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


     

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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhang, Z.
Right arrow Articles by Liu, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhang, Z.
Right arrow Articles by Liu, G.
Agricola
Right arrow Articles by Zhang, Z.
Right arrow Articles by Liu, G.
Related Collections
Right arrow Surface Water Quality
Published in J. Environ. Qual. 33:1403-1412 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Surface Water Quality

Potential Loss of Phosphorus from a Rice Field in Taihu Lake Basin

ZhiJian Zhang, YinMei Zhu*, PeiYong Guo and GuangSheng Liu

Environmental Technology Center, College of Environmental and Resource Sciences, Zhejiang University, Kuanxian 268, Hangzhou, Zhejiang 310029, P.R. China

* Corresponding author (zym2003{at}hzcnc.com).

Received for publication October 27, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSIONS
 CONCLUSIONS
 REFERENCES
 
Nonpoint-source pollution by phosphorus (P) poses a threat to waters in the Taihu Lake basin in China. The potential transfer of P in rice (Oryza sativa L.) fields through surface drainage and subsurface flow was investigated under simulated conventional irrigation–drainage management. Surface drainage events were conducted to avoid overflow across the plots after heavy rainfall and for rice harvest, at which time P losses were also investigated. This study was conducted in 2001 in a long-term rice field experiment. The experimental plots were treated with 0, 26, or 52 kg P ha–1 as superphosphate or 26 kg P ha–1 with equal parts of P supplied as superphosphate and pig manure. Phosphorus concentrations and loads in field floodwater on plots receiving P rapidly declined in a nonlinear manner before the first drainage, three weeks after fertilizer application. The combined application of fertilizer and manure P resulted in higher P transfer potential in field floodwater than with fertilizer P alone one week after P application. Phosphorus concentrations in interflow water sampled by Teflon suction cups inserted at a depth of 150 to 200 mm gradually increased within two weeks after P application, then declined. The concentration of P in interflow water was related to soil P buildup from long-term P application, as well as recently applied P. The 26 kg P ha–1 treatment (the conventional P rate in this region) resulted in a loss of 0.74 kg total phosphorus (TP) ha–1 and a drainage-weighted average concentration of 0.25 mg TP L–1 from the three surface drainage events. Results indicate that avoiding overflow drainage after P input and extending the time between P application and drainage may reduce P losses from rice paddies.

Abbreviations: DOP, dissolved organic phosphorus • DRP, dissolved reactive phosphorus • TDP, total dissolved phosphorus • TP, total phosphorus • TPP, total particulate phosphorus


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSIONS
 CONCLUSIONS
 REFERENCES
 
RESULTS OF WATER MONITORING have shown that less than 10% of the water body of Taihu Lake is mesotrophic, while more than 90% is eutrophic (Chen, 2001). Nonpoint-source pollution and release of P from lakebed sediment are two major sources of P contributing to eutrophication of this lake (Zhang et al., 2001). In agricultural systems, P mobilized from soil may be transported by drainage water to surface waters, where it accelerates the rate of eutrophication (Sharpley and Withers, 1994; Sims et al., 1998). In the Taihu Lake basin, rice fields account for more than 75% of the agricultural land. Therefore, understanding the mechanisms of P loss from rice fields is necessary in developing management strategies to protect water quality in this area.

Hydrology is the single most important factor in P transfer (Haygarth and Jarvis, 1999). Surface runoff is usually a horizontal pathway that is spatially limited and temporarily confined to rains of high magnitude and high intensity (Heathwaite and Dils, 2000). Interflow or subsurface runoff is a term commonly used to describe water moving vertically into surface soil and then laterally, such as at the interface of A and B horizons (Nash and Halliwell, 1999).

Compared with nonflooded soil systems, water management of rice fields is different, because it is largely controlled by irrigation and drainage practices. Theoretically, timing of surface drainage from rice fields depends on the stage of rice growth. Therefore, P in field floodwater (P load) can be considered as potential P for transfer, when considerable water is held in flooded rice fields by the surrounding earthen berms. However, poor water management, such as frequent deep irrigation and severe lateral loss of water through leaky berms, could change potential P transfer into actual P loss. Additionally, heavy rainfall or unexpected storm events can override the "normal" hydrological processes and result in major surface drainage events with significant transfer of P from the rice field into nearby waters (Wang et al., 2001; Zhang et al., 2002). In terms of environmental impacts, the potential P transfer due to P load in field floodwater, unexpected surface drainage caused by heavy rainfall, and seasonal drainage for rice growth should be examined. Unfortunately, few comprehensive field studies have been conducted.

