An Alternative Approach for the Determination of Soil Water Mobility

Pedosphere 18(3): 328 -334, 2008ISSN 1002-0160/CN 32-1315/PPEDOSPHERE⑥2008 Soil Science Society of ChinaPublished by Elsevier Limited and Science Presswww. elsevier. com/locate/pedosphereAn Alternative Approach for the Determination of SoilWater Mobility*lM. P. C. ENGLER', R. CICHOTA2, Q. DE JONG VAN LIER1,*2, E. M. BLOEM3, G. SPAROVEK1and E. SCHNUG31University of Sao Parlo, C.P. 9, Piracicaba (SP) 13418-900 (Brazil). E-mail: mpcampos@esalq. usp.br2Institute of Natural Resources, Massey University, P.0. Bor 11 222, Palmerston North (New Zealand)Institute for Crop and Soil Science, (JKI) Federal Research Centre for Cultivated Plante, Bundesallee 50, D-38116Braunschweig (Germany)(Received July 25, 2007; revised March 14, 2008)ABSTRACTA new laboratory method was proposed to establish an easily performed standard for the determination of mobile soilwater close to real conditions during the infiltration and redistribution of water in a soil. It consisted of applying a watervolume with a tracer ion on top of an undisturbed ring sample on a pressure plate under a known suction or pressurehead. Afterwards, 8oil water mobility was determined by analyzing the tracer-ion concentration in the soil sample. Soilwater mobility showed如be a function of the applied water volume. No relation between soil water mobility and appliedpressure head could be established with data from the present experiment. A simple one- or two parameter equation canbe fitted to the experimental data to parameterize soil water mobility as a function of applied solute volume. Sandy soilsshowed higher mobility than loamy soils at low values of applied solute volumes, and both sandy and loamy soils showedan almost complete mobility at high applied solute volumes.Key Words: laboratory method, mobile-immobile water, pressure head, soil textureCitation: Engler, M. P. C, Cichota, R, de Jong van Lier, Q, Bloem, E. M., Sparovek, G. and Schnug, E. 2008. Analternative approach for the determination of soil water mobility. Pedosphere. 18(3): 328- 334.The asessment of pesticide, nutrient, and heavy metal movement in soils depends on reliable modelsto describe solute transport. These models can improve insight into the behavior of ions in a porousmedium (difusion, dispersion, anion exclusion, adsorption, or exchange processes), and describe themedium itself (pore geometry, aggregation, and reactivity) (van Genuchten and Wierenga, 1976). So-lute movement also infuences the presence and amounts of pollutants in specific parts of the soil andin groundwater that has fowed through its profle (Van de Pol et al, 1977). Solute movement modelsare essential for estimating impacts of waste disposal, surface mining, and pesticide application and forpredicting fertilizer eficiency and environmental impacts.Preferential fow pathways play an important role in soil water and solute transport (Thomas andPhilips, 1979; Priebe and Blackmer, 1989; Kanwar et al, 1985). Preferential fow consists of a routefor solutes to move rapidly below the biologically active root zone (Casey et al, 1997). It leads to atemporary nonequilibrium condition of solute concentration or pressure head (Brusseau and Rao, 1990)that limits the ability to predict flow and transport processes in undisturbed media that do not considerthese pathways.In order to extend predictions of hydraulic and prefd中国煤化工modeling of solutetransport processes, models usually distinguish a mobile 8YHCNMH Gtion. It is assumedthat an essentially mobile fraction is active in solute trarJ. pauwowo, wuit a vomplementary im-*lProject Bupported by Coordenacao de Aperfeicoamento de Pessoal de Nirvel Superior (CAPNS), Brazil and DeutscheForschungsgemeinschaft (DFG), Germany-Bilateral Cooperation Project 017/04.*2Corresponding autbor. E-mail: qdjvlier@esalq.usp.br.SOIL WATER MOBILITY DETERMINATIONmobile fraction does not participate in mass flow (Gaudet et al., 1977). Mobile (or dynamic) water islocated inside the mobile pore domain (usually large and inter-aggregate pores). The fuid fow in themedium is assumed to occur only in this domain. Solute is transported in the mobile domain as anadvective- dispersive process, and a difusive process exchanges solute between the mobile and immobilefraction. Immobile (or stagnant) water is mostly located inside aggregates (small and intra-aggregatepores). Transfer by diffusion between the two liquid domains (mobile and immobile) is proportional tothe concentration difference between the mobile and immobile liquids.Recently hydrological sulfur model (HYSUMO) was developed. HYSUMO simulates the S cyclein the soil, the soil water fractions, and plant uptake, including the mobile and immobile soil waterfraction. The importance of the mobile/immobile soil water concept for S balance modeling has beendiscussed by Bloem et al. (2005) using simulation results under various scenarios. The authors showedthat the immobile soil water fraction affected leaching, nutrient availability, and groundwater pollution,storing a non- negligible part of nutrients and therefore contributing to plant nutrition. Since Coats andSmith (1964) suggested a mobile-immobile (dead-end pore volume) fraction for characterizing solutionmobility for petroleum engineering, several methodological improvements were made to determine soilwater mobility, but no standardized method is yet accepted.Equipment, tracers, and the necessary equations support for deterministic or stochastic assessmentof soil water mobility have been developed for both field and laboratory conditions. Laboratory studiesto investigate the solute movement through porous media have been made using tracer techniques incolumns with repacked soil (Padilla et al., 1999) and undisturbed soil samples (Jaynes et al, 1995;Lee et al., 2000; Ilsemann et al., 2002). Also, tracer techniques using disk permeameter and tensioninfiltrometer have been proposed for estimating the soil water mobility and the difusion cofficientunder field conditions (Clothier et al, 1992, 1995; Jaynes et al, 1995).Determination methods are usually applied under steady state conditions, which result in parametersthat do not contain all necessary information for modeling real, nonsteady state events. The solutemovement processes under non-equilibrium conditions tend to be more complex than those under steadystate conditions (Wierenga, 1977), and little information is available about them. To improve andvalidate a model like HYSUMO it is necessary to develop a nonsteady state method that allows thequantification of the mobile and immobile soil water fractions under non-equilibrium conditions. In thispaper, a laboratory method for this determination is described. First results from determinations infour soils difering in soil texture and groundwater conditions are reported.MATERIALS AND METHODSThe hydraulic pulse experimentBased on the assumption that soil water mobility depends on the intensity of the flow process andthe hydraulic conditions during the process, it is considered that mobile and immobile water contents arenonstatic soil parameters and that they should be expressed as a function of the applied water volumeand the initial and final pressure head. Different soil water mobility can be also expected for diferentways of applying the solution. A sudden application of the entire volume may be expected to lead tolower mobile soil water contents than a gradual application of the same volume, which is correlated toa more gradual water movement and involvement of smaller pores.These simple assumptions showed there were many variables involved in the process. To establishan easily performed standard for the determination of mobiconditions duringthe infiltration and redistribution of water in a soil, an amo1中国煤化工top of a sampleon a pressure plate under a known suction or pressure headYHC N M H Gished after waterapplication and the pressure head in the sample was expected to vary in time according to Fig.1. Theinitial pressure head became less negative immediately at the time of solute application (tpulse). If theapplied volume was sufficiently great, the pressure head would increase until 0, as it was the case for V2.330M. P. C. ENGLER et al.and V3 in Fig. 1, but for smaller volume (Vi) the sample would not reach saturation. The applied waterwould pass through the sample and be drained to the suction plate. After that, the initial pressure headwas reestablished.tpulosTime0V,V:V，Fig. 1 Pressure bead as a function of time during a bydraulic pulse experiment with three diferent pulse volumes (V1< V2 < V3). tulee is the time of solute application and hini is the initial pressure head.To allow calculating the soil water mobility, a non-reacting, non-absorbing and easy to detect ionwas 1used as a tracer. Chloride was chosen in this study and bromide or other ions can be employedalternatively. To assure that soil samples were totally equilibrated to a known concentration (Co) ofthe ion, initial tests were performed and showed that it took a considerable time for soil samples toequilibrate by difusion. Therefore, samples were left in a chloride solution for 28 days and ceramicplates were equilibrated to the same concentration for 24 h. Samples were put on the suction plate toequilibrate hydraulically to a pressure head (h) equal to -5, -10, -20, -40, and -60 cm. Filter paper(Schleicher & Schuell No. 593) was put on top of the samples and different volumes of solution (0.5,1.0, and 2.0 times of the mean sample pore volume) with a chloride concentration C1 were applied onthat. Samples and porous plates were then covered to minimize evaporation.Twenty-four hours after C1 application samples were analyzed. One third of the sample was usedto determine the soil water content and the remaining part to measure the chloride concentration [C1-]in the soil solution. To do so, the soil samples were shaken with 50.0 mL deionized water for 24 h. The .suspension was then filtered through Schleicher & Schuell No.593 filter paper and analyzed with a Clspecific electrode (S7, Mettler Toledo), from which [C1-] in the soil solution (Cg, mol L-1) could berecalculated.Supposing chloride difusion to be negligible, the values of Co, C, and C。allow to etimate themobile and immobile soil water fraction relative to the soil water content (θm and θm respectively, inm3 m~3) (Clothier et al, 1995):Co-Cs(1)Co-C10m=C-C(2)Values for mobile and immobile water contents were thus obtained as a function of pressure headand applied pulse volume.Soil samples中国煤化工Undisturbed soil samples (0.04 m high and 0.056 mi: from four Germansoils at a depth between 0.05 and 0.15 m. Two of the sMHC N M H Guntain region andhad a loamy texture. Loam 1 is a soil with a deep groundwater level (2.3 m below surface), while loam2 has a shallow groundwater level (1.5 m below surface), which is connected with a higher clay content.The other two soils with a sandy soil texture were ftom the Lineburger Heide. Sand 1 bhas a deepSOIL WATER MOBILITY DETERMINATION31groundwater level (about 5.5 m below surface), while sand 2 has a high content of organic matter withgroundwater close to the surface (0.8 m deep). The soils were described comprehensively by Bloem(1998). The physical properties of the top soils (0 -15 cm) were analyzed and the results are presentedin Table I.TABLE IPhysical characteristics of the top soils (0-15 cm) of the investigated four soilsBulk densityClaySiltSandkgm-3g kg-11492260470270Loam 21480310510180Sand 114261008820Sand 21 140120760From each soil, 45 undisturbed samples were used to perform the hydraulic pulse experimeat and 3to determine soil water retention at selected pressure heads(-5, -10, -20, - -40, -60, and -100 cm).One of the retention samples of the loam 1 soil showed very different results compared to the othersamples and was thus not considered in subsequent analyses. For hydraulic characterization, retentiondata were fitted to the van Genuchten (1980) equation:θ=[1 +|ah|川]”(3)Withθ-θ(4)6= 6。-0,where日is soil water fraction, 日, 0r, and 0g are water content, residual water content, and saturatedwater content (m3 m" -3), respectively, h is the pressure head (cm), and a (cm- ") and n are empricalparameters.RESULTS AND DISCUSSIONThe observed water retention curves are shown in Fig. 2. The sandy soils, especially sand 1, showeda high porosity in the lower suction range. Sand 1 released more than half of its water at a pressure headaround -50 cm. The loamy soils showed little variation in water content up to -100 cm. Thus, higherwater mobility as well as greater differences between the investigated pressure heads could be expectedfor the sandy soils. Mathematical analysis of Eq. 3 showed that a high value of a was associated to aninlection point (i.e, a maximum porosity) at low values of h, corresponding to a high macroporosity.However, values of n close to 1 led to a low inclination at the inflection point, smoothing the porositydistribution and making its maximum less significant. This was the case for loam 2. In general, highvalues of a are expected to be associated with higher water mobility.The experimentally determined mobile fractions are shown in Fig. 3. While the applied volumesVs equaled 0.5, 1.0, and 2.0 times of the mean sample total pore volume, results were expressed as arelative applied volume v, defned in function of the applied volume and the sample water content atthe respective pressure head:V。中国煤化工D= vOn(5)YHCNMHGwhere V is the total sample volume and On is the water content at a respective pressure head. Theretention data for the sandy soils showed a greater water content range on the studied pressure headintervals, corresponding to a higher dispersion of v values..332M. P. C. ENGLER et al.0.5Loam 1- +Sand 10.4++十0.3a= 0.02557 cm1a= 0.05428 cm'n= 2.0301n= 3.2955ε 0.2h0.1Loam 2Sand2- i+mh;a= 0.04400 cm'a= 0.02937 cm'n= 1.0996n= 2.38800.21-10-100 -1-100Pressure head (cm)Fig. 2 Water retention curves for the investigated loamy and sandy soils without (loam 1 and sand 1) and with (loam 2and sand 2) groundwater infuence.1.00.8| Loam 1Sand0.6:◆0.0Sand至gi▲▲■-5cm◆-10 cm▲-20 cm。-40 cm-60 cm.00.1.0 1.5 2.02.53.0 0.0 0.52.5 3.0Relative volumeFig. 3 Calculated mobile water fraction relative to the soil wate中国煤化工omy and sandy soiswithout (loam 1 and sand 1) and with (loam 2 and sand 2) ground:YC N M H Gof the relative appliedgolution volumes determined at five different pressure heads.The two sandy soils, but especially sand 1, showed mobile water fractions close to 1 even for smallrelative volumes of applied water. A distinct tendency for different pressure heads were not observedSOIL WATER MOBILITY DETERMINATION333from these data, indicating that almost all water fractions participated in movement within the studiedpressure head range. This is in agreement with the general idea that water in sandy soils is retained ininterconnected pores with large diameters. The loamy soils showed higher immobile water contents. Inthe case of loam 1, there seermed to be a tendency of increasing 0m at higher pressure heads. Higher0m in drier conditions is to be expected, as very large pores, responsible for almost all water transportand thus reducing θm under wet conditions are not available for transport at higher pressure heads.To quantify soil water mobility, a simple asymptotic equation is suggested to be fitted to data:0m=_上a+ vwhere the dimensionless parameter a determines the shape of the curve. Low values of a indicate a steepcurve, which reaches high values of θm at low values of v. No significant differences in a values werefound between different pressure heads, but regression to all data obtained for one soil led to signifcantdiferences between the sandy and loamy soils (Table II). This suggests a significant correlation betweena and n from Eq.3 and a from Eq.6, however, present data were insuficient to confirm this hypothesis.TABLE IIMean values and 95% intervals for parameter a (Eq. 6) for the investigated loamy and sandy soilsSoilMean95% minimum95% maximum0.53110.36650.6958Loam 20.49470.41180.5776Sand 10.16230.10100.2236Sand 20.22860.17220.2851Eq.6 yields 0m = 0 for v= 0 and tends to 1 for high values of u. Alternatively, a parameter mightbe added to Eq. 6 to account for the fact that, especially in soils with higher clay contents, a fraction ofsoil water might be immobile even at very high flow volumes. Thusem= u(7)a+vwhere 0m represents the high flow volume mobile fraction. Regressions to this equation with presentexperimental data resulted in the values for 0m close to 1 for all soils, indicating that there was (almost)no immobile water in these soils at very high fow conditions.CONCLUSIONSSoil water mobility was shown to be a function of applied water volume and could be parameterizedby fitting a simple one- or two-parameter equation to experimental data. However, no relation betweensoil water mobility and pressure head could be established with data ftom the present experiment.Sandy soils showed higher mobility than loamy soils at low values of applied solute volumes; both sandyand loamy soils showed an almost complete mobility at high applied solute volumes. As a consequence,under less intense leaching scenarios common to the studied region in Germany, nutrient retention inthe immobile water fraction should be taken into consideration in these soils. The establishment ofphysical relations between soil water mobility and other soil hydraulic functions should be an object offurther investigation, allowing the more routinely determined retention and conductivity properties tobe translated in soil water mobility.中国煤化工YHCNMHGREFERENCESBloem, E. 1998. Schwefel-Bilanz von Agrarokosystemen unter besonderer Bericksichtigung hydrologischer und boden-physikalischer Standorteigenschaften (in German). PhD Thesis, TU-Braunschweig. 156pp.334M. P. C. ENGLER et alBloem, E, Cichota, R., de Jong van Lier, Q, Sparovek, G. and Schnug, E. 2005. The importance of low mobile soil waterfor the S supply of plants. Landbauforsch. Voelkenrode Sonderh 86(Special Issue): 1-10.Brusseau, M. L. and Rao, P. s. C.1990. Modeling solute transport in structured soil: A review. Geoderna. 46: 169-192.Casey, F. x. M., Logadon, s. D., Horton, R. and Jaynes, D. B. 1997. Immobile water content and mass exchange coeficientofa field soil. Soil Sci. Soc. Am. J. 61: 1 030-1 036.Clothier, B. E, Heng, L, Magesan, G. N. and Vogeler, I.1995. The measured mobile-water content of an unsaturatedsoil as a function of hydraulic regime. Aust. J. Soil Res. 33: 397-414.Clothier, B. E, Kirkham, M. B. and McLean, J. E. 1992. In situ measurement of the effective transport volume for solutemoving through soil. Soil Sci. Soc. Am. J. 56: 733 -736.Coats, K. H. and Smith, B. D.1964. Dead-end pore volume and dispersion in porous media. Soc. Petrol. Eng. J. 4:73- 84.Gaudet, J. P., Jegat, H, Vachaud, G. and Wierenga, P. J. 1977. Solute transfer, with exchange between mobile andstagnant water, through unsaturated sand. Soil Sci. Soc. Am. J. 41: 665- 671.Ilsemann, J, van der Ploeg, R. R, Horton, R. and Bachmann, J. 2002. Laboratory method for determining immobile soilwater content and mass exchange coeficient. J. Plant. Nutr. Soil Sci.165: 332- 338Jaynes, D. B, Logsdon, S. D. and Horton, R. 1995. Field method for measuring mobile/immobile water content andsolute transfer rate coficient. Soil Sci. Soc. Am. J. 59: 352 -356.Kanwar, R. s., Baker, J. L. and Lafen, J. M.1985. Nitrate movement through the soil profile in relation to tllage systemand fertilizer application method. Trans. ASAE. 28: 1 802-1 807.Lee, J., Jaynes, D. B. and Horton, R. 2000. Evaluation of a simple method for estimating solute transport parameters:Laboratory studies. Soil Sci. Soc. Am. J. 64: 492- 498.Padilla, I. Y.,, Yeh, T. C. J. and Conklin, M. H. 1999. The efect of water content on solute transport in unsaturatedporous media. Water Resour. Res. 35(11): 3303-3 313.Priebe, D. L. and Blackmer, A. M.1989. Preferential movement of oxygen- 18-labeled water and nitrogen- 15-labeled ureathrough macropores in a Nicollet soil. J. Environ. Qual. 18: 66-72.Thomas, G. W. and Philips, R. E. 1979. Consequences of water movement in macropores. J. Environ. Qual. 8: 149-152.Van de Pol, R. M., Wierenga, P. J. and Nielsen, D. R. 1977. Solute movement in a field soil. Soil Sci. Soc. Am. J. 41:10-13.Van Genuchten, M. Th. and Wierenga, P. J. 1976. Mass transfer studies in sorbing porous media 1. Analytical solutions.Soil Sci. Soc. Am. J. 40: 473- 480.Van Genuchten, M. Th.1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. SoilSci. Soc. Am. J. 44: 892 -898.Wierenga, P. J. 1977. Solute distribution profiles computed with steady-state and transient water movement models. SoilSci. Soc. Am. J. 41: 1 050-1 055.中国煤化工MYHCNMHG