Paddy Water Environ (2003) 1:27–33DOI 10.1007/s10333-002-0007-7

A R T I C L E

Masaya Ishikawa · Toshio Tabuchi · Eiji Yamaji

Clarification of adsorption and movement by predicting ammonianitrogen concentrations in paddy percolation water

Received: 30 September 2002 / Accepted: 6 December 2002 / Published online: 15 February 2003� Springer-Verlag 2003

Abstract We noted that ammonia nitrogen was notadsorbed by the cultivated layers of highly permeablepaddy fields during the initial fertilization period, butreached the lower layers relatively early. In our study, weconsidered an exponential equation from an aqua-envi-ronmental perspective with the goal of obtaining goodgrowth of rice plants in order to estimate the concentra-tions and integrated volume of ammonia nitrogen accom-panying paddy percolation. Using this exponentialequation, we were able to derive a relation between timeand concentrations of paddy percolation water, andhypothesized that if percolation rates were less than10 mm/day, percolation would have no effect on ricegrowth, while simultaneously helping to maintain thegood water quality of the extra-paddy environment. Wealso clarified the differences between the potentialammonia nitrogen adsorption volume derived from theCEC value and the integrated amount of ammonianitrogen water in soil, and considered the causes fromthe perspectives of solute movement and water move-ment.

Keywords Water quality control · Paddy field irrigation ·Percolation rate · Pore volume · CEC

Introduction

Background

One of the most important forms of inorganic nitrogenneeded for the growth of rice plants is ammonia, and it isknown that the soil layers of rice paddies have thecapacity to adsorb ammonia nitrogen. This has beencorroborated by the fact that rice plants develop systemsfor utilizing ammonia nitrogen based on their physiolog-ical responses to such growth environments. According tothe CEC values, which represent the total volume ofpositive ions that the soil adsorbs, Japanese soils are in therange of 0.2~0.3 mol/kg (Okajima 1989). In more specificterms, at NH4

+, 1 mol=18,000 mg (N:14,000 mg);therefore, 1 kg of CEC 0.2 mol soil has been calculatedto have the ability to adsorb 2,800 mg of NH4

+-N.Assuming 1,000 tons of dry soil per 1 ha of cultivatedland gives us an adsorption of 2,800 kg of NH4

+-N for thisunit area. Since the volume of standard nitrogen fertili-zation in Japan is 100 kg/ha, we can see that there isconsiderable adsorption of NH4

+-N into the soil.However, various studies have reported that nitrogen

outflow load causes fluctuations in the percolationvolume (Takamura et al. 1979; Kondo et al. 1992;Tabuchi et al. 1992). In particular, studies on ammonianitrogen percolating into and running off from paddyfields showed that the ammonia nitrogen concentrationsin surface water were as high as 25~100 mg/L during theinitial periods of irrigation and fertilization (Takamura etal. 1976, 1977). This is nothing more than the loss offertilizer components to percolation during that year and/or the discharge of soil nutrients or fertilizer componentsfrom the previous year. These studies showed that inhighly permeable fields, ammonia nitrogen not adsorbedin the surface soil reaches the lower soil layers relativelyquickly; however, there is no clear understanding of howpermeability affects the movement of ammonia nitrogenin the surface soil of paddy fields (Ishikawa and Tabuchi1998). High ammonia nitrogen concentrations in soilduring irrigation periods may result in lodging of rice

M. Ishikawa ())Graduate School of Agricultural & Life Sciences,The University of Tokyo,1-1-1 Yayoi, Bunkyo-ku, 113-8657 Tokyo, Japane-mail: aimasa@mail.ecc.u-tokyo.ac.jp

T. Tabuchi,4630-104, Ami-machi, Ami, Inashiki-gun, 300-0331 Ibaraki, Japan

E. YamajiGraduate School of Frontier Sciences,The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,113-0033 Tokyo, Japan

plants due to root rot, and may lead to the leaching of highnitrate nitrogen when rainfall occurs during non-irrigationperiods (Takeda et al. 1991). To utilize rice fields forprotecting the aqua environment and stabilizing foodproduction, nitrogen applied to and contained in the soil(i.e., nitrogen mineralized from soil, nitrogen in irrigationwater or rainwater, and nitrogen fixed around rice roots)and in rice plants must be appropriately balanced bycontrolling irrigation and fertilization (Bolt and Bruggen-wert 1979, Ishikawa and Tabuchi 1998). Clarification ofthe appropriate balance of substances, especially nitrogen,is also needed.

