Increasing Sea Surface Temperatures Driving Widespread Tropicalization in South Atlantic Pelagic Fisheries

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Review for the paper “Increasing sea surface temperatures driving widespread tropicalization in South Atlantic pelagic fisheries” by Alícia Pereira and co-authors submitted to “Biology”.

 

The authors of this research paper conducted an analysis to investigate the effects of ocean warming on fisheries in the South Atlantic Ocean, focusing on pelagic species from 1978 to 2018. They found that the mean temperature of fish catches has significantly increased over this period, signaling a shift towards a dominance of warm-water species, particularly on the western side of the ocean. This transition was supported by the observed increase in the proportion of warm-water species within the catch composition. The authors employed ordination methods and estimates of beta diversity, which illustrated a significant shift in species composition from a system primarily composed of cold-water species to one increasingly characterized by warm-water species. The results of this study may have important implications for fisheries management and conservation strategies. As warm-water species become more prevalent, existing management practices may need to adapt to account for shifts in species distributions, which can affect local fishery yields and economic stability.

 

Recommendations.

 

Introduction.

 

The authors should clarify what "blue foods" are, and explain how they differ specifically from other food sources in the context of human diets.

 

The authors should mention how specific human activities beyond overfishing (pollution, habitat destruction, and climate change) contribute to the degradation of marine ecosystems.

 

The authors should explain what is meant by "metabolic rates" in the context of marine ecosystems and why changes in these rates are important.

 

The authors should provide a more comprehensive background. What previous studies have been conducted? What knowledge gaps currently exist? They should also highlight the novelty and importance of this study more clearly.

 

Materials and Methods.

 

Which statistical or analytical software was used to perform the linear regression models and ANCOVA?

 

Did the authors check the data for the necessary assumptions, such as normality and homogeneity of variance?

 

How did the authors adjust the time-series data for autocorrelations?

 

Results.

 

It would be useful to include a graph showing the catch dynamics of each fish category in the study area.

 

Figure S1 is cited on page 10. However, no supplementary materials were included.

 

Section 3.3. Were the clusters defined by PCoA different compared to each other? For example, were the overleaped clusters 1, 2, and 3  in Figure 4 significantly different?

 

Discussion.

 

The authors should discuss the biological and ecological role of albacore in these ecosystems to explain how the sensitivity of MTC to the albacore (T. alalunga) specifically impact broader ecological trends.

 

The authors should discuss the mechanisms that enable large pelagic predators to acclimate to thermal variability. They should provide information on the physiological adaptations or behavioral changes that enable acclimation, as well as the limits of these adaptations that lead to stress or decline.

 

What underlying factors might explain the stark contrast in the rate of MTC increase between SWAO (0.12°C per decade) and SEAO (0.04°C per decade)?

 

Figure S2 is cited on page 14, but no supplementary materials were included.

Author Response

[Comments 01] The authors should clarify what "blue foods" are, and explain how they differ specifically from other food sources in the context of human diets. The authors should mention how specific human activities beyond overfishing (pollution, habitat destruction, and climate change) contribute to the degradation of marine ecosystems.

[Response 01] Suggestions accepted. Please see the revised manuscript lines 44-49.

 

[Comments 02] The authors should explain what is meant by "metabolic rates" in the context of marine ecosystems and why changes in these rates are important.

[Response 02] Suggestions accepted. Please see the revised manuscript lines 57-61.

 

[Comments 03] The authors should provide a more comprehensive background. What previous studies have been conducted? What knowledge gaps currently exist? They should also highlight the novelty and importance of this study more clearly.

[Response 03] Suggestions accepted. Please see the revised manuscript lines 84-89 and lines 100-116.

 

[Comments 04] Which statistical or analytical software was used to perform the linear regression models and ANCOVA?

[Response 04] All analysis were conducted using the language and environment for statistical computing R 4.4.2 (R Core Team, 2024). The suggestion was included in lines 224-226.

 

[Comments 05] Did the authors check the data for the necessary assumptions, such as normality and homogeneity of variance?

[Response 05] Yes. The results for the assumptions were presented in the supplementary document and all the references were included a long the manuscript. Please see the revised manuscript lines 257-264 and respective supplementary material.  

 

[Comments 06] How did the authors adjust the time-series data for autocorrelations?

