Do Seabirds Control Wind Drift during Their Migration across the Strait of Gibraltar? A Study Using Remote Tracking by Radar

Round 1

Reviewer 1 Report

Manuscript ID: remotesensing-1722181
Type of manuscript: Article
Title: Do seabirds control wind drift during their migration across the Strait of Gibraltar? A study using remote tracking by radar.
Authors: Gonzalo Muñoz Arroyo *, María Mateos-Rodríguez
Submitted to section: Ecological Remote Sensing,

 

Summary: This manuscript tests whether seabird species compensate for wind experienced during migration by adjusting their heading, or whether they allow themselves to be drifted. Five species were studied in the Strait of Gibraltar, using a combination of radar and visual identification. The flight mode of the species affected whether compensation or drift was observed. Species with flapping flight compensated for wind during both spring and fall migration, regardless of wind direction. On the other hand, a gliding species, Cory’s shearwater, adaptively adjusted its strategy according to the direction of the wind, with drift observed when tail winds predominated in the spring, and compensation observed when wind directions were variable in the fall. This is a nice study that demonstrates how radar can be used to inform our understanding of seabird movement patterns. I have a few minor suggestions for improvement.

 

Comments:

 

1. In the Abstract and Introduction, it would be nice to discuss a bit more why you expect differences in whether compensation or drift is observed based on the mode of flight (flapping, gliding). Perhaps present this by the hypothesis regarding the difference between flapping and gliding birds?
2. L105: How many days were the data collected on during each period? What about variation in weather conditions other than wind?
3. Figure 1, panel 3. Is there a way to differentiate the flight paths of the different species?
4. Do the statistics account for potential non-independence of observations that occurred on the same day?
5. L206: Would you expect to see different patterns at higher wind speeds?
6. L201: Over a longer timeframe, might you see something different with respect to the amount of compensation versus drift? Did the time of recording affect your metric of drift?
7. L213: Would you expect anything different for flocks versus individual birds?
8. L308: Or beyond a certain wind speed, birds may be unable to avoid drift?
9. I suggest adding a concise conclusion.
10. L386: I am not sure about this because a bird could have the capacity for maneuverability, but yet choose not to exercise it as an energy conservation mechanisms, for example.
11. Finally, there are some problems with typos/language use in the manuscript, so I suggest carefully proof reading. E.g.

L38: should be no “to” before flight

L73: “develop” sounds strange, I suggest “display”

L288: change “to be” to “being”

L320-321: why would you not see this pattern?

L330-331: Does not exactly work

L336: “keystones” instead of “keystone”

Comments for author File: Comments.pdf

Author Response

Reviewer 1.

 

Comments and Suggestions for Authors

Manuscript ID: remotesensing-1722181
Type of manuscript: Article
Title: Do seabirds control wind drift during their migration across the Strait of Gibraltar? A study using remote tracking by radar.
Authors: Gonzalo Muñoz Arroyo *, María Mateos-Rodríguez
Submitted to section: Ecological Remote Sensing,

 

Summary: This manuscript tests whether seabird species compensate for wind experienced during migration by adjusting their heading, or whether they allow themselves to be drifted. Five species were studied in the Strait of Gibraltar, using a combination of radar and visual identification. The flight mode of the species affected whether compensation or drift was observed. Species with flapping flight compensated for wind during both spring and fall migration, regardless of wind direction. On the other hand, a gliding species, Cory’s shearwater, adaptively adjusted its strategy according to the direction of the wind, with drift observed when tail winds predominated in the spring, and compensation observed when wind directions were variable in the fall. This is a nice study that demonstrates how radar can be used to inform our understanding of seabird movement patterns. I have a few minor suggestions for improvement.

 

Comments:

 

1. In the Abstract and Introduction, it would be nice to discuss a bit more why you expect differences in whether compensation or drift is observed based on the mode of flight (flapping, gliding). Perhaps present this by the hypothesis regarding the difference between flapping and gliding birds?

 

We have tried to fit the reviewer's suggestion. In this way, we have expanded the information on the reasons why seabirds can show different behaviours in relation to wind drift compensation when crossing the Strait in the Abstract section. Additionally, we have alternated the order of the hypotheses stated in the introduction, remarking the expectations and their possible causes.

In any case, we are willing to expand this information, if necessary.

 

1. L105: How many days were the data collected on during each period? What about variation in weather conditions other than wind?

 

Now we indicate in the Materials and Methods section the number of days in each season, and the weather conditions (see lines 109-112 in the current version). The observations were generally made in good weather conditions, and under light or moderate winds, since the radar was not operative to detect birds under strong wind conditions (associated in the Strait of Gibraltar with bad weather conditions). We now indicate the maximum limit of wind speed thresholds for each season.  In fact, we have moved a paragraph referring to the limitation of radar operability from the results section to the material and methods section, to explain this point (originally in L. 206-207; moved to L. 113-114).

