Assessment of Soil Quality in Peruvian Andean Smallholdings: A Comparative Study of PCA and Expert Opinion Approaches

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. The weight assignment in Table 1 is primarily based on "expert judgment combined with literature support," but the manuscript does not specify the number of experts involved, their backgrounds, or the degree of agreement among their assessments (e.g., Cohen’s Kappa).
2. Although the authors compare the mathematical behavior of linear and nonlinear scoring functions, such as distributional skewness and kurtosis, they lack a deeper ecological interpretation. For instance, does the nonlinear scoring better reflect the "threshold effects" commonly observed in plant responses? Does it more accurately capture the "marginal benefits" of agronomic interventions? These questions remain unexplored, limiting the ecological relevance of the scoring method selection.
3. It is recommended that the authors explicitly address how data distribution influences the stability of the spatial model in the discussion section. Additionally, future work could benefit from a multi-scale sampling density analysis and model sensitivity testing to better assess the robustness and reliability of the spatial predictions.
4. It is recommended that the authors further discuss the "manageability" and "responsiveness" of the indicators included in the SQI. Introducing a weighting adjustment based on the proportion of "actionable indicators", those that can be directly influenced through agronomic practices, could enhance the practical applicability of the index.

Author Response

Comment 1: The weight assignment in Table 1 is primarily based on "expert judgment combined with literature support," but the manuscript does not specify the number of experts involved, their backgrounds, or the degree of agreement among their assessments (e.g., Cohen’s Kappa).

Response 1: We appreciate the reviewer’s observation. The weights assigned in Table 1 were not obtained through a new expert consultation but were adopted from the work of Lenka et al. (2022) (DOI: https://doi.org/10.3389/fenvs.2022.865473), complemented with information from relevant soil science literature. To clarify this in the manuscript, the paragraph has been modified to read: “In this approach, the selection of primary soil properties was based on the criteria described by Lenka et al. [41], as well as on their recognized influence on soil fertility [42]” (lines 176–177).

Comment 2: Although the authors compare the mathematical behavior of linear and nonlinear scoring functions, such as distributional skewness and kurtosis, they lack a deeper ecological interpretation. For instance, does the nonlinear scoring better reflect the "threshold effects" commonly observed in plant responses? Does it more accurately capture the "marginal benefits" of agronomic interventions? These questions remain unexplored, limiting the ecological relevance of the scoring method selection.

Response 2: 

We appreciate this observation. In the paragraph: “In contrast, non-linear scoring through sigmoidal functions (Equation 3) has been demonstrated to moderate the influence of extreme values and amplify mid-range differences. This result yields distributions that are more balanced and reflective of functional performance. This is illustrated by the flatter, less skewed distribution of EONLinear. Previous studies emphasize the superiority of non-linear transformations for environmental indices, noting that soil functions do not always improve proportionally in conjunction with increases in indicator values [6,11,91]. Consequently, scoring methods that incorporate thresholds or diminishing returns—such as sigmoid curves—more accurately reflect ecological realities"(Lines 604-612). We address the advantages of non-linear scoring in agronomic and environmental contexts. This discussion implicitly considers the “threshold effects” and “marginal benefits” mentioned in the reviewer’s comment, as these concepts are inherent to the idea of diminishing returns and threshold-based responses in soil–plant systems.

Comment 3:  It is recommended that the authors explicitly address how data distribution influences the stability of the spatial model in the discussion section. Additionally, future work could benefit from a multi-scale sampling density analysis and model sensitivity testing to better assess the robustness and reliability of the spatial predictions.

Response 3: We thank the reviewer for this observation. A paragraph has been added in the discussion highlighting how data distribution influences the stability of the spatial model, specifically addressing variability, skewness, and the presence of outliers in edaphic datasets, as suggested: “This consideration is critical in edaphic datasets, which often exhibit high variability, skewness, and occasional values near or below detection limits [92,93]. These data characteristics can hinder statistical analyses and spatial interpolations, especially when using methods such as GWRK that rely on residual modeling [94]. By attenuating the disproportionate effect of outliers and enhancing sensitivity in the mid-range, nonlinear scoring facilitates the development of more robust soil quality indices and spatial representations.” (Lines 613-619).

