JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 20 February 2008
Published in J Environ Qual 37:712-717 (2008)
DOI: 10.2134/jeq2007.0073
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Liang, X.-Q.
Right arrow Articles by Zhang, Z.-J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Liang, X.-Q.
Right arrow Articles by Zhang, Z.-J.
Agricola
Right arrow Articles by Liang, X.-Q.
Right arrow Articles by Zhang, Z.-J.
Related Collections
Right arrow Water Management
Right arrow Nitrogen

TECHNICAL REPORTS

Surface Water Quality

Nitrogen Loss through Lateral Seepage in Near-Trench Paddy Fields

Xin-Qiang Liang, Hua Li*, Ying-Xu Chen, Miao-Miao He, Guang-Ming Tian and Zhi-Jian Zhang

Dep. of Environmental Engineering, College of Natural Resources and Environmental Science, Zhejiang Univ., Hangzhou, 310029,China

* Corresponding author (liang410{at}zju.edu.cn).

Received for publication February 9, 2007.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Summary and Conclusions
 REFERENCES
 
A near-trench paddy field experiment with five urea application rates (0–360 kg N ha–1 in 90-kg increments) was conducted on a paddy soil in the Taihu Lake Region of China to elucidate N losses through lateral seepage during three rice (Oryza sativa L.) growing seasons. The total N (Nt), NH4+–N, and NO3–-N concentrations in the lateral seepage water increased with increasing N rates. The seasonal Nt fluxes by lateral seepage varied from 6.8 to 25.6 kg N ha–1 for urea application rates of 90 to 360 kg N ha–1. Lateral seepage accounted for 4.7 to 6.6% of the Nt applied, implying that lateral seepage was an important pathway of N loss from near-trench paddy fields. The cumulative N loss via lateral seepage was significantly related to N fertilization rate (P = 0.05). Floodwater level was also identified as a main factor affecting N losses via lateral seepage from paddy fields, as indicated by a positive linear relationship (R2 = 0.43) between floodwater level and daily lateral flow during the flooded period (P = 0.05). Under the conditions of these experiments, a shallow floodwater depth of 50 mm, urea application rates of 90 kg N ha–1 or less, and no rainfall within 1 wk after N application reduced N losses by lateral seepage from paddy fields.

Abbreviations: Nt, total N


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Summary and Conclusions
 REFERENCES
 
THE effective abatement of point-source pollution in recent years has highlighted the increasing relative contribution of nonpoint source agricultural pollution to surface water in China (Zhang et al., 2004). Nitrogen (N) fertilizers have been extensively used in China, especially in developed agricultural regions such as the Taihu Lake region of southeastern China, resulting in nonpoint source N pollution from agricultural fields (Guo et al., 2003; Liang et al., 2004; Wang et al., 2004). The mitigation of N fertilizer loss from croplands is important to agricultural and environmental research.

In general, dissolved and particulate N discharges from croplands into water systems occur via storm runoff (Owens et al., 1991; Fierer and Gabet, 2002) and vertical leaching (Panda et al., 1989; Zhu et al., 2000). In recent years, subsurface drainage or lateral seepage has been identified as an important hydrological pathway of N losses from croplands (Jaynes et al., 2001; Strock et al., 2004; Youssef et al., 2006). Paddy fields are typically flooded before rice seedlings are transplanted. After numerous cultivations, a hard pan of 50 to 100 mm thickness often emerges at the interface between the topsoil and subsoil. This hard pan, with a saturated hydraulic conductivity of 0.34 to 0.83 mm d–1 (Chen and Liu, 2002), has been reported as the least permeable soil layer in the rice paddy. The hard pan can limit the infiltration of ponding water into the subsoil, exaggerating the problem of N loss associated with lateral seepage. Paddy fields are surrounded by bunds, and overland runoff from the field occurs only when the height of floodwater exceeds that of the bunds. Thus, water balance and nutrient loss mechanisms from paddy fields are different from those in other types of agricultural fields. Walker and Rushton (1984) found that a water balance could not be achieved if evapotranspiration, storm runoff, surface drainage, and vertical percolation through the plow layer were considered the only sources of floodwater losses from paddy fields. Bouman et al. (1994) further showed that lateral seepage was the only other possible source for floodwater losses. Chen et al. (2002) used a three-dimensional model, FEMWATER, to simulate the process of lateral seepage in a terraced paddy field and found that significant lateral seepage would occur at the wet/dry boundary of the field. These studies showed that lateral seepage may be an important pathway for floodwater losses in paddy fields but did not evaluate the nutrient loss through lateral seepage.

