Introduction

In 2012, the industrial swine production in Brazil comprised 37.63 million animals, of which 59 % was produced in the southern region (Abipecs 2012). This high number of animals generates large waste volumes that are usually used as nutrient source for grain crops or pastures, normally in the proper swine production unit. According to Oliveira (1994), each swine produces an average of 7.0 L day−1 of slurry, i.e., approximately 155,000 m3 day−1 of swine effluent are produced in southern Brazil.

The use of swine effluent (SE) with significant amounts of phosphorus (P) as nutrient source can be an alternative to minimize the use of processed phosphate fertilizers, as the natural reserves of P will decrease in the coming decades (Cordell et al. 2009). In addition, the use of SE is a way to promote P cycling on farms (Girotto et al. 2013). However, swine farms in southern Brazil are typically small, and the application of SE favors repeated and excessive applications on the same area (Kang et al. 2011), causing P accumulation in soil surface layers, particularly in no-till cropping systems (Adeli et al. 2008; Payet et al. 2009; Guardini et al. 2012; Lourenzi et al. 2013).

Phosphorus accumulation in the soil may saturate surface functional groups of reactive soil particles, increasing P concentration in the soil solution and also in runoff water (Ceretta et al. 2005; Smith et al. 2007; Kleinman et al. 2009; McDowell et al. 2011; Hahn et al. 2012). P transfer in runoff occurs in soluble and particulate forms. According to Haygarth and Sharpley (2000), soluble or dissolved P is the solution filtered through <0.45 μm, and particulate P is the difference between total P in the unfiltered sample and soluble P (filtered through <0.45 μM). The transfer of P and sediment in surface runoff from cultivated soils are major sources of potential contamination of surface waters adjacent to croplands (Jiao et al. 2011; Hart and Cornish 2012; Sweeney et al. 2012).

In the areas treated with swine effluent applications, the risk of runoff is variable and depends on several factors, such as the rate and frequency of swine effluent applications, soil nutrient concentrations, amount and frequency of rainfall, type of cropping system, slope, soil texture, soil depth, effective field capacity, drainage class, as well as other less important factors, addressed by several authors in numerous studies about P in the world (Yu et al. 2006; Meng et al. 2008; Flores et al. 2012; Sweeney et al. 2012; Wu et al. 2012; Wang et al. 2013). The combination of these factors makes an area more or less susceptible to solution and nutrient transfer in runoff, being very important for the assessment of environmental contamination risks in areas fertilized with swine manure. In southern Brazil, where this study was developed, the prevailing management system is no tillage in more than 80 % of the entire cropland area (Amado et al. 2006), i.e., in these areas, pig slurry is predominantly applied to the soil surface and is not incorporated into the soil, which can increase nutrient loss by runoff.

This study aimed to evaluate P accumulation in the soil and transfer of P forms in surface runoff in a Hapludalf soil under no-tillage subjected to successive swine effluent applications.

Material and methods

Site description, treatment, and evaluations

The experiment was carried out at the Federal University of Santa Maria (RS), Brazil (latitude S 29° 43′, longitude W 53° 42′) in a Typic Hapludalf soil (Soil Survey Staff 1999), with a sandy loam surface texture and a 4 % slope. The area had been cultivated under no-tillage for 8 years until 2000 when this experiment began. In March 2000, the 0–10 cm layer of the area was sampled. The results of the analyses were the following: clay, 170 g kg−1; silt, 300 g kg−1; sand, 530 g kg−1; pHH2O, 4.7; soil organic matter, 16 g dm−3; Al, 0.8 cmolc dm−3 (extracted by 1 mol L−1 KCl); Ca, 2.7 cmolc dm−3 (extracted by 1 mol L−1 KCl); Mg, 1.1 cmolc dm−3 (extracted by 1 mol L−1 KCl); P, 15.0 mg dm−3 (extracted by Mehlich 1, 0.05 mol L−1 HCl + 0.0125 mol L−1 H2SO4); and K, 96 mg dm−3 (extracted by Mehlich 1).

From May 2000 to January 2008, the application of swine effluent rates of 0, 20, 40, and 80 m3 ha−1 was tested on a total area of 12 m2 (4 m × 3 m) in an experiment with a randomized block design with three replications. The rates were applied to the soil surface before each crop in the following sequence: black oats (Avena strigosa)/maize (Zea mays L.)/oilseed radish (Raphanus sativus L.) in 2000/2001 and 2001/2002, black oats/pearl millet (Pennisetum americanum L.)/black beans (Phaseolus vulgaris L.) in 2002/2003, black oats + common vetch (Vicia sativa L.)/maize in 2003/2004 and 2004/2005, black oats/black beans/sunn hemp (Crotalaria juncea L.) in 2005/2006, and black oats/maize/black oats in 2006/2007, with a total of 19 swine effluent applications.