Objectives

The purpose of this study is to predict the ammonianitrogen concentrations in percolating water before riceseedlings are planted. The results of laboratory experi-ments using soil columns are reported. In these experi-ments, water with a high concentration of ammonianitrogen (25~100 mg/L) was irrigated as initial fertilizerbefore transplantation under various percolation rates(5~160 mm/day), and uniform saturated steady-state flowconditions were generated. This study calculates theeffects of soil thickness, ammonia nitrogen concentrationsin irrigated water, and percolation rates on changes inammonia nitrogen concentrations in percolating water inpaddy fields. In particular, a new set of notations andequations was derived for predicting ammonia nitrogenconcentrations in percolating water. This study alsoestimates water quality using the equation and discussesa percolation rate appropriate for conserving waterquality.

Moreover, we consider the perspectives of adsorptionand movement of ammonia nitrogen in cultivated soillayers. From the new notation and equation for predictingconcentrations of ammonia nitrogen in percolated water,the relation between percolation rates and the amount ofammonia nitrogen integrated into soil water was inves-tigated. Then, appropriate percolation rates from theperspectives of aqua-environmental conservation andnitrogen demand of paddy rice plants were investigated.Finally, the differences between ammonia nitrogenadsorption volume derived from the measured CEC valueand the amount of ammonia nitrogen integrated into soilwater required for percolation volume to equal irrigatedammonia nitrogen water concentrations were clarified.

Materials and methods

A column experiment was conducted in a thermostatic chamber(20 �C, humidity 50%), using plow layer soil taken from a paddyfield (1.7 ha) in Kakurai, Sakura, Chiba, which uses water fromLake Inba (Ishikawa et al. 1992a, 1992b). The soil (Table 1) wasstirred and sieved through a 4.75-mm mesh. Prominent roots,stones, and other debris were removed. The soil was then pouredinto 50-mm transparent acrylic columns of 0.76 g/cm3 wet density,40% moisture content, and 80% porosity; all identical to the soil

conditions of the actual rice field (Table 1). The columns werecovered with aluminum foil to block the light. Soil thicknesseswere set at 1, 2, 5, and 10 cm. Figure 1 shows the experimentaldevice. Biocolumn filters of 50 GI were wrapped in gauze andinstalled between the rubber stopper and soil samples of eachcolumn to prevent soil particles draining away from the system. AMariotte bottle was installed to fix water depth. The upper ends ofthe columns were covered with Parafilm to prevent evaporation.Ammonium sulfate was dissolved in pure water to make 25, 50, and100 mg/L solutions to be irrigated to each column. Percolation rateswere adjusted with an instillator and a three-way valve to generateconstant flows of 0, 5, 10, 20, 40, 80, and 160 mm/day. An irrigatedtotal nitrogen concentration of 0 mg/L and a percolation rate of10 mm/day were used as the control. Although 84 combinations canbe created from 4 soil thicknesses, 3 nitrogen concentrations, and 7percolation rates, only 23 were tested, mainly to compare andcreate standards (Table 2). Before proceeding with the experiment,air bubbles in the columns were removed by washing and saturatingthe soil with deaerated pure water for 24 h.

Water was analyzed for NH4+-N and NO3

�-N content. Perco-lated water was collected in 0.4-mL centrifugal filter tubes, from avenous needle installed at the outlet of each column. NH4

+-N wasanalyzed immediately after water collection to prevent oxidization.A high performance liquid chromatograph [(C) HITACHI, Ltd.]was used in the water analysis. CEC measurement for the above-mentioned cultivated paddy field soil was performed on controlsamples using the Schollenberger method (Schollenberger andSimon 1945).