The Taihu Lake Region in Southeastern China has been a major rice-producing region for several centuries. The N (total N [Nt]) concentration in the Taihu Lake varies from 2.0 to 9.0 mg N L–1 (State Environment Protection Administration of China, 2005), which is the main limiting actor to improve the water quality of the Lake. The surface water in the lakes and rivers within the region do not reach the statutory limits for drinking (Zhu et al., 2003). Drainage trenches are common in rice-cultured areas of the Taihu Lake region. The water level in drainage trenches is generally lower than that in paddy fields, producing a gradient that may enhance lateral seepage. The seepage water could carry N and other nutrients directly into the trenches and nearby lakes or rivers, which can result in eutrophication. Lateral seepage from paddy fields near trenches might contribute significantly to agricultural nonpoint-source pollution. The agricultural nonpoint-source pollution investigation of the Taihu Lake Region concluded that lateral seepage contributed to water drainage from paddy fields but did not clearly quantify water and associated N loss through lateral seepage (Guo et al., 2003). Little information about N lateral seepage in near-trench paddy fields is available, and quantifying lateral seepage at the field scale is difficult. The objectives of this field experiment were to develop a method for monitoring lateral seepage in situ, to evaluate lateral seepage and N losses from near-trench paddy fields, and to assess the impacts of floodwater level and urea application rates on N lateral seepage.


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Summary and Conclusions
 REFERENCES
 
A long-term paddy field experiment was established in May 2003 at the Jiaxing Agricultural Research Station (120°40' E, 30°50' N) in the Taihu Lake Region of Southeast China. The location has a subtropical monsoon climate with an average temperature of 28°C in summer and an average annual rainfall of 1200 mm. The dominant soil type at the station is a gleyed paddy soil (clay loam, mixed, mesic Mollic Endoaquepts).

Fifteen 4 x 5 m plots were separated by field bunds, and all plots bordered a 1-m-deep trench on one side. The field bunds were built in 2003 and were made of native soil. The height of the water level was about 500 to 800 mm lower in the trenches than in the fields throughout the 3-yr experiment. Paddy fields were flooded and maintained with 50 to 70 mm floodwater, the common agronomic practice in this region (Cao and Zhang, 2004). During flooded periods, plots were irrigated at 0500 daily to obtain a ponding depth of 50 mm. Floodwater height was measured with a ruler fixed to a brick that was buried in each plot level with the soil surface. Lateral seepage was allowed only through the near-trench side of each plot. The three nontrench sides of each plot were lined with impermeable nylon to a depth of 600 mm. Consistent with area agronomic practices, the paddies were not irrigated during a 10-d drying period at the end of July. There was no artificial surface drainage during the observation periods.

The experiment had five N treatments (0–360 kg N ha–1 in 90-kg increments, namely, R0, R90, R180, R270, and R360) with three replicates. The 15 plots were laid out in a completely randomized block design. The urea fertilizers were applied on 1 July, 9 July, and 9 August, with the ratio of 3:1:1 in each season. All plots received 40 kg P2O5 ha–1 and 150 kg KCl ha–1 on 1 July. All the fertilizers were evenly broadcast by hand in each plot. Rice (Oryza sativa L.) seedlings (25 d old) were transplanted at 150 x 150 mm spacing on 1 July and harvested on 31 October in 2003, 2004, and 2005.

An in situ suck-container was specifically developed for collecting lateral seepage water (Fig. 1 ). The main vessel of the container was wedge-shaped with a suck-surface (0.4 m x 4 m). A fiber filter cloth was fixed at the suck-surface, and all the other surfaces were made of rot-resistant plastic. Each plot had a separate container. Each container was adjacent to the outside of the bund, with its suck-surface kept in contact with the bund so that all water in the container was from lateral seepage. Soil water that seeped laterally into the field bund and passed through the fiber cloth of the suck-surface sampler was stored in the container. During sample collection, all water was removed from the container using a syringe. Surface runoff was collected using one PVC pipe installed through the field bund 100 mm above the plot soil surface (Fig. 1). Runoff water drained into a bucket. Sampler installation was complete 3 mo before the start of the experiment in 2003.


