Distribution and causes of intricate saline--sodic soil patterns in an upland South Australian hillslope.
Soil mapping by traditional means (e.g. Gunn et al. 1988; Soil Survey Staff 1993) at a high scale (e.g. 1 : 10000) (Gallant et al. 2008) can be prohibitively expensive and laborious, and therefore impractical in many instances (Bishop and McBratney 2001). Yet soil maps are essential to understand the nature of soil spatial variability, and are pivotal in formulating on-ground farming and natural resource management decisions (e.g. Solie et al. 1999; Brejda et al. 2000; Kravchenko and Bullock 2000; McKenzie et al. 2000; Park and Vlek 2002).
Underlying the so-called soil survey paradigm (Hudson 1992) that is integral to traditional soil mapping are mental models (Hoosbeek and Bryant 1992; Bui et al. 1999; McKenzie and Ryan 1999; McKenzie et al. 2000). Such models are conceptualisations of often-complex environmental relationships that couple the soil-forming influences and processes apparent in the landscape (e.g. topography and parent material) with the distribution of soil types seen in the landscape. Once sufficiently tested, evaluated, and refined by the soil surveyor, they are then used to make maps (e.g. Arnold 2006). Mental models are rarely formally communicated by the soil surveyor (Bui et al. 1999; McKenzie and Ryan 1999; McKenzie et al. 2000; Bui 2004), although reports that accompanying soil maps often contain descriptive narratives of soil and landscape relationships (e.g. Stephens et al. 1945; Beattie 1972) and field keys (e.g. McKenzie 1992) that are abstractions of surveyors' mental models.
Conceptual toposequence models described in Fritsch and Fitzpatrick (1994) have been used to simplify or abstract mechanisms that occur in toposequences (geological--pedological cross-section) to allow better understanding of soil--regolith processes, and used to explain causes of land degradation. These models emphasise: controls on water movement (e.g. stratigraphy, A, B, C horizons and layers, and slope gradient); strength and direction of principal water flows (e.g. perched and deep groundwater systems); accumulation/dispersion zones (e.g. concretions, salinisation, and mottling); and soil types. For a given area, soil maps of soil distribution and conceptual toposequence models showing soil, regolith, and hydraulic properties and connections combine powerfully to assist in refining, communicating, and implementing perhaps complicated land management decisions (Fitzpatrick et al. 1994; Fitzpatrick and Merry 2002; Fitzpatrick 2005).
Modern soil survey methods that apply geophysical survey and terrain analysis offer a strong opportunity to address the labour and expense limitations of traditional soil survey (McBratney et al. 2003; Grunwald 2006; McKenzie et al. 2008). Widespread adoption of these survey techniques has increased in recent years, causing need for a major revision of the Australian 'Guidelines for Surveying Soil and Land Resources' (McKenzie et al. 2008). These changes in the survey approach reflect reliable access to: high-powered, everyday computing coupled with remote sensing, geographic information system (GIS), and geostatistical software; growing environmental/soil--landscape digital archives, including digital elevation models (DEMs) for terrain analysis (Burrough and McDonnell 2000; Wilson and Gallant 2000b); and a growing suite of high-sample intensity, geophysical survey instruments linked to global positioning systems (GPS). Of these geophysical survey methods, electromagnetic induction (EMI, or EM) has been used to identify landscape apparent electrical conductivity ([EC.sub.a]) patterns (e.g. McKenzie et al. 1997; Bennett et al. 2000; Eastham et al. 2000; Broadfoot et al. 2002; James et al. 2003; Hedley et al. 2004; Spies and Woodgate 2005), revealing often spatially correlated soil properties of clay content and clay mineral type, water content and salinity (Slavich and Petterson 1990; McKenzie and Ryan 2008). Magnetic susceptibility surveys show Fe-oxide patterns that are typically dominated by maghemite and magnetite content (Mullins 1977; Thompson and Oldfield 1986; Evans and Heller 2003). Magnetic susceptibility landscape patterns have been used to interpret ancient and contemporary pedogenic processes (e.g. Fine et al. 1989; Williams and Cooper 1990; de Jong et al. 2000; Grimley and Vepraskas 2000; Royali 2001; Grimley et al. 2004; Pennock and Veldkamp 2006).
Thomas et al. (2007) describe a soil survey conducted by traditional methods (e.g. Gunn et al. 1988; Soil Survey Staff 1993) during 1988 to investigate relationships between intricate stunted crop growth patterns, soil properties, and possible root disease in a 330-ha site comprising a Belalie Valley hillslope in the Midnorth district of South Australia. Soil survey revealed localised watertable perching and waterlogging, and excessive salt concentrations associated with saline--sodic soils; physicochemical rather than biological soil issues were implicated as the key causes for the observed spatially intricate crop growth patterns. However, at the survey intensity of 1 morphological description per 2.6ha, and 1 physicochemical analysis per 18.3 ha, it was not possible to resolve and explain the soil processes that give rise to the spatially intricate saline--sodic soil patterns that were indicated by the observed crop growth patterns.