[Response 06] The autocorrelation were not formaly structured inside the models. In this study we had focused on the method proposed by Zeileis (2006) in his R package called “dynlm” and discussed by Pfaff (2008) in the book, entittle “Analysis of Integrated and Cointegrated Time Series with R”. This approach is based on the application of linear regression with dynamic temporal structure. This method allow us to include a time-lag process in the explanatory variable side in a simple form, preserving the time-serie structure and enabling the inclusion and evaluation of different time lags. This procedure allowed us to understand how and at what level of delay changes in the ecosystem structure can influence the composition of the species, for example. The suggestion was included in the manuscript. Please see the revised manuscript lines 184-190.

 

[Comments 07] It would be useful to include a graph showing the catch dynamics of each fish category in the study area.

[Response 07] Suggestion accepted. Please see the revised manuscript line 238 and respective supplementary material.

 

[Comments 08] Figure S1 is cited on page 10. However, no supplementary materials were included.

[Response 08] We apologize for the error. It has been corrected.

 

[Comments 09] Section 3.3. Were the clusters defined by PCoA different compared to each other? For example, were the overleaped clusters 1, 2, and 3  in Figure 4 significantly different?

[Response 09] Suggestions accepted. Please see the revised manuscript at line 302 and respective supplementary material.

[Comments 10] The authors should discuss the biological and ecological role of albacore in these ecosystems to explain how the sensitivity of MTC to the albacore (T. alalunga) specifically impact broader ecological trends. The authors should discuss the mechanisms that enable large pelagic predators to acclimate to thermal variability. They should provide information on the physiological adaptations or behavioral changes that enable acclimation, as well as the limits of these adaptations that lead to stress or decline.

[Response 10] Suggestion accepted. Please see the revised manuscript lines 406-410

 

[Comments 11] What underlying factors might explain the stark contrast in the rate of MTC increase between SWAO (0.12°C per decade) and SEAO (0.04°C per decade)?

[Response 11] Suggestions accepted. Please see the revised manuscript lines 501-504.

 

[Comments 12] Figure S2 is cited on page 14, but no supplementary materials were included.

[Response 12] We apologize for the error. It has been corrected.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors conducted a well-sound study on the tropicalization trend in the pelagic fisheries of the South Atlantic Ocean in relation to ocean warming. The manuscript based on a 41-year dataset to analyze changes in species composition, mean temperature of the catch, and the influence of environmental drivers. An important asset of the study is the fact that the authors provide access to data repositories (e.g., ICCAT, GitHub), which will enhance reproducibility of the study outcomes.

Minor revisions are recommended to improve interpretability and further strengthen the robustness of the conclusions. For instance, the authors would further explore key species' responses beyond MTC – e.g., shifts in catch per unit effort (CPUE) or reproductive patterns for signature species like large palagic fish species. Also, at the conclusion section the manuscript could benefit from couple of lines developing policy recommendations.

Author Response

Reviewer 02

 

[Comments 01] The authors conducted a well-sound study on the tropicalization trend in the pelagic fisheries of the South Atlantic Ocean in relation to ocean warming. The manuscript based on a 41-year dataset to analyze changes in species composition, mean temperature of the catch, and the influence of environmental drivers. An important asset of the study is the fact that the authors provide access to data repositories (e.g., ICCAT, GitHub), which will enhance reproducibility of the study outcomes. Minor revisions are recommended to improve interpretability and further strengthen the robustness of the conclusions. For instance, the authors would further explore key species' responses beyond MTC – e.g., shifts in catch per unit effort (CPUE) or reproductive patterns for signature species like large palagic fish species. Also, at the conclusion section the manuscript could benefit from couple of lines developing policy recommendations.

[Response 01] Suggestions accepted and presented along the revised manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

Reviewer Comments – First Round

Thank you for the authors’ efforts in preparing this manuscript. The study compiles a valuable long-term dataset and addresses an important question—how South Atlantic pelagic fisheries respond to ocean warming—but key elements require substantial clarification and methodological reinforcement. In particular, the rationale for focusing on large predatory fish, the choice and evaluation of statistical models, data standardization, and the consistency of writing and interpretation must be strengthened before the manuscript can be considered for publication.

