Nitrogen and phosphorus leaching in a tropical Brazilian soil cropped with sugarcane and irrigated with treated sewage effluent

Agricultural Water Management 117 (2013) 115–122

Contents lists available at SciVerse ScienceDirect

Agricultural Water Management journal homepage: www.elsevier.com/locate/agwat

Nitrogen and phosphorus leaching in a tropical Brazilian soil cropped with sugarcane and irrigated with treated sewage effluent Julius Blum a,b,∗ , Adolpho José Melfi a,b , Célia Regina Montes b,c , Tamara Maria Gomes d a

Departamento de Ciência do Solo, Escola Superior de Agricultura Luiz de Queiroz (ESALQ), Universidade de São Paulo (USP), P.O. Box 09, Piracicaba (SP), Brazil Núcleo de Pesquisa em Geoquímica e Geofísica da Litosfera (NUPEGEL), USP, Piracicaba (SP), Brazil Laboratório de analise ambiental e Geoprocessamento, Centro de Energia Nuclear na Agricultura (CENA), USP, Piracicaba (SP), Brazil d Departamento de Engenharia de Alimentos, Faculdade de Zootecnia e Engenharia de Alimentos da USP, Pirassununga (SP), Brazil b c

a r t i c l e

i n f o

Article history: Received 24 January 2012 Accepted 17 November 2012 Available online 12 December 2012 Keywords: Drainage Groundwater pollution Wastewater Error propagation Soil solution

a b s t r a c t There are concerns about groundwater contamination with N and P from fertilizers and other anthropogenic wastes. Use of treated sewage effluent (TSE) for crop irrigation can reduce the use of mineral fertilizers; however, it may add more nutrients into the soil than are necessary for crops, increasing the possibility of leaching. Thus, knowledge of nutrient dynamics in TSE irrigated soils is important for the safe use of this resource. However, the reliability of studies regarding ion leaching is limited due the high propagated variance, as these studies involve independent measurements of variables related to soil and soil solution. The objective of this research was to quantify P and N leaching in a TSE-irrigated Brazilian soil and identify the main causes of variance of this quantification. The experiment consisted of a treatment without irrigation and treatments with TSE irrigation to meet 100% and 150% of the crop water demand (CWD). Soil physical properties and soil water potential gradient were used to calculate internal drainage, and nutrient concentration was measured in soil solution samples taken with ceramic suction cups at a depth of 1 m. Variance propagation was calculated by linearization, and the contribution of each variable to the total variance was isolated and quantified. Irrigation with TSE increased N leaching; however, when applied in dosages that met 100% of the CWD, it did not threaten the groundwater quality. P leaching was as low as 100 g ha−1 and was therefore not an environmental concern. N leaching can be estimated considering the total N input and the rainfall; however, long-term data are needed to improve the accuracy of this estimation. The variance propagation of the soil water potential measurements represented up to 70% of the nitrogen leaching variance. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Waste produced in urban areas and agricultural fertilizers are the major contributors that raise the concentration of nitrogen (N) in the groundwater (Serhal et al., 2009). Crop irrigation with treated sewage effluent can replace commercial fertilizers, providing economic benefits; however, such treatment may add more N into the soil than is necessary for the crop (Leal et al., 2010). The overapplied N can easily be leached below the root zone due to fast nitrification in well-aerated soils and weak interactions between N-nitrate and soil colloids. The leached N is transported into bodies of water, causing eutrophication and making the water unfit for human consumption, as well as affecting marine species (Bond, 1998). However, special focus must be given to the phosphorus (P) input into water bodies because it is usually the limiting nutrient,

∗ Corresponding author at: Avenida Mister Hull 2977, 60021-970 Fortaleza, CE, Brazil. Tel.: +55 8533669452; fax: +55 8533669690. E-mail addresses: [email protected], [email protected] (J. Blum). 0378-3774/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.agwat.2012.11.010

and P control is of prime importance in reducing the accelerated eutrophication of fresh water (Sharpley et al., 1994). Thus, there are concerns regarding groundwater contamination with N and P from fertilizers and other anthropogenic wastes. The magnitude of the nutrient losses by leaching in soil systems is proportional to the concentration of nutrients in the soil solution and the amount of drained solution (Ghiberto et al., 2009). Irrigation with wastewater can increase both the concentration of nutrients (Gloaguen et al., 2007) and the volume of drained solution (Barton et al., 2005), thus increasing the possibility of nutrient leaching into the groundwater. However, major fluxes in nutrient leaching are observed when the input or mineralization of nitrogen does not coincide with the nutrient uptake by plants (Oliveira et al., 2002; Sieling and Kage, 2006). Thus, we hypothesize that the higher evapotranspiration rates during the peak growth of the crop increases the necessity for irrigation, synchronizing nutrient input and nutrient uptake. Soil variability is the main problem related to the assessment of N leaching (Addiscott, 1996). The extent of all biological, chemical and physical processes responsible for N leaching vary due to soil

