Application of Plant Densities in Management Units in the Soybean Cultivation

The application of management units (MU’s) aims to make economically viable to precision agriculture, making the technique accessible to a greater number of producers. Using MU’s, the experimental area is divided into plots with different productive potentials. In this context, the objective of the experiment was to verify the effectiveness of the area division in MU’s and to define the soybean plant density that provides higher productive efficiency in each MU. For the formation of MU’s it was used the altitude variation and the soil penetration resistance 0-0.1 m in the experimental area, being that the area was divided into 2 MU’s, called MU1 and MU2, and each MU was composed of 8 plots. At planting, 2 plant densities were applied, 214 000 and 257 000 plants ha, and each density was applied in 4 plots per MU, using row spacing of 0.70 m. In relation to productivity, there was a significant difference, applying the t-Student test, between MU’s, and the MU2, unit with higher productive potential, located in the highest part in the area, achieved higher productivity; and there was an effect, using the Tukey test, on the application of the 2 different plant densities in the MU’s, being that the densities of 214 000 and 257 000 plants ha reached, respectively, higher productivity in MU2 and MU1.


Introduction
Precision agriculture (PA) is still a technique that finds barriers to its application due to its operational and deployment cost of this technique.Thus, new ways of apply PA with lower sampling costs, reducing the amount of inputs and the definition of management units (MU's) (Roudier, Tisseyre, Poilve, & Roger, 2008;Suszek, Souza, Uribe-Opazo, & Nobrega, 2012) comes to make the PA a technique more practical and economic.The MU's are plots within the field that have similar characteristics and productive potential, being that each MU is susceptible to receive the same agronomic practice in all its extension (Schepers et al., 2004).For the determination of MU's different methods can be used with the use of soil attributes or crop parameters (Blackmore, 2000;Fraisse, Sudduth, & Kitchen, 2001), in addition to attributes attached to the relief (Zhu & Lin, 2011).
The relief, most of the time, assumes an important role in the determining of MU's, because the impact of the topography in the field is important to explain the yield variability of the crops (Kumhalova & Moudry, 2014).In several studies authors received positive results applying MU's in the productive area (Fleming, Westfall, Wiens, & Brodahl, 2000;Anuar, Goh, Tee, & Ahmed, 2008;Diacono et al., 2012).
The identification of MU's in the area allows the site-specific management of important cultural practices in the formation of the final productivity of crop, such as the adjustment in the plant density in sowing.Being that, Ribeiro et al. (2017) reports that plant density in soybean cultivation is an important component to increase grain yield, thereby reducing production costs.
In the experiment the objective was to verify the effectiveness of the area division in MU's and to define between 2 selected soybean plant densities, the density that provides the highest productive efficiency in each MU and in the experimental area.jas.ccsenet.

Soil An
The delim 40 samplin physical a and after th The altitude variation in the area and the soil penetration resistance at depth 0-0.1 m were used as the basis for the configuration of the MU's, because showed higher spatial correlation with productivity in the experimental area in the analysis performed in the agricultural years 2011/2012 and 2012/2013 (Schenatto et al., 2016), the most recent period was also the soybean planting.Being that, the MU2 has the highest productive potential and was established in the highest region, while the MU1 is located in the lower part of the area.
It was used Moran's bivariate spatial autocorrelation statistic (Czaplewski & Reich, 1993) to evaluate the spatial correlation between the analyzed attributes and to establish the spatial correlation matrix, which makes it possible to analyze which attributes influence positively or negatively the productivity.In selecting the attributes used to generate the MU's: spatially correlated the attributes (chemical and physical soil attributes and area relief), eliminated the attributes less spatial correlation between themselves and between soybean yield (Bazzi, Souza, Uribe-Opazo, Nóbrega, & Rocha, 2013).

Plant Densities
In the agricultural year 2015/2016, 2 different plant densities were used in the area, being that the lowest density was suggested by the farmer and applied also in the agricultural years 2011/2012 and 2012/2013 in the total area.The highest density was defined in this work and is 20% higher than the lowest density.At planting each density was applied in 4 plots per MU.
Seeding rates (seeds ha -1 ) used in plots in the MU's were 10% higher than plant densities (plants ha -1 ) because the seed germination rate for Syngenta 1359 had 90% germination effectiveness.The seeding parameters are shown in Table 1.

Statistical Analysis
The statistical analysis were performed using Randomized block design (RBD), being that the treatments was represented by the plant densities and each MU was considered a block.The Tukey test was used to compare the productivity reached by the plant densities within of the MU's.For MU's, it was compared the yield reached by the MU's using t-Student test.

Planting, Harvesting and Productivity Estimate
The soybean seeds were planted on October 17, 2015 and the mechanized harvest was performed on February 25, 2016.
The estimate of the average productivity of the 2 MU's that would be reached using the most efficient plant density by each MU, among the 2 tested in the experiment, in the agricultural year of 2015/2016 was performed with the objective of determining the highest yield of the area.This average productivity was compared with the average achieved without the use of MU's in the agricultural years 2011/2012 and 2012/2013.

Analysis of Soil Attributes and Fertilizing
The analysis of soil attributes was carried out in the years of 2013 and 2015 (Table 2), there was no analysis of soil attributes in 2014, because the soil did not present adequate conditions for the drought presented in the period.
jas.ccsenet.The full seed stage corresponds from R5 to R7 stage, and the water requirement for the plants reaches the value of 9 mm d -1 (Steduto, Hsiao, Fereres, & Raes, 2012).In the agriculture year 2015/2016, there were a lot of days without rainfall in full seed stage, with 10 days without rainfall, this factor may have affected the soybean yield.

Conclusion
The division of the productive area in MU's considering the altitude variation and soil penetration resistance obtains success in the increase of productivity in the soybean cultivation.
The use of 2 plant densities provided inverse results according to the MU.In MU2, unit with higher productive potential, the higher plant density resulted in lower productivity, while in MU1, unit with lower productive potential, the effect was the opposite, resulting in higher productivity.
It was confirmed that the division of the area into 2 MU's, with the application of appropriate plant density to the productive potential of each MU, provides greater productivity in the cultivated area.

Table 1 .
Seeding parameters performed