Three Decades of Changes in Brazilian Municipalities and Their Food Production Systems

Round 1

Reviewer 1 Report

Please avoid repeating the same sentences in Abstract and Introduction.

The paper contains many tables and figures yet the main objective of the research remains unclear. What is the key message delivered? The title speaks of trade-offs yet this remains vague across the text.

It is unclear why Fig. 1 shows high trade-off in one direction only. Indeed, it should follow a U-shape.

The paper should avoid using unclear abbreviations (CropLabor etc.) that make the reading of the paper impossible. Also, notations used in, e.g., Eq. 3 are not defined.

The meaning of Table 2 is unclear. Why these patterns should be considered as valid?

Unclear wording "classes of municipalities". This is simply grouping.

The quality of pictures is not satisfactory.

The figures should have large enough labels inside.

The use of the figures and maps should be reduced. Currently, it looks more like a geographical atlas without any serious conclusion.

Figure 4 is completely unclear: why you need this for each year? why table is not enough? The text on Lines 338-342 is gibberish. Table 3 repeats the same data?

Figure 5 makes no sense. A simple table would be enough. also, grouping should be done by variables, not by years.

Ostapenko, R.; Herasymenko, Y.; Nitsenko, V.; Koliadenko, S.; Balezentis, T.; Streimikiene, D. 2020. Analysis of Production and Sales of Organic Products in Ukrainian Agricultural Enterprises. Sustainability, 12, 3416. Sorgho, R.; Mank, I.; Kagoné, M.; Souares, A.; Danquah, I.; Sauerborn, R. “We Will Always Ask Ourselves the Question of How to Feed the Family”: Subsistence Farmers’ Perceptions on Adaptation to Climate Change in Burkina Faso. Int. J. Environ. Res. Public Health 2020, 17, 7200.

Author Response

Reviewer 1

Please avoid repeating the same sentences in Abstract and Introduction. The paper contains many tables and figures yet the main objective of the research remains unclear. What is the key message delivered? The title speaks of trade-offs yet this remains vague across the text.

Responses from the authors: We appreciate the careful review. We have edited the manuscript to avoid repeating sentences in the abstract and the main text. Additionally we have improved the manuscript’s key message. We have changed the manuscript title to better represent the work done.

-

Reviewer 1: It is unclear why Fig. 1 shows high trade-off in one direction only. Indeed, it should follow a U-shape. The paper should avoid using unclear abbreviations (CropLabor etc.) that make the reading of the paper impossible. Also, notations used in, e.g., Eq. 3 are not defined. The meaning of Table 2 is unclear. Why these patterns should be considered as valid? Unclear wording "classes of municipalities". This is simply grouping.

Responses from the authors: The manuscript no longer includes the original figure 1 as to avoid any confusion. We have changed many of the abbreviations to plain text and table 2 has been removed from the manuscript to avoid any confusion. Notations in equation 3 have been defined. The “classes of municipalities” have been changed accordingly in the text.

-

Reviewer 1: The quality of pictures is not satisfactory. The figures should have large enough labels inside. The use of the figures and maps should be reduced. Currently, it looks more like a geographical atlas without any serious conclusion. Figure 4 is completely unclear: why you need this for each year? why table is not enough? The text on Lines 338-342 is gibberish. Table 3 repeats the same data? Figure 5 makes no sense. A simple table would be enough. also, grouping should be done by variables, not by years.

Responses from the authors: We appreciate these comments. Figure 3 presents the statistical results for each variable according to the year of the agricultural censuses and within each municipality group according to soybean farm size. Table 3 in the previous version, now Table 2, presents the summary statistics of the variables. Both are important to present results. The caption for Figure 3 that was previously on lines 338-342 was improved with a clearer title, but the technical text was kept for completeness and transparency. Figure 5 is no longer in the manuscript and it has been substituted by Table 3, grouped by variables (not by years) as suggested.

Reviewer 2 Report

About the submission with the title "Spatiotemporal dynamics of synergies and trade-offs in Brazilian food production systems" I have the following comments:

 

It could be interesting for the readers that the abstract clearly presents the following: main motivations, objectives, methodologies used and main insights. For example, the main approaches considered were superficailly presented in the abstract.

 

It is important to explain if the figure 1 (and the others) are own elaboration or was obtained from another source.

 

The approaches carried out in the subsections 2.1 and 2.2 need a better scientific support.

 

In general, this submission seems to be scientifically weak, because employ older methodologies and forget new approaches as the spatial autocorrelation analysis. For a submission with the title "Spatiotemporal dynamics..." i would be expected something more robust in these fields.