116

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

type (Barton et al., 2005), and different soil structures can accelerate the movement of solution to depths where the roots are unable to reach (McLeod et al., 1998). Therefore, spatial variations in the concentration of nutrients or in physical properties of the soil increase the variance of nutrients leaching, diminishing the reliability of the results. Ceramic cups have being commonly used to extract the soil solution in field experiments (Sieling and Kage, 2006; Ghiberto et al., 2009); however, Addiscott (1996) advised that this method is more reliable in sandy soils, because the soil space variability increases with the clay content. Clayey soils have higher range of water availability categories than sandy soils, and the soil matrix is by-passed by cracks and other channels. Besides, the volume of the soil from which the cups extract water is larger in sandy soils, resulting in a better space representativeness. The use of the suction cup method to extract the soil solution and the soil water potential gradient to estimate soil drainage require measurement of the following: (i) ion concentrations in the soil solution; (ii) soil physical properties; and (iii) the soil water potential at different layers. These measurements contain independent errors that must be propagated to calculate the ion leaching variance. The variability of the physical properties of the soil and the ion concentrations in the soil solution are widely described in the literature; however, the influence of the error of each measure on the composition of the ion leaching variance is not well known. The knowledge of the main source of variation will be useful for controlling the error. Because the knowledge of the N and P dynamics in wastewater irrigated systems is important for the safe use of this resource and because insufficient data regarding these issues have been produced (Duan et al., 2010; Sharpley et al., 1994), the objective of the present study was to quantify N and P leaching and identify the main cause of its variance during two years of sugarcane cultivation irrigated with treated sewage effluent (TSE) on an Oxisol soil located in southeastern Brazil.

2. Materials and methods The experiment was carried out in Lins County, São Paulo State, Brazil (latitude: 21◦ 38 56 S, longitude: 49◦ 44 43 W, altitude 422 m). The soil of the experimental plots was classified as Typic Haplustox (Soil Survey Staff, 1999), sandy clay loam (770 g kg−1 of sand and 140 g kg−1 of clay at the 0–0.2 m layer and 710 g kg−1 of sand and 210 g kg−1 of clay at the 0.2–0.8 m layer); the mineralogy was predominantly composed of quartz and kaolinite and subordinately of hematite, magnetite and maghemite (59 g kg−1 Al2 O3 and 23 g kg−1 of Fe2 O3 at a depth of 0–1 m). The F test showed no significant difference between treatments for sand and clay content, assuming homogeneity between experimental plots. A sugarcane crop (cultivar RB 72454) was planted in March 2005 and harvested every September from 2006 to 2010. N, P and potassium (K) fertilization was carried out every year after the harvest, following the regional recommendation (Raij and Cantarella, 1996). Because TSE irrigation is a source of N (Da Fonseca et al., 2007), during the studied period (2009–2010), half of the mineral nitrogen dosage suggested, 50 kg ha−1 year−1 of N, was applied as ammonium nitrate. In addition, 60 kg ha−1 year−1 of K was applied as potassium chloride, and 13 kg ha−1 of phosphorus was applied only in 2010 as triple superphosphate. Treatments consisted of the following: (i) without irrigation (WI), (ii) irrigation at 100% of crop water demand (CWD) (T100) and (iii) irrigation at 150% of CWD (T150). Three replications of each treatment were performed. Irrigation management was based on critical soil water tension at the 0–0.6 m soil layer as monitored by tensiometers. The irrigation was performed every time the soil matrix potential reached −40 kPa and was carried out

for sufficient time to raise the soil water potential to −10 kPa at T100, and the time of irrigation at T100 was multiplied by 1.5 at T150. The TSE used in the experiment resulted in the following analytical data in mg L−1 : dissolved organic carbon (33.7 ± 31.8), alkalinity as HCO3 − (302 ± 28), Cl− (57.7 ± 10.6), P (3.6 ± 2.9), N-NH4 + (21.0 ± 10.2), N-NO3 − (0.03 ± 0.1), Al (0.02 ± 0.01), Fe (0.12 ± 0.05), K (18.7 ± 13.4), Ca (7.90 ± 1.32), Mg (1.96 ± 0.58), Na (112.3 ± 43.14), S (51.2 ± 30.0), B (0.1 ± 0.02), Mn (0.02 ± 0.01) and Zn (0.02 ± 0.01). Cadmium, Cr, Cu, Ni and Pb were not detected. The electric conductivity was 0.89 ± 0.12 dS m−1 , the pH was 7.7 ± 0.3 and the calculated sodium adsorption ratio was 9.3 ± 3.2 (Blum et al., 2012). Inputs of N and P at T100 were 249 and 33 kg ha−1 , respectively. More detailed information about the experiment implementation, climate, irrigation management and wastewater treatment are reported in Leal et al. (2009). Measurements of saturated hydraulic conductivity, bulk density and soil water content at 0, −0.5, −2, −6, −10, −30, −100 and −1500 kPa soil water potentials were carried out using soil cores taken in triplicate at depths of 0.1, 0.3, 0.5, 0.7 and 0.9 m in all experimental plots. Residual and saturated volumetric water content, the ˛ and n empiric parameters of van Genuchten equation (Eq. (1)) (Van Genuchten, 1980), were fitted using Nonlinear Least Squares (nls) function with R software (R Development Core Team, 2008). Tensiometer readings were performed every 2 or 3 days at depths of 0.1, 0.3, 0.5, 0.7, 0.9 and 1.1 m. Soil volumetric water content down to a depth of 1 m was calculated using the van Genuchten equation and soil water potential at depths of 0.1, 0.3, 0.5, 0.7 and 0.9 m. i = r +

