Relationships between fabric, water retention, and strength of hard subsoils in the south of Western Australia.
Crop yield on the sandy soils of the Western Australian wheatbelt is influenced strongly by the plant-available water (PAW) and strength of subsoils. High soil strength may limit root growth and access to potentially available moisture. This research investigated mechanisms of hardening of subsoils in the wheatbelt of south-western Australia. In this region, soils have mostly formed in a dissected lateritic landscape following aeolian, colluvial, and alluvial redistribution of materials (McArthur 1991). The intact lateritic profile was mostly formed by weathering of granitoids and other rock types and these profiles commonly consist of sand and pisolitic gravels over lateritic duricrust (cemented ferruginous material), mottled and pallid zones (quartz and clay rich materials), saprolite and saprock (isovolumetrically weathered materials with retention of parent rock fabric), and rock (Anand and Paine 2002).
Erosion and dissection of the lateritic profile and removal of most of the upper ferruginous horizons created a gently undulating landscape (Finkl and Churchward 1973) with present day soils having formed from the various in situ or transported lateritic materials (McArthur 1991) (Figs 1, 2). The sand fraction of soils consists mostly of angular quartz except where materials have been derived from ancient sandstones or coastal dunes where the grains are rounded. Some clay may be present in soils as round and hard sand-sized clay aggregates known as spherites which have been deposited by wind after being derived from lacustrine sediments (Killigrew and Glassford 1976). Panayiotopoulos (1989) concluded that the size, shape, and roundness of sand particles affect the ability of a material to be compressed, thus affecting soil strength. The extent of clay bridging between quartz grains and cementing by iron oxides and amorphous silica may also affect soil strength (Mullins and Panayiotopoulos 1984). Strong subsoils are abundant in south-western Australia and their management for cereal cropping involves ripping. Yield increases in cereal crops of up to 30% have been achieved by ripping (Jarvis et al. 1986).
The nature of strong subsoils formed in sedimentary materials will be compared with in situ granitic saprolite that occurs as a subsoil in many locations. The effects of fabric on water retention and strength will be investigated.
Materials and methods
Study area and sampling
Strong subsoils, which included some with hard traffic pans, were sampled from 7 sites near Binnu, Yuna, Pindar, Mingenew, Buntine, Kojonup, and Grass Patch in the wheatbelt of Western Australia (Fig. 2). Soils at these locations are generally classified as Kandosols with massive or weakly structured B horizons that lack texture contrast, or Chromosols with strong texture contrast between A and B horizons (Isbell 1996). Rainfall at these locations ranges from 300 to 400 mm with most received in winter (June, July, August). The average daily minimum and maximum temperatures are 9 to 15[degrees]C and 21 to 27[degrees]C, respectively (Bureau of Meterology 2008). In situ saprolite samples were collected from laterite profiles in the Darling Range of Western Australia (Kew and Gilkes 2006).
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
Definitions of saprolite and subsoil
Saprolite is 'weathered bedrock in which the fabric of the parent rock, originally expressed by the arrangement of primary mineral constituents of the rock, is retained by the alteration products. Alteration has been essentially isovolumetric' (Eggleton 2001). For the purpose of this research, strong subsoil is defined as material that has formed in colluvial, alluvial, or aeolian deposits which has been compacted or indurated in situ within a soil profile (Northcote 1974).
Water retention and unconfined compression strength measurements
Intact clods measuring 0.30 m in height and length and 0.20 m in thickness were removed from soil profiles and were used to prepare intact cubes (8 [cm.sup.3]) for the measurement of water retention and unconfined compression strength at matric potentials of -10, -30, -60, -100, -300, -600, and -1500 kPa (Kew and Gilkes 2006).
Particle size analysis and fabric description
The pipette method was used to determine the particle size of the <2 mm fraction of soil samples. Organic matter was removed and samples dispersed using the method of McKenzie et al. (2002). Grain size analysis of the sand fraction was by mechanical sieving with screens spaced at 1/2 phi intervals. Graphic mean ([M.sub.z]), inclusive graphic standard deviation ([[sigma].sub.I]), inclusive graphic skewness ([Sk.sub.I]), and graphic kurtosis ([K.sub.G]) were determined (Folk 1974).
