Use of on-ground gamma-ray spectrometry to measure plant-available potassium and other topsoil attributes.
The incidence of potassium (K) deficiency is increasing in crops, pastures, and forestry in south-western Australia. Although soil K can be measured using soil sampling and analysis, [Gamma]-ray spectrometry offers a potentially cheaper and spatially more precise alternative. This could be particularly useful in precision agriculture, where inputs are applied according to need rather than by general prescription.
In a study of topsoils near Jerramungup, Western Australia, strong relationships ([r.sup.2] = 0.9) were found between on-ground counts of [Gamma]-rays derived from [sup.40]K ([Gamma]-K) and both total K and plant-available K. The success of [Gamma]-ray spectrometry in predicting available K relied on a strong relationship ([r.sup.2] = 0.9) between total K and available K which may not hold in all areas. Although the relationship between [Gamma]-K and available K held over the range of 36-1012 mg/kg, crop response to K fertilisers is only expected when the available K content is [is less than] 100 mg/kg. Estimates of available K from [Gamma]-K were unreliable at this lower end of the regression curve. Separate analysis with a subset of the data with available K [is less than] 100 mg/kg showed a poor relationship between [Gamma]-K and available K ([r.sup.2] = 0.05; d.f. 11). The usefulness of [Gamma]-ray spectrometry may therefore be restricted to defining areas where response to fertiliser K may occur, and where further soil sampling and analysis are required to predict the fertiliser requirement.
Strong relationships ([r.sup.2] = 0.9) were also found between [Gamma]-K and a range of other soil attributes, including clay, silt, and organic carbon content. These relationships depended on the locally strong relationship between total K and these soil attributes. Since such relationships do not hold everywhere, the utility of [Gamma]-ray spectrometry will likewise be limited. Site-specific calibrations are required if [Gamma]-ray spectrometry is to be used for soil property mapping.
Additional keywords: [sup.40]K, Colwell-K, precision agriculture, soil survey, clay, organic matter.
Growth limitations due to inadequate soil potassium (K) are an increasingly important land management issue in many soils of south-western Australia (M. T. F. Wong, N. Edwards, and N. J. Barrow unpublished data). Natural K reserves are often low because of prior intense weathering and leaching, resulting in a predominance of kaolinite and sesquioxides in the clay-sized fraction rather than K-enriched primary minerals (Robson and Gilkes 1981). For example, a study of 136 undisturbed sites across the south-western agricultural region (McArthur 1991) found that 55% of surface horizons had Colwell-extractable K values of [is less than] 100 mg/kg.
The export of K through agricultural production and leaching continues to deplete soil K reserves. The ratio of soil reserves to export in agricultural products is low, and K deficiency is now considered to be a threat to sustainable agriculture in the region. Responses to applied K have been reported in grain legumes and cereals (M. T. F. Wong, N. Edwards, and N. J. Barrow unpublished data), and tree plantations (J. F. McGrath, pets. comm.), whereas previously, K responses were mostly confined to annual pastures (Colwell and Grove 1976; Cox 1980). Potassium fertilisers will therefore be required as part of a balanced nutrition strategy for sustainable and profitable agricultural and forestry land use.
Efficient use of K fertilisers requires identification of areas where plant responses are likely to occur. In south-western Australia, this requirement is invariably determined through soil analysis using the Colwell (1963) sodium bicarbonate extraction method, which is also used for determining phosphorus requirement (Rayment and Higginson 1992). Soil Colwell-K contents can, however, be highly variable, both spatially and temporally. Spatial variation in available K can be ascribed to intrinsic variation in soil fertility, past fertiliser applications, land use, and redistribution by livestock. For example, in an intensively sampled area of 5000 ha near Jerramungup, Colwell-K values ranged from 21 to 1012 mg/kg soil, with clear spatial distribution patterns (Harper 1994). Similarly, K content varies seasonally; in one example on a sandy soil, exchangeable K values ranged from 20 mg/kg in spring to 55 mg/kg in summer due to cycling in the soil-plant system (Roberts 1968).