Existence of a plowpan, which develops below the cultivated horizon under long-term rice cultivation, favors lateral water movement and P transfer from the cultivated soil horizon (interflow), particularly when there is a significant water level difference between a flooded rice field during the growing season and a nearby ditch. Additionally, some cultivated horizon interflow will occur during planned drainage before rice harvest. Therefore, interflow may be another important means for transfer and loss of P from rice fields. As for P leaching in the soil profile of nonflooded land, P transport by preferential flow generally outweighs matrix flow (Stamm et al., 1998; Sims et al., 1998; Heathwaite and Dils, 2000; Nash and Halliwell, 1999). Until now, the characteristics and mechanisms of P transfer by interflow from the cultivated horizon in rice fields have been poorly understood.

The intensity and quantity of P losses have been found to vary as a function of numerous factors, including P application rate, rainfall intensity, form of P applied, and method of application (Sharpley et al., 1994; Carpenter et al., 1998; Sims et al., 1998). Although preliminary results on P change in floodwater of rice fields were reported (Wang et al., 2001), more information is needed on P potential transfer and drainage losses related to actual irrigation–drainage management in rice fields. The objective of this study was to develop an improved understanding of the potential transfer of P through surface floodwater and interflow under simulated conventional deep irrigation of rice fields in the Taihu Lake basin. Surface drainage events in this field experiment were conducted to avoid overflow across the plots after heavy rainfall and for rice harvest, at which time P losses were also investigated.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSIONS
 CONCLUSIONS
 REFERENCES
 
Study Area
A long-term rice field experiment was established to investigate P status and loss potential in rice fields receiving a repeated annual input of P fertilizer, with or without manure P. The field experiment was located at Yuhang Agricultural Research Station (30°30' N, 120°18' E) in the southeastern part of the Taihu Lake basin. The location has a subtropical monsoon climate with an average annual rainfall of 1550 mm. Daily rainfall from July to October consistent with the rice season was recorded in 2001 (Fig. 1). The rainfall pattern shows an annual rainy season that usually occurs in August. The dominant soil type at the station is a Blue-purple paddy soil (Mollic Endoaquepts). The cultivated surface horizon (200 mm) has 3% sand, 47% silt, and 50% clay. Long-term rice cultivation has resulted in the development of a plowpan below the cultivation horizon (350–600 mm). In this region, local farmers routinely apply approximately 26 kg P ha–1 in late July or early August, either as inorganic P fertilizer or as a mixture of fertilizer and manure to support one crop of rice and an over-wintering crop, such as wheat (Triticum aestivum L.) or rape (Brassica napus L.). The local rice fields are normally flooded to approximately a 100-mm depth by irrigation or from rainfall before rice harvest.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 1. Daily rainfall recorded at the study site from July to October, 2001.

 
Rice Field Plot Experiment
The field plot experiment was established in May 1997. Twelve 2.6- x 5.4-m plots were constructed in two parallel rows. The construction of the plots, including the ridges, trenches, berms, and inlets and outlets was described by Zhang et al. (2002). All plots were irrigated 5 d before P application to give a floodwater depth of 100 mm, into which rice was transplanted from 1998 to 2001, except in 1997 when P fertilizer was applied before irrigation due to direct seeding of rice. In 2001, one month before P application, Teflon suction cups (5-µm pore size) were inserted in the middle of each plot to a depth of 150 to 200 mm.

Phosphorus Application Records
The experiment was conducted as a completely randomized block design with three replicates for each treatment. From 1997 to 2000, P was annually applied at rates of 0, 53, and 106 kg P ha–1 as superphosphate, and 53 kg P ha–1 with equal quantities of P supplied as superphosphate fertilizer and as pig manure. In 2001, the P application rates were reduced by one-half, that is, 0 kg P ha–1 P (P-0), 26 kg P ha–1 (P-26), 52 kg P ha–1 (P-52), and 26 kg P ha–1 with equal P quantities from superphosphate and manure [P-26 (M)]. The rates of P-26 and P-26(M) are similar to those used by local farmers. Concentrations of TP and water-soluble P in dried manure were 0.52 and 0.32%, respectively. Phosphorus content in superphosphate fertilizer was 6.1% TP and 5.2% water-soluble P. The amount of P applied to each plot was calculated based on water-soluble P content of superphosphate or manure.

