| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
CSIRO Land and Water, GPO Box 1666, Canberra, ACT 2601, Australia
* Corresponding author (Hamish.Cresswell{at}csiro.au).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: EC, electrical conductivity • ESP, exchangeable sodium percentage • LSmod, modified linear shrinkage
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Campbell (1985) suggested the cause of the inaccuracy to be flow restriction within the soil sample once a narrow layer at the base of the sample has dewatered. Once the base has dewatered, a very small hydraulic conductivity can occur, preventing further drainage from the sample within a reasonable equilibration period. Campbell (1985) supported the theory through soil water simulation modeling based on Richards' equation. It was suggested that coarse-textured soil materials were especially susceptible to incomplete equilibration because of the very small unsaturated hydraulic conductivity at matric potentials attained in pressure plate apparatus. Loss of hydraulic contact between the soil sample and ceramic pressure plate, as sometimes caused by soil shrinkage on drying, is also known as a mechanism that can cause incomplete drainage. The blockage of pores in ceramic pressure plates by colloidal material or through biological growth or byproducts is a further mechanism with the potential to cause incomplete drainage of soil samples in pressure plate apparatus.
Madsen et al. (1986) evaluated the pressure plate method using samples from six Danish soil horizons ranging from 2 to 27% clay content (clay defined as <2-µm diameter). They used a thermocouple psychrometer to check the soil water potential after samples had apparently attained equilibrium in pressure plate apparatus at –1.5 MPa matric potential. They suggested that total water potential was consistently less negative (wetter) than the matric potential that would be attained (–1.5 MPa) if samples had properly equilibrated on the pressure plates.
Gee et al. (2002) also used a thermocouple psychrometer and analyzed three soils (a sand, a silt loam, and a clay) that had been on –1.5 MPa pressure plates for 10 d or longer. They found total water potential was always less negative (wetter) than –1.0 MPa, indicating hydraulic nonequilibrium. Following Campbell (1985), they predicted equilibrium times for samples on –1.5 MPa pressure plates using numerical modeling and suggested that equilibration time may extend to months or years, accentuated in coarse-textured soil, due to low plate conductance and decreasing soil hydraulic conductivity at high pressures. Gee et al. (2002) concluded that methods other than pressure plate apparatus might be required to measure equilibrium water contents at –1.5 MPa over reasonable time periods. Water contents of the soil samples when at apparent equilibrium on pressure plates were not reported or compared with water contents at –1.5 MPa as determined by thermocouple psychrometer.
The aim of this study was to evaluate the accuracy of pressure plate apparatus in measuring soil water retention at –0.5 and –1.5 MPa matric potential, in terms of both water potential and water content, on a variety of Australian soils. If proper equilibration was not reached with some soil materials, then the objective was to identify common characteristics of those soil materials to enable identification of existing data that might be inaccurate and to provide recommendations for soil water characteristic measurement so that future errors can be avoided.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
The measurements were repeated for eight replicates of each of the 35 soil samples at a matric potential of –1.5 MPa and again for three replicates at –0.5 MPa. Two of the replicates at –1.5 MPa were discarded after suspected evaporative water losses (leaving six acceptable replicates for analysis). A small number of other sample replicates were also omitted from the analysis where individual thermocouple psychrometer measurements of potential were obviously in error given the measured sample water content and the water content–matric potential results from other replicates of the same sample.
Subsequently, a subset of 14 samples (six replicates of each) were chosen for wetting with 0.01 mol L–1 CaCl2 solution before determining gravimetric soil water content after equilibration on pressure plates at –1.5 MPa. These 14 samples comprised four "control" samples not susceptible to dispersion and another 10 samples selected because they (i) were susceptible to dispersion as indicated by exchangeable Na percentage (ESP), or (ii) had not reached their equilibrium water content at –1.5 MPa when previously measured. Most samples with high modified linear shrinkage were not considered for the CaCl2 experiment because shrinkage away from the pressure plate was a more likely causal mechanism for incomplete sample equilibration in these instances.
