Socioeconomic and Environmental Dimensions of Agriculture, Livestock, and Fisheries: A Network Study on Carbon and Water Footprints in Global Food Trade

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper is a theoretical and reflexive article, making an interesting contribution to the analysis of standardization processes from the human factor perspective, utilizing the systems thinking approach. It is a current and relevant theme for the journal Standards. However, minor to moderate revisions are needed. The article would benefit from a clearer structuring of the results and a more rigorous delimitation between methods and discussions.

 

1. Introduction

The introduction is well documented and effectively explains the context and relevance of standardization processes, emphasizing the complex role of the human factor in these processes. The authors correctly point out the need for systemic approaches in understanding standardization.

Comments:

1. It is recommended to clearly formulate the purpose and objectives of the paper at the end of the introduction.
2. It would also be useful to explicitly delimit the original contribution of the paper in relation to the existing literature.
3. Some passages can be synthesized for better readability.

 

2. Materials and Methods

The authors coherently present the theoretical framework of the analysis method, based on the conceptualization of systems and the understanding of the soft elements of systems (soft systems methodology).

Comments

4. A clearer explanation of the method used is recommended (is it a conceptual study? a case study? an applied systems analysis?), because in its current form it is not clear whether the methodology is empirical or theoretical.
5. A conceptual scheme or model that clearly highlights the proposed systemic approach would be useful.
6. A description of how the data were collected or processed is missing – if it is a purely theoretical article, this aspect should be specified.

 

3. Results

The results are structured around an analysis of the dynamics of standardization from the perspective of the human factor, using rich picture and causal loop diagrams, which makes an interesting visual contribution.

Comments

7. A clearer demarcation between results and their interpretation is needed – in its current form, the analysis seems to overlap with the discussion.
8. The authors could include a short narrative description under each diagram to better guide the reader.
9. It is recommended to add concrete results obtained from the application of the theoretical model or at least some testable hypotheses.

 

4. Discussion

The discussion section reflects a deep understanding of the systemic implications of the human factor in standardization. The authors connect the specialized literature well with the proposed model.

Comments

10. A more applied discussion on the possible consequences of the proposed model in standardization practice (e.g. in industry, health, education) would be useful.
11. It is recommended to include some limitations of the research and concrete suggestions for future research.
12. The integration of practical examples would strengthen the validity of the discussion.

 

5. Conclusions

It summarizes the main findings and reaffirm the importance of integrating the human factor into the systemic analysis of standardization.

Comments

13. A section of “policy implications” or “practical recommendations” would be useful to emphasize the applicative relevance.
14. The conclusions can be developed to better reflect the theoretical contribution of the article and the directions for further development.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Abstract

• Clarity and Coverage: The abstract provides a good overview of the study’s context, objective, and key findings. It clearly states the use of complex network analysis to examine carbon and water footprints in global food trade and the association with economic welfare. The main results are summarized – for instance, it notes that the carbon and water footprints embedded in food trade increased from 1986 to 2020 and that countries’ network centrality is linked to higher GDP per capita. This gives readers an accurate snapshot of the study’s outcomes.
• Style and Wording: Overall, the language is clear, though a couple of sentences are quite long. For example, the sentence “countries with higher imports of carbon and water through global food trade obtain economic benefits” could imply direct causality. It is advisable to nuance this phrasing to clarify that the relationship observed is a correlation rather than a proven causal effect. For instance, you might rephrase to “…tend to exhibit greater economic welfare” instead of “obtain economic benefits”.
• Essential Content: The abstract includes the necessary components – background context, a brief mention of methods (e.g., network analysis and regression models), main results, and a broad implication. It is concise (around 200 words) and aligns well with the content of the paper. As a minor improvement, consider highlighting the novelty or unique contribution of the study in one sentence. For example, explicitly state that this is the first study linking environmental footprints and economic welfare in a global trade network context. However, even without this addition, the abstract is effective and correctly represents the article.

Introduction

• Context and Relevance: The introduction provides a comprehensive backdrop, touching on global issues of hunger, poverty, and environmental degradation and framing them as systemic issues interconnected via global production and consumption networks. The text emphasizes how globalization and trade can create local economic pressures and environmental burdens, which sets a clear stage for the research problem. This establishes the relevance of examining the sustainability of global food trade – readers can immediately see why this topic matters.
• Structure and Clarity: The introduction is logically structured, moving from broad global challenges to the specific gap the study addresses. The transitions are smooth – for instance, the text mentions the need to investigate “indirect hidden costs linked to production and trade impacts to support advances in equal access to resources”. This directly leads to the research objective. The objective of the study is explicitly stated at the end of the introduction: “Therefore, the objective of the study is to investigate the potential embodied impacts of net environmental pressure corresponding to the carbon and water footprints involved in global food trade networks between 1986 and 2020, and their potential effects on economic welfare.”. This clear statement of aim is excellent and lets the reader know exactly what to expect.
• Minor Suggestions: The introduction could more explicitly highlight the novelty and distinct contribution of the study, in contrast to prior work. For example, after summarizing past studies (including the authors’ own [7,8]), you might add a sentence to emphasize that this research uniquely links environmental footprints with economic welfare in a complex network framework – something that has not been directly addressed before. This helps readers and reviewers recognize the originality of your work. Overall, the introduction is well-written and sets a solid foundation for the paper, with a clear articulation of the research question and its importance.

