Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study aims to address an examination of the existing literature on the emerging trends in smart farming research. How these most relevant studies contribute to the advancement of the Sustainable Development Goals. And who are the key actors driving this knowledge production?The article is interesting and promising. This is a review article.

The introduction does not clearly state how the paper will address the interactions of the notions/concepts about the contribution to the Sustainable Development Goals (SDGs), Digital Transformation in Agriculture, The Internet of Things (IoT) and Cyber-Physical Integration in Smart Farming. Here several linkages are missing. More background and policy frameworks should be added for these concepts and understand the link. It's better insert at the end of this section the structure of the article.

-Methods: why the authors applied different classical bibliometric laws (p. 4 line 189)? What's mean classical? Please, justify. The authors could compare other methods.   p. 4-5 lines 189-195

-The conclusions are quite generic; they are not adequately discussed. This lack of an added value. I would encourage the author/s to reinforce their contributions by stating more explicitly why their findings make a significant contribution to literature. Likewise, they could include it as a potential avenue for further research in the conclusions.
Quality of Communication: Has attention been paid to the clarity of expression and readability, such as sentence structure, jargon use, acronyms, etc. Although this is a well-written document, please check the punctuation

Author Response

Dear Reviewer 1, we have highlighted all the changes requested by the review team in green. Below are our responses to each of your comments.

Comment 1: The study aims to address an examination of the existing literature on the emerging trends in smart farming research. How these most relevant studies contribute to the advancement of the Sustainable Development Goals. And who are the key actors driving this knowledge production? The article is interesting and promising. This is a review article.
Response 1: We appreciate your overall evaluation, and at the beginning of the methods section, we have clarified the differences between review methods and bibliometric methods.

Comment 2: The introduction does not clearly state how the paper will address the interactions of the notions/concepts about the contribution to the Sustainable Development Goals (SDGs), Digital Transformation in Agriculture, The Internet of Things (IoT) and Cyber-Physical Integration in Smart Farming. Here several linkages are missing. More background and policy frameworks should be added for these concepts and understand the link. It's better insert at the end of this section the structure of the article.
Response 2: We have completely rewritten the introduction, making it more coherent, fluid, and connected to the questions and objectives of this study.

Comment 3: -Methods: why the authors applied different classical bibliometric laws (p. 4 line 189)? What's mean classical? Please, justify. The authors could compare other methods.   p. 4-5 lines 189-195
Response 3: We have corrected the term “classic” to “eponymous,” which is currently used in bibliometric methodology literature. In addition, we have made several clarifications in the methods section (highlighted changes).

Comment 4: -The conclusions are quite generic; they are not adequately discussed. This lack of an added value. I would encourage the author/s to reinforce their contributions by stating more explicitly why their findings make a significant contribution to literature. Likewise, they could include it as a potential avenue for further research in the conclusions. 
Response 4: We have made significant changes to both the conclusion and discussion sections, rewriting them with greater depth and clarity. We have also added implications, limitations, and future research directions.

Comment 5: Quality of Communication: Has attention been paid to the clarity of expression and readability, such as sentence structure, jargon use, acronyms, etc. Although this is a well-written document, please check the punctuation.
Response 5: We appreciate your positive feedback and have reviewed the details again to improve the formal aspects.

Reviewer 2 Report

Comments and Suggestions for Authors

This is very interesting review. I have enjoyed reading it.

I have some questions and comments:

Line 41: as far as i know, the SDGs linked to agriculture are  - SDG 1, SDG 2, SDG 5, SDG 6, SDG 10, SDG 12, SDG 14, SDG 15. This need references to UN or FAO documents.

Line 120 -123: define Gap more properly? are there any existing reviews on SDG and Smart Farming? If yes, what makes your studies unique? After the description of the aim, you should definitely highlight the novelty of your manuscript.

Did you use PRISMA methodology? Why not?

In discussion/ conclusions – are there any limitations of your study?

Provide suggestions for further research.

Author Response

Dear Reviewer 2, we have highlighted all the changes requested by the review team in green. Below are our responses to each of your comments.

Comment 1: - This is very interesting review. I have enjoyed reading it.
Response 1: We are very grateful for your comment and are glad to hear that you enjoyed our article.