The prime objective of this paper, therefore, is to conduct more detailed investigations of the Belalie Valley site than previously practical in 1988 by applying modern soil survey methodology to: reveal intricate saline--sodic soil patterns using [EC.sub.a] and volume magnetic susceptibility ([kappa]) surveys, and terrain analysis; and construct an improved conceptual toposequence model based on methods described in Fritsch and Fitzpatrick (1994), which we use to explain and communicate the pedogenic causes of the observed intricate soil patterns.
Materials and methods
The study area, centred on the coordinates 138[degrees]37'37"E, 33[degrees]24'2"S, is situated 23km south of the regional centre of Jamestown and 18km north of the town of Spalding (Fig. 1). The study area combines an upland landscape dominated by Sodosols (Isbell 1996). The annual rainfall is 470mm, which falls predominantly during winter (Mediterranean climate). Agriculture in the survey area consists of rainfed mixed cropping in rotations of wheat (Triticum spp.), barley (Hordeum vulgate), canola (Brassica napus), lucerne (Medicago sativa), and sheep grazing.
As shown in Fig. 1, the survey area (128 ha) lies within the 1988 traditional survey area (330ha), and features an east-facing hillslope with an elevation range of 470 m (Ward Hill) to 395m (base of Freshwater Creek). Figure 2 shows the elevation and slope profile traced along the transect A-A" (Fig. 1), which runs tangential to prevailing slope. In cross-section, the hillslope profile is hyperbolic concave (Schoeneberger et al. 2002), and geomorphologically consistent with a typical colluviated--alluviated hillslope profile.
In the investigation described in Thomas et al. (2007), detailed morphological descriptions (McDonald et al. 1998) of 124 soil cores ([less than or equal to] 2m) were augmented by laboratory physicochemical analysis (Rayment and Higginson 1992) of 18 of the soil cores. This survey revealed saline--sodic subsoils (Northcote and Skene 1972).
Previous soil survey and laboratory data (Thomas et al. 2007) indicate the hillslope soils to be pedologically complex, reflecting variable parent materials of interbedded tillites, quartzite shales, and massive quartzites (upper slopes), and fine grained siltstone, mudstones, and shales (lower slopes) (Burra 1:250000 sheet, Geological Survey of South Australia 1964). Provisional soil mapping conducted in the traditional survey area (Thomas et al. 2007) applied field observation, and detailed layer-by-layer morphological description of 125 GPS-located, intact soil cores taken to a maximum depth of 2 m. A subset of 18 was selected for a suite of physicochemical analyses, including: texture, EC of saturated paste extract ([EC.sub.sc]), cation exchange capacity (CEC), and pH (Rayment and Higginson 1992; McKenzie et al. 2002). These observations and data revealed a toposequence of 6 landscape soil units (LSUs), which, in accordance with Australian Soil Classification (Isbell 1996), consisted of: LSU 1, a Lithocalcic Calcarosol occupying the crest; LSU 2, a Brown Sodosol occupying the shoulder slope; LSU 3, a Red Sodosol occupying the foot slope; LSU 4, a Brown Sodosol, occupying the toe slope; LSU 5, a Red Sodosol of the gully face; and LSU 6, a Salic Hydrosol at the base of the 12-m gully.
Survey methodology and soil-landscape data
Ground geophysical survey
The geophysical surveys were conducted on foot during March 2003. The survey area consisting of 800 by 1600m (128 ha) area and followed a near-regular 20-m interval along hillslope (i.e. east--west) transects separated by 50m across the slope (i.e. north--south) (Fig. 1). Within practical constraints, the survey was designed to achieve the greater intensity along the hillslope (i.e. east--west), considered the orientation of greatest pedogenic variability due to prevailing slope and interbedded geology. The following instrumentation were acquired simultaneously during the survey: (i) Geonics EM38 EMI meter operated in vertical dipole mode to measure depth-weighted [EC.sub.a] of the soil layers (Rhoades et al. 1999) to an effective depth of <l.5 m (McNeill 1980); (ii) Geonics EM31 EMI meter to measure bulked [EC.sub.a] of soil-regolith layers to a depth of ~6m (McNeill 1980); (iii) Bartington MS2-D loop sensor to measure the volume magnetic susceptibility ([kappa]) of the upper 2 mm of the soil surface when in contact (Dearing 1999a); and (iv) GPS, accurate to within a few metres of true ground location.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
All 979 measurements acquired during the survey by each instrument (Fig. 1) were georegistered in a GIS. Data from each instrument were interpolated by ordinary kriging (Burrough and McDonnell 2000; Webster and Oliver 2001) using ArcGIS Geostatistical Analyst (Johnston et al. 2001) to create GIS raster-grid survey coverages of 3 m ground resolution (i.e. 3 by 3 m).
Elevation and digital terrain models
A 5-m resolution DEM of the 2200-ha area surrounding the survey areas (Fig. 1) was photogrammetrically generated using 1 : 40 000 scale aerial photographic stereo pair. Digital terrain models (DTMs) including slope and topographic wetness index (TWI) were generated from the DEM according to Eqn 1:
W = ln([A.sub.s]/tan[beta]) (1)
where As is the specific catchment area ([m.sup.2]/m) and [beta] is the slope gradient in degrees (Wilson and Gallant 2000a).
