Soil Physical Quality Indicators and Refinement of the Evaluation Method through the Srelative

This study aimed to verify the efficiency of indicators of measure of physical attributes’ alterations and to refine the Srelative determination method in order to increase its sensitivity to soil physical alterations. Soils under Ficus carica L. cultivation (with 0, 20, 40 and 60% of liquid bovine biofertilizer in the irrigation depth) and under forest were used. Parameters evaluated included soil granulometry, soil bulk and particle density, soil water retention curve (SWRC), porosity and the indices S and Srelative. The experimental design was completely randomized with four replicates. For Srelative refinement, with the SWRC containing only textural porosity, the soil was dispersed in water and with the addition of 1 N sodium hydroxide (with and without removal of sodium through washing). An ANOVA was performed for 0, 20, 40 and 60% of biofertilizer in 0-10, 10-20 and 20-30 layers; Dunnett test was used to compare the mean values of S-index and Srelative-index. With respect to four methods to obtain the Srelative-index the means were compared by Tukey test. Tests of line parallelism and intercept were performed for the regressions between each of the soil physical variables and Srelative-index obtained. It was found that S and Srelative indices were sensitive to soil physical alterations caused by the application of the biofertilizer; Srelative-index was sensitive to variation in soil bulk density and total porosity and the Srelative-index obtained from the method of soil dispersion in water is more sensitive to soil physical alterations in comparison to Srelative-index obtained through ADFE.


Introduction
Soil quality has become a theme of interest to researchers concerned with the protection of the soil and the sustainability of the agricultural systems.Initially, efforts were made to define soil quality and later transform such concept into something measurable (Armenise et al., 2013).
In soil science, various approaches about the concept of quality have been used by researchers and most of them are based on Larson and Pierce (1994)'s definition of soil quality as a combination of physical, chemical and biological properties that provide the means for both plant and animal production, as well as to regulate the water flow in the environment and serve as an environmental filter in the attenuation and degradation of environmentally damaging or dangerous components.For Villamil et al. (2008), soil quality is a multidimensional concept, in which many studies are involved, in combination or individually, in order to improve the understanding of the dynamics of the system.Soil quality cannot be determined directly, but can be constantly monitored through the quantification of alterations in its attributes resulting from use and management systems (Neves et al., 2007;Obade & Lal, 2016a).Monitoring soil quality is a promising component to be used in management strategies of agricultural soils (Ripoche et al., 2010).
However, it should be pointed out that the simplification of information about soil quality may result in inconsistent conclusions, which can lead to damages to the studied system, since some of the analyzed attributes are subjective and insufficient to represent a complex environment that plays various functions, like the soil (Obade & Lal, 2016a, 2016b).Thus, the most sensitive attributes to changes in soil management practices are the most adequate to be used as indicators (Arshad & Martin, 2002).
In this perspective, researchers have attempted to identify, select and attribute quantitative values to quality indicators that best represent the performance of certain functions of the soil.Among the various indicators, an index of evaluation of soil structure-S relative -index-derived from the S-index proposed by Dexter in 2004, was presented by Freire (2012).The results observed so far have been promising; however, it was verified in Freire (2012), Alves (2013), andAssis Júnior et al. (2016) the incomplete individualization of the particles to obtain the soil water retention curve reference.Thus, the S relative -index requires refinement to improve its sensitivity to physical changes imposed to the soil.
In this context, the study was based on the following hypotheses: 1) soil physical alterations can be assessed by indices and interpreted under the qualitative aspect; and 2) the S relative -index obtained from a soil water retention curve determined as close as possible to the textural porosity (reference curve) is more sensitive to soil physical alterations in comparison to the S relative -index obtained from a curve determined using air-dried fine earth (ADFE).Therefore, this study aimed to verify the efficiency of indicators of the measure of physical attributes alterations and to refine the S relative -index determination method in order to increase its sensitivity to soil physical alterations.

