Published online 23 August 2007
Published in Vadose Zone J 6:548-553 (2007)
DOI: 10.2136/vzj2006.0095
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SPECIAL SECTION: GROUNDWATER RESOURCES ASSESSMENT UNDER THE PRESSURES OF HUMANITY AND CLIMATE CHANGE
Combined Effects of Rainfall Patterns and Porous Media Properties on Nitrate Leaching
Fumi Sugitaa,* and
Kazuro Nakaneb
a Chiba Univ. of Commerce, Konodai 1-3-1, Ichikawa, Chiba 272-8512 Japan
b National Institute for Earth Science and Disaster Prevention, Tennodai3-1. Tsukuba, Ibaraki 305-0006 Japan
* Corresponding author (fsugita{at}cuc.ac.jp).
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Received 5 July 2006.
 |
ABSTRACT
|
|---|
Laboratory experiments were conducted to investigate the effects of water input properties on the movement of nitrate in three different porous media. A series of artificial rainfall amounts from 4.0 to 26 mm were applied every 24 h in 10 min durations over a weighing lysimeter packed with either fine sand, field soil over fine sand, or fine sand with artificial macropores. Movements of nitrate, chloride, and water in the lysimeter were monitored. Significant nitrate leaching occurred with pistonlike flow caused by the highest applications in all porous media. The nitrate movement coincided with that of chloride in a homogeneous sand medium. In a layered medium with soil cover over sand, nitrate/chloride ratios in the field soil decreased with time, suggesting the presence of degradation processes. Some parts of the soil may be serving as spots for denitrification. High nitrate concentration remained in the topsoil layer under light to intermediate rainfall applications in the two-layered medium, causing large vertical spreading of nitrate under intermediate rains by which infiltration occurs. Double peaks in the concentration profiles were observed in the macroporous medium under 26-mm applications, indicating that part of the nitrate was transported by preferential flow while the rest was transported by matrix flow. We estimated that heterogeneity of the porous medium resulted in large dispersion under intermediate to heavy rains and denitrification under light rains. Heterogeneity also affected pathways of leaching under heavy rainfalls. We estimated that climate change could significantly increase the chance of nitrate leaching by at least 25% in the Pacific East area of Japan through increases of heavy rain frequency and precipitation intensity and decreased time for denitrification in the surface soil.
|
INTRODUCTION
|
|---|
Climate change likely to occur in the near future would result in changes in precipitation patterns, as well as the onset of snowmelt and runoff. These changes in the timing of available water may in turn affect the volume and frequency of water input over farmland. Groundwater contamination by nitrate from agricultural lands is a serious environmental issue in many countries, including Japan, and the pathways and the rates of nitrate leaching from agricultural lands are controlled by many factors, including precipitation, soil properties, and farming practices such as irrigation and fertilizer application.
Laboratory and small-scale field experiments have provided insights into micro- to small-scale mechanisms of nitrate leaching. Blevins et al. (1996) identified the fate of a single application of N fertilizer by means of small-scale field experiments. Mantovi et al. (2006) demonstrated that nitrate from slurry accumulated in the surface layer in warm periods and leached quickly after rain due to macroporosity. Many researchers have reported fast infiltration of nitrate from the ground surface to depth via macropores generated by earthworms and plant roots. Macropores are relatively large continuous openings in porous media that are distributed ubiquitously over the field and densely near the surface (Munyankusi et al., 1994). Omoti and Wild (1979) reported that earthworm channels ranged from 2 to 10 mm in diameter and had an average density of 100 pores m–2. They also reported that macropores played an important role in transporting nitrate vertically in the field. Sugita et al. (2005) found that the onset of macropore flow depends on rainfall intensity, while the rate of macropore flow depends on rainfall intensity and duration. Catchment studies have also pointed out the importance of preferential flow and subsurface storm flow, which lead to larger-scale subsurface heterogeneity in nitrate discharge into surface water (Soulsby et al., 2003; Schilling and Zhang, 2004).