Topographic wetness index is a terrain index used to predict cumulative flow of water in near-surface landscapes, taking into account upslope contributing area and local landform (Wilson and Gallant 2000b).
Data integration and 3D GIS investigations
A digital database of the traditional soil survey of 1988 comprising morphological and physicochemical data (Thomas et al. 2007) was created and georeferenced in GIS, and spatially coregistered with the geophysical survey and DTM coverages, and the study area digital aerial photograph. Using the ArcMap 3D Analyst GIS module (Booth 2000; Kennedy 2004), the aerial photograph was projected, or rendered, on the study area DEM, offering 3D perspective scenes of the landscape. The GIS scenes could be interactively altered using manual zoom, tilt, and pseudo-hillshade controls. Next, the various geophysical survey and DTM coverages were projected on the scenes. Transparency manipulation of the upper coverage, e.g. the TWI coverage, allowed TWI patterns to be visually compared with landform using the aerial photograph visible below to give positional context in the landscape. In this way, swapping the aerial photograph for another coverage, e.g. the coverage, allowed K patterns to be visually compared to TWI patterns and visually interpreted according to landscape/ landform context. Furthermore, the georeferenced digital database of the traditional 1988 survey was projected on the 3D perspective scenes to highlight selected morphological or physiochemical data at the soil survey locations. Such manually controlled visual GIS approaches allowed thorough desktop investigation of landform and survey coverage relationships to be made, and assisted development of pedogenic models (e.g. conceptual toposequence).
Conceptual toposequence model construction
Fritsch and Fitzpatrick (1994) describe the preliminary stage of conceptual toposequence model construction. This involves identifying and describing soil--regolith features including geology, horizons, nodules, concretions and structure. In this way, for example, grey-green-blue, leached hydromorphic, lower slope soil--regolith landscape units are separated from redder, upper slope hillslope soil--regolith landscape units. Horizontal horizon boundaries and vertical soil--regolith landscape units boundaries (consistent with soil unit boundaries) are drawn on the conceptual cross-section diagram to define the hilislope arrangement of soil--regolith landscape units. Using the interpretative skills of the developer, displayed horizons and features are linked to hillslope hydrological processes (e.g. water flow paths, salinisation, and sodification) by plotted arrows indicating prevailing water flow directions that have been interpreted from data (e.g. piezometric data or hydraulic conductivity, [K.sub.sat]). Additional information interpreted from soil--regolith colour and other morphological, chemical, and mineralogical indicators, and geology, are applied to gain understanding of prevailing water flow paths. Such soil--regolith information may also reveal ancient soil--landscape processes (e.g. hydromorphic and lateritic soil--regolith landscape units), whereas current physicochemical (e.g. [EC.sub.sc] and pH) and hydraulic measurements (e.g. piezometric, mottling patterns, and other pedotransfer functions) and field observations (e.g. the emergence of recent scalds, changing patterns in plant species) reveal contemporary (degrading) processes.
Whereas traditional conceptual toposequence models are presented as 2D (profile), mechanistic cross-section models (e.g. Fritsch and Fitzpatrick 1994), we modified presentation to incorporate 3D GIS scenes of survey area landform and geophysical survey to provide surface spatial context to hillslope processes depicted in the mechanistic model.
Results and discussion
Landform, electrical conductivity and magnetic susceptibility patterns
Figure 3 shows the survey coverages overlaid on the survey area landscape (Fig. 1). These scenes provide a 3D perspective view, which assists in the interpretation of the survey patterns in the context of local landform. The 3D perspective view shown in Fig. 3 is one of many possible views of the survey area. Several prominent hillslope landform features are identifiable from the 3D perspective view of the aerial photograph shown in Fig. 3a, which, when combined, help interpret survey area hydropedological patterns (Lin 2003), and underscore the influence of near-surface drainage on hillslope soil properties. The hillslope features identified include: (i) 2 prominent zones of near-surface drainage ('dz'), which are bound by 3 laterally projecting ends (laterally convex areas) in the hillside ('if', i.e. nose slope interfluves; Schoeneberger et al. 2002) causing divergent near-surface water flow; and (ii) the break-of-slope, which coincides with the boundary between the colluviated footslope soils ('COL') and the alluviated toeslope soils ('ALL').
EM31 apparent electrical conductivity patterns
Deep profile (~6 m) [EC.sub.a] patterns using a EM31 in the survey area hillslope are shown in Fig. 3b, and show a range of [EC.sub.a] 0.14-2.5dS/m. These hillslope patterns feature several high [EC.sub.a] responses in the survey area (i.e. dark red areas). In the upper and mid slope zones of the landscape (i.e. LSUs 1-3), the highest [EC.sub.a] responses are associated with the locally low-lying drainage areas ('dz'), while the lowest responses tend to be associated with locally elevated areas, i.e. on interfluve areas ('if'). According to the conceptual drainage and salt accumulation model shown in Fig. 4 developed here for the hillslope section based on standard methods (Fritsch and Fitzpatrick 1994), the relatively high [EC.sub.a] responses in mid slope areas are attributed to the combination of the accumulation of deep profile salts and the presence of conductive basement rock. The possible contribution of electrically conductive Fe-gravels accumulated in colluvial layers is likely to also contribute to the observed high [EC.sub.a]. Salts present are likely to have collected and concentrated in the drainage zones from upslope areas, or preferentially seeped from below in saline groundwaters under the influence of piezometric pressure of a confined aquifer.