Polished thin sections were prepared of resin-impregnated soil. The fabric and clay matrix distribution were described using optical microscopy and the terminology of Bullock et al. (1985). A Jeol 6400 scanning electron microscope operating at 15 kV with a 5 n[Angstrom] beam current was used to map the spatial distribution of elements and analyse selected points on each thin section using energy dispersive X-ray spectrometry (EDS). Scanning electron microscope (SEM) micrographs were analysed using Image-J software (Rasband 1999) to determine the shape of sand grains and to develop a descriptive fabric classification system.
Results and discussion
Representative saprolite and subsoil fabrics
The fabric of in situ saprolite (Fig. 3) provides explanations for the water retention and strength characteristics of this material (Fig. 4). Chemical weathering of plagioclase, alkali feldspar, and biotite has formed kaolin which mostly occurs as an anisotropic matrix which pseudomorphs primary minerals (Fig. 3a). The isotropic kaolin matrix contains grains of resistant primary minerals and particularly quartz (Fig. 3b). Non-isovolumetric weathering or pedoplasmation has also occurred in some saprolite, forming new fabrics at the expense of parent rock fabric (Anand and Paine 2002). The separated but matching adjacent grain boundaries in fractured quartz and residual alkali feldspar grains in saprolite are indicative of minor pedoplasmation. For less-altered saprolite, the kaolin formed from plagioclase fills fractures and voids between and within residual alkali feldspar and quartz grains (Kew et al. 2008) (Fig. 3c). The potassium (K) element maps show the distribution of residual alkali feldspar grains. The clay in fractures (between dashed lines in Fig. 3d) and micrometre-size pores is usually stained with iron oxides and devoid of quartz and other minerals. Saprolite that has undergone substantial pedoplasmation and clay illuviation may contain an increased concentration of disaggregated, angular quartz grains (Fig. 3e). Transport over micrometre distances does not result in physical alteration of resistant primary mineral grains so that quartz grains are not rounded or spherical.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
The hard subsoils investigated in this research are from soils that overlie in situ regolith mostly derived from igneous rocks (Trendall 1975), but the rounded and spherical quartz grains present in the subsoils reflect a history of fluvial and acolian transport and abrasion. The particle size sorting of quartz grains and clay distribution differ substantially from that of saprolite and also differ between subsoils (Fig. 4). The rounded, 0.1-0.5 mm quartz grains in subsoils from Yuna and Binnu (Fig. 4a, c) contrast with the large (up to I mm), angular quartz grains in the subsoil from Buntine (Fig. 4b). The distribution of kaolin clay is indicated by the aluminium (Al) maps and shows that much clay exists as partial (Fig. 4a) to complete (Fig. 4c) coatings on some of the quartz grains.
Further evidence of the addition of aeolian material to these soils is the presence of rounded kaolin spherites (Killigrew and Glassford 1976) in the subsoils from Grass Patch, Binnu, and Buntine (Fig. 5). Spherites may be stained throughout with iron oxides and appear bright in SEM micrographs (Fig. 5a) or be stained on outer surfaces only (Fig. 5b). They may be composed of concentric rings of clay with imbedded micrometre-size quartz grains (Fig. 5c) and they are usually of similar size to associated fine quartz grains (Fig. 5d). The presence of spherites at depths ranging to 0.80m may indicate the depth of aeolian deposition but they may also have been mixed through the profile by pedoturbation. Spherites do not occur in saprolite (Fig. 3), which is consistent with the findings of Killigrew and Glassford (1976). Those authors proposed that spherites are formed by aeolian mobilisation of sediments in drying lakes. However, Stace et al. (1968) considered that spherites in yellow sandplain soils near Merriden, Western Australia, formed in Tertiary laterites and had experienced minor colluvial transport.
