Compaction and Physical Attributes of the Soil After the Development of Cover Plants

Compaction problems in heavily tilled soils have been commonly mitigated with the use of cover plants. Aiming to evaluate the effects of compaction on the physical properties of a plyntic Haplic-Alitic Cambisol soil after development of different cover crops, a complete randomized blocks design experiment, with 3 × 3 factorial arrangement and four replications, was conducted. Treatments consisted of cultivation of two legume species, crotalaria (Crotalaria juncea L.) and stylosanthes cv. Campo Grande (Estilosantes capitata + Estilosantes macrocephala) and a grass species, brachiaria (Urochloa brizhantha cv. Marandu), subjected to soil compaction: CM–Conventional soil management (tillage) without additional compaction; CMc4 and CMc8–conventional soil management with additional compaction using a 6 Mg tractor in four and eight wheel passes. Conventional management with additional compaction does not affect significantly the physical attributes at a soil depth of 0.10-0.20 m, and only the soil moisture does not differ according to the soil management, irrespective of the depth and kind of cover plant. Traffic levels in four passes result in an increased soil bulk density and macroporosity in the 0.0-0.05 m, and in soil resistance to penetration and total porosity in the layer up to 0.10 m. Cover crops are important in maintaining soil physical quality to reduce the negative effects of compacting forces, especially to stylosanthes cv. Campo Grande that provided greater soil protection in systems with or without addition of compaction, conditioning the lowest values of bulk density and soil resistance to penetration.


Introduction
Characterization of the effects of management systems on soil degradation and physical quality is very important (Servadio, Marsili, Vignozzi, Pellegrini, & Pagliai, 2005;Martinkoski, Vogel, Jadoski, & Watzlawick, 2017) and can be assessed through the soil bulk density, macro and microporosity, aggregates stability, mechanical penetration resistance and water infiltration in soil (Vasconcelos, Souza, Cantalice, & Silva, 2014;Gomes et al., 2017).Most of the studies on Amazon degradation deal only with the removal of forest cover (deforestation), not computing alterations caused by farming, especially when we observe the physiognomy areas of natural fields (steppe savanna) that are largely present in southern Amazon.
Understanding and quantifying the impact of land use and management on the soil physical quality are vital for the development of sustainable cropping systems (Marsili, Servadio, Pagliai, & Vignozzi, 1998).G. Souza, M. Souza, Silva, Barbosa, and Araújo (2014) observed that compacted soils exhibit a decreased volume of macropores and water infiltration and an increased soil density and penetration resistance.Tillage, or mechanical soil management, causes changes in the pores shape, reduction of channel dimensions and, consequently, in soil water permeability (Castro, Vieira, Siqueira, & Andrade, 2009 with diagonal tires and 6 Mg of total mass, with inflation pressure of 124 kPa in the front tires (14.9-28R1) and 137 kPa in the rear tires (23.1-30R1) was used.Compaction was achieved by tractor wheeling on the entire surface of the plot so that the tires compressed parallel areas between each other.The number of passes varied according to the treatment, and the tractor passed over the former pass so that the entire surface of each plot was wheeled in same number of times.
Sowing of the cover crops was made on February 2, 2012, comprised of: crotalaria (Crotalaria juncea, L.), stylosanthes cv.Campo Grande (Stylosanthes capitata and S. macrocephala) and brachiaria (Urochloa brizantha).Legume plants and grasses were cultivated in plots measuring 5 m long and 4 m wide, in a total area of 20 m 2 , and were sowed with the aid of a sowing machine with the recommended density for each species (crotalaria = 30 kg ha -1 , stylosanthes Campo Grande = 3.0 kg ha -1 , and brachiaria = 6.0 kg ha -1 ), with spacing of 0.33 m between lines.The sowing machine drill was removed to prevent possible adverse impacts of compaction, using only the cutting disc of the seeds metering device.The cultures were kept free from competition by applying herbicides and did not receive mineral fertilizer.
To assess emergence, the normal seedlings that emerged were counted every day until all replications were stabilized, which in this case was 17 days after emergence.Height of the cover plants was measured during the flowering period from the ground to the plant top, and was carried out in a representatively and randomly in the plot.To determine fresh and dry mass of the cover plants, the entire above-the-ground (or aerial) part of the plants was collected randomly at the beginning of the flowering period in a 0.5 m 2 area (1.0 m × 0.5 m).The material was weighed and put into a forced-air circulation oven at 65 ºC to constant weight to determine the dry mass of the aerial part of the plant (DMAP).The results were extrapolated to one hectare and expressed in Mg ha -1 .