Methodology (Materials and Methods)

• Overview: The methodology section is detailed and organized into clear subsections (2.1 Study design, 2.2 Data sources, 2.3 Variables, 2.4 Network analyses, 2.5 Statistical analyses). This structured approach greatly aids comprehension. The authors clearly describe a two-pronged strategy: first, network analysis of carbon and water footprint flows in food trade; second, regression analysis of associations between countries’ characteristics (including network position) and economic outcomes. The fact that two distinct datasets were built for these purposes is explicitly explained using bullet points. The methodological design is sound and well-communicated.
• Data Sources and Variables: The data sources are reputable and appropriate. Food trade flows come from FAOSTAT’s Detailed Trade Matrix and World Bank’s WITS, which cover a broad range of products and ensure global coverage from 1986–2020. The World Bank’s income classification is used to categorize countries by income level, and footprint intensities for products come from the Su-EATABLE LIFE (SEL) database. These sources are up-to-date (FAO/WITS up to 2020, SEL published 2021) and credible. The variables included in the regression dataset are extensive and relevant: demographic structure (population by age/gender, urbanization, etc.), economic indicators (unemployment, GDP per capita in PPP), food dependency ratio, globalization indices (economic, social, political – de facto and de jure), and network metrics (indegree, outdegree, betweenness centrality for the footprint network). Tables 1 and 4 list these variables and their summary stats. Including such a broad set of controls demonstrates a strong effort to account for confounding factors and is a notable strength of the methodology.
• Methodological Detail: The authors provide enough detail to understand how key metrics were derived. They present a general formula (Equation 1, lines 169-170) describing the calculation of footprints transferred in trade: essentially F_x^ijt = E_x^ijt × F_x, meaning the environmental footprint flow from country i to j at time t equals the trade volume times the footprint intensity. This clarifies how carbon/water footprints are attached to trade flows.
• Network Analysis: Section 2.4 explains that the bilateral trade connections were represented as adjacency matrices for each year and that network metrics were computed. It would help to include brief definitions of key network metrics for readers who are not experts. While the manuscript references a network theory text for background, adding a sentence or two to define indegree centrality (number or weight of incoming connections for a country), betweenness centrality (extent to which a country lies on paths between others, indicating an intermediary role), etc., would improve clarity. As it stands, the metrics can be inferred from context and tables, but explicit definitions in text would be beneficial. Nonetheless, it’s clear that standard metrics (degree, density, clustering coefficient, path length, etc.) were calculated to characterize the network’s evolution (Table 2 shows these).
• Statistical Analysis: Section 2.5 outlines the regression approach. It’s described that regression models were used to identify effects of network position on GDP per capita, including controls for other characteristics. One commendable aspect is that the authors later clarify in the text that they included fixed effects for region and year and even region×year interaction terms. This indicates a panel data approach accounting for unobserved regional and temporal factors (like regional trade agreements or global trends), which is excellent for methodological robustness. They also mention using lagged GDP (0, 1, 3-year lags) to test short- and medium-term associations. This addresses potential reverse causality and ensures the findings are not just contemporaneous correlations. These methodological choices (fixed effects, lags, multiple controls) demonstrate a rigorous analytical strategy.
• Reproducibility and Transparency: The data sources are all public and properly cited [33–40], and the authors include a Data Availability Statement indicating where the data can be obtained (FAO, WITS, etc.). This transparency is good for reproducibility. One suggestion to further enhance reproducibility: specify the software or packages used for analysis (e.g., network analysis and regression). Mentioning if the analysis was done in R (perhaps using igraph, plm, etc.) or Python, and any key libraries, would be a helpful detail for interested readers, though it’s not mandatory. The current level of detail is likely sufficient for a knowledgeable reader to follow the methodology.
• Additional Consideration: One limitation inherent to the methodology, which you partly acknowledge later, is that the footprint intensities are not country-specific – they are average values from the SEL database. This means the analysis assumes, for example, that a ton of wheat traded carries the same carbon/water footprint regardless of origin. In reality, production practices differ by country. You hint at this in your discussion by saying footprint indicators are “notional concepts” and decisions are detached from local impacts. It might be worth explicitly mentioning in the methods or limitations that using generic footprint factors is a necessary simplification given data availability, and perhaps cite a source or two on how footprint values can vary regionally. This is a minor point and mostly covered by your third limitation, but making it explicit helps readers understand the assumptions of the footprint calculations.
• Conclusion on Methods: The methods section is one of the strengths of the paper, as it thoroughly explains data collection, preparation, and analysis steps. There are no major flaws apparent. The suggestions above (defining terms, acknowledging assumptions) are relatively minor clarifications. In summary, the methodology is appropriate for the research questions and is described with sufficient detail to instill confidence in the results.

