SPATIAL DISTRIBUTION, EVOLUTION AND STRUCTURE OF MAIZE AND SOYBEAN PRODUCTION SYSTEMS IN STATE OF PARANÁ

The Brazilian state of Paraná is one of the leading producers of maize. However, expansion of soybean cropping has caused a drop in maize production and could have impacted production systems. The aim of the study was to verify the evolution and identify the structure, spatial dynamics and transformation of maize and soybean production systems in the state of Paraná. Municipal Agricultural Production data from the Brazilian Institute of Geography and Statistics (IBGE) provided the basis for this study. The Location Quotient was analyzed to identify microregions specialized in the production of soybean and first and second crops of maize. Principal components and groups were analyzed in order to characterize the structure and dynamics of maize and soybean production systems in specialized microregions. The results show how maize and soybean production systems have been transformed; soybean is now cropped in areas previously occupied by first crop maize, and the area occupied by second crop maize has increased. This has led to the predominant use of the crop rotation system with first crop soybean followed by second crop maize.


2
Food supply and food security pose major challenges for the 21 st century, and Brazil plays a crucial role, with its capacity to expand agricultural production (Freitas & Mendonça, 2016). A key factor of this expansion is the intensification of areas cropped with soybean, especially since areas previously occupied by other crops, including first crop maize, have been turned over to soybean (Melo et al., 2012), evidence of competition for land (Caldarelli & Bacchi, 2012). However, regions with the highest concentration of soybean also have the highest concentration of maize (Dias et al., 2016). in Paraná has added value to various agricultural and livestock production chains, such as poultry, pork and dairy, which, according to Martin et al. (2011), consume large quantities of maize.
Second crop maize is grown by dry farming after the first crop and allows optimization of farm labor and machinery, reducing the impact of seasonality on production, supply and prices (Tsunechiro et al., 2006). Since the climatic conditions in second growing season are less favorable, this cropping system has lower production potential than the first growing season.
Another factor that has contributed to the growth of the maize cropping is the adoption of the notill system, where the crop is planted directly into the soybean cover, cutting the time between harvesting the first crop and sowing the second crop (Bicudo et al., 2009;Albrecht et al., 2009).
With regard to soybean, according to MAP data, state of Paraná was responsible for 17.68% of total Brazilian production in 2016. This result was obtained due to the dynamics of the soybean production chain, consisting of various stages: grain (agricultural production), brans and vegetable oils and oils related to input supply for upstream links (Caldarelli et al., 2009).
Maize production seems to be directly linked to soybean production. In Paraná, the expansion of soybean cropping raises a number of questions concerning maize production, such as: What changes have been brought about by the spatial interdependency between the maize and soybean crops? How much has first cropr maize lost in terms of area? By how much has the second crop maize area increased? Has production increased or decreased? These questions provide a basis for examining a possible spatial interdependency between maize and soybean production.
Due to the dynamics and complexity in Paraná, since spatial changes in production will depend on the provision of adequate infrastructure, a market for the produce, credit programs and technical support (research and extension).
Therefore, the aim of this study was to verify the evolution and identify the structure, spatial dynamics and transformations of maize and soybean production systems in the state of Paraná.