The transition between the footslope and toeslope areas, which corresponds strongly with the colluvially dominated (LSU 3) and alluvial (LSU 4) boundary, forms a prominent contrasting [EC.sub.a] feature in the hillslope profile. The relatively high [EC.sub.a] at depth (~6 m) of much of the toeslope zone (LSU 3) is likely to be related the combination of clay content and salt accumulation, which are often correlated in landscapes (Slavich and Petterson 1990; McKenzie and Ryan 2008). Here, as shown in Fig. 4, salt accumulation is caused by a hydraulic conductivity barrier at the contact between the LSU 3 and LSU 4 soils. Such a hydraulic conductivity barrier is caused by the moderately shallow, coarser textured topsoil of the upslope LSU 3 soils that abut generally heavy clay texture soils at the corresponding depths in the LSU 4 soil profile. The sodic clay associated with the B horizon of the LSU 4 unit is the cause of a retarded throughflow rate in the deep profile, causing salts originally from upslope positions to be concentrated deep in the LSU 4 profile. The salts that have concentrated deep in the LSU 4 profile seep out of the gully face after slow seepage. Evidence for this process is presented from the highest EM31 [EC.sub.a] values encountered in the survey area at the gully face, and effervescent salts (halite and gypsum) were observed from the wicking of saline water on the gully face. The rate of salt transport through the LSU 4 profile has increased during the past 50 years, caused by the changed hydraulic properties by the formation in 1941 of the 12-m erosional gully at the base of Freshwater Creek (Cresswell and Liddicoat 2004).
EM38 apparent electrical conductivity patterns
When visually compared, the EM31 (Fig. 3b) and EM38 (Fig. 3c) survey area patterns are inverted in many locations of the hillslope. For example, in areas where the EM31 survey coverage shows high [EC.sub.a] responses, the corresponding areas of the EM38 survey coverage generally show low [EC.sub.a] responses.
The EM38 [EC.sub.a] survey, in a value range of [EC.sub.a] 0.05-1.1 dS/m, shows a strongly conductive feature 'm' in the shoulderslope, on the boundary of the LSU 1 and LSU 2 soils. This feature corresponds to a locally intense conductivity zone according to the locally intense [EC.sub.a] pattern. Given the evidence of sulfur in this landscape (i.e. acid sulfate soil conditions and gypsum described in Fitzpatrick et al. 2003), the feature is possibly pyritic. The absence of a matching [EC.sub.a] pattern from the EM31 survey (Fig. 3b) indicates that the feature is restricted to the near-surface. Beyond this soil-landscape feature, the areas of highest [EC.sub.a] response in the survey area hillslope occupy the LSU 3. Within LSU 3, the EC, patterns correspond to drainage zones and interfluves, with the interfluves featuring relatively strong [EC.sub.a] pattern responses (sites 1, 2, and 3 in Fig. 3c), while the drainage zones feature the relatively low EC, responses. Site 1 in the south of the survey area occupies a prominent nose slope interfluve, site 2 occupies relatively elevated land juxtaposed between the 2 prominent drainage zones of the hillslope, while site 3 in the north occupies the lower fringe of a saddle landform, adjacent to one of the prominent hillslope drainage zones.
[FIGURE 3 OMITTED]
The alluvial footslope features a zone of moderate near-surface [EC.sub.a]. This pattern is caused by soils that contain moderate salt concentrations in combination with a high clay content. As for the EM31 survey, the EM38 survey shows a pattern that corresponds to the LSU 3/LSU 4 boundary.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
Volume magnetic susceptibility patterns The magnetic susceptibility ([kappa]) survey, in the range 94.5-791.0 x [10.sup.-8] SI, indicates a soil-landscape with strongly elevated values compared with other reported soil--landscapes, principally European (Dearing 1999b). Overall, the [kappa] survey patterns in Fig. 3d strongly match the EM38 survey patterns in Fig. 3c. Like the EM38 survey patterns, strong [kappa] patterns correspond with the locally elevated interfluve zones, whereas the response patterns are weakest in the adjacent drainage zones. However, [kappa] and EM38 patterns in the survey landscape do not match at: (i) the nearsurface zone (i.e. 'm', Fig. 3c), (ii) at the colluvial-alluvial (LSU 3/LSU 4) boundary, and (iii) in the upper slope zones (LSU 1 and LSU 2), where the K survey responses are generally of lower values.
The slope DTM in Fig. 5a shows the steepest slope gradient (14[degrees]) in the shoulderslope, and gradients of flat/near-flat in the toeslope. The drainage zones can be traced up the hillslope in the mid and upper slopes by patterns of lower local gradients that finger up the hillslope.
