Effect of macropore flow on the transport of surface-applied cow urine through a soil profile.
The movement of water through macropores in soil is not a new concept, and was first documented in the 18th century (Beven and Germann 1982). However, the importance of macropores in contributing to water and solute transport has only been studied in detail since the late 1970s (eg. Ball et al. 1979; Smettem and Collis-George 1985; White 1985; Watson and Luxmoore 1986; Timlin et al. 1994). Field-scale data for solute transmission through macropores is still limited. Whilst macropores may constitute only a small fraction of total porosity of a soil (eg. 0.1 to 5.0%) they can greatly influence the transport of water, solutes, and pollutants through the soil (Bouma and Dekker 1978; White 1985).
Luxmoore (1981) classified soil pores into 3 size classes and considered the voids with a pore diameter of [is greater than] 1000 gm as `macropores'. However, the definition of macropores is arbitrary, and the minimum equivalent diameter for macropores reported in the literature ranges from 30 to 3000 gm (Beven and Germann 1982).
Macropores are formed in several ways, such as chemical weathering, shrinking, frost action, plant root channelling, and earthworm and other microfauna burrowing (Beven and Germann 1982). Their size varies depending on their origins. Additionally, they may be confined to the topsoil or continue to several metres down the soil profile (Williams and Allman 1969; Omoti and Wild 1979). Therefore, most soils contain some macropores, the nature and volume of which depend upon a number of soil characteristics and environmental conditions.
One of the most widely used methods to quantify water transmission through different-sized soil pores involves the use of a tension infiltrometer (Clothier and White 1981; Ankeny et al. 1988; Timlin et al. 1994). This method allows the exclusion of macropores of selected sizes from transmitting water or solute. Tension infiltrometers have been used to maintain both positive and negative pressures to categorise macropores (Watson and Luxmoore 1986; Wilson and Luxmoore 1988; Reynolds 1993). These studies have shown that a significant proportion of water (up to 85%) can be transmitted through macropores.
Nitrate leaching and contamination of groundwater from cow urine patches in grazed dairy systems has been reported by several researchers (e.g. Ball et al. 1979; Field et al. 1985; Silva et al. 1999). Macropore flow was thought to be an important mechanism by which urinary N was transported (Ball et al. 1979; Monaghan et al. 1989; Williams et al. 1990). However, these studies were confined to the top 150-350 mm soil layer. Under these experimental conditions, surface macropores could have been continuous through the soil column, whilst in reality they would normally terminate deeper in the soil profile. There have been no studies that quantify the contribution of macropore flow to the leaching of urine N through deep soil profiles under realistic field dairy pasture conditions.
The main objective of this research was to quantify the contribution of macropore ([is greater than] 600 [micro]m) flow to the transport of surface-applied cow urine N through a large intact soil lysimeter (700 mm deep).
Materials and methods
The experiment was carried out using a lysimeter (500 mm diameter by 700 mm depth), installed in a field lysimeter facility at Lincoln University, Canterbury, New Zealand. The soil in the lysimeter was a free-draining Templeton fine sandy loam (Udic Ustochrept, coarse loamy, mixed, mesic), and had been under ryegrass (Lolium perenne L)-white clover (Trifolium repens L) pasture for [is greater than] 9 years. Figure 1 shows the soil total porosity and volumetric water content at 0.5 kPa suction. The C:N ratio varied, ranging from 12: 1 in the upper layer to 7: 1 in subsoils (Silva et al. 1999).
[Figure 1 ILLUSTRATION OMITTED]
A detailed description of the procedure for collecting lysimeters has been given elsewhere (Cameron et al. 1992). In brief, the lysimeter consisted of a steel cylindrical casing which was pushed into the soil to collect an undisturbed soil monolith. A cutting ring at the base of the cylinder created a 5-mm annular gap between the soil monolith and the casing, which was filled with liquefied petroleum jelly. Once the jelly solidified, it formed an effective seal to prevent edge flow (Cameron et al. 1992). The bottom 40 mm of soil was removed from the cylinder and replaced with a washed gravel and sand mixture to create a free-draining situation similar to that of the field soil.
