Identification of Thermal Refuges and Water Temperature Patterns in Salmonid-Bearing Subarctic Rivers of Northern Quebec

Round 1

Reviewer 1 Report

Identification of thermal refuges and water temperature patterns in salmon-bearing subarctic rivers of northern Québec, by M. Fakhari et al.

 

General comments

This paper submitted to the Journal Geographies reports on the identification of thermal refuges and so-called “cooling zones” in two rivers of high latitude subject to the presence of permafrost. The topic of thermal refuges detected using TIR remote sensing is expending, and the team at INRS is known for its major contributions to the topic over the years. Although I think the paper may contribute nicely to the field, there are several elements that should be considered before publication.

First of all, I respect the choice of the authors to submit this manuscript to Geographies, although I believe it would have been an appropriate contribution to the open Special Issue on River temperature: Perspectives, Applications, and Future Directions for Research in Canada planned to be released in the Canadian Water Resources Journal later this year. Having said that, the choice of targeting an audience of geographers is understandable given the tools and methods used. Yet, I believe that there is room for improvement. My concerns are not related to the quality of the dataset, but more to (1) a better description of the material and methods, (2) further (spatial) analyses of the data, and (3) an expanded discussion to go beyond the rather descriptive nature of the work so far.

 

Main remarks

Since you are targeting an audience of geographers, you should try to provide more spatial analyses. The paper feels very descriptive in places although there seems to be enough material for further analyses. I suggest a few possible options for further spatial analyses, some of which may not be successful nor find a strong interest to your eye but it is my role as a reviewer to give some feedbacks. These are open suggestions.

One of the main strengths of your work is the high latitude location and the presence of a gradient in permafrost conditions. Yes, the resolution of your permafrost maps might not be sufficient to look at fine-scale processes of each thermal refuge. But you can maybe provide some elements at the large/valley scale? With more details about the underlying processes, how permafrost can play a role in controlling the occurrence and distribution of thermal refuges? To help, perhaps you should give a short review of, from a mechanistic point of view, what are the underlying processes at play.

If I had to summarise your results, I would say this:

• B = continuous permafrost, low density of thermal refuges, warmer water temperature (but colder air temperature according to elements in “2. Study area”), lower valley confinement.

• K = discontinuous permafrost, higher density of thermal refuges despite colder water temperature (finding a thermal refuge in when the water is 8°C means that groundwater is really cold), higher valley confinement.

What does this tell us about processes in each river?

Re. sections 4.1.1 and 4.1.2. I think these sections bring a bit of confusion. Of course there is no correlation between the length of cooling zones and the number of thermal refuges in the Berard River because there are only 4 thermal refuges there. Looking at Figure 6, it seems that about half of the refuges are found in cooling zones, and half are not (both rivers included). I think this is important: it shows that the two processes are not necessarily related, and that both phenomenons might be a response to different factors. In addition, seeing that there is a relationship between the length of a cooling zone and the decrease in temperature seem very much logical. And it also means one thing: that the thermal gradient is the same in each cooling zone, and so it can be hypothesised that the processes are also similar. I think there is something to be done with regards to thermal gradients, especially within the cooling zones.

One thing lacking in your discussion, which links back to your introduction, is the relevance of this work for salmonids. And do to so, I think you need to give more information about the conditions in which the surveys were done: was it a very warm or typical summer? On a warm day? What can you say about water temperature on the days of the TIR surveys; is 9°C normal in such an environment? What having a cold-water refuge in a 10°C river mean for an arctic char? And then building on this, you can discuss what your results will mean in the future with climate change.

And this loops back to my final main point: I think you could expand in the discussion on how climate change will alter the actual processes driving the occurrence and distribution of thermal refuges and cooling zones in relation to permafrost. Give more emphasis to this, show more projection, relate it to fish species and life cycles, other in-stream processes, etc.

 

Detailed comments

Abstract

Ln 10: “thermal refuges” is not a process, but represent features in the landscape. Understanding the processes that control the presence and distribution of these features is the interesting part.

Ln 10: isn’t the role of “cooling zones” not also important?

Introduction

Ln 25: suggestion: you use the term “salmonids” in the abstract and “Salmonidae” here. I think Salmonids is better suited.

Ln 27: suggest addition to better distinguish between the past and today: “Today, fishing is still important…”

Ln 28: Salmonidae → Salmonid?

Ln 29: lethal temperature is not a range, but rather a threshold, even if “large” (25-27°C)

Ln 35: suggest reordering: “Extreme conditions in Québec rivers are likely to occur…”

Ln 41: delete “the”: “… thermal refuges in rivers is…”

Ln 42: expand a little to better link with the following paragraph. Suggestion: “… to understand fish habitat vulnerability but remains rather difficult to address at the large scale”.