Water content was also determined at –1.5 MPa matric potential for each sample as follows. Replicates of each soil sample were brought to different water potentials so that some replicates were drier than –1.5 MPa and some were wetter. This was achieved by using combinations of draining samples in pressure plate apparatus and air drying. Total water potential was determined for each replicate using a thermocouple psychrometer. The samples were then weighed and oven dried at 105°C for determination of gravimetric soil water content. Electrical conductivity was determined on each soil material, enabling soil matric potential to be calculated as described below. This process yielded multiple water content–matric potential data pairs covering a range of soil dryness. A subset of the data from the main pressure plate evaluation analysis was also included to increase replication. Any obviously inconsistent data pairs were removed before linear regression analysis was used to identify the interpolated water content at –1.5 MPa matric potential. With a small number of samples, nonlinear fitting and interpolation was more appropriate and therefore adopted. For most samples, eight water content–matric potential data pairs were used in the regression analysis. The same measurement and analysis procedure was repeated to determine the interpolated water content at –0.5 MPa matric potential for each sample.
These interpolated water contents at –1.5 or –0.5 MPa matric potential formed "target" water contents against which pressure plate apparatus results were evaluated. Soil samples apparently equilibrated on pressure plate apparatus had their water content compared with the "target" water content and differences were generally assumed to represent measurement error from the pressure plate method. Statistical significance of the differences between water contents of samples coming off pressure plates and the target water contents was assessed with t-tests using a weighted average of the sample variance and a 95% confidence level.
This experimental procedure enables soil samples apparently equilibrated to a set matric potential (–1.5 or –0.5 MPa) on pressure plate apparatus to be evaluated by comparing both water potential and water content actually attained against their respective "target" values.
Pressure Plate Operation
Standard commercial pressure plate apparatus (Soilmoisture Equipment Corp., Goleta, CA) were used to apply the specified pressures to the soil materials. New pressure plates were used at the commencement of this experiment; the plates were soaked in distilled water overnight before being loaded with soil samples. The plates were covered with around 1-mm depth of water before loading the air-dry soil. No separate contact material was used. Each soil sample was spooned into a Perspex ring 19 mm in diameter and 10 mm deep. Samples were wet on the ceramic plates to achieve good contact between the soil and plate. During wetting, the plates were loosely covered with a polythene sheet to reduce evaporation. The samples were allowed to wet up for 3 to 4 h before being placed in the pressure chamber. Only one plate (replication) was placed in the pressure chamber at a time and each plate was pressurized for 5 to 6 d. Separate tests were undertaken to measure the water content attained by replicated soil samples after different periods of equilibration on pressure plates at –1.5 MPa. Results confirmed that drainage of water from soil samples of these dimensions had become close to apparent equilibrium after 5 d. Although complete cessation of drainage might not have been reached, it is considered that leaving samples on plates for much longer periods is likely to result in less accurate measurement overall due to potential evaporation loss and biological growth in the soil, water, and ceramic plates. No weights were applied to soil samples inside the pressure plate apparatus, nor were the samples covered with paper toweling or any other sheeting. Plates were cleaned after each use by soaking in distilled water, gently brushing and rinsing, and then allowing them to air dry. No microbial inhibiters or chemical cleaning agents were used.
A sample of a standard reference soil was included on all pressure plate measurement replicates, as is regular practice in this laboratory. Where measured water contents at –0.5 or –1.5 MPa deviated significantly from the well-established mean value for the standard, then data from the other samples on the plate were discarded and the cause of the error investigated. There was no temporal trend in results from the standard sample during this experiment, suggesting no significant ceramic plate degradation.
Thermocouple Psychrometry
Thermocouple psychrometers determine the total water potential of the liquid phase of a soil sample through measurement of the relative humidity of the equilibrium vapor phase. Total water potential (
t) is the sum of osmotic potential (o) and matric potential (m). Total soil water potential is related to the relative humidity through use of the following equation (Rawlins, 1966):
 | [1] |
where t is total water potential, R is the gas constant, T is temperature (Kelvin), M is the molecular weight of water, and hr is the relative humidity of water vapor equilibrated with the soil water (R/M 
