Soil Science Society of America Journal 64:1603-1608 (2000)
© 2000 Soil Science Society of America
DIVISION S-1-SOIL PHYSICS
Water Holding Capacity of Ironstone Gravel in a Typic Plinthoxeralf in Southeast Australia
Joost Brouwera and
Heather Andersonb
a Dep. of Plant Science, Wageningen Univ. and Research Centre, Haarweg 333, 6709 RZ Wageningen, The Netherlands, present address: Brouwer Environmental and Agricultural Consultancy, Wildekamp 32, 6721 JD Bennekom, The Netherlands
b Dep. of Primary Industry, P.O. Box 219, Maryborough, Qld 4650, Australia
brouwbar{at}bos.nl
 |
ABSTRACT
|
|---|
Water retention by coarse fragments in the soil is often ignored in agronomic and water balance studies. Following the calculation of inexplicably high water retention by the fine earth fraction, water contents at -20 and -1500 kPa and apparent bulk density were determined for remnant pisolithic ironstone gravel samples isolated from soils on the Dundas Tableland in southeast Australia. The volumetric water content of the ironstone gravel at -1500 kPa was found to vary between 0.12 and 0.24 m3 m-3, while at -20 kPa it varied between 0.16 and 0.36 m3 m-3. Available water holding capacity (AWHC) of the ironstone gravel varied between 0.03 and 0.15 m3 m-3. Both the AWHC and the water content at -1500 kPa of the ironstone gravel showed significant increases with depth. Magnetic ironstone gravel, found almost exclusively in the A and E horizons, was much denser than nonmagnetic ironstone gravel (average 3.38 vs. 2.64 Mg m-3), but had similar water retention characteristics. Ignoring the water retention characteristics of the ironstone gravel would have led to overestimation of the AWHC of the bulk soil by a factor 1.08 to 1.67 for various horizons. For the combined top 1.0 m of the soil, ignoring the water held by the ironstone gravel would have led to an estimated AWHC of 162 mm, while in fact it was only 129 mm. Water balance studies of soils with ironstone gravel clearly need to take into account the water holding characteristics of that gravel.
Abbreviations: AWHC, available water holding capacity
|
INTRODUCTION
|
|---|
AVAILABLE WATER HOLDING CAPACITY is an important agronomic and hydrologic characteristic of soils. It expresses how much water can be stored in the soil for plants to use during periods without rain or irrigation. This gives an indication of the drought sensitivity of soils, a key criterion in the selection of suitable crop species and varieties. The AWHC can also be used to calculate the likelihood of deep drainage or groundwater recharge taking place. This is important where there is, for instance, concern about groundwater level or quality.
During the past 50 yr a number of studies have shown that coarse rock fragments can hold considerable volumes of water; such fragments, or indeed the weathering bedrock itself, should not be ignored when calculating the AWHC of a stony soil (e.g., Coile, 1953; Berger, 1976; Hanson and Blevins, 1979; Arkley, 1981; Flint and Childs, 1984; Montagne et al., 1992; Jones and Graham, 1993; Ugolini et al., 1996; Oyonarte et al., 1998).
While the studies mentioned were concerned with pieces of fresh and weathering rock, under intermittent saturated conditions coarse fragments in the soil can also be formed by Fe cementation of fine earth material. The ferruginous nodules or pisolithic ironstone gravel that can thus be formed are quite commonly found in hydromorphic soils and Paleosols of the tropics and subtropics, in horizons of a range of textures (Tardy, 1992).
To show that the water retention properties of ironstone gravel, too, can appreciably affect the AWHC of soils, we present data from a field study in western Victoria, Australia. Correlations are also shown to exist between AWHC on the one hand and the depth of occurrence and magnetic properties of the ironstone gravel on the other hand. To the best of our knowledge, somewhat similar data have previously only been published in a francophone journal from West Africa (Boa, 1990, 1993; Assa and Boa, 1990), based largely on an unpublished report (Boa, 1983). The francophone journal has a very limited distribution and is not included by abstracting services. In addition, Boa's approach is quite different from ours, and, for instance, does not examine correlations between AWHC and magnetic properties and depth of occurrence of the ironstone gravel. We only happened to hear about Boa's publications when the senior author moved to West Africa, well after our own measurements had been made.
