INTRODUCTION
Currently, the growth of agriculture has intensified the use of natural resources in general, and particularly, it has accelerated many processes of soil degradation, which has adversely influenced crops.
Among the factors that have the greatest impact on crop yields, those related to soil properties stand out. An important soil variable is mechanical resistance, a characteristic that interacts with other soil properties such as bulk density, texture, moisture content and porosity (Saffih et al., 2009).
Compaction, measured through the mechanical resistance offered by the soil, is produced by the traffic of agricultural equipment, as well as by the inadequate management of tillage operations and the action of rainfall on the bare soil, among others causes, constitutes a detrimental effect on agricultural yields (Botta et al., 2002, 2007, Rodriguez & Valencia, 2012 and Olivet & Cobas, 2013).
Arvidsson et al. (2004) suggest that, depending on the type of soil and texture, soil moisture should be below the plastic limit (LP) during agricultural work, coinciding with several authors (Mueller et al., 2003) and Barzegar et al. (2004), that the most appropriate moisture content for mechanized agricultural work corresponds to 0.7 - 0.9 LP.
Numerous researchers and manufacturers have developed sensors for continuous (On-the-Go) measurement of soil properties (Adamchuk et al., 2004; Hall & Raper, 2005; Herrera et al., 2011 and Hemmat et al., 2013). Based on the measurement methods, a wide variety of prototypes of probes have been developed, however, in all cases probes of different geometries have been used, without referring to which is the optimal one.
Likewise, Johnson (2003), Chung et al. (2004), Chung & Sudduth (2006) and Nader et al. (2013) developed different models of the probe-soil interaction, aimed at clarifying this interaction process and to relate the reading of the probes to the level of compaction of the soil under different humidity conditions. Some of these models could not be validated under different soil conditions and others have the drawback that they do not take into account some geometric parameters such as the length and areas of the lateral faces of the probes.
Other investigations have been aimed at detecting the degree of soil compaction using non-invasive methods, based on the application of electromagnetic fields to the soil (Martínez et al., 2010, 2011), however, they could not be applied due to the simultaneous influence of factors such as moisture and organic matter content of the soil.
In the particular case of the present study, a number of experimental investigations have been conducted by Chukwu & Bowers (2005), Hall & Raper (2005), Chung et al. (2006), Chung & Sudduth (2006) and Sharifi & Mohsenimanesh (2012), in order to obtain the probe of the best correlation with the apparent density and resistance to penetration. However, until now there is no precise information on the technical requirements for the design of probes, hence the objective of this investigation was to determine the type of probe that ensures possible levels of correlation between apparent density and resistance to penetration under different humidity conditions for a leached Red Ferrallitic soil, typical of Cuba.
METHODS
The experimental investigations were carried out at the Soil Channel Laboratory, at the Agricultural Mechanization Center (CEMA) of the Faculty of Technical Sciences at the Agrarian University of Havana (UNAH), located in San José de la Lajas Municipality at Mayabeque Province.
The soil under study is a Red Leachate Ferrallitic soil according to the latest classification in force in the country (Hernández et al., 2015), from the agricultural area of San José de las Lajas, in the Province of Mayabeque, Cuba, with a plasticity index of 30.4%, plastic limit of 30.7% and 3.01% of organic matter (González, 2008).
Design and Construction of Probes of Different Geometric Shapes
To carry out the experiments, eight types of probes were designed, five with a cone shape and three with a wedge shape. Their geometric characteristics are shown in Table 1.
Probe and stem lengths were kept constant in all cases. The cone with an angle of 30º and a base area of 130 mm2 corresponds to the standard cone of the American Society for Agricultural and Biological Engineering (ASABE). As material for the construction of the probes, 1045 steel was used according to the standard of the American Iron and Steel Institute (AISI) and a surface finish of 0.32 µm was applied.
TABLE 1.
Geometric characteristics of probes subject to experimentation
Sample Preparation and Measurement of Soil Penetration Resistance, Density and Moisture
The soil samples for each experimental variant were placed in seven metal boxes of uniform dimensions (Fig. 1). They were weighed to achieve a uniform amount of soil in each tank.
FIGURE 1.
Dimensions of the soil tanks.
The determination of the humidity and bulk density of the soil was carried out according to NC 67: (2000). The weighing of the samples before and after drying was carried out with a College electronic balance with an accuracy of 0.01 g. In addition, a grid (Fig. 2) was used in order to unify the taking of samples and three samples of soil were taken with Kopecki cylinders from each container following the diagonal and away from the edges.
The soil previously deposited in the boxes was compacted with a mechanical press until all the boxes had similar levels of humidity and apparent density. The boxes were divided into 25 quadrants to take the samples (Fig. 2 a). The three red dots indicate the grids for sampling bulk density and moisture content. In the rest of the squares, the penetration resistance obtained with the standard ASABE cone (green dots) was measured and also the penetration resistance obtained with the other geometries (blue dots).
The penetration resistance was determined using the CEMA-08 durometer (Fig. 2 b), designed to withstand loads up to 3 kN with an appreciation of 2 N. The penetration resistance was calculated by dividing the penetration force, observed on the digital indicator of the durometer, by the area of the cone base or wedge undergoing experimentation. The dimensions of the cone base or wedge were measured with a micrometer with an appreciation of up to 0.01 mm.