After grain harvest in the summer crops or flowering in the winter crops, SE was applied for the next crop on top of the leftover crop residues. Swine effluent was broadcast over the soil surface 1 day before sowing of each species, without incorporation into the soil, as usually done by producers in southern Brazil. Maize, pearl millet, and black beans were grown in spring and summer as cash crops. The cover crops, black oats, and common vetch were grown in the winter and oilseed radish and sunn hemp in spring. All stubble produced was left on the soil surface. For this study, evaluations were conducted from 2004 to 2007, and during this period, seven applications of SE were performed (Table 1).

Table 1 Characteristics of the swine effluents (SE) applied before each crop and quantify of phosphorus applied in each rate of swine effluents
Full size table

In the last six crops (black oats/black beans/sunn hemp (2005/2006) and black oats/maize/black oats (2006/2007)), the runoff water volume and levels of available, soluble, particulate, and total P transfer in surface runoff were evaluated. For this, polyvinyl chloride (PVC) boxes (0.75 m × 0.50 m) were installed on each plot. The PVC structure was inserted into the soil to a depth of 5 cm. The surface runoff was assessed with a hard-wired collecting gutter connected to a hose coupled to a collection container (20 L) (see Fig. 1). After each natural rainfall (not simulated) with sufficient runoff volume for analysis, the stored runoff volume was removed from the collectors, the volume measured, and a sample (± 300 mL) taken to the laboratory for analysis.

Fig. 1
figure 1

Collection system solution by runoff

Full size image

Phosphorus analysis

The samples were separated in two fractions: One (approximately 50 mL) was immediately filtered through a 0.45-μm membrane and another fraction was not filtered. In the filtered fraction, the soluble P concentrations, and the unfiltered fraction, the concentrations of available and total P were determined. Concentrations of soluble P in the runoff water were determined directly in the solution filtered through a 0.45-μm membrane. For extraction of available P, 50 mL of the collected field samples was placed in 100-mL snap-cap bottles with 1 mL of 0.11 mol L−1 HCl (PA) (extracting solution). The samples were shaken intermittently for 10 min and then left to stand for 16 h.

The total P of the runoff water was determined by acid digestion (H2SO4 + H2O2) in the presence of saturated MgCl2, according to the method of Brookes and Powlson (1982). The concentrations of particulate P were determined as the difference between the amounts of total P and soluble P. The amounts of soluble, available, and total P were determined by colorimetry (Murphy and Riley 1962). For the determination of total P in SE, samples were dried to constant mass in a laboratory oven at 65 °C. After drying, the samples were digested as described by Tedesco et al. (1995) and the P content was determined by the method of Murphy and Riley (1962). The dry matter content and total amounts of P applied in SE are shown in Table 1.

During the study, the soil was sampled twice. The first sampling occurred in October 2004, after 12 SE applications, in the soil layers 0–2.5, 2.5–5, 5–10, 10–20, 20–40, and 40–60 cm. The second was performed in the same soil layers, in January 2008, after 19 SE applications. Soil samples were analyzed using Mehlich 1 (Tedesco et al. 1995), and the extracts were analyzed for available P. At the end of the extraction process, P content was determined by the method of Murphy and Riley (1962).

Statistical treatment of the data

Analysis of variance (ANOVA) was performed to determine treatment effects on runoff and soluble P, particulate P, and total P losses in runoff. Differences between treatment effects were determined by the Scott Knott test (α = 0.05). Pearson statistical tests were performed to test correlations between cumulative amounts of P applied with SE, cumulative volume runoff, cumulative dry matter production by crops, and cumulative losses of available P, soluble P, particulate P, and total P in runoff in the 2005/2006 and 2006/2007 growing seasons. All statistical analyses were performed using SISVAR (Ferreira 2008).

