Soil physical attributes in integrated bean and sheep system under nitrogen levels 1

Integrated crop-livestock systems (ICLS) can promote higher grain and meat/milk yield as well as reduce soil degradation risks. Thus, this study aimed to evaluate the influence of grazing sheep with moderate intensity and nitrogen doses on soil physical attributes, and bean crop yield (Phaseolus vulgaris Lam.) under rotation with summer corn crop (Zea mays Lam.) and winter pasture with oats (Avena sativa Lam) and ryegrass (Lolium multiflorum Lam.) The experimental design was completely randomized blocks with split-plot; the plots were nitrogen doses and the splitplots were grazing presence or absence. The treatments were nitrogen doses of 0, 75, 150 and 225 kg ha-1 applied on ryegrass and white oat pasture in winter followed by bean crop. Soil density, macroporosity, microporosity and total porosity and bean grain yield were evaluated. When ICLS are managed under no-tillage with moderate grazing, sheep trampling did not affect negatively soil physical attributes, which did not reach critical levels for subsequent growth and development of the bean crop. The black bean yield in summer was not affected by antecedent presence of sheep grazing in winter.


INTRODUCTION
Integrated crop-livestock systems (ICLS) have developed significantly and shown great improvement on soil quality, environment with increased production yield (MACEDO, 2009;ANDREOLLA, 2010).In southern Brazil, ICLS have been managed as alternative to increase farmers' profitability.These systems have used plant species like black oat (Avena strigosa Schreb) and ryegrass (Lolium multiflorum Lam) integrated with a diversity of summer crops (BALBINOT JUNIOR et al., 2009).
According to Silva et al. (2011), annual pastures followed by annual grain crops become efficient the soil physical structure and chemical properties, thus, increasing plant growth, because species with different root systems and shoot morphologies promote greater nutrient cycling and can fix nitrogen from the air, leading to sustainable yields.
The animals' trampling during grazing can compact soil layers especially on moist soils and high rate of grazing stocking (GIAROLA; TORMENA; DUTRA, 2007;LANZANOVA et al., 2007).According to Conte et al. (2011) and Carassai et al. (2011), such areas can show some topsoil compaction when under moderate grazing, but it does not cause serious damage to the crop development.However, in another hand, higher grazing intensity decreases the soil cover, which can directly affect soil compaction level.
Integrated crop-livestock systems under high grazing intensity increase the soil bulk density and microporosity as well as decrease macroporosity; even though, this negative impact occurs only on topsoil.Carassai et al. (2011) and Silva et al. (2011) found out that both grazing intensity and annual cropping help on changing such compaction levels and promote a viable environment for achieving sustainable yields.Thus, ICLS are a reliable alternative because they can diversify farm environments and can associate soil-plant-animal compounds in a balanced way; therefore, they preserve and restructure the soil, increasing the organic matter and nutrients (LOSS et al., 2012).
Bean crop is a very promising alternative to be employed in ICLS, because it has a short cycle and its sowing date is easy to be managed, allowing longer periods for using pasture.Furthermore, when nitrogen is supplied to winter crops under moderate grazing stocking, it is possible to keep nitrogen residual effect for subsequent crop (REICHERT, 2012).This fact helps the maintenance of soil fertility, because it adds organic matter and facilitates the nutrients cycling (CARVALHO et al., 2010;LOSS et al., 2012).
This study aimed to evaluate how grazing and nitrogen affect the soil physical attributes and the black bean yield under summer corn and under winter oat/ryegrass in rotation with pasture.