I have some questions and comments:

Comment 2: - Line 41: as far as i know, the SDGs linked to agriculture are  - SDG 1, SDG 2, SDG 5, SDG 6, SDG 10, SDG 12, SDG 14, SDG 15. This need references to UN or FAO documents.
Response 2: We have significantly modified the introduction at the request of the team of four reviewers. However, in subsection 2.3.3, we have indicated that the SDGs reported correspond to the classification made by Clarivate's Web of Science. The classification is reported in generic terms between lines 308 and 314 (including Figure 4), and in detail in section 3.5, including Tables 6 and 7. In addition to the respective comments in the discussion and conclusion.

Comment 3: - Line 120 -123: define Gap more properly? are there any existing reviews on SDG and Smart Farming? If yes, what makes your studies unique? After the description of the aim, you should definitely highlight the novelty of your manuscript.
Response 3: The improvement of the introduction allowed us to better present the gap, given the lack of precision of the concept of smart farming and the evolution, structure, and impact of scientific production related to this topic. 

Comment 4: - Did you use PRISMA methodology? Why not?
Response 4: In the methods section (particularly lines 114-119), we have highlighted the difference between bibliometric methods and review methods. Where exclusion and screening protocols in reviews are replaced by bibliometric laws and analyses of subsamples free of insignificant data.

Comment 5: - In discussion/ conclusions – are there any limitations of your study?
Response 5: We have made significant changes to both the conclusion and discussion sections, rewriting them with greater depth and clarity. We have also added implications, limitations, and future research directions.

Comment 6: - Provide suggestions for further research.
Response 6: We have made significant changes to both the conclusion and discussion sections, rewriting them with greater depth and clarity. We have also added implications, limitations, and future research directions.

Reviewer 3 Report

Comments and Suggestions for Authors

1. Search Strategy and Sample Bias (Major Concern)
In Section 2.2 (Data Collection), the authors state that the search vector used was {TS=(Smart NEAR/O Farming)}. This is a critical limitation. In this academic domain, terms such as "Precision Agriculture," "Smart Agriculture," "Digital Agriculture," and "Agriculture 4.0" are often used interchangeably or as complementary concepts to "Smart Farming."

• Critique: By restricting the search exclusively to "Smart Farming," the study likely excludes high-impact papers that use "Precision Agriculture" in the title/abstract but do not explicitly use the phrase "Smart Farming." This creates a significant sample bias that affects the validity of the core journals list (Table 4) and the author clusters.

• Action Required: The authors must either:

  1. Expand the search query to include these related terms (e.g., using OR) and re-run the analysis to ensure a representative sample; OR

  2. Provide a strong, theoretically grounded justification in the Introduction or Methods section for why "Smart Farming" is distinct enough from "Precision/Digital Agriculture" to warrant an isolated analysis.

2. Methodology of SDG Classification
The paper highlights the contribution of Smart Farming to specific SDGs (Section 3.5 and Table 5) as a key finding. However, the methodology for this mapping is opaque.

• Critique: How were the 1,580 articles classified into specific SDGs? Was this done via manual coding, specific keyword strings (e.g., searching for "Zero Hunger"), or using the Web of Science's automated SDG filters?

• Action Required: Please add a detailed description of the classification procedure in the Methods section. If keyword matching was used, the specific keywords for each SDG should be provided as Supplementary Material.

3. Discussion Depth
The Discussion section (Section 4) currently tends to repeat the statistical findings from the Results section (e.g., restating the h-index of 79 or the growth rates) rather than interpreting them.

• Action Required: The authors should focus more on the "why" and "how." For example, the results show a strong link between Smart Farming and SDG 2 (Zero Hunger) and SDG 3 (Good Health). The discussion should elaborate on the technological mechanisms driving this (e.g., is the link to SDG 3 driven by pesticide reduction via precision spraying? Is SDG 2 driven by yield prediction models?). Moving beyond descriptive statistics to thematic interpretation will strengthen the paper.

4. Journal Concentration (Minor Comment)
Table 4 (Bradford’s Nucleus) shows a very high concentration of journals from a single publisher (MDPI) or Open Access outlets (e.g., Sensors, Agriculture, Sustainability).

• Suggestion: It would be insightful for the authors to briefly discuss in the text whether this trend indicates that the field of Smart Farming is primarily driven by Open Access publishing models compared to traditional agronomy journals.

Author Response

Dear Reviewer 3, we have highlighted all the changes requested by the review team in green. Below are our responses to each of your comments.