The P fertilizer and manure were broadcast and incorporated into soil to a depth of 100 mm in 1997 for direct seeding of rice and to a depth of 50 mm from 1998 through 2001 for transplanting of rice. All plots received 170 kg N ha–1 (as urea) and 50 kg K ha–1 (as KCl). In 2001, 20- to 28-d-old rice seedlings were transplanted at 150- x 150-mm spacing, P fertilizer was applied on 31 July, and rice was harvested 15 November.

Plot Irrigation–Drainage and Water Sampling
In 2001, irrigation water samples were collected for chemical analysis (Fig. 2) from the Lingzhi River, located approximately 100 m from the experimental site. During the growing season, plots were regularly irrigated to a depth of 100 mm to simulate the conventional deep-irrigation practice in this area, when floodwater in plots was less than 5 mm deep. Floodwater depth was measured with a ruler fixed to a brick that was buried in each plot flush with the soil surface. Details on time and amount of irrigation water applied in 2001 are given in Fig. 3.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. Concentrations of total phosphorus (TP), total dissolved phosphorus (TDP), dissolved reactive phosphorus (DRP), total particulate phosphorus (TPP), and dissolved organic phosphorus (DOP) in irrigation water samples collected from Lingzhi River on 3 Aug. to 20 Oct. 2001.

 


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 3. Irrigation and cumulative P input over the six irrigation events, August to October, 2001. Changes in total phosphorus (TP), total dissolved phosphorus (TDP), dissolved reactive phosphorus (DRP), total particulate phosphorus (TPP), and dissolved organic phosphorus (DOP) are shown.

 
At regular intervals during the 2001 season when field floodwater was maintained, six to eight water samples were taken at random from each plot using a syringe and were combined to form a composite sample. Simultaneously, the depth of floodwater in the plots was measured. The first two surface drainage events occurred on 23 August and 1 September due to excessive rainfall (Fig. 1), while the last event occurred on 20 October as the plots were prepared for rice harvest. The drainage and water sampling procedures in each plot were as follows: (i) depth of water in the plot was measured, then the plot drainage outlet (200 mm in width) was opened; (ii) samples of drainage water were collected in a 100-mL plastic beaker at the plot outlet every 30 s until the brick was exposed; and (iii) samples were bulked in a clean barrel and after mixing thoroughly, a 500-mL subsample was taken from the barrel for analysis.

Interflow water samples were collected in porous cups with 0.05 MPa vacuum supplied by a portable vacuum pump. Before sampling, 0.25 mL of 4.0 M HCl was injected into each flask to reduce the pH of the expected 150-mL water sample to 2 to preserve the sample. Sampling was conducted approximately every 10 d while the plots were flooded.

Soil Analysis
Soil samples were taken from the 0- to 200-mm layer of each plot before the application of P fertilizer in 2001. Samples were air-dried and sieved (<2 mm) before chemical analysis (Table 1). Soil and swine manure analyses were conducted as described by Zhang et al. (2002). Soil P adsorption index was determined by a modified procedure (Mozaffari and Sims, 1994) as follows: 25 mL of 60 mg P L–1 solution, prepared by dissolving KH2PO4 in 0.01 M CaCl2, was added to 1 g of soil and equilibrated for 18 h. The difference in the concentrations of dissolved reactive P in solution before and after equilibration was considered as sorbed by the soil.


View this table:
[in this window]
[in a new window]
  		Table 1. Selected parameters of soil samples from the 0- to 200-mm cultivated horizon. Soil samples were taken on 30 July 2001.