(s − r ) [1 + (˛ ×

i)

(1)

n 1−(1/n)

]

where  i is the volumetric soil water content at the depth i,  r is the residual water content,  s is the saturated water content, i is the soil water potential, and ˛ and n are empiric parameters of the van Genuchten equation. The soil solution flux density at a depth of 1 m was calculated using the Darcy–Buckingham equation (Eq. (2)), for which the hydraulic conductivity as a function of the volumetric water content was estimated with the Mualen–van Genuchten equation (Eq. (3)) (Van Genuchten, 1980), and the soil water potential gradient was calculated as the difference between the soil water potential measurements at depths of 0.9 ( a ) and 1.1 ( b ) m. The average between the soil water content at 0.9 and 1.1 m was considered to be the soil water content at 1 m (Eq. (4)). qw =



k m

−k(m ) ×  L



 = ks

×

 m =

m − r s − r

⎧ ⎨

0.5



1− 1−

⎩

a + b 2

(2)



m − r s − r

⎫ 1/(1−1/n) 1−1/n ⎬2 ⎭

(3)

 (4)

where qw is the soil solution flux density, k() is the hydraulic conductivity as a function of the volumetric water content,  m is the mean soil water content between depths a (0.9 m) and b (1.1 m),  is the soil water potential gradient, L is the depth difference between the measurement depths, ks is the saturated hydraulic conductivity, and  a and  b are the soil volumetric water content at depths of 0.9 m and 1.1 m, respectively.

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

The soil solution was sampled every other month at a depth of 1 m using ceramic suction cups for nitrate, ammonium, nitrite and phosphate quantification. Ion (N and P) fluxes were estimated from August 2008 to July 2010 by integrating the product of the soil solution flux density and the soil solution ion concentration (Eq. (5)).



tn

qN =

qw CN dt

(5)

t0

where qN is the ion flux (nitrogen or phosphorus) and CN is the soil solution ion concentration. The propagated standard deviation of the N and P flux ((qN )) was calculated by linearization of the variance of each component of the ion flux equation (Eq. (5)), considering the correlation between the ˛ and n coefficients and between soil water potential at 0.9 m depth ( a ) and 1.1 m depth ( b ) (Eq. (6).



(qN ) =

 +

∂f s ∂s

 +

∂f  ∂ a

2

2

+

a

 +

∂f r ∂r

 

∂f ∂f ˛n ∂˛ ∂n

+



∂f  ∂ b

2

 +

∂f ∂f  ∂ a∂ b

2



+

b

∂f ks ∂ks

2

∂f ˛ ∂˛

 +

  a

b

+

2

2

∂f n ∂n

∂f CN ∂CN

2

(6)

where ∂f/∂x is the partial derivative of each variable (x) in Eq. (5). The contribution of each variable to the propagated variance of qN was calculated by isolating each component of Eq. (6). 3. Results and discussion 3.1. Soil properties Average values of soil bulk density (BD) were 1.71, 1.72, 1.69, 1.65 and 1.57 g cm−3 at depths of 0.1, 0.3, 0.5, 0.7 and 0.9 m, respectively, and irrigation treatments had no effect on this variable. Up to a depth of 0.7 m, the BD was close, equal or even higher than the critical BD presented by Reichert et al. (2009), which is defined based on soil texture and considering the lowest limit of the water range, root development or crop yield. This high BD value is due to compaction derived from conventional soil management or due to traffic by loaded trucks in moist soil conditions during sugarcane harvest, as values below the critical BD were observed in an experiment carried out at the same area before the sugarcane planting (Gloaguen et al., 2010). The saturated hydraulic conductivity (ks ) and the parameters of the soil water retention curve ( r ,  s , ˛, n) were not affected by TSE irrigation (Table 1). When depths were considered to be subtreatments, the F test was not significant at 0.05 for irrigation or an interaction between irrigation and depth. One probable cause of this lack of effect is the high variability of soil hydraulic conductivity (the observed coefficient of variation varied from 5 to 160%). High variability in the physical properties of soil is frequently reported and is usually considered to be the major challenge in studies regarding nutrient leaching (Addiscott, 1996). At a depth of 0.3 m, the hydraulic conductivity and the value of the coefficient n were lower compared to other depths, which is a reflection of the higher bulk density. 3.2. Water drainage and nitrogen leaching At WI, water storage above field capacity (∼0.27 m3 m−3 ) occurred during only 18 days of the 691 total days evaluated (Fig. 1). However, at T100 and T150, the number of days that soil water