There are a number of definitions of AWHC as determined in the laboratory. In this note we define the AWHC of the soil as the water being held by an undisturbed soil core at tensions between -20 and -1500 kPa, in cubic meters per cubic meter of soil core. Similarly, the AWHC of the sieved out coarse fragments in the soil is the water being held by those fragments at tensions between -20 and -1500 kPa, in cubic meters per cubic meter of coarse fragment.
|
Materials and Methods
|
|---|
Site Description
The data we present here were collected during a dryland salinity-related project at Gatum in western Victoria, Australia. Gatum lies on the Dundas Tableland 300 km west of Melbourne and 30 km north of Hamilton, at 37°27' S, 141°58' E. The Dundas Tableland, 200 to 300 m above sea level, consists of old paeneplains with varying underlying geology. These peneplains have been deeply weathered and lateritized probably in the early Pliocene, domed and tilted during the late Pliocene or early Pleistocene, and then reincised (Spencer Jones, 1965). In what remains of the old lateritic profiles, yellow duplex soils (Northcote, 1979), Yellow Dermosols and Brown Chromosols (Isbell, 1996), or Typic Plinthoxeralfs (Soil Survey Staff, 1996) are now developing. Selected properties of a representative profile on a broad crest are presented in Table 1
. The entire toposequence is described in Brouwer and Fitzpatrick (2000). The large amount of remnant pisolithic ironstone gravel these soils contain (see also Gibbons and Downes, 1964) is mostly between 6 to 20 mm in size. In our study, the ironstone gravel contents of individual samples varied from 5 to 27% by volume and the average ironstone gravel contents per depth ranged from 6.7 to 27% (Table 2)
.
View this table:
[in this window]
[in a new window]
|
Table 2 Average and standard deviations of characteristics related to water retention of ironstone gravel, fine earth, and entire 75-mm-high soil cores from Gatum, western Victoria, Australia
|
|
|
Methods
|
|---|
During the course of the project undisturbed soil cores were collected from different depths for neutron moisture meter calibration and for the determination of water retention curves. The cores were collected by first carefully removing the overburden without disturbing the layer to be sampled, and then driving down into the soil steel rings 75 mm high and 67 mm across (total volume 265 cm-3). The rings had a leading edge beveled outwards to avoid guiding soil from outside the targeted sample into the ring. If the ring did not enter the soil with relative ease because it hit a piece of ironstone gravel, the sample was discarded and a new sample taken approximately 20 cm away in the same horizontal plane. The sampling depths mentioned in the tables correspond with the depths of the center of the 75-mm-high cores.
The soil water retention curves of the undisturbed cores were determined on a suction table for potentials down to -20 kPa. Although in the determination of AWHC the upper limit can vary between -10 and -33 kPa, the limit of -20 kPa was the standard used in the laboratories of the Land Protection Division in Victoria; it is the limit used here in the calculation of AWHC values.
Water contents at -1500 kPa were determined in a pressure chamber as described by Klute (1986)(p. 648), using the <2-mm fraction fine earth. The fine earth fraction was isolated by grinding the air-dry peds from the undisturbed cores with a mortar and pestle and then using a 2-mm sieve to separate the fine earth from the ironstone gravel. For a pressure potential of -1500 kPa, this can be an acceptable method for determining soil water content (Elrick and Tanner, 1955, as cited in Klute, 1986).
Water contents at -20 kPa were calculated on the basis of the oven-dry weight of the whole core minus the oven-dry weight (not the moist weight) of the ironstone gravel. In this standard calculation (see Klute, 1986), it is implicitly assumed that the ironstone gravel holds no water at -1500 kPa. This means that any water the ironstone gravel does hold at -1500 kPa is wrongly assumed to have been held by the fine earth at between -20 and -1500 kPa. In our case, in samples with large amounts of ironstone gravel (up to 27% by volume), this method of calculation resulted in inexplicably high AWHC values for the fine earth fraction. We therefore decided to separately measure the water content at -1500 kPa of some of the ironstone gravel that had not yet been oven dried. For each of the layers near the surface, ironstone gravel from three different cores was used for this purpose. At 0.58 m depth, ironstone gravel from only one core was available, and at 0.83-m depth ironstone gravel from only two cores was available.
