Modeling Plant Diversity Responses to Fire Recurrence in Disjunct Amazonian Savannas

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study examines the impact of fire frequency on the plant community structure, taxonomic composition, and functional diversity of Amazonian savannas. It elucidates the mechanisms through which fire affects taxonomic and functional diversity and underscores the significance of preserving these ecosystems amidst the backdrop of climate change.

 

The following are my recommendations for the authors to conduct a minor revision to align with the journal's publication criteria.

 

1- The clarity of Figure 1 is inadequate for identifying the names of each sampling point; it is recommended to substitute it with a more distinct version.

2- The caption for Figure 1 should be positioned on the same page as the figure.

3- Line 105: The term "savannahs" should be amended to "savannas," and this alteration should be uniformly applied throughout the manuscript.

4- Line 124: The text sequence following subsection 2.1 skips directly to 2.3, with 2.2 appearing subsequently—this numerical order is erroneous.

5- Line 136: The version of R software utilized is obsolete; it is advisable to upgrade to version 4 or a more recent iteration.

6- The resolution of Figure 2 is insufficient, necessitating its replacement with a version of higher clarity.

7- Lines 220-222: The inclusion of three blank lines is superfluous.

8- Lines 235-236: The insertion of two blank lines is unnecessary.

9- A more detailed description of the fire conditions in the study area is warranted, encompassing statistics on the causes of fires, the extent of the burned area, the duration of the burning episodes, and other pertinent details.

Comments for author File: Comments.pdf

Author Response

Comments 1- The clarity of Figure 1 is inadequate for identifying the names of each sampling point; it is recommended to substitute it with a more distinct version.

• Response 1: Thank you for your observation. We have replaced Figure 1 with a new version featuring improved resolution, enhanced color contrast, and clearer labeling of the sampling points. Additionally, we revised the figure caption to provide a more detailed description of the mapped elements and to reflect the visual improvements made.

 

Comments 2- The caption for Figure 1 should be positioned on the same page as the figure.

• Response 2: Thank you for the suggestion. The manuscript has been revised to position the caption for Figure 1 on the same page as the figure.

 

Comments 3- Line 105: The term "savannahs" should be amended to "savannas," and this alteration should be uniformly applied throughout the manuscript.

• Response 3: Thank you for the important observation. The term "savannahs" was corrected to "savannas" on line 124, and this revision has been applied consistently throughout the entire manuscript.

 

Comments 4- Line 124: The text sequence following subsection 2.1 skips directly to 2.3, with 2.2 appearing subsequently—this numerical order is erroneous.

• Response 4: Thank you for the important attention. The numbering error was a formatting mistake and has now been corrected.

 

Comments 5- Line 136: The version of R software utilized is obsolete; it is advisable to upgrade to version 4 or a more recent iteration.

• Response 5: Thank you for the valuable observation. The reference to an outdated version of R was the result of a citation oversight. This has been corrected, and the manuscript now specifies the use of R version 4.3.0.

 

Comments 6- The resolution of Figure 2 is insufficient, necessitating its replacement with a version of higher clarity.

• Response 6: We thank the reviewer for highlighting this issue. We have replaced Figure 2 with a higher-resolution version to improve clarity and readability.

 

Comments 7- Lines 220-222: The inclusion of three blank lines is superfluous.

• Response 7: We thank the reviewer for noting this formatting issue. The three blank lines have been removed in the revised version of the manuscript.

 

Comments 8- Lines 235-236: The insertion of two blank lines is unnecessary.

• Response 8: We thank the reviewer for noting this formatting issue. The three blank lines have been removed in the revised version of the manuscript

 

Comments 9- A more detailed description of the fire conditions in the study area is warranted, encompassing statistics on the causes of fires, the extent of the burned area, the duration of the burning episodes, and other pertinent details.

• Response 9- Thank you for your valuable comment. We appreciate the suggestion and have addressed it by adding information on the cumulative burned area over the study period, the peak fire occurrence season, and a citation of a relevant study on fire propensity in the region. These details have been included as a new paragraph in the Methods section.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for the opportunity to review your manuscript. The study addresses a relevant topic with potential to contribute to the understanding of fire dynamics and vegetation structure in savanna ecosystems.