Large lysimeter experiments allow us to assess the effects of subsurface heterogeneity on nitrate transport without sacrificing accuracy in quantification of water and solute fluxes. Owens et al. (1995) obtained an accurate nitrate mass balance using a large undisturbed lysimeter and proposed a strategy for reducing nitrate leaching from agricultural land to groundwater. Schoen et al. (1999) performed tracer experiments using a large lysimeter packed with undisturbed soil and found an occurrence of preferential flow during nitrate leaching.
Although laboratory studies have been conducted under constant water flux to investigate roles of macropore flow on solute transport in the vadose zone (Buttle and Leigh, 1997), actual rainfalls are not constant, which could result in even faster leaching. Our study investigated the combined effects of rainfall input patterns and subsurface heterogeneity on nitrate leaching in the vadose zone by means of laboratory tracer experiments using a weighing lysimeter equipped with an artificial rainfall generator. Following a nitrate tracer application to the surface, rainfalls of constant amount and constant intervals were repetitively applied over the weighing lysimeter packed with either homogeneous fine sand, soil over fine sand, or a macroporous medium, and the movements of water and solutes were monitored. The experiments were repeated at different rainfall application rates.
|
Materials and Methods
|
|---|
The weighing lysimeter was 3.0 m long, 1.0 m wide, and 0.6 m deep (Fig. 1
). The simulated groundwater table was maintained at 0.55 m below the surface by a Marriott tank connected to the lysimeter. Air flow over the lysimeter was controlled to have a constant wind speed of 1.5 m s–1, temperature of 20°C, and a relative humidity of 50%. The soil water was allowed to evaporate into the air during all experiments, and the evaporation rate was monitored by weighing the lysimeter.
The first set of experiments was conducted with the lysimeter packed with homogeneous fine sand (average diam. = 0.14 mm). Tracer solution containing 1000 mg L–1 of chloride and 1000 mg L–1 of nitrate was applied as 2 mm of rain before every set of experiments. Precipitation properties during 2001 to 2005 at Tsukuba, where our laboratory is located, are summarized in Table 1. On a daily basis, 38.5% of rainfall events recorded was less than 4 mm. Artificial rainfall of 4 mm per application was applied once every 24 h over the lysimeter for 16 applications. The set of experiments was repeated by changing the rainfall amount per event to 6.8, 13, and 26 mm. It took approximately 10 min to apply a single rainfall event over the lysimeter for all application rates. The experimental conditions and the number of rainfall applied in each case are summarized in Table 2. Each rainfall was applied every 24 h until total rainfall exceeded at least 52 mm.
View this table:
[in this window]
[in a new window]
|
TABLE 2. Amounts and duration of simulated rainfall events repeated in each experimental condition. The rainfall was applied every 24 h, and water samples were taken before the rainfall application and 24 h after the last rainfall application.
|
|
Water movement in the lysimeter was monitored every minute with pressure transducers (Daiki Rika Kogyo Ltd., Saitama, Japan) connected to tensiometers inserted at 0.05, 0.1, 0.15, 0.2, 0.3, 0.45, and 0.5 m below the surface in the lysimeter. The lysimeter and Marriott tank were weighed every minute to estimate the water input and loss from the system. The drainage from the lysimeter was collected in a plastic bottle and was measured manually once a day when drainage was observed. Tracer movement was monitored by taking 2- to 10-mL samples of soil water using suction samplers (Sankei Rika Ltd., Tokyo) placed at 0.05, 0.1, 0.15, 0.2, 0.3, 0.4, 0.5 and 0.6 m below the surface in the lysimeter. The water samples were analyzed for chloride and nitrate using ion chromatography (PIA-1000, Shimazu, Kyoto). The tracer was prepared each time for a tracer application, which resulted in a slightly different tracer concentration each time. To compensate for the tracer concentration variation effect, the observed concentrations were normalized by the concentration ratio of control tracer to tracer applied (normalized concentration = observed concentration x concentration of control tracer/concentration of tracer applied).