Ln 57-72: please revise this paragraph. The content is correct, but the phrasing can be improved.

Ln 62: delete “the”: “Cooling zones in rivers are…”

Ln 63-64: I agree about the necessity of also looking at cooling zones in your study, but (1) I believe the two are complementary and inform on slightly different processes, so they shouldn’t be opposed and you do not need to look at cooling zones only because the method will miss out on some refuges; and (2) there are other factors that can lead to TIR mapping not capturing some refuges, notably the presence of riparian trees.

Study area

Ln 82: “where the river is braided”

Ln 85: toward north → northwards

Ln 88: see comment on Figure 2.

Figure 1: for this figure, as well as the others, I suggest numbering or lettering each sub-plot to make it easier to detail each part in the caption. It would also avoid references such as “Figure 7 top” for instance.

Figure 2: perhaps this figure is not necessary. You could simply cite reference 29 when you mention the classification Ln 88. At least move the figure to supplementary material.

Ln 105: plural: “Koroc River and its floodplain are in the zone...”

Figure 3: adopted → adapted.

Figure 4: could you draw the river on these maps? We can guess where they are based on the presence of alluvial deposits but it would be easier for the reader with the channel drawn across the floodplain to visualise the geology in contact with the river.

Material and Methods

For the entire M & M section (and Results section too), I strongly suggest to revise the tense use. Please consider using past continuous instead of past perfect, it would lighten the text and make it easier to read.

Ln 133: put “Table 1” between brackets

Ln 134: this can be moved to the caption of table 1

Ln 140: how were the 3 points chosen? Did you manually select these 3 points? Are they only 3 pixels then, or are each point made of several pixels? Why no average the temperature of the entire wetted channel of each image for instance?

Ln 157-175: why choosing a different numbering system (A, B, C) from other sub-sections? (e.g. 4.1.1 below). In addition, you need to provide more details on how these parameters were gathered. Manual digitisation of orthophotos? Existing databases? It is important to ensure that the scale at which you extracted the data matches the scale of your analysis (e.g. slope or width will look very different if you extract them at the 100-m or km scales).

Ln 158: “used for description of the thermal refuges…”. Yes, but perhaps you have enough material to go beyond description.

Ln 164: like → such as?

Ln 174: proved

Ln 175: to occurrence → with occurrence

Results

Ln 177: Catalogue? Please try to find a more specific title

Ln 179-188: see Main remarks about considering the density of thermal refuges.

Ln 179-180: suggest rewording: “four thermal refuges were detected: two cold side-channels and two lateral seeps (Figure 6a)”.

Ln 183: to match past tense used elsewhere in the manuscript, please use “were GW-controlled” and “were tributary plumes”.

Ln 183: 94% → 93%?

Ln 186-187: remove “based on the observed numbers”

Figure 6: please consider numbering your subplots. You should also explain the meaning of “wall-base channel A” and “wall-base channel B” in the caption, since it is not mentioned in the text. It took me some time to find an explanation back in Table 1. In legend, remove “and its name”.

Ln 200: 4.1.3. “Correlation between thermal refuges and cooling zones”?

Ln 201-211: see Main remarks on this section.

Figure 8: what is the statistical and mechanistic relevance of fitting polynomial functions? What is the interpretation?

Figure 7: how many thermographs were used as control points in the Koroc River? There is no “red cross” on Figure 7b. You can also remove “and its name” in the legend.

Ln 217-220: show location and temperature of tributaries in Figure 7?

Ln 223: could it also be because of a limited discharge of the tributary?

Ln 233, 239, 240, 261, 265: wall-based → wall-base (according to Dugdale et al. (2012) and the rest of your manuscript).

Ln 233: “wall-base channels and spring books”.

Ln 234: “wall-base channels”

Ln 236-247: can you confirm the difference between spring brooks and wall-base channels of type A, then? I struggle to see the difference...

Ln 250: re-order: “… and the length of a cooling zone...”

Ln 262-271: spring brooks and wall-base channels of type A show very similar ranges in entrenchment ratio, and you seem to draw a conclusion about the distribution of spring brooks being controlled by entrenchment (i.e. average entrenchment; valley not too wide but not too large either) but not wall-base channels of type A. This adds to the confusion I was mentioning above: what is the difference between the two types? Is there any other information do be drawn from Figure 10? e.g. tributaries are found in wide areas? Lateral seeps are found across a wider range of entrenchment conditions, but how about distance to the valley wall for instance?

Ln 268: instead or in addition to citing [17], you can refer to your Figure 5.