To assist in better interpretation of TWI patterns in Fig. 5b, the survey area is stratified according to hillslope position. In the upper and mid slope positions, the elevated TWI values match the drainage patterns also traced in the slope DTM. According to the TWI, these patterns correspond to water accumulation zones, which indicate locations of focused surface and throughflows. Patterns of elevated TWI values are evident fingering up-slope in the drainage zones. Accordingly, the reduced TWI values in upper and mid slope positions indicate where surface and throughflows are dispersed, defining interfluves. In the lower slopes where the gradient is flatter, mid to lower range TWI pattern values indicate positions of net water accumulation, indicating where surface and throughflow rates are less, and hence areas of the landscape that are likely to be affected by seasonal waterlogging.
[FIGURE 6 OMITTED]
Descriptive hydropedological model
The [EC.sub.a] surveys using EM38 and EM31 indicated patterns that are dominated by the 3D distribution of salts and clay mineral or layer silicate content, with possible additional contributions in the higher elevation landscape from deep profile (~6 m) magnetic gravels (i.e. maghemite and hematite) indicated by the EM31 survey patterns. The EM38 patterns correlate with the [kappa] patterns, especially in the sloping LSU 3 soil zone. The similarity in these patterns can be hydropedologically explained, particularly when interpreted in combination with throughflow and water accumulation patterns identified in the TWI coverage. For example, in the LSU 3 soil locations where matching EM38 and [kappa] survey patterns are strongest, these areas tend correspond to perched landforms that receive lower rates of throughflow flushing, making these positions comparatively drier than other parts of the landscape. These areas of LSU 3 are highlighted in Fig. 6 (sites 1-3), in which the EM38 [EC.sub.a] contours strongly match to the TWI 'flushing' patterns. Under these hydraulic conditions, the near-surface layers experience a combination of: (i) reduced rates of leaching, resulting in higher residual salt concentrations; and (ii) pedogenic conditions that favour the preservation of maghemite and hematite. Thus, for different pedogenic reasons, the matching EM38 and [kappa] patterns found in the LSU 3 soils are the expression of near-surface freshwater flushing rates.
[FIGURE 7 OMITTED]
Soil survey data from Thomas et al. (2007) show that illuvial enhancement of the fine clay fraction has taken place in upper profiles of the toeslopes soils (i.e. LSU 4 soils). As such, a hydraulic barrier has formed at the contact of LSU 3 and LSU 4 sola, causing upslope freshwater throughflows to back up and create perched groundwater conditions during winter in the lower elevation LSU 3 areas. During summer when evaporation rates are high, the salts that have accumulated in the perched groundwater systems become concentrated (i.e. [EC.sub.sc] [greater than or equal to] 2 dS/m), forming non-groundwater salinity (Fitzpatrick 2005; Rengasamy 2006) in the subsoil.
Improved conceptual toposequence model
Knowledge accumulated of soil--landscape patterns and processes in the survey area using the 3D investigation of the survey area terrain, combined with the geophysical and morphological surveys, laboratory physicochemical data (Thomas et al. 2007), and DTMs (including TWI and slope), facilitated development of the conceptual toposequence model shown in Fig. 7, first presented in Fitzpatrick et al. (2003) and based on methods in Fritsch and Fitzpatrick (1994). However, the conceptual toposequence model constructed in the current study features improvements to the standard method of model development. The key improvement is the incorporation in the graphic model of the oblique view and geographical survey, in this case EM38 [EC.sub.a] (Fig. 7), overlaying the aerial photograph. Such treatment provides spatial context to soil--regolith features and processes depicted in the graphic profile representation of conventional conceptual toposequence models, and pictorially depicts the information we present concerning landscape, hydropedological, and degradation processes described in the descriptive (mental) model in the section above.
The improved conceptual toposequence model highlights our understanding of pedogenic relationships between landform, geology, and soil types (LSU) of the landscape. Furthermore, the morphological properties (e.g. layering, colour and redoximorphic conditions, nodules and concretions, and structure) of LSUs are shown, giving expression to pedogenic processes in the landscape. Overprinted on these are dominant water flow regimes, water perching, and salt-affected (saline and sodic) zones. Finally, superimposed on the model are nearsurface (<1.5 m) [EC.sub.a] patterns for a section of the landscape, showing electrical conductivity relationships of landform and soils, which we attribute to clay and salinity relationships. The model provides tangible expression of the metal model of soil--landscape processes at play, with particular emphasis on the formation of saline--sodic patterns.
In this study we have applied new soil--regolith survey approaches to reveal intricate saline--sodic soil patterns that were not revealed using traditional soil survey alone. We have demonstrated that, with the combination of geophysical soil survey techniques [EM31 of deep (<6 m) [EC.sub.a], EM38 for shallow (<l.5m) [EC.sub.a], and [kappa] for near-surface patterns of maghemite and hematite] and terrain analysis, it is possible to spatially locate near-surface water throughflow patterns, assisting in the interpretation of pedogenic processes and soil distributions. Such interpretations were strongly assisted by the use of 3D GIS, which allowed superior pedogenic models to be formulated and committed in the form of the improved conceptual toposequence model. The model serves to communicate complex pedogenic information, and we anticipate could be adapted in the development of a regional predictive method, particularly given that the hillslope is regionally representative.