[FIGURE 5 OMITTED]
Classification of fabric
A classification scheme for subsoil and saprolite fabrics that describes the arrangement of clay matrix and quartz grains was developed from 2D scanning electron micrographs and analyses using Image J software (Fig. 6, Table 1). The clay present in representative subsoils has been shaded black and the automated enhanced analysis confirms the highly varied distribution of clay matrix (Fig. 6a-c). A grain-support fabric characterised by quartz-to-quartz contact with minor amounts of clay coating or bridging between quartz grains is shown for a sandy loam sample at a depth of 0.10-0.30m on aeolian sandplain (Northcote et al. 1967) at Yuna (Fig. 6a). This latter material has been subjected to extensive aeolian transport, which has removed most of the free clay, and can be contrasted with a clayey sand subsoil from Buntine (0.10-0.60 m) (Fig. 6b). This material exhibits more extensive clay bridging between quartz grains, which are angular and similar to those of in situ granitic saprolite from the Darling Range (Fig. 3). A partial clay matrix support fabric (Fig. 6c) with rounded quartz grains is present in sandy loam subsoil from Binnu (0.80-1.60m) on the same sandplain as the Yuna sample, which had a grain-support fabric (Fig. 6a). In contrast to the dominant grain support fabric of subsoil materials, all saprolite samples (Fig. 6d-f) examined have clay matrix support fabrics. These observations of grain morphology are important for developing an explanation of soil strength. The roundness and sphericity of sand grains influences the strength of sandy-textured soils (Panayiotopoulos 1989). Henderson et al. (1988) found that penetration resistance of sandplain subsoils in Western Australia increased with density of packing. The diverse subsoil fabrics described above, which are sandy with sand grains of various shapes and sizes, would be expected to exhibit a wide range of strengths. The classification of subsoil fabric shown in Table 1 may be used to infer subsoil strength. Grain-support fabrics represented by classes 1 and 2 may have lower strength than clay-support fabrics represented by class 5, where strong capillary forces operate within the clay matrix (Gill and Alonso 2002).
The soil fabric classification was compared to the coarse fine size distribution terminology developed by Bullock et al. (1985). This latter system does not set limits on the abundances of quartz, clay, or void, which may lead to some ambiguity in allocating a sample to a class. Subsoils in class 1 and 2 (Table 1) have a monic-gerfuric-chitonic coarse-fine distribution, class 3 and 4 (Table 1) have a chitonic enaulic distribution, and class 5 (Table 1) has an enaulic porphyric distribution. Saprolite samples are all class 5 with an enaulic-porphyric distribution.
The strength of subsoils may be affected by the porosity of the clay that coats grains and forms structural bridges between quartz grains of varied size and shape. The internal porosity of the clay of both saprolite and subsoil can be determined from the concentration of chlorine from the impregnating resin (epichlorohydrin [C.sub.3][H.sub.5]OCl) within pores in the clay (Kew and Gilkes 2007). Non-saline regolith contains almost no chlorine; thus, the chlorine in resin in pores within the clay determined by microprobe (EMPA) analysis provides a direct measure of porosity. Signal loss during EMPA of porous samples due to the presence of resin-filled pores (Tretyakov et al. 1998) results in a total oxide value <100%. Thus, the total oxide weight percentage also provides a measure of clay matrix porosity.
[FIGURE 6 OMITTED]
The clay matrix in subsoils developed in sedimentary materials is mostly denser than for saprolite, the former subsoils having lower chlorine concentrations in the resin-impregnated matrix and a higher total oxide weight % (Fig. 7). Thus, the clay crystals are more densely packed in subsoil clay matrix than in saprolite. The mostly optically isotropic kaolin matrix of saprolite formed by alteration of feldspars to kaolin occupies the extent of the former feldspar grains (Kew et al. 2008); thus, a highly porous clay matrix is formed (Fig. 3b). The degree of porosity depends on the composition of the parent feldspar grain, with plagioclase producing a dense clay matrix and alkali feldspar a highly porous clay matrix. These differences reflect the Al concentrations in the feldspars and the retention of all Al during alteration. The dense clay matrix of subsoil is a consequence of pedogenic processes that have transported and packed clay crystals including eluviation and shrink-swell forces due to dry-wet conditions (Mullins and Panayiotopoulos 1984).
[FIGURE 7 OMITTED]
There are large differences in water retention by subsoils and saprolite at all matric potentials as shown by the water content v. pF relationships (Fig. 8). The fabric class (Table 1) of subsoils is predictive of these relationships. The PAW held in pores 30-0.2 [micro]m is similar for subsoil and saprolite and ranges from 0.12 to 0.16 [m.sup.3]/[m.sup.3]. The clay matrix of subsoils has been shown to be dense and less porous than saprolite but porosity in the PAW range is not reduced. Minor increases in PAW for class 1 and 2 subsoil (0.12 [m.sup.3]/[m.sup.3]) and clay-rich saprolite (0.16 [m.sup.3]/[m.sup.3]) may be partly due to micron-size voids present between poorly sorted, generally rounded quartz grains in subsoils and pores between structural ped faces in clay-rich saprolite.