Results

Soil available P

The successive SE applications promoted an increase in available P levels in the soil. In all soil layers, the levels were highest after the application of 80 m3 ha−1 of swine effluent, which demonstrates the migration of P into the soil profile (Fig. 2). The most significant increases were observed in the surface layer (0–2.5 cm) after the first 12 SE applications. Compared to the control, the increases in the 0–2.5 cm layer were 652, 1372, and 2827 %, with the addition of 161, 321, and 643 kg P ha−1 at rates of 20, 40, and 80 m3 ha−1 of swine effluent, respectively. In contrast, in the last seven SE applications, 464, 928, and 1856 kg P ha−1 were applied in 20, 40, and 80 m3 ha−1 of swine effluent, and the increases were 35, 25, and 60 %, respectively.

Fig. 2
figure 2

Levels of available P, extracted by Mehlich 1, in layers of soil after 12 (a) and 19 (b) swine effluents applications. The regions 1, 2, 3, 4, and 5, respectively, mean fertilization group, critical level for the plants, land use range as a means of recycling waste, high environmental risk category, and range of environmental pollution, according to Gatiboni et al. (2014)

Full size image

In southern Brazil, soil fertility under no-tillage is diagnosed in the 0–10 cm layer (CQFS-RS/SC 2004). In this layer, the increases in available P levels exceeded 2000 and 5000 % after 12 and 19 SE applications of 80 m3 ha−1, respectively (Table 2). However, considering the levels of available P in soil after 12 in comparison with the levels after 19 SE applications, the increases were less than 200 % with SE application of 80 m3 ha−1 (Table 2).

Table 2 Levels and percentage increases of available P in 0–10 cm layer of soil after 12 and 19 applications of swine effluents (SE)
Full size table

Water transfer in surface runoff

Water transfer in surface runoff varied in response to the different rainfall events in all seasons evaluated (Fig. 3). In addition, the rainfall/runoff ratio showed that, normally, the runoff water volume decreased with increasing SE rates (Fig. 4). Furthermore, water transfer in surface runoff was correlated with rainfall volumes for black oat in 2005, black beans in 2005/2006, maize in 2006/2007, and black oat in 2007 (Fig. 4). However, this was variable and depended on precipitation intensity, the time elapsed after the last rainfall, and soil moisture at the time of rainfall. The surface runoff, in general, tended to decrease with increasing SE rates (Table 3), as is evidenced by the high correlation between these variables (Table 4).

Fig. 3
figure 3

Volume of rainfall and surface runoff solution for black oat in 2005 (a), black beans in 2005/2006 (b), sunn hemp in 2006 (c), black oat in 2006 (d), maize in 2006/2007 (e), and black oat in 2007 (f) in a soil under swine effluents applications. P5 the means differ by the Scott-Knott test (α = 0.05), n.s. not significant

Full size image
Fig. 4
figure 4

Ratio of rainfall/runoff solution for black oat in 2005 (a), black beans in 2005/2006 (b), sunn hemp in 2006 (c), black oat in 2006 (d), maize in 2006/2007 (e), and black oat in 2007 (f) in a soil under swine effluents applications. P5 the means differ by the Scott-Knott test (α = 0.05), PC Pearson correlations, n.s. not significant; *p value < 0.001

Full size image
Table 3 Amount of P applied to the soil through swine effluents (SE) applications, volume of rainfall, volume of surface runoff and losses of available P, soluble P, particulate P, and total P by surface runoff in a soil under swine effluents application
Full size table
Table 4 Pearson correlations between cumulative amounts of P applied with swine effluents (PSE), volume runoff (Runoff), dry matter production by crops (DM), loss of available phosphorus (Pav), loss of soluble phosphorus (Psol), loss of particulate phosphorus (Ppart), and loss of total phosphorus (Ptotal) by runoff in agricultural years 2005/2006 and 2006/2007
Full size table

The greatest water loss was observed in the maize crop of 2006/2007, and in contrast, the smallest transfer occurred in sunn hemp (2006) (Table 3). In both cases, the rainfall volume was determinant for the runoff volume, as the largest amounts of rainfall were observed for maize (2006/2007) and the smallest amounts for sunn hemp (2006) (Table 3). This is clear when Fig. 3c (sunn hemp-2006) is compared with Fig. 3a, b, and d–f, where the lowest level rainfall resulted in smallest runoff volume in the three events evaluated. On average, during the 2-year evaluation period, solution transfer in surface runoff was 27.2, 21.7, 15.6, and 8.2 % of the total rainfall at SE rates of 0, 20, 40, and 80 m3 ha−1, respectively.

