stopped.

Results and Discussion:

289-308: In 2010-2023, did the increasing growth of mangroves include the S. apetala? If it did, was this species planted during this period, or is the increasing coverage because of its natural growth?

316: There are interesting data on 2013, 2015, 2016, 2017, 2019, and 2020. There is a temporal change in which they change every two years. Why did it happen? Is there any planting or other possible reason?

327-345: Please also explain other local species' growth or existence for comparison.

350-351: Invasive species typically exhibit strong tolerance, high competitive ability, and rapid growth rates. Does S. apetala also have these characteristics? Please explain.

371-374: What is the possible cause of this phenomenon where the local mangrove is able to compete with S. apetala?

375-384: Has S. apetala been planted in an area where local mangroves cannot grow? But did it then expand to the location where local mangroves grow?

387-395: But, in several years (see Figure 7), there was decreasing growth of S. apetala; why?

397-420: Was this phenomenon (Table 4 and Figure 12) in line with Figure 7?

452: The local mangroves were unable to adapt to the changing environment; please briefly describe these changes.

453-462: Reference 59-61 said that there is human interference and natural disaster, which means that it is not merely caused by the natural characteristics of S. apetala that make it dominant in the study area. What do you think about that?

Conclusion.

464-495: The conclusion could be improved, made shorter, and answered the study's hypothesis and objectives.

Best regards,

Reviewer

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Dear Authors, 

The quality of English language is good, but it could be improved.

Best regards,

Reviewer

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This is a study that tracked the growth of the invasive species Sonneratia apetala and its impact on local mangroves on the east coast of the Leizhou Peninsula in China. The study used Landsat and Sentinel-2 satellite data, along with the XGBoost machine learning algorithm, to classify mangroves with native and invasive species. The highlight and criticism of the work lies in the methodology, since the authors use a combination of Landsat and Sentinel-2A satellite data, historical Google Earth images and drone data (strengths), and the combination of data from different sources and resolutions is questioned, as the methodology has some steps that need to be clarified for a more complete and robust analysis of the study area.

The work deals with graphic elements, figures, tables, and formulas. All of them are explored in the text with clarity and precision.

I move on to some points that left me with questions.

1 - I think it would be interesting to contextualize how Sonneratia apetala was implemented in the region? Although it has ecological dominance, isn't there any information about this plant's history? Due to the preference for this species in the region? Did the implementers know it was an exotic species?

2 - In the processing stage. It is clear that the XGBoost algorithm was implemented using Python software, but the steps used are not sufficiently detailed. I am referring to better explaining the parameters and configurations. Particularly, I only know about the algorithm's application for numbers, I believe other readers will have the same curiosity about its use for images.

3 - Although the authors have expressive results with RMSE, the study does not mention other common precision metrics in image classification analyses (for example, F1-score). I reiterate, RMSE is an effective precision metric and is appropriate for the context of this study. However, it does not provide detailed information about the types of errors committed, such as false positives (when a pixel is incorrectly classified as belonging to a category), false negatives (when a pixel belonging to a category is incorrectly classified as not belonging to it), omission errors (when a class is left out of the analysis), or commission errors (when an area is incorrectly included in a class). Would it be possible to include this metric and discuss it?

4- In the discussion, I suggest a critical analysis of data validation. Google Earth images, being a composition of different sensors, can present challenges in processing, although this was not a significant problem in the studied region.

Given that the validation data for 2022 was derived from the same sources (drone images and Google Earth) used in model training, it would be important to include a discussion about the potential impact of this bias on classification accuracy and the generalization of results for other years or areas. This analysis would strengthen the methodology and offer an interesting approach for future study replications.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

(1) There is debate whether Sonneratia apetala is invasive, so you need references to support your points when using "invasive species"; (2) the introduction part is still not satisfactory, you lack the summary of how researchers used remote sensing to study Sonneratia apetala expansion, which is the main aim of your manuscript.  In addition, the authors only mention Sentinel-2 data in the last paragraph of the introduction, which is not consistent with the descriptions in "2.2 Data sources and preprocessing";(3) it's strange you mention Google Earth images in the conclusion part, because you write nothing about Google Earth images in "3 Results and Discussion".

Comments on the Quality of English Language

More improvements can be done. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have taken into account the observations made in the previous version. I agree with the current version.

Author Response

Thank you very much for your positive feedback. We are glad to hear that the revisions have addressed your concerns and that you are satisfied with the current version. We appreciate your valuable input and your time in reviewing the manuscript.