 
Water Analysis
One-half of each water sample was filtered through a 0.45-µm Millipore (Billerica, MA) filter within 12 h of collection. The other one-half was kept as an unfiltered sample. The dissolved reactive phosphorus (DRP) in the undigested filtered samples was directly determined spectrophotometrically at 880 nm (Murphy and Riley, 1962). When P concentrations were expected to be very low, P was determined by the method of Rao et al. (1997). Total P concentrations in both filtered and unfiltered waters were analyzed after samples were digested by the method of Haygarth and Jarvis (1997). The digestion process used 20 mL of sample, which was pipetted into a 75-mL digestion tube, mixed with 0.15 g K2S2O8 and 1.0 mL of 0.505 M H2SO4, and autoclaved (tube lid untightened) at 121°C and 0.14 MPa for 1 h. Phosphorus concentrations of acid-digested unfiltered and filtered water samples represent total phosphorus (TP) and total dissolved phosphorus (TDP), respectively. Total particulate phosphorus (TPP) is assumed to be the difference between TP and TDP, while the difference between TDP and DRP is assumed to be dissolved organic phosphorus (DOP) (Hooda et al., 1999). A Shimadzu (Kyoto, Japan) TOC-500 analyzer was used to quantify total organic carbon (TOC) of interflow water samples.

Statistical Analyses
Measured concentrations of different P forms were multiplied by the mass of surface floodwater in the rice field plots and the mass of surface drainage water to obtain P load and P loss, respectively. Total P losses were summed from individual P losses over three drainages, while net P losses were calculated by total P losses minus cumulative P inputs from irrigation events. To calculate a drainage-weighted average P, total P losses were divided by the summed water mass of drainage. Net P loss ratio was expressed as percentage of net P losses to P application rate. Since no data on the water mass of interflow were obtained in this rice season, the actual losses of TP or DRP from interflow were not presented.

One-way analysis of variance was conducted to test the effects of different P treatments on amounts of P forms in surface and interflow water samples. Statistical differences among individual treatments were determined using Duncan's least significant difference (LSD) test at the 0.05 probability level.


   RESULTS AND DISCUSSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSIONS
CONCLUSIONS
 REFERENCES
 
Phosphorus Transfer Potential in Field Floodwater
Phosphorus Concentrations in Floodwater
One day after P application (1 August), TP had increased to 7.59, 14.54, and 5.93 mg P L–1 for P-26, P-52, and P-26(M) treatments, respectively, compared with 0.21 mg P L–1 from plots without application P (Fig. 4a). Simultaneously, DRP increased to 4.87, 8.53, and 2.58 mg P L–1 for P-26, P-52, and P-26(M), as compared with 0.10 mg P L–1 for the control (Fig. 4b). These data indicate that increasing rates of chemical fertilizers or manure increased levels of P in field floodwater. By the fifth day (5 August) after P application, TP and DRP concentrations had declined significantly to 1.49, 2.88, and 1.89 mg TP L–1, and 0.90, 1.45, and 0.81 mg DRP L–1, for treatments P-26, P-52, and P-26(M), respectively. Concentrations of TP and DRP in P-0 plots remained nearly unchanged. Afterward, P concentrations continued to decline from Day 5 to Day 19 after application of P but at a decreasing rate of decline. Although P contents in soil sampled before P was applied in 2001 had increased to a range of 86 to 134 mg Olsen P kg–1 (Table 1), compared with 67 mg P kg–1 before the initiation of the experiment in 1997 (Wang et al., 2001), P in field floodwater due to P fertilizer application to topsoil could be rapidly removed over time due to continual sorption of P by soil (Schulthess and Sparks, 1991). Therefore, extending the time between P application and expected field drainage may reduce P losses.



View larger version (39K):
[in this window]
[in a new window]
  		Fig. 4. Concentrations of total phosphorus (TP), total dissolved phosphorus (TDP), dissolved reactive phosphorus (DRP), and total particulate phosphorus (TPP) in field floodwater after P application. P-0, 0 kg P ha–1; P-26, 26 kg P ha–1 as superphosphate; P-52, 52 kg P ha–1 as superphosphate; P-26(M), 26 kg P ha–1 with equal P quantities of superphosphate and manure. Vertical bars above and below each data point are standard error bars.

