Application of a two-dimensional model to simulate flow and transport in a macroporous agricultural soil with tile drains

It is essential that important field processes are taken into account to model water flow and chemical transport accurately in agricultural fields. Recent field studies indicate that transport through macropores can play a major role in the export of solutes and particulates from drained agricultural land into surface water. Non‐ideal drain behaviour may further modify the flow and transport. We extended an existing two‐dimensional flow and transport model for variably saturated soils (SWMS_2D) by adding a macropore domain and an additional Hooghoudt drain boundary condition. The Hooghoudt boundary condition accounts for an entrance head needed to initiate flow into the drains. This paper presents the application of the new model (M‐2D) to an agricultural field in Switzerland. To understand interactions between macropore flow and drains better we simulated water flow and bromide transport for four different field scenarios. We considered both collector drains only with an ideal drain boundary condition (with and without macropores) and collectors and laterals with a Hooghoudt boundary condition (also with and without macropores). For each scenario, inverse modelling was used to identify model parameters using 150 days of data on observed cumulative discharge, water table depth, and tracer concentration. The models were subsequently tested against a 390‐day validation data set. We found that the two additional components (macropore flow, drain entrance head) of the M‐2D model were essential to describe adequately the flow regime and the tracer transport data in the field.     

Identification of efficient transport pathways from the soil surface to field drains by smoke injection

Surface‐applied agrochemicals have been measured soon after application in drainage water from agricultural fields in various studies. The bypass flow from the soil surface into drains may result from direct macropore connections. In order to identify such macropores, smoke was injected into three 1.15‐m deep tile drains in a sandy loam soil using ‘smoke oil’ and an air blower. Smoke‐emitting macropores (SEM) at the soil surface were characterized as having either strong or weak plumes when compared with reference plumes from 3‐ and 6‐mm wide tubes. A total drain length of 93 m, at three separate sub‐drain lines, was investigated after harvest of wheat in autumn 2010 and in spring 2011. Smoke only reached the soil surface layer via earthworm burrows, located in a 1‐m wide belt directly above the drain lines. The distributions of SEM across the drain line were similar in autumn and spring. The average number of SEM along the drain lines was 2.3 SEM m^?1 drain line. Ponded water in 6‐cm wide rings was applied above 52 burrows, including 17 reference burrows that did not emit smoke, and 13 pathways in the soil were examined with dye tracer. Strong SEM marked the entrance of efficient transport pathways conducting surface‐applied water and dye tracer into the drain. Water infiltration rates were significantly larger (P < 0.05) in strong SEM (geometric mean, 195 ml minute^?1; n = 19) than in weak SEM (geometric mean, 63 ml minute^?1; n = 16) and for the reference burrows (geometric mean, 39 ml minute^?1; n = 17). The results suggest that the smoke injection technique is a valuable means of identifying potentially efficient pathways for surface‐applied contaminants to enter drains, such pathways being associated primarily with strong SEM.     

The influence of soil management on drainage hydrographs

Abstract. Drainage hydrographs from mole-drained plots having different tillage treatments (tractor tined, gantry tined, tractor ploughed and gantry ploughed) were measured for different rain events over a growing season. In the autumn just after tillage, a large rainfall produced peaky drain flows on the tined plots but a rather flat response from the ploughed plots. In the winter, the drain response to a small rain event showed less differences in peak flows between the treatments. The recession time constant of the hydrographs was used as an index of the structural macropore development in the soil above the water table. Hydrographs from the gantry plots recessed more quickly than those on the tractor plots and those on the tined plots recessed more quickly than those on the ploughed plots. Lack of soil compaction on the gantry plots and continuous vertical fissuring created by the non-inverting tillage tines resulted in the gantry tined treatment having the fastest drainage response. In the ploughed plots compaction and smearing of the soil at the base of the plough layer restricted the rate of downward movemenl of water. The work indicates that soil management practices can play an important role in the drainage and leaching of aggregated soils.     