Results

• Organization: The results section is clearly divided into two main parts corresponding to the research approach: 3.1 Network analyses and 3.2 Statistical analyses. This separation reflects the two-step study design and makes it easier for readers to follow. The authors present findings through a combination of text, tables, and figures, ensuring that important patterns are both described and illustrated.
• Network Analysis Results: The paper reports how the global food trade network (for embedded carbon and water footprints) evolved over time. Table 2 provides network metrics for selected years (1986, 1990, 2000, 2010, 2020), and the text highlights the trends. For example, the average degree nearly triples (from ~21.8 in 1986 to ~62.4 in 2020; Table 2), indicating a dramatic increase in connectivity of the trade network. Network density also rises significantly (0.086 to 0.246), meaning countries are more interconnected in 2020 than in 1986. The average clustering coefficient increases modestly (0.477 to 0.620), suggesting moderately stronger clustering (group formation) over time. The authors correctly interpret these metrics in the text, noting that the small increase in clustering combined with a slight decrease in average path length and stable diameter indicates that distances between countries in the network have decreased and connections form a bit faster (i.e., a more integrated network with low latency).
• Role of Countries by Income Level: Table 3 summarizes the share of carbon and water footprint trade accounted for by countries in different income categories over time. Key findings are that the share of high-income countries in both exported and imported footprints declined from 1986 to 2020 (e.g., carbon export share from ~69% to ~59%, carbon import share from ~66% to ~57%). Conversely, upper middle-income countries greatly increased their share (carbon export from ~9% to ~28%, import from ~11% to ~29%). The text points out these trends clearly: high-income countries, while still dominant, have a decreasing portion of the pie, whereas emerging economies have a substantial rise in participation. Meanwhile, low-income countries remain marginal in the network by 2020 (<2% share), which is an important observation about equity and is consistent with prior findings in literature. These results effectively answer part of the research question by showing who is driving global resource trade and how that has changed over time.
• Figures (Network Graphs): The paper includes network visualizations (Figures 1 and 2). Figure 1 (as described in the text) shows the global trade networks for carbon and water footprints in years 1986, 2000, and 2020, with node size proportional to degree and nodes colored by income group . Figure 2 appears to present perhaps aggregated connections or another perspective. These figures reinforce the quantitative results: the graphs for different years highlight the increasing intensity of connections and the prominent role of high-income countries (they are visually central and larger nodes in earlier years, though by 2020 one might see more mid-income country involvement). The legend indicating colors by income group is helpful. To ensure these figures are effective, the authors should verify they are of high resolution and clear in the final submission so that readers can discern the details. In the text, the authors already describe what the figures illustrate: e.g., “the central protagonism of high-income countries is highlighted” and the pattern that most connections involve high-income nodes persists over time. Perhaps the authors could add any specific insight visible in the graphs that isn’t already numeric; for example, if certain emerging economies became significantly more connected by 2020 (like China, India, etc.), it could be mentioned. But even without naming countries, the current description focusing on income categories is clear. Overall, the figures support the findings well.
• Statistical Analysis Results: Section 3.2 presents the associations between network position metrics and economic welfare (GDP per capita). The authors report that the regression coefficients show a significant positive association between a country’s indegree centrality and betweenness centrality in the trade network and its GDP per capita. In plain terms, countries that import more environmental resources (high indegree) and those that act as key intermediaries (high betweenness) in the global food trade tend to have higher income levels. This is a central finding and is explicitly stated (lines 688-693). The authors further note that this association was examined with lagged GDP values (0, 1, 3-year lags) and remained positive in the short- and medium-term, which suggests some robustness to the timing of the relationship.