Ln 276: rephrase: “In the case of cooling zones, …are highly variable (Table 3)”

Ln 277-279: have you tried looking at your data in a different way? For instance dividing your temperature profile into sections of a given length (e.g. 5 times the mean channel width), and compute for each sub-section a response variable (e.g. thermal gradient, whether T° increases or decreases) and explanatory variables (e.g. sinuosity, entrenchment ratio, etc.), perhaps calculated as relative values (given the large differences between the two rivers according to Table 3)? Only a suggestion but perhaps there is more to be drained from your dataset.

Ln 281: “… role in driving (or controlling) GW-SW interactions”

Ln 285: “… has a semi-confined channel...”

Table 3: so you have calculated the thermal gradients of each cooling zone; why not use it? To make it easier to read, you should report it as °C/km (and not m). How can you report a sinuosity <1 (K1)?

Ln 293: remove River: “Geology and land cover”

Ln 295-296: do you mean that the valley is made of coniferous forest, but that the watershed slopes are covered with low density vegetation and exposed bedrock typical of Tundra landscapes?

Ln 308-311: your two sentences are somehow contradictory, or you have to be more specific. In addition, you could show a plot to reinforce your statement, e.g. boxplot of thermal refuges density according to dominant soil type and/or land cover in the vicinity of the river.

Ln 310: “… and the presence of more…”

Ln 312-317: okay, this section is very disappointing, especially because: (1) you attracted the reader by saying the main strength of this work is to show results in rivers with contrasting permafrost conditions, (2) the source and resolution of your permafrost maps should be mentioned in the Material & Methods section and not just here, and (3) despite the low resolution, I think you should try to dig deeper into this permafrost story (see Main remarks).

Ln 316: maybe explain “talik” as this is quite a specific term

Ln 325: you should mention in the M&M section that you are using satellite images for further description of your observations. Also, you should state the source and year of the images used.

Ln 329: “… indicator of the absence of…”

Ln 331: “Both… show degradation”

Ln 331: “Decrease or increase in size depend…”

Ln 332: difference → different

Ln 332: add a comma after “permafrost, and…”

Figure 13: it is important here to give the season (month) at which the satellite images where taken. Also consider numbering each sub-plot (e.g. a), b) and c) ) and provide a quick description of c).

Ln 339: “… top of the river valley, which sits on bedrock.”

Ln 341: consider adding a box on Figure 13 to guide the reader into where s/he should look to find beaded streams.

Ln 342: “… upstream of its main tributary…”

Ln 343: “… and form pools”

Ln 343-344: can you prove/justify this? With some data or literature references?

Discussion

Ln 353: size of what? Please specify

Ln 374: “Also, the use of TIR aerial images seems a reliable way…” since you didn’t prove it yourself

ln 361: difference in which direction? Re-state here, e.g. “… between the two rivers, i.e. higher density in the Koroc River”.

Ln 376: in low resolution → from low resolution

Ln 377: remove “thermal refuges”, not necessary

Ln 379: GW feds channels → GW-fed channels

Ln 382-388: see Main remarks on this section.

Ln 392: influence in thermal… → influence on thermal

Ln 396: remove comma after “since”

Discussion: what does the difference in the density of thermal refuges tell you about the processes at stake in each river?

Conclusion

Ln 399: 71 refuges over what distance?

Ln 403-405: any reason why shading is not at all mentioned in your paper? Have you considered it? Especially sections that flow in the middle of coniferous forests?

Ln 406: “… and difference of the two…” → “… and differences between the two…”

Ln 407: “… direction connection to the existence…”

Ln 411: “… important parameters in driving the occurrence of…”

Ln 413-415: create a boxplot of density in thermal refuges in relation to permafrost continuity? This could serve as the basis for discussion.

Ln 415: least → less

Ln 418: suggested addition: “Thaw of permafrost will lead to higher GW recharge, thus potentially to higher GW flow”?

 

Author Response

Thank you for your constructive comments.

I tried to apply them to the revised version as much as possible, and add new data analysis and information considering the limitation of dataset of this study.

I hope you find the revised version acceptable.

Kind regards.

Author Response File: Author Response.docx

Reviewer 2 Report

The presented manuscript describes an interesting study on a relevant and important topic. The study is important due to current climate change. It is of practical importance for the conservation of salmon populations. The authors have done a great job of collecting and processing the original data. The study area was studied in detail and diversified. The research design is correct and corresponds to the tasks set. The data is processed using modern and efficient methods. The methods are fully and clearly described. The results are well presented and qualitatively illustrated.

I think that the article can be published in its current form.

Author Response

Thank you for your kind review.

Kind regards.

Author Response File: Author Response.docx