 

1. Figure 1, panel 3. Is there a way to differentiate the flight paths of the different species?

 

Figure 1 simply shows the type of track provided by the radar in the context of our study area, and is not intended to show differences in passage patterns between species. Modifying this figure would require intense work with Arcview software, so unless it would be considered essential, we would prefer not to modify it.

 

1. Do the statistics account for potential non-independence of observations that occurred on the same day?

 

In the statistical analyses, the independence of the observations as a function of day has not been explicitly tested. The analyses have been conditioned by the use of circular statistics methods, in this case the Watson-Williams test (Batschelet, 1981). This test has the disadvantage that it does not allow the addition of covariates (Cremer and Klugkist, 2018[1]). These authors proposed an alternative PN circular GLM, that allows for the addition of covariates, but using this alternative analysis in our study would mean redoing all the analyses. However, the track data set used in our study was evenly distributed across different days at different seasons, encompassing a representative range of wind conditions in our study area (with the limitation of strong wind conditions, where our radar could not operate). Therefore, we consider that the effect of this day-dependence, if existing, may be negligible, and that our results and conclusions are sound, according to the conditions of the study. If the editor and reviewers consider it relevant, this justification can be included in the manuscript (in the discussion section).

 

1. L206: Would you expect to see different patterns at higher wind speeds?

 

Because our study is limited by radar operability to light or moderate wind conditions, we cannot say with certainty that it would occur beyond that range. According to the literature, birds subjected to very strong winds can be overpowered by the winds and cannot avoid becoming drifted off their preferred track direction [15; this point is already mentioned in the introduction, L. 63-64 in the current version]. Therefore, we cannot rule out that this occurs in the Strait of Gibraltar, where winds of over 80 km/h are not uncommon.

In the revised version, we have included additional information in this regard in the discussion section (see L. 316-320).

 

1. L201: Over a longer timeframe, might you see something different with respect to the amount of compensation versus drift? Did the time of recording affect your metric of drift?

 

A priori, we do not find any reason why the duration of the track could influence the flight behaviour at a local scale. The duration data of the tracks is offered simply for information purposes. In this work, we are analyzing the local behaviour of the birds when crossing a bottleneck, where the wind conditions are particular, but we do not think that the birds experience important changes in flight behaviour in such a limited route. However, if the editor or the reviewers consider it relevant, we can expand the information in this regard.

 

1. L213: Would you expect anything different for flocks versus individual birds?

 

A priori, we do not expect different behaviour between flocks of birds and individual birds. Once again, the characteristics of the flocks are provided for information purposes, since they can help to better understand the operation of the radar. In our study, the statistical method used (Watson-Williams test) does not allow the addition of covariates. In any case, a significant effect of flock size on flight behaviour is not expected.

 

1. L308: Or beyond a certain wind speed, birds may be unable to avoid drift?

 

See above. In the revised version, we have included additional information in this regard in the discussion section (see L. 316-320).

 

1. I suggest adding a concise conclusion.

 

We have tried to fit this suggestion and, at the same time, that of reviewer 2 on the reduction of the final section in the discussion (Ln 352-370 (390) of the original MS; see below). Now we have reoriented this last section of the discussion, towards a new section of conclusions. This section is now focused on summarizing the main findings of our research, emphasizing the relevance of radar applications to remote monitoring of animal movement, in the current context of global change.

We expect this fit the reviewer’s suggestions, but we are open to new advices.

 

1. L386: I am not sure about this because a bird could have the capacity for maneuverability, but yet choose not to exercise it as an energy conservation mechanisms, for example.

 

This sentence has been removed in the revised section, following the suggestions of reviewer 2.

 

1. Finally, there are some problems with typos/language use in the manuscript, so I suggest carefully proof reading. E.g.

 

We have carried out a thorough revision of the typology throughout the text. We hope that all mistakes have now been resolved correctly.

L38: should be no “to” before flight

Corrected.

L73: “develop” sounds strange, I suggest “display”

Corrected

L288: change “to be” to “being”

Done

L320-321: why would you not see this pattern?

Corrected

L330-331: Does not exactly work

At this point, we understand that the reviewer is referring to the syntax of the sentence. We have tried to simplify this sentence to make it clearer (see line 339-341). However, do not hesitate to make new suggestions in this regard.

L336: “keystones” instead of “keystone”

Corrected

peer-review-19300164.v2.pdf

Submission Date

26 April 2022

Date of this review

01 May 2022 17:29:56

 

[1] Cremers, J., & Klugkist, I. (2018). One direction? A tutorial for circular data analysis using R with examples in cognitive psychology. Frontiers in psychology, 2040.