Two samplings were conducted in the experimental area, the first in 2011 (Table 1) and the second in 2012, when the cover plants were in full blossom.Two samples of each layer were collected for determination of soil bulk density (ρ b ), soil moisture (Ms), soil penetration resistance (SPR) and soil porosity (macro, micro and total).Samples with preserved structure were collected with metal cylinders measuring 4.0 cm high, 4.05 cm diameter and with a volume of 64.8 cm 3 from the following layers: 0.0-0.05m, 0.05-0.10m and 0.10-0.20 m.Afterwards, the samples were saturated gradually to reach approximately 2/3 of the ring height.After saturation, the samples were subjected to -6 kPa pressure for determination of micro, macro and total porosity using the tension table method (EMBRAPA, 1997).
Table 1.Physical characteristics of the experimental area in Humaitá/AM in 2011, in 0.0 -0.20 m layer, before cultivation with cover plans  When the samples attained balance in the referred pressure, soil penetration resistance was measured, using an electronic penetrometer with constant speed of 1 cm min -1 , base diameter of 4 mm and semi-angle of 30º, developed by Serafim et al. (2008).The penetration rod was introduced into the geometric center of each sample.
The results in the 5-mm higher and lower part of the samples were disregarded, aiming to eliminate the sample peripheral effect.The frequency of readings of the SPR corresponded to the values obtained every 0.25 seconds, resulting in a total of 800 readings per sample, and a mean value of the readings was used.After the SPR determination the samples were put in oven at 105-110 ºC for 48 h, and the soil volumetric moisture and density were obtained using the volumetric ring method.
Data were subjected to analysis of variance and, when significant, to the Tukey's test at 5% probability level to compare the means.After standardization of the variables with zero mean and unit variance (µ = 0, σ = 1), data were subjected to Principal Component Analysis (PCA).The criterion for choosing the number of components was by selecting those that had eigenvalues above 1.00, as proposed by Kaiser (1958), and were able to synthetize an accumulated variance above 70% (Hair, Black, Babin, Anderson, & Tatham, 2009).This analysis was carried out using the Statistica 7.0 software program.

Results and Discussion
Soil compaction caused by tractor increased the soil bulk density and penetration resistance, resulting in reduced soil moisture (Ms), reduced macroporosity and total porosity (Pt) (Figures 2 and 3).Soil moisture was the only property that did not exhibit difference in the averages between tractor passes, layers and cover plants (Figure 2b).On the other hand, ρ b and SPR exhibited a more varied behavior, differing statistically according to the number of passes, depths and cover plants.Emphasis is given to the 0.10-0.20 m layer, where tillage did not have a significant influence on the physical attributes assessed.
In general, the 0.0-0.05m layer exhibited lower ρ b values compared to the 0.05-0.10m and 0.10-0.20 m layers (Figure 2a).This was also pointed out in the study of Andrade, Stone, and Silveira (2009) in determining the effect of cover crops (crotalaria and stylosanthes) on the physical quality of a Dystrophic Red Latosol under no-tillage.The cover crops, due to each plant's growth habit, played a different role in reducing ρ b , and brachiaria and crotalaria were the cover plants that indicated the lowest vales of ρ b (1.15 and 1.16 Mg m -3 , respectively, in the 0.0-0.05m layer, without additional compaction.On the other hand, the exploratory aggressiveness and uniformity of the stylosanthes' dense fasciculate root system minimized the negative impact of compaction with CMc4 and CMc8.The stylosanthes' mechanical behavior that was found in this study validates its contribution as a soil biological decomposer (Negreiros  2c).So, the values slightly higher than 2 MPa for crotalaria and stylosanthes and higher than 3 MPa for brachiaria indicate limiting conditions to the root growth (Couto et al., 2013;Deperon Júnior, Nagahama, Olszevski, Cortez, & Souza, 2016).
In the 0.10-0.20 m layer, irrespective of the kind of cover plant, the treatments (CM, CMc4 and CMc8) did not differ from each other and exhibited lower SPR values (1.75 and 1.96 MPa) (Figure 2c).In the study conducted by Tavares Filho, Barbosa, Guimarães, and Fonseca (2001), they found SPR values higher than the critical limit but without limitation to the root development.On the other hand, Andrade, Stone, and Godoy (2013) considered that SPR values ≥ 1.9 MPa were indicative of compacted soils, limiting plant growth.These conflicts allow to state that any inference about the soil physical quality requires multiple on the main properties that are sensitive to soil management, e.g. a porous soil area, to avoid wrong conclusions and inadequate soil management.
As no significant differences were found in the ρ b , Ms and SPR properties in the 0.10-0.20 m layer, the same was observed for the Pt, Macro and Micro (Figure 3).Macro and Pt and were slightly affected in the 0.0-0.05m and 0.10-0.20 m layers in the CM treatment, a behavior that was in agreement with that found by Bergamin et al. (2010), which can be explained in the investigation conducted by Camargo and Alleoni (2006).According to these authors, tractor wheeling at a pressure of up to 110 kPa is not enough to change Pt and macro porosity on the soil surface, considering that the wheeling pressure on the soil surface is nearly the same as the internal tires pressure.Kondo and Dias Júnior (1999) found that the 183 kPa pressure did not compress a very clayed Red Latosol, because pressures in this range did not exceed the bearing capacity of soil.