Studied Area
The studied area is located in the Apodi Plateau, at the Unit of Teaching, Research and Extension (UEPE), one of the physical units of the Federal Institute of Education, Science and Technology (IFCE-Campus of Limoeiro do Norte), situated in the municipality of Limoeiro do Norte-CE, Brazil, at an altitude of 145 m.The experimental area, cultivated with fig (Ficus carica L.), has in its center the geographic coordinates of 5°10′57.64″S and 38°0′45.97″W. The secondary forest taken as a reference is located 400 m away from the cultivated area.The soil of the experimental area is a Cambissolo Háplico (Embrapa, 2013).Some soil physical attributes are shown in Table 1.Silt Clay ---cm ------------------------------g kg -1 --------------------------- For particle-size distribution analysis, clay (≤ 0.002 mm) content was determined through the pipette method, sand (2.00 to > 0.053 mm) content through sieving and silt (0.053 to > 0.002 mm) content was the difference between sand and clay fractions.Clay dispersed in water was determined using the same method adopted for particle-size distribution, but without the chemical dispersant.

Experimental Procedure
The experiment was carried out in an open field, cultivated with fig, which was applied different rates of biofertilizer.The experiment began in October 2010.The biofertilizer applied to the soil was produced through anaerobic process in a 200 L-plastic container.A hose was adapted and connected to the lid and its other tip was submerged in a container with water at the height of 20 cm for the outlet of gases.The proportion used for the production of the biofertilizer was 1 volume of fresh bovine manure for 1 volume of water fermented for 30 days.
The rates of the biofertilizer were formulated with the following proportions: T0 100% of water; T1 for 20% of biofertilizer and 80% of water; T2 for 40% of biofertilizer and 60% of water; and T3 for 60% of biofertilizer and 40% of water.Three liters of biofertilizer were applied to the soil per plant and twice per month from October 2010 to August 2012, totaling 23 months during the 4 crop cycles.
At the end of the experiment, the organic matter added to the soil through the biofertilizer at 20%, 40% and 60% were approximately 0.41 kg, 0.82 kg and 1.24 kg, respectively, per area available to the plant.Samples of the biofertilizer were analyzed in the Soil, Water and Plant Tissues (LABSAT) Laboratory of the IFCE for chemical characterization (Table 2).
For soil quality evaluation, five soil treatments were considered including, under fig cultivation, one treatment of each of the applications of 20%, 40% and 60% of liquid bovine biofertilizer in the irrigation depth, 100% water i.e.
without biofertilizer application and one without biofertilizer but under natural vegetation.Each of the 5 treatments was applied at each of the three following soils depths: 0-10 cm, 10-20 cm and 20-30 cm (Table 2).The experimental was a completely randomized design with four replicates.Disturbed and undisturbed soil samples were collected using an Uhland sampler, in steel rings with 0.05 m of height and 0.05 m of diameter.In the laboratory, soil samples were analyzed for granulometry, soil bulk and particle density, soil water retention curve, total porosity, S-index and S relative -index.
Table 2.Chemical characteristics of different rates of bovine biofertilizer

