NITRATE LEACHING UNDER INTENSIVE FARMING AND SUCH INFLUENCING FACTORS AS TYPE OF FERTILIZER AND SOIL

Transcription

1 Title Author(s) Citation NITRATE LEACHING UNDER INTENSIVE FARMING AND SUCH INFLUENCING FACTORS AS TYPE OF FERTILIZER AND SOIL Maeda, Morihiro Annual Report of FY 21, The Core University Program between Japan Soiety for the Promotion of Siene(JSPS) and National Centre for Natural Siene and Tehnology(NCST). P.18-P.25 Issue Date 23 Text Version publisher URL DOI rights

2 NITRATE LEACHING UNDER INTENSIVE FARMING AND SUCH INFLUENCING FACTORS AS TYPE OF FERTILIZER AND SOIL M. Maeda Deparll1lent of Soils and Ferlilizers, i'ia/ional Agriullural Researh Cenler, Kannondai, nukuba, Jbaraki , Japan ABSTRACT Nitrate (NO)) leahing from agriultural area is related to ground and surfae water ontamination. NO) leahing is greatly affeted by types of fertilizer. Exessive N for rops derived from two hemial fertilizers (ammonium nitrogen and polyolefin-oated urea) aused substantial N3 leahing, whih was predited by an OECD model that alulates the risk of groundwater ontamination. Organi N from ompost appeared to aumulate in topsoil over several years, but NO) leahing reahed the same level as that in the hemially fertilized in the following years. Sine preferential flow in soil also influenes N 3 leahing, soil types should be evaluated with leahing experiments and models. Preferential flow in lay monoliths was well desribed by a modified tanks-in-series model through the parameters n and fm' that is, in terms of the degree of mixing of water in soil and the ratio between mobile regions and the maximum amount of water in a monolith, when freely drained. Further studies are neessary to establish the best management of agro-eosystems to maintain agriultural produtivity and the environmental quality. KEYWORDS Coated urea; ompost; leahing; monolith lysimeter; nitrate; preferential flow fntroduction NO) leahing from arable land, whih auses ontamination of groundwater, has beome a matter of worldwide onern. In Japan, N 3 -N onentrations in 6.3% of the groundwater investigated in 1998 exeeded 1 mg Lot, a Japanese standard for groundwater and agriultural land was reported to be a major soure (Environment Ageny, 1999). Nitrogen is intensively applied to Japanese agriultural land as a major essential nutrient. Nitrogen forms in fertilizer are ammonium (NH4), NO), urea, et. Compost made by livestok waste or sewage sludge is also regarded as N fertilizer and its adequate appliation to arable land is reommended for reyling valuable nutrient resoures. NH4 is a produt of mineralization of organi N. NH4 derived from inorgani and/or organi fertilizer is easily oxidized into NO} by mirobes in upland fields. NO}, whih is an anion, is not absorbed on to soil and is freely leahed out of root zones, beause soil has usually negative harge. The limate in most of Japan belongs to the ategory of the Asian monsoon, in whih averaged annual rainfall reahes 18 mm (Kuwabara, 1989) and heavy rainfall events often our with typhoons in summer This relatively high preipitation may inrease NO} leahing from agriultural area. It is neessary to establish the best agriultural management to maintain both high yields/quality of rops and the environmental quality. The effets of different N fertilizers and soil types on NO} leahing were disussed in the present paper. -18-

3 NITRATE LEACHING FROM DIFFERENT T\'PES OF FERTILIZERS Slow-release fertilizers, suh as polyolefin-oated urea, have been used to save labor from frequent appliation of hemial fertilizer and to inrease N reovery rates for rops in Japan (Saigusa et ai., 1993). Substitution of slow-release fertilizer for readily available onventional fertilizer suppressed N leahing in short-term studies (Sakata et al., 1995; Matsumaru, 1997). Both of them noted that N derived from slow-release fertilizer remained in the topsoil at the end of the experiments, but its later leahing was not examined. Although NO}-N onentration in subsurfae flow under pastures treated with readily available fertilizer inreased more rapidly than that with slow-release fertilizer, N 3 -N onentrations from both types of fertilizer treatment ontinued to inrease even after 1 years (Owens et ai., 1992). These results indiate the importane oflong-term