Agricultural International Trade by Brazilian Ports: A Study Using Social Network Analysis

Abstract

Agribusiness trade is a complex network of commercial relations among countries, and it is influenced by on-shore and off-shore logistics. Therefore, it is essential to comprehend these relationships to improve decision-making regarding production and logistical development. This paper investigates Brazilian agricultural and livestock exports between 2013 and 2022 to understand logistical bottlenecks based on trade partners. To do so, we performed descriptive statistics and social network analysis (SNA) considering measures such as degree centrality, k-core, and tie strength. Our results indicate Brazil’s dependency on Asian markets whereby eight of ten are located on this continent. We observe an unexpected result regarding the low purchase of these products byimportant Brazilian partners such as the United States, the UK, and the European Union. Finally, the study confirms the Brazilian logistical bottleneck where two logistical corridors correspond to 76% of all agricultural exports in the period, with Santos, the busiest port, moving more than 46% of the cargo.

1. Introduction

Food production plays an essential role in our world because it is responsible to provide our nutritional needs on a daily basis. However, the global population is growing and there is a lot of concern whether the food available will be enough. According to United Nations, the global population could grow to around 8.5 billion in 2030, 9.7 billion in 2050, and 10.4 billion in 2100 [1]. In this sense, many developing countries have been seen as a source of agricultural land to attend to this demand and at the same time to face the challenge of malnutrition and hunger over the world.

Brazil is located in Latin America and has the fifth largest territorial size in the world. Currently, it is one of the most important agricultural producers in the international scenario and uses just 8% of its land area to do so [2]. Due to this fact, specialists believe that Brazil will be “the barn of the world” regarding food production. However, how is this production nowadays? Which are the main consumer markets? Is there a concentration or it is spread among many countries? Are there logistical bottlenecks affecting international trade?

To start to answer those questions it is first necessary to understand the agri-food supply chain (AFSC) concept and, second, to find ways to measure and study the networks behind international trade. An agricultural supply chain involves a series of processes and different actors that provide the raw materials for many food products and animal feeding in addition to beverages, cereal, fruits, vegetables, etc. An agri-food supply chain can also be considered a complex network where a variety of actors interact to make the right decisions in order to achieve end-customer requirements [3].

Agri-food supply chains have been studied in different contexts. Esteso et al. [4], investigated the impact of product perishability on an AFSC design using a mixed-integer linear programming model which considers capacity, planting, harvesting, transporting, and perishability constraints for a multiple-period horizon. Cao et al. [5] examined AFSC by building a game model and exploring optimal decisions of all entities in a decentralized and a centralized system. They designed a cost-sharing and a buyback contract to coordinate greening effort decisions and analyzed the impact of the green standards at different stages of the supply chain. Ciccullo et al. [6] investigated the range of the available technologies and the detailed objectives of such technologies for food loss and waste prevention in AFSC (i.e., forecasting, monitoring, grouping, shelf life extension, product quality, and value upgrading). Hu et al. [7] analyzed the quality decision in an AFSC composed of four actors (i.e., agricultural producer, processing enterprise, distributor, and consumers) using Stackelberg’s game theory and developing combined multiple strategies (profit sharing, quality commitment, and risk sharing) for coordinating quality control in the AFSC.

One way to measure relationships in a network is to count the volume of transactions between nodes. These transactions could be different variables such as information, data, material, and currency. Additionally, to perform an analysis of relationships we can use the social network analysis method.

The social network method, also known as social network analysis (SNA), is a research method based on graph theories, probability theory, and geometry [8]. It began in 1934, in the USA, with the development of new methods of analysis among small groups’ relationships in sociometric studies by Jacob L. Moreno. This method gained relevance between the 1940s and 1950s and theoretical and methodological properties were established between the 1960s and 1980s [9]. However, it was Wellman [10] who promoted studies with new concepts and techniques that were developed to support the capture, representation, and relationship between social actors, as well as the structural interpretation under the concept of networks. It consolidated into an advanced and robust technique that was explored over time and applied in several areas, such as political science [11], ecology [12], tourism [13], sports science [14], education [15], and manufacturing [16].

