Spatial Analysis of Sustainable and Agricultural Development in Southern Brazil

doi:10.21203/rs.3.rs-4939731/v1

Download PDF

Research Article

Spatial Analysis of Sustainable and Agricultural Development in Southern Brazil

https://doi.org/10.21203/rs.3.rs-4939731/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This work sought to create and analyze a Sustainable and Agricultural Development Index (SADI) for the municipalities that make up the Southern Region of the country. For this, the methodology used was the factor analysis through the analysis of principal components and, subsequently, the exploratory analysis of spatial data (ESDA) was performed, verifying the distribution of the indicator. The main results revealed that the values found for the indicators showed a medium-low degree, considering that the highest value found was 0.5885 and the others presented values below 0.5. The municipalities with the highest indicators, with the highest degree of sustainable and agricultural development, were Curitiba (PR) (0.5885); Capivari do Sul (RS) (0.4541); and Paranapoema (PR) (0.4536). While the lowest indicators, with a lower degree of sustainable and agricultural development, were: Bombinhas (SC), (0.1843); Governador Celso Ramos (SC) (0.2170); and Matinhos (PR) (0.2172). Regarding the ESDA, it showed strong evidence of a positive and significant spatial correlation between the municipalities analyzed, in addition, there was a high disparity in the sustainable and agricultural development of the municipalities that make up these states.

Agricultural Economics & Policy

Agronomy

Environmental Economics

Environmental Policy

Agriculture

Sustainability

Factor Analysis

Exploratory Spatial Data Analysis (ESDA)

One of the great challenges today is to align the increase in agricultural production while promoting sustainable development, reducing impacts on the use of natural resources, and assisting in social development. The motivation stems from international debates and social pressure that plead for a new development model that can reconcile economic growth, social development, and environmental preservation (Sambuichi et al., 2012).

For Froehlich (2014), the theme of sustainability is increasingly in evidence, however, it is a concept that is in the construction phase and needs to be studied in greater detail for its better understanding. Despite this, the most widely used concept of sustainable development is the one developed by the Brundtland report of the World Commission for Economic Development – WCED (1987), which states that sustainable development is a process of change where the exploitation of resources, the application of investments, technological development and institutional changes are carried out consciously and according to current and future needs. Sustainable development is defined based on a few dimensions, the three dimensions most indicated by the literature are: economic, social, and environmental (Werbach, 2010; Pawlowski, 2008; Spangerber & Bonniot, 1998; Sachs, 1993; Organisation for Economic Cooperation and Development – OECD, 1993).

The economic dimension refers to the macroeconomic plan, i.e., the more efficient allocation and management of resources and a constant flow of public and private investments of endogenous origin, aiming at expanding access to these resources and economic and social development. The social dimension refers to the search for a standard of living adequate to the present and the future, aligned with an improvement in the quality of life of the population, with greater equity in income distribution, improvement in health, education, job creation, etc. And the environmental dimension is the one that aims at: the rational use of natural resources, the consumption of fossil fuels, the use of renewable and non-renewable resources in general; reducing the volume of waste and pollutants; increased research for the development of new low-waste technologies that use resources that promote urban, rural, and industrial development; and the definition of standards for environmental protection. This dimension requires solutions to production processes that use large amounts of natural resources, reducing degradation and producing with a view to meeting the demand of the world population without degrading the environment (Sachs, 1993).

In this context, it should be noted that Brazil is an essentially agricultural country and, therefore, this sector is extremely important for the generation of wealth and income in the country. However, the development of this sector has not occurred in a homogeneous manner across regions and product specialization. According to Melo (2006), this unbalanced development occurred due to the evolution of the agricultural sector, which was based on the formation of agro-industrial complexes. The formation of these agro-industrial complexes led to the generation of wealth, accompanied by an increase in the concentration of income and land, rural exodus, unemployment, imbalances in fauna and flora, and exclusion of the less privileged, since the capital invested over the years benefited the employer producers more than the small family producers (Pereira et al., 2004).

Thus, agricultural establishments occupy a total area of 351.3 million hectares, corresponding to about 41% of the Brazilian territory, with large discrepancies in the percentage of occupied area between Brazilian regions (Brazilian Institute of Geography and Statistics - IBGE, 2017). Thus, because it mainly uses land and natural resources in its production processes, the impacts on the environment caused by the agricultural sector are considerable, which ends up directly or indirectly affecting the climate, the hydrological cycle, and the quality of the natural resources available in the country (Sambuichi et al. 2012). Thus, a major concern arises: the search for growth and development patterns in the agricultural sector aligned with sustainable development.

In this context, the following questions arise: What is the contribution of agricultural development to economic, social, and environmental sustainability? How to reconcile this? And what is the current situation in Brazil? Thus, the objective of this work is to create and analyze an index of sustainable and agricultural development for the municipalities that make up the Southern Region of the country. The specific objectives were to create an indicator based on the dimensions of sustainability and the current agricultural situation in the region; and, to analyze its spatial distribution, discussing the results found. The methodology used was factor analysis through principal component analysis and later the Exploratory Analysis of Spatial Data (ESDA) was performed, verifying the distribution of the indicator and the formation of clusters.

The justification for the choice of the theme is due to the constant academic and social debate on the subject and the great difficulty in reconciling sustainable and agricultural development. Thus, this study sought to verify the construction of an indicator based on these two areas, which go hand in hand, verifying the relationship of the variables in the construction of the indicator. The choice of the analyzed region is due to the fact that the South Region had the second largest planted area in the country (22,320,494 hectares), behind only the Midwest Region (31,803,034), is the largest producer of wheat in the country (85% of production) and rice (82%), and is the second largest producer of corn (22%) and soybeans (28.26%) (Municipal Agricultural Survey – Brazilian Institute of Geography and Statistics - PAM-IBGE, 2021). In addition, the agricultural sector has great prominence in the region, it is a great generator of employment and income, and due to its geographical location, favorable climate, and available natural resources, the region has great potential for agricultural production. Finally, it is highlighted that studying agriculture in the Southern Region of Brazil, aligned with sustainable development, can help in understanding how agricultural production can be carried out in a sustainable way, preserving the natural resources and biodiversity of this region.

