The forage harvested in the area sampled per plot were determined a fresh pasture mass. Subsequently, the identified sub-samples were dried in an oven with forced air circulation at 65°C for 72 h to determine the mass of dry matter over 0.10 m high (Gardner, 1986). The percentage of dry matter was then calculated and expressed in kg ha^-1 DM. Then, soil samples were crushed (Willey mill) to determine the total content of nutrients. To determine the height of grass from the ground level to the highest part a top were used a scale and an x-ray paper were used to standardize the height of the plants, in 10 replications. The bulk density of forage was obtained by dividing the dry mass weight/height plants that was expressed in kg ha^-1 cm^-1. The forage accumulation was calculated by subtracting the herbage mass in pre-grazing forage by the post-grazing mass. Forage accumulation rate (over 0.10 m) was expressed in kg ha^-1 day^-1 DM. It was calculated by dividing the accumulation of forage for 21 days grazing cycle by 21, the number of days in the cycle.

The pasture's carrying capacity was calculated considering an herbage allowance of 5 kg DM per 100 kg live weight. After the measurements of pasture, the cattle were put to graze, aimed at standardizing the grass height to 0.10 m. After 21 days of application of the last installment of SW, soil samples were taken at different soil depths such as 0.00-0.20 m and 0.20-0.40, originating six samples, randomly collected in the plot with a Dutch auger. The pH in water, exchangeable acidity (Al³⁺), potential acidity (H + Al) and soil organic matter (OM), phosphorus (P), potassium (K⁺), calcium (Ca²⁺), magnesium (Mg²⁺) copper (Cu²⁺), zinc (Zn²⁺), manganese (Mn²⁺) and iron (Fe) was analyzed, according to the methodology described by EMBRAPA (2009).

The aerial parts of the plants were subjected to analysis of the N, P, K, Ca, Mg, S, Fe, Cu, Mn, Zn, B according to methodologies EMBRAPA (2009) after each 21-day cycle. With the nutrient content of aerial parts of each cycle, the accumulation of these nutrients and recovery efficiency of each of the plant were evaluated. The accumulation of nutrients in the aerial part of each cycle was then used to define the nutrient uptake of the grazed pasture (kg ha^-1) to 105 days of experiment. All results were analyzed using the Barttlet and Jarque-Bera test (Jarque and Bera, 1980) to check the homogeneity of variances conditions and normality,  respectively. Analysis of  variance and regression analysis for the SW doses were carried out, to determine levels of significance, using the SAEG 9.1 program, 2007.

The higher carrying capacity, 10 AU ha^-1 was obtained with the dosage of 2500 m³ ha^-1, whereas grazing efficiency of 50% (Figure 2b). Lupatinie Hernandez (2006) showed, among various types of forage, the Tifton 85 grass with high fertilization responded better to irrigation, with a carrying capacity of up to 10 AU ha^-1. Which, showed the great forage production potential of grass associating intensive management, adequate fertilization and irrigation. The accumulation rate of dry matter per day obtained averages 138-219 kg ha^-1 day^-1 (Figure 2e). Aguiar et al. (2005) showed an average annual accumulation rate of  forage  of  172  kg ha^-1 day^-1. Drummond et al. (2006) reached herbage accumulation rate of 148.2 kg ha^-1 day^-1 of dry matter in grazing Tifton 85 fertilized with swine waste in the region of Uberaba - MG. Research works on pastures have presented linear increases of forage dry matter using doses of wastewater (Orrico junior et al., 2013; Andrade et al., 2014; Homem et al., 2016), with average accumulation doses of up to 170 kg ha^-1 day^-1 of DM in Tifton 85 grass using swine wastewater (Andrade et al., 2014).

Vielmo et al. (2011) used SW at doses of 0 to 320 m³ ha^-1, cycle of 28 days on Tifton 85, and verified a production of 151 kg ha^-1 day^-1 of dry matter at the highest dose. Gomes et al. (2018) obtained higher productivity, equal to 189 kg ha^-1 day^-1 of dry matter using dose of 300 m³ ha¹. Andrade et al. (2012) in an experimental area on the campus of Rio Paranaiba UFV evaluated the forage accumulation and managed in the intensive irrigated system. The ideal point of grazing and the chemical composition of forage produced in summer and autumn in pasture managed in intensive system with Tifton 85 grass found herbage accumulation rates of 140.0 kg ha^-1 day^-1 of DM in the summer and 122.2 kg ha^-1 day^-1 of DM in the fall. The height of the ideal grass for grazing was 25.4 cm.

