Review History

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.

View examples of open peer review.


  • The initial submission of this article was received on February 17th, 2022 and was peer-reviewed by 2 reviewers and the Academic Editor.
  • The Academic Editor made their initial decision on April 1st, 2022.
  • The first revision was submitted on June 15th, 2022 and was reviewed by 1 reviewer and the Academic Editor.
  • The article was Accepted by the Academic Editor on July 27th, 2022.

Version 0.2 (accepted)

· Jul 27, 2022 · Academic Editor


Thank you very much for your revision. Based on the reviewer's comments and after another read through the manuscript I am happy to accept it!

However, there are a few very minor issues that should to be fixed before the production process starts, so I kindly ask you to quickly fix these (see attached annotated pdf).

I look forward to seeing the published version of your paper!

[# PeerJ Staff Note - this decision was reviewed and approved by Dezene Huber, a PeerJ Section Editor covering this Section #]

[# PeerJ Staff Note: Although the Academic and Section Editors are happy to accept your article as being scientifically sound, a final check of the manuscript shows that it would benefit from further editing. Therefore, please identify necessary edits and address these while in proof stage. #]


Basic reporting

I was positive about the first version already, although I had many questions and comments. I am fine with all the responses and the way authors have worked on these comments. I think the paper has improved significantly. In particular, the structure of the paper itself is now clear and unambiguous.
About my point on the lack of validation: I am okay with the response.
I think the paper meets the PeerJ standard (from what I know of this journal).

Experimental design

no comment

Validity of the findings

no comment

Additional comments

Line 49 in the abstract : "...could inform improve...". Either one word should not be there, or another word is missing.

Version 0.1 (original submission)

· Apr 1, 2022 · Academic Editor

Major Revisions

Dear authors,

We have now received two reviews, both of which highlight the potential value of your study, but also list several issues that need to be addressed before I can make a final editorial decision. I encourage you to carefully consider all points raised and explain in a detailed point-by-point reply how you have addressed the comments.

From my point of view, several comments by reviewer 1 on (1) the missing validation of the approach, (2) the partly unjustified statements about novelty, and (3) the lacking discussion on limitations and caveats deserve special attention. But also all other comments by both reviewers should be carefully taken into account in a revised version of your manuscript, which I look forward to receiving soon!

Kind regards,

[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. #]


Basic reporting

This paper reports a framework to quantify the vulnerability of ectotherms to climate change based on the landscape heterogeneity of available microclimates (here, surface and air temperatures obtained by remote sensing and modelling) at a rather high resolution (few cm scale) and on the thermal limit of the species studied. The framework relies on the use of UAV (drones) equipped with IR cameras to derive the spatial heterogeneity of surface temperatures, which are translated latter into air temperature layers using a statistical modelling approach. In the paper, the framework is explained, and then applied to 3 lepidopteran species to generate maps of thermal risks as an illustration.
I am quite favorable about the idea of having this framework available. There is growing evidence and need for such approaches that integrate the spatial and temporal heterogeneity in temperatures at the landscape scale (microclimates) to assess the vulnerability of species – which is quantified here by the difference between the temperature in the microhabitat used by a species and its specific thermal limits. The use of UAVs allow indeed a rapid assessment of landscape heterogeneity with a high resolution (high enough for much species). Therefore, this paper is timely and useful.
About specific points for PEERJ:
- The english is excellent and professional.
- Introduction and background: sufficient, although see my comments for additional references.
- The structure of the article conforms to standards. See my comments on figures however.
- the paper is "self-contained". This paper reports a framework, therefore the hypotheses are not laid out like for a standard article with original research.