Nowadays, people live immersed in a significant amount of data, as a response to the mass use of the internet. The internet is responsible for the generation and centralization of data from different areas, such as social, economic, scientific, and political, becoming the epicenter of such activities [17]. Due to the amount of data generated, data analysis becomes a challenge for the scientific community. Thus, it is important to have software to facilitate the qualitative and quantitative analysis of these networks, as well as their categorization, tabulation, and manipulation. Those analyses are only possible through software developed as tools for the use and application of the SNA [18]. Among several software possibilities, for the present study, we chose to use UCINET 6 © and NetDraw 2.138 ©.

With these ideas in mind, we established two research questions for the present study:

RQ1. Is there a concentration of Brazilian agribusiness export flows regarding logistics and market countries?

RQ2. Is it possible to obtain knowledge of the Brazilian agri-food supply chain network that could guide the decision-making process of agricultural production, international trade, and logistics processes?

To answer those questions the objective of this paper is to investigate Brazilian agribusiness exports between 2013 and 2022, considering export routes and which countries are market partners in international trade.The analysis was performed using descriptive statistics and SNA techniques with computational tools to produce graphics and statistics.

The article is divided as follows: after this introduction, materials and methods are presented, results and discussion show the main findings and the conclusion points out the main ideas obtained from the research.

2. Materials and Methods

A social network analysis (SNA) approach was adopted to analyzed the Brazilian agricultural exports network between 2013 and 2022. SNA methods provide valuable information from a network produced based on a data set [19].

Several measures in a network are proposed by the SNA method. Among them, some of the most visible can be highlighted: network centrality, k-core, and tie strength relationship. Network centrality is divided into three folds in the literature: degree, closeness, and betweenness [20]. In this study, to investigate the Brazilian agricultural exports network, the relevant centrality measure was degree, due to direct relationships. The centrality degree is a measure of a node’s immediate adjacency and involvement in the network. It can be defined as the number of ties that a given type of node has [21].

According to Borgatti, Everett, and Johnson [19], there is not one single concept for measures of centrality, but a group of concepts. In general, we can understand the centrality measures in a relation to the node’s contribution to the network structure. In other words, centrality is considered a structural node [22]. This importance can be interpreted in different ways depending on the type of data analyzed. Central nodes are commonly interpreted in social relationships as the most prominent, most influential, or as leaders [23], or as having great autonomy, control, visibility, involvement, prestige, power, and so on [19].

Another measure adopted in this research was k-core. It is one of the most common methods of SNA and results in a subgraph where each actor has a degree k or greater than the other actors in the subgraph. Therefore, in a k-core where k = 2, each actor is connected to at least two other actors. A k-core analysis can be used as a criterion for decreasing nodes in a very complex network. By eliminating nodes with a low k value, it is possible to focus on the most relevant nodes [24].

When it comes to relationships among actors, the most relevant aspect is the volume of transactions. Many characteristics are possible, such as monetary value, load transferred, information exchange, and so on. One way to study these relationships is based on volumes in a table. However, the most comprehensive analysis is possible when a graph network is plotted. In the network, the lines among actors represent connections between nodes and are also important sources of data interpretation. The thickness of a line, for example, can be used to indicate the frequency of interaction [25], the strength of the connection between the pairs [12], or the volume of the connection [19].

2.1. Data Collection

The data were collected from the Brazilian Foreign Trade Portal [26]. The data are provided based on the Siscomex system that registers all the transactions of Brazil’s international trade, and are made available in an online platform. They refer to the volume of Brazilian exports of agricultural and livestock production by maritime cargo ports to importer countries in metric tonnes. For data extraction, the following parameters were adopted:

• We selected a ten-year period from January 2013 to December 2022.

• The extraction filters were international standard classification by economic activities, selecting the division of plant, animal, and hunting; maritime transportation; ports of exportation; and importer country.

The research returned 2353 records (lines) with 7 columns (year, ISIC division code, ISIC division description, via, port, countries, metric tonnes). We summed the volume of exports over the years for every country. Values lesser than 1000 tons per year were excluded. The final database was composed of 253 records (lines) that represented 96.8% of all data.