117

storage exceeded field capacity was 122 and 298, respectively, indicating higher potential of leaching in these treatments because the internal flux of the soil solution occurred when the water content of the soil reached or was close to the water content at field capacity at a depth of 1 m (Fig. 2). In the non-irrigated plots, the internal drainage occurred predominantly from February 2009 to middle April 2009 and from September 2009 to March 2010, only during periods of high rainfall. Even in irrigated treatments, the internal drainage occurred primarily during rainy periods; however, occurred during a more extended period compared with WI. The total amount of drained solution over the studied period was 320, 523 and 665 mm at WI, T100 and T150, respectively, indicating a higher potential for nutrient leaching in irrigated plots. Phosphate leaching was insignificant over the studied period, resulting in losses below 100 g ha−1 . This result was expected because P was added through the TSE irrigation at an amount that was relatively balanced with the plant uptake, and the amount of P applied trough mineral fertilization was only 12 kg ha−1 year−1 higher than the P plant uptake. Moreover, well-weathered soils with high aluminum and iron oxides contents have a high P adsorption capacity (Brennan et al., 1994; Fontes and Weed, 1996), which reduces the probability of its leaching. Nitrogen at 249 and 374 kg ha−1 were added through TSE irrigation at T100 and T150, respectively, mostly as ammonium (NH4 + ). However, 92.3% of the mineral N in the soil solution was found as nitrate (NO3 − ) (Table 2). This result is to due the fast nitrification of ammonium in well-aerated soils, which agrees with the results reported by Duan et al. (2010). Barton et al. (2005), studying N leaching through soil cores, found that 74% of the N in the leached solution was organic N. In contrast, we measured the total N in soil solution samples collected in December 2009 and found that mineral N represented 73% of the total. Thus, despite the significant percentage of organic N in the soil solution (∼27%), we prioritized mineral N analysis because the low volume of extracted solution in some samplings did not allow us to measure total N during the entire evaluated period. Thus, we will present only results referring to nitrate and ammonium leaching because the nitrite fraction was insignificant. Due to the low soil water content during the dry season, soil solution sampling, especially at WI plots, was possible only during rainy periods (Table 2), similar to the study of Ghiberto et al. (2009). However, this caused minimal bias in the N leaching assessment because during drought periods, the internal soil solution drainage was almost null (Fig. 2). The average ammonium concentration in the soil solution was 7.5%; however, the ammonium represented only 5% of the total leached N. This occurred because of the low volume of leaching during periods of higher concentration of ammonium in the soil solution (May/2009 and February–June/2010), as was also observed by Ghiberto et al. (2009, 2011). The N leached from WI amounted to 16.4 kg ha−1 after 691 days and a total of 320 mm of leached solution. This result revealed lower N leaching when compared with Ghiberto et al. (2009) after 205 mm of solution leached in a similar soil; however, Ghiberto et al. (2009) applied the entire recommended dose of N fertilizer (120 kg ha−1 ). Although they found that small amount of leached N came from the fertilizer, a higher abundance of N in the soil due to the fertilization may have indirectly enhanced N leaching. In other situations, Ghiberto et al. (2011) reported only 1.1 kg ha−1 of N leaching; however, these results were in conditions of very low drainage (91 mm). Nitrogen leaching was higher in the TSE-irrigated treatments compared to the control (WI), amounting to 39.5 and 88.8 kg ha−1 at T100 and T150, respectively (Table 3). The major concern with respect to N leaching is the increase in the N-nitrate concentration in groundwater, which can result in values above the

118

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

Table 1 Saturated hydraulic conductivity (ks ), saturated soil water content ( s ), residual soil water content ( r ), n and ˛ empiric soil water retention curve coefficients in the soil profile at plots with non-irrigated treatment (WI) or irrigated with treated sewage effluent to meet 100% (T100) and 150% (T150) of the crop water demand. Treatment

Depth, m 0.1

0.3

0.5

0.7

0.9

1.84 (0.26)a 1.21 (0.27) 0.90 (1.24)

0.23 (0.08) 0.41 (0.10) 0.52 (0.07)

1.14 (0.06) 5.89 (0.30) 1.44 (2.32)

2.24 (0.46) 1.73 (0.68) 1.71 (0.26)

3.42 (1.07) 3.96 (0.32) 3.32 (0.29)

0.41 (0.01) 0.41 (0.01) 0.38 (0.01)

0.43 (0.02) 0.43 (0.01) 0.41 (0.00)

0.45 (0.01) 0.45 (0.01) 0.44 (0.01)

0.49 (0.01) 0.47 (0.01) 0.44 (0.00)

0.50 (0.00) 0.52 (0.00) 0.46 (0.01)

0.12 (0.00) 0.11 (0.02) 0.12 (0.01)

0.14 (0.00) 0.14 (0.02) 0.13 (0.02)

0.14 (0.03) 0.14 (0.02) 0.14 (0.01)

0.13 (0.02) 0.13 (0.02) 0.13 (0.01)

0.13 (0.01) 0.12 (0.02) 0.12 (0.02)

1.63 (0.12) 1.68 (0.17) 1.66 (0.19)

1.50 (0.09) 1.54 (0.10) 1.64 (0.14)

1.61 (0.13) 1.61 (0.04) 1.79 (0.10)

1.76 (0.12) 1.72 (0.06) 1.79 (0.14)

1.77 (0.10) 1.79 (0.11) 1.84 (0.12)

4.69 (0.16) 4.10 (1.26) 2.41 (0.72)