The ironstone gravel was scrubbed to remove adhering soil, soaked in water at room temperature for 3 d, put in the pressure chamber at -1500 kPa until equilibrium water content had been reached, weighed, oven dried, and weighed again. Dry bulk densities were subsequently determined by rapid immersion of the oven-dried ironstone gravel samples of known weight into water in a graduated cylinder. The accuracy of these simple bulk density determinations can be affected by water entering the pores of the ironstone gravel. The bulk density determinations were therefore checked against the pore volume of the ironstone gravel, known from saturated water content determinations, and the specific densities of the constituent minerals of the ironstone gravel: if the bulk density determination is precise, it will equal (specific density) x (1 - fractional pore volume).
Because the water contents of the ironstone gravel at -1500 kPa were all found to be very high, it was decided to carry out further water retention determinations on the same ironstone gravel samples, this time separated into magnetic and nonmagnetic fractions. Water contents at -20 kPa were determined by embedding known weights of ironstone gravel in known amounts of diatomaceous earth of high uniformity and with known water retention characteristics. The uniformly fine pore system of the packed diatomaceous earth ensures good water supply to, and even withdrawal of water from, the whole surface of the ironstone gravel. The samples with the ironstone gravel and the diatomaceous earth were first saturated for several days (initially by capillary action to minimize air entrapment) and then put on the suction table at -20 kPa until equilibrium was reached. The samples were then weighed. Water contents (kg kg-1) of the ironstone gravel samples were calculated using the known water content of the diatomaceous earth at -20 kPa. Water contents at -1500 kPa and apparent bulk density of the magnetic and nonmagnetic ironstone gravel fractions were then determined using a pressure chamber, oven, and graduated cylinder as described above.
|
Results
|
|---|
The results of the initial water retention determinations are presented in Table 2. Striking are the high water contents of the ironstone gravel at -1500 kPa, namely 0.18-0.24 m3 m-3. In addition, the apparent bulk density of the ironstone gravel was much higher in the top 0.30 m of the soil than lower down (3.09 vs. 2.55 Mg m-3, P < 0.001). The ironstone gravel content of the soil increased with depth to 0.6 m. The volumetric water content of the ironstone gravel at -1500 kPa was much less above 0.5 m (average 0.18 m3 m-3) than below it (average 0.24 m3 m-3; P < 0.01). The water content at -1500 kPa of the fine earth also increased with depth. The latter is most likely related to the parallel increase in clay content of the soil (see the texture classifications in Table 1).
The results of the second set of water retention determinations are presented in Table 3
. The magnetic and nonmagnetic ironstone gravels showed clear differences in apparent bulk density (P < 0.01) and in distribution with depth: magnetic ironstone gravel was found almost exclusively in the A and E horizons, where it constituted up to one-half of the total ironstone gravel fraction. The high apparent bulk density of the magnetic ironstone gravel is the reason for the high average bulk density of 3.1 Mg m-3 for the combined ironstone gravel in the A and E horizons (0.0–0.30 m), compared with only 2.5 to 2.6 Mg m-3 in the B horizon (0.3 to >0.9 m): see Table 2. The AWHC values of both magnetic and nonmagnetic ironstone gravel varied between 0.03 and 0.15 m3 m-3.
View this table:
[in this window]
[in a new window]
|
Table 3 Average and standard deviations of properties of nonmagnetic and magnetic ironstone gravels in the soil profile at Gatum, western Victoria, Australia
|
|
|
Discussion
|
|---|
All 12 ironstone gravel samples showed a considerable AWHC, both for the nonmagnetic gravel fraction and, where present, for the magnetic fraction. This already suggests that the AWHC of ironstone gravel should not be ignored or dismissed out of hand.