Precision agriculture technologies are developing that allow land managers to vary both the rate and type of fertiliser for maximum profit while minimising adverse environmental effect. These methods rely on improvements in accuracy of on-the-go positioning offered by differential global positioning systems (DGPS) (Cannon et al. 1994) and variable rate fertiliser (VRF) applications. For these systems to provide site-specific K fertiliser applications, an estimate of the soil K supply and its spatial variation is needed. Although this information may be represented as a map, and such maps can be produced by intensive
sampling, analysis, and appropriate interpolation procedures (Wollenhaupt et al. 1994), expense is an obvious practical limitation to this approach.
The nature of soil K offers the possibility of minimising the cost of sampling and mapping through remotely sensed measurements. Potassium in the Earth's crust contains ~0.012% of the radioactive isotope [sup.40]K. This isotope emits "/-rays with a diagnostic peak at 1.46 MeV which can be measured along with emission from uranium (U), thorium (Th), and their radioactive decay products (Ward 1981; IAEA 1991). The method has been used to determine the total K content of soils (Smith and Talibudeen 1981) and its regional distribution (Duval 1990), and for geologic mapping (Ward 1981). The technique has been utilised in several recent studies on Australian soils (Bierwirth 1996; Cook et al. 1996; Wilford et al. 1996). The measurement is often made with airborne detectors. With these detectors, about 66% of the counts come from an area of radius equal to the ground clearance (Ward 1981). For an aircraft flying at 60 m above ground level, 66% of the counts are derived from an area of ~1.1 ha. With the aircraft flying at 200 km/h, this `footprint' would be travelling at 56 m/s. The airborne [Gamma]-count is therefore an area averaged count. This averaging is useful when the relationship between the [Gamma]-counts and the surface attribute is already known. It is a problem when such a relationship is being investigated because spatial variability leads to an imprecise ground-based estimate of the mean for the area measured by the [Gamma]-ray spectrometer. In such a case, static ground-based measurements are expected to give a more precise correlation between counts and the soil attribute being investigated since the `footprint' is only a few metres wide. Once such a relationship is established, it should then be possible to reduce the cost of data acquisition by using airborne measurements. Issues to be considered when making airborne measurements, such as attenuation of [Gamma]-rays, speed of aircraft, and the effect of topography are discussed in more detail by Ward (1981).
Attenuation decreases the intensity of [Gamma]-rays, and the contribution of the subsoil to [Gamma]-ray emission at the soil surface is negligible. It is estimated that 91% of [Gamma]-rays measured at the surface of typical soils originate from the top 10 cm, and [is less than] 2% of the K emission coming from sources below 30 cm reach the soil surface (Spielberg 1975). Water also increases this attenuation significantly (Carroll 1981). [Gamma]-ray emission at the soil surface due to [sup.40]K is thus mostly derived from the top 10 cm. This is fortuitous, as this layer coincides with the depth of soil sampled for analysis to determine likely response of crops and pastures to K, and for which soil analysis values and corresponding plant response data exist (Cox 1980).
The aim of this work was to test the ability of a portable [Gamma]-ray spectrometer to measure plant-available K in the topsoil in a typical dryland farming environment of south-western Australia. The utility of the technique for discerning likely fertiliser requirement and its value in precision farming were determined. As [Gamma]-ray spectrometers measure total K, subsidiary aims were to determine the relationships between total and available K in these soils and between [Gamma]-ray counts from [sup.40]K ([Gamma]-K) and other soil attributes that are also related to the total K content. We also assessed whether the technique could have a wider role in soil property mapping as suggested by Bierwirth (1996).