We found the cargo was exported via 20 ports. However, some represented less than 1% and we excluded these 10 from the sample, with 10 ports remaining. The data remaining represented 96.4% of the trade volume. After, we adopted the same procedure for countries, but now considering values greater than 0.5% to include important Brazilian partners in the international scenario, such as the United Kingdom and the United States that presented a surprisingly low volume of less than 1%; now 93.5% of data and 182 records (lines) remained. Finally, despite the fact that less important ports were removed from the database, this had no significant effect on the export volume of the countries selected.

2.2. Data Analysis

To analyze the data, we first imported the collected database to Microsoft Excel v.19©. Second, ports and countries were portrayed by their names, removing spelling errors and replacing spaces between words with underscores. Third, countries and ports were arranged in rows and columns considering name, origin, destination, and value.

After the data treatment, we created a vna file in txt notepad presenting actors, relations from–to, and volume of transaction. The vna file is one way to organize data to analyze them in a social network software (Figure 1).

In this study, we adopted UCINET 6.738 © to produce a quantitative analysis of the network, and used its counterpart NetDraw 2.178 © to produce a graphical analysis from the data [27]. Thus, the next step was to import the vna text file in both programs, and the analysis produced can be seen in the next section of this paper.

3. Results and Discussion

3.1. Descriptive Statistics

Before providing an SNA analysis of the data, we explored them using short descriptive statistics of the sample regarding importing countries and Brazilian ports, see Figure 2 and Figure 3.

The descriptive statistics were important to permit us a better comprehension of the results obtained. Note that the Figures present all the countries and ports before the final sample cut. Thus, it was possible to validate the importance of this final arrangement to produce an SNA analysis, using the correct players without outliers.

Note in Figure 2, that the main Brazilian partners are located in the Asian continent. As an example of this movement in direction of Asian countries, until 2000, the major buyer of Brazilian soybean was the European Union; however, China progressively increased soybean imports from Brazil, reaching approximately 75% of the Brazilian soy exports in 2016 [28].

Based on Figure 3, it is possible to verify the high concentration of cargo in Brazilian ports. Despite the country having 35 ports, only five account for more than 5% of the cargo, whereby Santos retains almost 50% of the flow. The Port of Santos is located in the state of São Paulo and is the most important port in Brazil, representing an economic influence of more than 50% of the Brazilian GDP, and 25% of the country’s foreign trade is carried out through it [29]. Figure 4 shows a map of Brazilian ports.

3.2. SNA Analysis

Centrality Degree

The centrality degree is a measure of the node’s immediate adjacency and involvement in the network. It can be defined as the number of ties that a given type of node has [21]. Moreover, the centrality degree illustrates which actors and organizations are at the center of interactions, i.e., when an actor has more links to other actors, he/she has a higher centrality degree [31]. Figure 5 presents the centrality degree network of the study and Table 1 quantifies these relationships.

Figure 5 connects Brazilian ports and importer countries. The larger the node size is, the greater its importance to the network. However, we need to take into account that the volume exported by ports is the same volume received by importers. for this reason, nodes of importer countries will appear small in relation to port ones. The ports are 10 and the countries are 27 in the network.

Our results show that ports are the main actors in the network, whereby Paranagua, Sao Luiz, Belem, and Santos are the most important ones. Despite Santos having the highest volume in ten years, this difference is not necessarily reflected in the number of country connections. However, the results confirm the four busiest ports in Brazil are among those with more import country connections—21 or 22 from a total of 27 possibilities.

From an importer country perspective, Egypt, Morocco, and Spain receive cargo from the ten ports in the sample in the 10 year period representing the use of different logistics corridors. Moreover, there is an indication that cargo depends more on port capacity and vessel availability than a developed logistics corridor.

Although the network presents an effect of two groups of importer countries and cargo ports, the network presents a reasonable centralization of 36.5%. We can infer, based on degree centrality, a tendency for the use of multiple logistics corridors based on the assumption that importers seek vessels available on different routes to minimize logistics’ costs. Importers seem to be interested in obtaining the best negotiation results regarding cargo freight rather than maintaining the use of a regular logistical corridor. International trade is a very important component of the shipping market and the maritime transportation is very sensitive and reacts to any change in any direction of world trade, meaning the demand for shipping space has to adjust to those oscillations which justifies different logistical corridors [32].

