Journalof SoilScience, 1988,39,639-648

The effect of sodium and calcium on physical properties and micromorphology of two red-brown

earth soils R. S. B. GREENE*, P. R E N G A S A M Y t , G. W. FORDS, C. J. CHARTRESg

& J. J. MILLAR**

*CSIRO Division of Wildlife & Ecology, Deniliquin, N.S. W. 2710, ?Institute for Irrigation and Salinity Research, Tatura, Victoria 3616, t Victorian Crops Research Institute, Horsham, Victoria

3400, gCSIR0 Division of Soils, Canberra, A.C.T. 2601, and **Bendigo College of Advanced Education, Bendigo, Victoria 3550, Australia

SUMMARY

The effect of treatment with either gypsum or sodium chloride on the saturated hydraulic conductivity (K,) of repacked soil columns and modulus of rupture (MOR) was studied on surface samples of two red-brown earth soils from SE wheat belt in Australia.

When the exchangeable sodium percentage (ESP) of the two soils was increased to > 80, K, was substantially reduced and MOR increased relative to the untreated soil; the values of the parameters were nearly equal for these pairs of high ESP soils. However, after treatment with gypsum the Raywood soil had a K, twice, and a MOR less than half, the corresponding values for the Glenloth soil.

Micromorphological and scanning electron microscope (SEM) observations suggest that the increase in K, following gypsum treatment is associated with an increase in visible macropores and reduced clay dispersion; Na treatment increased dispersion at the soil surface, with the clay particles forming an impermeable surface seal and illuviation argillans.

I N T R O D U C T I O N

Poor soil structure is a major problem on red-brown earth soils in the wheat belt of south-eastern Australia. During rain, surface soil aggregates slake and disperse, and on drying form a dense, hard crust which adversely affects soil properties. Gypsum is widely used to improve the physical properties of red-brown earth soils (Loveday, 198 1).

Gal et al. (1984), using a scanning electron microscope (SEM), studied the effect of ESP and gypsum treatment on the structure of soil crusts formed by rain. ESP observations were also used to explain the reduction in hydraulic conductivity (K,) that occurred when repacked columns of soil aggregates were leached with mixed NaCI-CaCI, solutions of low electrolyte concentration (Chen & Banin, 1974). Evans & Buol(I968) and Chen et a]. (1980), used photomicrographs and SEM images, respectively, to compare the properties of soil crusts.

However, none of the above experiments have involved a systematic study of the effects of different electrolyte treatments on soil microstructure and fabric, and related these effects to changes in soil physical properties. It is the aim of this paper to determine the effects of applying water with different electrolyte concentrations and Ca/Na balance and to compare the structural and physical properties of the surface crust formed on two red-brown earth soils. A combination of optical and SEM observations, and of physical measurements (modulus of rupture, hydraulic conductivity) were used. These physical properties were chosen because they are known to be sensitive indicators of changes in porosity in different soils (Loveday, 1974; Aylmore & Sills, 1982). The experiments were conducted on repacked soil columns under laboratory conditions.

639

640 R. S. B. Greene et al.

The soils used were both hard-setting red-brown earths (Stace et al., 1968) or Xeralfs (Soil Survey Staff, 1975). Although they have similar texture, the mineralogy differs: the clay fraction of the Glenloth soil is predominantly clay-mica, while the Raywood soil has mainly interstratified minerals (Greene & Ford, 1985). Chartres et al. (1985) reported that the effect on macroporosity of gypsum at 15 h ha-' was different at each site: the volume of macropores approximately doubled at Raywood, but there was only a minor increase at Glenloth. The differences were not related to changes in exchangeablecations at the two sites (Greene & Ford, 1985), but the contrasting effects of gypsum applications on macroporosity may be related to differences in clay mineralogy; this is discussed further in this paper.

E X P E R I M E N T A L M E T H O D S

Sites The soils at the two sites have been described by Greene & Ford (1985). Samples of the surface soil (CrlOcm) were taken in February 1979 from the control plots at each of the experimental sites, during the second wheat stubble phase of a pasture-fallow-wheat-wheat rotation. The soils were air dried and sieved to < 2 mm; some of the chemical and physical properties are given in Table 1.

The particle size analysis was obtained by suspending the soil in water using a high speed stirrer with sodium hexametaphosphate and sodium hydroxide as dispersing agents. A plummet balance was used to measure the amount of silt and clay. The amount of fine (2Cr200 lm) and coarse (20Cr2000pm) sand was measured using dry sieving. Organic carbon was determined using a modified Walkley Black method (McLeod, 1975). Exchangeable cations and cation exchange capacity were determined using a modification (Greene & Ford, 1985) of the leaching method described by Tucker (1974).