Table 1 Basic physical characteristics of soil samples

Depth (cm) 0–20Specific gravity of soil particles 2.61Solid phase (%) 20Liquid phase (%) 22Gas phase (%) 58Water content (%) 40Porosity (%) 80Dry density (g/cm3) 0.54Wet density (g/cm3) 0.76Degree of saturation (%) 28Saturated hydraulic conductivity (cm/s) 4.0�10�4

Fig. 1 Experimental device

28

Results and discussion

Figure 2 shows changes in ammonia nitrogen concentra-tions in water percolated from a 1-cm-thick soil sampleagainst pore volume when water with 50 mg/L ofammonia nitrogen was irrigated under percolation ratesof 5, 10, 20, 40, 80, and 160 mm/day. Pore volume is theintegrated volume of percolated water divided by the totalvolume of pore space (Nielsen and Bigger 1961; Bear1969; Boast 1973; Ishiguro and Iwata 1988). Therefore, inthe case of perfect exchange, one pore volume is the totalwater contained in a soil sample, i.e., one pore volume ofsolution saturates the entire soil sample, from the surfaceto the bottom. Here, pore volume is expressed as

Pv ¼ Vw

Vp

VP¼nV

VW ¼ qt

where pore volume is represented by Pv, total volume ofpercolated water Vw, total volume of pore space Vp,

porosity n (%), total volume of soil sample V, percolationflux q (mm/day) and time t (day).

As a result, a clear distinction was observed amongpercolation rates. When pore volume or time wasconstant, a higher percolation rate increased ammonianitrogen concentrations rapidly in percolated water,indicating that percolation rates affect not only concen-trations in percolating water, but also the capacity of thesoil to adsorb ammonia nitrogen. A lower percolation rateresults in longer contact time and thus enhances theammonia nitrogen adsorption ability of the soil. Figure 3shows changes in ammonia nitrogen concentration ratiosafter pore volumes were divided by and corrected with thepercolation rate. After this correction, all data fell on asingle curve. This curve, which is an exponential equationfor predicting ammonia nitrogen concentrations in per-colated water, was approximated as:

Cr ¼ 1� exp �aPv0ð Þa ¼ 4:2� 10�4

Pv0 ¼ Pvq

qf¼ Pvqr;Cr ¼ Cout

Cin

ð1Þ

Fig. 3 Correction for pore volume with percolation rate (irrigatedwater: 50 mg/L ammonia nitrogen, soil thickness: 1 cm)

Fig. 2 Changes in ammonia nitrogen concentrations in waterpercolated under different percolation rates (irrigated water:50 mg/L ammonia nitrogen, soil thickness: 1 cm)

Table 2 Column experiments under various conditions

Percolation rates 0 l

(mm/day) 5 z z

10 z z

20 l z

40 l l l z

80 l l l l l

160 l l l l l l l

Soil thickness(cm)

1 l l l l l l l l l l z z l

2 l l

5 l

10 z z z z l l l

Irrigated waterconcentration(NH4

+-N, mg/L)

0 l l

25 l l l

50 l l l z z l l l l z z z z l l

100 l l l

z: discontinue experiment in progressl: complete experiment

29

where Pv' is the pore volume corrected with percolationrate, qf (mm/day) is reference percolation flux (in thiscase, qf=1 mm/day), qr (mm/day) is percolation rate, Cr isthe concentration ratio, Cin (mg/L) is ammonia nitrogenconcentration in irrigated water, and Cout (mg/L) isammonia nitrogen concentration in percolated water. Thevalue for ‘a’ was determined by the least squares method.