Infiltration patterns into two soils under conventional and conservation tillage: influence of the spatial distribution of plant root structures and soil animal activity

The characteristics and properties of the soil macropore system may cause different infiltration behavior under different tillage practices. To evaluate the effect of a specific tillage system on infiltration and percolation with particular regard to the influence of crop structure and soil animal activity dye tracer irrigation experiments were conducted in a silty (Luvisol) as well as in a sandy loam soil (Podzolluvisol). The spatial distribution of water flow paths was experimentally examined at four square areas of 0.49 m², under conservation and conventional tillage. Natural rainstorms were simulated by irrigating the plots with 2.8×10⁻³ M methylene blue solutions. For both soils the root crowns of the agricultural crop, wormcasts and stained soil sections as well as macroscopic conduits were traced on plastic sheets. The investigated soil depths were 0, 5, 10 and 20 cm for the both soils. For the Luvisol, the 30, 40, 50, 80 and 120 cm depths were also studied.For the Luvisol, the conservation tillage plot revealed pronounced vertical connectivity and continuity of the macropore network (maximum depth of stained pores=120 cm), while at the conventional tillage plot, continuous macropores were observed to soil depths of 50 cm, but mainly restricted to the ploughed topsoil (0–30 cm soil depth).For the Podzolluvisol, at the conservation tillage site extensive mulch residues prevented water transport beneath 5 cm soil depth. In contrast, at the conventionally tilled site stained water reached a depth of 20 cm. For all investigated plots on both soil types, the location of the root crowns of agricultural crop and of wormcasts was not related to percolation patterns.The results suggest that conservation tillage on silty soils under agricultural landuse could induce an increased water retention capacity reducing the significance of fast runoff components.     

Solute recycling by crops and leaching in a drained arable soil

Preferential flow, as it bypasses the soil matrix, can greatly enhance the leaching of chemicals. When a soil is drained there is the risk that such short‐circuiting results in more or less direct passage of polluting chemicals from the soil to the groundwater. If the groundwater table is shallow the chemicals could be transferred back into the surface soil by hydraulic lift through roots and subsequent release by exudation or from decaying plant residues and again become exposed to leaching by preferential flow, thus strongly enhancing the chance of export via the drains. We investigated the leaching of bromide in a tile‐drained arable field over 2 years of crop rotation. The site was a former wetland, artificially drained a century ago for agriculture. Bromide was applied over 1.6 ha at a dosage of 10 g Br per m² in August 1995 after the harvest of wheat. During the 2 years 18% of the applied bromide was exported via the drainage system, most of it in preferential flow events and more than half of it in a single winter storm 5 months after the application. Within 7 months 56% of the applied tracer was leached out of the main root zone into the groundwater. Subsequently the tracer re‐emerged in water taken up by sugar beet in the following season. The beet accumulated 50% of the initially applied bromide in their leaves and released it again after harvest when the leaves were left as green manure on the field. Our results show that this recycling of solutes to the topsoil can have an important influence on their leaching as the solutes are thus again exposed to preferential transport into drains in the course of preferential flow events.     

Water flow characterization and test of a kinematic-wave model for macropore flow in a highly contrasted and irregular double-porosi medium

Laboratory experiments to characterize water flow through a column consisting of an artificial double-porosity medium are described. A gamma-attenuation device was used to measure water storage in the intra- and inter-aggregate porosity domains after successive infiltrations. Several days after macropore flow had ceased, a significant proportion of the water that had infiltrated was retained within the aggregates. After the column had been brought to its water retention capacity, several rains with different intensities were applied, and the drainage hydrographs were recorded and analysed in order to characterize water transport in the coarser pores and to test the kinematic-wave approach to macropore flow proposed by Germann (1985). Reasonable fits of calculated hydrographs to experimental data were obtained for high input intensities, but the model was unable to reproduce the hydrographs recorded at low intensities. In particular, the model could not simulate the high dispersion of the input signal observed at low rain intensities.     

Effect of water flow on the mobility of the root-knot nematode Meloidogyne incognita in columns filled with glass beads, sand or andisol

In the present study, the migration of nematodes was studied in columns filled with three materials of different textures and chemical properties. The role of soil pores that enable root-knot nematode (Meloidogyne incognita) second stage juveniles (J2) to escape rapid water flow in soil was demonstrated using columns filled with glass beads, sand or andisol that maintained a constant water flow. Under a constant flow flux of 36 cm h⁻¹, living J2, dead J2 or anion bromine tracer (Br⁻) was injected in the middle of the column and then drainage water equivalent to two pore volumes (PV) was collected. The passive transport of the anion tracer in water flow could be explained by a convection dispersion equation. The dead J2 showed a pattern similar to that of Br⁻. However, the living J2 resisted movement in the water flow and remained in the column even at the highest water flow rate of 93.3 cm h⁻¹ in glass beads. The mobility of living J2 was affected by the filling materials; the number of J2 passing through the column was much lower in the andisol-filled column than in the other two columns but the total number of J2 in drainage water was 5% or less of the number injected for all columns. We suggest that J2 were affected not only by soil water flow but also by soil pore structure and have the ability of withstanding or avoiding movement in soil water flow.     

