Spatial statistics applied to soybean production data from Paraná State for 2003-04 to 2009-10 crop-years

Abstract

In the current study, we performed a soybean production spatial distribution analysis in Paraná State. Seven crop-year data, from 2003-04 to 2009-10, obtained from the Paraná Department of Agriculture and Supply (SEAB) were used to develop a Boxmap for each crop-year, show soybean production throughout this time interval. Moran's index was used to measure spatial autocorrelation among municipalities at an aggregate level, while LISA index local correlation. For each index, different contiguity matrix and order were used and there was a significance level study. As a result, we have showed spatial relationship among cities regarding the production, which allowed the indication of high and low production clusters. Finally, identifying main soybeanproducing cities, what may provide supply chain members with information to strengthen the crop production in Paraná. . Elaborou-se um Boxmap para cada ano-safra, com o intuito de demonstrar como ocorreu a alternância dos valores de produção de soja neste intervalo de tempo. O Índice de Moran foi utilizado para apresentar os níveis de autocorrelação espacial a nível global entre os municípios e o índice de autocorrelação LISA, para estudar a autocorrelação espacial a nível local, com diferentes matrizes de contiguidade e com diferentes ordens de contiguidade, assim como o estudo dos níveis de significância. Através destes métodos, foi possível demonstrar como os municípios se relacionam espacialmente quanto à produção da soja, o que permitiu a indicação de clusters de altabaixa produção, possibilitando assim que se conheçam os principais municípios produtores de soja, subsidiando os membros da cadeia produtiva com informações para melhorar seus planos no fortalecimento da cultura no Paraná.

Analysis of Spatial Autocorrelation of Grain Production and Agricultural Storage in Paraná

“…The West mesoregions (comprising the municipalities of Cascavel, Toledo, Assis Chateaubriand and Terra Roxa), Western Center (comprising the municipalities of Ubiratã, Campo Mourão and Goioerê) and Eastern Center (comprising the municipalities of Castro, Tibagi and Ponta Grossa) showed clusters of municipalities with high production (red color in Figure 4). Prudente et al (2014) used the LISA to analyze the value of the spatial autocorrelation of soybean production for each municipality of Paraná through the LISA Cluster Map and found clustering values with high and low soybean production with a level of 5% significance in the studied harvest.…”

“…Grzegozewski et al (2017) analyzed the spatial variability of soybean production and agro-meteorological variables in the state of Paraná through the Moran (I) global autocorrelation index and verified different sowing periods between the regions and the great climatic variability in the state. The soybean producing regions were spatially associated in the production interval in the 2003/2004 to 2009/2010 harvests in the state of Paraná, where Prudente et al (2014) studied them from the global spatial autocorrelation. Araújo et al (2013) applied the bivariate spatial autocorrelation analysis in spatial groupings of soybean production in the state of Paraná and identified the formation of municipalities groups, through the similarity of the variables under analysis.…”

Spatial and Multivariate Analysis of Soybean Yield in the State of Paraná-Brazil

“…Soybeans are used in agribusiness (vegetable oil production, animal feed), in the food industry (flour, extracts, beverages, sprouts), and in the chemical industry (adhesives, coatings, water emulsion paper for paints) (Bergamaschi, 1992;Berlato, 1999;Prudente et al, 2014).…”

“…Studies using these indicators have shown for the state of Paraná : the soybean crop profile at different seeding dates (Dalposso et al, 2013, Cima et al, 2018; how municipalities relate spatially to soybean production; which are the main municipalities producing soy (Prudente et al, 2014); and the analysis of the spatial relationship of soybean yield with agrometeorological characteristics (Grzegozewski et al, 2017) In this context, regarding the simultaneous analysis of a set of variables, the multivariate analysis techniques can be helpful in finding patterns generated by the set of variables. In the Western region of Paraná , Araújo et al (2013) carried out a cluster analysis using information on the local Moran index (LISA) applied to agrometeorological and soybean yield data in crop year 2005/2006, and the formation of groups of similar municipalities was identified regarding their spatial distribution in relation to soybean yield and all agrometeorological elements analyzed.…”