Soil mapping guides the definition of management zones.

The adoption of Precision Agriculture (PA) techniques can contribute to achieving the maximum productive potential of crops. In Brazil, the use of these tools is increasingly frequent. Currently, producers can purchase machines equipped with systems for mapping crop productivity, applying inputs at variable rates, autopilot, and controlling the sowing and spraying process. Companies providing technical assistance and consulting services have been using sensors in their surveys to identify weeds, plant stand failures, quantify nitrogen content, measure soil resistance to penetration, and the apparent electrical conductivity of the soil, with the goal of enabling their clients to produce more using fewer inputs, which ultimately reduces production costs and the environmental impact caused by agricultural activity.

Among the tools used by precision agriculture service providers, sensors that measure the apparent electrical conductivity (EC) of the soil deserve special mention. Several studies have already proven the relationship between variations in soil EC and variations in crop productivity. This behavior can be explained because the soil attributes (chemical and physical) that influence electrical conductivity are also responsible for variations in crop productivity; therefore, it can be said that soil EC serves as an indirect indicator of the spatial variability of soil attributes in an agricultural field.

Mapping variations in soil electrical behavior can help identify areas within a field that have different nutrient levels. These variations in electrical conductivity values ​​can be organized into a zone map, delineating areas of high and low values, thus establishing management zones.

According to Dr. Thomas A. Doerge, a precision agriculture agronomist at Pioneer Hi-Bred International, management zones “are sub-regions of the field that present a combination of limiting productivity factors for which the same rate of input application is appropriate” (Doerge, n.d.). Soil management by zones allows for a reduction in analysis costs, since the use of sampling grids is no longer necessary. In this way, the number of samples is reduced, requiring only one soil sample composed of simple samples, collected in each zone, for characterization.

Professor and researcher Eduardo Leonel Bottega, PhD, from the UFSM campus in Cachoeira do Sul, in partnership with Campear Engenharia Agronômica and Falker Automação Agrícola Ltda, conducted a study in a commercial area intended for grain cultivation using a no-till system. The objective was to identify which soil attributes the delimitation of management zones, based on EC measurement, was able to differentiate. 

The study was conducted in the municipality of Cachoeira do Sul (RS), in an area of ​​25,8 ha, irrigated with a center pivot. Data collection took place on November 13, 2018, after the wheat harvest. Soil EC readings were taken using the Terram commercial equipment, manufactured by Falker Automação Agrícola Ltda. 

The equipment consists of a chassis on which four straw cutting discs are mounted, equally spaced at 0,25 meters. The inner discs act as potential electrodes and the outer discs as current electrodes. The electrical conductivity of the soil is obtained from the inverse of the electrical resistivity of the soil, measured by the potential difference between the emitted and received current. The collected data were recorded in the Fieldbox system, manufactured by the same company. The equipment was towed across the area using an ATV, and 4.485 readings of the soil's electrical conductivity were obtained.

After the readings, the data were subjected to outlier analysis to eliminate discrepant values ​​resulting from occasional lack of contact between the discs and the soil. Subsequently, geostatistical analysis was performed to quantify the spatial variability of soil EC and produce a map indicating its variation. The thematic map of the spatial distribution of soil EC was produced using ordinary kriging (Figure 2a). After performing ordinary kriging, the data were grouped into two classes to produce the management zone map. The grouping was performed using the fuzzy k-means algorithm. The soil sampling point map was overlaid on the management zone map (Figure 2b), thus making it possible to identify points belonging to classes 1 and 2 (ZM 1 and ZM 2).

Soil samples were collected at a depth of 0,0–0,2 meters and sent for routine laboratory analysis. At each of the 24 sampling points, five individual samples were collected within a three-meter radius of the sampling point. The samples were mixed, and a composite sample was taken to characterize the soil at the respective point. A t-test (p < 0,05) was performed between the values ​​of soil attributes, based on the management zones where each sampling point was located.

Figure 3 shows the Box-Whisker plots of soil attribute values ​​that showed statistical differences as a function of the management zone to which they belong.

The management zones delimited based on the spatial variability of soil EC, measured with the Terram commercial equipment, were able to differentiate the following soil attributes: Clay, Potassium (K), Calcium (Ca), Sum of bases (SB), effective CEC (t) and CEC at pH 7 (T). It was observed that the management zone with the highest average value of soil EC also presented the highest averages for the values ​​of soil attributes, demonstrating the potential of EC mapping in differentiating, within the same plot, areas with differences in the values ​​of soil attributes.

In practice, these results could be used, for example, in making decisions regarding the plant population to be established in the area, with population density varying according to the clay content in each management zone (ZM). Another factor to highlight is soil sampling, which could be carried out to characterize each management zone, requiring only two analyses of composite soil samples, representative of each delimited ZM. This would reduce the cost of analyses compared to the traditional grid soil sampling method. The results obtained in this study demonstrated the promising use of soil EC as a delimiting factor for management zones; however, further studies are recommended to effectively consolidate this new soil fertility management tool.

*By Eduardo Leonel Bottega, Alicia Baumhardt Dorneles, Cristielle König Marin e Zanandra Boff De Oliveira (UFSM), Rodrigo Franco Dias e Eder Luis Sari (Campear Agricultural Engineering)