• Interpretation of Coefficients: It is excellent that the authors included Table 5, which details the regression coefficients for the models at lag 0, lag 1, and lag 3. This table  provides the quantitative evidence behind the textual summary. For example, the coefficient for indegree centrality is ~0.122 (p<0.05) at lag 0 and increases to ~0.266 (p<0.01) at lag 1 (meaning a country that increases its indegree centrality sees significantly higher GDP per capita one year later, controlling for other factors). Betweenness centrality has a large coefficient (~9 to 11, p<0.001 across models), reflecting that a small change in betweenness (which is scaled 0–1 in network terms, I presume) corresponds to a large difference in GDP (understandable due to how betweenness is measured). The table also shows that most control variables behave as expected: e.g., a higher working-age population share, higher male population percentage (likely proxying overall population structure), higher urbanization, and longer life expectancy all correlate with higher GDP (significant positive coefficients). Unemployment has a small negative coefficient, and interestingly, food dependency ratio has a very small negative coefficient (though significant due to large N). The globalization indices show some mixed effects (social globalization strongly positive, economic globalization de facto slightly negative, etc.), which could be briefly remarked on if space allowed. However, the authors wisely focus their narrative on the network metrics in the text and leave these details in the table.
• Statistical Significance and Robustness: The authors indicate statistical significance by wording (“significant positive association”) and presumably used standard thresholds (p<0.05, p<0.01, etc. as shown by * in Table 5). They also mention performing checks to avoid bias in estimations, suggesting they did tests like perhaps multicollinearity checks or tried alternative model specifications. One suggestion: The results text could explicitly mention the inclusion of region and year controls (as footnoted in Table 5) because that is an important part of the model’s rigor. It is hinted at in the methodology (and footnote), but restating it in results or discussion – that the effect holds even after controlling for regional differences and global time trends – would underscore the robustness of the finding. Nonetheless, the discussion does articulate caution regarding confounding, so it’s acceptable.
• Addressing Research Questions: The results clearly address the research objectives. The first objective (to examine trends in embedded footprints from 1986–2020) is fulfilled by the network analysis results showing increased footprints and connectivity. The second objective (to explore associations with economic welfare) is directly answered by the regression outcomes indicating those associations exist and are positive. The authors do a good job not just presenting results, but also explaining what they mean. For instance, they interpret the indegree/GDP link as countries benefiting economically from participating in global trade, and they relate patterns like clustering and modularity to how the network is structured.
• Use of Tables/Figures: Each table and figure is utilized in the text. There is no extraneous data. Table 1 and Table 4, which are descriptive, are referenced indirectly when describing demographic and globalization trends (e.g., line 666-674 notes ongoing demographic transition, rise in urban pop, etc., reflecting data from Table 4). This helps ground the interpretation of regression results. All figures are referenced and help illustrate points (as discussed above). There is a nice balance between numerical reporting (in tables) and high-level interpretation (in text). The text does not drown in numbers, but instead highlights the key numbers that support the arguments.
• Minor Suggestions for Results: One important conclusion drawn is that network participation is linked to economic benefits. The authors should ensure not to overstate this as causal (which they mostly avoid). In the results section, they phrase it as a linkage (“are linked to higher income”, line 688-691) – this is appropriately cautious. They might add a phrase like “associated with” to reinforce that interpretation if needed. Another small suggestion: the authors might consider quantifying the magnitude of the increase in footprints over time in absolute terms (if data is available). For example, stating something like “the total carbon footprint exchanged grew by X% or from Y to Z metric tons CO2eq between 1986 and 2020” would give a concrete sense of scale to the statement that footprints increased. If such total numbers are easily obtainable from their data, mentioning them could strengthen the impact of the results. However, this is not strictly necessary as the focus is on network and share metrics.
• Overall Impression of Results: The results section is very well done. It is comprehensive and yet digestible. The findings are clearly presented and logically ordered. I do not see any unwarranted claims or unsupported statements – everything is backed by data from the study. The authors have effectively communicated the key patterns and relationships that emerged from the analysis. There are only minor clarifications or additions as noted, none of which require new analysis.