The highest production of dry matter was obtained with an average dose of 2500 m³ ha^-1 (Figure 2a), which was 66.5% higher than the average control, in the period of 105 days. The  linear  increase  of  production  and  grass height influenced the dry matter density, which increased from 92.4 kg ha^-1 cm^-1 DM in witness to 125.2 kg ha^-1 cm^-1 DM at the highest dose (2500 m³ ha^-1), an increase of 34.5% in the grass density (Figure 2c). With increasing doses of the SW, the pH did not show any variation in the two depths (Tables 1 and 4), due to the alkaline characteristics of SW.  According to the results this study, Silva et al. (2014) observed that soil attributes related to the acidity did not suffer influence of successive applications of SW. These differed from Queiroz et al. (2004) who observed a lowering of the pH with the application of SW. According to Bouwer (2000), in soils receiving wastewater, there may be a decrease in pH due to the mineralization of organic compounds of the SW, which facilitates the production of CO₂ and organic acids. Already, Lucas et al. (2013) observed the pH increased from 5.47 to 6.77 in the soil (0 to
0.60 m) at 1015 days of SW application. Increased pH values of soil were consistent with the high pH values of SW used, ranging from 7.08 to 7.70.

The available potassium content (K⁺) in the soil was similar at both depths with 101.0 and 102.0 mg dm^-3 levels (Table 4) and fall within the adequate availability class, according to CFSEMG Ribeiro et al. (1999). A significant increase in K⁺ content at a depth of 0.20-0.40 m, indicated a percolation of K⁺ for the deeper layers, because the soil characterization analysis before applying the SW doses (Table 1) the levels were 94.0 and 56.0 mg dm^-3 in these depths. By being a monovalent, K⁺ has low retention on soil colloids, being susceptible to being leached from the surface layers for the subsoil layers. Which with the application of SW  doses,  the  amount  of added K⁺ was high than extracted by Tifton 85 (Figure 3c), favoring an increase in the levels of K⁺ in the soil (Table 4).

The base saturation (V) showed a significant increase in the depth of 0.20-0.40 m (35.09%) with the application of SW doses (Table 4) in relation to the initial content (30%) (Table 1). This result is related to the increase in potassium and magnesium content at this depth. With the application of SW, the levels of organic matter that were 2.3% at a depth of 0.00 to 0.20 and 1.6% in the depth 0.20-0.40 m (Table 1), did not show any increase due to the application of the treatments (Table 4), with no significant differences. Mattias (2006) did not observe an increase of organic matter in the application of liquid effluent from swine farms.

The lack of response to the increase of organic matter can be explained by Assmann et al. (2006) who observed no increase in organic matter content by applying liquid swine waste. For, according to the authors, they must be considered inherent characteristics of manure used, where the quality of the organic compounds may determine a greater or lesser accumulation in the soil. The organic compounds present in the liquid manure of pigs have been easily digested by oxidizing in a few days or weeks and are favored for higher microbial activity. The microbial biomass is considered a vital part of the organic matter, composed of micro-organisms (bacteria, fungi and actinomycetes), comprising 2 to 5% carbon and up to 5% of total nitrogen (Moreira and Siqueira, 2003). There were no significant differences among treatments in nutrient contents Fe, Cu²⁺ and Zn²⁺ (Table 4). However, for Lucas et al. (2013) and Rosa et al. (2017a, b), the applications of swine wastewater favored the accumulation of copper and zinc.

According  to Girotto (2007), successive applications of SW to the soil cause accumulation of Zn²⁺ in the surface layers that were found significant for the accumulation of Zn²⁺ layer to the depth of 0.10 m. The Cu² + and Zn²⁺ elements are used in animal feed as a supplement, without being fully absorbed and therefore excreted in high amounts, remaining in SW (Rosa et al., 2017b). The high capacity of Oxisol adsorbs Cu and Zn and low mobility of these elements could be verified by (Lopes et al., 2014). Gomes Filho et al. (2001) reported poor copper mobility in soil, stating that this element among the heavy metals is the most strongly adsorbed or complexed by the soil. Furthermore, according to Lopes et al. (2014) high adsorption of Cu²⁺ and Zn²⁺ may be the result of an increase in soluble organic compounds, which have the ability to complex these nutrients. The average leaf content of nutrients N, P, K, S and Zn differed due to the application of increasing doses of SW (Table 5). Thus, the application of different doses of SW influenced the foliar contents of these nutrients.