Leaching with tension infiltrometer
During late summer-autumn (February-May) 1998, leaching of surface-applied cow urine N through the lysimeter was measured at 2 selected suctions (0.5 and 0 kPa) using a slightly modified Reynold's (1993) tension infiltrometer. Our tension infiltrometer consisted of 3 main components, namely an infiltrometer disc, bubble tower, and water reservoir (Fig. 2). The diameter of the infiltrometer disc was 495 mm with an effective diameter of 465 mm. The base of the infiltrometer disc was made of stainless steel mesh with 3-mm-diameter holes at 6-mm centers. This base was covered with 3 layers of 20-gm mesh nylon cloth.
[Figure 2 ILLUSTRATION OMITTED]
The infiltrometer disc was first filled with water to eliminate all air from the disc. The water reservoir (200-L barrel) was filled with water, leaving a small air space volume on the top. The bubble tower was calibrated to produce the correct suction in relation to the relative positions of the water-supply tube of the water reservoir and the soil surface of the lysimeter. The water-supply tube was used as a water manometer for this calibration. All air was removed from the water-supply tube while the infiltrometer was immersed in a water bath. The reservoir was closed air tight and water suction was maintained at 0.5 kPa using a vacuum pump attached to the top of the water reservoir (Fig. 2).
The effective water flow under 0.5 kPa suction was due to drainage through pores [is less than or equal to] 600 [micro]m diameter. This diameter was assumed to be the lower limit of macropores in our study. The equivalent pore radius (cm) corresponding to the applied negative pressure h (cm [H.sub.2]O) was calculated from the capillary rise equation:
r = 2[Gamma] cos[Alpha] / [Rho]gh
where [Gamma] is the surface tension of water (N/m), [Alpha] is the wetting angle of the water and the pore wall (assumed to be nearly zero, therefore cos [Alpha] [congruent] 1), p is the density of water (kg/[m.sup.3]), and g is acceleration due to gravity (m/[s.sup.2]). After the completion of leaching under 0.5 kPa suction, water flow and solute transport were also determined under 0 kPa suction on the same lysimeter to compare the effect of different suctions on solute transport. It was considered essential to conduct the two leaching events on the same soil lysimeter to keep the soil pore systems constant.
Prior to each leaching event grass in the lysimeter was trimmed to ground level and all debris was removed without disturbing the soil. A nylon cloth (20-[micro]m mesh) was laid on the surface of the lysimeter. A sand contact material (particle size 75-250 [micro]m) was poured on top of the nylon cloth to an average depth of 10-15 mm. The contact material was levelled and smoothed to ensure good contact with the infiltrometer. The infiltrometer disc was placed on top of the contact material and allowed to equilibrate under 0.5 kPa suction for about one week.
Following the equilibration period the tension infiltrometer, contact material, and cloth were removed from the lysimeter. Labelled urine was then applied to the lysimeter at 1000 kg N/ha by pouring it onto the soil surface. The urine had been collected from milking cows and analysed fresh for total N, and the rate of application (1000 kg N/ha) was similar to that under a typical cow urine patch. The urine was labelled with [sup.15]N urea and [sup.15]N glycine at a 9:1 ratio and with bromide at a rate of 60 kg Br/ha. Immediately after the urine was applied, the nylon cloth and contact sand material were replaced, and the infiltrometer disc was installed on the lysimeter with a 0.5 kPa suction.
Leachate samples were collected once the leachate volume reached 2-3 L or at 24-h intervals. Samples were then treated with 0.6 mL of 0.5 mg/L of phenylmercuric acetate to inhibit microbial activities and were stored in a freezer for subsequent analysis. Samples were then filtered through Whatman No. 42 filter paper and analysed for nitrate-N, nitrite-N, and bromide using ion exchange chromatography (Waters, USA), and ammonium-N and urea-N using flow injection analysis (Tecator, Sweden). Once the concentration of bromide in the leachate reached negligible levels (e.g. after 10 weeks), the infiltrometer, contact material, and cloth were removed and urine was again applied to the lysimeter at the same rate of 1000 kg N/ha labelled with 15N and bromide (120 kg Br/ha). The infiltrometer was then replaced and set to 0 kPa water suction. Leachate samples were collected and analysed as above.