4.93 (0.45) 4.92 (1.83) 2.30 (3.38)

3.01 (0.06) 3.60 (0.50) 2.13 (1.24)

2.71 (0.28) 3.03 (0.23) 2.18 (0.65)

3.16 (0.16) 3.17 (0.87) 2.42 (0.65)

0.23 0.22 0.25

0.27 0.26 0.28

0.29 0.28 0.29

0.29 0.27 0.28

0.27 0.27 0.27

0.17 0.15 0.18

0.21 0.20 0.20

0.21 0.20 0.19

0.19 0.19 0.19

0.18 0.17 0.17

−1

ks , cm h WI T100 T150  s , cm3 cm−3 WI T100 T150  r , cm3 cm−3 WI T100 T150 n WI T100 T150 ˛ WI T100 T150 (−10 kPa), cm3 cm−3 WI T100 T150 (−40 kPa), cm3 cm−3 WI T100 T150 a

Numbers represent the average of up to nine repetitions, and the respective standard deviations are in parenthesis.

limit of 11.3 mg L−1 recommended by the World Health Organization (WHO, 2011). In our study, the nitrate concentration in the soil solution frequently surpassed this limit at T100 and T150 (Table 2). However, the volume-balanced average N concentration in the leached solution (N leached (mg ha−1 )/volume of leached solution (L ha−1 )) over the entire evaluated period was 5.1, 7.3 and 13.4 mg L−1 at WI, T100 and T150, respectively, which was over the WHO limit only at T150. Thus, irrigation with 100% CWD did not threaten the groundwater quality with respect to N

concentration and the TSE irrigation beyond this limit must take into account the groundwater body dilution capacity. The balance between the N input, considering both irrigation and fertilization, and N removal by harvest was −25 kg ha−1 , +218 and +328 kg ha−1 over the entire evaluated period at WI, T100 and T150, respectively. The high positive balance found in the TSEirrigated plots represents an N leaching threat, as also reported by Leal et al. (2009). However, the unbalanced N addition does not seem to be the only reason for N leaching, as Barton et al. (2005)

Fig. 1. Water storage in the 1 m soil layer calculated with the van Genuchten equation and soil water potential tensiometer measurements during the period evaluated, at plots with non-irrigated treatment (WI), irrigated with treated sewage effluent to meet 100% (T100) and 150% (T150) of the crop water demand.

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

119

Fig. 2. Rainfall, volumetric soil water content (), volumetric water content standard deviation ( sd) and drained water during the period evaluated at a depth of 1 m at plots with non-irrigated treatment (WI) and irrigated with treated sewage effluent to meet 100% (T100) and 150% (T150) of the crop water demand.

reported an increase of 117 kg ha−1 of N leaching as result of TSE irrigation despite of a −180 kg ha−1 balance between N input and N uptake, which was most likely related with the increase of the rate of N mineralization due the higher soil moisture at irrigated plots. Buczko et al. (2010) also found that the N balance was a poor indicator of N losses during a single year and only slightly better considering longer term data. In their approach, N leaching was more related to the hydrology. However, Duan et al. (2010) verified that the cumulative application of nitrogen had good

correlation with N leaching. In our approach, we also observed good correlation between the N input and the N leached, finding that 53% of the N input was leached (Fig. 3a) (the two final months were excluded from this correlation to avoid bias due the high input of N and the absence of drainage in this period). However, we observed that the leaching occurred only in rainy periods; thus, a considerable improvement was achieved if the rainfall was considered in the model (Fig. 3b). Nevertheless, this model was fitted to a small data set for a short period of evaluation; a longer period of

120

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

Table 2 Nitrogen as nitrate (N-NO3 − ) and nitrogen as ammonium (N-NH4 + ) concentrations in the soil solution during the studied period at plots with non-irrigated treatment (WI) or irrigated with treated sewage effluent to meet 100% (T100) and 150% (T150) of the crop water demand. Period

August-08 September-08 October-08 November-08 December-08 January-09 February-09 March-09 April-09 May-09 June-09 July-09 August-09 September-09 October-09 November-09 December-09 January-10 February-10 March-10 April-10 May-10 June-10 Average a b

N-NO3 −

N-NH4 +

WI (mg L−1 )

T100 (mg L−1 )

T150 (mg L−1 )

WI (mg L−1 )

a

T100 (mg L−1 )

T150 (mg L−1 )

– – – 8.7 (0.6)b 7.8 (0.6) 6.3 (0.9) 4.3 (2.1) 4.8 (–) 7.7 (–) – – – – – 6.7 (5.4) 6.7 (3.7) 6.7 (3.5) 5.5 (2.3) 2.7 (1.1) – – – –

– – – 11.3 (–) 13.7 (–) 12.6 (7.3) 7.5 (3.4) 5.6 (2.4) 6.5 (3.3) 8.7 (–) 12.4 (–) 13.3 (–) 11.2 (–) 9.3 (–) 7.6 (2.6) 6.4 (1.9) 5.2 (2.0) 4.1 (1.4) 3.3 (0.8) 3.1 (0.5) 3.1 (0.1) 6.6 (–) 11.1 (–)

