Application of an original soil tillage resistance sensor in spatial prediction of selected soil properties

Highlights

• Implementation of novel three-point-hitch measuring system.

• Investigation of potential of new measuring system for better soil observation.

• Utilization of geostatistics for maps generation of soil physical properties.

• Study of interactions between soil tillage resistance and standard soil properties.

Abstract

Horizontal and vertical variability of soil properties, temporal dynamics, and the complexity of monitoring and prediction processes are the limiting factors for full implementation of all the postulates of site specific crop management. The aim of this research was to prove the possibilities for utilization of original soil tillage resistance measuring device in spatial assessment of some physical properties of soil.

The tillage resistance measurement and soil evaluation were performed in 2012 on a field of 0.78 ha, with calcic chernozem type of soil. The year before, the field was divided into three equal plots where three types of tillage were applied (moldboard plowing, disc harrowing and chisel plowing) in order to increase spatial soil differences and to ensure more reliable testing. Soil physical properties were observed at 30 points that were arranged on the nodes of a rectangular grid all over the field along with their geopositioning. Soil tillage resistance was measured in real time and space, during 30 passes that were made on the entire field by a standard tractor and plow.

The tillage management applied in 2011 resulted in the differences of some soil parameters. Significant differences between the field plots arose with respect to cone index, moisture content and soil tillage resistance. However, no differences arose regarding the texture (sand, silt + clay content) and bulk density. The correlation analysis showed a high positive correlation between soil resistance and cone index (R = 0.82) and a negative correlation between soil resistance and moisture content (R = −0.41). Fitting of variograms in the modeling of spatial correlation gave the highest prediction error for cone index (13.9%), then for tillage resistance (9.32%), sand content (3.95%), moisture content (3.34%), bulk density (2.4%), and finally clay + silt content (2.2%). The generated maps showed that the soil compaction, moisture and tillage resistance had certain orientations with respect to the field plots.

Introduction

Modern methods in crop production aim to achieve maximum productivity per unit of input and used surface area with minimal negative consequences on the sustainability of a biosystem. In order to realize that, all production steps need to be equally optimized on all parts of the field taking into consideration the specific requirements of the produced crop (Hemmat et al., 2008). It is not an easy task according to heterogeneity of soil that is described in numerous studies (Adamchuk et al., 2001, Lapen et al., 2002, Hanquet et al., 2004, Hemmat et al., 2008). Spatial variability of soil is in relation to the type of soil, topography and applied in-field management. The tendency of enlarging the fields by simply erasing the previous borders that used to divide naturally different types of soil contributes to the increased in-field variability (Oliver, 2010). The concept of “fertilization of plants instead of soil”, i.e., banding of fertilizers in the crop row, additionally increases variations in the nutrient content of soil (Hu et al., 2014). Therefore, in order to implement site-specific soil management, a sufficient amount of valid information about spatial variability is necessary for good decision making regarding every part of the field (Jabro et al., 2010, Hanquet et al., 2004). The quality of information depends on the applied method of data collection. Most of the standard methods used in the determination of soil conditions are spatially discontinuous, time consuming, and expensive, and they provide little accurate data (Oliver and Webster, 2014). The current trend in the advancement of systems for field data collection is the development of “on-the-go” measurement systems which are equipped with various sensors for proximal detection (Adamchuk et al., 2004). Considering the fact that those methods do not offer high accuracy data, proximal measurement systems are used for fast collection of vast geodata. In their research a great number of authors use the systems for measurement of soil mechanical resistance in order to assess the variabilities in soil physical conditions (Rockström et al., 1999, Gaston et al., 2001, Raper et al., 2005, Chung et al., 2006, Topakci et al., 2010). Van Bergeijk et al. (2001) indicated that soil tillage resistance could provide information about the locations of different types of soil. Lapen et al. (2002) concluded that if there is a relation between tillage resistance and soil compaction in the field, then the map of soil resistance can be used for the mapping of soil compaction. Understanding and modeling the variabilities in soil properties has less significance for precision agriculture unless they have some correlation with key parameters such as the soil moisture, soil type, and soil fertility. The maps of tillage resistance can be used as a criterion to locate the critical areas which require special treatments of site-specific management (Hanquet et al., 2004).

The objective of this research was to test the possibility of applying a new measuring system in spatial assessment of some soil physical properties like texture, bulk density, moisture content and cone index as a step further in the device development. The research is basically the extension of the development process of a system for soil resistance measurement according to the principles of precision agriculture (Kostić et al., 2014).