Monitoring environmental conditions and flow rates
The lysimeter had one temperature probe installed at 300 mm depth and 3 tensiometers at 150, 300, and 600 mm depths. The temperature probe (LM 35 CZ integrated circuit sensor) was inserted horizontally and tensiometers were inserted at 20 [degrees] downward angle through the lysimeter wall. Matric potential was recorded by a pressure transducer attached to each tensiometer and connected to a Campbell Scientific CR10 datalogger. The infiltration rate was also recorded via volume infiltrated, by using a pressure transducer (Sensym 0-1 PSI, SCX C) installed on the top of the water reservoir and connected to the same datalogger.
Results and discussion
Infiltration rate and drainage volume
The infiltration rate at 0 kPa was both higher (by about an order of magnitude) and more variable than at 0.5 kPa suction (Fig. 3). This result agrees with the findings of Clothier and White (1981) that an increase in suction, and therefore the removal of the contribution of macropores, decreases the rate of infiltration. The higher infiltration rate at 0 kPa is thus due to the higher proportion of functioning pores and/or increased pore continuity compared with 0.5 kPa suction.
[Figure 3 ILLUSTRATION OMITTED]
The time period to drain about 2 effective pore volumes was 10 weeks at 0.5 kPa suction whilst at 0 kPa it was only 2 weeks (Table 1). Thus, the drainage rate was about 5-fold slower at 0.5 kPa suction than at 0 kPa suction. These results demonstrate the significant contribution of macropores to water movement.
Table 1. Total amount of solute losses and percentage recoveries in the leachate at 0.5 and 0 kPa suctions
Suction 0.5 kPa 0 kPa Time taken for 2 effective pore 10 2 volumes to emerge (weeks) [NO.sub.3]-N losses (g/[m.sup.2]) 0.08 0.04 [NO.sub.3]-N losses (% of total N applied) 0.08 0.04 [NH.sub.4]-N losses (g/[m.sup.2]) 0.17 10.54 [NH.sub.4]-N losses (% of total N applied) 0.17 10.54 Urea-N losses (g/[m.sup.2]) nil 1.59 Urea-N losses (% of total N applied) nil 1.59 Total N losses (g/[m.sup.2]) 0.25 12.17 Total N losses (% of total N applied) 0.25 12.17 [sup.1]5N losses (% of total [sup.15]N applied) 0.07 2.99 Bromide losses (g/[m.sup.2]) 5.93 10.82 Bromide losses (% of total Br applied) 98.86 90.27 Average daily temperature ([degrees] C) 17 8
Soil matric potentials and temperature
The plots of matric potential ([[Psi].sub.m]) measured at 3 depths under both suctions (Fig. 4) indicate that these [[Psi].sub.m] values were more negative than the [[Psi].sub.m] imposed at the surface. However, once an equilibrium flow has been established through the lysimeter, then the horizontally averaged Darcy velocity v (z) must be constant with depth z.
[Figure 4 ILLUSTRATION OMITTED]
Here v = -K([[Psi].sub.m]) ([Delta]H/[Delta]z) is the product of the local hydraulic conductivity K([[Psi].sub.m]) at the local matric potential ([[Psi].sub.m]), and the local gradient [Delta]H/[Delta]z of hydraulic potential H. In practice, the profile flow regime adjusts itself so that the average of this product is constant with depth. Matric potential values more negative than the surface-imposed values (Fig. 4) simply reflect this self-adjustment in a vertically layered and heterogeneous soil profile. Nevertheless the soil matric potentials were about 1-4 kPa more negative at 0.5 kPa suction than at 0 kPa suction (Fig. 4). This indicates that the soil moisture contents at 0 kPa suction were higher and thus more macropores were conducting water.