Unconfined compression strength
The unconfined compression strength (UCS) of representative subsoils in transported materials and saprolite samples is responsive to changes in water content, with all materials becoming weaker as water content increases (Fig. 9a). The relationship is mostly well described by a negative exponential function with a conceptual dry strength coefficient A (UCS extrapolated to a water content of 0.0 [m.sup.3]/[m.sup.3]). The dry strength of subsoil classes 1 & 2 and 3 & 4 is much lower than fur saprolite, with subsoil class 5 being similar in dry strength to quartz-rich saprolite (Table 2). Responsiveness to water content is represented by the exponential coefficient B (Table 2) and there is no systematic difference between subsoil classes 3 & 4 and 5 and saprolite, for these averaged data for which values range from -10 to -20. The grain-supported fabric class 1 and 2 materials from Binnu shows much less response to changes in water content with a near zero coefficient B (-0.07). The relationship of UCS to matric potential is described by a positive exponential function, with subsoil and saprolite becoming stronger as matric potential increases (Fig. 9b). Wet strength (coefficient C, which is the extrapolated value of UCS at near saturation, pF 0) is much higher for both quartz-rich and clay-rich saprolite than for the various classes of subsoils in sedimentary materials (Fig. 9b). There is no systematic difference in the responsiveness of strength to matric potential (coefficient D, Table 2) between subsoil and saprolite samples.
[FIGURE 8 OMITTED]
Preservation of parent rock fabric in saprolite due to in situ isovolumetric weathering (Nahon 1991) may be responsible for the higher values of dry and wet strength. However, subsoils (not class 1 and 2) and saprolite are similarly responsive to changes in water content and matrix potential (coefficients B and D, Table 2). This result may be interpreted as indicating that the strength of the clay matrix in these materials is similarly affected by differences in water content.
Dry strength (A) increased with clay content for diverse in situ regolith materials and there was a wide range of dry strengths at any particular clay content (Kew and Gilkes 2006). Subsoils with quite diverse fabrics also show an increase in dry strength with clay content (Fig. 10a). There is no systematic relationship between wet strength and clay content for either subsoil or saprolite (Fig. 10b) as was also observed for lateritic regolith by Kew and Gilkes (2006), who suggested that the strength of very wet materials is not sensitive to texture. Possibly fabric type, grain size, or sorting and cementing by iron oxide or silica may have stronger influences on wet strength than does clay content.
Graphical and statistical analysis of subsoils
Graphical (Folk 1974) and statistical analysis of particle size data were used to determine particle size and shape properties that may affect the UCS of sandy subsoils. The data used in the analyses were derived from both dry sieving and image analysis of SEM micrographs (Table 2). The strong subsoil from Kojonup with a medium clay texture (52% clay) and matrix support fabric was excluded from the analysis. Remaining samples were loamy sand to clay loam textures (20% clay maximum) with grain support fabrics.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
Analysis of grain size data using Folk (1974) parameters based on the phi size scale for subsoils on transported materials shows much variation between the samples that contain very fine to medium sand ([M.sub.z]) which is poorly to very poorly sorted ([[sigma].sub.I]), coarse skewed ([Sk.sub.I]), and leptokurtic to platykurtic ([K.sub.G]). Factor analysis identified a strong positive relationship between graphic mean size in phi units ([M.sub.z]), oxide weight %, and UCS at matric potentials of -10 kPa (wet) and -600 kPa (dry) (Fig. 11). These fine materials (with high [M.sub.z] in phi units and high clay density as indicated by microprobe oxide wt % values) are associated with high UCS values. The distribution of samples within the factor diagram shows substantial variation between samples from the same location. Thus, sample 7 from Binnu, which has a high clay %, fine to very fine grain size, and high oxide weight % for its clay matrix indicating a dense clay matrix, is located far away in Fig. 11 from samples 6 and 8, also from Binnu, which contain medium-sized grains, low clay %, low oxide weight % for the clay matrix, but particles better sorted with a strongly peaked particle size distribution. Other Binnu samples (1, 2, 9) occupy intermediate positions in the factor plot.