 
Due to heavy rainfall on 22 August (78 mm) and 30 August (55 mm) (Fig. 1), the floodwaters were drained on 23 August and 1 September to avoid overflow across the plots. Within this period, TP was commonly lower than in samples collected before 23 August; however, TP on August 29 was slightly higher than on August 26 and was still consistent with P application levels (Fig. 4a). Notably, the TP concentrations in plots receiving P were higher than in irrigation water (0.12 mg TP L–1) on 27 August (Fig. 2). Between 23 and 27 August, heavy rainfall (Fig. 1) became the sole water supply to the plots; P in rainwater (not shown) was negligible. The DRP concentration (Fig. 4b) was almost the same as that of TP, implying that some soil-adsorbed fertilizer P was released into the floodwater. After 2 September, the TP concentrations in field floodwater ranged from approximately 0.1 to 0.5 mg P L–1 and were not linked to applied P levels.

The curves of TDP (Fig. 4c), TPP (Fig. 4d), and DOP (not shown) were quite similar to that of TP. The cause of the sharply reduced concentrations of TPP found on 26 and 29 August in the P-0 plots after first field drainage is unclear, but one possibility could be the removal of fine soil aggregates suspended in solution through field drainage (He et al., 1995). Another reason may be that the portion of soil particles suspended by rainfall normally has more P readily sorbed from fertilizer (Sharpley et al., 1994), resulting in higher TPP from the plots that received P than from the plots not receiving P.

To better understand the characteristics of P transfer potential during the period from P application to the first drainage event, nonlinear regression equations relating changes in P concentrations with respect to time were developed. Phosphorus concentrations in all forms in field floodwater of plots receiving P declined significantly in a nonlinear manner (Table 2). For the P-0 plots, this decline in concentration was significant only for TDP and TPP.


View this table:
[in this window]
[in a new window]
  		Table 2. Regression equations relating P concentration and P load in field floodwater with respect to time (t, days) from P application on 31 July to the first drainage event on 23 Aug. 2001.

 
The patterns of P loads for TP (Fig. 5a) and DRP (Fig. 5b) (TDP, TPP, and DOP not shown) in field floodwater of fertilized plots were similar to those of P concentrations. Before the first drainage, P loads in the forms of TP, TDP, and TPP after P application declined significantly with respect to time and could be characterized with nonlinear regression equations (Table 2). There were no significant nonlinear relationships between the P loads and time for the P-0 plots. An appreciable amount of the annual fertilizer P input may be lost, if either heavy rainfall occurs or floodwater is lost due to the poor water management closely following fertilizer application. These results are different from those studies of nonflooded land agricultural systems (Sharpley et al., 1994; Hooda et al., 1996; Heathwaite et al., 1998), where P transfer normally is associated with transport of soil particles through surface runoff caused by rainfall.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 5. Loads of total phosphorus (TP) and dissolved reactive phosphorus (DRP) in field floodwater after P application. P-0, 0 kg P ha–1; P-26, 26 kg P ha–1 as superphosphate; P-52, 52 kg P ha–1 as superphosphate; P-26(M), 26 kg P ha–1 with equal P quantities of superphosphate and manure.

 
Effect of Phosphorus Source on Transfer Potential
Within the first week after application (Fig. 4), TP, TDP, and DRP concentrations in floodwater of plots treated with inorganic fertilizer only (P-26) were higher than in those plots receiving the combined application of fertilizer and manure [P-26(M)]. Afterward, the trend was reversed. There are at least two possible explanations. First, P would be released from manure by mineralization (Sharpley et al., 1994), which gradually increased levels of dissolved P. Second, CaCl2–P in plots receiving combined application of P fertilizer and manure was significantly higher than that of fertilizer P only (Table 1), implying that P application combined with manure contributes to greater P transfer potential from soil. This is reasonable because soil CaCl2–P mainly represents soil water-soluble P and is easily transferred by runoff and drainage. The TPP concentrations for P-26(M) remained higher than for P-26 during the entire rice season (Fig. 4d). Phosphorus blended with organic matter leads to strong water solubility (Schulthess and Sparks, 1991; McDowell and Sharpley, 2001), which could contribute a relatively higher proportion of particulate P from the combined treatment than from fertilizer only. The combined application of fertilizer and manure P [P-26(M)] maintained higher P transfer potential in field floodwater than the fertilizer P only treatment (P-26) after the first week following P application (Fig. 4).