Materials and methods
This work was undertaken in an area adjacent to Lake Cairlocup, near Jerramungup, 400 km south-east of Perth. The coordinates of the south-west corner of the study area were 33 [degrees] 45'S and 118 [degrees] 45'E. The soils and geomorphology of this area, which has a mean annual rainfall of 350 mm, are discussed in detail elsewhere (Harper 1994), and are typical of much of the central and eastern wheatbelt of Western Australia. The area lies east of the lake and comprises lunettes and playas, bounded by gently undulating hills formed on deeply weathered granitic rocks, with up to 40 m relief. This mantle has been modified by localised stripping and deposition. Farming involves annual rotation of cereals (Triticum aestivum, Hordeum vulgate) or legume crops (Lupinus angustifolius) and pasture (Trifolium subterraneum, Lolium rigidum).
Soils and sampling
From the 5000-ha study area of Harper (1994), we revisited 32 sample sites with a range of Colwell-extractable K in the topsoil (0-10 cm) of 36-1012 mg/kg. Composite samples previously had been taken from areas up to 0.25 ha and comprised 20-30 subsamples from a depth of 0-10 cm, the cultivated Ap horizon. Soil profiles to depths of 2 m were described and characterised by Harper (1994).
The major soils from this area are: (a) deep sands (Mesonatric Grey Sodosols, Isbell 1996; Uc2.21, Northcote 1979; Typic Quartzipsamments, Soil Survey Staff 1987) on dunes, adjacent swales, and sand sheets; (b) sandy duplex soils on lunettes and swales derived from aeolian and lacustrine sediments (Mesonatric Yellow and Brown Sodosols; Dy2.83, 4.83; Typic Natrixeralfs); and (c) loamy surfaced, strongly alkaline duplex soils (Hypernatric Grey Sodosols; Dy2.13; Calcixerollic Xerochrepts) also on lunettes and swales immediately adjacent to playas. Despite deeper reserves of K in some of these soils, their contribution to ground [Gamma]-ray measurements was expected to be small due to attenuation (Spielberg 1975).
Samples were air-dried and gently crushed, and a range of analyses was performed on the [is less than] 2 mm fraction. Particle size analysis was performed by the pipette method (Gee and Bauder 1986) and the clay ([is less than] 2 [micro]m) and silt (2-20 [micro]m) fractions were determined. Soil organic carbon content was determined by wet oxidation using a dichromate sulfuric acid mixture (Nelson and Sommers 1982). Exchangeable K (Exch-K) was determined by the method of Gillman and Sumpter (1986). Estimates of plant-available K (Colwell-K) and phosphorus (Colwell-P) were made using extraction with 0-5 M sodium bicarbonate as detailed in Methods 18A1 and 9B1 by Rayment and Higginson (1992). Electrical conductivity and pH were determined on 1:5 soil: water suspensions (Rayment and Higginson 1992). Tamm's reactive iron was determined using an acid oxalate extract (Rayment and Higginson 1992). All results are reported on an oven-dry basis.
Total K was measured by X-ray fluorescence (XRF) on pressed powdered samples. The samples were measured using a Sc/Mo tube, a LiF analysing crystal, and a proportional flow detector. The measurements were made against standard monitoring samples.
[Gamma] emissions were measured using an Exploranium GR 256 portable [Gamma]-ray spectrometer with a 2-L NaI crystal detector. At each location, the detector was placed on the ground and emission measured for 100 s. Replicate counts (3-4) were taken depending on the count rate. The counts were made with the detector oriented in the same direction each time. The data were corrected for background [Gamma]-ray and Compton scattering and equipment dead time (IAEA 1991). The measurements were undertaken in late summer in February and March 1996 when the soil was at its driest and contained [is less than] 3% water on a gravimetric oven-dry basis.