0.461 MPa K–1).
The commercial Peltier-type thermocouple psychrometer used in this study (Tru-psi SC-10X, Decagon Devices, Pullman, WA) was calibrated using salt solutions of known water potential. Careful temperature control was observed, as is necessary for accurate psychrometer operation (Rawlins and Campbell, 1986). Each soil subsample for psychrometer measurement was collected directly from a soil sample drained to equilibrium on a pressure plate. A thin-walled cork borer slightly smaller in diameter than the psychrometer sample cup was pressed into the soil sample on the pressure plate, taking a vertical subsample to include any water content gradient that might be present. A rubber plunger was then used to push the soil plug into the psychrometer sample cup. A conical depression was then formed with a packing cone on the top of the sample as consistent with procedure for operating the psychrometer. Three sandy samples were not sufficiently cohesive to maintain the conical depression and these cups were flat-filled to a level just clear of the thermocouple junction. The sample cups were immediately placed in the sample changer or capped and held in a humid chamber. Despite operating in a constant-temperature room, it was found necessary to leave the cups in the sample changer for at least 1 h before the commencement of measurement. The required equilibration time was thought related to water redistribution in the soil after it has been pressed into the cup.
Following total water potential determination with the psychrometer, the water content of each sample was determined by oven drying the soil in the sample cup for at least 24 h at 105°C. Rawlins and Campbell (1986) suggested that the practical maximum precision obtainable with Peltier psychrometers under ideal laboratory conditions is 3 to 4 kPa. They noted that the measurement of total water potential to within 10 kPa would require temperature differences between the reference junction of the psychrometer and the liquid phase of the soil samples to be controlled to within <10–3 °C.
Electrical Conductivity Determination
To directly compare pressure plate apparatus and the thermocouple psychrometer, it is necessary to separate the total water potential determined by the psychrometer into its components. The osmotic potential of the soil solution (o) can be approximated from the electrical conductivity (EC) of the saturation extract as (Rawlins and Campbell, 1986)
 | [2] |
 | [3] |
where os is the osmotic potential of the saturated extract (kPa), 
and s are the soil water content and the water content at saturation, and EC has units of dS m–1. Once o is known, then m is determined as the difference between total and osmotic potentials. This assumes an ideal solution, and ignores anion exclusion and the precipitation of salts with low solubility.
Soil samples were added to a quantity of distilled water on the pressure plate as part of the laboratory procedure described above. The saturated soil was then drained, so much of the salt in the soil was probably leached out. Thus using the bulk soil EC was not appropriate for estimation of the osmotic potential. Electrical conductivity had to be determined on the actual subsamples used within the psychrometer. After samples from the psychrometer had been oven dried and weighed following total potential determination, they were shaken in 5 mL of distilled water for 1 h. The EC of the solution was determined with a Radiometer conductivity meter (Radiometer A/S, Copenhagen) and multiplied by –36 following Eq. [3]. The product was then multiplied by the ratio of the solution water content to the water content at the time of water potential measurement to estimate the osmotic potential of the sample.
Supporting Measurements
Attributes of the soil materials in Samples 1 to 30 and Sample 35 (Table 1) in this study had been previously determined as follows. Particle size distribution was measured using a Sedigraph 5100 particle size analyzer (Micrometrics Instrument Corp., Londonderry, NH) following ultrasonic dispersion and chemical pretreatment where necessary (Bowman and Hutka, 2002) to remove organic matter, soluble salts, and cementing agents. Soil shrinkage was determined using the modified linear shrinkage (LSmod) method of McKenzie et al. (1994), which uses sieved rather than remolded soil (McGarry, 2002, Method 518.03). Exchangeable cations and cation exchange capacity were determined using a 1 mol L–1 NH4Cl extracting solution at pH 7.0 with a pretreatment for soluble salts (Rayment and Higginson, 1992, Method 15A2). Total organic C was determined volumetrically using a high-frequency induction furnace following Method 6B2 in Rayment and Higginson (1992).
Soil attributes for Samples 31 to 34 (Table 1) were measured using different techniques. Particle size distribution was determined using the plummet balance technique described by McIntyre and Loveday (1974). Organic C was determined by wet oxidation using the method attributed to Walkley and Black, without any attempt to correct the values to total organic C (Rayment and Higginson, 1992, Method 6A1).