The present study is divided into four sections, the first of which consists of this introduction. Section two presents the methodology used; Section three analyzes and discusses the results and finally, the conclusions of the study are presented in Section four.

2.1 Statistical Foundations of the SADI

To capture the sustainable and agricultural development of the municipalities belonging to the Southern Region, a wide range of variables was used. Therefore, to reduce this broad set, the method of factor analysis was used, which transforms this set into a reduced number of factors. As a result, there is a more direct and parsimonious reduction of data, which facilitates the interpretation and analysis of data. Thus, factor analysis is a multivariate technique that allows the union of variables that follow the same behavior about correlation, thus, from a group of variables chosen to compose the model, the union of information occurs through factors or components (Hair Jr. et al., 2009). Therefore, to explain the factor analysis model, this study uses the principal components method, which extracts uncorrelated factors, maximizing their contribution to the common variance (Fávero & Belfiore, 2017).

A factor analysis model is given by the following equation (Mingoti, 2005):

$$\:{X}_{i}={\alpha\:}_{ij}{F}_{j}+{\epsilon\:}_{i}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(1\right)$$

where, $\:{X}_{i}$ = $\:{({X}_{1},\:{X}_{2},\:{X}_{3}\dots\:,{X}_{p})}^{t}\:$= a transposed vector of observable random variables; $\:{\alpha\:}_{ij}$ = a matrix $\:\left(p\:x\:m\right)\:$of fixed coefficients called factor loadings, which describe the linear relationship of $\:{X}_{i};\:e$ $\:{F}_{j}$ ; $\:{F}_{j}$ = ($\:{F}_{1}{,\:F}_{2}$,…,$\:\:{F}_{p}$) is a transposed vector $\:(m\:<\:p)$ is a transposed vector of latent variables, describing the unobservable elements of the sample used; $\:{\epsilon\:}_{i}$ = ($\:{\epsilon\:}_{1}$, $\:{\epsilon\:}_{2}$,…,$\:\:{\epsilon\:}_{p}$) represents a transposed vector of random errors, corresponding to measurement and variance errors not explained by common factors. Considering that the variables composing the indicator have different scales and principal component analysis aims to maximize variance, it may be sensitive to these different scales, potentially affecting the results. To address this issue, it is necessary to standardize the variables, expressing the data in comparable units (Lattin, 2011). This procedure can be achieved by the following equation:

$$\:Z=\frac{({X}_{i}-\stackrel{-}{X})}{S}\:,\:i=1,\:2,\:3\:...,n\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(2\right)$$

where, Z = the standardized variable; $\:{X}_{i}$ = the variable to be standardized; $\:\stackrel{-}{X}$ = the arithmetic mean of the variable $\:X$; S = to the sample standard deviation of the variable $\:X$. With the standardization of the observable variables $\:{X}_{i}$, these can be replaced by a standardized variable vector $\:{Z}_{i}$, which will solve the problem of difference in scale units.

The next step is to identify the number of factors suitable to compose the model, for this the measure called eigenvalue or commonly known as characteristic roots, which express the total variance explained by each observed factor. The determination of the number of factors required and that will represent the data set is defined by Kaiser's Rule, which recommends the use of those principal components whose characteristic roots have values greater than one (Lattin, 2011).

To facilitate the interpretation of these factors, orthogonal rotation is performed using the varimax method, which builds a simple structure based on the structure of the columns of the matrix of factor loadings, promoting the maximization of the variance of the factor (HAIR et al., 2009).

According to Mingoti (2005), the factorial scores are those values referring to each observation of the sample and situate them in the space of common factors, given by:

$\:{F}_{j\:}$ =$\:\sum\:_{j=1}^{k}{b}_{i}{X}_{ij}$, com i = 1, 2,..., p (3)

where, $\:{F}_{j\:}$= to the factorial scores; $\:{b}_{i}=\:$ the regression coefficients that represent the weighting weights of the variables $\:{X}_{ij}$ in the factor $\:{F}_{j\:}$; $\:{X}_{ij}$ = to the values of the variables for the k-th element that makes up the sample.

Thus, the Sustainable and Agricultural Development Index ($\:{SADI}_{m}$) can be given, according to Mingoti (2005), by the following expression:

$$\:{SADI}_{m}\:{=\:{\sum\:}_{i=1}^{p}\:\left(\frac{{\sigma\:i}^{2}}{{\sum\:}_{i=1}^{p}{\sigma\:}^{2}i}\:{F}_{im}\right)}_{\:}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(4\right)$$

Where $\:{\sigma\:}^{2}\:$is the variance explained by the fator $\:i$; $\:p$ is the number of factors chosen; $\:{\sum\:}_{i=1}^{p}{\sigma\:}^{2}i$ is the sum of the explained variances of $\:p$ and $\:{F}_{ie}$ is the factorial score of the m-th municipality in the Southern Region of Brazil, of the factor $\:i$. The results are standardized in the form:

$$\:{SADI}_{m}=\frac{{SADI}_{m\left(before\right)}-{SADI}_{min}}{{SADI}_{m\text{á}x}-{SADI}_{min}}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(5\right)$$

in which the individual values of $\:{SADI}_{m},\:$as well as the maximum and minimum values are used to adjust the results between 0 and1.$\:\:$

According to Fávero and Belfiore (2017), the results achieved can be evaluated through the Kaiser-Meyer-Olkin (KMO) test and the Bartlett sphericity test, the first is given by the following equation:

$$\:KMO=\frac{\sum\:_{l=1}^{k}\sum\:_{c=1}^{k}{\rho\:}_{lc}^{2}}{\sum\:_{l=1}^{k}\sum\:_{c=1}^{k}{\rho\:}_{lc}^{2}+\sum\:_{l=1}^{k}\sum\:_{c=1}^{k}{\phi\:}_{lc}^{2}}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(6\right)$$

where, KMO values between 0.6 and 0.7 are considered reasonable, however, it should be noted that the closer to 1, the better the overall adequacy of the model. The Bartlett test is given by:

$$\:{\text{{\rm\:X}}}_{Bartlett}^{2}=\left[(n-1)-\left(\frac{2k+5}{6}\right)\right]\text{l}\text{n}\left|\text{D}\right|\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(7\right)$$

where, the degrees of freedom are obtained by $\:\frac{k\:(k-1)}{2}$, being $\:n$ the sample size,$\:\:k$ the number of variables, and $\:D\:$the determinant of the correlation matrix $\:\rho\:$. When the value of the test is greater than the critical value, the identity matrix hypothesis is rejected.