However, I identified several issues regarding methodological clarity, data interpretation, and ecological reasoning. All my comments and suggestions have been inserted directly in the PDF file. I strongly recommend a careful revision, particularly regarding the treatment of categorical vs. continuous variables, and the need for more precise ecological framing.

Additionally, please note that the term “savannah” is used throughout the manuscript. The standard and scientifically accepted spelling is “savanna” (without the final 'h'). I recommend revising this for consistency with ecological literature.

I encourage the authors to revise the manuscript accordingly and provide a detailed response to each comment.

Best regards,

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Only minor improvements are needed in the English language. I suggested changes to the use of the terms "savanna" and "cerrado" for consistency with scientific standards.

Author Response

Reviewer 2

Comments 1- This sentence, as currently written, is vague and adds little value. All savannas, regardless of their location, have ecological importance. If the authors wish to emphasize a specific ecological role or a conservation concern unique to Amazonian savannas, they should be more precise and explain what makes these savannas particularly significant in the context of their study. “Overall, this study highlights the ecological importance of Amazonian savannas and underscores the need to integrate these ecosystems into fire management strategies, particularly under climate change scenarios that may intensify fire regimes.”

• Response 1: Thank you for your thoughtful comment regarding the need to clarify the significance of Amazonian savannas. We agree that the previous sentence was too general and did not clearly reflect the specific context of our study. In response, we revised the final sentence of the abstract to avoid vague language and to highlight the moderate resilience of these ecosystems to fire recurrence, as well as their relevance to fire management under climate change. The revised sentence now reads: “In conclusion, this study highlights the moderate resilience of Amazonian savannas to fire recurrence and emphasizes the need to incorporate these ecosystems into fire management plans under climate change scenarios.” (lines 36-39).

 

Comments 2 - In addition to the effects of fire and resource availability, species interactions—particularly competition—also mediate the coexistence of multiple life forms in savanna systems. I suggest incorporating this perspective into your introduction. See, for example, DOI: 10.1111/btp.13399, Abreu et al. (2023) (DOI: 10.1111/1365-2664.13994), which demonstrates how competitive interactions can promote the coexistence of trees, grasses, and subshrubs.

• Response 2- The introduction was revised to incorporate biotic interactions as additional drivers of vegetation dynamics in tropical savannas. The suggested references were very helpful and have been cited in support of this perspective. The authors thank the reviewer for indicating these relevant studies, which contributed to improving the ecological depth of the introduction. The revised text now reads: “Alongside environmental filters, biotic interactions such as competition among growth forms—such as grasses, trees, and subshrubs—also shapes vegetation structure and species coexistence (14). Fire, in turn, modulates these interactions by affecting recruitment and survival rates across functional groups, especially by limiting the establishment of certain woody species (15). (lines 50-54).

 

Comments 3- The text refers only to herbs and grasses as potentially favored by fire, but subshrubs may also benefit from fire regimes in savanna systems. I recommend including this functional group in your discussion. See, for instance, Zizka et al. (2015) – DOI: 10.1111/aec.12142, Bond (2016) – DOI: 10.1126/science.aad5132, which highlights the positive response of subshrubs to fire in neotropical savannas.

• Response 3- Thank you for the valuable observation. The text was revised to include subshrubs and shrubs as growth forms favored by frequent fire. Differences in fire response strategies between shrubs and trees were also incorporated, with emphasis on the greater resilience of shrubs due to traits like early reproduction and strong resprouting. The suggested references were cited to support these additions. The revised text now reads: “In Brazil's Cerrado—the largest tropical savanna in South America—frequent fires (annual or biennial) promote the coexistence of multiple growth forms and help sustain high biodiversity across herbaceous, grass, subshrub and shrub layer (16–18). According to (18) shrubs and trees exhibit distinct life-history strategies in response to fire, with shrubs showing an avoidance/adaptation strategy characterized by early reproduction and strong resprouting capacity after fire. Notably, fire-adapted shrubs tend to be more resilient than trees, showing a greater ability to withstand and recover from fire events (19).” (Lines 55-62).