Comment 1. Search Strategy and Sample Bias (Major Concern)
In Section 2.2 (Data Collection), the authors state that the search vector used was {TS=(Smart NEAR/O Farming)}. This is a critical limitation. In this academic domain, terms such as "Precision Agriculture," "Smart Agriculture," "Digital Agriculture," and "Agriculture 4.0" are often used interchangeably or as complementary concepts to "Smart Farming."
Critique: By restricting the search exclusively to "Smart Farming," the study likely excludes high-impact papers that use "Precision Agriculture" in the title/abstract but do not explicitly use the phrase "Smart Farming." This creates a significant sample bias that affects the validity of the core journals list (Table 4) and the author clusters.
Action Required: The authors must either:
Expand the search query to include these related terms (e.g., using OR) and re-run the analysis to ensure a representative sample; OR
Provide a strong, theoretically grounded justification in the Introduction or Methods section for why "Smart Farming" is distinct enough from "Precision/Digital Agriculture" to warrant an isolated analysis.
Response 1: 
- Our article is about “Smart Farming.” We have modified the introduction to explain the reasons behind this decision (lines 38 to 106). We have also detailed the objectives along with the research questions.
- In the methods section (particularly lines 114-119), we have highlighted the difference between bibliometric methods and review methods.
- Working with the search vector {TS=(Smart NEAR/O Farming)} yields results related to Smart farming: Precision agriculture (LS:101), Internet of things (LS:81), Machine Learning (LS:56), Deep learning (LS:54), Artificial intelligence (LS:46), and SDigital agriculture (38).
- Incorporating additional terms using an inclusive Boolean search (OR = +) would generate an overrepresentation of the terms suggested by the reviewer, and this selection bias would completely alter the results presented in section 3.4 “Relationships and Scientific Networks” (315-409) and the results of the five resulting thematic clusters. In addition to the collateral effects that this selection bias would generate on all the results presented.

Comment 2. Methodology of SDG Classification
The paper highlights the contribution of Smart Farming to specific SDGs (Section 3.5 and Table 5) as a key finding. However, the methodology for this mapping is opaque.
Critique: How were the 1,580 articles classified into specific SDGs? Was this done via manual coding, specific keyword strings (e.g., searching for "Zero Hunger"), or using the Web of Science's automated SDG filters?
Action Required: Please add a detailed description of the classification procedure in the Methods section. If keyword matching was used, the specific keywords for each SDG should be provided as Supplementary Material.
Response 2: In addition to the complete and detailed list of articles that Clarivate's Web of Science has classified as contributing to the 17 SDGs, this is specified in subsection 2.3.3.The classification is reported in generic terms between lines 308 and 314 (including Figure 4), and in detail in section 3.5, including Tables 6 and 7. In addition to the respective comments in the discussion and conclusion.

Comment 3. Discussion Depth
The Discussion section (Section 4) currently tends to repeat the statistical findings from the Results section (e.g., restating the h-index of 79 or the growth rates) rather than interpreting them.
Action Required: The authors should focus more on the "why" and "how." For example, the results show a strong link between Smart Farming and SDG 2 (Zero Hunger) and SDG 3 (Good Health). The discussion should elaborate on the technological mechanisms driving this (e.g., is the link to SDG 3 driven by pesticide reduction via precision spraying? Is SDG 2 driven by yield prediction models?). Moving beyond descriptive statistics to thematic interpretation will strengthen the paper.
Response 3: 
- We have made significant changes to both the conclusion and discussion sections, rewriting them with greater depth and clarity. 
- We have also added implications, limitations, and future research directions. 
- We have expanded and deepened the discussion, providing additional details from the references in contrast (458-563).

Comment 4. Journal Concentration (Minor Comment)
Table 4 (Bradford’s Nucleus) shows a very high concentration of journals from a single publisher (MDPI) or Open Access outlets (e.g., Sensors, Agriculture, Sustainability).
Suggestion: It would be insightful for the authors to briefly discuss in the text whether this trend indicates that the field of Smart Farming is primarily driven by Open Access publishing models compared to traditional agronomy journals.
Response 4: With regard to open access journals, we have modified Table 4 and paragraphs 259-265.

Reviewer 4 Report

Comments and Suggestions for Authors

Abstract

The abstract is well written, with the research objective clearly stated and the main results effectively highlighted.

Introduction

In subsection 1.1, the contribution of smart farming to the individual Sustainable Development Goals (SDGs) is presented only in a general manner. To improve clarity and strengthen the conceptual grounding of the review, it is necessary to provide a more detailed explanation of how smart farming supports each of the mentioned SDGs.