Here we build on a soil research legacy from the study area landscape initiated by Dr Albert Rovira of CSIRO Division of Soils during the 1980s. We are grateful for the ongoing support of the Cootes and Ashby families, the local farmers. Funding by the Cooperative Research Centre for Landscape Environments and Mineral Exploration (CRC-LEME), National Action Plan for Salinity and Water Quality (NAP), and SA Department for Water, Land and Biodiversity Conservation (DWLBC) is acknowledged. Finally, we thank Dr Richard Cresswell (CSIRO Land and Water) and the 2 anonymous referees for their advice in refining the manuscript.
Manuscript received 21 November 2008, accepted 15 December 2008
Arnold RW (2006) Soil survey and soil classification. In 'Environmental soil--landscape modeling: Geographic information technologies and pedometrics'. (Ed. S Grunwald) pp. 32-59. (Taylor & Francis Group: Boca Raton, FL)
Beattie JA (1972) Groundsurfaces of the Wagga Wagga Region, New South Wales. CSIRO Soil Publication No. 28, Melbourne.
Bennett DL, George R J, Whitfield B (2000) The use of ground EM systems to accurately assess salt store and help define land management options for salinity management. Exploration Geophysics 31, 249-254. doi: 10.1071/EG00249
Bishop TFA, McBratney AB (200 I) A comparison of prediction methods for the creation of field-extent soil property maps. Geoderma 103, 149-160. doi: 10.1016/S0016-7061(01)00074-X
Booth B (2000) 'Using ArcGIS 3D Analyst.' (Environmental Systems Research Institute, Inc.: Redlands, CA)
Brejda JJ, Moorman TB, Smith JL, Karlen DL, Allan DL, Dan TH (2000) Distribution and variability of surface soil properties at a regional scale. Soil Science Society of America Journal 64, 974-982.
Broadfoot K, Morris M, Stevens D, Heuperman A (2002) The role of EM38 in land and water management planning on the Tragowel Plains in northern Victoria. Exploration Geophysics 33, 90-94. doi: 10.1071/ EG02090
Bui EN (2004) Soil survey as a knowledge system. Geoderma 120, 17-26. doi: 10.1016/j.geoderma.2003.07.006
Bui EN, Longhhead A, Comer RJ (1999) Extracting soil landscape rules from previous soil surveys. Journal of Soil Research 37, 495-508. doi: 10.1071/S98047
Burrough PA, McDonnell RA (2000) 'Principles of geographical information systems.' (Oxford University Press Inc.: New York)
Cresswell R, Liddicoat C (2004) Application of airborne geophysical techniques to salinity issues around Jamestown, South Australia. Department of Water, Land and Biodiversity Conservation, SA-SMMSP Site Summary Report No. DWLBC 2004/37, Adelaide, S. Aust.
de Jong E, Pennock DJ, Nestor PA (2000) Magnetic susceptibility in different slope positions in Saskatchewan, Canada. Catena 40, 291-305. doi: 10.1016/S0341-8162(00)00080-1
Dearing JA (1999a) 'Environmental magnetic susceptibility: using the Bartington MS2 system.' (Chi Publishing: Kenilworth, England)
Dearing JA (1999b) Magnetic susceptibility. In 'Environmental magnetism: a practical guide'. (Eds J Walden, F Oldfield, J Smith) pp. 35-62. (Quaternary Research Association: London)
Eastham J, Gregory PJ, Williamson DR (2000) A spatial analysis of lateral and vertical fluxes of water associated with a perched watertable in a duplex soil. Australian Journal of Soil Research 38, 879-890. doi: 10.1071/SR99003
Evans ME, Heller F (2003) 'Environmental magnetism: Principles and applications of enviromagnetics.' (Academic Press: Burlington, MA)
Fine P, Singer MJ, La Ven R, Verosub K, Southland RJ (1989) Role of pedogenesis in distribution of magnetic susceptibility in two California chronosequences. Geoderma 44, 287-306. doi: 10.1016/0016-7061 (89) 90037-2
Fitzpatrick RW (2005) Hydro-pedologically based toposequence models as a powerful tool for managing salt-affected landscapes: case studies from geo-chemically variable saline environments. In 'International Salinity Forum. Managing Saline Soils and Water: Science, Technology and Social Issues'. Extended abstracts for oral presentations. Riverside Convention Center, CA. pp. 181-184.