3.3. K-Core

This central connection is important as it indicates that a small group has similar relationships in a network (Figure 6). Thus, a k-core is a group of nodes that are connected to at least k nodes in the group; consequently, a clique may be seen as a special condition of a k-core in which k equals the number of nodes in the group minus one [33].

According to k-core measurements, the main network is composed of the red nodes that have an indicator 8, followed by the black nodes with and indicator of 7, and the blue nodes with one of 6. The results illustrate that only one Brazilian port is out of the main network, Santarem; which means that the k-core analysis captured the most relevant part of network. In the centrality degree analysis Santarem was the last node among the ports with 13 connections, and this low impact was repeated in the k-core study as well.

We can infer that our results permitted the capture of the most influential part of the Brazilian agri-food supply chain where the red network presents the most valuable partners. The main network is composed for 9 ports and 10 importer countries. This result is remarkable because it allows decision-makers to pay more attention to these partners and ports, because they represent the main core of the Brazilian trade. At the same time, it proves that main partners do not only focus on one logistical corridor, confirming the conclusions in the centrality degree analysis.

3.4. Tie Strength

The strongness of the connections between importer countries and Brazilian ports can be seen in Figure 7. Table 2 and Table 3 compare the volume exported and the centrality degree.

Our results demonstrate that despite the differences in order of connections in the network (22 or 21 nodes), the ports of Santos, Paranagua, Belem, and São Luiz remain the most important Brazilian ports. However, in this analysis it is possible to characterize two logistics corridors in Brazil, the South/Southwest corridor composed of Paranagua e Santos (60.2%) and the North/Northwest corridor composed of Belem and Sao Luiz (15.8%). Both represent 76% of agricultural exports in the analyzed 10 years; however, the second one moved only 1/4 of the cargo movement of the first corridor.

Although these analyses are relevant, the numbers of the Port of Santos are out of scale and indicate the development of this port related to the others. However, this result indicates how fragile the logistical infrastructure of the country is, as almost 50% depend on only one corridor and one port. A well-developed transportation infrastructure is one of the reasons for the economic growth [34].

From the importer countries’ side only eight of then were responsible for around 74% of the movement. Note that only one country is on the European continent, Spain. On the other hand, other important Brazilian partners, according to the Brazilian government, is not listed in a relevant position—the United States and the United Kingdom. This result points out an important question in the agreements of Brazil with the US, the UK, and the EU. Their impact in the most prominent sector of the country is much lower than would be expected. Nevertheless, the results show the necessity of Brazilian leaders and producers to take note of the Asian market because eight of the ten main partners are located there.

The Asian market is very important for Brazil due to its growing market. China, for instance, faces problems in ensuring that there is food production for its population due to natural limitations. Its production capacity is reduced no matter how technologically advanced it may be. The growth of urban areas and the income of workers increased the demand for food. It represented an opportunity for Brazil to become one of the main suppliers of agricultural products to this Asian country [35].

Cariello et al. [36] pointed out that agribusiness has been the sector with the most relevant participation in Brazilian exports to China. In terms of value, between 2007 and 2017, shipments destined for this Asian country increased by 469%, with an average year-on-year growth of 21%. Nevertheless, it is fundamental to note that China, despite being the main Brazilian commercial partner, occupied only the third place in volume of agricultural and livestock items imported between 2013 and 2022.

4. Conclusions

This paper investigated the destination of Brazilian agricultural and livestock production between 2013 and 2022 considering ports and importer countries. Using social network analysis (SNA) method, we analyzed the cargo movement in volume. Regarding the results, it is possible to identify a higher concentration of on-shore logistics corridors connecting centers of production in Brazil to two main corridors with two ports each. Moreover, the corridor South/Southeast of the ports of Santos e Paranagua moves four times more than the second corridor identified as North/Northeast with the ports of Belem and São Luiz.

Our study presented other surprising results. The US, the UK, and the EU are not among the main partners. Our data show that together they represented 12.4% of the volume in the analyzed period. On the other hand, Asian countries are the main partners, and China, despite its importance and population, represents only the third place. This result was considered unexpected.