The daily average soil temperature at 300 mm depth decreased gradually over the experimental period from 17 [degrees] C to 8 [degrees] C (Table 1). The daily average temperature during the first leaching event did not drop below 10 [degrees] C. However, the daily average temperature was [is less than] 10 [degrees] C during the second leaching event. According to Poiseuille's law the water flux in a tube is inversely proportional to viscosity, and viscosity is inversely related to temperature. Hence, flow rate should increase with temperature. However, the infiltration data in this study showed relatively little effect of temperature (Fig. 3), indicating the predominant importance of pore size in affecting water movement. However, temperature can have a considerable impact on the forms and the amounts of N leached from the urine applied, as discussed below.
Different forms of solutes leached
Leaching data are presented in the form of breakthrough curves (BTCs) depicted as solute concentration v. cumulative drainage (Fig. 5). Only two forms of N ([NO.sub.3]-N and [NH.sub.4]-N) and bromide were detected at 0.5 kPa suction (Fig. 5a). However, under 0 kPa suction, urea-N was also detected in addition to these two N forms and bromide (Fig. 5c). The absence of urea-N at 0.5 kPa suction was probably due to the slower and delayed leaching allowing time for urea hydrolysis to occur in the soil.
[Figure 5 ILLUSTRATION OMITTED]
Bromide leaching losses
The greater volume of cumulative drainage under 0 kPa suction was because of the extended leaching period required to produce a complete bromide BTC. The peak concentration of bromide under 0.5 kPa suction emerged around 140 mm of cumulative drainage (Fig. 5b). Under 0 kPa, the peak concentration of bromide appeared around 200 mm of cumulative drainage (Fig. 5d). Water flow through soil macropores may have two distinct effects on solute movement: (i) preferential flow where surface-applied solutes are transported rapidly through the macropore system; and (ii) `by-pass flow' where solutes that are resident in the soil aggregate pores (or other micropores) are bypassed by water flowing through the macropore system. Where a solute (e.g. Br) has diffused into the soil micropore system it can take longer to be removed under bypass flow conditions (e.g. 0 kPa suction) compared with more uniform flow conditions (eg. 0.5 kPa suction). The comparatively greater drainage volume required for the complete bromide peak to emerge at 0 kPa suction was probably due in part to greater bypass flow, and in part to the greater pore volume than at 0.5 kPa suction. Under 0.5 kPa suction, soil pores [is greater than] 600 [micro]m were excluded from transmitting water and solutes. Therefore, there was a smaller effective pore volume as well as less bypass flow.
Nitrate leaching losses
The concentrations of [NO.sub.3]-N at both suctions were very low (Fig. 5a). The low [NO.sub.3]-N concentrations were probably caused by different mechanisms for the 0.5 kPa and 0 kPa suctions. For the 0.5 kPa suction, [NO.sub.3]-N was probably denitrified or immobilised because of the long residence time of N in the soil under reasonably high temperatures (Table 1). For the 0 kPa suction, there was probably insufficient time for all of the N to be transformed from urea-N to [NO.sub.3]-N, thus allowing N to be leached as urea-N or [NH.sub.4]-N (Fig. 5c). Any [NO.sub.3]-N produced may have been denitrified under the saturated conditions. It must also be noted, however, that the cool temperatures during the leaching period under 0 kPa suction may also have slowed down the N transformations into [NO.sub.3]-N.
As a consequence of the low [NO.sub.3]-N concentration, the total leaching losses of nitrate were also low (0.08 and 0.04 g/[m.sup.2], equivalent to 0.08 and 0.04% of the total N applied) at 0.5 kPa and 0 kPa suctions, respectively (Table 1).