Fitzpatrick RW, Cox JW, Fritsch E, Hollingsworth ID (1994) A soil-diagnostic key to manage saline and waterlogged catchments in the Mt Lofty Ranges, South Australia. Soil Use and Management 10, 145-152. doi: 10.1111/j.1475-2743.1994.tb00477.x
Fitzpatrick RW, Merry RH (2002) Soil-regolith models of soil-water landscape degradation: development and application. In 'Regional water and soil assessment for managing sustainable agriculture in China and Australia'. (Eds TR McVicar, L Rui, J Walker, RW Fitzpatrick, L Changming) pp. 130-138. (Australian Centre for International Agricultural Research: Canberra)
Fitzpatrick RW, Thomas M, Davies PJ, Williams BG (2003) Dry saline land: an investigation using ground-based geophysics, soil survey and spatial methods near Jamestown, South Australia. CSIRO Land and Water, Technical Report No. 55/03, Adelaide, S. Aust. Available at: www.clw.csiro.au/publications/technica12003/tr55-03.pdf
Fritsch E, Fitzpatrick RW (1994) Interpretation of soil features produced by ancient and modern processes in degraded landscapes. I. A new method for constructing conceptual soil-water-landscape models. Australian Journal of Soil Research 32, 889-907. doi: 10.1071/ SR9940889
Gallant JC, McKenzie N J, McBratney AB (2008) Scale. In 'Gudelines for surveying soils and land resources'. (Eds NJ McKenzie, MJ Grundy, R Webster, AJ Ringrose-Voase) pp. 27-43. (CS1RO Publishing: Collingwood, Vic.)
Geological Survey of South Australia (1964) Burra Sheet: S.A. Geological Atlas Series Sheet 1 54-5 Zones 5 & 6. (Department of Mines: Adelaide, S. Aust.)
Grimley DA, Arruda NK, Bramstedt MW (2004) Using magnetic susceptibility to facilitate more rapid, reproducible and precise delineation of hydric soils in the midwestern USA. Catena 58, 183-213.
Grimley DA, Vepraskas MJ (2000) Magnetic susceptibility for use in delineating Hydric soils. Soil Science Society of America Journal 64, 2174-2180.
Grunwald S (2006) 'Environmental soil landscape modeling: Geographic information technologies and pedometrics.' (Taylor & Francis Group: Boca Raton, FL)
Gunn RH, Beattie JA, Reid RE, van de Graft RHM (1988) 'Australian soil and land survey handbook: Guidelines for conducting surveys.' (Inkata Press: Melbourne and Sydney)
Hedley CB, Yule IJ, Eastwood CR, Sheperd TG, Arnold G (2004) Rapid identification of soil textural and management zones using electromagnetic induction sensing of soils. Australian Journal of Soil Research 42, 389-400. doi: 10.1071/SR03149
Hoosbeek MR, Bryant RB (1992) Towards the quantitative modeling of pedogenesis: a review. Geoderma 55, 183-210. doi: 10.1016/0016-7061 (92)90083-J
Hudson BD (1992) The soil survey as a paradigm-based science. Soil Science Society of America Journal 56, 836-841.
Isbell RF (1996) 'The Australian Soil Classification.' (CSIRO Publishing: Collingwood, Vic.)
James IT, Waine TW, Bradley RI, Taylor JC, Godwin RJ (2003) Determination of soil type boundaries using electromagnetic induction scanning techniques. Biosystems Engineering 86, 421-430. doi: 10.1016/j.biosystemseng.2003.09.001
Johnston K, Ver Hoef JM, Krivoruchko K, Lucas N (2001) 'Using ArcGIS geostatistical analyst.' (Environmental Systems Research Institute, Inc.: Redlands, CA)
Kennedy H (2004) 'Data in three dimensions: A guide to ArcG1S 3D analyst.' (Thomson Delmar Learning: Clifton Park, NY)
Kravchenko AN, Bullock DG (2000) Correlation of corn and soybean grain yield with topography and soil properties. Agronomy Journal 92, 75-83. doi: 10.1007/s100870050010
Lin H (2003) Hydropedology: Bridging disciplines, scales, and data. Vadose Zone Journal 2, 1-11. doi: 10.2113/2.1.1
McBratney AB, Mendonca ML, Minasny B (2003) On digital soil mapping. Geoderma 117, 3-52. doi: 10.1016/S0016-7061(03)00223-4
McDonald RC, Isbell RF, Speight JG, Walker J, Hopkins MS (1998) 'Australian soil and survey handbook.' (Australian Collaborative Land Evaluation Program: Canberra)
McKenzie NJ (1992) Soils of the Lower Macquarie Valley, New South Wales. CSIRO Division of Soils Divisional Report No. 117, Melbourne.
McKenzie NJ, Coughlan K, Cresswell H (2002) 'Soil physical measurement and interpretation for land evaluation.' (CSIRO Publishing: Collingwood, Vic.)
McKenzie NJ, Gessler PE, Ryan PJ, O'Connell DA (2000) The role of terrain analysis in soil mapping. In 'Terrain analysis: principles and applications', pp. 245-265. (John Wiley & Sons: New York)
McKenzie N J, Ryan PJ (1999) Spatial prediction of soil properties using environmental correlation. Geoderma 89, 67-94. doi: 10.1016/S0016-7061(98)00137-2
McKenzie NJ, Ryan PJ (2008) Measuring soil. In 'Guidelines for surveying soils and land resources'. (Eds NJ McKenzie, MJ Grundy, R Webster, AJ Ringrose-Voase) pp. 263-284. (CSIRO Publishing: Collingwood, Vic.)