Comment 4: It is recommended that the authors further discuss the "manageability" and "responsiveness" of the indicators included in the SQI. Introducing a weighting adjustment based on the proportion of "actionable indicators", those that can be directly influenced through agronomic practices, could enhance the practical applicability of the index.

Response 4: We appreciate the reviewer’s valuable suggestion. A paragraph has been added in the discussion section providing further detail on the manageability and responsiveness of the indicators used in the SQI: “Most of the evaluated indicators are commonly included in agricultural soil assessments. Some parameters, such as sand, silt, and clay content, exhibit high temporal stability and are not easily influenced by management interventions [81]. In contrast, other indicators—such as SOC, EC, pH, CEC, and aggregate stability—are more responsive to medium-term management practices [82]. Finally, certain variables, such as the balance of exchangeable cations and the availability of micronutrients, can respond more rapidly to targeted soil fertilization strategies [83].”(Lines 559-565). Regarding the second part of the comment, we note that the weighting assignment via PCA is purely data-driven and does not account for the environmental or agronomic significance of each indicator. In contrast, the expert opinion approach inherently incorporates such significance when assigning weights, and this aspect has also been further developed in the discussion section.

 

Reviewer 2 Report

Comments and Suggestions for Authors

(1) Line 1: Please indicate the type of the manuscript.

(2) In the first two sentences of the abstract, the research background and significance of this article have not been clearly introduced.

(3) In lines 14-15, how many indicators were the initial 24 indicators finally reduced to?

(4) At present, the logical structure of the abstract is not clear enough. Please reorganize and revise the content of this part. Meanwhile, the core research findings of this paper are not clearly presented in the abstract.

(5) Generally speaking, there should be a good correspondence between the abstract and the keywords, therefore, it would be better if the keyword “Labile soil carbon” and “Particulate organic matter” could be mentioned in the abstract.

(6) In the first paragraph of the introduction, the logical structure of this passage is chaotic. Please reorganize and improve it. In particular, the content in lines 41-43 seems more appropriate to be placed at the beginning of the paragraph.

(7) The important research methods or models to be adopted in this study and the innovativeness of this study have not been necessarily elaborated and discussed in the introduction.

(8) Subtitles need to be set for Figure 1 to clearly introduce the specific content of each sub-figure. Figure 1 also lacks necessary legends to label and explain essential geographical elements.

(9) Given that there are many research processes in this paper, it is recommended to draw a technical flowchart and supplement corresponding text explanations to illustrate the research ideas of this paper.

(10) There is a sharp contrast between 3.2 and 3.3. Is the analysis of expert opinions in this paper relatively simpler compared to the analysis of PCA?

(11) For the 2.5 section, necessary references should be provided in the statistical method part to enhance the reproducibility of scientific research.

(12) The four evaluation methods mentioned in 3.6 are not clearly explained in the data and methods section.

(13) In Figure 6, each subfigure is not fully displayed. Please readjust the layout of each figure.

(14) There is a large amount of content in the discussion section. It is recommended to use subheadings for narration to enhance the logic and coherence of the paper writing. Meanwhile, most of the content in the conclusion section seems more appropriate to be placed in the discussion section. Therefore, it is suggested to reorganize and rewrite the content of the conclusion section.

Author Response

Comment 1: Line 1: Please indicate the type of the manuscript.

Response 1: This observation has been addressed; the manuscript type has been specified in the revised version (Line 1).

Comment 2: In the first two sentences of the abstract, the research background and significance of this article have not been clearly introduced.

Response 2: This observation has been addressed; the first two sentences of the abstract have been reformulated to more clearly present the research background and significance (Lines 11-15).

Comment 3: In lines 14-15, how many indicators were the initial 24 indicators finally reduced to?

Response 3: This point has been addressed; we have added a specification indicating the final number of indicators used in both the PCA and Expert Opinion approaches (Line 18).

Comment 4: At present, the logical structure of the abstract is not clear enough. Please reorganize and revise the content of this part. Meanwhile, the core research findings of this paper are not clearly presented in the abstract.

Response 4: This observation has been addressed; the abstract has been reorganized to ensure a clearer logical structure, and the main research findings are now explicitly presented (Lines 11-33).

Comment 5: Generally speaking, there should be a good correspondence between the abstract and the keywords, therefore, it would be better if the keywords “Labile soil carbon” and “Particulate organic matter” could be mentioned in the abstract.