Saturated hydraulic conductivity measurements Duplicate 200 g samples of air dry soil (< 2 mm) were packed to a bulk density of 1.3 g cm-3 in Perspex permeameters (5.08 cm diameter, 12 cm high) to form a column of soil - 7.3 cm high. The soil was retained using a piece of nylon net supported by fine wire gauze at the bottom of the column, and a piece of nylon net was also placed on the soil surface to protect it from splash. The soil was allowed to wet slowly overnight by placing the permeameter into a beaker of distilled water. The saturated hydraulic conductivity was determined by maintaining a constant head of water above the soil surface and water percolation was measured (water treatment).

By-product gypsum ( <0.25 mm) was uniformly mixed with the soil at the equivalent of 15 t ha-' CaSO, 2H,O, and packed into the permeameter column: K, was again measured (calcium treat- ment). The concentration of calcium in the soil solution after leaching for 7 h was measured by analysing a sample of the effluent, using atomic absorption spectrophotometry. To study the effect of ESP > 80% on K,, 200 g of soil in a column was first leached with 250 ml of 1 .O M NaCI. After the excess NaCl solution had been leached out with distilled water, the hydraulic conductivity with distilled water was measured (sodium treatment).

Table 1. Selected chemical and physical properties of the surface 10 cm of the two red-brown earth soils studied

Property Glenloth Raywood

pH (1:5w/v soil-water) 7.0 7.2 Clay, < 2 pm (YO) 26.5 21.5 Silt, 2-20 pm (Y) 27.0 10.5 Fine sand, 2Cb-200 pm (Y) 42.5 49.0 Coarse sand, 20&2000 pm (YO) 3.5 17.0 Organic carbon (%) 0.72 1.08 ESP (Yo) 9.3 4.9

Efect of Nu and Cu on micromorphology 64 1

For all three treatments K, was measured after 7 h of leaching. This period was selected arbitrarily to avoid experimental difficulties caused by the K, of the sodium treatment continually declining with time, while there was still adequate concentrations of electrolyte to maintain floccu- lation and stable values of K, in the gypsum treatments.

After 24 h, the soils were drained, air-dried, crushed to < 2 mm and analysed for exchangeable cations. The spontaneous dispersion of the < 2 mm material following treatment with either gypsum or sodium chloride followed by water, was measureu (Rengasamy et al., 1984).

Preparation of samples for thin-section and SEM observation Further sets of soil samples from Glenloth and Raywood were leached in the way described with water, or with gypsum, or with sodium chloride followed by water. One set of treated columns were dried at <4O"C and impregnated with polyester resin; thin sections were cut from the upper 30 mm to examine crust formation at the soil surface. A second set of samples were air-dried and pieces (- 20 mm diam.) of the upper 10 mm of the soil crust were carefully broken from the columns. These pieces of the crust were glued to microscope stubs to show either the soil surface or a section normal to the soil crust. The samples were vacuum coated with gold and examined with a Cambridge S150 SEM at a range of magnifications.

Determination of modulus of rupture Modulus of rupture (MOR) determinations closely followed the method of Richards (1 953), except that the drying temperature used was 45°C (Aylmore & Sills, 1982). Samples were replicated six times. Soils of a particular cation status were prepared as above by leaching with either (i) water, (ii) gypsum or (iii) NaCl followed by water.

R E S U L T S

Saturated hydraulic conductivity Table 2 gives the exchangeable &ions, K,, MOR and dispersible clay for both soils after the three treatments. With water, K, for Glenloth was lower than for Raywood; both decreased with NaCl treatment, but with the gypsum treatment (= 15 t ha-') K , for the Raywood soil was about double

Table 2. Properties of surface soil samples after water, gypsum and NaCl treatments in permeameter columns. ESPis based on the sum ofcations. K, is mean of duplicate values (with YO variation of mean); MOR is mean (and

SE) of six determinations

Exchangeable cations (meq kg-'1

MOR Dispersed clay* Soil and ESP K, treatment Na+ K+ Ca2+ Mg2+ (%) (cmh-') (kPa) (Yo)