Experimental design

I have several remarks at that level that should help improve the paper. The methods are not explained with enough details overall.
1) In the methods, I miss information about how many sites, how many loggers total, how many loggers per site, how many loggers per category of microhabitat elements ... The lack of information at that level prevents the reader from getting a comprehensive view of what should be done and how if he/she has to apply this framework. More specifically: (1A) line 178: ibuttons were placed on the ground. Please be explicit as to what this logger measured exactly. I suppose it measured soil surface temperature in the shade. (1B) Line 184 and elsewhere: the text should make clear that the remote sensing is measuring SOIL SURFACE temperature, and mostly the surface that is exposed to solar radiation - indeed shaded areas at the level of soil likely cannot be 'seen' by the UAV. Therefore, I wonder to what extent the surface temperature obtained by the UAV can really be compared to the temperature collected by the loggers in the air and at the soil surface in the shade, although I agree that the statistical (regression between surface and air temperatures) sounds reasonable but such a relationship may not be that solid for all ecosystems (this should be acknowledged in the discussion, but see point 2 below). (1C) Lepido are exposed mostly to air temperature. This is likely to be true for adults but not necessarily for eggs and young caterpillars, which are mostly exposed to surface temperatures depending on their body size (see Pincebourde et al. 2021 in Functional Ecology or Woods 2013 in Functional Ecology as well). (1D) Line 206: specify the dates / period of those 29 surveys. The mosaic of thermal environments is likely to vary much across months/seasons. Therefore this study may reflect only what happens during the specific period that was surveyed (again this should be acknowledged in the discussion, see below). (1E) the lack of information on number of loggers etc. impedes one to comprehend how the air temperature was modelled: line 246, which ibutton? the ones that were deployed at the surface of the ground or those that were included into PVC pipes? how this modelling of air temperature took into account the differences between different elements in the landscape? A different linear regression was done for each element/microhabitat? (1F) it is very unclear to me what was done with those 3 lepido species: how data were collected concerning their presence in a given microhabitat?

Validity of the findings

2) The discussion totally lacks a paragraph on limitations of the framework. This is important since acknowledging the limitation will boost new development to improve rapidly this same framework. The discussion is relatively short and focuses mostly on the strength and potential applications of the framework. Among the numerous limitations, I note that in figure 6, the regression analysis is significant but the dispersion around the linear model is large. For instance, when remote sensing indicates 30°C, the actual air temperature can as low as 5 and as high as 35°C. This magnitude of error can considerably challenge the overheating index and the thermal positioning maps. Finally, while I like the idea of applying the framework as an illustration, The result on vulnerability is not validated in any way and nor discussed in regards to what is known for the 3 species in terms of habitat restriction due to climate change for instance. This is another limitation here: the approach has not been validated yet.
3) I am afraid that extracting monthly average is too coarse temporal resolution relative to the high spatial resolution reached by the use of UAV. I expect that thermal limits will be considerably underestimated when using such coarse temporal resolution because the average considerably buffer the extreme temperatures.
4) The Results section is quite confusing for several reasons that follow. (4A) Results line 313: it is difficult to make the link between this sentence and the corresponding figure. Thermal diversity is the explanatory variable in the sentence but it is the explained variable (Y axis) in the figure. In addition, the figure displays an inverted U shape relationship while the sentence here seems to suggest a linear regression analysis? (4B) Line 317: I am confused here: this result relates overheating index with air temperature which was measured by hand with a thermometer during each drone survey. BUT the overheating index itself results from the matrix of air temperatures inferred for all pixels of the map. Therefore, what is the pertinence of using air temperature on the x axis in the fig 5? (4C) Line 320: I would have positioned this result before the results on overheating maps. In addition, I expected to see also the values for historical thermal limits of the 3 species. I expected to see more details on the spatial distribution of the thermal positioning index.
5) Line 335: “this is the first model that directly assesses microclimatic variation using high resolution, drone-based remote sensing, and also the first to assess those temperature measurements relative to species’ observed thermal limits”. I disagree with this statement of novelty. See the few papers by Emile FAYE et al. (incl the one you cite in Method in Ecology and Evolution, but also several others). One can also argue that there are other studies reporting microclimatic variations at much higher spatial resolution (not only mine, but also on lizards), although not all inferred from drones.
6) Figure 1. Where are the loggers along this pipe? one may worry that no holes were created in the pipe at the location of the logger to facilitate air circulation. Even if white, the pipe may overheat slightly when in the sun if air circulation across the pipe is not enabled.
7) Figure 4: the inverted U-shape makes sense to me intuitively: low foliage height diversity means an overall homogeneous and flat 'surface', therefore relatively homogeneous thermal landscape. The same for very diverse foliage height: at some point of extreme complexity the top surface of vegetation becomes thermally homogeneous. Maximal thermal diversity is obtained at intermediate vegetation height diversity at which probably we get a mixture of bare ground and spots with dense vegetation. That said, I am surprised by the rather high thermal diversity index which varies only little in the range 0.93-0.96. Moreover, this U relationship is obtained thanks to only one point on the left side with low thermal diversity. Without this point, the inverted U shape would have been compromised. I am unsure what to do with this result, and again this is only little discussed in the discussion. For instance, to what extent such relationship depends on the ecosystem and plant species?