 

Comments 4 - I am not sure whether the issue stems from the uploaded figure or from the PDF conversion process, but the quality of the map is currently poor. Please improve the resolution and overall clarity of the map to ensure it is readable and publication-ready. Savannas instead savannahs!!

• Response 4- Thank you for your observation. We have replaced Figure 1 with a new version featuring improved resolution, enhanced color contrast, and clearer labeling of the sampling points. Additionally, we revised the figure caption to provide a more detailed description of the mapped elements and to reflect the visual improvements made. The term "savannahs" was corrected to "savannas" on line 111, and this revision has been applied consistently throughout the entire manuscript.

 

Comments 5 - Throughout the manuscript, please revise the use of the term Cerrado. The capitalized form is typically used when referring to the biome as a whole. However, when referring to specific phytophysiognomies or vegetation types within the biome, the correct usage is lowercase (cerrado). Please ensure consistency and accuracy in terminology.

• Response 5- Thank you for the important observation, this revision has been applied consistently throughout the entire manuscript.

 

Comments 6 - The decision to transform originally continuous and informative variables into three categorical classes requires clearer justification. In my view, this transformation likely resulted in a loss of information and analytical power. The authors should explain the rationale behind this choice more thoroughly. Was the transformation motivated by interpretability, statistical constraints, or another reason? Please clarify how this decision contributes to the robustness or clarity of the analysis.

• Response 6- We thank the reviewer for the valuable comment and clarify that the transformation of continuous traits into ordinal categories was motivated by both ecological and analytical considerations. Several traits in our dataset exhibited highly contrasting distributions across plant growth forms. For example, traits such as plant height or stem diameter (DBH) naturally vary over orders of magnitude between trees and herbaceous species. Additionally, some traits are structurally exclusive — grasses lack a stem with measurable diameter, while woody species do not form a continuous herbaceous cover. These intrinsic biological differences posed challenges for direct quantitative comparison and increased the risk of overemphasizing traits with broader numerical ranges. By converting continuous traits into ecologically meaningful ordinal classes, we enabled more balanced and interpretable comparisons among species. The revised sentence now reads: “Although continuous traits are generally preferred for their higher resolution (56), we transformed them into ordinal categories (low, medium, high) to reduce scale imbalances, accommodate structural differences among growth forms, and improve ecological interpretability and analytical consistency.” (Lines 118-121).

 

 

Comments 7 - According to your description, the maximum value per pixel should be 26—representing areas burned every year during the 26-year period analyzed. However, in Figure 3, I observed values such as 6000 on the x-axis, which do not align with this logic. This raises concerns about how the values used in the analysis were derived. Please provide a detailed explanation of the methodology used to calculate these values. Without this clarification, your results are not replicable. Additionally, revise the x-axis scale in Figure 3 to accurately reflect the data and ensure consistency with your methods.

• Response 7- We thank the reviewer for raising this important point. The high fire recurrence values observed in Figure 3 do not represent individual pixel values, but rather the summation of all pixel fire frequencies within each circular buffer (1 km to 5 km radius) around each sampling point. We have now clarified this in the text: “To capture this broader influence, we reconstructed the fire history at five spatial scales by delineating circular buffers with radii of 1 km, 2 km, 3 km, 4 km, and 5 km from the central coordinates of each sampling point. Within each buffer, we summed the number of years in which each pixel burned over a 26-year period. Therefore, the fire recurrence values used in the analysis correspond to the total number of fire events across all pixels within each buffer.” (Lines 167-172).

 

Comments 8 - In the richness model (fire1), a similar pattern emerges: a single data point appears to strongly influence the observed unimodal distribution. I encourage the authors to carefully examine this point, as it may be disproportionately affecting the model's interpretation.

• Response 8- We thank the reviewer for the insightful comment regarding the richness model and the potential influence of a single data point. We carefully re-examined the results and agree that the decline in richness at high fire recurrence is driven in part by one site with notably low species richness. To address this, we revised the corresponding paragraph in the Results section (Lines 244-253).