 

Specific Comments:

1. Introduction Focus: The Introduction flows well until it abruptly shifts to large predatory fish such as tuna and sharks. The ecological or management reason for spotlighting this group is unclear. If the only justification is that their indicators changed the most, that alone may not engage readers. Please clarify, from an ecological standpoint, why these top predators matter, for example, their role in top-down control of pelagic food webs or their heightened sensitivity to warming—and how this justifies making them the focal taxa of the study.
2. Catch-Data Standardization: The manuscript appears to analyze raw catch tonnage directly. Have you considered standardizing these data (e.g., Z-scores, CPUE, or converting to estimated numbers of individuals)? For example, if an average tuna weighs ~10 kg but an average shark weighs ~50 kg, five tuna equals one shark in weight. Using tonnage alone may therefore distort each species’ contribution when analyzing diversity or temperature preference (MTC). Please explore a standardization approach or run a sensitivity test to show how body-mass differences influence your results.
3. Model Choice (Linear vs Non-linear): You analyze SST–MTC trends solely with linear regression. Evaluate non-linear options and justify retaining a linear form based on residual patterns or information criteria.
4. Model Evaluation and Overfitting: R², AIC, and p-values are listed, but the priority order for model selection and any overfitting checks (validation set, cross-validation, adjusted R²) are missing. Detail the procedure and provide residual diagnostics.
5. Data/Results Consistency: Lines 8–10 in Methods 2.1 and the first sentence of the Results report the same statistic with different values. Remove duplication and ensure the numbers match.
6. Result Visualization: While the numerical presentation on page 6 (lines 2–7) is understandable, including a visual representation—such as a boxplot or trend-line figure—would enhance the reader’s grasp of the patterns.
7. Slope Notation: In the regression on page 6 (lines 12–17), it is unclear whether a or b represents the slope. Please specify the exact form (e.g., y = a x + b where a = slope).
8. Best-Fit Criterion: R² alone cannot define the “best” model. Present AIC/BIC, p-values, and residual checks (normality, homoscedasticity, autocorrelation) to justify the final model.
9. SEAO Results Omission: Section 3.2 explains SWAO trends in detail but scarcely discusses SEAO, even though it appears in Table X. Provide the same level of interpretation for SEAO.
10. Discussion Focus: The paper is framed as an assessment of warming-driven change, yet the Discussion repeatedly foregrounds human influences (over-fishing, market-driven selective harvest) and notes that these may obscure the environmental signal. This could give the impression that the study’s findings are overshadowed by human drivers, potentially diluting the intended focus on climatic impacts. Please separate anthropogenic pressures from climatic factors, discuss their respective roles, and ensure the Conclusion directly answers the original research question about warming effects.
11. Over-confident Statements: Lines 3–5 of the Discussion (p12) claim that large predators such as tuna and sharks “cannot occur outside 24 °C.” These species are generally eurythermal and can track prey into both slightly cooler (~23 °C) and warmer (~26–27 °C) waters. Shifts in their proportional catch could therefore reflect prey-driven movements, not a strict temperature ceiling. Please soften this wording and provide supporting evidence (e.g., published temperature ranges) or acknowledge alternative explanations.
12. Formatting Consistency:

• • Font style mismatch for last author – The last author’s name appears in a different font; please unify the style.
  • Terminology consistency – Standardize throughout the manuscript (e.g., p vs p-value, species names written as T. albacares in some places and Thunnus albacares in others). Figures and text should use identical notation.
  • Table 1 unit check – The final column “BCt range” is labelled in °C, but if it represents the transport volume of the Brazil Current this unit is inappropriate; verify and correct the unit.

Author Response

Reviewer 03

 

[Comments 01] Introduction Focus: The Introduction flows well until it abruptly shifts to large predatory fish such as tuna and sharks. The ecological or management reason for spotlighting this group is unclear. If the only justification is that their indicators changed the most, that alone may not engage readers. Please clarify, from an ecological standpoint, why these top predators matter, for example, their role in top-down control of pelagic food webs or their heightened sensitivity to warming—and how this justifies making them the focal taxa of the study.

[Response 01] Suggestions accepted. Please see revised manuscript lines 100-116.

 

[Comments 02] Catch-Data Standardization: The manuscript appears to analyze raw catch tonnage directly. Have you considered standardizing these data (e.g., Z-scores, CPUE, or converting to estimated numbers of individuals)? For example, if an average tuna weighs ~10 kg but an average shark weighs ~50 kg, five tuna equals one shark in weight. Using tonnage alone may therefore distort each species’ contribution when analyzing diversity or temperature preference (MTC). Please explore a standardization approach or run a sensitivity test to show how body-mass differences influence your results.

[Response 02] We understand that the reviewer suggests that altering the MTC index, expressing fish abundance in numbers, not in biomass as originally proposed by Cheung et al. (2019), could produce results that would change MTC time series trends obtained in this study and therefore our general conclusions. We understand this is a valid possibility, never addressed in the numerous studies that utilized this method after its original description, and deserving a proper methodological study.