In lines 76–78, the authors state that in Brazil agricultural digitalization has increased productivity in more than 84% of producers adopting digital technologies. While this is a valuable example, it would be beneficial to include additional empirical evidence from other countries or production systems to strengthen the argument and illustrate the broader global impact of digital transformation in agriculture.

In lines 114–118, the authors mention that machine learning is used for crop-yield prediction, weed detection, and animal-health monitoring; however, these statements remain generic and insufficiently supported. It would strengthen the manuscript if the authors provided concrete examples from empirical studies, including information on the type of datasets used, the accuracy or performance of the applied models, and the specific crops or production systems examined.

In lines 120–123, the authors list several research questions regarding emerging trends, contributions to the SDGs, and key actors in smart farming research. While these questions provide a useful orientation, it would be beneficial for the authors to explicitly state the overall objective of the study in a clear and concise sentence.

Methods

The methodology section lists inclusion criteria but does not provide any exclusion criteria. The authors should explicitly state how duplicate records were handled, whether non-article document types were excluded, and whether any additional screening procedures were applied to ensure the relevance of the final corpus.

Results

Section 3.1 provides a clear descriptive overview of publication growth and appropriately applies Price’s Law, demonstrating a strong exponential trend in Smart Farming research. However, the interpretation remains overly general. The authors should more explicitly explain the factors driving the sharp increase in publications after 2014, clarify the methodological procedure used to determine the contemporary semi-period, and address potential sources of bias such as database expansion and the proliferation of open-access journals.

Section 3.2 presents the application of Bradford’s and Lotka’s laws to analyze journal and author productivity in Smart Farming research, but several important elements require clearer explanation. The interpretation of Bradford’s zones is limited, and the authors should elaborate on what the reported “low concentration” implies and why nine journals constitute the core. Likewise, the Bradford multipliers and percentage error are provided without interpreting whether they indicate a good fit to the theoretical model. Regarding Lotka’s Law, although the authors report an exponential distribution (R² > 99%), they do not present key model parameters.

The description of Figure 3 is unclear, particularly regarding the meaning of the green and orange curves and how they support the argument about field maturation.

Section 3.4 provides a useful descriptive overview of the keyword co-occurrence network and the five thematic clusters, but the analysis remains largely narrative. The authors do not specify key VOSviewer parameters. Moreover, the cluster interpretation lacks quantitative support, such as link strengths or centrality measures, making it difficult to assess the actual relationships between clusters.

Discussion

The Discussion section provides a coherent summary of the main findings of the bibliometric analysis and effectively highlights the rapid growth of Smart Farming research and its technological drivers. However, the section remains highly descriptive and largely reiterates the results without offering deeper critical interpretation. The authors should strengthen the Discussion by explicitly contrasting their findings with previous bibliometric studies in the field, identifying similarities, differences, or novel contributions.

Conclusions

The Conclusions section provides a clear summary of the main results and reiterates the relevance of Smart Farming within the context of the SDGs. However, the section largely repeats descriptive findings without offering additional synthesis or reflecting on the broader implications of the study.

 

Author Response

Dear Reviewer 4, we have highlighted all the changes requested by the review team in green. Below are our responses to each of your comments.

Comment 1: Abstract
The abstract is well written, with the research objective clearly stated and the main results effectively highlighted.
Response 1: We are very grateful for your comment.

Comment 2: Introduction
In subsection 1.1, the contribution of smart farming to the individual Sustainable Development Goals (SDGs) is presented only in a general manner. To improve clarity and strengthen the conceptual grounding of the review, it is necessary to provide a more detailed explanation of how smart farming supports each of the mentioned SDGs.
In lines 76–78, the authors state that in Brazil agricultural digitalization has increased productivity in more than 84% of producers adopting digital technologies. While this is a valuable example, it would be beneficial to include additional empirical evidence from other countries or production systems to strengthen the argument and illustrate the broader global impact of digital transformation in agriculture.
In lines 114–118, the authors mention that machine learning is used for crop-yield prediction, weed detection, and animal-health monitoring; however, these statements remain generic and insufficiently supported. It would strengthen the manuscript if the authors provided concrete examples from empirical studies, including information on the type of datasets used, the accuracy or performance of the applied models, and the specific crops or production systems examined.
In lines 120–123, the authors list several research questions regarding emerging trends, contributions to the SDGs, and key actors in smart farming research. While these questions provide a useful orientation, it would be beneficial for the authors to explicitly state the overall objective of the study in a clear and concise sentence.
Response 2: We have completely rewritten the introduction, making it more coherent, fluid, and connected to the questions and objectives of this study.The improvement of the introduction allowed us to better present the gap, given the lack of precision of the concept of smart farming and the evolution, structure, and impact of scientific production related to this topic.