The source of nitrate in natural systems is most often located on the ground surface. Thus, it is likely for nitrogen to come in contact with an organic soil layer before infiltrating to the groundwater. Field soil was taken from a woodlot where vegetables used to be grown near the laboratory, and the top 0.2 m of sand layer in the lysimeter was replaced by the soil. The soil was a dark Andosol with a high porosity (66%) and organic content (on average, 3%), and a well-developed double porosity structure. The two-layered system represents actual field conditions better than the homogeneous medium, although it still is a highly idealized and artificially packed porous medium. The series of experiments were repeated over the two-layered medium.
Three hundred nonconnected vertical macropores were artificially generated over the lysimeter (100 pores m–2) by inserting and removing a stainless steel rod (2 mm diam.). Half of the macropores (150 pores) extended to 0.2 m from the surface, while 75 pores extended to 0.3 m, and the remaining 75 pores extended to 0.6 m from the surface, since macropores were more densely distributed near the ground surface in actual fields (Munyankusi et al., 1994). Although many inlets of the macropores were collapsed during rainfall applications, they remained as they were until the end of the set of the experiments.
|
Results and Discussion
|
|---|
Water and Nitrate Movement in Response to Rainfall Applications
Homogeneous Medium
Concentration profiles of chloride and nitrate after 52 mm of total rain at different rainfall amounts per event in the homogeneous medium are shown in Fig. 2
. It shows the profiles observed 24 h after 13 daily events of 4-mm rain, after 8 events of 6.8-mm rain, after 4 events of 13-mm rain and after 2 events of 26-mm rain. Evaporation from the lysimeter was almost comparable to or exceeded rainfall during the 4-mm applications. Thus, no significant downward solute movement was observed under this application rate. Approximately 80% of the rain was lost to evaporation during the 6.8-mm applications, while 47 and 23% of the rain was lost to evaporation during the 13- and 26-mm applications, respectively.
View larger version (28K):
[in this window]
[in a new window]
|
FIG. 2. Chloride and nitrate profiles after 52 mm of total rain by repeating rainfall of different amounts per event in the homogeneous medium. The concentration was normalized by the concentration ratio of control tracer to the tracer applied.
|
|
Figure 3
shows the typical water content profiles at 60 and 600 min after single 4- and 13-mm applications. No significant change in water content was observed during 4-mm applications. Although an increase of water content in the vadose zone was observed under 13-mm application, no significant downward solute infiltration took place under either 4- and 13-mm application rates.
View larger version (23K):
[in this window]
[in a new window]
|
FIG. 3. Volumetric water content profiles 60 min and 600 min after the 4- and 13-mm rainfall applications.
|
|
Apparent solute leaching took place only during the highest application rate (26 mm). The solute profiles after 52 and 104 mm of total rain at 26-mm applications are shown in Fig. 4
. The profiles of chloride and nitrate coincide with each other, and piston-type infiltration was clearly observed.
View larger version (24K):
[in this window]
[in a new window]
|
FIG. 4. Chloride and nitrate profiles after 52 and 104 mm of total rain by repeating 26-mm rainfalls in the homogeneous medium.
|
|
In the homogeneous medium, chloride and nitrate profiles coincide with each other under all experimental conditions, indicating that nitrate behaved as a conservative tracer. No degradation of nitrate was observed during the experimental period. Both solutes reached the groundwater table as piston flow only by 26-mm applications. The properties of the profiles observed are summarized in Table 3. On a daily basis, 12.4% of the rainfall events from 2001 to 2005 in Tsukuba, Japan, exceeded 26 mm per day (Table 1). Thus, nitrate applied from the ground surface would not be leached by 87.6% of rainfall events if the porous medium is homogeneous.