Results and discussion
Comparisons of measures of soil potassium
The different measures of soil K were strongly related to one another giving [r.sup.2] values ranging from 0.89 to 0.96 (Fig. 1). Colwell-K and exchangeable K values were also linearly related (Fig. 1a; [r.sup.2] = 0.96); when these data were presented in the same units, the slope of the relationship was close to unity, indicating that in these highly weathered soils both methods may have extracted the same soil K fraction in spite of the difference in size of the two extracting cations. Re-analysis of the study of Toms and Fitzpatrick (1961), who measured 87 soil samples representative of the medium rainfall (400-600 mm) areas of Western Australia, showed that K recovered with hydrochloric acid and K exchangeable with 1 M [NH.sub.4] were also linearly related ([r.sup.2] = 0.88). For Western Australian soils, Colwell-K thus appears to be as good a measure of plant-available K as exchangeable K or acid-extractable K. Colwell-K values are reported in this paper, as this method is most widely used as a measure of plant-available K in Western Australia.
[Figure 1 ILLUSTRATION OMITTED]
[Gamma]-ray spectrometry measures total K content irrespective of forms in which the K fractions occur. Log-linear relationships were obtained between total K and [Gamma]-K (Fig. 1b; [r.sup.2] = 0.94) and between total K and Colwell-K (Fig. 1c; [r.sup.2] = 0.89), the latter relationship suggesting that the proportion of non-exchangeable K increases as the total K content is increased. A strong linear relationship was obtained between Colwell-K and [Gamma]-K (Fig. 1d; [r.sup.2]= 0.93).
Despite the strong linear relationship between [Gamma]-K and Colwell-K, the practical utility of the [Gamma]-ray spectrometric method for estimating K fertiliser requirement may be limited. This is because crop response to K is only expected to occur at Colwell-K values [is less than] 100 mg/kg (Cox 1980). This represents the lower end of the regression curve which extended from 36 to 1012 mg/kg; and in this lower region, estimates of Colwell-K from [Gamma]-K were unreliable. Separate regression analysis with a subset of the data with Colwell-K [is less than] 100 mg/kg showed that the regression between Colwell-K and [Gamma]-K only accounted for 5% of the variation (d.f. 11).
The basis for the relationship between Colwell-K and [Gamma]-K was the strong relationship found between Colwell-K and total K in this study. This does not normally hold in soils in many other regions (Birtsch and Thomas 1985) and the [Gamma]-radiometric technique will likewise have limited utility in measuring plant-available K in those areas. Attempts to use [Gamma]-ray spectrometry to measure Colwell-K on a dissected landscape at Yornaning, in south-western Australia, for example, failed due to emission from exposed rocks and soil-forming materials that do not contribute to the pool of plant-available K (R. Glencross, pers. comm). Similar problems are expected with soils containing significant amounts of feldspars, since total and exchangeable K are not related in those soils. Soils at a Busselton site, for example, contained 1% K due to feldspars present in the sand fraction, but only small amounts of exchangeable K resulting in acute K deficiency in pastures (Drover 1961).
The value of [Gamma]-ray spectrometry in precision farming will lie in identifying areas where soil sampling and analysis should proceed, rather than in providing site-specific K fertiliser recommendations. In this role, it will allow cost savings relative to sampling and analysing the whole area. The accuracy of [Gamma]-ray spectrometry may be improved by using larger detector crystal sizes and longer counting times. However, this is not compatible with on-the-go measurements. Reductions in sensitivity would also occur if measurements were made from airborne rather than ground platforms, and if measurements were made of moist soils such as would occur with variable rate fertiliser technology attached to seeders. It should be recognised, however, that even with ultimate precision in defining soil K concentrations, the relationships between soil K supply and plant fertiliser response are imprecise (Colwell and Grove 1976; Cox 1980).
Relationships between [Gamma]-radiometric counts for potassium and other topsoil attributes
At Jerramungup, the total K content was dependent on the clay content [log total K (%)= 1.17 log clay (%) -1-59; [r.sup.2]= 0.87; d.f. 21]; hence, [Gamma]-ray emissions also enabled clay content to be determined. The regression between total K and clay content indicated that the clay contained approximately 5% K. The predominant clay minerals in these soils were illite and kaolinite (Harper 1994) with the illite being the principal source of non-exchangeable K.
