All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Dear author,
Thank you for your resubmission. One of the previous reviewers was kind enough to revisit the manuscript and has recommended acceptance. After considering your revisions in light of the first-round comments—including those of the second reviewer—I am in full agreement. I believe your work will make a valuable contribution to the literature.
I would also like to offer my apologies for the rather brusque tone of my initial decision letter. It can be difficult to strike the right tone in editorial correspondence, and in aiming for formality, I fear I over-corrected. I recognise that the publication process can be discouraging at times, and an Associate Editor’s role includes noting reviewer comments with which they may disagree. What I should have clarified was that the concern regarding novelty could be set aside—the manuscript did require additional work, but its novelty was not in question.
Congratulations and best wishes,
Anthony
[# PeerJ Staff Note - this decision was reviewed and approved by Richard Schuster, a PeerJ Section Editor covering this Section #]
After having read the authors' new or revised documents given in the zip file (in total there were 3 separated files), I think they have done better.
The article looks enhanced and I am glad for this.
Hopefully, the new article would bring new knowledge for those working in the area of animal behavior or ecology. Congrats for the author team..
Good, it has become more clear for us.
It seemed that team has improved it.
I am sure with the validity of the findings, where they brought new ideas or novelty regarding their research; esp. the application of UAV in ecological science and primate science.
No comment.
Thank you for your submission and for your patience while I sought appropriate reviewers.
After careful consideration I regret to inform you that I am unable to recommend your manuscript for publication in its current form. I find myself in agreement with one of the reviewers that while your dataset has potential and you highlight some interesting findings of relevance to continued progress in the field, the manuscript requires significant revision and conceptual reconsideration before it could be considered suitable for publication. That said, the work does not contain signficant flaws that render it unpublishable.
I encourage you to consider the reviewers' feedback carefully and resubmit when you are ready. I look forward to reading the revision.
**PeerJ Staff Note**: Please ensure that all review, editorial, and staff comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
1. It is a well-written and structured manuscript.
2. The following link provided for Raw data/code is broken, does not exist, or is not accessible for some reason: https://github.com/songguan26/Bornean-Orangutan-Nest
However, the link in the manuscript is functional, directing users to the GitHub repository where they can view the various Jupyter notebooks containing code. The raw data is not available.
Line 204 – 224: Highly inadequate information on data collection. How was the flight mission conducted exactly? Was it a pre-determined path or free flight? Was the video recorded or random images taken? What is the resolution of the images? No such specifics are given. It is crucial for readers to understand in detail how data is collected for any scientific study.
Line 216:218: The sentence is very ambiguous. Please make it as clear as possible.
Line 227”-231: It is not clear how these images with and without nests are selected from the total images, which the authors report as 1720 and 1911
Identifying nests from aerial imagery is difficult, as stated by the authors. As you can see in Figure 1. Only experts can reliably identify nests. For a naïve person, the identified nests in Figure 1 are difficult to distinguish from other dry patches in the same image in the bottom right corner, as they appear very similar. Then, using the DL or computer vision on such data becomes a proxy of a proxy. This is a significant drawback of this method. I believe it will require ground truthing (researchers visiting the location and verifying if the nest exists) to determine the error in manually labelled data by experts and ground truth. Based on that, we will then know the actual error in the nests detected by the object detection models and reality. The current results only demonstrate how different object detection models compare to manually labelled data.
The findings are valid. However, I find nothing novel in this study. Object detection is now a common practice in many fields, including ecology and wildlife biology. This study only compares different object detection models, which is a common practice and not novel in itself. Generally speaking, any of these object detection models, with sufficient training data, can accurately identify any objects.
Thus, merely identifying orangutan nests using any of these models and comparing them is not a step forward. However, authors were able to show which features were considered necessary by different models for object detection, which could be an essential insight in computer science.
However, generally, I believe, authors could have used their data to obtain some truly novel results. As the Authors have stated, the population size of orangutans is known in their study area. They could have checked if the total number of nests they found using DL matches the actual population estimates, as many preceding studies show how to use nests for estimating population. This or any other such application could have been novel.
Overall, I believe the study provides no significant step forward as automatic detection of an organ-gutan nest is already achieved by the studies cited by the authors.
General Comments :
Line 102 – 153: This primer on computer vision and DL is not necessary and can be shortened to a single short paragraph
I have read the article and overall, I think this is well-written and sequentially explained. However, I suggest some minor revisions or comments, with the hope that this could improve the clarity and better understanding for the reader. Please look at my extra separate file, which I sent here and I hope that author could follow the suggestions from our side.
Literature references are enough and author has written professionally.
Moreover, the raw data shared are also good and understandable.
Original primary research is in line with the aims and scope of the journal.
Research questionnaires was well-defined, however, I highly encourage author to highlight more about the theory of nest density in its relationship with the estimation of the orangutan's number. So far, unfortunately, we haven't heard about this issue in the paper. Although, author have provided Table 1, where it contained the numbers of "with nest" and "without nest". Please make it clearer, how many at the end the estimation of the orangutan still live in your sampling areas?
Second: please have a look at my comments regarding the Materials and Methods.
Study site: lines 182-199.
Further details: please have a look at my separate sheet.
The validity of the findings (highlights):
- The idea and novelty are good and useful. Author has tried to propose new ideas about the measurement of nest density with Drone and validate the results with the AI and computational arrangements and calculations. They are great.
However, there are two major concerns what I was thinking during I was reviewing:
Study Duration (after Materials and Methods's Section): lines 201 - 206
The Sepilok survey was done in July 2015, while the other area (Bukit Piton) was surveyed in January 2016.
The field study and the use of the drone for aerial images were conducted in 2014, with the permission from the local authority in Sabah (lines 220-221).
My questions:
- why are there two different years? from the study duration with the data relying from drone? Could you please clarify these?
- what about the validity of the data, if this was taken about almost a decade ago, with the current situation (year 2025)? Please give some arguments...to support the validity of your collected data.
Please consider to add more informations about the nest density theory and story. As we read, the article is more or less has been more intensely discussing about the "technical" aspects (Drone, IT, and DL, etc.);
Discussion Section: (lines 369-371)
---while counting nests from the ground is -------
this challenge has been highlighted.. how would you estimate the strength of this method (using drone) vs. manual counting? what are the pro's and con's of each? Please write down in the paragraph, so that reader could estimate the relevancy and efficacy of the methods;
for further details please have a look at my separate files. I hope that this would be helpful for author to improve their writing.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.