This study is not free of limitations. We considered the volume exported and the occupancy of the logistics service; however, the products have different values and therefore, the positions of importer countries can be different in position from our results. These differences are less than in the case of industrial products. Usually, Brazil exports commodities with low added value based on volume, so we considered that is more important to verify the impact of logistics flows than a trade balance, which was not the proposal of this study. Therefore, we believe this study is valid and contributes to an analysis of Brazilian logistics and to interpretation of the international connections with the aim to create policies for diversification of logistical corridors providing more profitability to growers and to the government and population via international trade taxes.

We suggest the use of the same method on different products for further studies, to see if this brings similar insights. It also can be used to better understand a single product. Another approach could be the use of cross-section analysis or the comparison of these results with previous years or decades. Finally, one could check if the results could be different using monetary values. Note that both can be used because monetary value and volume are not linear. A certain value received could represent a cargo moved in another year and vice-versa.

Author Contributions

Conceptualization, D.L.d.J.X., J.G.M.d.R., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C.; methodology, D.L.d.J.X., J.G.M.d.R., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C.; software, D.L.d.J.X., G.S.R. and P.F.d.C.C.; validation, D.L.d.J.X. and J.G.M.d.R.; formal analysis, D.L.d.J.X. and J.G.M.d.R.; investigation, D.L.d.J.X., J.G.M.d.R., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C.; resources, D.L.d.J.X., J.G.M.d.R., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C.; data curation, D.L.d.J.X.; writing—original draft preparation, D.L.d.J.X., J.G.M.d.R., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C.; writing—review and editing, D.L.d.J.X. and J.G.M.d.R.; visualization, D.L.d.J.X., J.G.M.d.R., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C.; supervision, J.G.M.d.R.; project administration, D.L.d.J.X. and J.G.M.d.R.; funding acquisition, D.L.d.J.X., A.H.I., A.C.D., G.S.R., J.S.d.S. and P.F.d.C.C. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Data Availability Statement

Data available upon request.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Vna text file.

Figure 2. Importer countries by importance in cargo volume (%).

Figure 3. Brazilian ports by importance in cargo volume (%).

Figure 4. Map of Brazilian ports. Source: [30].

Figure 5. Degree centrality network connecting countries (red nodes) and ports (blue nodes).

Figure 6. k-core.

Figure 7. Tie strength countries (red nodes) and ports (blue nodes).

Table 1

Centrality degree.

Actor Name	Degree	NrmDegree	%
PORT_OF_PARANAGUA	22	61.111	6.145
PORT_OF_SAO_LUIZ	22	61.111	6.145
PORT_OF_BELEM	21	58.333	5.866
PORT_OF_SANTOS	21	58.333	5.866
PORT_OF_MANAUS	17	47.222	4.749
PORT_OF_PORTO_DE_RIO_GRANDE	17	47.222	4.749
PORT_OF_BARCARENA	16	44.444	4.469
PORT_OF_SAO_FRANCISCO_DO_SUL	15	41.667	4.190
PORT_OF_VITORIA	15	41.667	4.190
PORT_OF_SANTAREM	13	36.111	3.631
Egypt	10	27.778	2.793
Morocco	10	27.778	2.793
Spain	10	27.778	2.793
Algeria	9	25.000	2.514
Iran	9	25.000	2.514
Japan	9	25.000	2.514
China	8	22.222	2.235
Saudi_Arabia	8	22.222	2.235
South_Korea	8	22.222	2.235
Vietnam	8	22.222	2.235
Colombia	7	19.444	1.955
Malaysia	7	19.444	1.955
Netherlands_(Holand)	7	19.444	1.955
Portugal	7	19.444	1.955
Bangladesh	6	16.667	1.676
Dominican_Republic	6	16.667	1.676
Ireland	6	16.667	1.676
Israel	6	16.667	1.676
Mexico	6	16.667	1.676
United_States	6	16.667	1.676
Venezuela	6	16.667	1.676
Indonesia	5	13.889	1.397
Italy	5	13.889	1.397
Guatemala	4	11.111	1.117
Jordan	3	8.333	0.838
UK	2	5.556	0.559
India	1	2.778	0.279
Network Centralization = 36.19%

Table 2

Ports outdegree.