            We, however, are unable to produce this analysis as part of this study, firstly because we expect to have comparable results with global analysis such as William Cheung’s and other studies that only applied MTC with species’ biomass in the catches, but most importantly because we do not have access to the necessary individual weight data of all species in the catches to properly convert biomass in numbers. Applying a “generic” mean weight, available in other databases, as suggested by the reviewer, to proceed such conversion would build a considerable uncertainty around the true numbers in the catches, that would fail to test the sensibility of our results to the effect of individual fish size on MTC.

            Additionally, the structured sensitivity analysis on iteratively removing one species at a time also helps in understanding the possible individual impacts of each species on the trends observed throughout the time series evaluated. And, in general, such effects tended to be maintained for almost all species. Still regarding the use of other quantities, such as the proposed case of using CPUE, in the form of a relative abundance index, for example, this approach has already been verified by Cheung et al. (2013), where the MTC was estimated based on capture and abundance indices of 55 demersal species, and the result of the comparison showed that there was no significant evidence to support the hypothesis of a difference between the trends observed in the MTC estimated on the different databases used. Corroborating the use of capture data as used in the present study.

 

[Comments 03] Model Choice (Linear vs Non-linear): You analyze SST–MTC trends solely with linear regression. Evaluate non-linear options and justify retaining a linear form based on residual patterns or information criteria.

[Response 03] Suggestion accepted and analysis were included in supplementary material.

 

[Comments 04] Model Evaluation and Overfitting: R², AIC, and p-values are listed, but the priority order for model selection and any overfitting checks (validation set, cross-validation, adjusted R²) are missing. Detail the procedure and provide residual diagnostics.

[Response 04] Suggestion accepted. Please see revised manuscript lines 255-264 and respective supplementary material.

 

[Comments 05] Data/Results Consistency: Lines 8–10 in Methods 2.1 and the first sentence of the Results report the same statistic with different values. Remove duplication and ensure the numbers match.

[Response 05] Suggestion accepted. Please see revised manuscript line 142.

 

[Comments 06] Result Visualization: While the numerical presentation on page 6 (lines 2–7) is understandable, including a visual representation—such as a boxplot or trend-line figure—would enhance the reader’s grasp of the patterns.

[Response 06] For consistency between Reviewers suggestions, two figures were included in supplementary material as a form attend to the suggestion made.

 

[Comments 07] Slope Notation: In the regression on page 6 (lines 12–17), it is unclear whether a or b represents the slope. Please specify the exact form (e.g., y = a x + b where a = slope).

[Response 07]. Suggestion accepted. Please see the revised manuscript lines 244-248.

 

[Comment 08] Best-Fit Criterion: R² alone cannot define the “best” model. Present AIC/BIC, p-values, and residual checks (normality, homoscedasticity, autocorrelation) to justify the final model.

[Response 08] The best fit were defined on the combination criterias as described in the Materials and Methods. For the simple regressions, the AIC criteria were not presented once the idea is not compare models. For these cases, the R² were just presented as a measure of the explicability of the model. As suggested, the residual checks were presented in the supplementary document.

 

[Comment 09] SEAO Results Omission: Section 3.2 explains SWAO trends in detail but scarcely discusses SEAO, even though it appears in Table X. Provide the same level of interpretation for SEAO.

[Response 09] Suggestion accepted. Please see the revised manuscript lines 491-504.

 

[Comments 10] Discussion Focus: The paper is framed as an assessment of warming-driven change, yet the Discussion repeatedly foregrounds human influences (over-fishing, market-driven selective harvest) and notes that these may obscure the environmental signal. This could give the impression that the study’s findings are overshadowed by human drivers, potentially diluting the intended focus on climatic impacts. Please separate anthropogenic pressures from climatic factors, discuss their respective roles, and ensure the Conclusion directly answers the original research question about warming effects. Over-confident Statements: Lines 3–5 of the Discussion (p12) claim that large predators such as tuna and sharks “cannot occur outside 24 °C.” These species are generally eurythermal and can track prey into both slightly cooler (~23 °C) and warmer (~26–27 °C) waters. Shifts in their proportional catch could therefore reflect prey-driven movements, not a strict temperature ceiling. Please soften this wording and provide supporting evidence (e.g., published temperature ranges) or acknowledge alternative explanations.

[Response 10] Suggestions accepted. Please see the revised manuscript lines 375-381.