The determination of the cone index (ASABE standard) of the soil was measured with a FIELDSCOUT Digital Penetrometer Model: SC 900 SN: 328, with an appreciation of ± 1,25 cm, ± 15 PSI (± 103 kPa).
FIGURE 2.
a) Scheme for taking samples of resistance to penetration, bulk density and humidity. b) CEMA-08 Durometer.
Experimental Design
Soil moisture and bulk density were defined as independent variables, taking two moisture levels: intermediate (28%) and high (35%) and three levels of bulk density (1, 1.1 and 1.2 g cm-3), resulting in a 2x3 factorial design for a total of six treatments to be performed during the experimental runs.
For each treatment, three repetitions were performed. The maximum and minimum levels of bulk density were selected, from a pre-experiment, in order to reach values of resistance to penetration in a wide range between 0.5 and 6 MPa. As a dependent variable, the resistance to penetration measured with the probes of different geometric shapes was established.
RESULTS AND DISCUSSION
Table 2 shows the results of experimental investigations aimed at determining the relationships between resistance to penetration, measured with probes of different geometric dimensions, with the cone index ASABE and with the dry bulk density of the soil.
TABLE 2.
Relationship between penetration resistances obtained with probes of different geometry with ASABE cone index and dry volumetric density, for two moisture levels
*Leyenda para el grado de relación entre las variables: FF-significativamente fuerte (R2 ≥ 0,90); F-fuerte (R2 ≥ 0,80); MF- medianamente fuerte (R2 ≥ 0,60); D- débil (R2 ≤ 0,59).
* Legend for degree of correlation: FF-significantly strong (R2 ≥ 0.90); F-strong (R2 ≥ 0.80); moderately strong MF- (R2 ≥ 0.60); D- weak (R2 ≤ 0.59).
As it can be seen in the previous table, most of the geometries tested presented a strong degree of correlation with the ASABE cone index for the two humidity levels experienced, with the exception of the wedge with a base area of 260 mm2 at humidity of 35% and the wedge of 130 mm2 at humidity of 28%, which presented a moderately strong degree of correlation. The 60o cone, whose degree of correlation was weak for humidity of 35%, was also an exception.
The plot of the experimental points is shown in Fig. 3, as well as the line of best fit for the case of the cone and the wedge with angles of 30º and base area 520 mm2, for which the strongest degrees of correlation with the cone index ASABE were obtained for both humidity levels, resulting determination coefficients between R2 = 0.93 and R2 = 0.95.
FIGURE 3.
Regression equation and coefficient of determination between the wedge indexes (IW) obtained with a wedge of 30ox520 mm2 and the cone index (IC) ASABE for each of the moisture levels under study.
The correlation analyzes between the resistance to penetration, obtained with the different probes, and the apparent density, presented a more different behavior, obtaining strong degrees of correlation only in the case of the 30o x 520 mm2 wedge and the 60o cone x 260 mm2, where the determination coefficients R2 were 0.84 and 0.81, respectively, for the humidity level of 28%. In treatments with a high humidity level (35%), only moderately strong degrees of correlation were obtained with the apparent density for the cases of the cone of 30o x 520 mm2 and the wedge of 30o x 260 mm2.
Regarding the ASABE cone index and its correlation with the apparent density, the results of the experiments showed a weak correlation (R2 = 0.16) for the high humidity level and a moderately strong correlation (R2 = 0.60) for the humidity of 28% (Fig. 4). Similar results were obtained by Vega (2008) during the continuous sampling of the cone index in Red Ferrallitic soils in cane areas of Mayabeque Province, while Hall & Raper (2005) report, a similar determination coefficient between the cone index and the bulk density (R2 = 0.55), but for a sandy-silty soil (71.6% sand; 17.4% silt; 11% clay).
FIGURE 4.
Regression equation and coefficient of determination between the ASABE cone index and the volumetric density for each of the moisture levels under study.
These results contrast with those obtained for another type of soil by Ramírez & Salazar (2006), who report, for an Andisol (Marinilla-La Montañita, Colombia) a close relationship (R2 = 0.95) of the apparent density with resistance to the penetration obtained with a cone penetrometer 30o and 10 mm in diameter at the base, within an experimental range with densities between 0.3 and 1.0 g ∙ cm-3 and resistance to penetration between 2.0 and 4.2 MPa.
In the same way, the results obtained show that the resistance to penetration obtained with the 30o x 520 mm2 wedge-shaped prismatic probe presented the highest degrees of correlation, both with the ASABE cone index, for both humidity levels, as with the bulk density for the humidity level of 28% (Fig. 5).
FIGURE 5.
Regression equation and coefficient of determination between wedge indexes (IW) obtained with a wedge of 30 x 520 mm2 and the volumetric density for each of the moisture levels under study.
Similar results, although in a type of sandy-silty soil (71.6% sand; 17.4% silt; 11% clay), reported Hall & Raper (2005), who obtained higher values of the coefficient of determination (R2 = 0,74) with a wedge-shaped prismatic probe 30o x 620 mm2 in base area.
CONCLUSIONS
The wedge-shaped prismatic probe with an angle of 30o and a base area of 520 mm2, presented the highest levels of correlation with the ASABE cone index (R2 = 0.95) and the apparent density (R2 = 0.84), for a humidity of 28%. Hence, this probe has the best characteristics for detecting the resistance to penetration of a leached Red Ferrallitic soil, typical of Cuba, since it would provide information, not only regarding resistance to penetration, but also referred to soil apparent density, in this case measuring at humidity levels, about 28%.