View this table:
[in this window]
[in a new window]
|
TABLE 3. Summary of nitrate profile properties observed at the end of each experimental case. Depths of the concentration peak and that of nitrate front are in meters below the surface. Two peaks are in one case for macroporous medium.
|
|
Two-Layered Medium
In the two-layered system in which fine sand was covered by 0.2 m of organic soil, evaporation between events exceeded rainfall after daily 4-mm applications, as observed in the homogeneous medium. No significant downward solute movement was observed.
Figure 5
shows the chloride and nitrate concentration profiles after 52 and 104 mm of total rain from 13-mm applications. The profiles were obtained 96 h (4 rainfalls) and 192 h (8 rainfalls) after the tracer application, respectively. The nitrate concentration in the soil water was inherently high (range, 55.1–104.0 mg L–1) probably due to preceding land use (vegetable cultivation). Nitrate did not leach beyond 0.3 m below the surface during the 13-mm applications. However, both nitrate and chloride concentrations increased in the soil (0.1- to 0.2-m depth) during the first four applications and then started to decrease, indicating that solute leaching followed by dilution due to the repeated rainfall applications.
View larger version (22K):
[in this window]
[in a new window]
|
FIG. 5. Chloride and nitrate profiles after 52 and 104 mm of total rain by repeating 13-mm rainfalls in the two-layered system.
|
|
The ratio of nitrate concentration to chloride concentration should be constant if both behave conservatively. The nitrate/chloride ratio decreased from 2.5 to 2.2 at 0.15 m below surface and 0.88 to 0.83 at 0.2 m below the surface in the soil layer between 96 and 192 h after the tracer application. In contrast, the nitrate/chloride ratio was constant or slightly increased in the sand layer below the top soil. The decrease of the ratios suggests the occurrence of nitrate degradation. Although the denitrification rate was small, hot spots where denitrification takes place are often found in the unsaturated zone (Parkin, 1987). The topsoil layer is inherently heterogeneous and has structure in which there are anaerobic parts. These anaerobic parts may have acted as hot spots of denitrification in our lysimeter.
Figure 6
shows components of the average water balance (soil-water storage, evaporation, and discharge) and estimated chloride and nitrate distributions after 104 mm of total rain for three different application rates. More than 80% of water was lost to evaporation during 6.8-mm rains, and 100% of the chloride remained in the porous medium, while 8.2% of nitrate was estimated to have undergone denitrification. Approximately 50% of the cumulative rainfall water was lost to evaporation under 13-mm application cases. A slight loss of mass was estimated for nitrate, whereas 100% of the chloride remained in the porous medium. During 26-mm applications, approximately one-third of the rainfall was lost to evaporation, and another one-third was lost to discharge. The distributions of chloride and nitrate were quite similar, and no evidence of denitrification was observed. The solute residence time (the experimental period over which the water balance was estimated) was 16 d for the 6.8-mm application, 8 d for the 13-mm application, and 4 d for the 26-mm application. A short residence time did not allow nitrate to undergo denitrification during 26-mm application.
View larger version (37K):
[in this window]
[in a new window]
|
FIG. 6. Average water, chloride, and nitrate distributions after 104 mm of total rain as percentage of total input by mass.
|
|
In the two-layered system, water and a significant amount of solute reached the groundwater table only at the 26-mm applications. On the other hand, denitrification was observed in the top soil layer under light rains.
Macroporous Medium
Effects of rainfall patterns on nitrate transport in a macroporous medium were examined using the fine sand with artificially generated macropores. Soil water near the surface drained or evaporated quickly after the rain in the macroporous medium, so that it was not possible to obtain water samples at depths shallower than 0.2 m, under 4- and 6.8-mm application-rate conditions. Concentration profiles of chloride and nitrate after 52 mm and 104 mm of total rain at 13-mm application rates are shown in Fig. 7
. Considerable increases in both chloride and nitrate concentrations were observed up to 0.3 m below the surface, indicating fast transport of solutes via macropores followed by quick infiltration of solute from macropores to the matrix. Dilution is estimated to be the main mechanism of the observed concentration decrease.