Figure 1
View larger version (50K):
[in this window]
[in a new window]

  		Fig. 1. The collection system for water samples of lateral seepage. B is a gravel-packed buffer zone between the bund and the trench, C is a container for collecting lateral seepage, R is a bucket with a pipe for collecting runoff, and T is a 1-m-deep trench. Lateral seepage water is pumped out from tube 1 using a syringe, and tube 2 is for air connection.

 
Lateral seepage and surface runoff samples were collected daily in the first week after each N application. Thereafter, samples were collected at 3- to 5-d intervals. Water samples were acidified with HCl solution and kept on ice in the field. Samples were filtered to remove particles of diameter >11 µm, and the filtrates were frozen for later analyses. A continuous-flow analyzer (BRAN+LUEBBE, AA3, Germany) was used to determine concentrations of NH4+–N, NO3––N, and Nt in the water samples (Mulvaney, 1996).

The bund soil was sampled in the summer of each year, before rice seedlings were transplanted. Samples were collected from the surface to a depth of 40 cm. The mean sand, silt, clay, and organic carbon contents; bulk density; and soil hydraulic conductivity were determined for the 0- to 40-cm soil (Table 1 ). The vertical percolation rate of the field bunds was nearly 0 mm d–1 in each year.


View this table:
[in this window]
[in a new window]

  		Table 1. Mean sand, silt, clay, organic carbon, bulk density, and saturated hydraulic conductivity of bund soil (0–40 cm depth) in different years.

 
Daily lateral seepage (mm d–1) was calculated by dividing the volume of lateral seepage water removed from the container at each sampling time (mm3) by the suck-surface area (1.6 x 106 mm2) and dividing this value by the time interval (d). Surface runoff (mm d–1) was similarly calculated from the volume of runoff water divided by the area of the plot and the time interval. Seasonal Nt loss (g N m–2 season–1) by lateral seepage and surface runoff were calculated using:

Formula 1[1]
where C is the Nt concentration in lateral seepage or surface runoff, V is the volume of lateral seepage through suction-surface or surface runoff, Aplot is the area of field plot, and i means on the i-day after the rice seedlings transplanting (i = 1~90 in each season).

Statistical analysis of the data was accomplished by standard ANOVA, and paired values were compared by LSD at the 5% level using the SPSS software package (SPSS, 2000).


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Summary and Conclusions
 REFERENCES
 
Water Lateral Seepage
During the periods of field drying (21 July to 31 July in each year), no lateral seepage was observed. However, before and after field drying, lateral seepage occurred every day (Fig. 2 ). The lateral water seepage rate before field drying varied from 0.2 to 25.0, 1.3 to 20.9, and 0.1 to 22.8 mm d–1 in 2003, 2004, and 2005, respectively. The lateral seepage observed after field drying varied from 0.2 to 25.7, 0.2 to 16.0, and 1.1 to 23.8 mm d–1 in 2003, 2004, and 2005, respectively. These variations were mostly dependent on rainfall, which increased the floodwater level during a particular day (Fig. 3 ). The high precipitation on 16 July 2003 and 20 Aug. 2004 increased floodwater height to about 100 mm (Fig. 3). After 28 September, seepage water was recorded as "zero" because there was little input of irrigation or rainwater. The final cumulative lateral seepage was about 350 mm, 293 mm, and 309 mm at the end of the rice season in 2003, 2004, and 2005, respectively (Fig. 2). However, three sides of each paddy field in this study were impervious, so these values might not represent lateral seepage from unlined paddies. Besides, the seepage of the paddy field moved as subsurface flow, which was unlike to the terrace paddies because the lateral seepage from the terraced paddy was mostly unsaturated and flows along the sloping surface (Huang et al., 2003).


Figure 2
View larger version (56K):
[in this window]
[in a new window]

  		Fig. 2. Dynamics of lateral seepage during three rice-growing seasons. Vertical bars above each data column indicate SE.

 

Figure 3
View larger version (59K):
[in this window]
[in a new window]

  		Fig. 3. Dynamics of rainfall, irrigation, and floodwater level during three rice-growing periods.