Thus, in the next topic, the second methodology used in this work will be described, which is the exploratory analysis of spatial data (ESDA). After the presentation of the econometric procedure adopted for the development of the SADI proposed in this research, as well as its form of spatial distribution, the description of the variables used is presented, along with the source of the data used to compose the SADI of the municipalities belonging to the Southern Region of Brazil.

2.2 A Synthesis of Exploratory Spatial Data Analysis (ESDA)

The Exploratory Analysis of Spatial Data (ESDA) is given through the interaction of agents in space, that is, the value of a certain variable of interest in each region $\:i$ depends on the value of this variable in neighboring regions $\:j$. The inclusion of location in the study makes the results more consistent, according to Anselin (1999) spatial econometrics is a subfield of econometrics that deals with spatial interaction (spatial autocorrelation), in addition to spatial structure (spatial heterogeneity) in regression models.

Thus, to obtain more consistent results, the ESDA is used, verifying whether there is spatial autocorrelation between the municipalities and their neighbors, and whether they influence the sustainable and agricultural development of the latter. Autocorrelation is obtained by Moran's statistics $\:I$, which indicate the degree of linear association between the vectors of the values observed over time, in addition to the weighted average of the values of the neighbors (Parré, 2014; Almeida, 2012). Moran (apud Almeida (2012)), states that the formula for this statistic can be expressed by the following equation:

$$\:I=\frac{n}{\sum\:\sum\:{\omega\:}_{i}}\frac{\sum\:\sum\:{\omega\:}_{ij}\:\left({y}_{i}-\stackrel{-}{y}\right)({y}_{i}-\stackrel{-}{y})}{{(\sum\:{y}_{i}-\stackrel{-}{y})}^{2}}$$

where, $\:n$ = the number of spatial units, $\:{y}_{i}$ = the variable of interest; $\:{\omega\:}_{ij}$ = the spatial weight for the pair of spatial units $\:i$ and $\:\:j$ measure the level of interaction between them.

Considering this, spatial dependence is one of the main characteristics of spatial data. However, to determine the degree of spatial autocorrelation, one must consider the degree of neighborhood in which one intends to analyze the spatial dependence. By adopting the neighborhood criterion, it is possible to construct the matrix of space weights (Sabater, Tur & Azorín, 2011). These matrices are based on contiguity, and can be determined according to neighborhood, geographic or socioeconomic distance, or even by a combination of both (Almeida, 2012). The most common forms of spatial weight matrices used are the queen convention and the tower as shown in Fig. 1, where the neighbors of regions A and B are the highlighted regions. About the queen contiguity convention, it considers beyond the boundaries with non-zero extension, the vertices (nodes), in the presentation of a map, as contiguous. The Torre contiguity, on the other hand, considers only the physical borders with non-zero extension between the regions (Almeida, 2012).

The values obtained for Moran's $\:I$ can vary between − 1 and 1. Thus, negative values indicate negative spatial autocorrelation, demonstrating that there are dissimilarities between the values of the studied attribute and its spatial location; positive values, on the other hand, indicate positive spatial autocorrelation, demonstrating similarities between the values of the studied attribute and its spatial location (Lins et al., 2015). In view of this, Fig. 2 presents in a generic way the scatter plot of Moran's $\:I$ statistic with the possible types of clusters of spatial linear association. It is noteworthy that this dispersion analysis is valid for both the univariate and the multivariate Moran's $\:I\:$statistics$\:.$

On the horizontal axis, the variable of interest ($\:X$) is placed, and on the vertical axis, the spatial lag of the variable of interest ($\:W\_X$) is presented. Thus, it is possible to verify the concentration pattern of the data, which are divided into four types of associations: Q1 - high-high (HH), Q2 - high-low (LH), Q3 - low-low (LL), Q4 - high-low (HL). If the spatial association is positive, the regression line will be increasing and the municipalities will tend to be grouped in the first and third quadrants, however, if the relationship is negative, the line will be decreasing, and the units will be grouped predominantly in the second and fourth quadrants (ALMEIDA, 2012).

2.3 Variables and Data Source

The data used in the multivariate model were extracted from the database of the Brazilian Institute of Geography and Statistics (IBGE), the Department of Informatics of the Brazilian Unified Health System (DATASUS) and the transparency portal of the Office of the Comptroller General of the Union. The sample comprises 1,191 municipalities belonging to the Southern Region of Brazil. The list of variables used is shown in Table 1.