Forms of P transfer in runoff

The highest concentrations of P forms (available P in Fig. 5, soluble P in Fig. 6, particulate P in Fig. 7, and total P in Fig. 8) were observed in the first rainfall events after applying SE and there was an inverse correlation with surface runoff volume (Table 4), e.g., with the events after 29 and 34 days of black oat crops of 2005 and 2006, respectively. In contrast, when rainfall levels were high and when a large runoff volume was produced (Fig. 3), the P concentrations of all forms (Figs. 5, 6, 7, and 8) were relatively low. This was the case in maize (2006/2007) and black oat (2007), where there were the largest number of events with high rainfall and large runoff volumes was largest (Fig. 3e, f), with P concentrations not exceeding 30 mg L−1 of total P in all events (Figs. 5e, f; 6e, f; 7e, f; and 8e, f).

Fig. 5
figure 5

Concentration of available P in the surface runoff solution for black oat in 2005 (a), black beans in 2005/2006 (b), sunn hemp in 2006 (c), black oat in 2006 (d), maize in 2006/2007 (e), and black oat in 2007 (f) in a soil under swine effluents applications. P5 the means differ by the Scott-Knott test (α = 0.05), n.s. not significant

Full size image
Fig. 6
figure 6

Concentration of soluble P in the surface runoff solution for black oat in 2005 (a), black beans in 2005/2006 (b), sunn hemp in 2006 (c), black oat in 2006 (d), maize in 2006–2007 (e), and black oat in 2007 (f) in a soil under swine effluents applications. P5 the means differ by the Scott-Knott test (α = 0.05), n.s. not significant

Full size image
Fig. 7
figure 7

Concentration of particulate P in the surface runoff solution for black oat in 2005 (a), black beans in 2005/2006 (b), sunn hemp in 2006 (c), black oat in 2006 (d), maize in 2006/2007 (e), and black oat in 2007 (f) in a soil under swine effluents applications. P5 the means differ by the Scott-Knott test (α = 0.05), n.s. not significant

Full size image
Fig. 8
figure 8

Concentration of total P in the surface runoff solution for black oat in 2005 (a), black beans in 2005/2006 (b), sunn hemp in 2006 (c), black oat in 2006 (d), maize in 2006/2007 (e), and black oat in 2007 (f) in a soil under swine effluents applications. P5 the means differ by the Scott-Knott test (α = 0.05), n.s. not significant

Full size image

The P concentrations were lowest in the solutions where no SE was applied (SE rate of 0 m3 ha−1), and values did not exceed 5 mg L−1 for all P forms. The P levels in the surface runoff after a SE application of 20 m3 ha−1 was also low; only in the events after 8 and 34 days on black oat (2006), the values exceeded 20 mg L−1 in available and total fractions (Figs. 5d and 8d). In contrast, the P levels in surface runoff after SE applications of 40 and 80 m3 ha−1 were high, exceeding, in many events, a concentration of 30 mg L−1 P. The P concentration in the runoff was highest after a rainfall after 34 days of black oat cultivation (2006) in soil treated with 80 m3 ha−1 of SE (88, 81, 22, and 103 mg L−1 P for available, soluble, particulate, and total fractions).

The transfer of total P was highest in the 2006/2007 growing season (Table 3). On maize (2006/2007) and black oat (2007), rainfall was high (Table 3), resulting in total P transfer in surface runoff that was 2.8, 3.0, 2.0, and 4.6 times higher in the 2006/2007 than the 2005/2006 growing season at the four tested SE rates.

On average, in the 2-year evaluation, total P transfer in surface runoff was 9.0, 6.1, and 3.7 % of the total applied in the SE rates of 20, 40, and 80 m3 ha−1, respectively. Of the total transfer, 45, 70, 62, and 66 % was in the available form (extracted with 0.1 mol L−1 HCl); 31, 57, 49, and 52 % in the soluble form; and 69, 43, 51, and 48 % in the particulate form in the SE rates of 0, 20, 40, and 80 m3 ha−1.

Loads of P transfers in runoff

The cumulative transfer of P forms in surface runoff was correlated with the phosphorus applied with SE (Table 4); this factor is important to consider in areas treated with swine manure. For example, in agricultural years 2006/2007, the highest transfer forms of P were observed at the highest rates of swine effluent (Table 3). Furthermore, cumulative transfer of P forms was inversely correlated with cumulative transfer of runoff (Table 4). For example, in the soil treated with 20 m3 ha−1 of SE, there was a transfer in runoff of 5375 m3 ha−1 observed in the 2006/2007 growing season, providing a 30.75 kg ha−1 transfer of total P, while a transfer of 2276 m3 ha−1 observed in the soil treated with 80 m3 ha−1 of SE resulted in a transfer of 50.35 kg ha−1 available P (Table 3).