Herbicide sorption, degradation, and leaching in three Swedish soils under long-term conventional and reduced tillage

Soil tillage has the potential to influence water flow and solute transport through the soil by cutting continuous macropores which connect the soil surface to the subsoil. Tillage also affects soil organic carbon sequestration which may lead to different sorption and degradation properties depending on the long-term tillage practices. The objective of this study was to quantify the differences in sorption, degradation and leaching of the herbicides bentazone and isoproturon between conventional tillage (CT) and reduced tillage (RT) under Swedish conditions. Three sites, Ultuna (silty clay), Säby (loam), and Lönnstorp (sandy loam moraine till), where replicate plots had been under either CT or RT for at least 9 years, were included in the study. A higher organic carbon content had developed in the top 5 cm of RT plots compared to the 10–20 cm depth and CT plots since the establishment of the experimental treatments. Adsorption and degradation were studied in laboratory experiments and solute transport was studied in undisturbed column experiments using non-reactive tracers and herbicides. The results from the column experiments were not significantly different between tillage treatments for Säby and Lönnstorp. For Ultuna, RT resulted in a more pronounced preferential tracer transport pattern and isoproturon leaching was twelve times larger compared to CT columns. This indicates that the tillage treatment had affected the macropore connectivity only at Ultuna. Freundlich adsorption coefficients for both bentazone and isoproturon were larger (though not always significantly) in the top 5 cm of RT soil compared to the 10–20 cm depth and to CT, reflecting the higher organic carbon content. The degradation rate was also generally larger (though not always significantly) in the top 5 cm of RT soil. These results show that RT has the potential to reduce pesticide leaching. However, any such reduction may be counter-balanced by enhanced preferential flow for soils where RT results in improved macropore connectivity.     

Localization of denitrification activity in macropores of a riparian wetland

Soil structure heterogeneity in the form of macropores and preferential flow channels can complicate efforts to quantify the physical and biological characteristics of wetland systems. We collected soil cores from two riparian wetlands to determine whether soil associated with macropores had elevated denitrification potentials compared to bulk soil from the same core. Cores were inspected for obvious macropores, which were distinguished as visible holes in the core, sometimes with decaying root matter, or as highly unconsolidated layers that appeared to have a substantially lower bulk density than the surrounding soil. Denitrification potentials were significantly higher in pores (P<0.05) for six of the 16 cores that were obtained from the Cheraw State Park site. In cores obtained from a second site, denitrification potentials were significantly higher in pores for six of 20 cores and the trend of higher denitrification in pores was present in the majority of cores that had measurable activity. In cores with significant differences, denitrification was often 1-2 orders of magnitude greater in soil surrounding the macropore than in the bulk soil. Denitrification potentials of the bulk soils were similar in magnitude to the potentials measured in composited cores from previous studies. It is possible that the difference between macropore and bulk denitrification rates developed due to preferential flow of nitrate-rich water through the macropores. Previous work showed that water entering these riparian systems in groundwater and storm runoff had elevated levels of NO₃⁻.     

Identification of Lateral Macropore Flow in a Forested Riparian Wetland through Numerical Simulation of a Subsurface Tracer Experiment

Understanding wetland hydrogeology is important as it is coupled to internal geochemical and biotic processes that ultimately determine the fate of potential contaminant inputs. Therefore, there is a need to quantitatively understand the complex hydrogeology of wetlands. The main objective of this study was to improve understanding of saturated groundwater flow in a forested riparian wetland located on a golf course in the Lower Pee Dee River Basin in South Carolina, USA. Field observations that characterize subsurface wetland flow critical to solute transport originating from storm-generated runoff are presented. Monitoring wells were installed, and slug tests were performed to measure permeabilities of the wetland soil. A field-scale bromide tracer experiment was conducted to mimic the periodic loading of nutrients caused by storm runoff. This experiment provided spatial and temporal data on solute transport that were analyzed to determine travel times in the wetland. Furthermore, a 3-D numerical, steady-state flow model (MODFLOW) was developed to simulate subsurface flow in the wetland. A particle tracking model was subsequently used to calculate solute travel times from the wetland inlet to the outlet based on flow modeling results. It was evident that observed tracer breakthrough times were not typical of these measured wetland soil matrix conductivity values. Based on surface water sampling results at the wetland outlet, tracer arrival time was about 9 h after the injection of the tracer. These results implied an apparent mean K value of 2,050 m/day, which is 152 times larger than the mean of the measured values using slug tests (13.4 m/day). Modeling efforts clearly demonstrated this implied preferential flow behavior; particle travel times resulting from the calibrated flow model were in the order of hundreds of days, while actual travel times in the wetland were in the order of hours to a few days. This significant difference in travel times was attributed to the presence of macropores in the form of dead root channels and cavities forming a pipe-flow network. The analyses presented in this study resulted in an estimate of the ratio of matrix permeability to matrix plus macropore permeability of approximately 1/150. Eventually, the tracer test and resulting travel times between various points in the wetland were critical to understanding the true wetland flow dynamics. The final conceptual model of the hydraulic properties of the wetland soils comprised a low permeability matrix containing a web of high K macropores. Simulation of tracer transport in this system was possible using a flow model with significantly elevated K values.     