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The overall performance of the sandy soils was good, with Samples 17, 18, and 27 (clay contents of <1, 2, and 8%, respectively) showing very little error in water content determination. These samples had very small –1.5 MPa water contents (<1.9%) and the majority of the water was expected to drain at comparatively large matric potentials. One of the sands (Sample 1) did not reach equilibrium. The reason for this error could include dispersion of the small amount of clay in this sodic material, and restricted drainage after initial dewatering because of very small unsaturated hydraulic conductivity. This sample was probably more prone to error in target water content determination than some others, the volume of the air cavity in the psychrometer sample cup was harder to keep consistent, there was a large osmotic potential correction, and the matric potential–water content relation was very steep.
There was a small number of soil materials (e.g., Samples 23, 29, and 30) giving errors in water content on pressure plates but without initially obvious causal mechanisms. Although not having high ESPs, the performance of these samples was shown later to respond to the use of CaCl2 for reducing dispersion.
When the replicates of 35 soil materials were removed from the pressure plates at apparent equilibrium (of –1.5 MPa), they were found to have actually attained an average matric potential of –1.10 MPa (Table 2). Eight samples had matric potentials equal to or less negative than –1.00 MPa (i.e., an error of 0.5 MPa or greater). Some of those samples not equilibrating close to –1.5 MPa are those with the larger errors in water content as discussed above. Others have small water content errors and just reflect very steep water retention curves around –1.5 MPa. In these cases, small changes (errors) in water content correspond to large changes in matric potential (e.g., Sample 17).
Considering just the matric potential discrepancy between the pressure plate method and the thermocouple psychrometer method is potentially misleading. For example, here some samples had <0.05% error in "equilibrium" water content off pressure plates, but that corresponded to large deviations from the target –1.5 MPa matric potential (e.g., Samples 4 and 7). In practice, draining a sample to within 0.05% of its water content at –1.5 MPa is a very good result and it is unlikely that greater accuracy is routinely achieved with vapor equilibrium methods.
Analysis of Samples off Pressure Plates at –0.5 MPa
Soil samples on pressure plates at –0.5 MPa equilibrated, on average, to 0.2% wetter than the target water content at –0.5 MPa matric potential as determined by thermocouple psychrometer (Table 3
). Of the 35 samples, 29 had water contents that were not significantly different (t-test, 95% confidence interval) from the target values. Eleven of these samples were either <0.05% wetter or else slightly drier than the target water contents. There were, however, seven samples with errors 
0.4% ; although not statistically significant at the 95% level, these are still of concern.
|
One soil material (Sample 31) was >1.0% wetter than the target water content and four samples were between 0.5 and 1.0% (inclusive) wetter. In general, the soil materials giving the larger equilibrium water content errors at –0.5 MPa were also samples with larger than average errors at –1.5 MPa (e.g., Samples 26, 29, 30, and 34, although there were exceptions).
Soil samples on pressure plates at –0.5 MPa attained an average matric potential of –0.48 MPa as determined using a thermocouple psychrometer. Of those, out of the 35 soil materials that were wetter than target, nine samples had errors between 0.10 and 0.19 MPa, and two samples had errors 0.20 MPa (Samples 20 and 30). Again, some of these samples had not fully drained while others were very close to the target water content but have steep water retention curves. Ten samples had mean matric potentials less than –0.50 MPa, that is, were apparently too dry (giving a matric potential error >0). Nine of these 10 samples had water content errors <0.1%; however, the corresponding matric potential errors were large in some cases, often in sands (e.g., Samples 1, 17, 18).
The mechanisms discussed above as contributing to incomplete drainage of soil samples on pressure plates at –1.5 MPa are diminished in their contribution to measurement error at –0.5 MPa pressure, as could be expected, but still present. Soil materials with LSmod of 5% or above, and Samples 31, 33, and 34, which are classified as Vertosols in the Australian soil classification (Isbell, 2002) but don't have LSmod data available, all showed incomplete drainage. There is little evidence of coarse-textured soils and soils with high ESP having reduced drainage in this instance. The performance of the sandy soils was good, with Samples 17, 18, and 27 (clay contents of <1, 2, and 8%, respectively) again showing very little error in water content determination.