Ammonium leaching losses
As with [NO.sub.3]-N, the concentration of [NH.sub.4]-N was very low at 0.5 kPa suction (Fig. 5a). However, at 0 kPa suction the breakthrough concentration of [NH.sub.4]-N remained high even after 3.5--4 pore volumes of leachate (Fig. 5c). The lower concentration of [NH.sub.4]-N at 0.5 kPa suction was again due to the longer residence time of the applied N in the soil and the warmer temperatures (17 [degrees] C), which stimulated N transformations and denitrification. Under 0 kPa suction the lower temperature and short residence time of N in the soil meant that the [NH.sub.4]-N produced from urea hydrolysis was not nitrified.
Despite cation exchange reactions which would have retained some [NH.sub.4]-N in the soil profile, the main form of N lost in this experiment was [NH.sub.4]-N. Table 1 shows that the loss of [NH.sub.4]-N at 0.5 kPa suction was 0.2 g/[m.sup.2], equivalent to 0.2% of the total N applied. At 0 kPa suction this increased to 10.5 g/[m.sup.2], equivalent to 10.5% of the total N applied. The high flow rate of water and solute under 0 kPa may have resulted in a lowered efficiency of cation exchange to retain [NH.sub.4]-N. However, significant concentrations of [NH.sub.4]-N remained in the soil profile at the end of the experiment (Fig. 6).
[Figure 6 ILLUSTRATION OMITTED]
Urea leaching losses
Figure 5a shows that no urea-N was detected at 0.5 kPa suction. However at 0 kPa suction the concentration of urea-N reached a maximum of 16 mg/L (Fig. 5c). The higher leaching rate at 0 kPa provided less time for urea-N to be hydrolysed to [NH.sub.4]-N and hence higher concentration of urea-N leached at 0 kPa suction. The location of peak concentration of urea-N was similar to that of the bromide peaks (Fig. 5). Since urea is a highly soluble neutral solute, it would have behaved as a non-reactive solute and quickly leached through the soil. During macropore flow, urine N may be leached through the soil in the form of unhydrolysed urea.
Total N leaching losses
The total N leaching losses show that 12.2 g/[m.sup.2], equivalent to 12.2% of the total N applied, was leached at 0 kPa suction, whilst only 0.25% of the N applied was leached at 0.5 kPa suction (Table 1). This indicates that 98% (as a proportion of total N recovered) of the total N leached was via macropores [is greater than] 600 [micro]m. This was confirmed by the [sup.15]N results which showed that the proportion of [sup.15]N lost through soil pores [is greater than] 600 [micro]m was 98%. It must be noted that as the two leaching events, first at 0.5 kPa and then at 0 kPa suction, were conducted on the same lysimeter (which was necessary for comparison purposes), some of the N leached under 0 kPa suction may have come from the urine applied in the first leaching event. This, however, does not invalidate the contrasting leaching patterns under the two suctions.
Transport through soil macropores [is greater than]600 [micro]m was found to be responsible for 98% of the N leaching losses. The forms and amounts of N leached were affected by the interactions of macropore flow and N transformation processes. The rapid transport of water and solute through macropores decreased the residence time of N in the soil. The N from applied urine was leached as urea-N or [NH.sub.4]-N before it was converted to [NO.sub.3]-N. When macropores were prevented from conducting (under 0.5 kPa suction), the urine N remained in the soil profile for an extended period allowing the N to be transformed from urea-N to [NO.sub.3]-N, which was then denitrified or immobilised. Interpretation of these results has implications for dairy farm management. It is recommended that stock should be removed 1-2 days before irrigation water is applied as this will allow animal urine to diffuse into soil micropores and thus decrease N leaching by macropore flow.
We thank the New Zealand Foundation for Research, Science and Technology and the New Zealand Dairy Board for funding this research. The authors are grateful to John Milne of Lincoln Ventures Ltd for building the infiltrometer. We also thank Trevor Hendry, Stephen Moore, Angela Reid, Roger Cresswell, Gaye Bruce, and Maureen McCloy of Soil, Plant and Ecological Sciences Division, for excellent technical support. Thanks also to Lynne Clucas for proof reading the manuscript.
Ankeny MD, Kasper TC, Horton R (1988) Design for an automated tension infiltrometer. Soil Science Society of America Journal 52, 893-896.