Glenloth Water 7.0 8.6 28.4 31.2 9.3 0.1 (20%) 211.2 (5.9) 0.8

NaCl 62.6 4.2 2.8 1.2 88.4 0.03(17%0) 385.8(24.5) 7.2 Gypsum 0 6.8 67.5 3.5 0 1.13 (8%) 136.8 (2.2) 0

Water 4.4 7.8 41.7 35.3 4.9 0.3 (lO%o) 97.7 (1.6) 0.2 Gypsum 0 5.9 75.3 5.2 0 2.18 (8%) 61.2 (1.2) 0

Raywood

NaCl 73.2 4.2 7.2 2.1 84.4 0.04(13%) 382.3(21.4) 8.7

*Rengasamy et al. (1984)

642 R. S. B. Greene et al.

that for the Glenloth soil. At 7 h, the measured concentrations of calcium in the leachate (Glenloth, 18.9 and Raywood 1 1.7 meq I-’) were sufficient to maintain the clay colloids in a flocculated state (Rengasamy et al., 1984).

Modulus of rupture, dispersible clay and exchangeable cations Following treatment with water, the Raywood soil had a lower MOR value, and less dispersible clay, than the corresponding soil from Glenloth (Table 2). After treatment with gypsum, the MOR of both soils decreased considerably with the Raywood soil still having the lower value; dispersible clay was reduced to an undetectable level for both soils. When the Glenloth and Raywood soils were leached with NaCl to give ESP > 8O%, their MOR’s increased to similar values (- 384 kPa), while the amount of dispersible clay in each increased dramatically to 7.2% and 8.7% at Glenloth and Raywood, respectively.

Micromorphology Thin Sections. Micromorphological observations on the Glenloth samples leached with water

did not show any indications of clay dispersion (e.g. void argillans). Nodules in the water-treated sample had sharp external boundaries. In the NaCl treated samples some clay nodules were noted to have diffuse boundaries indicative of some clay dispersion. A poorly oriented layer of clay on the upper surface of the sample was also observed. Illuviation ferri-argillans, both filling narrow (50pm) fissures and as menisci around the necks of larger pores, were also noted (Fig. I(a))

Fig. 1. (a) Glenloth, sodium treated; illuviation void ferri-argillans (arrowed) are visible in an in-filled void approximately 10 mm below the top of the column, indicating dispersion, translocation and deposition of clays. (b) Raywood, sodium treated; note a 6W30 pm thick crust formed by the dispersion and then sedimentation of clay is apparent on the upper surface of the column. (c) Raywood, water treated; note that margins of the central void (arrowed) observed 10 mm below the upper surface of the column are free of illuvial clay. (d) Raywood, calcium treated; again no visible evidence of clay movement or dispersion of microaggregates either at the column surface or within the matrix.

EfSect of Nu and Cu on micromorphology 643

throughout the upper 30 mm of the column sample. No surface clay layer, void ferri-argillans, or other evidence of clay dispersion was noted in the calcium treated samples.

In the NaCl treated Raywood samples, part of the surface was covered by a 6 6 8 0 pm thick layer of strongly oriented clay which contained some coarser particles (silt) towards its base (Fig. l(b)). Illuviation ferri-argillans, including one consisting of strongly oriented clay, > 300 pm thick, were common in fissures and vughs throughout the upper 30 mm of the column. No evidence of clay crusts or illuviation argillans was observed in either the water (Fig. l(c)) or gypsum (Fig. l(d)) treated samples.

SEM Observations. In Figs 2 and 3, SEM micrographs are given of the leached soils from Glenloth and Raywood, respectively. For each of the three leaching treatments, two micrographs depict the same area of the soil surface (at x 100 and x 1000 magnification) and one micrograph is of the soil in cross section (at x 1000 magnification). The micrographs are representative of the type of structures observed in extensive examination of 1620 mm2 of each surface and about 15 mm of the length of each section at magnifications up to x 5000.

Glenloth The differences in morphology of the Glenloth soil due to treatment with either water, gypsum or NaCl are marked. In the x 100 micrographs, silt and sand grains coated with clay particles are visible at the surface of soil treated with either water or gypsum. Preliminary measurements on the x 1000 micrographs indicate that the average diameter of clay aggregates in the gypsum treatment is larger (4.02 pm) than the water treatment (2.26 pm). Visual observations also suggest that the average pore size is larger in the gypsum treatment. In vertical section, the upper 56100 pm of both the gypsum and water leached soils consist of silt and sand grains coated with clay particles, and numerous large voids. Examination of several specimens indicates that the coating by clay is less continuous in the soil treated with water than with the gypsum treatment.

