4 Discussion
The physical properties studied did not present great variation in the profile, as demonstrated through a comparison of the three layers. L2 layer showed slightly lower values for porosity but higher values for BD and PR than the other layers; however, this layer presented the normal values for S index (Fig. 1). The PR values increased with increasing soil depth, as demonstrated by a comparison of L2 with the surface layer of Oxisols in the Brazilian Cerrado. Similar trends have also been reported by Silva et al. (2012) and Cherubin et al. (2015). Further, de Moraes et al. (2016) also reported the reduction in hydric properties and porosity at soil depth of 0.10-0.20-m.
The subdivision of the layered profile (L1, L2 and L3) was important in this study due to direct influence of the variables on soybean productivity.. Nevertheless, variables only at L1 were significantly related with the soybean yield (Table 1). However, the soil properties at L2 indicated a lower quality compared with those of L1 and L3, nonsignificant and negative influence of this lower quality was observed on the soybean yield. These trends might be attributed to the strong stratification of the chemical properties in the surface layer under long-term NT (Caires et al., 2015; Martínez et al., 2016) because L1 contains most of the roots of the cultivated plants (Calonego et al., 2017). Therefore, any fluctuation in variables atL1, which contains most of the roots of soybean, would have greater effects on productivity than changes in L2 and L3.
The development of the root system of plants depends on various factors, including the physical conditions of the soil, such as BD, PR and particle size (Rogers et al., 2000). The ability of the soybean’s root system to provide a sufficient amount of water to the entire plant depends on its abundance, diameter and root length. These characteristics of the root system increase its area of ​​absorption, its transport capacity and the volume of soil exploited (He et al., 2016). However, the benefits of favorable soil physical conditions, such that the genetic potential is expressed and the root growth in the profile is adequate, might not result in productivity gains in relation to those found in lower-quality areas if the rains are well distributed or if the chemical properties are stratified in the surface layer (Conte et al., 2009). Although the architecture of the root system depends on several factors, the distribution of nutrients in the profile is also important; for example, phosphorus stimulates lateral growth of the root system, and nitrogen stimulates an improved depth distribution. Thus, the surface application of nutrients might restrict the depth distribution of roots (He et al., 2016).
The direct positive relation of SOM and its fractions to the clay content might be attributed to favorable conditions for the formation of aggregates that arise from the combination of clay and SOM, which is intensified in high-fertility soils due to the presence of divalent cations (Denef and Six, 2005). The formation of aggregates increases the chemical stabilization of SOM by cations (Kögel-Knabner et al., 2008), increases the sorption of SOM on clay surfaces (Vogel et al., 2014) and decreases the action of microorganisms due to physical insulation (Baldock and Skjemstad, 2000). However, in this study, the opposite trends were also observed in sandy soils (Fig.7), a negative relationship shown.
The non-significant relationship of PAW and S index with SOM and its fractions was expected because NT increases the AWC to plants with greater efficiency (in cases with the same water content (PAW) and the same pore distribution (S index)) compared with that obtained under conventional plow-based systems (Le Quére et al., 2015). The significant relationship of PR with OLF is important which should further be investigated to potentially alleviate the soil compaction under NT systems.
The PR and BD are strongly correlated (Mota et al., 2014), and both exerted a negative direct effect on soybean yield. The values of PR and BD for which the line of adjustment of the model crossed the line of the productivity. The average values of PR and BD in studied plots were approximately 1.6 MPa and 1.25 Mg m-3, respectively. The value of PR below than 2.0 MPa was proposed by Taylor et al. (1966) and the values of 2.5 MPa and 1.64 Mg m-3 proposed by Tormena et al. (2017) for PR and BD, respectively. Furthermore, the PR was measured at the soil moisture content corresponding to a potential of 6 kPa. Both PR and BD might have adversely affected the root growth at 0-0.07 m soil layer (Tormena et al., 2017), but an increment in soil moisture can alleviate these negative effects.