Discussion

• Summary of Key Findings: The discussion begins by summarizing the main findings of the study, noting the intensification of global food trade networks and the shifting participation (emerging economies rising, etc.). This recap helps reinforce the take-home messages for the reader. It is stated that the evolution of the trade network involved more intense bilateral exchanges and that these findings align with expectations. Immediately linking these findings with broader implications, the authors note that their results confirm patterns identified by previous studies, such as the minimal participation of low-income countries.
• Interpretation and Theoretical Implications: The authors go beyond restating results; they explore what these results mean in the larger context of food system sustainability. For example, they discuss the synergies and trade-offs between different sustainability dimensions. One interesting point is the mention that improvements in one sustainability indicator can exacerbate another – specifically, that healthier diets with lower greenhouse emissions can increase water footprint. This acknowledges that the problem is complex, which is very relevant to policy discussions. By bringing this up, the authors hint that simply encouraging shifts in diet may have unintended environmental trade-offs, underscoring the need for integrated assessments. They then connect this to their findings by implying that national data systems need to integrate environmental and economic indicators to identify such trade-offs. This shows a deep understanding of the subject matter and places the study’s findings within the context of multi-dimensional sustainability challenges.
• Connection to Literature: The discussion is rich with references to existing literature, demonstrating thorough engagement with other research. The authors compare their findings to prior network analyses of global food trade and related topics. For instance, they mention that what they found regarding low-income countries’ negligible role reinforces findings of earlier studies on global trade networks. They also reference studies about diet acceptability and consumer behavior to discuss that even if sustainable diets are affordable, cultural acceptance can be a barrier. They highlight contradictory effects of international relations on economic, environmental, and health outcomes. These references show that the authors are aware of the multifaceted nature of the problem and that their work touches on, and is informed by, various strands of research (network theory, sustainability science, nutrition, economics).
• Acknowledgement of Limitations: A particularly strong aspect of the discussion is the transparent acknowledgement of the study’s limitations. The authors systematically list several key limitations:
• Data limitations: They acknowledge that their data sources, while comprehensive, may have missing data or incomplete information. They mention potential issues like unreported trade flows or lack of detail in commodity descriptions. Importantly, they note they addressed some of these issues by verifying import-export symmetry and standardizing procedures. Recognizing these data issues and explaining how they mitigated them gives credibility to the analysis.
• Scope of analysis (food for human consumption only): The authors highlight that their study focuses on food for human consumption, using supply-to-food ratios from FAO. They admit imported food might be used differently than domestic production (e.g., feed vs. food), but due to lack of data, they assumed similar distribution. This is a reasonable assumption and they justify it well. By stating this limitation, they clarify that the findings pertain to edible food trade and not to other agricultural trade like feed or biofuel, which is helpful for context.
• Notional footprint concept: The discussion points out that the footprint indicators are conceptual approximations of environmental pressure. The analysis looks at environmental flows embedded in trade from a supply perspective. They correctly note that a country’s net importer or exporter status of footprints is somewhat abstracted from on-the-ground impacts, and that they did not delve into drivers of food demand or internal distribution of impacts. This is a crucial limitation: their macro-level approach cannot capture which populations within a country benefit or suffer. They explicitly call out that they didn’t identify economic, cultural, and social drivers of food demand or inequalities within countries – leaving that for future research. This candor delineates the boundaries of their conclusions.
• Ecological nature of regression & confounding: They state that their regression analysis is essentially an ecological correlation (country-level association) and must be interpreted with caution regarding causation. They mention that although they included major determinants of GDP (from trustworthy sources), some confounders (education, infrastructure) were absent due to data limitations. This is a very important admission, as it reminds readers that correlation doesn’t prove causation and that omitted variable bias is possible. They also reassure that they attempted robust strategies to control for potential confounders and bias. Indeed, as we saw, they included many controls and fixed effects, which is as much as one can do with available data, but they appropriately remain cautious.
• Time frame of GDP data: They explain why their panel is effectively from 1990 onwards (economic data constraint) and note the varying panel lengths for lagged models. They then point out that despite using shorter panels for lagged GDP, the short- and medium-term associations they found were consistent. This addresses any concern that not using data prior to 1990 might have skewed results.
• Addressing Limitations: The authors not only list these limitations but also often discuss how they addressed or mitigated them and how future work could overcome them. For example, they suggest that future research should explore the underlying drivers of food demand and distributional issues at the individual level, perhaps using agent-based models or system dynamics. This forward-looking statement indicates that the authors see their study as part of a continuing inquiry. It’s very positive in a review process to see authors proactively considering the next steps and alternative methods (like agent-based modeling) for capturing complexity that their current approach cannot.
• Practical and Policy Implications: The discussion (and the conclusions) articulate what the findings mean for policy and practice. The authors emphasize that their investigation highlights connections between environmental footprints in trade and countries’ economic welfare. They suggest that the findings could help set priorities in transforming food systems towards sustainability at a global level. They also mention that evidence from the study can inform policy strategies for promoting economic welfare in a sustainable way. In essence, they argue that acknowledging these trade-offs and connections is important for policy-makers to craft balanced strategies (like ensuring continued access to healthy diets while maintaining sustainability standards). This is a prudent and relevant interpretation: it avoids prescriptive specifics (since that would be beyond the scope) but still draws a meaningful message for decision-makers.
• Alignment with Data: The discussion stays well-aligned with the study’s data and findings. The authors do not claim anything that their data can’t support. For instance, when they talk about consumer preferences or long-term diet patterns, they cite external sources [59,60] – they are not deriving these from their own results but rather situating their results among those considerations. All statements about their results (like those five limitations or the contributions) are directly tied to what was found or the methodology used. This scientific honesty is appreciated.
• Style and Coherence: The discussion is written in an academic yet accessible style. Each major point (limitations, implications, future research) is handled in its own segment, which makes the text easier to follow. The use of ordinal terms (“First, … Second, … Finally, …”) to enumerate limitations is effective. The flow from discussing findings relative to literature into limitations and then into broader implications is logical and coherent. One minor stylistic point: the paragraph starting with “Third, the footprint indicators…” is somewhat dense, combining conceptual and practical limitations and ending with suggestions for future research. Consider breaking that into two paragraphs or using clearer transition language to separate the idea of notional footprints and the lack of identifying demand drivers – they are related but distinct points. However, this is a relatively small issue, and overall the discussion is well-structured.
• Conclusion on Discussion: The discussion is comprehensive and insightful. It not only interprets the results in depth but also reflects on the study’s credibility and situates it in the broader scientific discourse. The authors have shown critical thinking by identifying limitations and not overstating their findings. They have also demonstrated the relevance of their work by connecting to policy and suggesting future studies. This section meets the standards expected for a high-quality journal. Only minor adjustments (as mentioned: slight rephrasing or splitting of paragraphs for clarity) could be made, but no major revisions are needed here.