Phosphorus Transfer Potential in Field Interflow
Phosphorus Concentrations in Interflow
Phosphorus concentrations in interflow water varied widely, ranging from 0.08 to 0.58 mg TP L–1 (Fig. 6a) and from 0.06 to 0.55 mg DRP L–1 (Fig. 6b). These concentrations were lower than concentrations published in other studies. For examples, Smith et al. (1998) found that the volume-weighted concentrations of molybdate reactive phosphorus (MRP) in Teflon suction cups installed in grassland soils ranged from 0.22 to 12.20 mg MRP L–1. In a similar study, P concentrations sampled by a suction lysimeter at a 200-mm depth in grazed land ranged from 0.05 to 1.26 mg TP L–1 and from 0.02 to 1.03 mg DRP L–1 (Heathwaite and Dils, 2000). Further analysis showed that the dominant P form was DRP in all P treatments over eight sampling events with the means of the ratio of DRP to TP ranging from 0.72 to 0.87. Our findings are similar to those of Heathwaite and Dils (2000), who reported that more than 60% of TP was DRP in soil water obtained by a suction lysimeter at a 200-mm depth.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 6. Concentrations of total phosphorus (TP) and dissolved reactive phosphorus (DRP) in interflow water samples at the 150- to 200-mm depth. Samples collected from 30 June to 10 Oct. 2001; P applied on 31 July. P-0, 0 kg P ha–1; P-26, 26 kg P ha–1 as superphosphate; P-52, 52 kg P ha–1 as superphosphate; P-26(M), 26 kg P ha–1 with equal P quantities of superphosphate and manure. Vertical bars above and below each data point are standard error bars.

 
Temporal changes in the P concentrations in interflow (Fig. 6) were distinctly different than that of P in field floodwater. These data show that both TP and DRP concentrations in interflow water increased from 30 July to 13 August regardless of P treatment, and followed the same order as both P application rates and soil test P (Table 1). Two factors were particularly responsible for the initial increase. First, approximately 20 d after flooding, reduced conditions in the cultivated layer were completely developed. Redox potential of seepage waters throughout the rice growing season showed that Eh (approximately 80 mV) was initially positive in the paddy soil to a depth of 20 cm immediately after flooding, then rapidly decreased to highly reduced conditions (around –280 mV) on Day 20 after flooding (Maruyama and Tanji, 1997). Second, more P is generally released from flooded soil than from drained soil (Sanyal and De Datta, 1991; Sims et al., 1998; Pant et al., 2002). These factors contributed to P concentrations reaching a peak at about 20 d after tillage and flooding of the experimental plots. From 13 August to 4 September, P concentrations in interflow gradually declined, and afterward reached relatively constant values that were lower than original concentrations on 30 July. Phosphorus concentrations in the interflow were reduced in part due to uptake by rice that had entered the stooling stage approximately two weeks after transplanting.

As for water movement, hydrological pathways in the soil profile may have changed in the flooded field for rice growth. The formation of cracks by shrinkage of the soil matrix during dry periods allows for preferential flow through the soil profile (Heckrath et al., 1995; Stamm et al., 1998; Beauchemin et al., 1998; Simard et al., 2000), which can readily transfer fertilizer P from topsoil to subsurface water. However, the plots were puddled to a depth of 200 mm after being continuously flooded. Thus, the dominant water movement in the cultivated horizon was matrix flow rather than preferential flow, otherwise a peak in P concentrations would have occurred within a few days after P application rather than 14 d after application (Fig. 6). Phosphorus concentrations in the interflow water were consistent with Olsen- and CaCl2–extractable soil P (Table 1), although soil P was not measured after application of P fertilizer in this year. Notably, whether P was applied to plots or not, there were individual peaks in TP and DRP concentrations. Based on these observations, it can be concluded that P concentration and P transfer potential in interflow water were related to soil P buildup from long-term P application, as well as recently applied P.