Table 1

Variables used to calculate the Sustainable and Agricultural Development Index
Dimension	Variable	Description	Source	Year
Sustainable and social development	x1	Geometric Growth Rate	IBGE	2021
	x2	Birth Rate	Datasus and IBGE	2021
	x3	Number of hospital beds/thousand inhabitants	Datasus	2022
	x4	Illiteracy rate	IBGE	2017
	x5	Number of households with access to mains water supply/TF	Datasus	2015
	x6	Number of households with access to Garbage/TF collection	Datasus	2015
	x7	Number of households with other garbage disposal/TF	Datasus	2015
	x8	Number of households with access to the sewer/TF network	Datasus	2015
	x9	Number of households that do not have access to the sewer/TF network	Datasus	2015
Environmental	x10	Number of agricultural establishments with Springs or Streams - protected by forests/NE	IBGE	2017
	x11	Number of agricultural establishments with silviculture species/NE	IBGE	2017
	x12	Forestry production - planted forests/TA	IBGE	2017
	x13	Forest production - native forests/TA	IBGE	2017
	x14	Number of agricultural establishments that use chemical fertilization/NE	IBGE	2017
	x15	Number of establishments that use pesticides/NE	IBGE	2017
Economic Dimension	x16	Value of resources allocated to municipalities (Union and states)	Office of the Comptroller General of the Union	2022
Economic Dimension	x17	GDP per capita	IBGE	2020
Agricultural	x18	Production of temporary/HV crops	IBGE	2017
	x19	Production of permanent crops/HV	IBGE	2017
	x20	Livestock and other animal husbandry/TA	IBGE	2017
	x21	Number of agricultural establishments that make use of Irrigation/NE	IBGE	2017
	x22	Number of tractors in agricultural establishments/NE	IBGE	2017
	x23	Number of agricultural establishments with employed staff/NE	IBGE	2017
	x24	Amount of expenses incurred by agricultural establishments with Medicines for animals/NE	IBGE	2017
	x25	Amount of expenses incurred by agricultural establishments with fuels and lubricants/NE	IBGE	2017
	x26	GVA of agriculture/TA	IBGE	2020

Source: Prepared by the authors, 2023.

Note: TF = Total Families; AT = Total Area; and NE = Number of Agricultural Establishments.

Finally, after the presentation of the methodology used, the results are presented, analyzed, and discussed.

3.1 Presentation and Interpretation of Factors

When analyzing the adequacy of the sample used, it was found that in the Kaiser-Meyer Olkim (KMO) test the value obtained was 0.6568, indicating the good adequacy of the sample used, and the Bartlett sphericity test was significant and with a statistical of 26528.302, rejecting the null hypothesis, which indicates that the correlation matrix is an identity matrix. Thus, based on the values obtained through the tests performed, it was possible to conclude that the sample used is adequate to the factor analysis procedure and, therefore, allows the continuation of the study.

By performing the factor analysis using the principal components method, it was possible to extract seven factors with characteristic roots greater than one, synthesizing the information contained in the twenty-six variables analyzed, as shown in Table 1. The contribution of the seven factors that explain the total variance of the indicators used is significant, representing 67.09% of the total variance of the data set. For Hair et al. (2009), the use of an accumulated variance of 60% is satisfactory in social sciences.

Table 1

Characteristic root, percentage explained by each factor and accumulated variance
Factor	Characteristic Root	Variance explained by the factor	Cumulative variance
Factor 1	4.66137	18.65%	18.65%
Factor 2	3.26799	13.07%	31.72%
Factor 3	2.87585	11.50%	43.22%
Factor 4	1.86501	7.46%	50.68%
Factor 5	1.54686	6.19%	56.87%
Factor 6	1.4244	5.70%	62.57%
Factor 7	1.132	4.53%	67.09%

Source: Survey results, 2023.

After presenting the factors that will compose the constructed indicator, the orthogonal rotation of these factors is carried out using the varimax method. Thus, Table 2 shows the factor loadings and commonalities for the factors considered in this research. In its interpretation, only factor loadings with values above 0.5 (bold italics) were considered, to indicate the variables that are most strongly associated with a given factor.

Table 2

Factor loadings and commonality after orthogonal rotation of factors
Variable	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7
x1	0.0975	0.212	0.0128	0.5292	0.2588	0.0318	0.235	0.5422
x2	0.0521	0.1921	0.0853	0.0001	0.0303	0.7754	0.0482	0.3486
x3	0.1024	0.0161	0.0053	0.12	0.0414	0.7737	0.0537	0.3716
x4	-0.0248	0.0644	-0.0943	0.0572	-0.1544	-0.1544	-0.6473	0.5164
x5	0.9901	0.0312	0.0073	0.0217	0.0092	0.0324	0.0238	0.0165
x6	0.99	0.0383	0.0031	0.0314	0.0075	0.0401	0.0282	0.015
x7	0.5477	0.167	0.1686	0.3016	0.0799	0.0499	0.0302	0.6225
x8	0.9321	0.0058	0.0241	0.0092	0.0121	0.0752	0.0129	0.1246
x9	0.6208	0.0999	0.0645	0.0305	0.0618	0.3252	0.109	0.4781
x10	0.0465	0.3635	0.0403	0.5714	0.3192	0.1822	0.0316	0.4015
x11	0.0685	0.0494	0.1979	0.0004	0.8918	0.0064	0.0397	0.1568
x12	0.0952	0.274	0.7204	0.1874	0.1595	0.0106	0.1243	0.3207
x13	0.1093	0.7879	0.2081	0.2014	0.1083	0.029	0.2314	0.2173
x14	0.0371	0.7983	0.1175	0.0104	0.1548	0.0069	0.0591	0.3199
x15	0.0131	0.2874	0.0778	0.018	0.8338	0.0713	0.0922	0.202
x16	0.7973	0.1396	0.0426	0.2857	0.0476	0.1643	0.0729	0.2267
x17	0.0414	0.0999	0.2114	0.0701	0.0194	0.0224	0.6688	0.4906
x18	0.131	0.0428	0.0742	0.7251	0.0206	0.0648	0.0207	0.4447
x19	0.1485	0.6159	0.0451	0.3753	0.0587	0.0945	0.0687	0.4386
x20	0.0493	0.7899	0.0034	0.1803	0.2619	0.1031	0.2057	0.2196
x21	0.0071	0.6469	0.2533	0.3827	0.2416	0.012	0.2591	0.2452
x22	0.0074	0.1812	0.7751	0.1034	0.0288	0.0191	0.039	0.3529
x23	0.03	0.5336	0.3481	0.4139	0.0487	0.0132	0.007	0.4193
x24	0.0491	0.1259	0.6215	0.2062	0.0458	0.0115	0.2508	0.4878
x25	0.0106	0.0403	0.7932	0.0879	0.08	0.0253	0.0404	0.3528
x26	0.113	0.1057	0.1788	0.2122	0.1581	0.0144	0.6803	0.4111

Source: Survey results, 2023.