Discussion

Effect of swine effluent applications on available P in soil

Increases in the levels of available P, especially in the soil surface layers (Fig. 2), occurred due to the high amount of P added to the soil, the form of P found in the SE and the soil management system. Manures have high concentrations of nutrients, including of P (Guo and Li 2012; Ye et al. 2014), contained in manures in an inorganic form (Cassol et al. 2001). This form is readily available to plants, and this is the predominant form of P accumulation in manure-fertilized soils (Chardon et al. 2007; Adeli et al. 2008; Gatiboni et al. 2008; Lourenzi et al. 2013).

In addition to the increment in available P in the upper soil layers, an increase of available soil P in the 40–60 cm layer was also observed after 19 applications of 40 and 80 m3 ha−1 of pig slurry. This shows the increased migration of P in the soil profile when the application of phosphates doses exceeded the amounts exported by the system (Lehmann et al. 2005). Phosphorus added via the slurry will gradually saturate the adsorption sites of the inorganic soil colloids, particularly the oxides, increasing the amount of P in the soil solution. In addition, the dissolved organic compounds can compete for the adsorption sites of the phosphate anion, decreasing its adsorption capacity to soil colloids, which increases the P concentration in surface runoff solution (Holford et al. 1997).

At the end of this study, 19 SE applications had been made, and with the 80 m3 ha−1 rate, 2500 kg ha−1 P was applied (Lourenzi et al. 2013). This resulted in an available P content in the 0–10 cm soil layer of 754 mg dm−3. This level is 62 times higher than the one observed in the same soil layer when no swine effluent was applied and was much higher than the level considered excessive (>42 mg of available P kg−1 in soil with ≤20 % clay) by the Committee of Chemistry and Soil Fertility of the States of Rio Grande do Sul and Santa Catarina in southern Brazil (CQFS-RS/SC 2004). These results show that successive SE applications promoted excessive P availability in the soil (at levels above those needed for normal plant growth) and that P can be transported by runoff, which raises the risk of eutrophication of surface waters adjacent to cultivated areas (Smith et al. 2001; Daverede et al. 2004; Ceretta et al. 2010b; Hahn et al. 2012).

In this sense, Gatiboni et al. (2014) developed a study to determine the risk of environmental contamination by P in soils treated with SE applications in southern Brazil. For this, these authors considered the levels of available P (extracted by Mehlich 1) in different clay classed of soils (<20, 20–40, 40–60, 60–80, and >80 % clay) and attributed five regions of interpretation of available P levels (Fig. 2). Region 1 represents the P levels where fertilization is necessary for an adequate plant development, region 2 represents adequate P levels in soil for adequate plant development, region 3 represents the levels of P with a low risk of environmental contamination, region 4 represents levels of available P of soil with high environmental risk, and region 5 represents environmental pollution.

In this study, the soil contained <20 % clay, and for this clay class, Gatiboni et al. (2014) determined a threshold of 59 mg P L−1 as limit of recycling waste with low risk of environmental contamination (limit region 3 to region 4). Considering this value, after 19 SE applications, all rates induced P levels with high environmental risk to a depth of 10 cm, and at SE rates 40 and 80 m3 ha−1, the P levels represent environmental pollution to a depth of 20 cm (Fig. 2b). These results indicated that for use of swine effluent as nutrient source, it is very important to consider the nutrient content in the soil, especially of P, because the use of high SE rates can raise the P levels in soil and the environmental risk.

Effect of swine effluent applications in solution transfer in surface runoff

The smaller transfer of surface runoff with increasing SE rates can be explained by the high nutrient input at higher application rates. This increases nutrient levels in the soil (Adeli et al. 2008; Scherer et al. 2010; Lourenzi et al. 2013) and provides the capacity for greater dry matter production by crops (Table 1). As a consequence, water infiltration in the soil profile under a no-till cultivation system is favorable (Ceretta et al. 2005; Krutz et al. 2009).