Conclusions

• Effective Summary: The conclusion section succinctly reiterates the main findings: it confirms the increasing trend in carbon and water footprints in global food trade from 1986 to 2020 and acknowledges the link between network position and economic welfare. These statements directly mirror the study’s objectives and findings, ensuring consistency. They serve as a clear take-home summary for readers, which is exactly the purpose of a conclusion.
• Recommendations and Broader Implications: The conclusions also include forward-looking statements. The authors advise that national governments should consider the possible consequences of food production, trade, and consumption policies on both the environment and the economy to find optimal choices that balance these priorities. Furthermore, they emphasize that global efforts to transform food systems should prioritize sustainable development standards to guarantee continued access to healthy and sustainable diets worldwide. These recommendations are well-grounded in their findings: since the study revealed that trade can confer economic benefits but also entails environmental costs, it follows that policy should aim to optimize both outcomes.
• Alignment with Data: Importantly, the conclusions do not introduce new data or claims not supported by the study. The recommendation for sustainable food system transformation is general enough to be supported by the notion that unsustainable trends were observed and need addressing. The call to consider environmental-economic trade-offs in policy is a direct inference from their results. Thus, the conclusions are justified and appropriately scoped.
• Tone and Clarity: The tone of the conclusions is appropriately assertive about what was found and cautious about what should be done. Phrases like “should consider” and “should prioritize” are suggestions rather than over-confident directives, which is fitting. The writing is clear and free of jargon. One possible enhancement could be to add a specific quantitative highlight from the results to emphasize the significance of the trends (e.g., mention how many-fold the network connectivity or the total footprint increased), but given the word count constraints typical for conclusions, the present qualitative summary is acceptable.
• No Overreach: It’s worth noting that the conclusion stays within the boundaries of the study’s findings. The authors do not claim causality or specific policy measures, only that considerations and priorities should shift given the evidence. This is prudent and maintains scientific integrity.
• Final Elements: After the formal conclusions, the manuscript includes author contributions, funding, ethics statements, and so on, which are all appropriately addressed. The contributions statement is detailed; funding sources are acknowledged; and ethical statements note not applicable, which makes sense for this type of study. The data availability and conflicts of interest are properly stated. These indicate compliance with journal standards and transparency.
• Conclusion on Conclusions: The conclusion section effectively closes the paper by summarizing insights and pointing to their importance in a bigger context. It aligns perfectly with the rest of the paper. No substantive changes are needed here. Perhaps just ensure that it remains consistent if any tweaks are made earlier (e.g., if in revision some numeric specifics are added to results, you might optionally echo one in conclusions). Overall, it’s a strong finish to the manuscript.

References

• Recency and Relevance: The reference list has around 60+ entries and notably includes a large fraction of recent works (from the last 5 years). This indicates the authors have incorporated contemporary research.
• Quality of Sources: Almost all references are from peer-reviewed journals or authoritative sources (FAO, World Bank).
• Reference Formatting: The references appear to follow the required style (likely MDPI style, numerical order). They include titles, journal names, volumes, pages, DOIs, which is good. A minor formatting check: ensure consistency in capitalization (it looks consistent) and that et al. usage is correct (some references list many authors fully which is fine in MDPI style if under a certain number).
• Potential Additions: I don’t see any obvious missing references. Perhaps if there are any brand new 2024 or 2025 studies specifically on environmental impacts of food trade networks, you could include them, but it appears you have covered up to early 2024.
• Conclusion on References: The reference list is current, pertinent, and properly curated. It provides strong support for the manuscript’s content. No significant changes are needed, aside from routine final checks for formatting consistency and perhaps updating any “forthcoming” references if they’ve been published by the time of final revision.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Overall, this is an extremely good and valuable paper which deserves to be published. There are no significant comments, but this reviewer recommends some small, mainly cosmetic, changes and some clearer explanations of important concepts and technical terms in order to facilitate better access for the interested lay-reader. At least comments 6, 7 and 12 should be addressed. This would improve the article’s impact.

1) Line 52: It would be useful to outline briefly what the “sustainable development standards” referred to are? E.g. related to the SDGs and/or FAO standards, etc.?