紫色土坡耕地土壤大孔隙流的定量评价

为阐明大孔隙丰富且孔径呈两极分化的紫色土坡耕地土壤大孔隙流的运移规律,通过室内土柱试验获取耕作层0~20 cm、非耕作层20~40 cm原状土柱和填装土柱的穿透曲线,分析饱和条件下土壤大孔隙流发生规律,并采用解析法CXTFIT软件拟合了水分优先运移参数,PFSP指标(大孔隙流引起的穿透曲线延展量与水动力弥散作用及两区作用引起的延展量的比值)定量评价土壤大孔隙流的贡献率。研究结果表明:1)以填装土柱水流为平衡基质流计算,耕作层0~20 cm原状土柱中大孔隙流的导水贡献率为66.2%~68.5%,而Br-累积淋出量占总淋出量的62.3%~66.1%。对于非耕作层20~40 cm,土壤大孔隙流导水贡献率为0.2%~1.7%,而Br-随大孔隙流运移的比例却达14.5%~20.5%。说明耕作层土壤中大孔隙流现象远比在非耕作层土壤中更为显著;2)PFSP值结果表明大孔隙流作用对穿透曲线延展量的贡献率最大,两区交换运移作用次之,水动力弥散作用的最小。即PFSP值越大,大孔隙流对总水流通量的贡献率越大。     

The fate of cadmium in field soils of the Danubian lowland

The susceptibility of soils to deep penetration of cadmium was assessed by measuring cadmium adsorption on soil particles <0.01 mm, easily mobile in soil macropores, and bypassing ratio. The latter is defined as a ratio of the rate of macropore flow to the rate of total (macropore and matrix) flow. Measurements were made on soils from the Danubian lowland, which is a large (1260 km²) agriculturally managed area situated in the south-west of Slovakia, with a shallow (0.5–3.8 m deep) underlying aquifer containing about 10 km³ of freshwater. In this study, the susceptibility of soils to deep penetration of cadmium was assessed on light, loamy-sand soil in Kalinkovo, medium heavy, loamy soil in Macov, and heavy, clay soil in Jurova. It was found that when the interaction between soil and cadmium lasted 1 min, more than 35, 32, and 48% of cadmium was adsorbed on the particles <0.01 mm in soils from Kalinkovo, Macov, and Jurova, respectively. In the case of ponding infiltration, more than 50% of water can flow via topsoil macropores in Kalinkovo, about 70% in Macov, and 96% in Jurova. This value of bypassing ratio can be met during an irrigation/rain with higher intensity then the infiltration rate into the soil matrix of studied soils. As the rains resulting in the macropore flow can occur 24 times on average in south-western Slovakia during the vegetation season, the probability of deep penetration of cadmium is very high, mainly during an initial stage of rainfall. For this reason, some mitigative agricultural practices (e.g. subsurface fertilizer banding or shallow ploughing) should be used in this region to prevent soils from the deep penetration of cadmium.     