Manure Application and Phosphorus Potential Transfer in Interflow
Unlike the P in floodwater, P concentrations in interflow water from plots receiving the combination of manure and fertilizer [P-26(M)] were higher than those from plots receiving only P fertilizer at the same rate (P-26), on most sampling dates (Fig. 6). Additionally, the overall mean of DRP to TP ratio in P-26(M) (0.87) was significantly (P < 0.05, Duncan's test) higher than that of P-26 (0.74). After four years (1997–2000) of continuous P application, soil total P, Olsen P, and P sorption index did not differ between the P-26 and P-26(M) treatments before field activities in 2001 (Table 1). However organic C accumulation in P-26(M) was significantly higher than that of P-26 (Table 1). Concentrations of total organic carbon (TOC) in the interflow water from treatment P-26(M) were higher than those of P-26 (Fig. 7). Organic matter may increase P concentrations in soil solution because of its competition with P for sorption sites in ligand exchange (Schulthess and Sparks, 1991), and its capacity to decrease the chemical activity of iron or aluminum, which would precipitate P from solution (Bloom et al., 1979). In addition, the fact that soil CaCl2–P levels in plots receiving P and manure [P-26(M)] were greater than the fertilizer only treatment at the same P rate (P-26) (Table 1) suggests that P solubility in paddy field plots receiving P-26(M) was greater than that of P-26. Therefore, although there were no data on the P losses from interflow, the potential for P transfer in interflow of rice fields following long-term applications of a combination of manure and fertilizer was higher than that of fields receiving only P fertilizer at the same rate.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 7. Concentrations of total organic carbon (TOC) in interflow water samples at the 150- to 200-mm depth. Samples collected from 30 June to 10 Oct. 2001; P applied on 31 July. P-0, 0 kg P ha–1; P-26, 26 kg P ha–1 as superphosphate; P-52, 52 kg P ha–1 as superphosphate; P-26(M), 26 kg P ha–1 with equal P quantities of superphosphate and manure. Vertical bars above and below each data point are standard error bars.

 
Phosphorus Loss during Drainage and Implication for Minimizing Phosphorus Losses
Phosphorus Losses in Surface Drainage
Concentrations of P in drainage water from rice paddy plots (Table 3) exceeded the threshold of 0.035 mg L–1, the minimum level needed to trigger eutrophication (Organisation for Economic Co-Operation and Development, 1982). Total P losses over three drainage events were not significantly different between P-26 and P-0, but increasing P application rate (P-52) significantly increased total TP and TPP losses. For dissolved P (e.g., TDP and DRP), significant differences were observed with P-52 > P-26 > P-0. However, total DOP losses increased in plots receiving P, while no significant difference was found between P-26 and P-52. Similar characteristics were observed for drainage-weighted average P (Table 3). Discounting cumulative P input from irrigation (Fig. 3), significant differences were found following P-52 > P-26 > P-0 for net P losses in terms of TP, TDP, and DRP. Except for DOP, net P loss ratio for TP, TDP, DRP, and TPP followed P-52 > P-26. Total P losses, net P losses, drainage-weighted average P, and net P losses were significantly higher for P-26(M) than for both P-26 and P-52.


View this table:
[in this window]
[in a new window]
  		Table 3. Phosphorus concentrations and losses of total phosphorus (TP), total dissolved phosphorus (TDP), dissolved reactive phosphorus (DRP), total particulate phosphorus (TPP), and dissolved organic phosphorus (DOP) over three surface drainages from an experimental rice field.

 
Approximately 70 to 80% of total TP losses, 59 to 81% of total TDP losses, 53 to 80% of total DRP losses, 52 to 87% of total TPP losses, and 53 to 83% of total DOP losses occurred during the first drainage event (23 August), 24 d after P application (Table 3). Since P concentration and P load decline decreased curvilinearly (Table 2), more P would probably have transferred to water, if the first surface drainage had occurred earlier than 23 August.

Net P losses from plots receiving no P (P-0) gave negative values (i.e., gains) for all forms except TPP (Table 3). Moreover, application of 26 kg P ha–1 (P-26) also resulted in net P gains for TDP and DOP and nearly zero net losses for TP and TPP. Local P practices (26 kg P ha–1) resulted in 0.74 kg TP ha–1 of total P losses, 0.001 kg TP ha–1 of net P losses, and 0.25 mg TP L–1 of drainage-weighted average P over three surface drainage events (Table 3). The relatively high level of P in irrigation (Fig. 2) and the first two surface drainages conducted exclusively to avoid the overflow contributed to the low value of net TP losses from the local P application rate.