Material and Methods
The study was based on MAP data published by the IBGE, and data on the National Family  (Greene, 2008). This rate of variation was expressed as a percentage, since the harvested area and the quantity produced verified the acceptability of the assumptions measured by the Student's t-test, at a significance of 5%. The data series for analyzing the maize crop were defined as a function of the period during which production was split between the first and second crops, and during the years compatible with soybean cropping periods. Thus, for soybean there were four periods (1997 to 2001, 2002 to 2006, 2007 to 2011 and 2012 to 2016), and for maize two periods (2007 to 2011 and 2012 to 2016).
the dependent variable relating to the harvested area, quantity produced and productivity in the i th year; µ random error; and i the number of years.
Next the location quotient (LQ) was calculated. It indicates the region's specialization in maize and soybean cropping. Based on the basic aggregate, LQ is used to measure and compare regional specialization for a given activity. This parameter was applied to estimate the specialization in first and second crop maize, and soybean for the microregions of Paraná, based on Gross Production Value (GPV) averages for the period. In accordance with Equation 2 (Isserman, 1977), LQ was obtained based on the proportion between the GPV for each crop and the agricultural activity GPV, which includes permanent and temporary farming. (2) Where is the GPV for agricultural activity in region ; is the total GPV of all agricultural activities in region is the GPV of agricultural activity in all regions; and is the total GPV of all activities in all regions.
In other words, the numerator corresponds to the way in which production for an agricultural activity is split in the microregion and the total agricultural activity in the same microregion.
Similarly, the denominator corresponds to the split in production for a given agricultural activity in Paraná and the total agricultural production in Paraná. If the result obtained is greater than 1, the activity is a specialization of the microregion.
Thus, non-specialized microregions have an LQ < 1, and specialized regions an LQ ≥ 1. Note that, for data availability reasons, the LQs for first and second crop maize were based on data from 2007 to 2016.
Once the microregions specialized in growing first and second crop maize and soybean had been defined, principal component analysis and is not correlated with the first, and so on (Fávero & Belfiore, 2015). The components were defined based on the Kaiser criterion, selecting components with eigenvalue ≥ 1 (Kaiser, 1960).
PCA was performed based on the following variables: LQ, credit earmarked for agricultural activity (PRONAF, PRONAMP and CSV), and number of cooperatives. Based on the results of PCA, cluster analysis was performed taking into account specialized microregions and using the Ward method (hierarchical clustering). This is an interdependent statistical technique allowing variables to be marshaled into homogeneous groups, according to a similarity or distance measurement (Fávero & Belfiore, 2015).
Monetary restatement was based on the extended national consumer price index (IPCA) for December 2017. SPSS software was used to process the data and ArcGIS 10.2 to produce maps. Table 1 gives the results relating to changes in harvested area, quantity produced and productivity have been adopted in this mesoregion, helping to manage these natural resources.

Results and Discussion
In terms of first crop maize productivity, the Northwest mesoregion had the worst indices, well below the average for Paraná.
However, the West Central, East Central, West and South-Central mesoregions showed the best productivity indices, with an average higher than the average for state of Paraná.
Productivity gains reflected the efficiency and intensification of productivity, i.e. higher yield per unit area.
Between 2012 and 2016, there was a 5% increase in the first crop maize harvested area in Paraná. In this period, the Southwest and Southeast mesoregions showed higher growth, at 87.9% and 58.5% respectively. However, the West and North Central mesoregions made the highest contributions in terms of harvested area, at around 37% and 25% respectively. This increase is related to the pursuit by rural producers of alternatives to obtain higher economic profitability, one of the options being to crop maize after harvesting soybean, i.e. adopting a soybean-maize as crop rotation system (Bicudo et al., 2009;Albrecht et al., 2009). Table 1. Changes in productivity, harvested area and quantity produced for maize in Paraná mesoregions between 2012 and 2016.   The West mesoregion was the main producer, accounting for over 40% of total yield in Paraná.

State and
Note that the maize produced in the West region is used mainly in poultry and pig breeding, to satisfy demand from the region's agroindustrial facilities (Martin et al., 2011). Maize production for animal feed has added value to the produce of this region. Furthermore, over this period there was an increase in average second crop of maize productivity in Paraná. This was due, in part, to research on improving and developing seeds,     is an essential input for animal feed (Alves et al., 2009    The State of Paraná is one of the largest producers and exporters of soybean in Brazil, and has ample installed infrastructure to cope with logistics, storage and processing of soybean (Caldarelli et al., 2009). Between 1997 and 2016, production of soybean in Paraná rose by 43.34%, with annual average growth of 4.98%.
In 2016, the West mesoregion produced the most soybean, accounting for around 21% of Paraná's total soybean production. This increase in production is related to growth in global demand for soybean, exportation and a rise in the price of soybean and derivatives, increasing returns for producers and thus driving expanded production  and livestock cooperatives that buy soybean for processing into other products or for sale to trading companies, manufacturers or international buyers.
Furthermore, the fact that soybean is easy to trade explains why farmers produce this commodity, and the expansion of cropping areas.
Principal component analysis (PCA) was applied to identify three main components that together accounted for 84% of the variability in the data (Table 3)   Group 5 is the microregion of Toledo.
The region is known for its high degree of specialization in the production of soybean and second crop maize (highest producer in Paraná).
It accounts for 10.55% of all rural credit raised by farmers to finance soybean and maize crops, with most financing raised through the PRONAF program. Toledo is also known for pig and chicken production. According to data in the Municipal