 

[Comments 11] Font style mismatch for last author – The last author’s name appears in a different font; please unify the style. Terminology consistency – Standardize throughout the manuscript (e.g., p vs p-value, species names written as T. albacares in some places and Thunnus albacares in others). Figures and text should use identical notation. Table 1 unit check – The final column “BCt range” is labelled in °C, but if it represents the transport volume of the Brazil Current this unit is inappropriate; verify and correct the unit.

[Response 11] Suggestions accepted.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Reviewer Comments – 2nd Round

This manuscript offers a valuable contribution by examining long-term pelagic fisheries data in the South Atlantic Ocean to explore climate-driven tropicalization. I appreciate the authors’ effort to revise and supplement the manuscript following the first-round review. Upon reviewing the updated version, I suggest the following additional clarifications and improvements to further enhance the manuscript’s clarity, consistency, and analytical rigor.

 

Specific Comments:

• Clear Rationale for Target Species Selection and Effects Description: Lines 99–114 describe environmental pressure on tuna and tuna-like species, emphasizing their mobility and ecological role. However, it remains unclear why these species were selected as focal taxa, despite the presence of many thermally sensitive marine species. The rationale should go beyond general traits (e.g., mobility, trophic role) and explain why these species, specifically, are more suitable indicators of climate impacts in this region. Additionally, the description of ecological responses remains too vague. Phrases like “under considerable pressure,” “alterations in species composition,” and “trophic structure” are not sufficiently concrete. Please consider adding specific examples such as:
• Documented poleward shifts in tuna/shark distribution
• Observed changes in spawning behavior or migratory routes
  • Measurable trophic consequences from predator loss or warm-affinity species dominance

Such examples would enhance clarity and help the reader understand the significance of the biological changes described.

• Applicability of MTC to Mixed-Size Functional Groups: At the individual species level, it is plausible to approximate MTC using biomass data. However, when aggregating across functional groups like warm-affinity vs. cold-affinity species—which include a wide range of body sizes—the assumption that biomass-based MTC can reliably reflect abundance becomes more problematic. Because readers may differ in how they interpret such metrics, it is important that the authors clearly explain in the Discussion why they believe their biomass-based MTC is still valid in this context. If the authors are confident in this approach, a transparent discussion of its limitations and justification would help align readers’ expectations.
• Inconsistency in Nonlinear Regression Implementation and Interpretation: The authors mention that linear and nonlinear regression models were compared, but the exact types of nonlinear models used (e.g., quadratic, polynomial, logarithmic, logistic) are not specified. Moreover, in Supplementary Table 01, linear and nonlinear models show nearly identical AIC and R² values across most comparisons, which is statistically unusual and suggests that the same or equivalent model formulations may have been used. If nonlinear regression was properly tested using different model forms, the manuscript should clearly state which models were used, and if any outperformed the linear models, explain why the linear version was ultimately chosen. This justification should be included explicitly in the Methods section.
• Inconsistent Use of Model Selection Criteria: In the Methods section (L190–195), the authors indicate that AIC is the primary criterion for model selection. However, in the Results and Discussion, only R² values are consistently reported and interpreted. To maintain consistency, please clarify whether model selection was actually based on AIC, R², or both. If AIC was computed but not used, this should be stated transparently. Presenting AIC/BIC values only in the Supplement without referencing them in the main text may confuse readers.
• Over-Reliance on Supplementary Figures Without Explanation: While the supplementary file includes multiple figures (e.g., model diagnostics, non-linear fits, residual plots), none of these are accompanied by textual explanations in either the Supplement or the main manuscript. Figures alone—without captions, interpretation, or discussion—do not allow readers to understand why they are relevant or what conclusion can be drawn from them. At minimum, please provide brief text in either the Supplement or main text to explain:
  • What each figure is showing
  • What it implies about model fit or residuals
  • Whether the diagnostics supported the model assumptions

Author Response

Reviewer Comment 01:  Clear Rationale for Target Species Selection and Effects Description

• Lines 99–114 describe environmental pressure on tuna and tuna-like species, emphasizing their mobility and ecological role. However, it remains unclear why these species were selected as focal taxa, despite the presence of many thermally sensitive marine species. The rationale should go beyond general traits (e.g., mobility, trophic role) and explain why these species, specifically, are more suitable indicators of climate impacts in this region. Additionally, the description of ecological responses remains too vague. Phrases like “under considerable pressure,” “alterations in species composition,” and “trophic structure” are not sufficiently concrete. Please consider adding specific examples such as:

• Documented poleward shifts in tuna/shark distribution

• Observed changes in spawning behavior or migratory routes

  • Measurable trophic consequences from predator loss or warm-affinity species dominance

Such examples would enhance clarity and help the reader understand the significance of the biological changes described.