Ball PR, Keeney DR, Theobald PW, Nes P (1979) Nitrogen balance in urine-affected areas of a New Zealand pastures. Agronomy Journal 71,309-314.
Beven K, Germann P (1982) Macropores and water flow in soils. Water Resources Research 18, 1311-1325.
Bouma J, Dekker LW (1978) A case study on infiltration into dry clay soil. 1. Morphological observations. Geoderma 20, 27-40.
Cameron KC, Smith NP, Mclay CDA, Fraser PM, McPherson RJ, Harrison DF, Harbottle P (1992) Lysimeter without edge flow: an improved design and sampling procedure. Soil Science Society of America Journal 56, 1625-1628.
Clothier BE, White I (1981) Measurement of sorptivity and soil water diffusivity in the field. Soil Science Society of America Journal 45, 241-245.
Field TRO, Theobald PW, Ball PR, Clothier BE (1985) Leaching losses of nitrate from cattle urine applied to a lysimeter. Agronomy Society of New Zealand Proceedings 15, 137-141.
Luxmoore RE (1981) Micro-, meso and macroporosity of soil. Soil Science Society of America Journal 45, 671.
Monaghan RM, Cameron KC, McLay CDA (1989) Leaching losses of nitrogen from sheep urine patches. New Zealand Journal of Agricultural Research 32, 237-244.
Omoti U, Wild A (1979) Use of fluorescent dyes to make the pathways of solute movement through soils under leaching conditions. 2. Field experiments. Soil Science 128, 98-104.
Reynolds WD (1993) Unsaturated hydraulic conductivity: field measurement. In 'Soil sampling and methods of analysis'. (Ed. M. R. Carter) pp. 633-644. (Canadian Society of Soil Science: Ottawa)
Silva RG, Cameron KC, Di H J, Hendry T (1999) A lysimeter study of the impact of cow urine, dairy shed effluent and nitrogen fertiliser on nitrate leaching. Australian Journal of Soil Research 37, 357-369.
Smettem KRJ, Collis-George N (1985) Statistical characterisation of soil biopores using a soil peel method. Geoderma 36, 27-36.
Timlin DJ, Ahuja LR, Ankeny MD (1994) Comparison of three field methods to characterise apparent macropore conductivity. Soil Science Society of America Journal 58, 278-284.
Watson KW, Luxmoore RJ (1986) Estimating macroporosity in a forest watershed by use of a tension infiltrometer. Soil Science Society of America Journal 50, 578-582.
White RE (1985) The influence of macropores on the transport of dissolved and suspended matter through soil. Advances in Soil Science 3, 95-120.
Williams PH, Gregg PEH, Hedley MJ (1990) The effect of preferential flow on dairy cow urine and simulated rainfall on movement of potassium through undisturbed topsoil cores. Australian Journal of Soil Research 28, 857-868.
Williams RE, Allman DW (1969) Factors affecting infiltration and recharge in a loess-covered basin. Journal of Hydrology 8, 265-281.
Wilson GV, Luxmoore RJ (1988) Infiltration, macroporosity, and mesoporosity distribution on two forested watersheds. Soil Science Society of America Journal 52, 329-335.
Manuscript received 17 March 1999, accepted 23 September 1999
R. G. Silva, K. C. Cameron, H. J. Di, N. P. Smith, and G. D. Buchan Centre for Soil and Environmental Quality, PO Box 84, Lincoln University, Canterbury, New Zealand.
|Printer friendly Cite/link Email Feedback|
|Author:||Silva, R. G.; Cameron, K. C.; Di, H. J.; Smith, N. P.; Buchan, G. D.|
|Publication:||Australian Journal of Soil Research|
|Date:||Jan 1, 2000|
|Previous Article:||Effect of antecedent soil water regime and rice straw application time on [CH.sub.4] emission from rice cultivation.|
|Next Article:||Irrigation of an allophanic soil with dairy factory effluent for 22 years: responses of nutrient storage and soil biota.|