Reviewer 2 Report

Review ‘Do Seabirds Control Wind Drift during their Migration Across the Strait of Gibraltar? A Study using Remote Tracking by Radar.’ Arroyo and Mateos-Rodríguez

 

The authors describe seabird passages through the narrow Strait of Gibraltar during their spring and autumn passage and how wind influences their behavior or not. They found that 4 out of 5 species fully compensate for wind drift, while the one species able to use a gliding flight style is able to drift (Spring) or fully compensate for wind drift (Autumn). These results nicely fits the theory of wind compensation under moderate wind condition and undoubtedly deserve attention and the opportunity to be published. Nice written paper!

A few minor points:

Ln 57: Reference in two different styles

Ln 60: should or should be able to compensate for wind drift?

Ln 62: remove ‘getting’

Ln 63: ‘in high wind speed conditions’

Ln 65: remove ‘to’

Ln 65-68: This sentence is an extreme of the text above and does make the text unnecessary difficult to read. Suggest to remove this sentence or rewrite

Ln 127: remove ‘, respectively’ since no other reference is given

Ln 179: would use another word for ‘oppositely’, e.g. contrarily or reversely

Ln 182-192: Why did the authors use T1 and T2 instead of T_L and T_R? Furthermore where is B_track coming from? A bit more explanation is necessary here. And the formula should be: (T1 – T2) / (α1 – α2) or as formula 4 in [REF 16]. Is a reference missing, probably [16]?

Ln 204-205: Remove “Wind data within the same hour were available for all tracks, permitting calculations of air speed and heading direction in all the cases.” This is not a result, or move to methods section

Ln 210-212: not really a result and partly mentioned in methods section

Table 2: layout of species name; do you really need one decimal place in all those degrees (only SD)?

Ln 213-214: If 74% are flocks is mentioning 26% single birds necessary?

Ln 233-235: please rewrite sentence; maybe use the word ‘partial ‘?

Ln 247-251: Isn’t this part of the discussion?

Ln 311: ‘According to theory’ would suggest to add REFERENCE here

Ln 320-322: awkward sentence

Ln 324-326: remove ‘they show’, but words are missing in this sentence

Ln 330-332: this sounds reasonable, but why would Cory’s not respond the same in both seasons? Is this a lack of enough measurements/wind conditions? But see explanation in next para, thus a biased selection of those birds that fly more north through the Straight?

Ln 335: replace migrating by migration

Ln 352-370 (390): I don’t understand this chapter in this paper. The paper is on flight behavior during different wind conditions. Radar is ‘just’ a tool to measure this. Suggest to shorten this for at least 50%

 

Author Response

Reviewer 2.

Comments and Suggestions for Authors

Review ‘Do Seabirds Control Wind Drift during their Migration Across the Strait of Gibraltar? A Study using Remote Tracking by Radar.’ Arroyo and Mateos-Rodríguez

 

The authors describe seabird passages through the narrow Strait of Gibraltar during their spring and autumn passage and how wind influences their behavior or not. They found that 4 out of 5 species fully compensate for wind drift, while the one species able to use a gliding flight style is able to drift (Spring) or fully compensate for wind drift (Autumn). These results nicely fits the theory of wind compensation under moderate wind condition and undoubtedly deserve attention and the opportunity to be published. Nice written paper!

A few minor points:

Ln 57: Reference in two different styles

Corrected

Ln 60: should or should be able to compensate for wind drift?

We change the sentence as the reviewer suggested.

Ln 62: remove ‘getting’

Done

Ln 63: ‘in high wind speed conditions’

Corrected

Ln 65: remove ‘to’

Done

Ln 65-68: This sentence is an extreme of the text above and does make the text unnecessary difficult to read. Suggest to remove this sentence or rewrite.

Following the reviewer suggestion, we have removed this sentence. Then, we have moved the reference #20 to support the previous sentence.

Ln 127: remove ‘, respectively’ since no other reference is given.

Actually, mean directions refers to main vector directions, while scatter refers to angular deviations. However, we have rewritten the phrase (see lines 133-134), for clarity.

Ln 179: would use another word for ‘oppositely’, e.g. contrarily or reversely

Done.

 

Ln 182-192: Why did the authors use T1 and T2 instead of T_L and T_R? Furthermore where is B_track coming from? A bit more explanation is necessary here. And the formula should be: (T1 – T2) / (α1 – α2) or as formula 4 in [REF 16]. Is a reference missing, probably [16]?

Indeed, we have followed Green and Alerstam 2002 [16] to calculate drift and compensation magnitudes, and we use the same nomenclatures. In fact, the formula included in our manuscript refers to equation #4 of the article by these authors. We have now included this reference, following the reviewer's suggestion.