Treatment with NaCl and then water to remove the excess electrolyte, produces a completely different structure in this soil. At low magnification, the surface appears as a relatively flat, coherent layer with few pores (Fig. 2: compare (a)(i) and (b)(i) with (c)(i)). There are no silt or sand grains visible, and analysis of the surface by energy dispersive X-ray analysis confirms that it consists almost entirely of clay particles. These clay particles would have concentrated on the surface as a result of dispersion from soil aggregates (following NaCl treatment) and redeposition. At high magnification, these clay particles are seen to be arranged in layers mainly parallel to the soil surface. In section, the upper 10-20 pm of the soil is very dense, and the parallel alignment of the clay particles is clearly evident (Fig. 2(c)(iii)). The structure below the dispersed clay layer has similar structure to that of the calcium or water treated samples.

Raywood As with the Glenloth soil, the effect of NaCl treatment was very different from that ofeither gypsum or water. The upper surface of both the gypsum and the water leached soil consists of silt and sand grains coated with clay particles; this results in an open structure with common meso- and macro- pores (Figs 3(a)(i) and 3(b)(i)). There was evidence, again, that the clay coating was less continuous in the water treated than in the gypsum treated soil. In section, both the gypsum and water treated samples have a very open, well structured appearance.

The surface of the NaCl treated soil has a compact, coherent appearance, and even at x 1000 there are few visible pores (Fig. 3(c)(ii)). The ‘tide marks’ obvious in Figs. 3(c)(i) & (ii) are very fine clay particles which tend to remain in this pattern on drying. In section, the upper 1 6 2 0 pm of the surface is very similar to the Glenloth soil following NaCl treatment. The clay particles appear to have dispersed from the remainder of the soil matrix and formed a layered structure (crust) on the upper surface of the soil. The photomicrograph of the NaC1-treated Raywood soil (Fig. l(b)) also shows the layer of strongly oriented clay on the soil surface.

Fig

. 2. S

EM

mic

rogr

aphs

of

the

Gle

nlo

th so

il af

ter

trea

tmen

t wit

h (

a) c

alci

um

, (b) w

ater

, an

d (c

) sod

ium

. For

eac

h tr

eatm

ent,

mic

rogr

aphs

(i)

and

(ii)

dep

ict t

he s

oil

surf

ace

and

mic

rogr

aph

(iii

) is

of the s

oil i

n c

ross

-sec

tion

.

Fig.

3. S

EM m

icro

grap

hs of

the R

ayw

ood

soil

afte

r tre

atm

ent w

ith (a

) cal

cium

, (b) w

ater

, and

(c) s

odiu

m ch

lori

de. F

or ea

ch tr

eatm

ent, m

icro

grap

hs (i

) and

(ii)

depi

ct

the s

oil s

urfa

ce an

d m

icro

grap

h (iii

) the

soil

in cr

oss-

sect

ion;

note

the s

moo

th su

rfac

es in

c(i)

and

c(ii)

show

ing t

he cl

ay se

dim

enta

tion

as n

oted

in F

ig. l

(b).

646 R. S. B. Greene et al.

D I S C U S S I O N

In the experimental approach used, great care was taken to eliminate any effects arising from mechanical impact of water on the soil surface. Also, because the soil aggregates in the cylinders were allowed to wet slowly, the likelihood of any aggregate breakdown due to slaking would be minimal (Collis-George & Laryea, 1972). Therefore, the differences in physical properties of these soils following treatment with different electrolytes was attributed to varying degrees of physico- chemical dispersion of soil aggregates resulting from both their exchangeable cation composition and the nature and concentration of the electrolyte in the leaching solutions.

Furthermore, as a minimum ofmechanical disturbance was applied in the leaching treatments, it is appropriate to measure the spontaneous dispersion of the samples following the various leaching treatments. The K, and MOR values obtained after treatments clearly show the effect of the amount of spontaneously dispersible clay in controlling the physical properties of the Glenloth and Raywood soils.

The macroporosity of both soils is stabilized by treatment with gypsum, and the concentration of calcium electrolyte maintained in the soil solution prevents the disruption of aggregates and the occlusion of pores by dispersed clay particles. The K , values are, therefore, higher than their corresponding values following treatment with water. The stabilization of macroporosity (with fewer inter-aggregate contacts) that occurs in the gypsum treated soils also explains why they have a lower MOR compared with the water treated soils. Ford (1981) also showed that gypsum applied in the field caused a marked decrease in the crust strength of both soils. The changes in the composition of exchangeable cations in this field experiment (reported by Greene & Ford, 1985), are similar to those produced by treating columns of repacked soil with gypsum (Table 2).