 

Another aspect is about the concepts of trade-off and synergies adopted in the submission and seems to me not well linked with the methodologies used. In other words, how the approaches considered quantify the synergies and the trade-offs. This for me seems hard to understand in this article. The interlinkage seems forced.

Author Response

Reviewer 2

About the submission with the title "Spatiotemporal dynamics of synergies and trade-offs in Brazilian food production systems" I have the following comments:

It could be interesting for the readers that the abstract clearly presents the following: main motivations, objectives, methodologies used and main insights. For example, the main approaches considered were superficailly presented in the abstract. It is important to explain if the figure 1 (and the others) are own elaboration or was obtained from another source.

Responses from the authors: We have improved the abstract and removed Figure 1 from the manuscript (based on another reviewer’s comment). But to note, that figure and all other figures and tables were developed by the authors of the present submission.

 -

Reviewer 2: The approaches carried out in the subsections 2.1 and 2.2 need a better scientific support. In general, this submission seems to be scientifically weak, because employ older methodologies and forget new approaches as the spatial autocorrelation analysis. For a submission with the title "Spatiotemporal dynamics..." i would be expected something more robust in these fields.

 Responses from the authors: Thanks for these comments. Regarding the distribution of soybean farm size there is spatial autocorrelation pattern (confirmed by Moran’s I test—and now included in the manuscript). In the introduction section, we explained the existence of this pattern when describing the history of the formation of food production areas in Central and North Brazil and the formation of agricultural frontiers. As in this study we did not develop a model based on “traditional” regression analysis, but instead developed statistical test analysis to assess the value of each variable according to municipality classes (of farm size). Therefore, focus changes from a spatiotemporal approach to a social-economic outcomes from different farm sizes. The spatial autocorrelation does not jeopardize our results, but in fact reinforces the notion of what is happening in agricultural frontiers in Brazil—large land concentration by few commodity producers. We revealed that expansion frontiers are leading to the formation of large farms, few producers and resulting in food systems with low human labor demand and producing almost exclusive commodities for the international market. Sections 2.1 and 2.2 are intended to show how we developed the variables used in the study. In this regard, these two sections are complete with literature and describe a set of key socio-agricultural variables of interest in food production systems. Section 2.3 contains more information on the tests used to draw conclusions from our data set. We have included new edits to make our approach clearer. While straightforward, our methods are scientifically sound and enable us to present new results that build on the literature in the field by providing empirical findings from official statistical data. However, we acknowledge and agree that the level of sophistication of our analysis likely does not merit the use of ‘Spatiotemporal Dynamics’ in the title and we have changed the title accordingly.

-

Reviewer 2: Another aspect is about the concepts of trade-off and synergies adopted in the submission and seems to me not well linked with the methodologies used. In other words, how the approaches considered quantify the synergies and the trade-offs. This for me seems hard to understand in this article. The interlinkage seems forced.

Responses from the authors: In section 2.3 of the manuscript we have improved the text and made some clarifications about the approach used in the study. We hope that the new edits to our text make our approach and conclusions clearer. In our study, we looked at the different classes of municipalities (based on average soybean farm size) to evaluate the variables of interest. The Kruskal-Wallis tests provide knowledge of when and where results are significantly different according to class group. But also we are able to examine where variables have preferable values at the same time (synergies) or some values preferable vs other that are not (trade-offs). In this regard, we were able to draw conclusions about whether municipalities are in synergistic or trade-off situations.

 

Reviewer 3 Report

I read with great interest the manuscript with the title “Spatiotemporal dynamics of synergies and trade-offs  in Brazilian food production systems”. The paper is deals with the quite interesting issue of food production systems, tracing likely synergies (or trade-offs) in Brazil. The manuscript is well prepared, the structure and the language are fine.  My main concerns are the following:

• Some arguments /claims are abused by the too many references used to back up them.
• I do not understand the (1) especially the lines 190-191. Do you mean that RPit represents the total land?
• Lines 194-195 are confusing. What do you want to say?
• Line 205, why (2) has to be weighted average? What are the weights?
• Lines 205-209, do you want to say that PA represents the gross revenues net of taxes and transportation costs?
• Line 236, what is the meaning of effective number of crops?
• Lines 238-239: Why do you have to transform the Shannon Index?
• Lines 241-279: You do not explicitly say what you are trying to examine and how? First, you do not have to convert a continuous variable to categorical in order to discriminate it (create classes). Second, the choice of parametric vs non parametric methods requires a normality test. You mentioned it (line 255, without providing the evidence). Then, you should opt for Kruskal-Wallis and Spearman as opposed to ANOVA and Pearson.
• It is not a good idea to “infect” your discussion with the p values
• Line 411, BTW, the p-values reflect the reliability of an estimate and it is not the statistical significance level.
• Lines 427-428, Correlation coefficients should be given in a Table along with the critical value rather than the size.