The accuracy of the results of the water retention calculations on a volume basis depend of course on the accuracy of the apparent bulk density determinations for the ironstone gravel. The apparent bulk densities can be checked against total pore volume and the specific densities of the minerals present in the ironstone gravel. In the ironstone gravel, with almost no large pores, total accessible pore volume will be somewhat, but not much, greater than volumetric water content at -20 kPa. On the basis of visual inspection of scanning electron microscope images of a number of pieces of ironstone gravel from the same site, we have estimated a conversion factor of 1.1. The volumetric water contents at -20 kPa of 0.16 to 0.36 m3 m-3 (Table 3) thus convert to total pore volumes of 0.17 to 0.40. The average apparent dry bulk densities of the nonmagnetic ironstone gravel of 2.3 to 2.9 Mg m-3 then convert to specific densities of 3.3 to 4.2 Mg m-3. This is very similar to the specific density of 3.6 to 4.0 Mg m-3 of goethite, the dominant cementing Fe mineral in the nonmagnetic ironstone gravel from Gatum (R.W. Fitzpatrick, 1995, unpublished data). In the same way, the apparent bulk density of the magnetic ironstone gravel of 3.4 Mg m-3 and the volumetric water contents at -20 kPa of 0.13 to 0.15 m3 m-3 (Table 3) convert to specific densities of 4.2 to 4.8 Mg m-3. This is close to the specific density of 4.9 to 5.3 Mg m-3 of maghemite and hematite, the dominant Fe minerals in the magnetic ironstone gravel, which also contains some lighter goethite (R.W. Fitzpatrick, 1995, unpublished data). Assa and Boa (1990) also reported apparent bulk densities of up to 2.9 for ironstone gravel from Ivory Coast, without distinguishing between magnetic and nonmagnetic ironstone gravel.
The increase with depth in water content of the ironstone gravel at both -20 and -1500 kPa suggests an increase in porosity with depth. Similar increases in ironstone gravel porosity with depth can be inferred from the data of Tiessen et al. (1991): ironstone gravel taken by them from >0.65-m depth at a site in Ghana had an effective cation exchange capacity that was twice as high as that of ironstone gravel taken from the 0.3- to 0.4-m depth: 32 vs. 16 mmol kg-1. Mineralogy being the same (kaolinite dominated), a greater effective cation exchange capacity indicates a greater reactive surface and thus arguably a greater porosity of the ironstone gravel from greater depth.
The data in Table 2 show that, when calculating AWHC for the gravelly soils, ignoring the water held in the ironstone gravel at -1500 kPa led to overestimation of the AWHC by 0.012 to 0.066 m3 m-3. Put differently, the real AWHC of the bulk soil (including ironstone gravel) in individual soil layers, was only 60 to 93% of the AWHC calculated while ignoring the water held by the ironstone gravel at -1500 kPa. As mentioned above, the latter calculations are based on the incorrect, but commonly held, assumption that the water still held by the ironstone gravel at -1500 kPa or less had been held by the fine earth fraction at between -20 and -1500 kPa. Using the data in the 13th column of Table 2, the real AWHC of the top 1.0 m of an average profile at our field site in Gatum is estimated to be 
129 mm. Ignoring the water held by the ironstone gravel at less than -1500 kPa would have led one to assume that it was 162 mm (15th column of Table 2).
The data in Table 3 show that the AWHC values of both the magnetic and nonmagnetic ironstone gravels varied between 0.03 and 0.15 m3 m-3. Boa (1983), studying ironstone gravels from Ivory Coast, also measured AWHC values of up to 0.15 m3 m-3 (and volumetric water contents at saturation of up to 0.34 m3 m-3). This is similar to the AWHC values reported for rock fragments by Hanson and Blevins (1979), Flint and Childs (1984), Montagne (1992), and Jones and Graham (1993). Given the formation process of ironstone gravel (very slow precipitation, initially as bridges between particles in the soil, often followed by the formation of a weathered rind on the outside), significant porosity of ironstone gravel is not unexpected. However, one would expect the porosity of the ironstone gravel to be less than the porosity of the fine earth it is formed in. Comparison of the AWHC values for whole cores in Table 2 with those for the ironstone gravel only in Table 3 and comparison of the water contents of gravel and fine earth at -1500 kPa indicate that the ironstone gravel is indeed less porous than the fine earth.