View larger version (22K):
[in this window]
[in a new window]
|
FIG. 7. Chloride and nitrate profiles after 52 and 104 mm of total rainfall by repeating 13-mm rainfalls in the macroporous medium.
|
|
The concentration profiles of chloride and nitrate under 26-mm applications are shown in Fig. 8
. Double peaks are clearly observed in both profiles, suggesting fast macropore flow followed by matrix flow, which resulted in large vertical spreading of the solute.
View larger version (20K):
[in this window]
[in a new window]
|
FIG. 8. Chloride and nitrate profiles after 104 mm of total rain by repeating 26-mm rainfalls in the macroporous medium.
|
|
Fast nitrate leaching under intermediate applications and double peaks under 26-mm applications were observed due to macropore flow, whereas no effect of macropores was found during light rains.
Expected Change in Rainfall Patterns and Its Effect on Nitrate Leaching
The Japan Meteorological Agency (2006) estimated the precipitation change over Japan due to global warming using the regional climate model (RCM20) with the A2 scenario (IPCC, 2000). They reported that a slight increase of the average annual precipitation and pronounced increases in the variation and intensity of precipitation are expected for the Eastern Pacific region of Japan, which includes Tsukuba and Tokyo. It is also expected that the region will receive precipitation of more than 50 mm d–1 three more days per year than under historical conditions. Our laboratory experiments indicated that significant nitrate leaching occurred only under the precipitation of more than 26 mm in all porous media. Tsukuba received daily precipitation of more than 26 mm for 12 d per year on average during 2001 to 2005, which accounted for 45.3% of the total precipitation (Table 1).
Three more days of precipitation exceeding 50 mm under the A2 scenario will result in at least a 25% increase of intensive rainfalls that cause nitrate leaching. It is known that denitrification occurs in the vadose zone in actual fields, and denitrification in organic soil was observed during our experiments. Andosol in which soil structure is developed covers broad areas of Japanese fields, and it offers spots for denitrification before the nitrate reaches the groundwater table or surface water. Increased precipitation variation will also increase soil-water residence time variation, which may result in increased amount of soil water reaching groundwater without denitrification processes. Sugita et al. (2005) observed that the onset of macropore flow that causes nitrate leaching depends on precipitation intensity and antecedent soil-water conditions. Increased precipitation intensities under climate change will also increase the onset of nitrate leaching Therefore, overall, climate change will significantly increase the chance of nitrate leaching through increased precipitation intensity, frequency of heavy rain, and decreased soil-water residence time.
|
Conclusions
|
|---|
Nitrate movement in response to repeated artificial rainfall events in a homogeneous, layered, and macroporous media were observed and summarized in Table 3. No vertical leaching of the solute was observed under 4-mm applications in all porous media due to water loss to evaporation processes. Significant amounts of nitrate leached to the groundwater only at the highest (26 mm) application rates in the homogeneous and layered media. In a layered medium, with soil cover over sand, the highest nitrate concentration remained in the topsoil under light to intermediate applications, resulting in large vertical spreading of nitrate under intermediate rainfall application cases. Double peaks in the soil-water concentration profile were observed in a macroporous medium under the highest (26 mm) applications, indicating that some nitrate was transported by preferential flow while the rest was transported more slowly by matrix flow. Traces of the nitrate reached groundwater under intermediate (13 mm) rainfalls in a macroporous medium due to preferential flow, which led to broad vertical distribution of nitrate.
Nitrate movement coincided with that of chloride in the homogeneous sand medium. The nitrate/chloride ratio in the topsoil of the layered medium, however, decreased with time under light rains, suggesting the presence of degradation processes. Part of the soil that has structure may be serving as spots for denitrification in the lysimeter when the solute residence time is long.
It was found that heterogeneity of the porous medium causes large dispersion under intermediate to heavy rains and denitrification under light rains. It also affects pathways of leaching under heavy rainfalls.