Many subsoils in the wheatbelt of Western Australia have formed from aeolian or other sediments. They consequently have a distinctive morphology compared with those of subsoils that are saprolite formed by in situ isovolumetric weathering. Transported materials contain rounded quartz grains, while those in saprolite are angular. Aeolian transport of clay has formed rounded clay aggregates or spherites that occur in subsoils. In situ saprolite has not been transported but may have been subjected to some illuviation and mixing (pedoplasmation); however, the parent rock fabric is generally retained and quartz grains have not become rounded. Scanning electron microscopy and electron microprobe analysis has shown that transported subsoil has a less porous clay matrix than saprolite, possibly due to closer alignment of clay crystals. These morphological differences can be classified and are reflected in differences in water retention and strength between subsoils in transported materials and saprolite. Subsoils in transported materials have lower dry and wet strengths than saprolite and the water retention at all matric potentials is lower; however, the clay matrix of both types of material is equally responsive to changes in water content and matrix potential. The size, shape, and degree of sorting of quartz grains, the distribution of dense clay matrix, and cementing by iron oxides or amorphous silica may combine to affect the strength of subsoil materials. These factors help to explain why adjacent subsoils of similar texture may nevertheless differ significantly in their suitability (in their natural state) for crop production.
Cereal cropping and pasture production have been improved when dense, strong subsoils have been ripped. Fracturing the subsoil allows plant roots to access water and nutrients that are otherwise unattainable. The longevity of ripping in transported subsoils is expected to be reduced by further trafficking of farm vehicles. However, subsoils formed in saprolite may contain fragile pseudomorphs of primary minerals resulting in minor fabric alteration by ripping. The addition of gypsum or lime during ripping may improve the longevity of ripping in subsoil materials that are sodic. This study has shown that subsoils in saprolite are quite different to those in transported material, so that field investigations of ripping and other subsoil amelioration practices requires that the subsoil type should be clearly identified.
We wish to acknowledge the valuable comments and assistance of Steve Davies from Geraldton and staff from the Northam district office of the Department of Agriculture and Food WA and UWA - CMM. This research was funded by Grains Research and Development Corporation (GRDC).
Manuscript received 26 April 2009, accepted 17 September 2009
Anand RR, Paine M (2002) Regolith geology of the Yilgarn Craton, Western Australia: implications for exploration. Australian Journal of Earth Sciences 49, 3-162. doi:100.1046/j.1440-0952.2002.00912.x
Bullock P, Fedoroff N, Jongerius A, Stoops G, Tursina T, Babel U (1985) 'Handbook for soil thin section description.' (Waine Research Publications: Wolverhampton, UK)
Bureau of Meterology (2008) Climate Data Western Australia. National Climate Centre, Melbourne, Australia.
Eggleton RA (2001) 'The regolith glossary.' (Cooperative Centre for Landscape Evolution and Mineral Exploration, National Capital Printing: Canberra, ACT)
Finkl CW, Churchward HM (1973) The etched landsurfaces of southwestern Australia. Journal of the Geological Society of Australia 20, 295-307.
Folk RL (1974) 'Petrology of sedimentary rocks.' (Hemphill Publishing Company: Austin, TX)
Gili JA, Alonso EE (2002) Microstructural deformation mechanisms of unsaturated granular soils. International Journal for Numerical and Analytical Methods in Geomechanics 26, 433-468. doi:10.1002/nag.206
Henderson CWL, Levett A, Lisle D (1988) The effects of soil water content and bulk density on the compactibility and soil penetration resistance of some Western Australian sandy soils. Australian Journal of Soil Research 26, 391-400. doi:10.1071/SR9880391
Isbell RF (1996) 'The Australian Soil Classification.' (CSIRO Publishing: Collingwood, Vic.)
Jarvis RJ, Hamblin AP, Delroy ND (1986) Continuous cereal cropping with alternative tillage systems in Western Australia. West Australian Department of Agriculture Technical Bulletin No. 71.