Response 5: This observation has been addressed; the keywords have been modified to ensure better alignment with the abstract content (Line 34).

Comment 6: In the first paragraph of the introduction, the logical structure of this passage is chaotic. Please reorganize and improve it. In particular, the content in lines 41-43 seems more appropriate to be placed at the beginning of the paragraph.

Response 6: The first paragraph of the introduction has been restructured, and the content originally in lines 41–43 has been moved to the beginning of the paragraph, as suggested by the reviewer (Lines 39-43).

Comment 7: The important research methods or models to be adopted in this study and the innovativeness of this study have not necessarily been elaborated and discussed in the introduction.

Response 7: We appreciate the reviewer’s comment. Following the suggestion, we have expanded the Introduction to elaborate on the research methods and the innovativeness of this study. Specifically, we have added the following sentence: "Moreover, the choice of normalization method plays a critical role, as it helps address common issues associated with raw data, such as non-normality and skewness [23]. This becomes particularly relevant when the construction of SQIs is complemented by spatial interpolation techniques like GWRK, which rely on the spatial structure of residuals to refine model predictions [24,25] and support the development of high-resolution soil quality maps." (Lines 80-85).

Comment 8: Subtitles need to be set for Figure 1 to clearly introduce the specific content of each sub-figure. Figure 1 also lacks necessary legends to label and explain essential geographical elements.

Response 8: A description for each sub-figure in Figure 1 has been added, and the map has been slightly modified. (Lines 125-127)

Comment 9: Given that there are many research processes in this paper, it is recommended to draw a technical flowchart and supplement corresponding text explanations to illustrate the research ideas of this paper.

Response 9: A figure (Figure 2) containing the flowchart summarizing the main steps followed in the research has been added, along with corresponding text explanations.(Lines 234-243)

Comment 10: There is a sharp contrast between 3.2 and 3.3. Is the analysis of expert opinions in this paper relatively simpler compared to the analysis of PCA?

Response 10: It is correct that the PCA section is more extensive, as the steps involved are more complex. In the case of the EO approach, the weighting of each indicator was based on the methodology described by Lenka et al. (2022), which made the process more straightforward. (Lines 176-178).

Comment 11: For the 2.5 section, necessary references should be provided in the statistical method part to enhance the reproducibility of scientific research.

Response 11: The necessary references have been added to Section 2.5 (statistical analysis) to enhance the reproducibility of the research (Lines 246-252).

Comment 12: The four evaluation methods mentioned in 3.6 are not clearly explained in the data and methods section.

Response 12: The Methods section was improved by adding a clearer description of how the four SQIs were developed, the suggested flowchart was incorporated to visually summarize the process.(Lines 234-243)

Comment 13: In Figure 6, each subfigure is not fully displayed. Please readjust the layout of each figure.

Response 13: Figures 6 and 7 (now Figures 7 and 8) were modified to improve their layout and ensure that each subfigure is fully displayed (Lines 462 and 495).

Comment 14: There is a large amount of content in the discussion section. It is recommended to use subheadings for narration to enhance the logic and coherence of the paper. Meanwhile, most of the content in the conclusion section seems more appropriate to be placed in the discussion section. Therefore, it is suggested to reorganize and rewrite the content of the conclusion section.

Response 14: Subheadings were added to the discussion section to improve its logical structure and coherence. In addition, the conclusion section was rewritten to be more concise, and some content was moved to the discussion section as suggested.(Lines 670-683)

Reviewer 3 Report

Comments and Suggestions for Authors

This work focus on the evaluation of soil quality in a mountain region at Peru via important soil parameters data processing via PCALinear, PCANLinear, EOLinear and EONLinear. It is very meaningful for the local farming industry. However, there are serious issues for the authors to address:

1) improper writing and serious typos. For instance, in the abstract Line 13, What is SQI. We have to provide the whole name of a concept when we introduce it the first time in the manuscript. In line 16, When the full name of PCA is introduced, it is better to add their abbreviation immediately behind the full name with parentheses. Such issues are quite common in this manuscript. Others are not specified but the team has to check and correct all of them. In Line 44, why the word font is bold here? In line 45 Why the reference citation format keeps changing? Please keep it consistent and use Arabic number instead as the majority do. Also, the references are not enough to support your claim that most of the studies only focus on field scale. At least 5-6 highly cited articles to support your claim. Otherwise it is not convincing at all. In Line 105, If the method GWRK is not developed by your team, it is a must to cite the original paper. In line 196, 199, SL, SNL, please use subscript for L and NL to keep consistent with your equations. It should be l not 1 in Line 196. In line 482, why there are two figures labeled as Figure 6?