researh. Exessive use of livestok manure is also related to groundwater ontamination (Thomsen et al., 1993; Adams et al., 1994; Chang and Entz, 1996). NO} leahing is affeted by manure types with different amounts of available N (Bekwith et ai, 1998). Compost, whih has a great advantage in handling, reduing offensive odors, and eliminating pathogens, is a major produt of livestok waste in Japan. Beause most livestok farmers do not have suffiient farm land for reyling waste, exessive waste must be transported to other farmers (Harada and Yamaguhi, 1998). Aordingly, livestok ompost has been applied not only to pastures but also to vegetable fields, where fertilization is usually more intensive. Residual N from previously applied manure is mineralized and may be available for leahing in ontinuously managed fields (Angle et al., 1993; Thomsen et al., 1993; Bergstbm and Kirhmann, 1999). NO} leahing was studied in an Andisol treated with 4 N fertilizers (AN: ammonium N at 4 kg N ha-1y"1, CD: polyolefin-oated urea at 4 kg N ha- 1 /, Se: swine ompost at a rate of 8 kg N ha-1y"1, or NF: no fertilizer) for 6 years (Maeda et al., 2Ia). The appliation rate of SC was determined on the assumption that 5% ofn in SC is available for rops (Harada, 1997). Andisols (volani ash soils) over more than half of the upland fields in Japan. Sweet orn was grown in summer, followed by Chinese abbage or abbage in autumn eah year. NO:l-N onentrations in soil water at I-m depth rose markedly in the summer of the seond year and then flutuated between 3 to 6 mg L- 1 in the CD and AN plots (Fig. 1). N 3 -N onentration in the SC plot began to inrease in the fourth year, reahing the same level as that in the CD and AN plots in the beginning of the sixth year. In the 1'.'F plot, NO:l-N onentration was about 1 mg L- t for the first 4 years and dereased to 5 mg L- t toward the end to the experiment. 'I 1"\,.J \.. J E.s J I 1 -i==='======i===i:===i=====:-15 :E.-: 8 -.s '" 6 " t5 4 g 'i' 5 z AN 4/1 4/1 4/1 4/ / { / ILOOO Sampling date Fig. 1. NOrN onentration at I-m depth under 4 different N treatments and umulative preipitation during the experimental periods: AN. CU. Se. and NF represent ammonium nitrogen. oated urea. swine ompost. and no fertilizer. respetively. The highest NO}-N onentrations were observed in the spring of 1998 and the lowest in the winter of 19-

4 in the AN and CU plots during the experimental period after the summer of the seond year ( ) when the influene of fertilization learly appeared. The highest preipitation, reorded in , may have aused the lower N3-N onentration in the following year of 1999/2, beause more humid onditions ontribute to water dilution ofn 3 -N and denitrifiation. To evaluate the N3-N onentration in soil water at I-m depth, a simple equation suggested by OECD was used (OEeD, 1999). OECD authorized the following method alulating the risk of water ontamination by nitrogen, based on N and water balane on the soil surfae. This model is used for omparing the risk among ountries or loal areas in a ountry (OECD, 1999). PNC =PNP EW where PNC (mg L- I ) is the potential N 3 -N onentration, PNP (mg ha-i) the potential N3 present, and EW (L ha-i) exess water. N surplus in topsoil an be used as PNP at a preliminary step (OECD, 1999). PNP was obtained by subtrating the mean N uptake from the mean N appliation in the present study. As a result, PNCs were 38, 44, and 19 mg L- I for the CU, AN, and SC plots, respetively. The PNCs satisfatorily predited N3-N onentration in the AN and CU plots (Fig. 1). In ontrast, the alulated PNC in the SC plot was substantially overestimated, presumably beause little organi N from SC was available for N 3 leahing during the experimental period. In other words, exessive N for rops from AN and CU would ause N3 leahing in due time. On the other hand, organi N from SC appeared to aumulate in topsoil during the first several years; yet the aumulation of the organi N may ause late N mineralization in soil, resulting in an inrease in N3-N onentration after the fourth year of the experiments and later (Fig. 1). To examine N3 fate in deeper soil near groundwater, N 3 distribution in deeper soil was examined in the AN plot after 6-year ontinuous fertilizer appliation (Maeda et at. 21b). N 3 -N onentration in soil