Laboratory Analyses
In the granulometric analyses, clay was determined through the pipette method, sand through sieving, and silt through the difference between clay and sand fractions (Gee & Bauder, 1986).Soil particle density ( p ) was determined through the volumetric flask method (Blake & Hartge, 1986a) and soil bulk density ( s ) using undisturbed samples, collected in cylinders with a known volume and dried at 105 ºC until constant mass (Blake & Hartge, 1986b).Soil porosity was obtained by the equation: where, TP is porosity (m 3 m -3 ), and  p and  s are in kg dm -3 .
In the determination of the soil water retention curve, the water content at saturation was considered as equal to soil porosity (TP); for low tensions (2, 4, 6, 8 and 10 kPa), the points were obtained using Haines' funnel.The other points (33, 100, 300, 700, 1000 and 1500 kPa) were obtained using Richards pressure plate apparatus (Klute, 1986).
The curve was fit using the mathematical model proposed by van Genuchten (1980)'s Equation ( 2), (2) where,  r and  s are, residual and saturation water contents (m 3 m -3 ), respectively,  the soil water matric potential (kPa),  (scaler of ), m and n (related to the shape of the curve).The software SWRC, version 2.0, was used and the variables  s and  r were fixed with soil moisture values measured in the laboratory at saturation and at the tension of 1500 kPa, respectively.The parameters , m and n were fitted using the iterative method of Newton-Raphson, with no dependence between the parameters m and n (Dourado-Neto et al., 2000).
Based on the parameters of van Genuchten's equation, the slope at the inflection point (S-index) was determined according to the following equation (Dexter, 2004a): (3) S relative -index was determined as a ratio between the S value obtained with the soil water retention curve for the considered management and the S of the reference curve.The S-index used as reference was obtained through the water retention curve for the soil of the secondary forest, with a disturbed soil sample, using a sample of air-dried fine earth (ADFE), placed in rings with 0.05 m of height and 0.05 m diameter, which was prepared in such a way that the particles were normally arranged without the necessity to pre-establish a value of soil bulk density (Freire, 2012).
As to the refinement of the S relative -index, the objective was to obtain a reference S-index value from the water retention curve determined using soil with disturbed structure as close as possible to the textural porosity, since the perception of the studies of Freire (2012), Alves (2013), and Assis Júnior et al. ( 2016) is that the curve determined from the ADFE still contains part of its porosity associated with the microstructure.In the refinement process, the S relative -index was obtained according to the same procedure for the arrangement of the soil in the rings.The difference is that the material tested was dispersed in water then added 1 N sodium hydroxide (with and without subsequent washing for the removal of salts, particularly sodium).Dispersion was performed following the pipette method established by Gee and Bauder (1986).As previously mentioned, the dispersion aimed to arrange soil particles according to the textural porosity.After the process of chemical and physical dispersion, the samples containing sand, silt and clay in solution were dried in an oven at 45 ºC until constant mass.
After drying in the oven, the material was analyzed for the size distribution of the fractions.For comparison purposes, the analyses were performed using 20 g of ADFE, 20 g of material dispersed in water and 20 g of material dispersed with addition of 1 N sodium hydroxide (with and without subsequent washing).Each material was sieved through a set of five sieves (1 mm; 0.5 mm; 0.25 mm; 0.105 mm, 0.053 mm), totaling six size classes (≤ 2 mm to > 1 mm; ≤ 1 mm to > 0.5 mm; ≤ 0.5 mm to > 0.25 mm; ≤ 0.25 mm to > 0.105 mm; ≤ 0.105 mm to > 0.053 mm; ≤ 0.053 mm).
After dispersion and analysis of distribution of the fractions according to the diameter, the material was used to generate the soil water retention curve and, from the curve, the S index value was obtained.As described in Freire (2012), the S relative -index was obtained by the Equation ( 4): (4) remembering that the values of S, in both forms of soil structure, mathematically derived from equation ( 1), but replaced the volume-based moisture by the gravimetric moisture.The program Microsoft Excel® was used for data processing.The Kolmogorov-Smirnov test at 0.05 probability level was applied to verify data normality.

Statistical Analyses
The granulometric distribution data received statistical treatment using the microcomputer program PHI, developed by Jong van Lier and Vidal-Torrado (1992), which uses the statistical parameters of Folk and Ward (1957) to establish comparisons between the classes of soil particle size present in the sample that was used to generate the reference soil water retention curve.In the PHI program, the input data correspond to the absolute percentages of each granulometric fraction in the sample and its respective diameter in the phi scale [φ = -log2 D (mm)].Diameter classes were transformed to a phi scale whereby  from ≤ 2 to > 1 mm led to φ from ≤ -1 to > 0;  from ≤ 1 to > 0.5 mm to φ from ≤ 0 to > 1;  from ≤ 0.5 to > 0.25 mm to φ from ≤ 1 to > 2;  from ≤ 0.25 to > 0.105 mm to φ from ≤ 2 to > 3.32;  from ≤ 0.105 to > 0.053 mm to φ from ≤ 3.32 to > 4.32;  from ≤ 0.053 mm to φ from ≤ 4.32.
The experiment had two controls, one as a reference for biofertilizer rates and the other as a reference for the soil cultivated.Analysis of variance was performed by F test considering five treatments and four replications.The comparison the means of the other treatments (biofertilizer 0, 20, 40 and 60%) in relation to soil under forest was performed by Dunnett test at 0.005 probability level.With respect to four methods to obtain the S relative -index   Given these results, with significant differences between the lines for all analyzed relationships and greater sensitivities observed for SR Water and SR NaOH , the fact that obtaining SR Water is simpler suggests that this indicator is the one that must be used to evaluate soil structural quality.

Conclusions
S and S relative indices were sensitive to soil physical alterations caused by the application of the biofertilizer.Srelative-index was sensitive to variation in soil bulk density and total porosity.The S relative -index obtained from the method of soil dispersion in water is more sensitive to soil physical alterations in comparison to S relative -index obtained through air-dried fine earth.

Table 1 .
Soil physical characteristics