The values found for the commonalities point to how much of the seven factors explain each variable, and it is also observed that there is a positive relationship between the variables that make up the factors. Commonality is presented next to each factor loading. In view of this, it is possible to observe that Factor 1 is strongly related to the following variables: Number of families with access to public water supply/TF (x5); Number of households with access to garbage/TF collection (x6); Number of households with other garbage disposal/TF (x7); Number of households with access to the sewer/TF network (8); Number of households that do not have access to the sewer/TF network (x9); and, Transfers of public resources (x16). Factor 1 presents the highest explained variance, corresponding to 18.65% of the total accumulated variance. This factor is related to public investment, infrastructure, and basic sanitation.

Factor 2 is related to the following variables: Forest production - native forests/TA (x13); number of agricultural establishments that use chemical fertilization/NE (x14); Production of permanent crops/TA (x19); Livestock and other animal husbandry/TA (x20); Number of agricultural establishments that make use of Irrigation/NE (x21); Number of agricultural establishments with employed personnel/ NE (x23). This factor is associated with technology and land use, and has the second highest explained variance, corresponding to 13.07% of the total accumulated variance.

Factor 3 is related to the following variables: Forest production - planted forests/TA (x12); Number of tractors in agricultural establishments/NE (x22); Amount of expenses incurred by agricultural establishments with Medicines for animals/NE (x24); Value of expenses incurred by agricultural establishments with fuels and lubricants/NE (x25). This factor is associated with environmental degradation, and has the third highest explained variance, corresponding to 11.50% of the total accumulated variance.

Factor 4 is related to the following variables: Geometric growth rate (x1); Number of agricultural establishments with springs or streams - protected by forests/NE (x10); Production of temporary crops/TA (x18). This factor is associated with the increase in agricultural production, and has the fourth highest explained variance, corresponding to 7.46% of the total accumulated variance.

Factor 5 is related to the following variables: Number of agricultural establishments with silviculture/NE species (x11); and Number of establishments that use pesticides/NE (x15). This factor is associated with pest control and combat and has the fifth highest explained variance, corresponding to 6.19% of the total accumulated variance.

Factor 6 is related to the following variables: Birth Rate (x2); and Number of hospital beds/thousand inhabitants (x3). This factor is associated with Hospital and Health Structure and has the sixth highest explained variance, corresponding to 5.70% of the total accumulated variance.

Factor 7 is related to the following variables: Illiteracy rate (x4); GDP per Capita (x17); and GVA of agriculture/TA (x26). It is noteworthy that the variable x4 has a negative impact on the composition of the factor, i.e., on both agricultural and per capita income. This factor is associated with income and has the seventh highest explained variance, corresponding to 4.53% of the total accumulated variance.

Thus, the twenty-six variables used were synthesized into seven factors, namely: Factor 1, Public Investment and Basic Sanitation; Factor 2: Technology and Land Use; Factor 3, Environmental Degradation; Factor 4, Agricultural Production; Factor 5, Pest Fighting and Control; Factor 6, Hospital Structure and Health; and Factor 7, Income, which together explain 67.09% of the total variance of the indicators analyzed.

3.2 Analyzing and Discussing the SADI

From the factorial scores, it was possible to construct the Sustainable and Agricultural Development Index (SADI) for the municipalities of the Southern Region of Brazil. This index varies between 0 and 1, and the closer it is to 1, the greater the degree of sustainable and agricultural development of the municipality that makes up the sample. The values found for the indicators showed a medium-low degree, considering that the highest value was 0.5885 and the others had values below 0.5. Thus, the municipalities with the highest indicators, with the highest degree of sustainable and agricultural development, were Curitiba (PR) (0.5885); Capivari do Sul (RS) (0.4541); and Paranapoema (PR) (0.4536). While the lowest indicators, with a lower degree of sustainable and agricultural development, were: Bombinhas (SC), (0.1843); Governador Celso Ramos (SC) (0.2170); and Matinhos (PR) (0.2172), as shown in Table 3.

Table 3

Calculated SADI Result
Major Indicators				Lower Indicators
Municipality	SADI	GDP (thousand USD)	GDP Agriculture (thousand USD)	Municipality	SADI	GDP (thousand USD)	GDP Agriculture (thousand USD)
Curitiba (PR)	0.5885	17602600.86	3277.99	Adrianópolis (PR)	0.2371	56.98	4.46
Capivari do Sul (RS)	0.4541	52.23	13962.88	Itaperuçu (PR)	0.2305	114.35	5.69
Paranapoema (PR)	0.4536	15.48	6873.11	Garopaba (SC)	0.2277	136.53	3.70
Turvo (SC)	0.4348	128.48	22505.38	Gravatal (SC)	0.2274	57.66	3.66
Ituporanga (SC)	0.4333	221231.86	51600.62	Pontal do Paraná (PR)	0.2256	118.68	0.81
Imbuia (SC)	0.4296	42.40	17631.16	Altamira do Paraná (PR)	0.2222	16.70	8.53
Faxinal do Soturno (RS)	0.4287	45.19	3520.97	Pescaria Brava (SC)	0.2173	21.46	1.06
Carazinho (RS)	0.4262	618106.56	25205.71	Matinhos (PR)	0.2172	163.97	0.38
Mafra (SC)	0.4240	419950.17	54590.97	Governador Celso Ramos (SC)	0.2170	75.70	9.94
Vacaria (RS)	0.4221	506691.12	95583.64	Bombinhas (SC)	0.1843	156.99	5.06

Source: Survey results, 2023.

Figure 3 shows the spatial distribution of the SADI proposed in this study. Therefore, the areas were colored according to the values presented in this index, so that the lighter shades refer to the municipalities in the South Region of Brazil that had the lowest SADIs, while the darker shades represent the states with the highest SADIs.