 

Comments 9 - The interpretation presented seems inaccurate. When savanna physiognomies (vegetation types) are included as a random effect, the aim is not to assess the differences among them, but rather to control for their variability in order to estimate the general effect of fire across all savanna types. Please revise this statement accordingly, as the current wording suggests the opposite.

• Response 9- We thank the reviewer for this clarification. We agree that the original statement may have suggested an incorrect interpretation of the random effect structure. We have revised the sentence accordingly: “By including vegetation type (phytophysiognomy) as a random effect, the analysis controlled for structural and compositional differences among savanna formations, allowing a more generalized estimation of fire effects across all vegetation types” (Lines 293-295).

 

Comments 10 - Additionally, what are your considerations regarding fire exclusion? Since you report a unimodal pattern of encroachment and diversity, it is important to also discuss how fire exclusion may influence these dynamics.

• Response 10- We thank the reviewer for this pertinent suggestion. We have now incorporated this important consideration in a dedicated paragraph in the Discussion section (lines 326-333).

 

Comments 11 - Citation

• Response 11- The appropriate citations were added as requested. The reviewer’s suggestions were appreciated and helped strengthen the manuscript.

 

Comments 12 – Text removed

• Response 12- The sentence was removed as suggested.

 

Comments 13 – Text removed

• Response 13- The sentence was removed as suggested.

Reviewer 3 Report

Comments and Suggestions for Authors

Attached

Comments for author File: Comments.pdf

Author Response

Comment 1 – Introduction, lines 88-90. Is this statement a hypothesis? If not, it should be removed; if it is, then the authors should rephrase it to align with the study's focus.

• Response 1- The sentence was removed as suggested, as it did not reflect a hypothesis explicitly tested in the study.

 

Comment 2: Why 10 × 100 m plots, not 10 × 10 m or 10 × 20 m plots?

• Response 2: We thank the reviewer for the thoughtful comment. The 10 × 100 m plot configuration was chosen to enhance the detection of horizontal variation in vegetation structure, a common feature of Amazonian savannas. This elongated shape increases the likelihood of capturing the full diversity of life forms present across the physiognomy, particularly when woody individuals are sparsely and unevenly distributed.

 

Comments 3- Statement not necessary

• Response 3- We thank the reviewer for this observation. The sentence in lines was removed.

 

Comments 4- Results – Lines 183–185: Interpretation of “diverse and compositionally rich” communities

• Response 4- We thank the reviewer for raising this important point. The analyses were based on floristic surveys that included all recorded species; however, only two exotic species were identified, representing less than 1% of the total taxa. These species occurred sporadically and did not dominate the vegetation structure in any of the sampled physiognomies. Therefore, the higher richness and diversity observed in CR, CSU, and CSS reflect predominantly native plant communities, and not the result of compositional enrichment by alien species. A clarification has been added to the results section to address this concern (lines 196-204).

 

Comments 4- Referencing style should be aligned with the journal’s style

• Response 4- The comment regarding reference formatting was fully addressed. All references were revised using the Elsevier American Chemical Society citation style via Mendeley to ensure consistency and compliance with formatting standards.

 

 

Comment 5- Final remark

• Response 5- We thank the reviewer for the suggestion. A conclusion section was added to synthesize the main findings and highlight their ecological and management implications, as recommended.

Reviewer 4 Report

Comments and Suggestions for Authors

This study presents the results of an analysis based on vegetation census data and Generalized Additive Mixed Models to evaluate the relationships between fire frequency and diversity metrics in the state of Amapá in Brazil. I congratulate the authors on this
excellent research, which has identified some findings that suggest that recurrent fires may reduce taxonomic diversity without significantly altering functional structure.

The paper is very well written, and I have no major recommendations, except that it should include a conclusion section, in which the authors should indicate the significance of the study's main findings.

Author Response

Comments 1: Conclusion section

Response 1: We thank the reviewer for the suggestion. A conclusion section was added to synthesize the main findings and highlight their ecological and management implications, as recommended.