Comment 3: Methods
The methodology section lists inclusion criteria but does not provide any exclusion criteria. The authors should explicitly state how duplicate records were handled, whether non-article document types were excluded, and whether any additional screening procedures were applied to ensure the relevance of the final corpus.
Response 3: In the methods section (particularly lines 114-119), we have highlighted the difference between bibliometric methods and review methods. Where exclusion and screening protocols in reviews are replaced by bibliometric laws and analyses of subsamples free of non-significant data.

Comment 4: Results
Section 3.1 provides a clear descriptive overview of publication growth and appropriately applies Price’s Law, demonstrating a strong exponential trend in Smart Farming research. However, the interpretation remains overly general. The authors should more explicitly explain the factors driving the sharp increase in publications after 2014, clarify the methodological procedure used to determine the contemporary semi-period, and address potential sources of bias such as database expansion and the proliferation of open-access journals.
Response 4: We have made further clarifications to the presentation of the results of Price's Law (Paragraphs 237-249). 
The contemporary semi-period of Price's second law is simply a temporary median. In addition to what has already been pointed out in the methods (138-142), we have clarified this in the footnote to Figure 1 (234).
Exponential growth is natural in a topic that arouses interest in the global scientific community. With regard to open access journals, we have modified Table 4 and paragraphs 259-265.

Comment 5: Section 3.2 presents the application of Bradford’s and Lotka’s laws to analyze journal and author productivity in Smart Farming research, but several important elements require clearer explanation. The interpretation of Bradford’s zones is limited, and the authors should elaborate on what the reported “low concentration” implies and why nine journals constitute the core. Likewise, the Bradford multipliers and percentage error are provided without interpreting whether they indicate a good fit to the theoretical model. Regarding Lotka’s Law, although the authors report an exponential distribution (R² > 99%), they do not present key model parameters.
Response 5: We have improved the explanation in the application of both laws. For Bradford, we have corrected Table 3 and interpreted the percentage error of -1.03% (259-261). In the case of Lotka, we have detailed the statistical results in paragraphs 279-290 and added a new Table 5.

Comment 6: The description of Figure 3 is unclear, particularly regarding the meaning of the green and orange curves and how they support the argument about field maturation.
Response 6: We have reworded the explanation of Figure 3, between lines 295 and 303.

Comment 7: Section 3.4 provides a useful descriptive overview of the keyword co-occurrence network and the five thematic clusters, but the analysis remains largely narrative. The authors do not specify key VOSviewer parameters. Moreover, the cluster interpretation lacks quantitative support, such as link strengths or centrality measures, making it difficult to assess the actual relationships between clusters.
Response 7: We have explained the five clusters more in detail, indicating occurrence and centrality results (324-395), and an additional sheet with details of occurrences, centrality, and link strengths has been added to the Excel file of supplementary material. We have detailed the parameters with which VOSviewer generated Figure 5 (396-402), and we have regenerated Figure 5 (403). We have indicated the highest link strengths in Figure 5 (405-409).

Comment 8: Discussion
The Discussion section provides a coherent summary of the main findings of the bibliometric analysis and effectively highlights the rapid growth of Smart Farming research and its technological drivers. However, the section remains highly descriptive and largely reiterates the results without offering deeper critical interpretation. The authors should strengthen the Discussion by explicitly contrasting their findings with previous bibliometric studies in the field, identifying similarities, differences, or novel contributions.
Response 8: We have expanded and deepened the discussion, providing additional details from the references in contrast (458-563).

Comment 9: Conclusions
The Conclusions section provides a clear summary of the main results and reiterates the relevance of Smart Farming within the context of the SDGs. However, the section largely repeats descriptive findings without offering additional synthesis or reflecting on the broader implications of the study.
Response 9: We have made significant changes to both the conclusion and discussion sections, rewriting them with greater depth and clarity. We have also added implications, limitations, and future research directions.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript in the new version is improved. The authors cleared all the aspects. Also, the paper has been meticulously revised.