Here, soil thickness is represented by L (cm). Since theamount of water percolating during a given time t (days)is qt/10 (cm), pore volumes corrected with percolationrate Pv' at t are:

Pv0 ¼ 10q2t

Lnð2Þ

From Eq. (1), the relation between time and ammonianitrogen concentrations in percolated water may beexpressed as:

Cout ¼ Cin 1� exp�10aq2

Lnt

� �� �ð3Þ

by substituting Eq. (2) in Eq. (1) (Fig. 4). As Fig. 4b, cillustrates, close agreement between observed and pre-dicted values was obtained by using Eq. (3). A significantdifference was observed in a 10-cm-thick sample, but itdisappeared after 10 days (Fig. 4a).

In “Standards for Nitrogen and Phosphorus Dischargedinto Lakes and Ponds” (Japanese Water Pollution ControlLaw), the standard total nitrogen concentration is 1 mg/Lor less for irrigated water, environmental conservation,etc. (Environment Agency Government of Japan 1995).The time required for ammonia nitrogen concentrations inpercolated water to reach the 1 mg/L level was calculatedfrom Eq. (3) for various percolation rates. Table 3illustrates the relations between percolation rates and timerequired for ammonia nitrogen concentration in percolat-ing water from 20-cm-thick plowed soil to reach 1 mg/L,for different irrigation concentrations. A lower percola-tion rate results in less ammonia nitrogen penetrating intothe soil, a longer time for passing through the soil, andthus larger adsorption to the soil. Therefore, a lowerpercolation rate requires a longer time for the ammonianitrogen concentration in percolating water to reach 1 mg/L. If 7 days, i.e., the number of days usually used forinitial fertilization in Japan, are required for the ammonianitrogen concentration to reach 1 mg/L in water perco-lating from 20-cm-thick plowed soil, percolation rates are47, 33, and 23 mm/day, respectively, when 25, 50, and100 mg/L of ammonia nitrogen concentrations areirrigated. Ammonia nitrogen concentrations in the surfacewater of most rice fields in Japan do not remain at 100 mg/L for 7 days, i.e., throughout the initial fertilization period(Ishikawa et al. 1991). According to the exponentialequations, even percolation rates of 30~40 mm/day maypurify water to a level that does not adversely affect theaqua environment in this soil (Table 3). A percolation rateof 30 mm/day or lower may sufficiently eliminatefertilizer nutrients from water by percolation during ricecultivation in this soil (Table 3).

An attempt was made to calculate the amount ofammonia nitrogen that remained entirely in the soilsolution (hereinafter, we shall express this as amount ofammonia nitrogen in the soil water) using the relationshipequation derived from the exponential Eq. (1). Given T(t)as the total amount of ammonia nitrogen in the soil water(mg) to day ‘t’, total irrigated ammonia nitrogen load(mg) as I(t), total percolated ammonia nitrogen load (mg)as S(t), if we express T(t) as:

T tð Þ¼ I tð Þ�S tð ÞWe can use the following equation to express the

relationship between time and total amount of ammonianitrogen in the soil water; by converting with total

Fig. 4a–c Time changes in ammonia nitrogen concentrationsplotted under various soil thicknesses, irrigated water and perco-lated rates. (a percolation rate: 160 mm/day; b percolation rate:80 mm/day; b, c soil thickness: 1 cm; a, c irrigated water: 50 mg/Lof ammonia nitrogen)

30

nitrogen volume per 100 g of dry soil, T(t) can beexpressed as follows:

T tð Þ ¼ nCin

10aqr1� exp

�aq2t

10Ln

� �� �ð4Þ

where r is dry density (g/cm3). Here, q = constant (mm/day) and Cin = constant (mg/L). Based on Eq. (1), therelationship equation that shows relations between timeand the total ammonia nitrogen volume in the soil waterwas introduced.

