Review Reports - Evaluating the Efficacy of Agricultural Interventions in Northeast Madagascar

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper “Evaluating the efficacy of agricultural interventions in northeast Madagascar” addresses an important and timely topic: evaluating agroecological interventions (market vegetable gardening and poultry husbandry) in rural Madagascar. The study is based on a large sample (over 500 participants) and combines quantitative and qualitative methods, which is a clear strength. The article is well organized, generally clear, and offers practical recommendations for program design.

While the article may not offer strong theoretical innovation, it provides a well-documented case-study evaluation that could be of interest to practitioners, applied researchers, and policymakers working in similar contexts. The paper has merit and could be publishable after addressing the following issues:

1) Clarify the focus on agroecology

Define more explicitly what regenerative agroecology means in the context of this project.In the section DLC-SAVA Regenerative Agroecology Intervention Design, clearly list the techniques implemented so that readers can better understand the results.

2) Compare the adoption of different techniques

Present more systematically which agroecological practices were more successfully adopted and which faced resistance. Provide interpretation of why certain practices were more appealing or feasible for farmers (e.g., cost, labor, cultural preference).

3) Discuss project sustainability

In the paper, authors note that many development projects fail in the medium term due to lack of capacity and follow-up. It would strengthen the article to describe in more detail what measures were taken in this case to ensure the project results sustainability.

4) Acknowledge focus group limitations

Clearly state that the focus group included only eight participants, limiting generalizability. Position these insights as illustrative rather than representative.

5) Improve presentation of quantitative results

In the Quantitative evaluation of the Data Analysis section, summarize key adoption rates in tables rather than long narrative passages. This would make the findings more transparent and accessible.

6) Clarify gender-related findings in poultry husbandry

In the Results (lines 574–575), the authors report no significant gender differences in poultry husbandry adoption, but in the Discussion (lines 622–623), they state that men were significantly more likely to adopt. This appears contradictory and should be clarified. If the gender difference applies only to specific practices (e.g., coop construction) rather than overall adoption, this distinction should be made explicit.

Author Response

1) Clarify the focus on agroecology

Thank you for this suggestion, in the second paragraph around line 48 we revised the paragraph for more detail on regenerative agriculture and agroecology. In the section on our intervention design lines 136 we revised to include more detail on the techniques implemented

2) Compare the adoption of different techniques

We added more details about practice adoption in the discussion starting on line 655

Evidence shows that participation in the workshop increased knowledge and skills, with understanding and adoption especially high for in-situ composting (95% of adopters). Most respondents reported a change in behavior with variable adoption of different techniques. In particular, in-situ composting was one of the most commonly adopted technologies, which is the incorporation of compostable materials into the soil as a direct amendment. When compared to hot composting, in-situ composting was preferred because it was perceived to be faster; rather than waiting three months and turning the compost pile every two weeks, participants appreciated that they could add the amendments directly before planting and immediately plant afterwards. Similarly, case studies in Tanzania revealed that demonstration and hands-on practice of in-situ composting methods involving direct application of crop residuals and manures to the fields produced high yields of quality vegetables and the methods were readily adopted. We observed moderate adoption of rainwater capturing earthworks, like small swales and pocket ponds around the garden beds. While some participants did not perceive a need for such earthworks, others remarked at how effective they were for harvesting rainwater directly in the garden. Women were more likely to adopt earthworks than men, and it likely helped to decrease the workload of watering the garden. Adoption of new agricultural technologies was increased when participants perceived advantages related to climate change adaptation and for soil productivity; in this case, adapting to more seasonal and unpredictable rainfall by capturing rainwater in the garden. Other practices, such as intercropping and the creation of hot compost, were least often adopted. For intercropping, women were significantly more likely to adopt than men, but those farmers that did not adopt expressed they were less likely to use the practice due to habit and preference for growing each crop in its own bed. For hot compost, participants expressed less interest because of the added labor required to turn the pile every two weeks and the long time investments (~three months).