Data from IBGE (2020) indicate that the Northwest Mesoregion of Rio Grande do Sul is responsible for 41.89% of the GVA of agriculture, being the region with the highest agricultural production in the state. In view of this, Lisbinski et al. (2020) analyzed the degree of rural development of this Mesoregion and found that the municipalities belonging to the South and Midwest of this region had a higher concentration of high and medium degrees of rural development. On the other hand, the municipalities in the northern part of this Mesoregion showed a low degree of development. These findings agree with the results found in this research, because when observing the map, it is possible to verify that the highest concentration of high SADIs in the state of Rio Grande do Sul are found in the Northwest Mesoregion.

It was also observed that the municipality in the state of Rio Grande do Sul with the highest SADI was Capivari do Sul (0.4541). The result corroborates the work of Fujimoto et al. (2006) who, when developing an indicator of socioeconomic performance of the northern coast of the state of Rio Grande do Sul, identifying environmental problems, found that the municipality with the highest degree of development in this region was Capivari do Sul. In addition, the authors found that the municipality's GVA has a predominance in agriculture and the population is predominantly urban, however, according to IBGE data (2022), there was a decrease of 1.36% in the population between the years 2020 and 2021.

Regarding the state of Santa Catarina, the highest indicator was found in the municipality of Turvo located on the southern coast of the state, while the other high-performance indicators were found in the Central Region, in the Alto Vale do Itajaí. Silva and Rosa (2020) analyzed the sustainable development indicators of the Santa Catarina mesoregions, the results pointed out that the state has a sustainable performance considered medium-low in its five mesoregions, and one mesoregion presented a median sustainable performance, which is the Itajaí Valley, with an IDMS of 0.625, which corroborates the results found in this work. The authors justify this by the fact of the low levels of environmental management, economic dynamism, and distribution of wealth, pointing to the need for greater attention to the Economic and Environmental dimensions by municipal public managers.

About the state of Paraná, it was possible to verify that the largest indicator of the observed sample belongs to Curitiba, in addition, it can be observed that the highest indicators of SADIs are in the east of Paraná, in the municipalities close to Curitiba. Eberhardt and Lima (2012) analyzed the profile and stage of economic development of the regions of the state of Paraná and found similar results. The authors were able to verify that of the regions analyzed, the micro-region of Curitiba was the one that presented the best performance, since its indicator was higher than the others, in addition, the region was considered the most dynamic in economic terms, when compared to the other micro-regions of the state.

Turra and Lima (2018) created a Sustainable Development Index (SDI) for the micro-regions of the state of Paraná, the results pointed out that the Micro-region of Curitiba occupied the first place, having its performance associated with the economic and institutional dimensions, a characteristic present in places that have already undergone an intense urbanization process. While the worst indicators were concentrated in micro-regions located in the Center-South Mesoregion of the state, which is characterized by low temperatures and rugged terrain, inhibiting the development of certain grain crops (soybeans and corn, for example), favoring livestock activities and reforestation.

In view of this, it was possible to verify the high degree of heterogeneity in the distribution of the SADI among the municipalities of the Southern Region. According to Belik (2015), in view of this existing heterogeneity, especially in the rural sector, policies must be differentiated according to the type of clientele, making it impossible to adopt a single agricultural and sustainable development policy. It is noteworthy that it is not only a question of the technological gap, but also of the gap in relation to access to services, transportation and commercialization, and income. Thus, if the objective is to promote sustainable development and Brazilian agribusiness jointly, it is necessary to work on some policies capable of facilitating the insertion of more appropriate and sustainable technologies, as well as the construction of markets for these products and services offered by rural producers.

3.3 Spatial Distribution of Data

From now on, we move on to the analysis of the ESDA, which occurs from the interaction of the municipalities in space, that is, whether the value of a variable observed in one municipality $\:i$ depends on the value of this variable in the other neighboring municipalities.

To perform the Moran's $\:I\:$autocorrelation test, the Queen and Tower spatial weight matrices were tested, the highest result was achieved by using a Queen-type spatial weight matrix, therefore, the analyses were made using this spatial configuration, considering that it better represents the interaction between the regions. Thus, Moran's $\:I$ autocorrelation test presented a statistic of 0.870, demonstrating strong evidence of positive spatial correlation, presenting high statistical significance, rejecting the null hypothesis, demonstrating the existence of spatial correlation between the municipalities that make up the sample. Cluster analysis grouped the indicator into five groups: High-High, Low-High, High-Low, Low-Low and non-significant. Thus, it is possible to observe that the municipalities that presented high indicators and that have neighbors with high SADI indicators (High-High) are mainly concentrated in the state of Rio Grande do Sul and the South and Center-South of Santa Catarina.

The aspects that lead to the highest degree of sustainable and agricultural development in these regions are the climate and soil favorable to the production of various crops such as soybeans, corn and wheat, for example; the high degree of technology and innovation used in these regions, which indicated high indicators, since in the Southern Region of Brazil there are several institutions for research and development of agricultural technology, which contribute to the improvement of productivity and sustainability of agriculture in these areas; the organization of farmers in these regions into cooperatives, allowing greater bargaining capacity and access to broader and more lucrative markets; and, invested in public policies aimed at agriculture, such as tax incentives, rural credit, technical assistance and rural extension, which has contributed to the sustainable development of the agricultural sector (Tiecher, 2016; Tomazzoni & Schneider, 2017; Lisbinski et al., 2020).

The Low-Low clusters, those with low indicators, surrounded by low indicators, are mainly found in the Central and Western regions of Paraná. According to Leiva (2018) and Gioia, Barros, and Barros (2017), the Central and Western regions of Paraná still face problems related to basic sanitation in the region, where most of them do not have satisfactory data on access to sewage networks, which can negatively affect the quality of life of rural producers and limit the potential for development in the region. In addition, about solid waste collection, it was evidenced that the cities in this region still have dumps that do not comply with specific legislation. Finally, it should be noted that some parts of these regions have soils with low fertility, which can hinder the cultivation of some agricultural crops, in addition to limiting productivity, and the region is also affected by droughts and intense heat, which affects agricultural production and increases the risk of loss of some crops (Tiecher, 2016).