Here, we will consider T(t) when tfi+¥. From Eq. (4),we can obtain

limt!þ1

T tð Þ ¼ nCin

10aqrð5Þ

Note that the unit here is mg/100 g of dry soil.According to Eq. (5), total ammonia nitrogen volume inthe soil water can change with irrigated ammonia nitrogenwater concentrations and percolation rate. In other words,when the percolation rate is constant, total ammonianitrogen volume in the soil water will simply keepincreasing as long as irrigated ammonia nitrogen concen-trations are increasing, and decrease when irrigatedammonia nitrogen concentrations decrease. However,there is a limit to the ammonia nitrogen adsorptioncapacity of the soil, which has been determined by theCEC value. Generally, the CEC value is the result ofsufficient contact between liquid and soil particles, and isderived when all reaction groups manifest this function. Itis also the potential ion conversion volume. Form thisCEC value, we can determine the range of application forEq. (5).

The CEC value is represented as E (mol/kg of drysoil). Since 1 mol=14,000 mg at ammonia nitrogen, E isable to adsorb 1,400�E (mg/100 g of dry soil) of ammonianitrogen. Therefore, Eq. (6) can be derived as follows:

limt!þ1

T tð Þ ¼ nCin

10aqr� 1; 400E ð6Þ

Expressing Eq. (6) in terms of q, the range ofapplications can be given by the following equations:

1ð Þ0 < q � a�1Cinr�114; 000�1E�1n Cin : constant:ð Þ

limt!þ1

T tð Þ ¼ nCin

10aqrð7Þ

2ð Þq � a�1Cinr�114; 000�1E�1n Cin : constant:ð Þlim

t!þ1T tð Þ ¼ 1; 400E

ð8Þ

Subtracting the percolated ammonia nitrogen loadfrom the irrigated ammonia nitrogen load provides theamount of ammonia nitrogen in the soil water that wascalculated for various times until the two loads becameequal. The resulting values were called the ammonianitrogen adsorption limits. Since the upper ends of theexperimental columns were covered with Parafilm, nitro-gen fixation and ammonia volatility were roughlyneglected. Nitrate nitrogen could not be detected inpercolation water over one pore volume. Therefore,denitrification activity was also omitted. The CEC valueof this soil, as a control, was measured to be 0.262 mol/kgof dry soil. Converting this into ammonia nitrogenadsorption gives a value of 367 mg/100 g of dry soil.The percolation rates for each irrigated ammonia nitrogenwater concentration and the ammonia nitrogen adsorptionlimits were plotted as shown in Fig. 5. Since the unit forthe ammonia nitrogen adsorption limit was (mg/100 g ofdry soil), the soil thickness was of no consequence.Equations (7) and (8) were used here. Since the potentialammonia nitrogen adsorption volume derived from themeasured CEC value was 367 (mg/100 g of dry soil), thiscondition was also taken into consideration. A porosity of80% and dry density of 0.54 g/cm3 were also substituted.As a result, the relation between percolation rate q (mm/day) and ammonia nitrogen adsorption limit T(q) (mg/100 g of dry soil) for each irrigated ammonia nitrogen

Table 3 Percolation rates satis-fying water quality standardlevel (plowed soil thickness:20 cm)

Percolation rates (mm/day) Time necessary for Cout=1(mg/L) (day)

Cin=25 (mg/L) Cin=50 (mg/L) Cin=100 (mg/L)

2 3,887 1,924 9575 622 307 153

10 155 76 3820 38 19 930 17 8 440 9 4 280 2 1 0.6

160 0.6 0.3 0.1

Fig. 5 Percolation rate and ammonia nitrogen limit when NH4+-N

concentration of percolation water was equal to that of irrigationwater (irrigated water concentrations: 25, 50 and 100 mg/L NH4

+-N)

31

water concentration was as follows: at an irrigatedammonia nitrogen water concentration of 25 mg/L,T(q)=8,818/q (24fq), T(q)=367 (0fq<24); at 50 mg/L,T(q)=17,637/q (48fq), T(q)=367 (0<qf48); and at 100 mg/L, T(q)=35,273/q (96fq), T(q)=367 (0<qf96) (Fig. 5).