Tillage and plastic mulch effects on soil properties and growth and yield of cocoyam (Colocasia esculenta) on an ultisol in southeastern Nigeria

Protected cultivation, mainly represented by plastic-film mulching, has greatly improved crop production worldwide since the 1950s. However, despite its widespread use in tropical USA, Europe and China, its use in sub-Saharan Africa is not widespread. A field experiment was conducted using cocoyam (Colocasia esculenta L. Schott) to evaluate the effects of two tillage systems (tilled and no-till) and plastic-film mulch (black and clear plastic-film mulch) on soil properties and cocoyam growth and yield in 2003 and 2004 planting seasons on a Typic paleudult in southeastern Nigeria. The experiment comprised six treatments and was laid out in the field using randomized complete block design replicated three times. Results showed that 70–80% of the corms emerged 7–8 days (21 days after planting [DAP]) earlier in both tilled and no-till plastic-film mulched plots when compared to the unmulched plots. At later stages of crop development, the plants in the tilled black plastic-film mulched plots were taller by 61–67% than those in the unmulched no-till plots, which had the lowest plant height (27–30 cm). At 98 DAP, there were no significant treatment differences in leaf area index (LAI) between tilled and no-till mulched plots with LAI of 15.5–19.8. However, LAI was reduced in both unmulched plots by 35–54% when compared to the mulched plots. On the average soil temperature was higher in plastic-film mulched plots than that under plots without mulch by about 2 °C. Results show significantly lower soil bulk density (between 1.10 and 1.26 Mg m⁻³) in both tilled clear and black plastic-film mulched plots when compared to the corresponding no-till clear or black plastic-film mulched plots (1.40–1.45 Mg m⁻³). For the two seasons studied volumetric water content (VWC) in tilled black plastic-film mulched plots were significantly higher than VWC in other mulched plots by between 10 and 38% in 2003 and between 17 and 30% in 2004. At harvest (270 DAP) the highest corm yield was obtained in tilled black plastic mulched plots (29.1 Mg ha⁻¹). This was higher (P = 0.05) than yields obtained in no-till, no mulch plots by 72%. Yields were also higher in tilled black plastic mulched plots when compared to tilled clear plastic mulched plots, no-till black plastic mulched plots and no-till clear plastic mulched plots by 29, 47 and 59%, respectively. These findings suggest that plastic mulched plots provide a better soil environment for cocoyam than unmulched plots and that tilled mulched plots especially tilled black plastic mulched plots provide superior edaphic environment for cocoyam when compared to other treatments used.     

Parameterization of the MACRO model to represent leaching of colloidally attached inorganic phosphorus following slurry spreading

Abstract. The dual porosity soil water and contaminant transport model MACRO was tested for its suitability to represent water flows and leaching of phosphorus (P) through field drains following spreading of slurry. These flows are characterized by very high loadings of P, including a high proportion in colloidally attached form, for about one week following winter spreading of slurry, followed by quite a rapid decline to the low background level. Use was made of the option in MACRO for representing colloid facilitated contaminant transport. The model simulates transport through macropores and soil matrix pores (micropores) of contaminant carrying colloids, as well as trapping of colloids by straining and filtration using an adaptation of standard filtration equations. Calibration involved selecting soil hydraulic parameters, colloid filtration coefficients and P sorption characteristics for two soils from measured and literature values. Both P in solution and P attached to colloids were represented in simulated outputs. Reasonable agreement was found between simulated and measured water and leached P flows. Work with the model suggests that macropore flow through the soil to field drains of colloidally transported P is an important component of water pollution associated with slurry spreading     

Effects of tillage and liming on macropore networks derived from X‐ray tomography images of a silty clay soil

Soil structure influences water infiltration, aeration and root growth and, thereby, also the conditions for sustainable crop production. Our objective was to quantify the effects of different soil management methods and land uses on the topsoil structure of a silty clay soil. We sampled 32 intact soil columns (18 cm high, 12.7 cm diameter) from an experimental silty clay field with four treatments: conventional tillage (CT ), conventional tillage followed by liming (CTL ), reduced tillage (RT ) and unfertilized fallow (UF ). The columns were analysed using 3‐D X‐ray tomography. The samples were taken in autumn after harvest, 7 yr after quick lime was applied to the CTL plots. Despite a relatively large number of replicates per treatment (8, 8, 8 and 6 (two UF samples were excluded), respectively), there were no significant differences between any of the investigated macropore network properties related to tilled treatments. The UF treatment, in contrast, exhibited more vertically oriented macropores, which were also better connected compared to the other treatments. This confirms previous findings that tillage may disrupt the vertical continuity of macropore clusters. The impact of liming on soil pore network properties may have been limited to pores smaller than the resolution in our X‐ray images. It is also possible that the effects of lime on soil structure were limited to a few years which means that any effect would have diminished by the time of this study. These matters should be further investigated in follow‐up studies to understand better the potential of lime amendments to clay soil.     