And line 691:

In general, the poultry husbandry intervention achieved many of the desired short-term and medium-term outcomes. Almost 45% of participants trained adopted new techniques, especially building a chicken coop (61% of adopters) and separating the adults from younger chickens (30%). While men and women did not differ in their likelihood of making a chicken coop, women were significantly less likely to build the coop according to the standards shared in the workshops than men, because the labor investment of building a chicken coop is especially challenging for women, as revealed by the qualitative results. Our results show that participation increased knowledge and skills in how to create a chicken coop and biosecurity focused on chicken health, hygiene, and vaccinations. Building the chicken coop was a crucial step in the training that distinguishes the intensified approach from traditional husbandry practices because the chickens are secured at night in a clean space protected from the elements. Separating younger and older chickens is also key because it decreases competition for food and speeds growth. These clear advantages may be the main reason the techniques were adopted. In contrast, most participants reported in semi-structured interviews that they lacked the knowledge or financial ability to produce balanced chicken feed as they were instructed, and about 74% of participants are providing a mixed diet. The production of chicken feed is also challenging because of the quantity required and access to enough raw materials for daily feed (e.g., corn, beans, cassava, peanuts, other protein sources). For populations facing challenges of food security, having enough food for the household is already challenging, and the added burden of feeding a flock of chickens is prohibitive. For these reasons, some techniques were more likely to be adopted than others, which can assist in future workshops to emphasize the importance of those techniques less frequently adopted.

3) Discuss project sustainability

We revised the discussion to give more details starting on line 753

4) Acknowledge focus group limitations

Clearly state that the focus group included only eight participants, limiting generalizability. Position these insights as illustrative rather than representative.

We added statements in the results line 305:

We recognize that this limited sample size is not representative of the views of all participants in the program, but rather illustrates deep insights and nuanced information from a subset of those who were trained.

5) Improve presentation of quantitative results

In the Quantitative evaluation of the Data Analysis section, summarize key adoption rates in tables rather than long narrative passages. This would make the findings more transparent and accessible.

We added Table 2 and revised the results throughout, lines 241-296 to correspond with this suggestion; i.e., not writing the percentages in the text but rather summarizing the results and referring to the table.

6) Clarify gender-related findings in poultry husbandry

Revised, line 693:

While men and women did not differ in their likelihood of making a chicken coop, women were significantly less likely to build the coop according to the standards shared in the workshops than men, because the labor investment of building a chicken coop is especially challenging for women, as revealed by the qualitative results.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript uses the Duke Lemur Center SAVA Conservation project in northeast Madagascar as a case study to evaluate the short- and medium-term impacts of two agricultural interventions (market vegetable gardening and poultry husbandry). The study combines quantitative follow-up surveys and qualitative interviews/focus groups, with a sample covering 27 communities and over 500 farmers. The paper explores gender differences, regional variations, and provides practical improvement suggestions. Overall, the data collection is solid, and the study has high application value. However, there are some shortcomings in terms of mechanistic explanations and writing standards. Specific comments are as follows:

The paper introduces agricultural and food security issues in Madagascar but lacks a summary of existing studies on similar interventions, making the introduction feel more like a project report rather than a scholarly paper. Additionally, the references are outdated, with nearly half of the citations published over 2020years ago. It is recommended to update the references with more recent studies (within the past 5 years) to enhance the timeliness and academic rigor of the background.
In the qualitative analysis section, there are too many direct quotes from the original interviews, which reduces the readability of the manuscript and lacks concise, objective expression. It is suggested to simplify the interview excerpts and only retain the most representative quotes, with the rest moved to the Appendix to reduce redundancy in the main text.
Currently, the manuscript mixes Materials and Methods, Results, Discussion, and Conclusion, which makes the structure seem chaotic and lacks the clarity of a research paper. It is recommended to rewrite the manuscript following the standard academic paper structure, with clearly numbered sections such as Introduction, Methods, Results, Discussion, and Conclusion.
The paper only evaluates short- and medium-term effects, with no consideration of long-term impacts on livelihood improvements, income increases, or nutritional changes. It is recommended to acknowledge this limitation in the discussion and expand on future research directions, especially with regard to long-term outcomes.
The quantitative analysis results need to be supported by statistical tables, variance analysis tables, and other data formats that allow for clearer presentation of the main statistical results. Directly writing the results in the text makes the paper too lengthy and reduces reader interest. It is recommended to present results in clear and concise graphs/tables.
The paper presents quantitative results (e.g., adoption rates, gender differences) and large amounts of qualitative interview data, but there is no clear connection between the two. It is suggested to integrate both data types using a comprehensive model or logical framework. This will allow for a deeper analysis of the underlying mechanisms and establish clearer causal relationships.
The conclusion section should be more data-driven, presenting quantifiable impact indicators such as: “The study found that female adoption rates were 1.68 times higher than male rates, and the market vegetable intervention had significant short-term effects.” Additionally, it should clearly highlight the paper's novelty and compare the findings with existing studies to emphasize its academic gap-filling and practical contributions.