– 18.3 (4.1) 23.8 (1.5) 28.1 (1.0) 31.0 (2.7) 27.0 (3.2) 15.0 (3.9) 10.2 (4.2) 12.3 (5.0) 13.5 (5.5) 13.8 (6.9) 14.0 (7.4) 14.1 (4.4) 14.2 (1.7) 14.1 (1.7) 13.9 (4.7) 13.5 (7.5) 9.9 (4.9) 5.9 (1.7) 4.5 (–) 3.9 (–) 4.9 (–) 6.5 (–)

– – – 0.2 (0.0) 0.4 (0.1) 0.4 (0.1) 0.2 (0.0) 0.3 (–) 0.8 (–) – – – – – 0.5 (0.4) 0.6 (0.3) 0.6 (0.3) 0.6 (0.3) – – – – –

– – – 0.3 (–) 0.8 (–) 0.9 (–) 0.4 (0.1) 0.7 (0.3) 1.8 (1.0) 1.8 (–) 0.6 (–) 0.3 (–) 0.7 (–) 1.2 (–) 1.2 (0.9) 0.9 (0.5) 0.6 (0.3) 0.5 (0.2) 0.6 (0.2) 0.7 (0.2) 1.0 (–) 3.0 (–) 5.6 (–)

– 0.3 (0.2) 0.2 (0.1) 0.1 (0.1) 0.1 (0.1) 0.1 (0.1) 0.1 (0.0) 0.6 (0.7) 1.5 (2.0) 1.7 (2.0) 0.9 (0.9) 0.5 (0.5) 0.5 (0.3) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.6 (0.2) 0.9 (0.6) 1.3 (1.3) 1.4 (1.1) 1.3 (14.8)

6.2 (2.3)

8.1(2.3)

14.2 (4.0)

0.5 (0.2)

1.2 (0.4)

0.7 (1.2)

Absence of soil solution in the sampling period. Numbers represent the average of up to three repetitions, and the respective standard deviations are in parenthesis.

evaluation with a wider range of N input is recommended to improve the reliability of the N leaching predictions. Contrasting with the results observed by Duan et al. (2010), in our study, the N accumulated in the soil during the dry periods. This difference partly results due to the higher water storage capacity due the larger soil layer in our study (1 m vs. 0.3 m). One other difference

between the experiments was the irrigation management, as in our case the applied water was calculated to raise soil water content up to field capacity, whereas Duan et al. (2010) applied a surplus of effluent to force leaching. Another factor that contributed to N leaching is that the period of largest crop growth occurred between December and January,

Table 3 Nitrogen as nitrate (N-NO3 − ) and nitrogen as ammonium (N-NH4 + ) flux during the studied period at plots with non-irrigated treatment (WI) or irrigated with treated sewage effluent to meet 100% (T100) and 150% (T150) of the crop water demand. Positive numbers indicate gains and negative numbers losses of N at the depth of 1 m. Period

N-NO3 − WI (kg ha−1 )

August-08 September-08 October-08 November-08 December-08 January-09 February-09 March-09 April-09 May-09 June-09 July-09 August-09 September-09 October-09 November-09 December-09 January-10 February-10 March-10 April-10 May-10 June-10 Total a b

a

– – – −0.9 (0.9)b −0.2 (0.2) −2.2 (1.2) −2.9 (2.0) −1.1 (–) −0.5 (–) – – – – – −0.8 (0.3) −2.7 (1.1) −2.6 (1.8) −0.9 (0.4) −0.3 (0.1) – – – – −15.2 (8.0)

N-NH4 + T100 (kg ha−1 ) – – – −1.8 (–) 0.0 (–) −3.1 (3.4) −9.0 (2.9) −2.7 (0.8) −0.3 (0.2) 0.1 (–) 0.1 (–) 0.1 (–) −0.6 (–) −9.4 (–) −1.6 (0.4) −2.2 (0.8) −1.9 (0.7) −3.1 (0.8) −0.8 (0.3) 0.0 (0.1) 0.0 (0.0) 0.1 (–) 0.1 (–) −36.0 (10.6)

T150 (kg ha−1 )

WI (kg ha−1 )

T100 (kg ha−1 )

T150 (kg ha−1 )

– −0.7 (0.3) −1.3 (0.6) −6.1 (4.4) −2.3 (1.1) −3.8 (1.2) −15.9 (1.9) −9.6 (1.4) −1.9 (0.3) −0.4 (0.1) −0.4 (0.2) −0.2 (0.2) −1.1 (0.4) −13.9 (1.7) −2.6 (0.4) −7.5 (1.1) −8.2 (1.4) −5.4 (0.8) −3.5 (0.6) −0.4 (–) −0.2 (–) −0.2 (–) −0.1 (–)

– – – 0.0 (0.0) 0.0 (0.0) −0.1 (0.0) −0.1 (0.1) −0.1 (–) −0.1 (–) – – – – – −0.1 (0.0) −0.2 (0.1) −0.3 (0.1) −0.1 (0.0) −0.1 (0.0) – – – –