Soil Dispersion and Equilibration on Pressure Plates
The results in Table 4
confirm the hypothesis that addition of CaCl2 would reduce soil dispersion and any associated mechanism impeding sample equilibration (e.g., blocking of pores in the ceramic pressure plates). The four "control" samples (Samples 8, 9, 12, and 14) returned water contents off the pressure plates that matched the target water contents determined by thermocouple psychrometer (at –1.5 MPa matric potential) and were nearly identical to the previous (distilled water) results. Of the 10 remaining samples, all drained to smaller water contents than they had achieved previously when they had been wetted with distilled water. Average water content error of the 10 noncontrol samples was 0.07% when wetted with CaCl2 solution, compared with 0.52% when previously measured after wetting with distilled water. Four samples equilibrated to water contents less than the target, illustrating measurement error associated with the thermocouple psychrometer method. Of the six samples wetter than the target, two had errors significantly different at the 95% level. Five samples were 0.2 or 0.3% wetter, and one sample was 0.4% wetter than the target water contents.
|
General Discussion
Loss of soil–plate contact due to soil shrinkage on pressure plates remains a difficulty. The previous experience in this laboratory is that loading samples on pressure plates or using secondary contact material has not been very successful. Gee et al. (2002) came to the same conclusion after a systematic assessment of both techniques.
Previous modeling (Campbell, 1985; Gee et al., 2002) has supported the assertion that coarse-textured soil materials are particularly susceptible to nonequilibrium error due to low unsaturated hydraulic conductivity restricting the drainage of water from samples on pressure plates. Our results suggest good drainage from the majority of samples with <10% clay content at –0.5 and –1.5 MPa matric potential. The apparent discrepancy is probably reflecting the emphasis placed on reaching equilibrium water potentials, rather than water contents, in the earlier studies. In this study, water contents were measured, thus highlighting the very small water contents and steep water retention curves around –1.5 MPa matric potential in coarse-textured soil. These soil materials release the majority of stored water at large (less negative) matric potentials. Although equilibrium matric potential is not always attained on pressure plates, the error in water content was shown to be very small in most of the coarse-textured soils analyzed here.
In this study, soil samples were left 5 to 6 d to equilibrate in pressure plate apparatus. For our approximately 10-mm-high samples, this was sufficient time for most outflow to have ceased and for soil water contents in the nonswelling and nondispersing samples to approach the –1.5 MPa water content as determined by thermocouple psychrometer. Equilibrium time requirements are strongly influenced by the height of the samples (time for reaching equilibrium is proportional to the square of the height of the sample) and the nature of the soil materials. There are differing views in the literature regarding what constitutes adequate equilibration time. McIntyre (1974) reported equilibration times of 4 to 7 d for 7- to 10-mm-deep samples (depth of material prewetting) on pressure plates at –1.5 MPa. Klute (1986) wrote that establishment of equilibrium on pressure plates requires 1 to 7 d and that for core samples of 2 to 3 cm high, an equilibration time of 2 to 3 d had been found sufficient. Burke et al. (1986) indicated that equilibration time is usually from 2 to 14 d, again depending on sample height, soil material, and imposed pressures. Gee et al. (2002) observed that 15-mm-high soil samples on –1.5 MPa pressure plates had not reached equilibrium matric suctions after 10 d. Equilibrium times have been estimated, for example, by fitting outflow volume data to an exponential function. Topp et al. (1993) used this approach to estimate equilibration times of up to 25 d for 76-mm-high soil cores.
It appears that complete cessation of drainage may take weeks or longer in some circumstances. The usual laboratory pressure plate procedure is to observe water outflow in a burette and when the rate becomes too low to measure in a reasonable time, assume the samples adequately approximate equilibrium. This is a pragmatic approach given the usually exponential relation between outflow volume and time. That the nonswelling samples in this study attained water contents after 5 to 6 d close to equilibrium water contents as determined by thermocouple psychrometer reflects: (i) that we used samples only 10 mm high, and (ii) that the outflow volume (and corresponding change in soil water content of samples on the plates) that typically occurs in the latter part of the outflow process (i.e., in the days leading up to complete cessation of drainage) is small.
The use of samples higher than 10 mm is likely to accentuate problems of nonequilibrium on pressure plates at –1.5 MPa. Leaving samples for long time periods on pressure plates in an effort to get more complete equilibrium increases the chances of other errors, such as sample desiccation through vapor exchange with the air inside the chamber. Any air leaks will accentuate this. We recommend using soil samples no more than 10 mm high for determination of water content at –1.5 MPa, which is consistent with the procedure described by Dane and Hopmans (2002).