As mentioned earlier, the appropriate percolation ratesfor maintaining good water quality during the initial 7-day fertilization period were 47 mm/day at an irrigatedammonia nitrogen water concentration of 25 mg/L;33 mm/day at 50 mg/L; and 23 mm/day at 100 mg/L.The appropriate percolation rates are those in which thepercolation water concentration of ammonia nitrogen is1 mg/L at a depth of 20 cm in cultivated soil on theseventh day after percolation. Using Eq. (4) to derive thetotal ammonia nitrogen volume in the soil water for the 7-day initial fertilization period for each irrigated ammonianitrogen water concentration gives us the followingvalues: 7.50 mg/100 g of dry soil at an irrigated ammonianitrogen water concentration of 25 mg/L; 10.7 mg/100 gof dry soil at 50 mg/L; and 15.3 mg/100 g of dry soil at100 mg/L. Since the cultivated depth was 20 cm,converting the total ammonia nitrogen volume in the soilwater to kg/ha gives us the following values; 81.1 at anirrigated ammonia nitrogen water concentration of 25 mg/L; 115.4 at 50 mg/L; and 165.6 at 100 mg/L. It hasgenerally been stated that the most suitable nitrogencontent for Japanese Koshihikari rice during young earformation is 50 kg/ha (Miyama 1990), so our values areslightly above the nitrogen demand of paddy rice plants.Therefore, these values are the upper limit of percolationrates that can still maintain water quality conservation,making it necessary to obtain better percolation rates forpaddy rice growth. With this in mind, we used Eq. (4) totry to calculate percolation rates that can maintain theoptimum nitrogen content for Koshihikari rice (50 kg/ha)throughout the 7-day initial fertilization period. As aresult, we obtained the following percolation rate values:28.8 mm/day at an irrigated ammonia nitrogen waterconcentration of 25 mg/L; 14.3 mm/day at 50 mg/L; and7.15 mm/day at 100 mg/L. These values were 39~69%below the optimum percolation rate for maintaining waterquality conservation. Furthermore, using Eq. (3) at thistime to calculate the percolation ammonia nitrogen waterconcentrations at –20 cm in cultivated soil on the seventhday, provided the following estimated values: 0.38 mg/Lat an irrigated ammonia nitrogen water concentration of25 mg/L; 0.18 mg/L at 50 mg/L; and 0.093 mg/L at100 mg/L. These values are 62~91% lower than the 1 mg/L water quality standard for lakes and ponds in Japan, andshould therefore help to maintain good water environ-ments.

In addition, the 50 kg/ha nitrogen volume is 1.26%ofthe 3,963.6 kg/ha of the potential ammonia nitrogenadsorption volume derived from the CEC value of thesample soil. Therefore, under these percolation rates andirrigated water concentrations of ammonia nitrogen, theammonia nitrogen input during the 7-day period is almostall adsorbed and maintained in the soil, and can be usedincrementally during the initial growth of rice.

Finally, the experimental measurements showed that,until the irrigated water concentrations and percolationwater concentrations of ammonia nitrogen became equal,the ammonia nitrogen adsorption limit ranged from 58.1to 363 mg/100 g of dry soil (Fig. 5) due to the differencesbetween irrigated ammonia nitrogen water concentrationsand percolation rates. The relationship equation gives afairly good agreement with the experimental results(Fig. 5). Therefore, the relationship equation derivedfrom Eq. (1) may describe relations between time and thetotal ammonia nitrogen volume in the solid and liquidphase. Furthermore, a comparison of these values withpotential ammonia nitrogen adsorption of the soil derivedfrom the CEC value shows values for the ammonianitrogen adsorption limit (mg/100 g of dry soil) within arange of 16 to 99%. The reason for this is believed to bedifferences in contact area and contact time with the soilparticles, which are caused, for example, by deviations inflow rate distribution of soil water and the occurrence offlow before dispersion. Even when the amount ofpercolation of the disturbed soil was kept constant,deviations in the flow rate distribution of water in thesoil occurred. Therefore, it is believed that percolatingwater did not come into contact with all the clay-soilparticles involved in adsorption.