Following treatment with sodium, the extreme sodicity (ESP > 80%) results in severe dispersion of clay particles from both soils, with similar reductions in K, and increases in MOR.

The changes in the two soil physical properties (K, and MOR) selected as indicators of porosity changes following treatment with NaCl and gypsum are consistent with the micromorphological changes seen in the optical and SEM micrographs. These changes in visible pores are assumed to reflect changes in total porosity. The SEM micrographs indicate the presence of larger, more open voids in the gypsum treated soils compared with the water treatment, particularly on the upper surfaces of the samples. However, since it is probable that most of the effect of gypsum is via a reduction in macroscopic swelling whilst the columns are saturated, it is not possible to demonstrate this effect in the SEM phot6graphs of the dried soil.

Both the SEM and thin section observations indicate that dispersed clay formed a partial surface layer in both the Glenloth and Raywood NaCl treated samples. This layer presumably develops as clay dispersed on wetting sediments out in the water maintained on the soil surface. These layers are probably similar to natural depositional layers of oriented clay observed in puddles and furrows at the field sites after rain, and once formed would considerably reduce the infiltration rate into the soil. ‘Washed-in’ layers as reported by McIntyre (1958), were not apparent in either the Glenloth or Raywood NaCl treated samples, although other evidence of clay dispersion was plentiful. This raises the question whether the so-called washed-in layer results from illuviation of dispersed clay, or may in fact be a depositional lamella of clay-rich material which has been subsequently buried. Other investigations of crusts (Chen et al., 1980; Norton et af., 1986) have similarly failed to find washed-in layers resulting from day illuviation.

Although the changes in morphological and physical properties induced by the different treat- ments were similar on both soils, their respective magnitude may well be the result of differences in initial packing in the permeameter columns, or the result of minor differences in organic matter and texture between the soils. However, Chartres et al. (1985) also noted that samples from field plots treated with and without gypsum exhibited differences in pore-size distributions, with Raywood developing a more open structure and greater macroporosity than Glenloth following gypsum treatment. The Raywood soil contains some smectite and its total clay content has a considerably higher surface area than the mica-rich clay at Glenloth (337.7 and 167.8 m2 g-’, respectively). Furthermore, spontaneous dispersion at Raywood is greater than at Glenloth upon treatment with high levels of sodium electrolyte. Consequently, it is suggested that the differences in magnitude of

Effect of Na and Ca on micromorphology 647

the effects of the water, gypsum and NaCl treatments on structural properties may be partially the result of differences in clay mineralogy between the sites.

There is another possible explanation for the difference in response: in soils where physico- chemical forces (double layer effects) are more important than the mechanical effects involved in consolidation and reorientation of clay particles, a change in either type or concentration of electrolyte in the pore fluid will drastically affect the void ratio (Olson & Mesri, 1970). From their consolidation experiments on clays, Olson & Mesri (1970) deduced that physico-chemical effects dominate in high surface area smectitic soils, whilst in sand, muscovite and kaolinite, which have low surface areas, mechanical effects are dominant. In illites (clay-micas), which have an intermediate surface area, physico-chemical and mechanical effects will both operate.

In the Raywood soil, where smectite is dominant, flocculating conditions would tend to cause the clay particles to move in groups (or domains) and produce macrovoids. In the Glenloth soil, however, which is dominant in clay-mica, additional mechanical effects would produce a different pattern of orientation and consolidation (Smart, 1975). Thus, the difference in clay mineralogy between the Raywood and Glenloth soils could partially account for the observed difference in void size distributions, following treatment with gypsum in the field, observed by Chartres et al. (1985). The possible differences in average pore size noted between the water and gypsum treatments at Glenloth may be a response to electrolyte affecting the dispersion and flocculation properties of the clay minerals present. Future measurements will aim to quantify these differences on the Glenloth and other soils.

A C K N O W L E D G E M E N T S

The authors gratefully acknowledge the assistance given by Professor L. A. G. Aylmore, Depart- ment of Soil Science, University of Western Australia and Dr D. Atkinson, Centre for Colloid Science, Swinburne Institute of Technology, in the measurements ofmodulus of rupture and surface area, respectively.