[Author's Answer]  Suggestion accepted and included in the new version of the manuscript between lines 70 and 75.

---

Reviewer Comment 02:  Applicability of MTC to Mixed-Size Functional Groups

At the individual species level, it is plausible to approximate MTC using biomass data. However, when aggregating across functional groups like warm-affinity vs. cold-affinity species—which include a wide range of body sizes—the assumption that biomass-based MTC can reliably reflect abundance becomes more problematic. Because readers may differ in how they interpret such metrics, it is important that the authors clearly explain in the Discussion why they believe their biomass-based MTC is still valid in this context. If the authors are confident in this approach, a transparent discussion of its limitations and justification would help align readers’ expectations.

[Author's Answer] Suggestion accepted. A rationale about the applicability of the MTC was included in the new version of the manuscript between lines 170 and 176. Although the aforementioned distinctions may potentially have an effect, it is essential to note that the MTC index was developed to address this type of problem. Since its proposal by Cheung et al. (2013), who applied this index globally, all subsequent studies have also been structured in the same way, including those based on catch weighting. Other indices in the same family, such as those proposed by the SeaAroundUs researchers, for example, also derive their weightings from catches of different functional groups, which essentially aim at representing biomass balance fluctuations in the ecosystem. In this sense, using the index as proposed does not seem erroneous. However, future studies may be considered to test the hypothesis proposed by the reviewer.

---

Reviewer Comment 03: Inconsistency in Nonlinear Regression Implementation and Interpretation

The authors mention that linear and nonlinear regression models were compared, but the exact types of nonlinear models used (e.g., quadratic, polynomial, logarithmic, logistic) are not specified. Moreover, in Supplementary Table 01, linear and nonlinear models show nearly identical AIC and R² values across most comparisons, which is statistically unusual and suggests that the same or equivalent model formulations may have been used. If nonlinear regression was properly tested using different model forms, the manuscript should clearly state which models were used, and if any outperformed the linear models, explain why the linear version was ultimately chosen. This justification should be included explicitly in the Methods section.

[Author's Answer]  Suggestion accepted and included in the new version of the manuscript between lines 189 and 213.

--- 

Reviewer Comment 04: Inconsistent Use of Model Selection Criteria

In the Methods section (L190–195), the authors indicate that AIC is the primary criterion for model selection. However, in the Results and Discussion, only R² values are consistently reported and interpreted. To maintain consistency, please clarify whether model selection was actually based on AIC, R², or both. If AIC was computed but not used, this should be stated transparently. Presenting AIC/BIC values only in the Supplement without referencing them in the main text may confuse readers.

[Author's Answer]  Suggestion accepted and included in the new version of the manuscript at lines 189 and 230. The Akaike information criterion was used as described in the manuscript only for the models that require comparisons between models and selections. We separate both cases in the material and methods presented in the manuscript. The first part is talking about the independent regressions of each isolated variable over time. And these results are reflected in Table 01 of the manuscript. For these cases, there is no reason to use AIC; just the R-squared and statistical significance of the slope parameter seem to be reasonable in the discussions of these results, once there are no comparisons to be made or selections. On the other hand, for the time-lagged linear models, here the AIC, R-squared, and significance were used for the analysis. And these results are reflected in Table 03. The inclusion of the AIC criteria for the simple regressions was done based on the reviewer's last request. The description of the steps taken was improved in the manuscript to allow a better interpretation of the methodological process conducted in this study.

--- 

Reviewer Comment 05:  Over-Reliance on Supplementary Figures Without Explanation

• While the supplementary file includes multiple figures (e.g., model diagnostics, non-linear fits, residual plots), none of these are accompanied by textual explanations in either the Supplement or the main manuscript. Figures alone—without captions, interpretation, or discussion—do not allow readers to understand why they are relevant or what conclusion can be drawn from them. At minimum, please provide brief text in either the Supplement or main text to explain:

  • What each figure is showing

  • What it implies about model fit or residuals

  • Whether the diagnostics supported the model assumptions

[Author's Answer]  Suggestion accepted. Supplementary materials has now a complete description and interpretation of each Table and Figure presented.

--- 

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript presents a clear research design, robust data interpretation, and meaningful academic and practical contributions. It is suitable for publication in its current form; I recommend final acceptance.