Conclusion

Assuming increasing surface water concentrations inpaddy fields during early irrigation in the initial fertili-zation period, laboratory experiments were conducted onsoil layer percolation using columns. As a result, a higherpercolation rate resulted in a faster rise in ammonianitrogen concentrations in percolated water, even for aconstant pore volume. Therefore, a lower percolation ratemay possibly extend soil contact time, thus enhancing thesoil’s ability to adsorb ammonia nitrogen. A newcorrection method (correcting pore volume with percola-tion rate) was also introduced. All data for the relationsbetween concentration ratio and pore volume correctedwith percolation rate fell on a single curve. This curvewas used as an exponential equation for predictingammonia nitrogen concentrations in percolating water.This exponential equation could determine the relationbetween time and ammonia nitrogen concentration inpercolating water and also the number of days necessaryfor ammonia nitrogen concentrations in percolating waterto reach the 1 mg/L level. This calculation showed alower percolation rate requiring a longer time. Forexample, when 50 mg/L is irrigated to paddy fields, thepercolation rate of the soil should be 30~40 mm/day sothat the ammonia nitrogen concentration at a depth of20 cm in plowed soil is 1 mg/L in 7 days, i.e., the averageduration of initial fertilization in Japan, indicating that apercolation rate of 30 mm/day or lower can sufficientlyeliminate fertilizer nutrients from water during ricecultivation in the soil.

32

Moreover, the ammonia nitrogen volume in the soilwater (mg/100 g of dry soil) calculated from theconcentration of ammonia nitrogen in percolation showedvarious values depending on the differences betweenirrigated ammonia nitrogen water concentrations andpercolation rates. Then, using the exponential equation topredict the percolation water concentrations of ammonianitrogen, a new relationship equation was derived toobtain the ammonia nitrogen volume in the soil water.Using this relationship equation, we were able to explainchanges in integrated ammonia nitrogen volume in thesoil water using differences in irrigated ammonia nitrogenwater concentrations and percolation rates. Furthermore,this equation was also used to calculate the optimumpercolation rate that can satisfy the nitrogen demand ofrice plants during the 7-day initial fertilization period. Forexample, the percolation rate (7.15~28.8 mm/day) thatcould maintain 50 kg/ha of nitrogen in cultivated soilduring this period decreased as irrigated ammonia nitro-gen water concentrations (25~100 mg/L) increased.Furthermore, when irrigated ammonia nitrogen concen-trations were 25~100 mg/L, the percolation ammonianitrogen concentrations on the seventh day at a depth of20 cm in the cultivated soil decreased as irrigatedammonia nitrogen water concentrations increased, withestimated values ranging from 0.093~0.38 mg/L. There-fore, for this type of paddy soil, we were able to derive apercolation rate that would have no effect on rice growthwhile simultaneously helping to maintain the good qualityof the water environment.

Finally, the potential ammonia nitrogen adsorptionvolume derived from the measured CEC value differedfrom the integrated ammonia nitrogen volume in the soilwater until irrigated water concentrations and concentra-tions of percolated water of ammonia nitrogen becameequal. A comparison with the potential ammonia nitrogenadsorption volume derived from the CEC value shows arange of 16~99%. The reason for this is believed to bedifferences in contact area and contact time with theparticles, which are caused, for example, by deviations inthe flow rate distribution of soil water and the fact thatflow occurs earlier than dispersion.

Acknowledgments Thanks are due to Prof. Dr. Sato Yohei,Graduate School of Agricultural & Life Sciences, The University ofTokyo, for valuable advice. Special thanks are extended to Prof. Dr.Bill A. Stout, Texas A&M University, Department of AgriculturalEngineering; Dr. Ishiguro Munehide, Okayama University, Depart-ment of Environmental Management Engineering; and Dr. V.Anbumozhi, Graduate School of Frontier Sciences, The Universityof Tokyo, for helpful comments on an earlier draft of this paper.