MATERIAL AND METHODS
This experiment was carried out at the experimental area inside the Universidade Estadual do Centro-Oeste, in Guarapuava, Paraná State (PR), Brazil (51° 29' longitude and 25° 33' latitude).Regional climate is mesothermal humid subtropical with cool summers and mild winter (Cfb), according to Köppen (MAAK, 1968), and mean annual temperature range from 18 °C to 22 °C.Mean annual rainfall range from 1.400 mm to 1.800 mm and altitude is 1.100 m (IAPAR, 1994).
Soil is classified as Oxisol typical Distroferric (EMBRAPA, 2013).The studied area was under notillage since 2004 and the rotation was implemented (Table 1).S o i l c h e m i c a l a n a l y s i s w a s c a r r i e d o u t f o r undertaking a correct cropping management (Table 2).The experiment was carried out in randomized block design with split-plots and three replicates.The plots were composed by nitrogen doses while the split-plots were treatments with and without sheep grazing at moderate intensity.The total experimental area had 3.1 ha and 2.6 ha were used and divided in three blocks of 0.88 ha.Each block was divided into four plots of 0.22 ha; and inside each plot, 96 m 2 area was isolated and remained without grazing.The remaining 0.5 ha was used to keep the regulator animals for maintaining the same grazing pressure among the experimental units.The treatments consisted of four nitrogen (N) doses applied under urea form (450 g N kg -1 U) (0, 75, 150, and 225 kg N ha -1 ) on system's winter phase.After corn harvesting, the pasture sowing was carried out on May 20, 2008, with 17 cm among rows.The fertilization occurred at the pasture sowing time with 0.0 N, 60 P 2 O 5 , and 60 K 2 O kg ha -1 , according to Comissão de Fertilidade do Solo (Comission of Soil Fertility) (1995).
The grazing occurred during 158 days, from July 22 to November 18, 2008; 72 weaned lambs were used for testing.Lamb breeds from Texel and Ile de France were used and their crossbreeds also, and they started grazing at nearly two months old and with initial live weight of 24.7 kg ± 0.57 kg.
The animals were randomly divided into 12 lots of six lambs in the plots and kept under continuous grazing system with ryegrass.The load was variable to keep the pasture at 14 and 15 cm height (MOOT; LUCAS, 1952) according to Freitas (2003).Stocking adjustment was periodically made at three days intervals considering the ratio between height and mass forage according to Carvalho, Ribeiro Filho, and Poli (2001).
The stocking rate was obtained by the sum of the average weight of the testing animals and the average weight of the regulator animals multiplied by the number of days that they remained into experimental unit.The obtained value was divided by the number of grazing days and expressed in kilogram of live weight per hectare (kg Lw -1 ).After removing the animals, the pasture desiccation was carried out on December, 2008; then, beans were sown under no-tillage system with 40 cm among rows and 15 seeds density per linear meter.
Bean emergency occurred on December 15, 2008.The FTS Sovereign bean which belongs to the black trade group was the used bean genotype with indeterminate growth habit, upright, and medium blooming cycle.It is adapted to the mechanical harvesting and it is also moderately resistant to major crop diseases.
The bean crop was fertilized with 100 kg P 2 O 5 ha -1 (triple superphosphate) and topdressing with 190 kg K 2 O ha -1 (potassium chloride).Nitrogen was not applied.Control of weeds, pests, and diseases during the crop cycle was carried out according to Araújo et al. (1996).
The bean yield was determined in a 3.2 m 2 area (2 rows x 0.4 m x 4 m); after moisture correction to 13% and the value was converted to kg ha -1 .The bean was harvested on March 30, 2009.
Samples were taken at 0-5, 5-10, and 10-15 cm deeps with two replicates for evaluating soil physical attributes.Samplings were taking at the following periods: (a) after corn sowing (2008); (b) after winter phase (2008); and (c) after beans harvesting (2009).Soil bulk density (BD) was determined by the volumetric ring method, while total porosity, macroporosity, and microporosity were determined by the voltage table method (EMBRAPA, 1997).

Results presented homogeneous variances by
Lilliefors test and were submitted to analysis of variance by F test.When the results differ at 5%, the averages of qualitative factor (grazing/no grazing) were compared by Tukey test, and quantitative factor (nitrogen) was compared by regression analysis with SISVAR software (FERREIRA, 2000).

RESULTS AND DISCUSSION
There was no interaction (p>0.05)among factors evaluated on physical soil attributes.It was observed that BD values (   (2009).However, it is remarkable that the BD value increased after grazing and such value decreased at all nitrogen doses after beans harvesting, indicating that there are no significantly changes on soil physical attributes when ICLS are carried out with no-tillage and moderate grazing stocking.et al. (2009) and Flores et al. (2007) evaluated animals' grazing as well as different management on soil physical attributes and reported similar results.The low BD was attributed to high aggregate stability and to soil structure preservation, which had been used with grazing in winter for eight years before this trial.It also indicates that controlled grazing during the winter does not affect soil physical quality (SILVA et al., 2011;COSTA et al., 2009) and demystifies that the animals' presence in farming areas causes problems to the subsequent crop.Stocking rate was 498, 1.014, 953 and 1.208 kg Lw -1 on average, for respective treatments with 0, 75, 150, and 225 kg N ha -1 .

Costa
Regarding macroporosity (Table 4), there was similar trend of BD, with a decrease on values after animals' grazing on the surface layer; those values increased at all nitrogen doses after bean cropping.Macroporosity was higher in areas without grazing after winter period.Flores et al. (2007) observed similar results.This pattern usually occurs.However, the  , 2010;CARASSAI et al., 2011).
Regarding microporosity (Table 5), it was found out that there was insignificant increase in value after animals' grazing.Changes in soil microporosity observed on ICLS under no-tillage with moderate grazing (SPERA et al., 2009;FLORES et al., 2008) are generally small in magnitude and did not reach critical levels for root growing of cultivated plants.
This behavior was expected and was due to radicular renewal ability and biopores' formation, which promoted the soil decompaction by using forage plants in ICLS.Castagnara et al. (2012) andBalbinot JR. et al. (2009) also studied different soil managements and did not find out any significant differences on macroporosity between areas with or without grazing.
Regarding total porosity, there were no differences among treatments.Thus, it can be concluded that a well-structured management system affects directly soil response as well as its physical, chemical, and biological properties.
Soil under cropping for a certain period can recover its initial conditions concerning the physical attributes and it can prove that animals' grazing presence does not affect soil quality when rationally used.Loss et al. (2012) and

Table 1 -
Crop rotation managements from 2004 to 2009

Table 2 -
Chemical parameters of soil in the experimental area, Guarapuava (PR) OM: organic matter.P: phosphorus extracted by Mehlich-1 solution.CEC: cation exchange capacity at pH 7.0.V: base saturation

Table 3 )
did not reach limiting values V. R. M.Andreolla et al.

Table 4 -
Macroporosity (m 3 m -3 ) in 0-5, 5-10 and 10-15 cm layers in three sampling periods with and without grazing, Guarapuava, PR Mean in the same column and layer does not differ by F test (p>0.05) *