 
The lateral seepage rate from paddy fields is mostly dependent on characteristics of field bunds (Wickham and Singh, 1978; Chen and Liu, 2002; Huang et al., 2003), especially the age of the bund. Huang et al. (2003) reported that the bund age determined flow direction of lateral seepage and noted that the lateral seepage flow was horizontal in newly built bunds. The bunds at this study site were newly built in 2003. In addition, a hard pan might be present at the bottom of the bund (Fig. 1), suggesting that lateral seepage might be primarily by horizontal flow. Seasonal rainfall in 2004 (290 mm) was about 60% of that in 2003 (446 mm) and 2005 (458 mm), whereas seasonal water output by lateral seepage was similar in 2004 and 2005 and greatest in the first year of the study (Fig. 2). These results suggest that the cumulative lateral seepage was not strongly dependent on seasonal rainfall but may be related to bund age. A significant increase in bulk density (P = 0.05) was observed in the top 40 cm of the bund during the study period, with a concomitant decrease in organic carbon and clay contents (Table 1).

Figure 4 shows the relationship between floodwater level and lateral seepage rate during the flooded period of three rice-growing seasons. The average rate of lateral seepage varied from 0 to 20.9 mm d–1 during flooded nonrainy days (floodwater level = 50 mm), in agreement with previously reported results. Huang et al. (2003) reported that the lateral seepage rate from the bund of a clay paddy soil was about 12.4 mm d–1. The increase in lateral seepage rate with increasing floodwater level was probably a result of conditions affecting the hydrology of this site, including (i) a lower water level in drainage trenches than in paddy fields produced a gradient that encouraged lateral seepage, and (ii) the presence of a hard pan at the bottom of the bund (about 40 cm depth) limited the infiltration of ponded water into the subsoil, which could exaggerate lateral seepage. Walker and Rushton (1984) showed that lateral seepage could be greatly reduced by maintaining shallow depths of floodwater in flat paddy fields. Chen and Liu (2002) also found that differences in the initial floodwater level could influence subsequent seepage. These results suggest that maintaining a shallow floodwater level of 50 mm may be helpful to minimize lateral seepage and associated N loss from near-trench paddy fields.


Figure 4
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 4. Relationship between lateral seepage and floodwater level in flooded period of three rice seasons.

 
N Concentrations in Lateral Seepage
Total N, NH4+–N, and NO3––N concentrations in lateral seepage water responded to the fertilizer applications in 2003, with the peak values appearing 2 or 3 d after each application (Fig. 5 ). The concentration of Nt, NH4+–N, and NO3––N in seepage water increased with increasing fertilizer application rate (e.g., the Nt concentrations at R0, R180, and R360 in 2003 ranged from 0.1 to 5.1, 0.2 to 26.8, and 0.3 to 47.6 mg N L–1, respectively) (Fig. 5). In 2004 and 2005, similar variations were found, although the peak values of Nt, NH4+–N, and NO3––N concentrations were lower than those in 2003 (Fig. 5). The rapid appearance of nitrate and ammonium in lateral seepage water after fertilizer application indicated that urea was rapidly hydrolyzed in the floodwater, in agreement with previous observations that 40 to 80% of the applied urea was hydrolyzed to NH4+ and HCO3– within 24 h after broadcast application to a flooded field (Chowdary et al., 2004). These results suggest that the time interval between fertilizer application and rainfall and/or irrigation was an important factor in controlling N losses via lateral seepage. Application of fertilizer at least 1 wk before rainfall and/or irrigation may reduce N concentrations in lateral seepage and trench water.


Figure 5
View larger version (40K):
[in this window]
[in a new window]

  		Fig. 5. Temporal variations of total-N, NH4+–N, and NO3––N concentrations in lateral seepage water during three rice seasons. Vertical bars above each data point indicate SE. Arrows represent N fertilizer application dates.

 
The mean proportion of NH4+–N in seepage water in all samples collected each year are given in Table 2 . The mean NH4+–N /Nt proportions ranged from 0.57 to 0.66 over the five N rates, indicating that NH4+–N was the main N form in the seepage water. Lateral seepage water is expected to come mainly came from paddy floodwater in which NH4+ is the main N form (Wang et al., 2003). Nitrate-N was also detected in lateral seepage water during all three observation seasons of this study (Fig. 5). Similar to the results for NH4+–N, NO3––N concentrations were greatest shortly after fertilization. This was partly attributed to the release and nitrification of previously fixed NH4+–N. Schneiders and Scherer (1998) reported that the release and mobilization of newly fixed NH4+–N was high, particularly in the rhizosphere of rice plants where the redox potential could be high due to the secretion of O2 from the roots.


View this table:
[in this window]
[in a new window]

  		Table 2. Mean ratio of NH4+–N to total-N in the lateral seepage water. Values in parentheses are 1 SD for triplicate plots.