water delined with depth (Table 1). N 3 -N onentrations in deeper than 3 m were almost equal to that in groundwater (Table 1). Groundwater table flutuated between 12 m and 4 m during the experimental period. On the sampling date, the groundwater table existed around 3.6 m below the ground. The groundwater table rose to within 2-m depth a few times in summers with heavy rainfall Probably, rising groundwater washed away N 3 in the soil solution and resulted in derease in N 3 onentration in the deeper layers. Table 1. NO)-N and Cl onentrations, mole ratios of NOrN to Cl, and S 15 N values in soil water sampled from different soil depths and groundwater in the ammonium nitrogen (AN) plot NO,-N Cl N 3 -N/CI i51 m (mgl I) (mg L"I) (1111 I 1111) (%) loo R ± R ± ± ± ± ± ± ±.6 _O.5 t 3 :;g ::!: ±.8 OA );.9' 35 :u ± ± ll.7 OA ± 1.5 OA ± 1.5 OA 9.1 ± O?-rolludwater 3.9 ± ± ±.6 All water samples were obtained on 18 April 2. N=2-5. t: No repliate beause of insuffiient amount of water. Cl onentrations dereased with depth as well as N 3-N but the rate of the deline for N 3 -N was larger than that for Cl (Table 1). The (ilsn for soil water deeper than 3 meters also showed higher values than that at I-m depth (Table 1). (ilsn value (%) is expressed as the following equation (1) (2) where R is the ratio of lsn;i4n and the standard is atmospheri N 2. (ilsn value of the substrate beomes 2-

5 higher through mirobial reation beause the heavy isotope ofn, 15N, beomes relatively enrihed in the residual fration of the substrate. Namely, olsn inreases through reations suh as denitrifiation (Yoneyama, 1987). Botther et al. (199) found dereasing N 3 onentrations with groundwater depth orrespond to the inrease in olsn values, whih reveals the signifiane of denitrifiation in the deease of N 3 onentration. Hene, anaerobi onditions resulting from rising groundwater may have aused denitrifiation in the present study. Botther et al. (199) attempted to investigate the ontribution of denitrifiation to the derease in N 3 -N onentration by use of the following Rayleigh equation (Mariottti et ai., 1981): where onini is the initial value of olsn for the substrate (%), 8-an enrihment fator (%), and!no3,n the fration of unreated residual substrate, whih is obtained to divide substrate onentration by initial substrate onentration. The enrihment fator & in groundwater was -16% (Botther et al., 199), although this fator in the previous studies ranged from -7.8 to -3% (Arai and Tase, 1992). Assuming that & is -16% and N3-N at 2-m depth (47 mg L'l, 8J5Nini = -.7%) moved to that at 3-m depth (8 J5 N= -8.9%), N 3 -N onentration at 3-m depth will be expeted to be 26 mg L'l, whih is by far greater than the measured value of 3.8 mg L'l, indiating the ontribution of denitrifiation was probably small in this study. (3) SOIL SUTRUCTURE AND NITRATE LEACHING N 3 leahing is affeted by preferential water flow in soil. Preferential flow refers to maropore flow, fingering, and funneled flow (Steenhuis et al. 1995). Maropores, whih reflet soil struture, root deay, worrriholes, et., onstitute preferred flow pathways for infiltrating water in soil. Fingering ours as a result of wetting front instability suh as hange in hydrauli ondutivity with depth and ompression of air ahead of the wetting front. Funneled flow takes plae when sloping geologial layers ause pore water to flow laterally, aumulating at a low region. To evaluate N 3 leahing in soils where preferential flow may our, undisturbed soil should be used and exposed to natural field onditions. A new drilling method for olletion of soil monoliths up to 1 m depth was developed in Sweden (Persson and Bergstrom, 1991) and modified by my group in Japan to attah it on any trators (Photo 1). The drill onsists of a steel ylinder with four mounted utting teeth at the bottom, into whih a PVC tube (286 mm i.d.) is inserted. The drill was attahed on the rear of a trator. The hydrauli power of the trator was used for pushing, lifting, and rotating the drill. Photo 1. A devie, whih uses hydrauli system of a trator, for olleting soil monoliths. Leahing rates of N were measured in 5-mm deep monolith lysimeters ontaining a heavy lay soil (Maeda and BergstOm, 2). Nitrogen was applied at a rate of 1 kg ha'l (5 kg N3-N ha'! and 5 kg NH4-N ha'i). Bromide was applied at a rate of 1 kg ha'l to provide information on water movement -21-

6 through the profiles. Three preipitation regimes were used in tripliate: (i) N-C and N-A, whih were exposed to natural preipitation during the period 22 July 1998 to 25 January 1999; (ii) D-A, whih was exposed to 'double' natural preipitation from 22 July to 9 November 1998, and then natural preipitation only until 25 January 1999; and (iii) T -A, whih was exposed to 'triple' natural preipitation from 22 July to 9 November 1998, and then natural preipitation until 25 January At the end of Otober for T-A and in the middle of January for D-A, the umulative amounts of leahate reahed a level, whih was equivalent to the amount of water (218 mm) that was present in eah monolith after being freely drained. The N-C and N-A monoliths did not reah this water saturation level during the period (i.e., it was only 68 % of the field apaity moisture ontent at the end of the experiment). If the water entering the soil moved downwards as matrix flow, the Br peak onentration would appear when the aumulated leahate reahed an amount of water orresponding to the water ontent at field apaity in a monolith. However, in all treatments, the Br peaks eluted before the aumulated leahate reahed this level, whih is a lear indiation of preferential flow (Fig. 2). Nitrogen leahing ourred in the form ofn 3 -N, whih was also largely displaed through preferential flow (Fig. 2). Conentrations ofn 3 -N in leahate in the treatments whih reeived the element often exeeded 1 mg L- 1 NH 4 1eahing aounted for only a few perent of the total input ofnh 4 -N. T-A O-A 5 1 i ')B' r- 4 4 '\:h...l Ol.s 3J r 3 J 'It 1.9 T / 11 \ C C) 2J 1 2 u () j/l\ \\ l I" J\4 ] 11'" A S N D J J A S N D Date 5 1 Date b) N 3 -N Fig. 2. Conentrations of a) Br and b) NOrN in leahate through lay soil monolith lysimeters exposed to natural preipitation (N-A. treated with hemials: : ontrol (N-C not treated: e). 'double' natural preipitation in summer (D-A, treated with hemials: ). or 'triple' natural preipitation in summer (T-A. treated with hemials: ). Arrows indiate dates when the umulative amounts of leahate reahed a level equivalent to the amount of water that was present in eah monolith after being freely drained. Although many studies have suggested the ontribution of preferential flow ourring in soil on rapid leahing of solutes by examining dyed flow paths, few quantifiation approahes have been disussed by use of a non-reative traer. In my study (Maeda and Bergstbm, 2), a modified version of the 'tank-in-series model' was used to desribe a distribution funtion of non-reative traer (Br) travel time and ompare the results with experimental data obtained in monolith lysimeters. The theory of the model is based on the tanks-in-series model (Fogler, 1992). Distribution of travel time (residene time) of a non-reative traer is determined. by injetion of the ompound into the feed stream entering the reator, and then the number (n (-» of equally sized ompletely mixed flow tanks in series is obtained as a parameter. The parameter n shows the degree of mixing of flow in the reator, and varies from one for ideal ompletely mixed flow, to infinity for ideal plug flow The following assumptions were made for soil systems: (i) when water ontent in soil exeeds the maximum amount of water held in a freely drained monolith, infiltration ours and its rate is equal in all tanks; (ii) soil has stagnant flow regions and their ratio to the maximum amount of water in a monolith, when freely drained, does not hange during the experiment; and (iii) rainfall does not ontain the studied traer. After some mathematial steps, the distribution funtion of travel time (E(r) was obtained as the following equation: 22-

7 ()(T)"-l " CV. t: E(r)=---"-= )111 m e.1m M f(n) ---,--r where en (mg L- 1 ) is the traer onentration in leahate,jm (-) is the ratio between water in mobile regions and the maximum water ontent in a freely drained monolith, M (mg m- 2 ) is the mass of the traer applied on a lysimeter surfae, V (mm) is the maximum water ontent in the freely drained monolith, r is alled a gamma funtion, and normalized time (T (-)) is defined as the umulative leahate divided by the field apaity water ontent. When determining parameter values, the required experimental period was made shorter by using a peak oordinate of distribution of non-reative traer travel time, instead of the moment analysis_ (4) la --- N-A(i) GJ 1.2 (f m=.b5. n=1.5).. Q) --- E.. N-A(ii) (f m=.65. n=l.l) Qj > x N-A(iH) O.B (f m=.75. n=1.2) ID.6 '.. :g OA x D Vi.2 i:5 'X.f r---:--- Freeze-thaw period la -- D-Abi) GJ ;;< (f m =1.. n=2. ) 1.2,>I<' Q) E --- D-A(ii) :;:; (f m =.75. n=1a) Qj > x r; D-A(iii Qj O_B (f m =_6. n=1.3).r;.6 '. 1 x, :; OA "' "- D." Vi.' : f!< ","" i:5 ;(-',--. - r---:--- Freeze-thaw period 1A ---T-A(i) GJ (f m=.6. n=1.5) Q) --- T-A(ii) : (f Qj m =.65. n=1.5 > x T-A(iH).r; Qj O.B (f m=.b5. n=1.5 u.r;.6 ' :. 1 :; OA D Vi.2 i:5 Fig_ 3, Measured and simulated distributions of traer travel time (E(r) vs. normalized time (r): (i), (ii), and (iii) represent the three respetive repliates. Monolith lysimeters were exposed to natural preipitation (N-A), 'double' natural preipitation in summer (D-A), or -triple' natural preipitation in slimmer (T-A). -23-

8 Preferential flow is aused mainly by two mehanisms. One is the existene of stagnant regions (dead spae). This mehanism an be expressed in terms of the parameter 1,,,. The other is due to uneven water-flow rates. Some regions with rapid flow rates, suh as maropores, and matrix flow regions will result in a large variability of flow rates. This ondition may be expressed in terms of the parameter 11. Experimental results under 3 different water regimes are plotted in Fig. 3, together with fitting urves for different values on the parameters /,,, and n used in the model. Generally, experimental data was well predited by the modified tanks-in-series model. The values off,,, ranged from.6 to 1 and those of n from 1.1 to 2.. These results suggest that the heavy lay soil used in this study might have -4% of stagnant regions in the profile. If the degree of mixing and stagnant regions were smaller without preferential flow, the peak of E(r) would appear at r = 1. The peaks of E(r) were, however, observed far earlier than r = 1 in all treatments (Fig. 3). In other words, the mixing onditions in the monoliths were pronouned. This high degree of mixing suggests that matrix flow, and different kinds of preferential flow, ourred simultaneousl y in this lay soil. During later periods of the experiment, measured values of traer travel time were lower than simulated values in all monoliths (Fig. 3). These periods oinide with freeze-thaw periods. The flow onditions during suh periods seemed to be quite different from those ourring during the summer months. Low onentrations of Br (low distributions of normalized traertravel time) were seen in the first disharge after a long period of snow aumulation, whih is similar to an observation by Bergstrom and Jarvis (1993). They explained this in terms of redistribution of the traer to smaller unfrozen pores on freezing and that, upon melting, perolating water would be pratially free from the traer. Overall, this approah an be a helpful tool when hemial leahing through soils with different preferential flow proess are ompared. Combining the model for a non-reative traer with hemial reation models (inluding sorption, transformation, and biodegradation), will then ontribute to give us a omplete piture of hemial leahing. CLOSING REMARKS Monitoring data on N3 leahing under omplex environmental onditions should be better understood for predition or management ofn3 leahing. The author introdued the approahes, for this purpose, based on simple mathematis. Although preise predition for N fate in the environment is required in some ases, the risk evaluation of agriultural management for NO, leahing will also provide enough information for politial deision or regulation. REFERENCES Adams P L., Daniel T. C, Edwards D. R., Nihols D. J., Pote D. H., and Sott H. D. (1994) Poultry litter and manure ontributions to nitrate leahing through the vadose zone. Soil Sei. So. Am. 1., 58, Angle J. S, Gross C M, Hill R. L. and MIntosh M. S. (1993) Soil nitrate onentrations under orn as affeted by tillage, manure, and fertilizer appliations. 1. E11viron. Qual., 22, Arai H. and Tase N (1992) Evaluation of self-purifiation in Tamagawa Jousui hannel by using 15N natural isotope. (In Japanese, with English abstrat.) Environ. Sei., 5, Bekwith CP., Cooper J., Smith K. A. and Shepherd M. A. (1998) Nitrate leahing loss following appliation of organi manures to sandy soils in arable ropping: 1. Effets of appliation time, manure type, overwinter rop over and nitrifiation inhibition. Soil Use Manage., 14, Bergstrom L. F. and Kirhmann H. (1999) Leahing of total nitrogen from nitrogen-15-labeled poultry manure and inorgani nitrogen fertilizer. 1. Enviro17. Qual., 28, Bergstrom L. F. and Jarvis N. J. (1993) Leahing of dihlorprop, bentazon, and 3GCI in undisturbed field Iysimeters of different agriultural soils. Weed Sei, 41, Botther J, Strebel, Voerkelius S, and Shmidt H. L. (199) Using isotope frationation of -24

9 nitrate-nitrogen and nitrate-oxygen for evaluation of mirobial denitrifiation in a sandy aquifer. J Hydro!., 114, Chang C. and Entz T (1996) Nitrate leahing losses under repeated attle feedlot manure appliations in Southern Alberta. J Environ. Qual., 25, Environment Ageny (1999) Annual report on the environment (Detail:,) 2. (In Japanese.) Gyousei, Tokyo, Japan. Fogler H S (ed.) (1992) Elements of hemial reation engineering, 2nd edn. Prentie-Hall International, Elglewood Cliffs, NI Harada Y (1997) Appliation rates of livestok manure. (In Japanese.) In: Manual for Treatment and Utilization of Livestok Wastes, K Haga et al. (ed.), Livestok industry's environmental improvement organization, Tokyo, Japan, pp Harada Y and Yamaguhi T (1998) Properties of animal waste omposts in Japan. In: Pro. Int. Workshop on Environmentally Friendly Manage. of Farm Animal Waste, T Matsunaka (ed.), Sapporo, November 1997, Kikanshi Insatsu, Japan, pp Kuwabara H (1989) Hydrologial yle. (In Japanese.) In: Handbook of agriultural engineering, H Shimura et al. (ed.), Shin-nihon-insatsu, Tokyo, Japan, p Maeda M., Zhao B., and Ozaki Y (21a) Long-term study on nitrate leahing in a Japanese Andisol treated with two types of hemial fertilizer and swine ompost. In: Groundwater Quality 21: Third Int. Con! on Groundwater Qual., Sheffield, June 21, UK Maeda M., Zhao B., and Ozaki Y (21b) Nitrogen Leahing and Balane in Volani Ash Soil Affeted by Different Fertilizers. In: International Symposium on Nitrogen Fertilization and the Environment in East Asian Countries, Tsukuba, 5-6 February 21, Japan, p. 2. Maeda M. and Bergstbm L. F (2) Leahing patterns of heavy metals and nitrogen evaluated with a modified tanks-in-series model J Contam. Hydro!., 43, Mariotti A, Germon I C, Hubert P., Kaiser P., Letolle R., Tardieux A, and Tardieux P (1981) Experimental determination of nitrogen kineti isotope frationation: Some priniples; illustration for the denitrifiiation and nitrifiation proesses. Plant Soil, 62, Matsumaru T (1997) Suppression of nitrate leahing in upland fields by use of oated fertilizer. (In Japanese, with English abstrat.) Jpl1. J Soil Si. Plant Nutr., 68, OECD (1999) Environmental indiatorsfor agriulture, Volume 2: Issues and design, "The York workshop ". OECD PubL, Paris. Owens L. B, Edwards W. M., and Van Keuren R. W. (1992) Nitrate levels in shallow groundwater under pastures reeiving ammonium nitrate or slow-release nitrogen fertilizer. J Environ. Qual., 21, Persson L. and Bergstrbm L. (1991) Drilling method for olletion of undisturbed soil monoliths. Soil Si. So. Am. J, 55, Saigusa M, Kodama H, Shibuya K, and Abe T (1993) Single basal appliation of nitrogen fertilizer on dent orn ultivation using ontrolled release urea. (In Japanese, with English abstrat) J Jpn. Grassland S;, 39, Sakata N., Yamamoto K, Nakahara H, and Marumoto T (1995) Moving of nitrogen from oating fertilizer in soil (In Japanese, with English abstrat.) Jpn. 1. Soil Si. Plant NutI'., 66, Steenhuis T S, Parlange I -Y, and Aburime S. A (1995) Preferential flow in strutured and sandy soils: onsequenes for modeling and monitoring. In: Handbook of vadose zone haraterization and monitoring, Wilson et at. (ed.), CRC Press, In, Boa Raton, FL, pp Thomsen I K, Hansen 1. P, Kjellerup V, and Christensen BT (1993) Effets of ropping system and rates of nitrogen in animal slurry and mineral fertilizer on nitrate leahing from a sandy loam. Soil Use Manage., 9, Yoneyama T (1987) Natural abundane of 13C, 15N, 18, D and 34S in soils and plants: Variations, impliation and utilization. (In Japanese.) Jpn. J Soil Si. Plant Nutr., 58,