– – – −0.1 (–) 0.0 (–) −0.3 (0.1) −0.5 (0.1) −0.3 (0.1) −0.1 (0.1) 0.0 (–) 0.0 (–) 0.0 (–) −0.1 (–) −1.2 (–) −0.3 (0.1) −0.3 (0.1) −0.2 (0.1) −0.4 (0.1) −0.1 (0.1) 0.0 (0.0) 0.0 (0.0) 0.0 (–) 0.1 (–)

– 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) −0.1 (0.0) −0.5 (0.1) −0.2 (0.1) −0.1 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0) −0.5 (0.1) −0.1 (0.0) −0.3 (0.0) −0.3 (0.0) −0.3 (0.0) −0.3 (0.0) −0.1 (–) −0.1 (–) −0.1 (–) 0.0 (–)

−85.7 (18.1)

−1.2 (0.3)

−3.5 (0.9)

−3.1 (0.6)

Absence of soil solution in the sampling period. Numbers represent the average of up to three repetitions, and the respective standard deviations are in parenthesis.

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

121

Fig. 3. Correlation between total N leached and total N input (a) and between N leached and a function considering the total N input and rainfall (b). N: nitrogen input; acrfl: accumulated rainfall; rfl: monthly rainfall.

Table 4 Percentage of nitrogen leaching variance due to each component. Component of variance

WI (% of total variance)

T100 (% of total variance)

T150 (% of total variance)

Average (% of total variance)

Residual water content ( r ) Saturated water content ( s ) Alpha coefficient (˛) n coefficient (n) Alpha and n interaction Potential in the superior point (A) Potential in the inferior point (B) A and B interaction Saturated hydraulic conductivity (ks ) Nitrogen concentration

0 0 2 5 14 55 3 13 5 3

0 0 9 1 17 29 28 12 1 2

0 0 10 2 39 26 6 12 2 3

0 0 7 3 23 37 12 12 3 3

simultaneously with the rainy season, and the higher irrigation intensity occurred between March and August. Thus, the period of higher N uptake by plants coincided with the rainy season and not with nitrogen input through the TSE. 3.3. Composition of the variance Nitrogen leaching through the soil was calculated based on the N concentration in the soil solution, the soil physical properties, the soil water potential and the soil water potential gradient. Addiscott (1996) considered that the problems related to the quantification of N leaching were primarily due to the variability of the N concentration in the soil solution. However, despite the high coefficient of variation of the N concentration (Table 2), this variation caused only 2.6% of the variance of the N flux (Table 4). The parameters  s and  r did not affected N leaching, and the parameters n and ˛ played only a small role in the composition of N leaching variance; however, the interaction between n and ˛ lead to a significant increase in the N leaching variance. The variability of the water potential measurements was responsible for up to 70% of the N leaching variance. The importance of these measurements to the N flux variance is due to the soil water potential at different layers governs the hydraulic conductivity and the soil water potential gradient, and consequently, the flux of solution. Furthermore, due the logarithmic effect of the soil water potential on the soil water content, a small increase in the soil water potential results in a significant increase in the soil water content and consequently in the hydraulic conductivity. The soil water potential at a depth of 0.9 m ( a ) was more important to the variance composition than the soil water potential at a depth of 1.1 m ( b ). This higher contribution of a to the variance composition occurred because the internal drainage takes place at positive  ( a > b ), and usually, in this situation, the matrix potential is higher in the upper layer, resulting in a high water content and consequently in a higher flux variation.

4. Conclusions The leaching of P was as low as 100 g ha−1 throughout the experiment, and despite the fact that irrigation with TSE increased the N leaching, when applied in dosages to meet 100% of the CWD, it did not threaten groundwater quality. Thus, the use of TSE for irrigation to meet the CWD of sugarcane is not an environmental concern with respect to groundwater contamination with N and P. In addition, it prevents the direct disposal of these contaminants in surface water bodies. Nitrogen leaching can be estimated by considering the total N input and rainfall; however, long-term data are needed to improve the accuracy of these estimates. The variance propagation of the soil water potential measurements represented up to 70% of the nitrogen leaching variance. Acknowledgements The authors are grateful to professor Harold M. van Es from Cornell University for providing the idea regarding the isolation of the components of the variance. Our sincere gratitude is extended to FAPESP (Fundac¸ão de Amparo à Pesquisa do Estado de São Paulo), CAPES (Coordenac¸ão de Aperfeic¸oamento de Pessoal de Nível Superior), SABESP (Companhia de Saneamento Basico do Estado de São Paulo) and to the agroindustry EQUIPAV for financial and logistic support. References Addiscott, T.M., 1996. Measuring and modeling nitrogen leaching: parallel problems. Plant Soil 181, 1–6. Barton, L., Schipper, L.A., Barkle, G.F., McLeod, M., Speir, T.W., Taylor, M.D., McGill, A.C., van Schaik, A.P., Fitzgerald, N.B., Pandey, S.P., 2005. Land application of domestic effluent onto four soil types: plant uptake and nutrient leaching. Journal of Environment Quality 34, 635–643. Blum, J., Herpin, U., Melfi, A.J., Montes, C.R., 2012. Soil properties in a sugarcane plantation after the application of treated sewage effluent and