R E F E R E N C E S

AYLMORE, L.A.G. & SILLS, L.D. 1982. Characterisa- tion of soil structure and stability using modulus of rupture-exchangeable sodium ‘percentage relation- ships. Austrailian Journal of Soil Research 20,

CHARTRES, C.J., GREENE, R.S.B., FORD, G.W. & RENGASAMY, P. 1985. The effects of gypsum on macroporosity and crusting of red duplex soils. Australian Journal of Soil Research 23,467-79.

CHEN, Y., & BANIN, A. 1974. Scanning electron microscope (SEM) observations of soil structure. Changes induced by sodium-calcium exchange in relation to hydraulic conductivity. Soil Science 120, 428436.

CHEN, Y., TARCHITZKY, J., BROUWER, J., MORIN, J. & BANIN, A. 1980. Scanning electron microscope observations on soil crusts and their formation. Soil Science 130,49-55.

COLLIS-GEORGE, N. & LARYEA, K.B. 1972. An exam- ination of the wet aggregate analysis, the moisture characteristic, and infiltration-percolation methods of determining the stability of soil aggregates. Australian Journal of Soil Research 10,15-24.

EVANS, D.D. & BUOL, S.W. 1968. Micromorphologi- cal study of soil crusts. Soil Science Society of America Proceedings 32,19-22.

FORD, G.W. 1981. Some effects on soil properties of applying gypsum to two red duplex soils. 2. Development of surface crusts. Soil Management Conference, Dookie, May 1981, Australian Society of Soil Science-Victorian Branch, pp. 276-82.

2 13-24. *

GAL, M., ARCAN, L., SHAINBERG, I. & KEREN, R. 1984. Effect of exchangeable sodium and phosphogyp- sum on crust structure-scanning electron micro- scope observations. Soil Science Society of America Journal 48,872-878.

GREENE, R.S.B. & FORD, G.W. 1985. The effect of gypsum on cation exchange in two red duplex soils. Australian Journal of Soil Research 23,61-14.

LOVEDAY, J. 1974. Recognition of gypsum-responsive soils. Australian Journal of Soil Research 12, 87-96.

LOVEDAY, J. 1981. Soil Management and amelio- ration. In National Soils Conference 1980 Review Papers’ (Eds T. S . Abbott, C. A. Hawkins & P. G. Searle), pp. 39-57. Australian Society of Soil Science Incorporated: Sydney.

MCINTYRE, D.S. 1958. Permeability measurements of soil crusts formed by raindrop impact. Soil Science

MCLEOD, S. 1975. Studies on wet oxidation pro- cedures for the determination of organic carbon in soils. Notes on soil technique, 1973, CSIRO Australia, Division of soils.

NORTON, L.D., SCHROEDER, S.A. & MOLDENHALJER, W.C. 1986. Differences in surface crusting and soil loss as affected by tillage methods. In Assessment of Soil Surface Sealing and Crusting. Proceedings of the International Soil Science Society Symposium, Ghent, 1985 (Eds F. Callebaut, D. Gabriels & M. de Boodt), pp. 64-71. Flander’s Research Centre for Soil Erosion and Soil Conservation.

85,185-9.

648 R. S. B. Greene et al.

OLSON, R.E. & MESRI, G. 1970. Mechanisms control- ling compressibility of clays. P.A.S.C.E.96, SM6, 1863-1878.

RENGASAMY, P., GREENE, R.S.B., FORD, G.W. & MEHANNI, A.H. 1984. Identification of dispersive behaviour and the management of red-brown earths. Australian Journal of Soil Research 22, 41 3 4 3 1.

RICHARDS, L.A. 1953. Modulus of rupture as an index of crusting of soils. Soil Science Society of America Proceedings 17,321-3.

SMART, P. 1975. Soil Microstructure. Soil Science 119.385-93.

SOIL SURVEY STAFF. 1975. Soil Taxonomy. A basic system of soil classijkation for making and interpreting soil surveys. Agriculture Handbook No. 436, United States Department of Agriculture, Washington, DC.

STACE, H.C.T., HUBBLE, G.D., BREWER, R., NORTHCOTE, K.H., SLEEMAN, J.R., MULCAHY, M.J. &HALLSWORTH, E.G. 1968. Handbook ofAustralian Soils Rellim, Glenside, S.A.

TUCKER, B.M. 1974 Laboratory procedures for cation exchange measurements on soils. Technical Paper No. 23, CSIRO Australia, Division of soils.

(Received 12 February 1987)