Pos	Actor Name	Outdeg	Outdeg Volume	nOutdeg	Volume
1	PORT_OF_SANTOS	21	136,155,406,336	0.1358	46.4%
2	PORT_OF_PARANAGUA	22	40,576,499,712	0.0405	13.8%
3	PORT_OF_BELEM	21	24,558,422,016	0.0245	8.4%
4	PORT_OF_SAO_LUIZ	22	21,694,980,096	0.0216	7.4%
5	PORT_OF_VITORIA	15	15,140,585,472	0.0151	5.2%
6	PORT_OF_SAO_FRANCISCO_DO_SUL	15	14,871,945,216	0.0148	5.1%
7	PORT_OF_PORTO_DE_RIO_GRANDE	17	13,077,299,200	0.0130	4.5%
8	PORT_OF_MANAUS	17	11,531,336,704	0.0115	3.9%
9	PORT_OF_SANTAREM	13	10,253,905,920	0.0102	3.5%
10	PORT_OF_BARCARENA	16	5,735,461,888	0.0057	2.0%

Table 3

Importer countries indegree.

Pos	Actor Name	Indeg	Indeg Volume	nIndeg	Volume
1	Iran	9	44,369,219,584	0.0443	15.1%
2	Japan	9	32,002,746,368	0.0319	10.9%
3	China	8	31,351,230,464	0.0313	10.7%
4	Vietnam	8	30,467,119,104	0.0304	10.4%
5	Egypt	10	23,318,702,080	0.0232	7.9%
6	South_Korea	8	23,044,272,128	0.0230	7.8%
7	Spain	10	20,430,151,680	0.0204	7.0%
8	Malaysia	7	12,081,029,120	0.0120	4.1%
9	Saudi_Arabia	8	8,704,136,192	0.0087	3.0%
10	Indonesia	5	7,316,413,440	0.0073	2.5%
11	Mexico	6	6,507,229,184	0.0065	2.2%
12	Morocco	10	6,453,629,440	0.0064	2.2%
13	Netherlands_(Holand)	7	6,071,270,912	0.0060	2.1%
14	Algeria	9	5,686,460,928	0.0057	1.9%
15	Bangladesh	6	5,559,529,472	0.0055	1.9%
16	Colombia	7	4,852,723,200	0.0048	1.7%
17	Portugal	7	4,462,600,192	0.0044	1.5%
18	Dominican_Republic	6	4,263,524,608	0.0042	1.5%
19	Venezuela	6	2,640,897,536	0.0026	0.9%
20	Italy	5	2,535,391,488	0.0025	0.9%
21	India	1	2,197,663,232	0.0022	0.7%
22	Israel	6	2,126,153,216	0.0021	0.7%
23	Ireland	6	2,119,811,840	0.0021	0.7%
24	United_States	6	1,669,986,816	0.0017	0.6%
25	Jordan	3	1,619,229,184	0.0016	0.6%
26	Guatemala	4	1,003,094,656	0.0010	0.3%
27	UK	2	741,621,120	0.0007	0.3%

References

1. United Nations. World Population Prospects 2022: Summary of Results; United Nations: New York, NY, USA, 2022.

2. FAO. Faostat. Available online: https://www.fao.org/faostat/en/#country/21 (accessed on 24 March 2023).

3. Hernandez, J.E.; Mortimer, M.; Panetto, H. Operations management and collaboration in agri-food supply chains. Prod. Plan. Control; 2021; 32, pp. 1163-1164. [DOI: https://dx.doi.org/10.1080/09537287.2020.1796141]

4. Esteso, A.; Alemany, M.M.E.; Ortiz, A. Impact of product perishability on agri-food supply chains design. Appl. Math. Model.; 2021; 96, pp. 20-38. [DOI: https://dx.doi.org/10.1016/j.apm.2021.02.027]

5. Cao, Y.; Tao, L.; Wu, K.; Wan, G. Coordinating joint greening efforts in an agri-food supply chain with environmentally sensitive demand. J. Clean. Prod.; 2020; 277, 123883. [DOI: https://dx.doi.org/10.1016/j.jclepro.2020.123883]