Solution transfer in surface runoff is directly associated with rainfall volume (Ceretta et al. 2010b). However, even in treatments of 80 m3 ha−1 of swine effluent with a high rainfall volume, the solution transfer was small (Fig. 3) and the rate was justified by higher dry matter production of the crops (Table 1). Furthermore, the use of crop rotation also has significant effects on reducing surface runoff, as observed by Jiao et al. (2011). These authors found increases in water infiltration into the soil and up to 58 % reduction in runoff volume in a study evaluating different crop rotations. Another factor that interacts with solution transfer in surface runoff is soil moisture (Vadas et al. 2011), which may fluctuate according to the distribution of rainfall throughout crop cycles. In years with higher rainfall, as in the crops of maize (2006/2007) and black oat (2007), the soil remains moist with much of its porosity occupied by water, hindering water infiltration and favoring surface runoff. High soil moisture may result in an increased runoff volume for each event (Jiao et al. 2011), as well as an increase in the number of runoff events (Udawatta et al. 2006).

Effect of swine effluent applications in forms of P transfer in runoff

The highest concentrations of P forms in the runoff were observed in the first rainfall events after SE applications. This is because most of the P present in SE is in inorganic form (Cassol et al. 2001). Moreover, organic P present in SE is rapidly mineralized after soil application (Gatiboni et al. 2008). Thus, the levels of available P in the soil rapidly increase (Adeli et al. 2008; Ceretta et al. 2010a; Guardini et al. 2012). Furthermore, the presence of organic acids in manure can saturate the adsorption sites of surface functional groups, increasing the availability of soil P (Andrade et al. 2003).

The highest concentrations of soluble P in the first rainfall events after SE application may also be due to the reaction of P with soil particles or the formation of co-precipitates with metals, such as Cu, which would decrease P concentrations in the solutions of subsequent events. Furthermore, soil surface sealing, occurring with the addition of large amounts of swine effluent, may favor higher concentrations of phosphorus in the runoff, especially in early rainfall events after SE application, as noted by Bundy et al. (2001) and Allen and Mallarino (2008).

Concentrations of available P in the surface runoff, in most events, exceeded the maximum permissible concentration for drinking water (higher than 0.15 mg L−1) and was class 3 (class 3 is water that can be used for human consumption after proper treatment) (Conama 2005). Available P values were as high as 88.3 mg L−1 with the application of 80 m3 ha−1 of SE (Fig. 5). These results show the high pollution potential of SE applications on the environment, especially in the first rainfall events after application when the P concentrations in the solution can reach extremely high values. This situation increases eutrophication of waters, as reported by Smith et al. (2007), Allen and Mallarino (2008), Ceretta et al. (2010b), Hahn et al. (2012), and Wang et al. (2013). The concentrations of particulate P were higher in the runoff with high rates of SE application (Fig. 7), and this may be due to the transport of P-enriched organic particles in the runoff (Bertol et al. 2010).

Amounts of P transfer in runoff

The amounts of forms of P transferred by surface runoff were correlated with rainfall occurring during the period of evaluation. This is because in 2005/2006, the rainfall volume was 7317 m3 ha−1 lower than in the 2006/2007 growing season (Table 3). This resulted in smaller amounts of P forms transferred by surface runoff, especially at the SE rate of 80 m3 ha−1, which showed similar transfer of P forms as the rate of 20 m3 ha−1 (Table 3). However, in the 2006/2007 growing season, when larger rainfall volumes were observed, the amounts of P forms transferred by surface runoff were higher, with the largest transfer being observed at the highest SE rate (Table 3).

Transfer of total P in surface runoff was in the mean 9.0, 6.1, and 3.7 % of the totally applied SE at rates of 20, 40, and 80 m3 ha−1, respectively. These results are similar to those of Bertol et al. (2010), who reported that P transfer in runoff was less than 15 % of the total applied P in SE and mineral fertilizer. Sharpley et al. (1994) stressed that low P transfer, typically less than 5 % of the total applied, is usually observed in conservation systems. However, in the 2006/2007 growing season, the amounts of total P transferred reached 50 kg ha−1 after the SE application of 80 m3 ha−1 (Table 3). These quantities are high, and in soils treated with successive manure applications, practices to increase water infiltration into the soil are required, e.g., crop rotation and no tillage systems.

Conclusions

The high amounts of P applied in swine effluent caused significant increases in available P levels in the soil, resulting in transfer in runoff in the following order: available P > soluble P > particulate P. The result indicated that successive swine effluent applications raised the environmental risk of runoff P, and in areas treated with swine effluent over several years, it is highly recommended to consider the nutrient levels in the soil at the time of setting the swine effluent rates to be applied.