The significance of error in the measurement of soil hydraulic properties is dependent on the required accuracy or precision for the intended application. It appears that the error in equilibrium water contents determined on pressure plate apparatus increases as matric potential decreases (becomes more negative). A dry-end error in the soil water characteristic curve is unlikely to translate into significant analysis errors in some applications. For example, the prediction of unsaturated hydraulic conductivity using statistical pore models (e.g., Childs and Collis-George, 1950; Jackson, 1972; Mualem, 1976) is much more sensitive to error at the wet end of the soil water characteristic input or in the near-saturated hydraulic conductivity match point used with these methods (Stephens and Rehfeldt, 1985; van Genuchten and Nielsen, 1985; Cresswell, 1992). We suggest that analysis of sensitivity to soil hydraulic property inputs is a good way to determine the practical significance of errors in soil hydraulic property measurement in specific applications (e.g., Cresswell and Paydar, 2000).
Where the use of pressure plate apparatus with particular soils does not provide sufficient accuracy in measurement of the dry end of the soil water characteristic, then the primary alternatives are vapor equilibrium methods such as the thermocouple psychrometer. Use of a thermocouple psychrometer also necessitates the measurement of the EC of the soil sample on which total soil water potential is determined, and the subsequent calculation of osmotic potential. Osmotic potential was significant for some of the soil materials analyzed here (Table 2) and high EC is common in Australian soils. If a water content measurement is desired at a specific matric potential (e.g., –1.5 MPa to approximate the drained lower limit) then multiple measurements of total soil water potential and conversion to matric potential are required together with an interpolation procedure. These processes are time consuming and give the potential for additional experimental error. Furthermore, soil samples used with thermocouple psychrometers are often small in comparison with those used on pressure plate apparatus, necessitating more replicates for representative description of a particular soil material.
In this analysis, the thermocouple psychrometer was adopted as the "benchmark." The assumption was made that both direct matric potential determination by thermocouple psychrometer, and interpolated water contents at –1.5 and –0.5 MPa using psychrometer-measured potentials, are accurate benchmarks for assessing the accuracy of pressure plate apparatus. Measurement with the psychrometer, however, and the associated osmotic potential correction is clearly not error free. On average, standard errors for the interpolated target water contents at –1.5 or –0.5 MPa were 0.3% (w/w), approximately double the standard error associated with measurement of water content for the (larger) samples coming off pressure plates. Consistent with this are some samples coming off pressure plates at –0.5 MPa, or else at –1.5 MPa after CaCl2 wetting (Tables 3 and 4), with water contents less than their target values. Hence some of the apparent error being attributed to pressure plate apparatus, and interactions between the pressure plate method and specific soil attributes (e.g., texture, linear shrinkage, and dispersion), is in fact due to error in the use or performance of the thermocouple psychrometer method.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Swelling soil materials are not likely to equilibrate fully on pressure plates at either –0.5 or –1.5 MPa pressure. It seems likely that drainage is impeded through loss of soil–plate contact as samples shrink on drying. This mechanism correlated especially with soils exhibiting modified linear shrinkage of 5% or more. It is recommended that vapor equilibrium methods of soil water characteristic measurement be used with these soils.
Existing pressure plate data from swelling soils or soils prone to dispersion that have not been treated with CaCl2 or CaSO4 should be used with caution.
In coarse-textured soils, equilibrium matric potential was not always attained, probably because of impeded drainage due to very low unsaturated hydraulic conductivity. The corresponding error in water content observed is very small, however, reflecting the slope of the soil water characteristic. Error of this nature and magnitude will not be of practical concern for many applications.
For nonswelling soil materials, and providing that CaCl2 or CaSO4 is used to control dispersion, water contents of samples removed from pressure plates at –1.5 MPa were consistent with water contents at –1.5 MPa as determined using a thermocouple psychrometer. That is, pressure plate apparatus provide good estimates of "true" –1.5 MPa water content in such soil materials.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication May 10, 2006.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Bittelli and M. Flury Errors in Water Retention Curves Determined with Pressure Plates Soil Sci. Soc. Am. J., July 14, 2009; 73(5): 1453 - 1460. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
.