6. Ciccullo, F.; Cagliano, R.; Bartezzaghi, G.; Perego, A. Implementing the circular economy paradigm in the agri-food supply chain: The role of food waste prevention technologies. Resour. Conserv. Recycl.; 2021; 164, 105114. [DOI: https://dx.doi.org/10.1016/j.resconrec.2020.105114]

7. Hu, J.y.; Zhang, J.; Mei, M.; Yang, W.m.; Shen, Q. Quality control of a four-echelon agri-food supply chain with multiple strategies. Inf. Process. Agric.; 2019; 6, pp. 425-437. [DOI: https://dx.doi.org/10.1016/j.inpa.2019.05.002]

8. Zaw, T.N.; Lim, S. The military’s role in disaster management and response during the 2015 Myanmar floods: A social network approach. Int. J. Disaster Risk Reduct.; 2017; 25, pp. 1-21. [DOI: https://dx.doi.org/10.1016/j.ijdrr.2017.06.023]

9. Procopiuck, M.; Frey, K. Redes de políticas públicas e de governança e sua análise a partir da websphere analysis. Rev. Sociol. E Política; 2009; 17, pp. 63-83. [DOI: https://dx.doi.org/10.1590/S0104-44782009000300006]

10. Wellman, B. Structural analysis: From method and metaphor to theory and substance. Social Structures: A Network Approach; Structural Analysis in the Social Sciences Cambridge University Press: New York, NY, USA, 1988; Volume 2, pp. 19-61.

11. Klijn, E.H.; Koppenjan, J.F.M. Public Management and Policy Networks. Public Manag. Int. J. Res. Theory; 2000; 2, pp. 135-158. [DOI: https://dx.doi.org/10.1080/14719030000000007]

12. Farine, D.R.; Whitehead, H. Constructing, conducting and interpreting animal social network analysis. J. Anim. Ecol.; 2015; 84, pp. 1144-1163. [DOI: https://dx.doi.org/10.1111/1365-2656.12418] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26172345]

13. Luo, Q.; Zhong, D. Using social network analysis to explain communication characteristics of travel-related electronic word-of-mouth on social networking sites. Tour. Manag.; 2015; 46, pp. 274-282. [DOI: https://dx.doi.org/10.1016/j.tourman.2014.07.007]

14. Loughead, T.M.; Fransen, K.; Puyenbroeck, S.V.; Hoffmann, M.D.; Cuyper, B.D.; Vanbeselaere, N.; Boen, F. An examination of the relationship between athlete leadership and cohesion using social network analysis. J. Sport. Sci.; 2016; 34, pp. 2063-2073. [DOI: https://dx.doi.org/10.1080/02640414.2016.1150601] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26902840]

15. Cooc, N.; Kim, J.S. Peer influence on children’s reading skills: A social network analysis of elementary school classrooms. J. Educ. Psychol.; 2017; 109, pp. 727-740. [DOI: https://dx.doi.org/10.1037/edu0000166]

16. Yip, W.S.; To, S. Identification of stakeholder related barriers in sustainable manufacturing using Social Network Analysis. Sustain. Prod. Consum.; 2021; 27, pp. 1903-1917. [DOI: https://dx.doi.org/10.1016/j.spc.2021.04.018]

17. Castells, M.; Cardoso, G. Do Conhecimento à Acção Política. A Sociedade em Rede; Paz e Terra: São Paulo, Brazil, 1999; 47.

18. Apostolato, I.A. An overview of Software Applications for Social Network Analysis. Int. Rev. Soc. Res.; 2013; 3, pp. 71-77. [DOI: https://dx.doi.org/10.1515/irsr-2013-0023]

19. Borgatti, S.P.; Everett, M.G.; Johnson, J.C. Analyzing Social Networks; SAGE: Los Angeles, CA, USA, 2013.

20. Vignery, K.; Laurier, W. A methodology and theoretical taxonomy for centrality measures: What are the best centrality indicators for student networks?. PLoS ONE; 2020; 15, e0244377. [DOI: https://dx.doi.org/10.1371/journal.pone.0244377]

21. Tabassum, S.; Pereira, F.S.F.; Fernandes, S.; Gama, J. Social network analysis: An overview. WIREs Data Min. Knowl. Discov.; 2018; 8, e1256. [DOI: https://dx.doi.org/10.1002/widm.1256]