Water movement and macroporosity of an Australian Alfisol under different tillage and pasture conditions

We assessed the effect of different tillage practices (i.e. conventional tillage and direct drilling) and pasture conditions on the infiltration and distribution of infiltrated rain water in an Australian Alfisol. Bromide was used as a tracer for the infiltrated rain under simulated rainfall conditions. The different infiltration patterns were then related to the macroporosity of the soils.

A 25-year-old permanent pasture was found to have the highest density (number per area) of macropores and percentages of transmitting macropores. A 9-year-old pasture phase in a pasture/crop rotation did not fully restore the macroporosity of the soil. Conventional cultivation by scarifying to 0.1 m for 4 years significantly reduced macropore density as well as continuity when compared with the pasture soil. The reduced macroporosity led to increased run-off by reducing preferential flow and altered the pathway of infiltrated water movement. As a consequence, the increase in water content below 0.1 m in the cultivated soil was predominantly from downward displacement of original soil solution, resulting in more leaching. The infiltrated rain water largely remained on the surface 0.1-m layer. In contrast, macroporosity found under direct drilling was similar to that of the pasture soil.     

Laboratory experiments on drought and runoff in blanket peat

Global warming might change the hydrology of upland blanket peats in Britain. We have therefore studied in laboratory experiments the impact of drought on peat from the North Pennines of the UK. Runoff was dominated by surface and near‐surface flow; flow decreased rapidly with depth and differed from one type of cover to another. Infiltration depended on the intensity of rain, and runoff responded rapidly to rain, with around 50% of rainwater emerging as overland flow. Drought changed the structure of the peat and the subsequent behaviour of the peat in response to rain. Surface runoff was reduced, infiltration increased and flow increased within the deeper peat layers. Old and new water produced from the peat during simulated storms was identified by bromide tracing; the amount of old mobile water flushed out of the top few centimetres was small and there was less from deeper peat layers. No significant difference in the old and new water mixing processes could be identified between the control plots and the drought treatment plots. Lissamine staining showed preferential bypass flow through macropores in the peat, though only in the top 5 cm. Following drought, however, macroporosity increased within the upper peat layers, and preferential flow extended deeper than in controls. Peat structure recovered somewhat after drought, but the effects of the drought were long‐lasting. If these effects extend to the field during drier summers then we can expect changes to the hydrology and associated chemistry of blanket peat catchments in the British uplands.     

北京昌平区农地土壤大孔隙特征

研究在利用染色示踪法对北京昌平区农地的优先流发生区进行判断的基础上,采用Photoshop软件和土壤水分穿透曲线对该农地的大孔隙数量与分布特征进行量化分析。结果表明:试验农地的土壤大孔隙半径主要集中在0.5～2.8mm之间,平均半径为0.695～0.711mm,大孔隙率为5.10%～22.06%。随着土壤深度的增加,染色区在土壤剖面上呈现出集中分布的特征,同时,染色面积比例逐渐减小。各土层染色区的稳定出流速率是未染色区的1.39～2.05倍,在大孔隙各孔径范围内,染色区的数量是未染色区的1.33～3.57倍。大孔隙的垂直分布表现出上层多、下层少的特点,其中半径小于1.5mm的孔隙占98%以上。染色区在大孔隙密度、大孔隙连通性上的优势能够使其更快地进行水分运输并更早达到稳定,因而也就更易成为优先流发生区。     

Seasonal dynamics of soil–water pressure in a cracking clay soil

This paper describes the results of an investigation of changes in soil water pressure head (ψ) and its relationship to the macropore network in a cracking clay soil. Four vertical nests, each consisting of three tensiometers positioned at depths of 30 cm, 60 cm and 90 cm below the surface, were monitored continuously over a two-year period to study changes in ψ. On one occasion an anionic tracer (Br-) was applied to investigate the extent of macropore flow. The results revealed considerable temporal variation in ψ with consistent variations between adjacent tensiometer nests. Variations in ψ indicated the seasonal development of a soil macropore system, followed by its subsequent decay and demonstrated the significant effect of rainfall intensity, duration and timing on percolation pathways. Differences in ψ were examined for individual summer rain events which were characterised by differences in precipitation amount and intensity. A total of 79 rain events extending across the period of study were analysed to assess the degree to which time-invariant parameters can be used to describe changes in ψ at a depth of 30 cm below the surface. The results indicated that individual regression models had considerable success in predicting ψ, although the residuals in the regression models were high for the specific case of large summer rain events, and in particular for three events.