Knowledge of the leaf content of an intensive system of production, mainly in a very demanding nutritionally species, such as Tifton 85 grass, is very important for determining the amount of nutrient to be restored, reaching thus the mass production desired dry matter. Silva (1999) described the range of suitable concentrations for Tifton as follows, relative to macronutrients: C = 20.0-26.0; P = 1.5-3.0; K = 15.0-30.0; Ca = 3.0-8.0; Mg = 1.5-4.0; S = 1.5-3.0 g kg^-1. For optimal range micronutrient is: B = 5.0-30.0; Cu = 4.0-20.0; Fe = 50.0-200.0; Mn = Zn = 20.0-300.0 and 15.0-70.0 mg kg^-1.

In this experiment the nutrient contents fall within the proper range suggested by Silva (1999), except Zn, wherein the average of treatments was 144 mg kg^-1. The extraction of nutrients by dry weight of the shoot (above 10 cm) was high, demonstrating that the Tifton 85 grass has high nutrient extraction capacity (Table 6). The extraction of macro and micronutrients of shoots of grass Tifton 85 follows the following order: N> K> Ca> P> Mg> S> Mn> Fe> Zn> B> Cu (Table 6). As the DM output followed a linear growth projection (Figure 2a), all nutrients obtained the same trend (Figures 3 and 4).

The dose of 500 m³ SW ha^-1 provided the soil 411.08 kg ha^-1 of N and extraction was 358.01 kg ha^-1 of N, which corresponds to 87.09% of the applied level. These results show that the synchronism between the availability of N  from  SW culture and demand generated a beneficial effect on the nutrient recycling process. This balance can reduce the mineral N concentration in the soil solution, promoting the sustainability of the production system. Surely, this dose is the most appropriate in the environmental point of view with respect to N (Figure 3a). Sharpe and Harper (2002) further justify about 35% of N SW is lost in the form of NH₃ during application. The phosphorus provided by increasing doses of SW was lower than that extracted by the plant in all treatments (Figure 3b), since the nutrient content   in  this  SW  is  low (20.46 mg L^-1). Silva   et   al. (2012) conducted an experiment on the Bonsucesso farm and found that the phosphorus content in SW was reduced by approximately 80% to pass the Biodigester and the lagoon decantation, thus justifying the low content in SW. According to Mattias (2006), the accumulation of P in soils treated with SW correlates with a lower N/P ratio of waste, since it meet the required amounts of N by plants it is necessary to simultaneously apply greater amounts of P resulting in accumulation and organic P moving in the environment. At a dose of 500 m³ ha^-1 potassium  applied  via  wastewater  (254.69 kg ha^-1) was lower than that extracted by the plant (Figure 3c). Higher doses left nutrient residues in the soil (Figure 3c and Table 4).

The calcium extractions (Figure 3d) and magnesium (Figure 3e) by Tifton 85 grass were lower than the amount provided by the SW dose 2.000 m³ ha^-1. In this dose, the SW has provided the soil 103.8 kg ha^-1 of calcium and magnesium 67.06 kg ha^-1 and extracting the Tifton 85 was 86.16 kg ha^-1 of calcium and 55.92 kg ha^-1 magnesium, this being the SW dose sufficient to provide the amount extracted at 85 Tifton. There is a tendency to accumulate copper (Cu) in soil with SW (Figure 4a). Cu average extraction obtained in the dry matter of grass Tifton 85 was 0.2 kg ha^-1. The lowest dose of SW added tem times more than the Cu extracted by forage, while the highest dose added 40 more times. Figure 4 illustrates the low extraction by Cu Tifton 85 grass.

Lima and Miyada (2003) conducted an experiment with Cu in the form of cupric citrate, a more soluble source with greater utilization by the animal and less waste left in the excretions. Thus, they concluded that cupric citrate may replace the Cu sulfate, contributing to reducing this nutrient in the SW. SW in any of the doses could supply the manganese (Mn) needed for the production of Tifton 85 grass (Figure 4c). It is necessary to monitor this nutrient in the soil and the plant, as part of nutrients extracted by grass after being consumed by cattle and returned to the soil through animal waste. Zn supplied by the dose of 500 m³ ha^-1 was 2.86 kg ha^-1 approaches the extracted amount of and 2.72 kg ha^-1 at Tifton 85 (Figure 4d). The view on organic waste recycling needs to be diversified, being recovery and recycling of nutrients from organic wastes are a possible solution. When organic waste recycling is complemented by nutrient extraction, some nutrient loops within society can be closes, enabling more sustainable agricultural production in future (Kirchmann et al., 2017).