McKenzie N J, Webster R, Ringrose-Voase AJ (2008) 'Guidelines for surveying soil and land resources.' (CSIRO Publishing: Collingwood, Vic.)
McKenzie RC, George RJ, Woods SA, Cannon ME, Bennett DL (1997) Use of electromagnetic-induction meter (EM38) as a tool for managing salinisation. Hydrogeology Journal 5, 37-50. doi: 10.1007/ s100400050110
McNeill JD (1980) Electromagnetic terrain conductivity measurement at low induction numbers. Geonics Ltd. Technical Note TN-6, Ontario, Canada.
Mullins CE (1977) Magnetic susceptibility of the soil and its significance in soil science. Journal of Soil Science 28, 223-246. doi: 10.1111/ j.1365-2389.1977.tb02232.x
Northcote KH, Skene JKM (1972) Australian soils with saline and sodic properties. CSIRO, Soil Publication No. 27, Melbourne, Vic.
Park S J, Vlek PLG (2002) Environmental correlation of three-dimensional soil spatial variability: a comparison of three adaptive techniques. Geoderma 109, 117-140. doi: 10.1016/S0016-7061 (02)00146-5
Pennock DJ, Veldkamp A (2006) Advances in landscape scale soil research. Geoderma 133, 1-5. doi: 10.1016/j.geoderma.2006.03.032
Rayment GE, Higginson FR (1992) 'Australian laboratory handbook of soil and water chemical methods.' (Inkata Press: Melbourne)
Rengasamy P (2006) World salinization with emphasis on Australia. Journal of Experimental Botany 57, 1017-1023. doi: 10.1093/jxb/ erjl08
Rhoades JD, Chanduvi F, Lesch S (1999) 'Soil salinity assessment: methods and interpretation of electrical conductivity measurements.' FAO Paper 57. (FAO: Rome)
Royall D (2001) Use of mineral magnetic measurements to investigate soil erosion and sediment delivery in a small agricultural catchment in limestone terrain. Catena 46, 15-34. doi: 10.1016/S0341-8162(01) 00155-2
Schoeneberger PJ, Wysocki DA, Benham EC, Broderson WD (2002) 'Fieldbook for describing and sampling soils.' (Natural Resources Conservation Service, National Soil Survey Center: Lincoln, NE)
Slavich PG, Petterson GH (1990) Quantifying water movement processes strategic sampling from EM-38 surveys. In 'Soils and their land use in the 1990s'. Bendigo, Vic. pp. 20-26. (Australian Society of Soil Science Inc., Victorian Branch: Warragul, Vic.)
Soil Survey Staff (1993) 'Soil survey manual.' (U.S. Government Printing Office: Washington, DC)
Solie JB, Raun WR, Stone ML (1999) Submeter spatial variability of selected soil and Bermudagrass production variables. Soil Science Society of America Journal 63, 1724-1733.
Spies BR, Woodgate P (2005) 'Salinity mapping methods in the Australian context.' (Commonwealth of Australia: Canberra)
Stephens CG, Herriot RI, Downes RG, Langford-Smith T, Adcock AM (1945) A soil, land-use, and erosion survey of part of County Victoria, South Australia. Council for Scientific and Industrial Research Bulletin No. 188, Melbourne.
Thomas M, Fitzpatrick RW, Cannon ME, McThompson J (2007) Soil patterns along a hillslope in the Belalie Valley, Midnorth, South Australia. CSIRO Land and Water 33/07, Adelaide. Available at: www.clw.csiro.au/publications/science/2007/sr33-07.pdf
Thompson R, Oldfield F (1986) 'Environmental magnetism.' (Allen & Unwin: London)
Webster R, Oliver MA (2001) 'Geostatistics for environmental scientists.' (John Wiley & Sons: Chichester, England)
Williams RD, Cooper JR (1990) Locating soil boundaries using magnetic susceptibility. Soil Science 150, 889-895. doi: 10.1097/00010694-199012000-00009
Wilson JP, Gallant JC (2000a) Digital terrain analysis. In 'Terrain analysis: principles and applications'. (Eds JP Wilson, JC Gallant) pp. 1-27. (John Wiley & Sons Inc.: New York)
Wilson JP, Gallant JC (2000b) 'Terrain analysis: principles and applications.' (John Wiley & Sons Inc.: New York)
M. Thomas (A,B,C), R. W. Fitzpatrick (A,B), and G. S. Heinson (B)
(A) CSIRO Land and Water, PMB 2, Glen Osmond, Adelaide, SA 5064, Australia.
(B) School of Earth and Environmental Sciences, University of Adelaide, North Terrace, Adelaide, SA 5001, Australia.
(C) Corresponding author. Email: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Thomas, M.; Fitzpatrick, R.W.; Heinson, G.S.|
|Publication:||Australian Journal of Soil Research|
|Date:||May 1, 2009|
|Previous Article:||Genetic characterisation and symbiotic proprieties of native sinorhizobia trapped by Medicago sativa on Tunisian soils.|
|Next Article:||Carbon storage in a Ferrosol under subtropical rainforest, tree plantations, and pasture is linked to soil aggregation.|