References

Bear J (1969) Flow through porous media. In: De Wiest JM (ed)Academic Press, New York, pp 109–199

Boast CW (1973) Modeling the movement of chemicals in soils bywater. Soil Sci 115(3):224–230

Bolt GH, Bruggenwert MGM(1979) Soil chemistry A. Elsevier,Amsterdam, pp 54–72

Environment Agency Government of Japan (1995) Water PollutionControl Law. Standards for Nitrogen and Phosphorus Dis-charged into Lakes and Ponds

Ishiguro M, Iwata S (1988) Mass transfer into soil (4). Adsorptionof ion exchanges into soil. J JSIDRE 56(10):1017–1024

Ishikawa M, Okada K, Tabuchi T, Yamaji E (1991) Changes innitrogen concentrations in ponded water after the dissolution offertilizer, Proc Conference on Kanto JSIDRE, pp 57–58

Ishikawa M, Tabuchi T, Yamaji E, Nakajima J (1992a) Field test ofwater quality purification using underdrains—study of waterquality purification by soil layer in paddy fields (1). TransJSIDRE 159:81–89

Ishikawa M, Tabuchi T, Yamaji E, Nakajima J (1992b) Influence ofconcentration of nitrogen and percolation on water qualitypurification and rice growth—study of water quality purifica-tion by soil layer in paddy fields (2). Trans JSIDRE 159:91–99

Ishikawa M, Tabuchi T (1998) Practical approach to water qualityimprovement in agricultural areas, Agric Mech Asia Afr LatinAm 29(4):43–52, 55

Kondo T, Misawa S, Toyoda M (1992) Characteristics of effluentload of nutrient salts (N, P) from paddy fields located in thealluvial and lower areas in Hokuriku. Trans JSIDRE 159:17–27

Miyama M (1990) The applied fertilizer with the suitable amountof nitrogenous possession of the paddy rice and the soilnitrogenous absorption pattern. The applied fertilizer andnitrogenous inorganic substance of the paddy field soil.Hakuyu-sha, Tokyo, Japan, pp 63–97

Nielsen DR, Bigger JW (1961) Miscible displacement III. Theo-retical condition. Soil Sci Soc Am, Proc, 26, pp 216–221

Okajima H (1989) The function and structure of the soil.Noubunkyo Press, Tokyo, Japan, pp 64–66

Schollenberger CJ, Simon RH (1945) Determination of exchangecapacity and exchangeable bases in soils. Ammonium acetatemethod. Soil Sci 59:13–24

Tabuchi T, Yamafuji I (1992) Effect of puddling on percolation rateand nitrogen concentration in percolating water. Soil PhysicalConditions Plant Growth Jpn 66:47–54

Takamura Y, Tabuchi T, Suzuki S Cho T, Ueno T, Kubota H(1976) Study on material balance in paddy fields (1).s Nitrogenand phosphorus balance in paddy fields at Kasumigaurawatershed. Jpn JSoil Sci Plant Nutr 47(9):398–405

Takamura Y, Tabuchi T, Cho T, Ohtsuki H, Suzuki S Kubota H(1977) Study on material balance in paddy fields (2). Nitrogenand phosphorus balance and drainage in ill-drained paddy fieldsat Tonegawa watershed. Jpn J Soil Sci Plant Nutr 48(9,10):431–436

Takamura Y, Tabuchi T, Cho T, Nishimura N, Ohtsuki H, KubotaH, Suzuki S, Ohsaki K (1979) Study on material balance inpaddy fields (3). Nitrogen and phosphorus balance and drainagein dry paddy fields at Kasumigaura watershed. Jpn J Soil SciPlant Nutr 50(3):211–216

Takeda I, Kunimatsu T, Kobayashi S, Maruyama S (1991) Pollutantbalance of a paddy field area and its loading in the water. TransJSIDRE 153:63–72

33