2) Based on Table 7 and Fig 5, there is no significant difference in mean and median data between PCALinear, PCANLinear, EOLinear, EONLinear. Why the authors keeep claiming there is big difference in Line 428-434?

3) Based on the explanation of SQI, the higher, the better quality of soil is. For the 20 or so soil parameters considered, it seems that one important parameter is missing: the elevation of the farm land, or the height of the land. Because in many cases, the fertility of the soil or quality of soil is strongly affected by the elevation. The trend in the final Figure 7 may be mainly affected by the missing parameter height of the land. What do you think?

Author Response

Comment 1: Improper writing and serious typos. For instance, in the abstract Line 13, What is SQI. We have to provide the whole name of a concept when we introduce it the first time in the manuscript. In line 16, When the full name of PCA is introduced, it is better to add their abbreviation immediately behind the full name with parentheses. Such issues are quite common in this manuscript. Others are not specified but the team has to check and correct all of them.

Response 1: We have carefully revised the entire manuscript to correct improper writing, typos, and inconsistent abbreviation use. All concepts are now presented with their full name at first mention, followed immediately by their abbreviation in parentheses. This correction was applied consistently throughout the abstract, main text, figures, and tables to ensure clarity and avoid confusion.

Comment 2: In Line 44, why the word font is bold here?

Response 2: The unintended bold formatting in line 44 has been corrected in the revised manuscript (Line 45)

Comment 3: In line 45 Why the reference citation format keeps changing? Please keep it consistent and use Arabic number instead as the majority do. Also, the references are not enough to support your claim that most of the studies only focus on field scale. At least 5-6 highly cited articles to support your claim. Otherwise it is not convincing at all.

Response 3: The statement in line 45 has been rephrased to be less restrictive and now reads: "The application of the SQI as a tool for assessing management practices has been explored across various spatial scales, including crop-level, experimental field, and regional contexts [7–10]". Additionally, the citation format has been corrected to maintain consistency with Arabic numbering throughout the manuscript (Lines 45-47)

Comment 4: In Line 105, If the method GWRK is not developed by your team, it is a must to cite the original paper.

Response 4: A paragraph has been added where the first mention of the GWRK method now includes citations to the original articles in which it was developed.(Lines 80–85)

Comment 5: In line 196, 199, SL, SNL, please use subscript for L and NL to keep consistent with your equations.

Response 5: All necessary modifications have been made so that L and NL now appear as subscripts (SL, SNL) in lines 197 and 201, ensuring consistency with the equations.

Comment 6: It should be l not 1 in Line 196.

Response 6: The necessary corrections have been made to replace the number “1” with the lowercase letter “l”. (Line 197)

Comment 7: In line 482, why there are two figures labeled as Figure 6?

Response 7: The figure numbering has been corrected, and the figure previously mislabeled as "Figure 6" is now correctly labeled as Figure 8 (Line 495)

Comment 8: Based on Table 7 and Fig 5, there is no significant difference in mean and median data between PCALinear, PCANLinear, EOLinear, EONLinear. Why the authors keeep claiming there is big difference in Line 428-434?.

Response 8: Although the absolute differences in mean and median values between PCALinear, PCANLinear, EOLinear, and EONLinear are small, both the Friedman test and the Nemenyi post-hoc test (Table 7) indicate that these differences are statistically significant. This is further supported by the grouping letters shown in Figure 5, where the methods fall into distinct statistical groups.

Comment 9: Based on the explanation of SQI, the higher, the better quality of soil is. For the 20 or so soil parameters considered, it seems that one important parameter is missing: the elevation of the farm land, or the height of the land. Because in many cases, the fertility of the soil or quality of soil is strongly affected by the elevation. The trend in the final Figure 7 may be mainly affected by the missing parameter height of the land. What do you think?