The value of the α parameter of the Van Genuchten model was strongly affected by water loss at high potentials close to zero. The increase in the value of α has a positive and direct relationship with the presence of macropores with diameters close to the lower limit of the class, and drainage starts at high potentials. The parameter n depends on the rate of water loss from the point of inflection of the curve, and increases the rate of water loss with increase in the value of n. Therefore, to obtain an increase in the modulus of n, an equilibrium of the proportion of pores with each diameter class is required. In this context, the interventions that promote soil mobilization or the use of a seed drill can lead to a strong disturbance in the seeding row, increases the volume of macropores, and reduces the value of n, leading to advesre effects on productivity.
The matric potential at the inflection point of the curve, as indicated by I, reveales the continuity of the pore quality which is given by n, and its value depends on the pore size distribution in the micropore class. A lower loss of water in this category is associated with a lower potential. The distribution of pores that increases the values ​​of n and I depends on the structural arrangement of the soil, which can only occur in the absence of plowing, and the combined effects of the SOM, plant root system, biological activity and other mineralogical and chemical components that moderate soil aggregation.
The significant positive effect of total porosity and S index at L1 on the soybean yield indicates the need to balance the capacity between water and air (gas exchanges) to harvest high yields. Aeration, when reduced by water in the pores, linearly affects the root growth (Benjamin and Karlen, 2014). The soybean yield under Brazilian edaphic conditions is strongly dependent on biological nitrogen fixation by symbiotic bacteria. However, the fixation efficiency depends on satisfactory aeration and gaseous exchange (Tewari and Arora, 2016). A pore size distribution with balanced proportions in different diameter classes ensures the desired combination of aeration and water availability, as indicated by the S index (Dexter, 2004).
The non-significance of the relationship between the soybean yield and PAW/AWC might be attributed to a favorable rainfall distribution in quantities greater than the minimum soybean demand during the three harvest cycles of soybean. In the case of frequent rains, the role of soil capacity to supply water is minimal (Calonego et al., 2017). In this study, the soils under NT management had moderate to high PAW values.
For soils of sandy and clayey texture, NT management can affect the PAW and the productivity. The factors influenced by NT to reduce the difference in productive potential between sandy and clayey soils are increase in the SOC concentration, soil fertility (Sá et al., 2015) and water use efficiency by plants (Le Quére et al., 2015). The presence of crop residue mulch on the soil surface reduces the water loss through evaporation. The straw mulch also creates micro dams that delay water runoff and increase infiltration into the soil (Zhao et al., 2013). These effects add to the productive potential of sandy soils, which can then approach that of clayey soils.
Minasny and McBratney (2018) reported a modest SOM effect from increases in the PAW; specifically, only 1.4 and 1.9% increase in the PAW were obtained for every 10 g.kg−1 (1%) increase in the SOC in clayey and sandy soils, respectively. However, other benefits of conservation agriculture are obtained through the indirect effects of mulching, which can reduce evaporation and runoff and improve soil aggregation and biogeochemical functioning. Nonetheless, the review was primarily based on data from soils in temperate regions of the north-central USA and Europe or in dry regions of Australia. Under these conditions, predominant soils contain high-activity clays with strong affinity for physiochemical interaction with water. In contrast, the soils of the tropical Brazilian Cerrado region contain predominantly low-activity clays and under these conditions, SOM could be most critical for soil aggregation, structure (Vezzani and Mielniczuk, 2012), and CEC (Ciotta et al., 2002) and replaces the effect of clays on physiochemical interactions with water. In contrast, to the conclusions drawn by Minasny and McBratney (2018) reported a strong and positive effect of SOM on the PAW.
The data reported in this study regarding the relationship of SOM and its fractions to the PAW, PR and S depended on the predominance of clayey soils in the study area: 85% of the plots contained more than 40% clay, and the high levels of SOM masked the effects. Tavares Filho et al. (2012) reported a negative relationship between the PR and SOM in a red Oxisol, with a SOM content ranging from 12 to 40 g.kg-1. The S index, affected by the pore size distribution, was positively related with the FLF, and this relationship constitutes the basis for microbial activity, leading whose action is critical to the formation and stabilization of aggregates (Six and Paustian, 2014) and to increases in the S index.