The Low-High clusters, with low indicators, but surrounded by high indicators, are found in the states of Rio Grande do Sul and Santa Catarina; while the High-Low clusters, with high indicators, but surrounded by low indicators, are mainly found in the state of Paraná, which demonstrates the high disparity in the sustainable and agricultural development of the municipalities that make up these states. Among the factors that lead to this, we can mention the diversity in terms of geographical characteristics, such as relief, climate, soil, and water availability, which affects soil fertility, agricultural productivity, and sustainable development (Medeiros, 2011). In addition, investment in infrastructure among municipalities is skyrocketing, so that some receive more investments than others, impacting the availability of roads, storage, and transportation (Gazolla & Schneider, 2013; Manica, 2017). The municipalities of the Southern Region of Brazil have different economic activities, so that some may be more prone to sustainable development than others (Rossato, File & Lily, 2010; Fochezatto & Tartaruga, 2016). And finally, there is a difference in the implementation of public policies among municipalities, so that those that implement public policies aimed at access to health, education and sanitation services, sustainable agriculture and invest in agricultural research and technology will present a more advanced level of development (Schneider & Waquil, 2001; Smile, Gazolla & Schneider, 2010).

Therefore, it is possible to observe that the regional inequalities between the analyzed states and within the analyzed states are large, so that in the same region there are the highest and lowest indicators, and this is due to the various factors mentioned above. Thus, to face these challenges, a joint and coordinated effort between producers, government and civil society is necessary, aiming at the implementation of sustainable agricultural practices, improving infrastructure and investments in research and technology, in addition to promoting public policies aimed at regional and social development.

The objective of this study was to create and analyze an index of sustainable and agricultural development for the municipalities that make up the Southern Region of the country. For this, factor analysis was used as a methodology through principal component analysis and later the exploratory analysis of spatial data (ESDA) was performed, verifying the distribution of the indicator and the formation of clusters.

The main results revealed that the values found for the indicators showed a medium-low degree, considering that the highest value found was 0.5885 and the others presented values below 0.5. The municipalities with the highest indicators, with the highest degree of sustainable and agricultural development, were Curitiba (PR) (0.5885); Capivari do Sul (RS) (0.4541); and Paranapoema (PR) (0.4536). While the lowest indicators, with a lower degree of sustainable and agricultural development, were: Bombinhas (SC), (0.1843); Governador Celso Ramos (SC) (0.2170); and Matinhos (PR) (0.2172).

Regarding the ESDA, it showed strong evidence of a positive and significant spatial correlation between the municipalities analyzed. It was observed that the municipalities with clusters of the High-High type are mainly concentrated in the state of Rio Grande do Sul and the South and Center-South of Santa Catarina. The Low-Low clusters are mainly found in the Central and Western Regions of Paraná. The Low-High clusters are found in the states of Rio Grande do Sul and Santa Catarina. And, finally, the High-Low clusters are mainly concentrated in the state of Paraná, which demonstrates the high disparity in the sustainable and agricultural development of the municipalities that make up these states.

Therefore, from this work, it was possible to verify which municipalities present problems of sustainable and agricultural development, as well as in which regions these problems are concentrated, suggesting the adoption of more efficient and effective strategies by the public authorities, with the objective of leveraging sustainable and agricultural development, aligning both purposes. Bearing in mind that this is the great challenge of Brazilian agriculture. With this, it will be possible to promote the improvement of the infrastructure of these municipalities, as well as the quality of life of the population residing there.

Finally, in view of the complexity and breadth of the theme of the present work, as there is still no consensus on the concept and dimensions of sustainable development, in addition to the difficulty in defining variables that can capture effects related to both agricultural and sustainable development, further research is suggested with a view to deepening the theme of agricultural and sustainable development. either through an increase in the number of analysis variables, or through regional, state, and national analysis.