Additional comments

Minor issues:
-The paragraph line 91-105: The paper by Potter et al. 2013 (in Global Change Biology: microclimatic challenges in global change biology) appears extremely central here in this paragraph.
- Line 98: “Few studies have explored microclimate dynamics at scales < 100 m2”. Such studies are indeed numerous – see the review of Pincebourde et al 2016 in Integrative and Comparative Biology.
- Line 105: unclear why measurements require models? What kind of model? To predict what if the measurements are ... measured? Need to be reformulated I think.
- Line 192: “hottest and coldest locations” : in terms of which variable? The temperatures collected by the UAV and/or by loggers, surface and/or air?
- Line 267: clarify the equation. P seems to be a sum of the thermal positioning index across all pixels in a given map. Therefore I expected to see P to be divided by the number of pixels to be compared between maps? I probably misunderstand something here, indicating that the text may not be clear enough.

Reviewer 2 ·

Basic reporting

I think some background or mentioning of current microclimate models can put the papers in the right perspective. There are a lot of efforts around to develop physical models:
Compared to these past studies, what is the benefit of using statistical models like in the current paper?

The introduction needs work on flow and direction. Each paragraph should lead to the next to structure a complete story of the challenges in the field and the importance of the current work. For example, the first paragraph of the introduction is misleading since it’s about extreme events but the manuscript is more about modeling thermal refugia. Also, in line 125, the mention of satellites seems out of context since it is not clear that satellites are the main source of TIR images to date.

Experimental design

There is no research question but the authors clearly define the aim of the paper: to present a framework with an interesting case study. I think the framework is really relevant given the current state of microclimate models and the general surprising lack of using drone data.

Lines 125-126 - What does the satellite imagery mean? I thought images were acquired by a drone.
Lines 171-174 – It is not clear why these measures were in place.

Validity of the findings

Although I commend the authors for their hard work, especially gathering the data, the statistical analysis does not seem to fit the data. In particular, the work has two important flaws:
1) Measurements – the air temperatures were measured in the shade but ground temperatures can be significantly different between shade and open grounds, and so are the air temperatures close to the ground. Thus, the model might only work for shaded microhabitats.
2) The statistical model seems to be oversimplified - The problem is that the air temperature also depends on slope azimuth and shade, especially close to the ground.

Lines 313-316 – what model was used in figure 4? Clearly, a linea model with a quadric relationship is needed based on figure 4
Lines 317-319 – In the figure, it looks significant… What were the p-values of the slopes?

Additional comments

This is a very interesting paper that aims to develop and showcase a method for high-resolution mapping of air temperature. The authors have used drone mapping of ground thermal temperatures and locally measured air temperature to develop a statistical model between ground and air temperatures. However, I have significant concerns about the data collection of air temperatures and statistical analysis.
The paper is also disorganized. For example, the last paragraph of the introduction states 4 stages in the framework but the methods section has 5. The results section is quite short and does not include many details that can explain the strength and sensitivity of the framework.

Figure 2
The bottom scales miss intermediate values.

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.