The results of the two determinations of volumetric water content of ironstone gravel at -1500 kPa (Tables 2 and 3) show good agreement. That the water contents measured the second time are slightly lower than those measured the first time, particularly for samples from surface horizons, may be due to partial rubbing off of the weathered rind of the ironstone gravel during repeated handling. The second time the ironstone gravel had also already been oven dried once. The changes in bulk density and water content of the ironstone gravel with depth indicate that not all ironstone gravel is the same: differences in genesis, mineralogy, and weathering can cause differences in bulk density and water retention. This is also demonstrated by the differences in bulk density and water retention between the magnetic and nonmagnetic ironstone gravel fractions. Similar observations about the effects of genesis on ironstone gravel water holding characteristics were also made by Boa (1990) about ironstone gravels from Ivory Coast.
|
Conclusions
|
|---|
Clearly, ironstone gravel can retain significant amounts of water (0.03–0.15 m3 m-3) at potentials at which plants can take it up (-20 to -1500 kPa) and at potentials at which it is not available to plants (0.17–0.24 m3 m-3 at less than -1500 kPa). Magnetic and nonmagnetic ironstone gravel had similar water retention characteristics. However, both AWHC and water content at -1500 kPa were greater for ironstone gravel from below 0.5 m than for ironstone gravel from the top 0.5 m.
In traditional determinations of soil water retention curves, where ironstone gravel is assumed not to hold any water at all, the water actually held by the ironstone gravel at less than -1500 kPa is wrongly assumed to be held by the fine earth fraction at -20 to -1500 kPa. If there are significant amounts of ironstone gravel present in the soil, these wrong assumptions can lead to considerable overestimation of the AWHC of the bulk soil. In our own most extreme example, a horizon with 27% ironstone gravel by volume, the actual AWHC was 0.097 m3 m-3. Ignoring the water held by the ironstone gravel at less than -1500 kPa would have led to the conclusion that the AWHC was 0.163 m3 m-3, an overestimation by 67% (factor 1.67). Similarly, the real AWHC of the top 1.0 m of an average profile was calculated to be 129 mm. Ignoring the water held by the ironstone gravel at less than -1500 kPa would have led one to estimate that it was 162 mm. This difference of 33 mm in water storage capacity can greatly affect perceived options for improving land management at our field site in the Dundas Tableland area (average rainfall 630 mm yr-1): severe dryland salinity problems at the site are thought to be caused by an increase in annual groundwater recharge of only 20 to 25 mm.
The above facts demonstrate that when soils containing ironstone gravel are present, the water retention characteristics of such ironstone gravel should not be ignored in field studies or simulation modeling of the water balance. In land evaluation studies, one should also not automatically assume that high ironstone gravel contents will have a negative effect on crop growth because of reduced water availability.
It seems unlikely that changing the upper tension limit for the AWHC determination from -20 to -10 or -33 kPa would have led to substantially different conclusions.
|
ACKNOWLEDGMENTS
|
|---|
Robert van de Graaff initiated the Gatum project on the property of Neil and Sue Lawrance. Bruce Trebilcock, Phil Miles, and Ian Dreher assisted with the collection of the samples. Anne Jackman, Sonia Procter, and Ursula Piedrzak assisted with the water content determinations, on which David Cummings and Rob Fitpatrick gave advice. Soil profile information was provided by the CSIRO Division of Soils. Funding for the project came from the National Soil Conservation Programme, the Australian Wool Corporation, and the Salinity Programme of the Government of Victoria. The library staff at the Department of Conservation, Forests and Lands, at the CSIRO Division of Soils and at Alterra-DLO assisted with the references. Ray Isbell, Robert van de Graaff, Dr. A. Rogowski, and two anonymous referees provided much appreciated comments on drafts of this paper. To all these our sincere thanks.
|
NOTES
|
|---|
This study was carried out while the two authors were staff members of the Land Protection Division, Dep. of Conservation and Natural Resources, Victoria, Australia.
Received for publication July 23, 1999.
|
REFERENCES
|
|---|
- Arkley R.J. Soil moisture use by mixed conifer forest in a summer-dry climate. Soil Sci. Soc. Am. J. 1981;45:423-427.[Abstract/Free Full Text]
- Assa, A., and D. Boa. 1990. Caractéristiques porales des gravillons des sols ferralitiques du nord-ouest de la Cote d'Ivoire. [Pore characteristics of gravel from ferralitic soils in north-west Ivory Coast]. Agronomie Africaine (Int. J. Ivorian Assoc. for Agric. Sci.) 2:95–104.
- Berger E. Partitioning the parameters of stony soils, important in moisture determinations into their constituents. Plant Soil 1976;44:201-207.