122

J. Blum et al. / Agricultural Water Management 117 (2013) 115–122

phosphogypsum in Brazil. Agricultural Water Management 115, 203–216, http://dx.doi.org/10.1016/j.agwat.2012.09.010. Bond, W.J., 1998. Effluent irrigation – an environmental challenge for soil science. Australian Journal of Soil Research 36, 543–555. Brennan, R.F., Bolland, M.D.A., Feffery, R.C., Allen, D.G., 1994. Phosphorus adsorption by a range of Western Australian soils related to soil properties. Communications in Soil Science and Plant Analysis 25, 2785–2795. Buczko, U., Kuchenbuch, R.O., Lennartz, B., 2010. Assessment of the predictive quality of simple indicator approaches for nitrate leaching from agricultural fields. Journal of Environment Management 91, 1305–1315. Duan, R., Fedler, C.B., Sheppard, C.D., 2010. Nitrogen leaching losses from a wastewater land application system. Water Environment Research 82, 227–235. Da Fonseca, A.F., Melfi, A.J., Monteiro, F.A., Montes, C.R., Almeida, V.V., Herpin, U., 2007. Treated sewage effluent as a source of water and nitrogen for Tifton 85 bermudagrass. Agricultural Water Management 87, 328–336. Fontes, M.P.F., Weed, S.B., 1996. Phosphate adsorption by clays from Brazilian Oxisols: relationships with specific surface area and mineralogy. Geoderma 72, 37–51. Gloaguen, T.V., Forti, M.C., Lucas, Y., Montes, C.R., Gonc¸alves, R.A.B., Herpin, U., Melfi, A.J., 2007. Soil solution chemistry of a Brazilian Oxisol irrigated with treated sewage effluent. Agricultural Water Management 88, 119–131. Gloaguen, T.V., Gonc¸alves, R.A.B., Forti, M.C., Lucas, Y., Montes, C.R., 2010. Irrigation with domestic wastewater: a multivariate analysis of main soil changes. Revista Brasileira de Ciência do Solo 34, 1427–1434. Ghiberto, P.J., Libardi, P.L., Brito, A.S., Trivelin, P.C.O., 2009. Leaching of nutrients from a sugarcane crop growing on an Ultisol in Brazil. Agricultural Water Management 96, 1443–1448. Ghiberto, P.J., Libardi, P.L., Brito, A.S., Trivelin, P.C.O., 2011. Nitrogen fertilizer leaching in an Oxisol cultivated with sugarcane. Scientia Agrícola 68, 86–93. Leal, R.M.P., Herpin, U., Fonseca, A.F., Firme, L.P., Montes, C.R., Melfi, A.J., 2009. Sodicity and salinity in a Brazilian Oxisol cultivated with sugarcane irrigated with wastewater. Agricultural Water Management 96, 309–316.

Leal, R.M.P., Firme, L.P., Herpin, U., Fonseca, A.F., Montes, C.R., Dias, C.T.S., Melfi, A.J., 2010. Carbon and nitrogen cycling in a tropical Brazilian soil cropped with sugarcane and irrigated with wastewater. Agricultural Water Management 97, 271–276. McLeod, M., Schipper, L.A., Taylor, M.D., 1998. Preferential flow in a well drained and poorly drained soil under different overhead irrigation regimes. Soil Use and Management 46, 96–100. Oliveira, M.W., Trivelin, P.C.O., Boaretto, A.E., Muraoka, T., Moratti, J., 2002. Leaching of nitrogen, potassium, calcium and magnesium in a sandy soil cultivated with sugarcane. PAB 37, 861–868. van Raij, B., Cantarella, H., 1996. Outras culturas industriais. In: van Raij, B., Cantarella, H., Quaggio, J.A., Furlani, A.M.C. (Eds.), Recomendac¸ões de adubac¸ão e calagem para o Estado de São Paulo. , 2 ed. Instituto Agronômico de Campinas, Campinas, pp. 233–243. R Development Core Team, 2008. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3900051-07-0 http://www.R-project.org Reichert, J.M., Suzuki, L.E.A.S., Reinert, D.J., Horn, R., Håkansson, I., 2009. Reference bulk density and critical degree-of-compactness for no-till crop production in subtropical highly weathered soils. Soil & Tillage Research 102, 222–254. Serhal, H., Bernard, D., Khattabi, J., Sabine, B.L., Shahrour, I., 2009. Impact of fertilizer application and urban wastes on the quality of groundwater in the Cambrai Chalk aquifer, Northern France. Environmental Geology 57, 1579–1592. Sharpley, A.N., Chapra, S.C., Wedepohl, R., Sims, J.T., Daniel, T.C., Reddy, K.R., 1994. Managing agricultural phosphorus for protection of surface waters: issues and options. Journal of Environment Quality 23, 437–451. Sieling, K., Kage, H., 2006. N balance as an indicator of N leaching in an oilseed rape – winter wheat – winter barley rotation. Agriculture, Ecosystems & Environment 115, 261–269. Van Genuchten, M.Th., 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society of America Journal 44, 892–898. WHO, 2011. Guidelines for Drinking-Water Quality, fourth edition. World Health Organization, p. 541.