Kew GA, Gilkes RJ (2006) Classification, strength and water retention of lateritic regolith. Geoderma 136, 184-198. doi:10.1016/j.geoderma. 2006.03.025
Kew GA, Gilkes RJ (2007) Properties of regolith beneath lateritic bauxite in the Darling Range of South Western Australia. Australian Journal of Soil Research 45, 164-181. doi:10.1071/SR06128
Kew GA, Gilkes RJ, Mathison CI (2008) Nature and origins of granitic regolith in the Darling Range, Western Australia. Australian Journal of Earth Sciences 55, 473-492. doi:10.1080/08120090801888602
Killigrew LP, Glassford DK (1976) Origin and significance of kaolin spherites in sediments of south-western Australia. Search 7, 393-394.
McArthur WM (1991) 'Reference soils of south-western Australia.' (Australian Soil Science Society of Australia Inc. (WA Branch): Perth, W. Aust.)
McKenzie N, Coughlan K, Cresswell H (2002) 'Soil physical measurement and interpretation for land evaluation.' (CSIRO Publishing: Collingwood, Vic.)
Mulcahy MJ (1973) Landforms and soils of south-western Australia. Journal of the Royal Society of Western Australia 56, 16-22.
Mullins CE, Panayiotopoulos KP (1984) Compaction and shrinkage of sands and sand-kaolin mixtures. Soil & Tillage Research 4, 191-198. doi:10.1016/0167-1987(84)90047-3
Nahon DB (1991) 'Introduction to the petrology of soils and chemical weathering.' (John Wiley: New Jersey)
Northcote KH (1974) 'A factual key for the recognition of Australian soils.' (Rellim Technical Publications: Glenside, S. Aust.)
Northcote KH, Bettenay E, Churchward HM, McArthur WM (1967) 'Atlas of Australian Soils. Explanatory data for sheet 5. Perth-Albany-Esperance.' CSIRO. (Melbourne University Press: Carlton, Vic.)
Panayiotopoulos KP (1989) Packing of sands A review. Soil & Tillage Research 13, 101-121. doi:10.1016/0167-1987(89)90085-8
Rasband W (1999) 'Image J version 1.62.' (Research Services Branch, National Institute of Health: Bethesda, MD)
Stace HCT, Hubble GD, Brewer R, Northcotte KH, Sleeman JR, Mulcahy MJ, Hallsworth EG (1968) 'A handbook of Australian soils.' Rellim Technical Publication 206. (Rellim Technical Publications: Glenside, S. Aust.)
Trendall IR (1975) Precambrian. In 'The Geology of Western Australia'. Western Australian Geological Survey, Memoir 2. pp. 25 32.
Tretyakov VV, Romanov SG, Fokin AV, Alperovitch VI 0998) EMPA of the composition of opal-based nanostructured materials. Mikrochimica Acta 15(Suppl.), 211-217.
G. A. Kew (A,C), R. J. Gilkes (A), and D. Evans (B)
(A) School of Earth and Geographical Sciences, Faculty of Natural and Agricultural Sciences, University of Western Australia, Crawley, WA 6009, Australia.
(B) Bioworks, 16 Hines Rd, O'Connor, WA 6163, Australia.