22. Dey, P.; Chaterjee, A.; Roy, S. Influence maximization in online social network using different centrality measures as seed node of information propagation. Sādhanā; 2019; 44, 205. [DOI: https://dx.doi.org/10.1007/s12046-019-1189-7]

23. Vega-Oliveros, D.A.; Gomes, P.S.; Milios, E.E.; Berton, L. A multi-centrality index for graph-based keyword extraction. Inf. Process. Manag.; 2019; 56, 102063. [DOI: https://dx.doi.org/10.1016/j.ipm.2019.102063]

24. Kong, Y.X.; Shi, G.Y.; Wu, R.J.; Zhang, Y.C. k-core: Theories and applications. Phys. Rep.; 2019; 832, pp. 1-32. [DOI: https://dx.doi.org/10.1016/j.physrep.2019.10.004]

25. Saqr, M.; Fors, U.; Nouri, J. Using social network analysis to understand online Problem-Based Learning and predict performance. PLoS ONE; 2018; 13, e0203590. [DOI: https://dx.doi.org/10.1371/journal.pone.0203590] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30235227]

26. Brazilian Trade Foreign Portall. Comexstat. Available online: http://comexstat.mdic.gov.br/pt/home (accessed on 21 March 2023).

27. Borgatti, S.P. NetDraw Software for Network Visualization; Analytic Technologies: Lexington, KY, USA, 2002.

28. Martinelli, L.A.; Batistella, M.; Silva, R.F.B.d.; Moran, E. Soy Expansion and Socioeconomic Development in Municipalities of Brazil. Land; 2017; 6, 62. [DOI: https://dx.doi.org/10.3390/land6030062]

29. Watai, R.; Ruggeri, F.; Tannuri, E.; Santos, N.; Tavares, B.; dos Santos, J. An analysis methodology for the passing ship problem considering real-time simulations and moored ship dynamics: Application to the Port of Santos, in Brazil. Appl. Ocean Res.; 2018; 80, pp. 148-165. [DOI: https://dx.doi.org/10.1016/j.apor.2018.08.012]

30. Maps Brasil. Brazil Ports Map. Available online: https://maps-brazil.com/maps-brazil-transports/brazil-ports-map (accessed on 31 March 2023).

31. Celik, S.; Corbacioglu, S. From linearity to complexity: Emergent characteristics of the 2006 Avian Influenza Response System in Turkey. Saf. Sci.; 2016; 90, pp. 5-13. [DOI: https://dx.doi.org/10.1016/j.ssci.2016.01.006]

32. Jugović, A.; Komadina, N.; Perić Hadžić, A. Factors influencing the formation of freight rates on maritime shipping markets. Pomorstvo; 2015; 29, pp. 23-29.

33. Martínez-López, B.; Perez, A.M.; Sánchez-Vizcaíno, J.M. Social Network Analysis. Review of General Concepts and Use in Preventive Veterinary Medicine. Transbound. Emerg. Dis.; 2009; 56, pp. 109-120. [DOI: https://dx.doi.org/10.1111/j.1865-1682.2009.01073.x]

34. Luz, J.; Reis, J.; Araújo Leite, F.; Araújo, K.; Moritz, G. Effects of Transport Infrastructure in the Economic Development. Proceedings of the Advances in Production Management Systems. Initiatives for a Sustainable World; Nääs, I.; Vendrametto, O.; Mendes Reis, J.; Gonçalves, R.F.; Silva, M.T.; von Cieminski, G.; Kiritsis, D. Springer International Publishing: Cham, Switzerland, 2016; pp. 633-640.

35. Bispo, S.Q.A. NT DINTE 26—China: Importação dos Principais Subsetores do Agronegócio e o Market-Share Brasileiro. Notas Técnicas; Ministério da Saúde: São Paulo, Brazil, 2021; pp. 1-25.

36. Cariello, T.; Fragoso, G.; Pimenta, J.U.R.; Nacif, G.A.L. Oportunidades de Comércio e Investimento na China. Alimentos e Bebidas; Technical Report CEBC: São Paulo, Brazil, 2018.