Compaction exerted by the tractor wheeling in the CMc4 and CMc8 treatments caused a total reduction of the pores volume in the 0.0-0.05m and 0.05-0.10m layers, reflecting immediately in the reduction of Macro (Figures 3a and 3b).It can be seen that there was no statistical difference for Macro in the soils with four and eight passes, which indicates that the soil has already attained a maximum macropores reduction.However, the difference only occurred in the cover crops, showing that stylosanthes, brachiaria and crotalaria present specific biological characteristics in the soil physical quality (Andrade et al., 2009).In quantitative terms, except for the CM treatment in the 0.0-0.05m layer and irrespective of the cover plants, the compaction-induced Macro values are quite below the value that is considered critical, i.e., the minimum aeration value of 0.10 m 3 m -3 that is necessary for root development, as pointed out by Gupta and Allmaras (1987).
It was found that in the CMc8, the low Macro values tended to remain the same in the 0.05-0.10m and 0.10-0.20 m layers, where they did not differ significantly (Figure 3b).This behavior, coinciding with the highest ρ b and SPR values indicate that the pressure exerted in the CMc8 or over resulted in a new rearrangement of the soil particles, reaching maximum soil compaction.The results lead us to interpret that, under such mechanical conditions imposed by traffic, cover crops have difficulty to develop and lose their biological soil scarification potential (Andrade Neto, Miranda, Duda, Góes, & Lima, 2010;Villa Secco, Tokura, & Pilatti, 2017).contributed to reduce compaction of sandy soil, as is the case of Entisol, which is an expected behavior, since the species is native to the Cerrado (savanna), predominant in sandy and well-drained soils (Andrade, Assis, & Sales, 2010).
Soil compaction had an influence on the parameters of the cover plants assessed (Table 2).According to Modolo, Trogello, Nunes, and Modernel (2011), this happens because compaction reduces water infiltration, temperature and soil aeration and, consequently, the plant emergence and growth.Emergence suffered a reduction of 16, 12 and 11% for brachiaria, crotalaria and stylosanthes, respectively, in the CMc4 compared to CM.With the CMc8, such decrease reached 23%, 39% and 12% for brachiaria, crotalaria and stylosanthes, respectively.In turn, emergence of the cover plants did not differ significantly from each other.Considering a higher emergence rate, 85% on average, for the stylosanthes, 82% for crotalaria, and 80% for brachiaria, it confirms the preferable growth habit in more sandy soils (Andrade et al., 2010;Castagnara et al., 2013), as is the case of the plyntic Haplic-Alitic Cambisol of the present study, classified as a loam-clay-silt soil. (1)CM: conventional soil management (tillage) without additional compaction; CMc4 and CMc8: conventional soil management with additional compaction by traffic of 6 Mg tractor in four and eight passes, respectively.Means followed by same letter did not differ statistically from each other, uppercase in column and low case in row, by the Tukey's test at 5% probability level.CV: coefficient of variation.
Studies by Modolo, Fernandes, Schaefer, and Silveira (2008) found lower emergence rates in soybean cultivated under wheel loads ranging from 0 to 50 N, making the contact between soil and seed difficult, as well as water absorption, delaying the germination process.However, the lowest emergence rates observed were under wheel loads of 140 N due to surface sealing, little oxygen entry, causing the seedling to consume much more energy to emerge.
When analyzing the production of dry matter of the aerial part (DMAP) of the crotalaria in different soil managements (Table 2), it is possible to observe that there was no significant difference and yet it produced a lower amount of DMAP (1.70 to 1.41 Mg ha -1 ) than the other cover crops assessed.In the plots cultivated with brachiaria under average ρ b of 1.4 Mg m -3 , there was a significant reduction only in the CMc8 treatment compared to the CMc4 and CM.Studies conducted by Medeiros, Soares, and Guimarães (2005) Note.(1) ρ b : bulk density; SPR: soil penetration resistance; Mv: volumetric moisture; Pt: total soil porosity; Macro: macroporosity; Micro: microporosity; CM: conventional soil management (tillage) without additional compaction; CMc4 and CMc8: conventional soil management with additional compaction by 6 Mg tractor traffic in four and eight passes, respectively. ).

Table 2 .
Plants emergence, height, fresh mass of aerial part (FMAP) and dry mass of aerial part (DMAP) of three cover crops grown in different stages of soil compaction explain that a decreased DMAP is due to changes in the soil physical-hydro properties caused by soil compression, reducing