On a daily basis, 38.5% of rainfall events that Tsukuba received between 2001 and 2005 were less than 4 mm per day, and 12.4% of rainfall events exceeded 26 mm (Japan Meteorological Agency, 2006). Therefore, it is suggested that nitrate did not move downward at all for 38.5% of the rains and experienced minimal downward movement for 87.6% of rainfall events in Tsukuba under current climate conditions. Climate-change scenarios that result in higher frequencies of heavy rains could cause increased nitrate leaching by at least 25%. An increase of dispersion and decreased chance of denitrification in the subsurface system are also expected due to potential climate change.
|
REFERENCES
|
|---|
- Blevins, D.W., D.H. Wilkison, B.P. Kelly, and S.R. Silva. 1996. Movement of nitrate fertilizer to glacial till and runoff from a claypan soil. J. Environ. Qual. 25:584–593.[Web of Science]
- Buttle, J.M., and D.G. Leigh. 1997. The influence of artificial macropores on water and solute transport in laboratory soil columns. J. Hydrol. 191:290–314.[CrossRef]
- IPCC. 2000. Special report on emissions scenarios (SRES): A special report of working group III of the intergovernmental panel on climate change. Cambridge Univ. Press, Cambridge, UK.
- Japan Meteorological Agency. 2006. Climatic statistics. Available at http://www.data.jma.go.jp/obd/stats/data/en/index.html. Japan Meteorological Agency, Tokyo.
- Mantovi, P., L. Fumagalli, G.P. Beretta, and M. Guermandi. 2006. Nitrate leaching through the unsaturated zone following pig slurry applications. J. Hydrol. 316:195–212.[CrossRef]
- Munyankusi, E., S.C. Gupta, J.F. Monrief, and E.C. Berry. 1994. Earthworm macropores and preferential transport in a long-term manure applied typic Hapludalf. J. Environ. Qual. 23:773–784.[Abstract/Free Full Text]
- Omoti, U., and A. Wild. 1979. Use of fluorescent dyes to mark the pathways of solute movement through soils under leaching conditions: 2. Field experiments. Soil Sci. 128(2):98–104.
- Owens, L.B., W.M. Edwards, and M.J. Shipitalo. 1995. Nitrate leaching through lysimeters in a corn–soybean rotation. Soil Sci. Soc. Am. J. 59:902–907.[Abstract/Free Full Text]
- Parkin, T. 1987. Soil microsites as a source of denitrification variability. Soil Sci. Soc. Am. J. 51:1194–1199.[Abstract/Free Full Text]
- Schilling, K., and Y.K. Zhang. 2004. Baseflow contribution to nitrate nitrogen export from a large agricultural watershed, USA. J. Hydrol. 295:305–316.[CrossRef]
- Schoen, R., J.P. Gaudet, and T. Bariac. 1999. Preferential flow and solute transport in a large lysimeter, under controlled boundary conditions. J. Hydrol. 215:70–81.[CrossRef]
- Soulsby, C., J. Petry, M.J. Brewer, S.M. Dunn, B. Ott, and I.A. Malcom. 2003. Identifying and assessing uncertainty in hydrological pathways: A novel approach to end member mixing in a Scottish agricultural catchment. J. Hydrol. 274:109–128.[CrossRef]
- Sugita, F., T. Kishii, and M. English. 2005. Effects of macropore flow on solute transport in a vadoze zone under repetitive rainfall events. IAHS Publ. 297:122–129.
This article has been cited by other articles:
|
|
|
|
 
W. E. Riedell, J. L. Pikul Jr., A. A. Jaradat, and T. E. Schumacher
Crop Rotation and Nitrogen Input Effects on Soil Fertility, Maize Mineral Nutrition, Yield, and Seed Composition
Agron. J.,
July 7, 2009;
101(4):
870 - 879.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
T. R. Green, M. Taniguchi, and H. Kooi
Potential Impacts of Climate Change and Human Activity on Subsurface Water Resources
Vadose Zone J.,
August 23, 2007;
6(3):
531 - 532.
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