Author Response

Reviewer 3

I read with great interest the manuscript with the title “Spatiotemporal dynamics of synergies and trade-offs in Brazilian food production systems”. The paper is deals with the quite interesting issue of food production systems, tracing likely synergies (or trade-offs) in Brazil. The manuscript is well prepared, the structure and the language are fine.  My main concerns are the following:

 

Responses from the authors: We thank this reviewer for the positive and encouraging comment. In this new version, we have addressed all of your comments.

-

Reviewer 3: Some arguments /claims are abused by the too many references used to back up them.

Responses from the authors: We have re-examined the manuscript and where appropriate selected fewer references to avoid any abusive use of too many previous studies.

-

Reviewer 3: I do not understand the (1) especially the lines 190-191. Do you mean that RPit represents the total land?

Responses from the authors: Thanks for the question. In fact, the RP_i,t represents the number of Rural Properties (i.e., RP) in municipality i in a given year t.

-

Reviewer 3: Lines 194-195 are confusing. What do you want to say?

Responses from the authors: We have rephrased the sentence to make our arguments clearer. The idea of the sentence is to say that the results of farm size and its distribution across regions of Brazil are parallel with what has been discussed in previous literature.

-

Reviewer 3: Line 205, why (2) has to be weighted average? What are the weights? Lines 205-209, do you want to say that PA represents the gross revenues net of taxes and transportation costs?

Responses from the authors: The production value information is provided and developed by the IBGE (Brazilian Institute of Geography and Statistics). According to the IBGE, production value information is prepared by the following explanation: “Production value: A variable that is derived/calculated by the weighted average of the quantity of production and average of current prices paid to the producer, according to the periods of the harvest and commercialization of each product. Shipping charges, fees and taxes are not included.” Therefore, as we described in the manuscript, the production value represents the cash inflow generated for all crops produced in a given municipality, resulting from trade between producers and markets (national and international), but not accounting for taxes, fees and transportation costs.

-

Reviewer 3: Line 236, what is the meaning of effective number of crops? Lines 238-239: Why do you have to transform the Shannon Index?

Responses from the authors: The Shannon indices were transformed to its exponential exp(SDI) to express the effective number of crops (ENC), a more suitable index because the resulting values represent the number of crops dominating the production in a given municipality. Thus, ENC provides real numbers of how many crops a given municipality has in fact cultivated. This is more meaningful for reading and interpretations.

-

Reviewer 3: Lines 241-279: You do not explicitly say what you are trying to examine and how? First, you do not have to convert a continuous variable to categorical in order to discriminate it (create classes). Second, the choice of parametric vs non parametric methods requires a normality test. You mentioned it (line 255, without providing the evidence). Then, you should opt for Kruskal-Wallis and Spearman as opposed to ANOVA and Pearson.

Responses for the authors: In the current version of the manuscript we have improved some sentences and parts to make clear what we are examining. The decision of having classes for our analysis is explained in section 2.3 of the methods and the rationale to support that. Also in section 2.3 we provided the statistical significance of Shapiro-Wilk tests to justify the use of non-parametric tests instead of ANOVA and Pearson correlation.

-

Reviewer 3: It is not a good idea to “infect” your discussion with the p values

Responses from the authors: Thanks for the advice. We carefully reviewed the discussion to avoid the use of p values.

-

Reviewer 3: Line 411, BTW, the p-values reflect the reliability of an estimate and it is not the statistical significance level.

Responses from the authors: Thanks for this comment. We have clarified our references to p-values through the text to avoid any problem of interpretation.

-

Reviewer 3: Lines 427-428, Correlation coefficients should be given in a Table along with the critical value rather than the size.

Responses from the authors: The p-values were lower than 0.01 and this information was provided in the previous sentence, now in lines 409—411. As other reviewers argued to avoid the use of more tables or figures, we believe the necessary information provided in the text is enough for a correct read and interpretation of the results.

Round 2

Reviewer 1 Report

The paper requires serious proofreading. The presentations of the results is not satisfactory. My previous comments were not addressed (I found responses to the three comments only).

Author Response

Thank you for the comments. The manuscript was reviewed by a native English speaker and during the first revision round all the comments were addressed. In the new version of the manuscript we added a new paragraph in the introduction to increase the clarity of our approach.

 

Reviewer 2 Report

I suggest the authors present scientific support for the information presented for the Moran's statistics. For example it is important to cite authors as Anselin. 

Author Response

Thank you for the comments. The new version of the manuscript brings the Anselin (1995) reference to the Moran's I statistics.   Anselin, L. Local indicators of Spatial Association-LISA. Geographical Analysis 1995, 27, 93–115.