2) Line 129, given the reasoning here and that WITS data was used for fisheries, why wasn’t it used for all aspects of the first database? This is simply a question of clarification assuming that using data from the same source is likely to be more consistent/comparable than when mixing sources. There is probably a perfectly rational explanation/answer, so perhaps this could be stated.

3) Following on from this, it is not stated anywhere what the rational for concluding the study in 2020 is when the data sources for the second dataset stretch to 2024 and 2022. The first dataset, FAOSTAT, does only run to 2020, but no date information is given in the text for the WITS dataset which might potentially replace FAOSTAT? Is the reason to conclude the study in 2020 because global food chains changed dramatically from 2020 (see final comment), or again is there a simpler explanation?

4) Line 165: states the focus is on food traded for human consumption. This is an important detail, given that a great deal of traded agricultural production is also for animal consumption or for various industrial processes, which also has carbon and water footprints. The reason is almost certainly legitimate, but a brief explanation here would be useful, though this is discussed later.

5) Table 1: GDP is mentioned as measured by PPP, but there is no reference to “income” in Table 1, despite “income” being mentioned in lines 193 and 195. I presume in these two lines, income is measured in Table 1 as GDP per cap in terms of PPP. It would be useful to clarify and explain why these three labels are used seemingly interchangeably. Such use is 100% legitimate, but is likely to be confusing for readers unfamiliar with these terms.

6) Line 265: 255 countries? According to the UN there are 195 countries in 2025?

7) Table 2 and accompanying text: It would be useful to provide brief but clear explanations for the metrics in lay-person terms, plus terms like “latency”. This should be easy to do. I have done a lot of network analysis, but now some time ago, so I had to look up some of these terms to quickly understand the text.

8) Figure 1 caption: it would be useful to make it clear that the nodes are countries and the lines (edges) are trade flows. (This is mentioned earlier, but the reader should not have to refer back). Also, the country acronyms should be noted as Alpha-3 codes from, e.g. https://www.iban.com/country-codes

9) Figure 2 caption: is the same as the Figure 1 caption. It would be useful to add to the Figure 2 caption that this is a simplification, or generalisation, of Figure 1 in order to highlight the main issues – or words to that effect. Is Figure 2 a qualitative simplification, or is it generated directly from data averages – maybe parts of Table 3?

10) Just above the Table 3 caption: it might be useful to insert a brief note that the lower-middle-income percentages exhibited only small increases, except for carbon footprint exports where the percentage fell marginally, while both low-income and undefined fell a lot.

11) Section 3.2: As in comment 5, some technical terms the lay-reader is unlikely to understand might usefully be briefly but clearly explained, including “indegree” and “outdegree centrality”.

12) Table 5 needs some attention:

• The “GDP” label at the top of the first column presumably means all regressions are related to this, so this should be reflected in the caption. Also, the single asterisk at the end of the label should be changed (maybe to an #) so as not to confuse with the significance designations.
• Does the top row (t₀, etc.) mean time, i.e. here the three groups should presumably be 1986, 2000 and 2020?
• In the explanation below the table, the asterisk combinations should be explained, presumably * is p<0.10; ** is p<0.05; ***p<0.01 ?

13) Lines 345 and 357: it would be useful to clarify your use of the term “communities”. In line 357 you contrast this with groups of countries with economic similarities, so does your use of the term imply without economic similarities? This would make sense to me given the distinction between your study on human consumption (i.e. where product quality needs to be high) whilst Torreggiani incorporates feed for animal consumption where product quality is normally prioritised much less. This clearly reflects different trade patterns in these two types of commodity. However, the “geographic proximity” of your countries in communities based much more on trade in animal feed does not make sense to me, e.g. Brazil is a major exporter of animal feed ingredients to Europe, particularly soybean meal? Note: this comment is just an attempt at dialogue and does not need a response :-).

14) Section 4, discussion, mainly in relation to future food system developments and future research, it might be worth considering how the following are having impacts that your article does not address:

• Global food chains changed dramatically in 2020 because of Covid (maybe the reason for concluding your study in 2020?), although some of these changes have now been reversed, but only partially. There are many recent studies on these changes,
• There are many new innovative and fast growing changes in food production techniques, such as precision, vertical, cell-based, etc., mainly in high income countries but this is disseminating – again many recent studies. These developments have the potential both to reduce environmental footprints (e.g. by using fewer inputs and less space) and allowing greater self-sufficiency through ‘friend-shoring’ but also at national and even local level – potentially reducing trade? This could increase food sovereignty, resilience, self-sufficiency, etc., as briefly mentioned in lines 84-85.
• Again, initially in high income countries but spreading, consumption patterns are changing to greater awareness of food, prioritising both a greater focus on healthier and more nutritious food (e.g. more organic and plant-based) and less environmental impact through more local food production and less food waste.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Nowadays, production in the agri-food sector is subject to the principle of sustainable development. The currently applicable regulations oblige manufacturers to obtain raw materials of plant and animal origin, process them, and distribute finished food products not only with the overriding goal of ensuring the appropriate quality and health safety of food. Manufacturers are obliged, first of all, to systematically identify and monitor key environmental aspects, including the carbon footprint during the entire process of production, distribution, and sale of food. After reading the content of the article, the following changes should be made:

1. The abstract should take into account the purpose of the research conducted.
2. The introduction should be supplemented with an approach and research perspective consistent with the subject of the article, i.e. concerning carbon footprint emissions and water use in the food trade process. In addition, the introduction should contain research questions.
3. The article should be supplemented with a literature review section; the content of the research conducted so far in this area should be transferred from the introduction to it, which should be supplemented with a reliable analysis of the literature on the subject. On this basis, the authors should then formulate research hypotheses, which in the results chapter should be verified using appropriate research tools.
4. Chapter 2. Materials and Methods should be supplemented with a description of the criteria for searching the FAOSTAT and WITS databases. In addition, information on the data collection procedure (including terms) should be included, along with a justification for their selection with the subject and purpose of the article. The methodology for calculating the analyzed indicators can be moved to the appendices or removed from the text. The chapter should theoretically describe the regression analysis method, indicating the type of its use for data analysis (whether it was logistic, nonlinear, linear, polynomial regression, etc.). The theoretical description of network analysis should be supplemented in the same way; it is currently unknown which type will be used for data exploration. Will it be analysis by network type (including dynamic, multilayer networks, etc.), or perhaps the scope of analysis (local, global, modules), or the purpose (clustering, path analysis, etc.)?
5. Chapter 3. Results should be supplemented with a reliable indication of the results achieved. The analyses did not mention the countries that make up a given node; the statement that countries are moderately, poorly, or highly developed or with high incomes is not a precise analysis (the authors did not make such a classification). In subchapters 3.1. Network analyses and 3.2. For statistical analyses, the titles should be clarified to indicate what these analyses concern. In addition, the results achieved should be reliably described; much more can be read from the visible graphs and tables than is included in the text of the article. In general, the chapter was prepared very cursorily.
6. In Chapter 4. Discussion, the authors refer, among others, to the "strategic positioning of countries", which was not indicated and distinguished in the content preceding the discussion. In addition, this chapter should confirm/reject the research hypotheses that must be determined on the basis of the literature review.
7. Chapter 5. Conclusions should be supplemented with further research plans of the authors related to the subject of the impact of sustainable development principles resulting from the reduction of the carbon footprint and the limitation of water use on the food trade. This chapter should contain information on the limitations of the research, i.e. what the authors wrote in the discussion should be moved to section 5. In addition, this chapter should contain answers to the research questions specified in the introduction. Additionally, the practical and theoretical significance of the obtained results should be indicated.
8. The article contains citations of the authors' works; if the journal's policy prohibits such practices, they should be eliminated.
9. The language of the article requires correction by a native speaker, mainly prepositions, definite articles, sentence structures, punctuation marks, etc.

Comments on the Quality of English Language

The language of the article requires correction by a native speaker, mainly prepositions, definite articles, sentence structures, punctuation marks, etc.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 5 Report

Comments and Suggestions for Authors

This paper used network analysis to study the interactions and trend overtime for carbon and water footprint worldwide on food trade. The work filled the gap of lacking multidimensional study focusing on the food trade. However, the study needs some rework to improve the scientific soundness and overall merit of the work. I have the following comments:

 

Major comments:

More explanation is needed for Figure 2. I assume between each two nodes, one line represent import from i to j and the other line represent the import from j to i. Is there a way to distinguish the two directions in Figure 2? 

In line 240-241, the author stated that the lagged dependent variables were generated using data 0, 1, and 3 years after capture of the countries' positions in network metrics and other country characteristics. Please explain why z=0, 1 and 3 were used or add a reference for this selection.

For the parameters listed in Table 2, please explain how each of the parameters were calculated (eqn or algorithm used). Please add these details in section 2.4.  

I am having trouble following how the author reached out to the conclusions/statements in the discussion session. This session needs to be reworked: for the conclusions made, please refer to the table or specific parameter that supports the statement. 

Minor comments:

For eqn.1, the descriptions of F and E are missing a subscript x. 

Please add the explanation for N in Table 1. I assume that's the number of data points?

In Table 5, the asterisk symbol was double used to explain the model and to classify significance value. Please change one of them to another symbol. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 5 Report

Comments and Suggestions for Authors

All my comments have been addressed and I am happy with the current version.

Author Response

We would like to thank the reviewer for the time to read our manuscript and provide valuable suggestions to the study.