Implication for Minimizing Phosphorus Losses
Local farmers around the Taihu Lake basin commonly perform deep irrigation by applying more than 100 mm at one time following rice transplanting and P application. This practice appears to reduce the amount and frequency of irrigation, but in fact, it probably causes both higher intensity and greater quantity of field drainage, by which P would be transferred. Phosphorus application before early August coincides with the rainy season (Fig. 1). Irrigation water data near the experiment location (Fig. 2) showed that P concentration, particularly for total P (0.19 mg TP L–1) and soluble P (0.09 mg DRP L–1), decreased markedly after early August. After 8 September, P levels remained less than 0.08 mg TP L–1 and 0.04 mg DRP L–1. Meanwhile, more frequent rainfall occurred during August than the rest of the rice season (Fig. 1). Obviously, higher P in the irrigation water during early August probably came from the rice field rather than other sources. Additionally, the P transfer potential characterized by concentrations (Fig. 4) and loads (Fig. 5) in the floodwater sharply decreased, consistent with changes of P in irrigation water. Since loss of P from rice fields is a risk only if field drainage occurs, an improved water management should be lower ponded water depths or alternating wetting and drying rather than conventional deep irrigation–drainage. Frequent surface drainage after P application could be avoided and savings in irrigation achieved if rainfall was exploited and deep irrigation abandoned. Potential for P loss through interflow would also be reduced by decreasing water level differences between rice fields and nearby ditches and by alternating wetting and drying. This alternative approach deserves further investigation.

Although P concentrations (Fig. 4) and P loads (Fig. 5) were initially lower in plots receiving P as a combination of P fertilizer and manure than in plots receiving the same P rate from fertilizer only, the trend reversed after a few days. Application of manure and fertilizer P resulted in significantly greater P losses over three surface drainage events (Table 3). Similarly, P transfer potential in the cultivated horizon was higher than if only P fertilizer was applied (Fig. 6), although initial soil P levels were almost the same (Table 1). With the rapid development of animal production in recent years in the Taihu Lake basin, the potential for degradation of water resources by land application of animal manure is increasing. To decrease P losses, alternative utilization of animal manure should be encouraged.

Since P may be conserved by sorption processes, increasing the length of time between P application and runoff can reduce P concentrations and runoff losses (Sharpley, 1997; Shuman, 2002). One study showed that losses of P compounds in a paddy field watershed could be reduced by using a circular irrigation system (Takeda et al., 1997). Nutrient retention rate reached 97.8% for total P when a multipond system was used (Yan et al., 1998). These results indicate that P nonpoint-source pollution can be reduced by using improved water distribution in large-scale rice areas. More research is needed on nutrient retention by hydrological management in the Taihu Lake basin.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSIONS
 CONCLUSIONS
REFERENCES
 
This study describes the potential transfer and drainage loss of phosphorus in 2001 from a rice field experiment in Yuhang around Taihu Lake basin. Phosphorus from inorganic P fertilizers or manure contributes to elevated P concentrations in floodwater after P application. Phosphorus concentrations and loads in field floodwater from plots receiving P declined curvilinearly before the first drainage, three weeks after P application. The combined application of fertilizer and manure P maintained higher P transfer potential in field floodwater one week after P application than when only P fertilizer was applied. Phosphorus concentrations in interflow water at a depth of 150 to 200 mm gradually increased and reached a peak about two weeks after P application, then declined. Phosphorus transfer potential in interflow water by soil matrix was related to soil P buildup from long-term P application, as well as recently applied P inputs. Local P practices (26 kg P ha–1) resulted in 0.74 kg TP ha–1 of total P losses, 0.001 kg TP ha–1 of net P losses and 0.25 mg TP L–1 of drainage-weighted average P over three surface drainage events. The results indicate that avoiding overflow drainage after P input and extending the time between P application and field drainage may reduce P loss due to soil sorption mechanics. Improved water management should focus on minimizing P losses by lowering ponded water depths or by alternating wetting and drying and eliminating conventional deep irrigation–drainage.


   ACKNOWLEDGMENTS
 
This study was supported by Zheijiang Science and Technology Program on Control of Nonpoint Source Pollution from Agricultural Production (Project 99-2-030). We also express our appreciation to Hongfou Zhen, Shungeng Zhou, and Jin Cheng (Yuhang Agricultural Research Station) for field assistance.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSIONS
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 1177-1182. [Full Text]  




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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhang, Z.
Right arrow Articles by Liu, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhang, Z.
Right arrow Articles by Liu, G.
Agricola
Right arrow Articles by Zhang, Z.
Right arrow Articles by Liu, G.
Related Collections
Right arrow Surface Water Quality

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