Ln 204-205: Remove “Wind data within the same hour were available for all tracks, permitting calculations of air speed and heading direction in all the cases.” This is not a result, or move to methods section

This sentence has been removed as it is already explained in the method section.

Ln 210-212: not really a result and partly mentioned in methods section

This paragraph has been removed

Table 2: layout of species name; do you really need one decimal place in all those degrees (only SD)?

In table 2, we have removed the first decimal from the mean values for track, heading and wind directions; the decimal values for the angular deviations and for the wind speed (mean and SD) have been kept.

Ln 213-214: If 74% are flocks is mentioning 26% single birds necessary?

This sentence has been rephrased as follows (L. 219-220): “Most tracks (74%) were of flocks, the rest corresponded to single bird”

Ln 233-235: please rewrite sentence; maybe use the word ‘partial ‘?

We try to fit the reviewer suggestion, and rewrite the sentence as follows (L.239-241):

“Seabirds were exposed to crosswinds in 13.2% of the tracks; in the rest of the cases, they were flying with following or opposing winds with partial crosswind component”.

Ln 247-251: Isn’t this part of the discussion?

At this point, we consider that the description of the magnitude of the effect (drift or compensation) reinforces the exposition of the previous results, showing that both the compensation and the drift are practically total in the respective species. Therefore, if the editor agrees, we would prefer to keep this paragraph in the results section.

Ln 311: ‘According to theory’ would suggest to add REFERENCE here

We have added [7, 10] as the most relevant reference according to general theory.

Ln 320-322: awkward sentence

We have removed this sentence and then reworded the entire paragraph for clarity. The proposed new paragraph is as follows (L.327-333):

“Therefore, Cory’s shearwaters may prefer to drift when birds experience following winds because there probably is a certain wind angle when they take optimal advantage of the wind energy, increasing its flight speed and reducing its energy use during a journey through this region. As we predicted in our hypotheses, the capacity to adapt the flight pattern to wind conditions can be constrained by flight type and morphological characteristics of the different species.”

Ln 324-326: remove ‘they show’, but words are missing in this sentence

We have rephrased this sentence, as follows (L.333-335):

Mateos-Rodríguez and Bruderer [41] for the same range of species in our study, showed that these species responded differently to the wind, in terms of air speeds, depending on their morphological characteristics and flight behaviour.

Ln 330-332: this sounds reasonable, but why would Cory’s not respond the same in both seasons? Is this a lack of enough measurements/wind conditions? But see explanation in next para, thus a biased selection of those birds that fly more north through the Straight?

At this point, our study shows differences between species and between migration periods (particularly in the Cory's shearwater), but we cannot go much further in determining causality. We argue that the different wind conditions faced by the birds (with prevailing tailwinds in spring; L. 324-331), along with different migration patterns in both seasons (L. 339-351), may be underpinning the adaptive response in Cory's shearwaters. However, other possible causes cannot be ruled out.

We are aware of the limitations of our study. The range of winds conditions in which the radar can operate is limited to light or moderate winds, due to sea clutter that preclude low flying seabird detection [39, 40]. In this sense, seabirds can be overpowered under strong wind conditions, and then they cannot avoid becoming drifted off their preferred track direction [15]. Moreover, the detection range of the radar was limited to almost seven kilometres [Mateos-Rodríguez et al. 2010, this reference is not included at the moment], thus covering only the northern sector of the Strait. However, this range covers somewhat less than half of the passage channel in the Strait (minimum width, 14.7 km from the Island of Tarifa), which constitutes a considerable part of the local flight path for seabirds. Some of these issues are already discussed in different places throughout the discussion. If the editor and reviewers consider it necessary, we may place more emphasis on these aspects, which in any case we think do not detract from the validity of the results obtained.

Ln 335: replace migrating by migration.

Done

Ln 352-370 (390): I don’t understand this chapter in this paper. The paper is on flight behavior during different wind conditions. Radar is ‘just’ a tool to measure this. Suggest to shorten this for at least 50%

We had included this section in order to highlight the possibilities of radar technology as a remote system to describe in detail the flight behaviour of seabirds, showing some relevant applications in the current context of global change. We thought that these aspects could arouse the interest of potential readers beyond the environmentalists of the animal movement.

Now, we have reoriented this paragraph, following your suggestion, but also the one made by Reviewer #1. We have built a new section of Conclusions, focused on summarizing the main findings of our research, emphasizing the relevance of radar applications to remote monitoring of flight behaviour, in the current context of global change. Thus, we have also shortened this paragraph (from 492 to 244 words).

We expect this fit the reviewer’s suggestions, but we are open to new advices.

 

 

Submission Date

26 April 2022

Date of this review

10 May 2022 11:20:24