(C) Corresponding author. Email: firstname.lastname@example.org
Table 1. Classification of clay matrix fabric based on analysis of SEM micrographs using Image J software Class % Description 1 <30 Of an individual quartz grain coated with clay <10 Structural bridges (clay) between quartz grains 2 30-50 Of an individual quartz grain coated with clay 10-20 Structual bridges (clay) between quartz grains <30 Isolated clay matrix 3 >50 Of an individual quartz grain coated with clay 20-30 Structual bridges (clay) between quartz grains <30 Isolated clay matrix 4 50-100 Of an individual quartz grain coated with clay 30-60 Structual bridges (clay) between quartz grains 30-60 Isolated clay matrix 5 100 Of an individual quartz grain coated with clay >60 Structual bridges (clay) between quartz grains <30 Isolated clay matrix Class % Bullock et al. (1985) 1 <30 Manic, only fabric units of one size group or <10 amorphorous material present Cefiiric, coarser units linked by finer material 2 30-50 Chitonie, coarser units are surrounded by a cover of 10-20 smaller units <30 3 >50 Chitonic 20-30 Enaulie, a skeleton of larger units with aggregates <30 of smaller units in the interstitial spaces 4 50-100 30-60 30-60 5 100 Enaulic >60 Porphyritic, larger units occur in a dense <30 groundmass of smaller units Table 2. Variables included in the statistical analyses of data for subsoils on transported material [M.sub.z], Graphic mean; [[sigma].sub.1] inclusive graphic standard deviation; [Sk.sub.I], inclusive graphic skewness; [K.sub.G], graphic kurtosis Sample Image J data SEM data Circularity Class Porosity Oxide No. Cl wt% wt% Casurina s1 0.60 3 0.425 75 Casurina s2 0.60 2 0.365 76 Buntine s3 0.57 4 0.330 72 Buntine s4 0.62 4 0.320 75 Buntine s5 0.63 5 0.350 77 Binnu s6 0.62 1 0.345 71 Binnu s7 0.55 5 0.290 80 Binnu s8 0.63 2 0.380 70 Binnu s9 0.60 5 0.565 75 Grass Patch s10 0.64 4 0.435 80 Yuna s11 0.64 2 0.340 77 Yuna s12 0.55 4 0.395 78 Casurina s13 0.62 4 0.250 73 Kojonup s14 0.49 4 0.520 74 Grain size (Folk 1974) [M.sub.z] [[sigma].sub.1] [Sk.sub.I] [K.sub.G] phi Casurina 2.29 2.28 0.20 0.94 Casurina 2.25 2.29 0.25 0.95 Buntine 1.45 2.25 0.40 0.94 Buntine 2.18 2.71 0.41 0.70 Buntine 2.12 2.65 0.45 0.78 Binnu 1.79 1.31 0.32 1.65 Binnu 2.48 1.79 0.43 1.53 Binnu 1.87 1.34 0.29 1.81 Binnu 1.83 1.32 0.29 1.72 Grass Patch 3.23 1.93 0.39 0.84 Yuna 1.72 1.66 0.30 1.47 Yuna 1.67 1.68 0.20 1.29 Casurina 1.61 2.03 0.37 1.19 Kojonup 1.48 2.27 0.42 0.85 UCS Clay (%) -10 kPa -600 kPa (kN/[m.sup.2]) Casurina 14 13 98 Casurina 13 27 114 Buntine 8 25 81 Buntine 20 20 50 Buntine 17 15 104 Binnu 6 7 15 Binnu 16 27 357 Binnu 7 10 28 Binnu 6 21 172 Grass Patch 16 61 291 Yuna 9 11 35 Yuna 8 15 91 Casurina 13 9 98 Kojonup 8 15 75 Fig. 9. Plots of unconfined compression strength v. (a) water content and (b) matric potential for representative subsoil and saprolite. Exponential function coefficients and [R.sup.2] Class Rep Location Coefficient no. sample A ** 1 & 2 s6 Binnu 16 3 & 4 s12 Yuna 70 5 s7 Binnu 2029 Quartz-rich Huntly * 2162 Clay-rich Huntly * 7541 Class Coefficient [R.sup.2] no. B 1 & 2 -0.07 0.63 3 & 4 -12 0.57 5 -20 0.78 Quartz-rich -13 0.96 Clay-rich -10 0.98 * Saprolite from Darling Range. * Dry strength. Exponential function coefficients and [R.sup.2] Class Rep Location Coefficient no. sample C *** 1 & 2 s6 Binnu 1.9 3 & 4 s12 Yuna 1.5 5 s7 Binnu 2.4 Quartz-rich Huntly * 10 Clay-rich Huntly * 20 Class Coefficient [R.sup.2] no. D 1 & 2 0.58 0.93 3 & 4 1.4 0.99 5 0.91 0.91 Quartz-rich 0.77 0.95 Clay-rich 0.82 0.98 * Saprolite from Darling Range. *** Wet strength
|Printer friendly Cite/link Email Feedback|
|Author:||Kew, G.A.; Gilkes, R.J.; Evans, D.|
|Publication:||Australian Journal of Soil Research|
|Date:||Mar 1, 2010|
|Previous Article:||High clay contents, dense soils, and spatial variability are the principal subsoil constraints to cropping the higher rainfall land in south-eastern...|
|Next Article:||Claying and deep ripping can increase crop yields and profits on water repellent sands with marginal fertility in southern Western Australia.|