Almeida E (2012) Econometria espacial. Alínea, Campinas–SP, p 31
Anselin L (1999) The future of spatial analysis in the social sciences. Geographic Inform Sci 5(2):67–76
Belik W (2015) A heterogeneidade e suas implicações para as políticas públicas no rural brasileiro. Revista de Economia e Sociologia Rural 53:9–30
Controladoria-Geral da União (2022) Portal da Transparência, Detalhamento de Recursos Transferidos por UF e Município. Retrieved from https://portaldatransparencia.gov.br/transferencias/consulta?ordenarPor=mesAno&direcao=desc accessed on: April 12, 2024
Departamento de Informática do Sistema Único de Saúde do Brasil, DATASUS (2023) Tabnet. Retrieved from https://datasus.saude.gov.br/informacoes-de-saude-tabnet/ accessed on: April 12, 2024
Eberhardt PHDC, De Lima JF (2012) Evolução e estágio do desenvolvimento regional: o caso das regiões do Paraná. DRd-Desenvolvimento Reg em debate 2(1):189–203
Fávero LP, Belfiore P (2017) Manual de análise de dados. Elsevier, Rio de Janeiro
Fochezatto A, Tartaruga IGP (2015) Atividades econômicas potencialmente inovadoras e desenvolvimento regional no Rio Grande do Sul, vol 15. Revista Brasileira de Economia de Empresas, 1
Froehlich C (2014) Sustentabilidade: dimensões e métodos de mensuração de resultados. Desenvolve Revista de Gestão do Unilasalle 3(2):151–168
Fujimoto NSVM et al (2006) Litoral norte do estado do Rio Grande do Sul: indicadores socioeconômicos e principais problemas ambientais. Desenvolvimento e Meio ambiente, p 13
Gioia T, Barros MVF, Barros ONF (2017) Índices de vulnerabilidade da saúde aplicados aos municípios do Paraná. Confins. Revue franco-brésilienne de géographie/Revista franco-brasilera de geografia, (33)
Grisa C, Gazolla M, Schneider S (2010) A produção invisível na agricultura familiar: autoconsumo, segurança alimentar e políticas públicas de desenvolvimento rural. Agroalimentaria 16:65–79
Hair JF et al (2009) Análise Multivariada de Dados. 6. ed. Porto Alegre: Bookman. Brazilian Institute of Geography and Statistics, IBGE. (2017). Censo agropecuário 2017. Retrieved from https://sidra.ibge.gov.br/pesquisa/censo-agropecuario/censo-agropecuario-2017/resultados-definitivos accessed on: April 12, 2024
Brazilian Institute of Geography and Statistics, IBGE (2021) Pesquisa Agrícola Municipal, 2021. Retrieved from https://sidra.ibge.gov.br/pesquisa/pam/tabelas accessed on: April 12, 2024
Brazilian Institute of Geography and Statistics, IBGE (2022) Produto Estimativas de População – EstimaPop, 2022. Retrieved from https://sidra.ibge.gov.br/tabela/6579 accessed on: April 12, 2024
Brazilian Institute of Geography and Statistics, IBGE (2020) Produto Interno Bruto dos Municípios, 2020. Retrieved from https://sidra.ibge.gov.br/tabela/5938 accessed on: April 12, 2024
Lattin J, Carroll JD, Green PE (2011) Análise de Dados Multivariados. 1°. Cengage Learning, ed. São Paulo
Leiva JRL (2018) Saneamento Básico na Região Oeste do Paraná. Revista Gestão Sustentabilidade Ambiental 7(2):510–524
Lins J et al (2015) Análise espacial da evolução do índice de desenvolvimento humano nos municípios da região nordeste. Revista Economia e Desenvolvimento 14(1):81–96
Lisbinski FC et al (2020) Análise Espacial do Desenvolvimento Rural da Mesorregião Noroeste do Rio Grande do Sul. Revista Brasileira de Estudos Regionais e Urbanos 14(1):79–101
Medeiros M (2011) Diversidade de saberes em situações de interface: A emergência da agricultura de base. ecológica entre agricultores familiares no Sul do Rio Grande do Sul. Universidade Federal do Rio Grande do Sul. Faculdade de Ciências Econômicas. Dissertação de mestrado em Desenvolvimento Rural, p 156
Melo CO (2006) Caracterização do desenvolvimento rural dos municípios paranaenses: uma análise com base na estatística multivariada. Dissertação de Mestrado. Universidade Estadual de Maringá (UEM). Departamento de Economia, p 113
Mingoti SA (2005) Análise de dados através de métodos de Estatística Multivariada: uma abordagem aplicada. UFMG, Belo Horizonte
Organization for Economic Cooperation and, Development OECD (1993) OECD Core Set of Indicators for Environmental Performance Reviews: a synthesis report by the group on the state of the environment. Environment monographs
Parré JL (2014) Interpretando o espaço rural: desenvolvimento, recursos naturais e infraestrutura. Anais do XLI Encontro Nacional de Economia. ANPEC-Associação Nacional dos Centros de Pós-Graduação em Economia
Pawlowski A (2008) How many dimensions does sustainable development have? Sustain Dev 16(2):81–90
Pereira AA (2004) Agricultura de Goiás: análise & dinâmica. Editora da UCG
Rossato MV, Lima JE, Lírio VS (2010) Condições econômicas e nível de qualidade ambiental no estado do Rio Grande do Sul. Revista de Economia e Sociologia Rural 48:587–604
Sabater LA, Tur AA, Azorín JMN (2011) Análise exploratória dos dados espaciais (AEDE). In: Costa JS, Dentinho TP, Nijkamp P (eds) Compêndio de economia regional: métodos e técnicas de análise regional. Principia, Cascais, pp 237–298
Sachs I (1993) Estratégias de Transição para o Século XXI: desenvolvimento e meio ambiente. Studio Nobel e Fundação de Desenvolvimento Administrativo (Fundap), São Paulo
Sambuichi RHR et al (2012) A sustentabilidade ambiental da agropecuária brasileira: impactos, políticas públicas e desafios. Texto para Discussão
Schneider S, Waquil PD (2019) Caracterização socioeconômica dos municípios gaúchos e desigualdades regionais. Revista de Economia e Sociologia Rural 39(3):117–142
Silva LC, Rosa FS (2020) Indicadores de desenvolvimento sustentável das mesorregiões catarinenses: uma análise comparativa. REVISTA AMBIENTE CONTÁBIL-Universidade Federal do Rio Grande do Norte, 12(1), 273–294
Spangenberg J, Bonniot O (1998) Sustainability indicators: a compasso the road towards sustainability. Wuppertal Institute, p 81
Tiecher T (2016) Manejo e conservação do solo e da água em pequenas propriedades rurais no sul do Brasil: práticas alternativas de manejo visando a conservação do solo e da água. UFRGS, Porto Alegre
Tomazzoni GC, Schneider S (eds) (2022) A presença do cooperativismo na agricultura do Sul do Brasil: uma breve caracterização a partir dos dados do censo agropecuário 2017. DRd-Desenvolvimento Regional Em Debate, 12(ed. esp. Dossie), 65–88
Turra S, Lima AC (2018) Desenvolvimento Sustentável das Microrregiões do Estado do Paraná: uma análise pelo método Dashboard of Sustentability, vol 14. Revista Brasileira de Gestão e Desenvolvimento Regional, 5
Werbach A (2010) Estratégia para sustentabilidade: uma nova forma de planejar sua estratégia empresarial. Elsevier, Rio de Janeiro
World Commission on Environment and Development, WCED (1987) OUR COMMON FUTURE. Retrieved from < https://sustainabledevelopment.un.org/content/documents/5987our-common-future.pdf accessed on: April 12, 2024

The authors declare no competing interests.

DeclarationofConflictofInterest1.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Spatial Analysis of Sustainable and Agricultural Development in Southern Brazil

Status:

Version 1

Abstract

Figures

Introduction

Methodology

2.1 Statistical Foundations of the SADI

2.2 A Synthesis of Exploratory Spatial Data Analysis (ESDA)

2.3 Variables and Data Source

Analysis and Discussion of Results

3.1 Presentation and Interpretation of Factors

3.2 Analyzing and Discussing the SADI

3.3 Spatial Distribution of Data

Conclusion

References

Additional Declarations

Supplementary Files

Status:

Version 1