 
N Loss Flux through Lateral Seepage
In this study, the seasonal cumulative N flux by lateral seepage from a near-trench paddy field ranged from 0.68 to 2.56 g N m–2 (6.8–25.6 kg N ha–1) for five treatments receiving 0 to 360 kg N ha–1 (Fig. 6 ), accounting for 4.7 to 6.6% of the total urea fertilizer applied. These seasonal N loss rates were of the same order of magnitude as those found for vertical leaching in this region. Zhu et al. (2000) reported that seasonal N losses by vertical leaching beyond 60 cm were 5.5 kg N ha–1 and 7.0 kg N ha–1, amounting to 3.1 and 4.3% of the Nt loss, when 300 kg N ha–1 urea was applied to the single- and double-cropped rice land. Moreover, Fig. 6 shows that N losses via lateral seepage were significantly higher than that via surface runoff (P = 0.05) at each N application rate. Only 0.02 to 0.10 g N m–2 season–1 was transported by surface runoff. Additionally, a significantly positive correlation was observed between fertilizer application rates and seasonal Nt losses by lateral seepage at the P = 0.05 level (Fig. 6). Urea fertilizer application increased rice grain yields, but when N fertilization was supra-optimal, no advantage of additional urea was observed (Fig. 7 ). For each season, the crop grain yield significantly increased with the increase of N rates from 0 to 90 kg N ha–1 at p = 0.05 (Fig. 7). However, grain yields were not significantly different for urea application rates of 90 to 360 kg N ha–1. No significant difference in grain yields was observed between years, even in plots receiving zero N fertilizer application (Fig. 7). Therefore, a urea application rate of 90 kg N ha–1 could be recommended to reduce N losses by lateral seepage while maintaining maximum rice production potential.


Figure 6
View larger version (31K):
[in this window]
[in a new window]

  		Fig. 6. Comparison of mean seasonal flux of total nitrogen in lateral seepage and surface runoff under a series of urea fertilizer application rates. *Significant differences (P = 0.05) between lateral seepage and surface runoff at each N rate. Error bars represent 1 SD.

 

Figure 7
View larger version (34K):
[in this window]
[in a new window]

  		Fig. 7. Mean rice grain yield at different urea application rates in 2003, 2004, and 2005. Vertical bars above each data column indicate SE. Yields labeled with the same letter are not significantly different (P = 0.05).

 

   Summary and Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Summary and Conclusions
REFERENCES
 
Hydrology and N management play important roles in determining N loss flux by lateral seepage from flooded paddy fields. Hydrology provides the energy and the carrier for N transfer by subsurface interflow, leaching, and other pathways (Liang et al., 2004; Strock et al., 2004; Zhu et al., 2000). There was a positive linear relationship between floodwater level and daily lateral flow during the flooded period (P = 0.05), suggesting that maintaining shallow floodwater levels such as 50 mm may be helpful to minimize lateral seepage and associated N loss. Although N loss by lateral seepage only accounted for 4.7 to 6.6% of the total urea fertilizer applied, higher fertilizer application rates resulted in higher concentrations of N in lateral seepage water. Under these experimental conditions, an N application rate of 90 kg N ha–1 was observed to provide optimum yields with decreased N loss by lateral seepage compared with higher fertilizer application rates.

Peak concentrations of ammonium and nitrate in lateral seepage water appeared rapidly after fertilizer applications, within 2 to 3 d after urea fertilizer application. Fertilizing at least 1 wk before a rainfall or irrigation may provide the agronomic benefits of supplying the crop with nutrients while minimizing the risks of lateral seepage adversely affecting water quality.


   ACKNOWLEDGMENTS
 
The authors thank the National Key Basic Research Project of China (2002CB410807) and provincial natural science foundation of Zhejiang (Y504247) for funding this study. The authors also thank the associate editor and three anonymous reviewers whose comments significantly improved the manuscript.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Summary and Conclusions
 REFERENCES
 
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.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Summary and Conclusions
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Liang, X.-Q.
Right arrow Articles by Zhang, Z.-J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Liang, X.-Q.
Right arrow Articles by Zhang, Z.-J.
Agricola
Right arrow Articles by Liang, X.-Q.
Right arrow Articles by Zhang, Z.-J.
Related Collections
Right arrow Water Management
Right arrow Nitrogen

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome