Dual-domain solute transfer and transport processes: evaluation in batch and transport experiments

Soil macropore networks establish a dual-domain transport scenario in which water and solutes are preferentially channeled through soil macropores while slowly diffusing into and out of the bulk soil matrix. The influence of macropore networks on intra-ped solute diffusion and preferential transport in a soil typical of subsurface-drained croplands in the Midwestern United States was studied in batch- and column-scale experiments. In the batch diffusion studies with soil aggregates, the estimated diffusion radius (length) of the soil aggregates corresponded to the half-spacing of the aggregate fissures, suggesting that the intra-ped fissures reduced the diffusion impedance and preferentially allowed solutes to diffuse into the soil matrix. In the column-scale solute transport experiments, the average diffusion radius (estimated from HYDRUS-2D simulations and a first-order diffusive transfer term) was nearly double that of the batch-scale study. This increase may be attributed to a loss of pore continuity and a compounding of the small diffusion impedance through macropores at the larger scale. The column-scale solute transport experiments also suggest that two preferential networks exist in the soil. At and near soil saturation, a primary network of large macropores (possibly root channels and earthworm burrows) dominate advective transport, causing a high degree of physical and sorption nonequilibrium and simultaneous breakthrough of a nonreactive (bromide) and a reactive (alachlor) solute. As the saturation level decreases, the primary network drains, while transport through smaller macropores (possibly intra-ped features) continues, resulting in a reduced degree of nonequilibrium and separation in the breakthrough curves of bromide and alachlor.     

Concentrations of carbon monoxide and methane at two heights above a grass field and their deposition onto the field

To investigate whether wind is a significant driving force in the diffusion of CO and CH₄ from the atmosphere into soil, we measured the concentrations of these two gases at two heights above a temperate grass field in Japan and estimated their deposition velocities using micrometeorological techniques. The concentrations were inversely correlated with wind speed, indicating that the local concentrations were influenced by ground sources. The CO and CH₄ concentrations at 0.33 m were usually lower than those at 1.3 m. Although nocturnal data are suspected to be non-stationary, by selecting several periods when the changes of the concentrations were small but larger than analytical precision, we obtained a CO velocity of 2.9 and 3.9×10⁻² cms⁻¹, agreeing with a CO deposition velocity, 3.4×10⁻² cms⁻¹, obtained by applying a method using CO₂ as a tracer. The CH₄ influx obtained by the method using CO₂ as a tracer was 13 ngm⁻² s⁻¹. The ranges of the CO deposition velocity and CH₄ influx were similar to those obtained in previous studies in grassfields and in a nearby arable field using a closed-chamber technique. This shows that light winds do not greatly accelerate CO and CH₄ uptake by soil.     

Influence of tidal regime on the distribution of trace metals in a contaminated tidal freshwater marsh soil colonized with common reed (Phragmites australis)

A historical input of trace metals into tidal marshes fringing the river Scheldt may be a cause for concern. Nevertheless, the specific physicochemical form, rather than the total concentration, determines the ecotoxicological risk of metals in the soil. In this study the effect of tidal regime on the distribution of trace metals in different compartments of the soil was investigated. As, Cd, Cu and Zn concentrations in sediment, pore water and in roots were determined along a depth profile. Total sediment metal concentrations were similar at different sites, reflecting pollution history. Pore water metal concentrations were generally higher under less flooded conditions (mean is (2.32 ± 0.08) × 10⁻³ mg Cd L⁻¹ and (1.53 ± 0.03) × 10⁻³ mg Cd L⁻¹). Metal concentrations associated with roots (mean is 202.47 ± 2.83 mg Cd kg⁻¹ and 69.39 ± 0.99 mg Cd kg⁻¹) were up to 10 times higher than sediment (mean is 20.48 ± 0.19 mg Cd kg⁻¹ and 20.42 ± 0.21 mg Cd kg⁻¹) metal concentrations and higher under dryer conditions. Despite high metal concentrations associated with roots, the major part of the metals in the marsh soil is still associated with the sediment as the overall biomass of roots is small compared to the sediment.     

Adsorption studies of the herbicide simazine in agricultural soils of the Aconcagua valley, central Chile

Simazine is a s-triazine herbicide that has been applied worldwide for agriculture. This herbicide is the second most commonly detected pesticide in surface and groundwater in the United States, Europe and Australia. In this study, simazine adsorption behaviour was studied in two agricultural soils of the Aconcagua valley, central Chile. The two studied soils were soil A (loam, 8.5% organic matter content) and soil B (clay-loam, 3.5% organic matter content). Three times higher simazine adsorption capacity was observed in soil A (68.03 mg kg⁻¹) compared to soil B (22.03 mg kg⁻¹). The simazine adsorption distribution coefficients (K_d) were 9.32 L kg⁻¹ for soil A and 7.74 L kg⁻¹ for soil B. The simazine adsorption enthalpy in soil A was −21.0 kJ mol⁻¹ while in soil B the adsorption enthalpy value was −11.5 kJ mol⁻¹. These results indicate that simazine adsorption process in these soils is exothermic, governing H bonds the adsorption process of simazine in both the loam and clay-loam soils. These results and the potentiometric profiles of both soils, suggest that simazine adsorption in soil A is mainly governed by simazine–organic matter interactions and in soil B by simazine–clay interactions. The understanding of simazine sorption–desorption processes is essential to determine the pesticide fate and availability in soil for pest control, biodegradation, runoff and leaching.     

Shift in ectomycorrhizal community composition in Scots pine (Pinus sylvestris L.) seedling roots as a response to nickel deposition and removal of lichen cover

Scots pine seedlings were exposed to wet-deposited nickel (Ni) and removal of lichen cover in a dry heath Scots pine forest. Ni deposition affected the colonization of roots by indigenous ectomycorrhizal fungi in contrasting ways in intact and skimmed quadrats. Highest frequencies of tubercle morphotypes of ectomycorrhiza were found in quadrats exposed to 100 mg m⁻² year⁻¹ Ni in lichen covered treatment, while in skimmed quadrats these peaked after the treatment with 10 mg Ni m⁻² year⁻¹. Removal of the lichen layer increased the value of diversity index (H′) of ectomycorrhizal fungal community, probably due to the increase in the evenness of the morphotype distribution. Lichen removal seemed also to improve the condition of the short roots, as the frequencies of poor and senescent short roots were decreased by the removal.     

Column- and catchment-scale transport of cadmium: effect of dissolved organic matter

In the Ellen catchment on the Pinjarra Plain, NE of Perth in Western Australia, cadmium from fertilisers is starting to leach from soils. About 70% of surface soils in the Ellen catchment are sandy and often on top of a shallow ephemeral water table. Adsorption of Cd in the sandy soils of the Ellen catchment was studied by batch adsorption and by leaching small columns of soil. Adsorption of Cd increases linearly with increasing soil organic matter content and exponentially with increasing pH. Cadmium is significantly mobilised in the sandy soils by dissolved organic matter.The capacity of most of the sandy soils in the Ellen catchment to adsorb phosphate from fertiliser has been saturated. Resulting concentrations in Ellen Brook average 500 μg L⁻¹ P. Cadmium is adsorbed more strongly in the sandy soils than phosphate and is just starting to leach into Ellen Brook. From a comparison of Cd/P ratios in water, soils and fertiliser, cadmium concentrations in Ellen Brook are estimated to be at 10–30% of their maximum for complete breakthrough from soils. Present concentrations of Cd in Ellen Brook average 0.1 μg L⁻¹ and are estimated to approach the maximum for complete breakthrough in 100 yr. Maximum Cd concentrations in Ellen Brook could range from 0.6 to 2 μg L⁻¹, depending on rates of input with fertiliser and future increases in agricultural land use in the catchment.Breakthrough curves, resulting from leaching Cd through small columns of sandy soil, indicate that adsorption significantly increases the effective hydrodynamic dispersion of Cd. Longitudinal dispersivities, measured at pore-water velocities of 0.7–14 m day⁻¹, were 5 cm for Cd and 0.1–0.2 cm for chloride. The much greater dispersion of Cd in the sandy soils than of chloride is shown not to be caused by non-equilibrium adsorption.     

Time aggregation of nitrogen leaching in relation to critical threshold values

Leaching of nitrate contributes to the deterioration of groundwater and can consequently have a negative influence on the quality of our drinking water. Critical threshold values for nitrogen leaching are established to preserve groundwater quality. A critical threshold value for nitrate leaching of 50 mg 1⁻¹ (11.3 mg N 1⁻¹), similar to the drinking water standard, serves as a threshold value for European countries. However, the temporal aggregation scale on which this threshold value should be considered is unknown. A well tested simulation model was used to evaluate the exceedance of the threshold value at different time aggregation levels, ranging from one day till 30 yr. For three different soil structure types within one soil type and a selected fertilisation regime, the aggregated nitrogen leaching over 30 yr was 11.4, 19.2 and 10.6 mg 1⁻¹. Considering an aggregation level of one day, the critical threshold value of 11.3 mg N 1⁻¹ was exceeded 2973, 5801 and 2556 times, respectively, for the three structure types during 30 yr. By considering other time aggregation levels, a clear relation resulted between time aggregation level and the number of time elements during which the critical level was exceeded. Results strongly indicate that a critical threshold value for leaching should include an associated time-aggregation level.     

Multispecies transport of metal–EDTA complexes and chromate through undisturbed columns of weathered fractured saprolite

Laboratory-scale tracer experiments were conducted to investigate the geochemical and hydrological processes that govern the fate and transport of organically chelated radionuclides and toxic metals in undisturbed saturated columns of weathered, fractured shale saprolite. Three long-term, reactive contaminant injections were pulsed onto three separate soil columns, with the following influent mixtures: (1) ¹⁰⁹CdEDTA²⁻, (2) ¹⁰⁹CdEDTA²⁻ and ^57,58Co(II)EDTA²⁻, and (3) ¹⁰⁹CdEDTA²⁻, ⁵⁷Co(III)EDTA⁻, and H⁵¹CrO₄⁻. Both single and multiple species experiments were conducted to determine the importance of interaction between the contaminants and competition for surface sites. Flow interruption was used to identify physical and chemical non-equilibrium (PNE and CNE) which were caused by multiple pore-region flow and rate-limited chemical reactions, respectively. Reactive contaminant transport through the fractured, weathered shale was affected by sorption, redox, and dissociation reactions, which were mediated by soil organic matter and surficial oxides of Fe, Mn, and Al. The transport of CdEDTA²⁻ was significantly influenced by ligand-promoted dissolution of subsurface Fe and Al sources, resulting in the liberation of Cd²⁺, Al(III)EDTA⁻ and Fe(III)EDTA⁻. Flow interruption confirmed that the surface-mediated dissociation reaction was time-dependent, with the stability of the CdEDTA²⁻ complex dependent on its residence time within the soil. The migration of Co(II)EDTA²⁻ was dominated by oxidization to the highly stable Co(III)EDTA⁻ species, and elevated effluent Mn²⁺ suggested that surficial Mn(IV) oxides likely catalyzed the redox reaction, though Fe-oxides may have also contributed to the reaction. Dissociation (12%) of the Co(II)EDTA²⁻ complex was first observed during flow interruption, indicating that rate-limited dissociation of the complex by Fe-oxides may be significant under equilibrium conditions. The transport of HCrO₄⁻ was significantly altered by the reduction of mobile Cr(VI) to irreversibly bound Cr(III). The reduction reaction was catalyzed by surface-bound natural organic matter and flow interruption confirmed that the reaction was time-dependent. There was little evidence of competitive effects between the various contaminants in the multispecies experiments, since each was influenced by a different geochemical process during transport through the soil. The results of this study further support research findings that suggest anionic toxic metals and radionuclide–organic complexes can be significantly influenced by soil geochemical processes that can both enhance and impede the subsurface migration of these contaminants.     

Nitrate reduction during ground-water recharge, Southern High Plains, Texas

In arid and semi-arid environments, artificial recharge or reuse of wastewater may be desirable for water conservation, but NO₃⁻ contamination of underlying aquifers can result. On the semi-arid Southern High Plains (USA), industrial wastewater, sewage, and feedlot runoff have been retained in dozens of playas, depressions that focus recharge to the regionally important High Plains (Ogallala) aquifer. Analyses of ground water, playa-basin core extracts, and soil gas in an 860-km² area of Texas suggest that reduction during recharge limits NO₃⁻ loading to ground water. Tritium and Cl⁻ concentrations in ground water corroborate prior findings of focused recharge through playas and ditches. Typical δ¹⁵N values in ground water (>12.5‰) and correlations between δ¹⁵N and ln C_NO⁻3–N suggest denitrification, but O₂ concentrations ≥3.24 mg l⁻¹ indicate that NO₃⁻ reduction in ground water is unlikely. The presence of denitrifying and NO₃⁻-respiring bacteria in cores, typical soil–gas δ¹⁵N values <0‰, and decreases in NO₃⁻–N/Cl⁻ and SO₄²⁻/Cl⁻ ratios with depth in cores suggest that reduction occurs in the upper vadose zone beneath playas. Reduction may occur beneath flooded playas or within anaerobic microsites beneath dry playas. However, NO₃⁻–N concentrations in ground water can still exceed drinking-water standards, as observed in the vicinity of one playa that received wastewater. Therefore, continued ground-water monitoring in the vicinity of other such basins is warranted.     

The potential inhalation hazard posed by dioxin-contaminated soil

Conservative models were used to estimate the airborne concentrations of 2,3,7,8 tetrachlorodibenzo-p-dioxin (TCDD) vapor and particulates originating from soil containing 100 ppb TCDD. The upper-bound estimates were 3.25 pg/m³ of airborne TCDD vapor on-site and 0.51 pg/m³ for TCDD vapor 100 meters downwind. The TCDD air concentration on-site due to suspended particulate is estimated to be 1.4 pg/m³, based on a TSP level of 0.07 mg/m³. Assuming 70 years of continuous exposure to these concentrations, the upper-bound cancer risks determined from the Jury model were estimated to be 9.4 × 10⁻⁶ to 1.1 × 10⁻⁴ and 1.5 × 10⁻⁶ to 1.7 × 10⁻⁵ for inhalation of on- and off-site vapor, respectively, and 4.1 × 10⁻⁶ to 4.6 × 10⁻⁵ for dust inhalation. Since few sites have average soil concentrations as high as 100 ppb TCDD, this worst-case analysis indicates that inhalation will rarely, if ever, be a significant route of exposure to TCDD-contaminated soil. Experimental results support this claim and point to much lower risk estimates (8.4 × 10⁻⁹ to 9.9 × 10⁻⁸), suggesting that the parameters used in the Jury model are likely to overestimate the actual airborne levels of TCDD at contaminated sites.     

Adsorption of uranium(VI) on amorphous ferric oxyhydroxide at high concentrations of dissolved carbon(IV) and sulfur(VI)

Amorphous ferric oxyhydroxide is being used to treat groundwater contaminated with uranium(VI); the compound also has potential for use as a component in in situ chemical barriers. To quantitatively evaluate its effectiveness in such applications, adsorption of uranium(VI) onto amorphous ferric oxyhydroxide was experimentally investigated under a wide range of uranium(VI) (8.40·10⁻⁷−2.10·10⁻³ mol L⁻¹; 0.2–500 mg L⁻¹), sulfur(VI) (0–0.07 mol L⁻¹; 0–2240 mg L⁻¹) and carbon(IV) (0–0.0195 mol L⁻¹; 0–234 mg L⁻¹) concentrations, and pH-values (4–9.6). The adsorption behavior of uranium(VI) (uranyl ion and its complexes) is similar to that of other cations; it exhibits a sharp rise in the extent of adsorption with increasing pH. Interactions among uranyl complexes and surface sites are interpreted using a site complexation model. Although the model does not incorporate electrostatics and includes only a single type of adsorption site, it provides a reasonable match to measured adsorption and proton exchange data. The simplicity of the model and the resulting reduction in computational demand allow its efficient incorporation into coupled reaction-transport models.     

Kinematic wave approximation to solute transport along preferred flow paths in soils

A routing procedure is introduced which accounts for the loss of a conservative solute tracer from preferred paths during macropore flow. Water flow is treated as a series of kinematic waves from which the tracer is lost due to mixing previously stored soil water, and an expression for solute loss is added to a previously developed model. The model parameters are estimated through experiments at three different input rates applied to a column of a macroporous forest soil.The results of seven experimental runs indicate that solute losses are consistently highest at the early stages of infiltration and drainage flow. An empirical relationship is proposed which links the frequency distribution of the flow parameter with that for solute loss from the preferred path during transient water flow and solute transport.     

Contamination in Ontario farmstead domestic wells and its association with agriculture:: 1. Results from drinking water wells

Groundwater provides about 30% of water requirements in Ontario, but farm families depend almost entirely on private wells. Major potential contaminants on farms are nitrate (NO₃⁻), pathogenic microorganisms, pesticides and petroleum derivatives. A survey of farm drinking-water wells was conducted throughout the Province of Ontario, Canada, in 1991 and 1992 and tested for these contaminants. The main objectives of the survey were to determine the quality and safety of drinking water for farm families, and determine the effect of agricultural management on groundwater quality at a provincial scale. Four farm wells were chosen in each township where >50% of the land area was used for agricultural production. Elsewhere one well per township was usually sampled. Within each township the types of farming activity and dominant soils were additional criteria for selection. The network comprised 1292 of the estimated 500,000 water-wells in Ontario, and the study conformed to a stratified random survey. A subset of 160 wells, chosen by farm type, soil, and the presence or absence of a fuel storage tank, was investigated for the presence of petroleum derivatives: benzene, toluene, ethyl benzene, and xylene. About 40% of farm wells tested contained one or more of the target contaminants above the maximum acceptable concentration; 34% of wells had more than the maximum acceptable number of coliform bacteria, 14% contained NO₃⁻-N concentrations above 10 mg l⁻¹ limit and about 7% were contaminated with both bacteria and NO₃⁻. Only six wells contained pesticide residues above the interim maximum acceptable concentration (IMAC), but pesticides were detected in 7% of wells in winter and in 11% in summer. No wells contained detectable petroleum derivatives. These results for NO₃⁻ contamination were not significantly different from those reported for a survey of Ontario wells for the period 1950–1954, but the frequency of contamination by Escherichia coli was greater in the present study. None of the point sources investigated contributed significantly to the NO₃⁻ contamination. The percentage of wells contaminated by coliform bacteria decreased significantly with increasing separation of the well from the feedlot or exercise yard on livestock farms. A full statistical model including the type of well construction, depth, age and soil hydrologic group was developed to describe the frequency of NO₃⁻ contamination.     

Transient analysis of volatile organic compound concentrations for estimating emission rates

While emission rates of volatile organic compounds (VOCs) have been obtained for building materials, furnishings and processes in chambers, field measurements are more difficult. Procedures to estimate emission rates using transient analysis of VOC concentrations are described and applied in a two-story classroom/office building. The analysis employs semi-real-time VOC concentrations determined with a portable GC/FID and simultaneous air change rate measurements using tracer gas decay. The results of the analysis yield consistent values of emission rates for building materials ranging from 0.20 to 0.40 mg m⁻² h⁻¹ when normalized by floor area. Occupancy-related emissions were more difficult to estimate and covered a wider range from roughly 0.1 to 1.5 mg m⁻² h⁻¹. The test data were also analyzed in an attempt to determine sink parameters, but these efforts were not particularly successful. Furthermore, in these tests, the inclusion of sink effects did not significantly impact the estimated emission rates. While this paper offers a transient analysis approach that may lead to improved field estimates of VOC emission rates, it is not presented as a definitive methodology. Nevertheless, transient analysis has potential for use in other buildings, but simultaneous air change rate measurements are critical in its application in estimating VOC emission rates in the field.     

Development of steady-state phosphate concentrations in septic system plumes

Long-term monitoring of PO₄⁻³ behaviour in a well-defined septic system plume on calcareous sand (Cambridge site) shows that, after 17 yr of system operation, a distinct PO₄⁻³ plume (PO₄⁻³−P > 1 mg L⁻¹) is present extending 20 m downgradient from the infiltration bed. The PO₄³⁻ plume migration velocity is 1 m yr⁻¹, reflecting retardation by a factor of 20 compared to the groundwater velocity. During monitoring between years 10 to 17, an expanding steady-state zone was noted below the infiltration bed where PO₄³⁻ −P levels remained consistently near 4 mg L⁻¹, a value 25% lower than the average effluent value (6.3 mg L⁻¹). The pattern of attenuation — a 25% mass loss in the 2-m-thick vadose zone, then little further attenuation along the flowpath — is suggestive of a condition of equilibrium with a controlling phosphate mineral phase. Chemical equilibrium modelling shows supersaturation with respect to hydroxylapatite and variscite. Four other field sites are identified from the literature and from our work where similar steady-state PO₄³⁻ zones are present in septic system plumes. In these, steady-state levels range from 15% to 68% of effluent values, with lower concentrations observed in the more acidic plumes, again indicative of a mineral solubility control, possibly variscite.PO₄³⁻ behaviour in these plumes suggests that, although P migration velocity is controlled by the processes of sorption, the magnitude of PO₄³⁻ that is present is governed by the constraints of phosphate mineral solubility. When septic systems on sands are located relatively close to sensitive surface water bodies and when long-term downgradient impact is the primary concern, more attention should be focused on the geochemical conditions that control PO₄³⁻ mineral solubility rather than only on the sorption characteristics of the sediment.     

Analyses of locally measured bromide breakthrough curves from a natural gradient tracer experiment at Krauthausen

Solute travel time distributions were derived from breakthrough curves (BTCs) of bromide concentrations, which were measured during a large-scale tracer experiment in a quaternary fluviatile aquifer at Krauthausen. Travel time distributions to a specific point in the aquifer were derived from locally measured BTCs, using averaged absolute concentrations ?abs(x1,t), normalized concentrations ?norm(x1,t), and velocity-weighted normalized concentrations ?vw(x1,t). The travel time distributions were characterized in terms of equivalent convective-dispersive transport parameters: the equivalent solute velocity and equivalent dispersivity. Parameters were derived from BTCs using moment analyses and least-squares fits of the 1-D convection-dispersion equation (CDE). Both local and averaged BTCs showed pronounced tailing which was not well described by the 1-D CDE and which indicates the presence of macroscopic regions with low velocities in the aquifer. Therefore, dispersivities derived from CDE fits were significantly smaller than those derived from time moments. The BTCs of ?abs(x1,t) were dominated by only a few local BTCs with high concentrations and were less representative for the travel time distribution than BTCs of averaged normalized concentrations. Dispersivities derived from ?norm(x1,t) and ?vw(x1,t) were very similar. Finally, estimates of dispersivities and vertical correlation length of lnK, gamma 3, from BTCs were in agreement with a first-order estimate of the dispersivity and gamma 3 based on grain size data and flow meter measurements.     

Depleted N carryover, leaching and uptake for three years of irrigated corn

Deep percolation of nitrate can contribute to the deterioration of groundwater resources. Leaching of nitrate is a complex process affected by fertilizer and irrigation practices, efficiency of N use by the crop, and how the soil's water holding capacity and water transmission properties are affected by soil texture. Depleted (¹⁵NH₄)₂SO₄ fertilizer at N rates of 0, 125, 250 and 375 kg ha⁻¹ was applied annually for 3 years to continuous corn grown within three different water regimes. This time period and the labeled N permitted an evaluation of N use efficiency by the crop and NO₃ leaching and carryover on a Weld silty clay loam, a fine-textured soil, typical of the “hardland” soils of the semi-arid Great Plains. Three water regimes, W₁ ( 1.5 ET), W₂ ( ET) and W₃ ( 0.8 ET), were used. Beneath each plot within each water regime, Duke-Haise vacuum trough extractors were installed under undisturbed soil profiles at 1.22-m depth to measure weekly percolate and the NO₃ concentration in the percolate. The corn was harvested in the fall in the dent stage to measure the total above-ground biomass N uptake. Soil profiles (1.8 m) were sampled annually in the fall after crop harvest to determine NO₃---N in the soil or carryover.Great variability was encountered in measuring the amount of extractor water and its NO₃ content under each water regime, which made estimates of N03 leaching losses unreliable. Also, the variability demonstrates formidable problems in quantifying percolation losses with vacuum trough extractors under undisturbed fine-textured soil profiles. With the highest N rate of 376 kg ha⁻¹ yr⁻¹ and within the water regime W₁, where leaching was expected to be greatest, only 1% of the cumulative labeled N applied was found in extractor waters and most movement of the labeled N into extractors occurred the third year. The 125-kg-ha⁻¹ yr⁻¹ fertilizer N rate significantly increased the crop yield over the unfertilized plots without increasing residual NO₃---N accumulation; whereas fertilizer N rates of > 125 kg ha⁻¹ yr⁻¹ did not appreciably increase plant yields over the 125-kg-ha⁻¹-N rate, but did appreciably increase residual NO₃.     

Flux density and breakthrough times for water and tracer in a spatially variable, compacted clay soil

Flux density values, computed from observed infiltration and outflow measurements at 184 locations in a 0.3-m-thick, 9m × 23m layer of compacted clay subsoil, are compared to effective flux density values that are based on breakthrough time distributions for water and Br⁻ tracer over the same area. Results suggest that both water and tracer move at similar rates, but considerably faster than expected, on the basis of flux density alone, and that only a small fraction of the total pore space is involved in active transport. The ramifications of these findings are explored against the background of effective porosity, degree of compaction, and observed changes in bulk density with time.