Response 9: We appreciate the reviewer’s comment. The experimental design for SQI construction was intentionally restricted to edaphic properties, excluding climatic or topographic variables such as elevation from the index construction in order to keep the SQI as a purely soil-property-based measure. However, elevation, together with NDMI, aspect, and slope, was indeed incorporated in the spatial modeling stage used to generate the soil quality maps (Figure. 8). Therefore, while elevation did not directly influence the SQI values, it was considered in the geospatial prediction of SQI distribution, helping to capture its potential influence on soil quality patterns.

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

In comparison to its previous iteration, we are pleased to observe that the author has implemented modifications and enhancements across the majority of the content. Nevertheless, there remain some aspects which necessitate further refinement.

(1) The second sentence in the introduction carries little significance and lacks sufficient relevance to the core content of the article; it is recommended to be deleted.

(2) Consistent with the previous review comments, the important research methods or models to be adopted in this study and the innovativeness of this study have not been necessarily elaborated and discussed in the introduction. For example, the purposes, advantages, and disadvantages of the two scoring methods mentioned in the last paragraph of the introduction were not elaborated upon in the introduction.

(3) Although the author has refined the flowchart, the introduction to the flowchart could be placed within the main text rather than entirely in the caption of Figure 2.

(4) Please revise and improve the title of Section 4.2 in Discussion. At the same time, it is recommended that the discussion section should also provide a thorough elaboration and analysis of the limitations and future prospects of this study. What is the scientific theoretical value of this research, and what reference value does it hold for the development of other similar regions? These aspects should also be deeply explored in the discussion.

Author Response

Comment 1: The second sentence in the introduction carries little significance and lacks sufficient relevance to the core content of the article; it is recommended to be deleted.

Response 1: We appreciate the reviewer’s observation. The second sentence in the introduction has been removed as suggested.

Comment 2: We appreciate the reviewer’s insightful comment. In line with the suggestion, we have expanded the introduction to provide more details on the purposes, advantages, and disadvantages of the two scoring methods. This addition highlights the rationale for their selection and better frames the methodological and innovative aspects of the study (Lines 79-96).

Comment 3: Although the author has refined the flowchart, the introduction to the flowchart could be placed within the main text rather than entirely in the caption of Figure 2.

Response 3: We thank the reviewer for the suggestion. The explanation of the flowchart has been moved from the caption to the main text (Lines 245-250).

Comment 4: Please revise and improve the title of Section 4.2 in Discussion. At the same time, it is recommended that the discussion section should also provide a thorough elaboration and analysis of the limitations and future prospects of this study. What is the scientific theoretical value of this research, and what reference value does it hold for the development of other similar regions? These aspects should also be deeply explored in the discussion.

Response 4: We thank the reviewer for the suggestion. The title of Section 4.2 has been revised to “Comparative Analysis of SQI Construction”, and additional discussion has been incorporated to address the scientific value, limitations, and reference value of this study for similar regions. (Lines 666-685)

 

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed most of my comments well except the last one. It is not so convincing and in soil studies, one of the most important parameter is elevation. It may dominate the team's SQI analysis of the region. How to convince the readers of the data when such important parameter is missing? Please either address it seriously or put a claim: elevation is an important factor in determing SQI, and the SQI result shown in this work excluding this factor due to xxx.

Author Response

Comment 1: The authors have addressed most of my comments well except the last one. It is not so convincing and in soil studies, one of the most important parameter is elevation. It may dominate the team's SQI analysis of the region. How to convince the readers of the data when such important parameter is missing? Please either address it seriously or put a claim: elevation is an important factor in determing SQI, and the SQI result shown in this work excluding this factor due to xxx.

Response 1: We appreciate the reviewer’s comment. Elevation itself is not a soil property, but rather an external factor that indirectly influences soil conditions through variables such as soil carbon fractions, which were directly measured and included among the 24 soil parameters used in our SQI construction. These measurements already capture the influence that elevation may exert on soil quality. For this reason, we consider that incorporating elevation as an additional factor is unnecessary. In studies with fewer direct soil measurements, elevation may indeed serve as a useful proxy; however, in the present work, the extensive set of soil variables already accounts for its potential effects. Therefore, no additional reference to elevation in relation to SQI construction will be included.