- Boa, D. 1983. Caractéristiques hydriques des gravillons ferrugineux dans les sols ferralitiques. [Water retention characteristics of ironstone gravel in ferralitic soils]. ORSTOM, Abidjan (Ivory Coast) and Paris.
- Boa, D. 1990. Caractérisation morphologique des gravillons ferrugineux sur le bassin versant de Booro-Borotou. [Morphological characterisation of ironstone gravel from the Booro-Borotou catchments]. cAgronomie Africaine (Int. J. Ivorian Assoc. for Agric. Sci.) 2:83–94.
- Boa, D. 1993. Modèle de conversion du taux ponderal des gravillons en teneur volumique dans les sols gravillonnaires. [Model for the conversion of gravimetric water content to volumetric content in gravelly soils] Agronomie Africaine (Int. J. Ivorian Assoc. for Agric. Sci.) 5:145–153.
- Brouwer J., Fitzpatrick R.W. Combining the structural approach and the Australian Soil Classification in the interpretation of soil morphological features along a salt-affected toposequence in western Victoria: methods, profile data and a discussion of results. Glen Osmond, SA, Australia: CSIRO Division of Land and Water, 2000 Tech. Rep. 54..
- Coile T.S. Moisture content of small stone in soil. Soil Sci. 1953;75:203-207.
- Flint A.L., Childs S. Physical properties of rock fragments and their effect on available water in skeletal soils. In: Box J.E., Jr., ed. Erosion and productivity of soils containing rock fragments. Madison, WI: SSSA, 1984:91-103 SSSA Spec. Publ. 13..
- Gibbons F.R., Downes R.G. A study of the land in south-western Victoria. Kew, VIC, Australia: Soil Conservation Authority, 1964.
- Hanson C.T., Blevins R.L. Soil water in coarse fragments. Soil Sci. Soc. Am. J. 1979;43:819-820.[Abstract/Free Full Text]
- Isbell R.F. The Australian Soil Classification. Collingwood, VIC, Australia: CSIRO Publ, 1996.
- Jones D.P., Graham R.C. Water-holding characteristics of weathered granitic rock in chaparral and forest ecosystems. Soil Sci. Soc. Am. J. 1993;57:256-261.[ISI]
- Klute, A. 1986. Water retention: laboratory methods. p. 635–662. In Methods of soil analysis. Part 1. Agron. Monogr. 9. 2nd ed. ASA and SSSA, Madison, WI.
- Montagne C., Rudell J., Ferguson H. Water retention of soft siltstone fragments in a Ustic Torriorthent, central Montana. Soil Sci. Soc. Am. J. 1992;56:555-557.[Abstract/Free Full Text]
- Northcote K.H. A factual key to the recognition of Australian soils. Glenside, SA, Australia: Rellim Technical Publ, 1979.
- Oyonarte C., Escoriza I., Delgado R., Pinto V., Delgado G. Water-retention capacity in fine earth and gravel fractions of semiarid mediterranean montane soils. Arid Soil Res. Rehabil. 1998;12:29-45.
- Soil Survey Staff. Keys to soil taxonomy, 7th ed Washington, DC: USDA, 1996.
- Spencer Jones, D. 1965. The geology and structure of the Grampians area. Geol. Survey Vic. Memoir 25. Melbourne, VIC, Australia.
- Tardy Y. Diversity and terminology of lateritic profiles. In: Martini I.P., Chesworth W., eds. Weathering, soils and paleosols. Developments in earth surface processes 2. Amsterdam, the Netherlands: Elsevier, 1992:379-405.
- Tiessen H., Frossard E., Mermut A.R., Nyamekye A.L. Phosphorus sorption and properties of ferruginous nodules from semi-arid soils from Ghana and Brazil. Geoderma 1991;48:373-389.
- Ugolini F.C., Corti G., Agnelli A., Piccardi F. Mineralogical, physical, and chemical properties of rock fragments in soil. Soil Sci. 1996;161:521-542.
This article has been cited by other articles:
|
|
|
|
 
T. K. Tokunaga, T. K. Tokunaga, K. R. Olson, and J. Wan
Moisture Characteristics of Hanford Gravels: Bulk, Grain-Surface, and Intragranular Components
Vadose Zone J.,
August 1, 2003;
2(3):
322 - 329.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