Author Response

The paper introduces agricultural and food security issues in Madagascar but lacks a summary of existing studies on similar interventions, making the introduction feel more like a project report rather than a scholarly paper. Additionally, the references are outdated, with nearly half of the citations published over 2020years ago. It is recommended to update the references with more recent studies (within the past 5 years) to enhance the timeliness and academic rigor of the background.

We made more thorough literature review with more recent literature, specifically lines 48-69

In the qualitative analysis section, there are too many direct quotes from the original interviews, which reduces the readability of the manuscript and lacks concise, objective expression. It is suggested to simplify the interview excerpts and only retain the most representative quotes, with the rest moved to the Appendix to reduce redundancy in the main text.

We removed over 30 quotes throughout the results section, moving them to the Appendix as suggested. We streamlined the qualitative results section to be more concise and readable

Currently, the manuscript mixes Materials and Methods, Results, Discussion, and Conclusion, which makes the structure seem chaotic and lacks the clarity of a research paper. It is recommended to rewrite the manuscript following the standard academic paper structure, with clearly numbered sections such as Introduction, Methods, Results, Discussion, and Conclusion.

We cleaned the discussion and conclusion to reduce mixing results and discussion. We revised the manuscript structured according to Sustainability guidelines numbering 1. Introduction, 2. DLC-SAVA Regenerative Agriculture Intervention Design, 3. Materials and Methods, 4. Results, 5. Discussion, and 6. Conclusion sections clearly delineated

The paper only evaluates short- and medium-term effects, with no consideration of long-term impacts on livelihood improvements, income increases, or nutritional changes. It is recommended to acknowledge this limitation in the discussion and expand on future research directions, especially with regard to long-term outcomes.

We added a paragraph on this point to the discussion line 683. In fact, we have a study designed to measure these longer-term outcomes (though still medium term for most true outcomes).

We examined short- and medium-term outcomes following training in regenerative agroecology principles and technologies. Long-term outcomes, such as changes in yields, income, nutritional status, soil health, and on-farm biodiversity are the next phase of our focus in evaluating the impacts of our interventions. For a subset of participants, we conducted pre-intervention household interviews that measured indicators of socioeconomics, agricultural productivity for subsistence compared to market integration, food security, dietary diversity, and nutritional health of mothers and infants. We also assessed indicators of soil health including soil respiration (a measure of microbial life), compaction, texture, and other metrics. Our design entails repeated post-intervention measurements of these parameters at annual intervals for five years to test for changes over time compared to pre-intervention and compared to control individuals who did not adopt the new technologies. Such research is timely because few studies examine outcomes beyond exit interviews, and those that do follow longitudinal cohorts often demonstrate the improvements in livelihoods and soil health [17]. As of the writing of this article, we are in the fourth year of our cohort study and plan to conduct the final resample in 2026.

The quantitative analysis results need to be supported by statistical tables, variance analysis tables, and other data formats that allow for clearer presentation of the main statistical results. Directly writing the results in the text makes the paper too lengthy and reduces reader interest. It is recommended to present results in clear and concise graphs/tables.

We reformulated the supplementary table which had more of the requested statistics into the main text as suggested, Table 3 and 4. We really do appreciate this suggestion because we recognize it will help the reader and it even helped us to realize some of the more striking results that were somewhat buried before.

The paper presents quantitative results (e.g., adoption rates, gender differences) and large amounts of qualitative interview data, but there is no clear connection between the two. It is suggested to integrate both data types using a comprehensive model or logical framework. This will allow for a deeper analysis of the underlying mechanisms and establish clearer causal relationships.

We included a logical framework to begin the discussion section, 640-649, to clearly connect the quantitative and qualitative data.

The conclusion section should be more data-driven, presenting quantifiable impact indicators such as: “The study found that female adoption rates were 1.68 times higher than male rates, and the market vegetable intervention had significant short-term effects.” Additionally, it should clearly highlight the paper's novelty and compare the findings with existing studies to emphasize its academic gap-filling and practical contributions.

We edited the conclusion on lines 818-833 and 836-838 to follow the reviewer’s suggestions

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I accept the manuscript in present form. Thank you for your effort.