Table 1 summarises the relationships between [Gamma]-radiometric counts for K and an array of other soil attributes. Spearman rank correlation (Snedecor and Cochran 1978) was used as a means of ready comparison between the bivariate relationships, as those between different attributes are often non-linear. Strong relationships between [Gamma]-K values and each of the other attributes are evident, with rank correlation coefficient ([Rho]) ranging from 0.78 to 0.93. Selected bivariate relationships are illustrated in Fig. 2; here, parametric statistics were used to determine correlation coefficients.
[Figure 2 ILLUSTRATION OMITTED]
Table 1. Spearman rank correlation coefficients ([Rho]) between various soil properties (0-10 cm sample) and ground-measured [Gamma]-ray counts for [sup.40]K
OC, organic carbon; EC, electrical conductivity; Exch. Na, exchangeable sodium
K Na P OC Fe EC [Gamma]-K (counts/100 s) 1.00 Exch. Na (cmol/kg) 0.86 1.00 Colwell-P ([micro]g/g) 0.82 0.81 1.00 Organic carbon (%) 0.88 0.93 0.83 1.00 Tamm's Fe ([micro]g/g) 0.87 0.87 0.89 0.89 1.00 EC (dS/m) 0.81 0.92 0.85 0.91 0.89 1.00 pH 0.78 0.79 0.83 0.79 0.82 0.82 Silt (%) 0.93 0.92 0.84 0.94 0.92 0.90 Clay (%) 0.92 0.93 0.84 0.92 0.89 0.89 pH Silt Clay [Gamma]-K (counts/100 s) Exch. Na (cmol/kg) Colwell-P ([micro]g/g) Organic carbon (%) Tamm's Fe ([micro]g/g) EC (dS/m) pH 1.00 Silt (%) 0.76 1.00 Clay (%) 0.77 0.96 1.00
These strong relationships between [Gamma]-K values and other soil attributes suggest a potential role for [Gamma]-ray spectrometry in soil property mapping. Soil clay content was strongly related to [Gamma]-K (Fig. 2a; [r.sup.2]= 0.92). [Gamma]-ray activities in the upper soil layer are often positively correlated with clay content as clays can retain radio-isotopes (Martz and de Jong 1990). In this study, the clay itself was the source of the isotope. Clay content is in turn related to an array of other factors including water retention (Harper 1994), water repellence (Harper and Gilkes 1994b), wind erosion risk (Harper and Gilkes 1995), and hardsetting (Harper and Gilkes 1994a). Thus, in this environment, the technique may have a wider role in predicting the distribution of the risk of various forms of land degradation.
Organic carbon content is an important indicator of soil fertility. As clay stabilises soil organic matter and allows larger amounts to be maintained in soils, the relationship between clay content and organic carbon content meant that soil organic matter could also be measured by [Gamma]-ray spectrometry (Fig. 2c; [r.sup.2] = 0.89). Tamm's reactive iron content is used to determine the reactivity of the soil to phosphorus and the amounts required to correct deficiency (Lewis et al. 1981). The Tamm's reactive iron content could also be measured because of its relationship with the clay content of those soils (Fig. 2d; [r.sup.2]= 0.89). Strong relationships also occurred between [Gamma]-K and Colwell-P (Fig. 2e; [r.sup.2]= 0.68) and pH (Fig. 2f; [r.sup.2]= 0.69); however, the latter relationship is spurious.
[Gamma]-K was strongly related with other soil attributes in this study as the samples were taken along an aeolian deposition gradient from a local playa, Lake Cairlocup. Silt, clay, carbonate, and exchangeable sodium deflated from the playa floor were co-deposited from aeolian dusts (`parna'), and amounts decreased exponentially with distance from the source (Harper 1994). [Gamma]-K values decreased similarly (Fig. 3). Such erosion/deposition systems are a common feature of the broad inland drainage lines in the central and eastern wheatbelt of Western Australia (Bettenay 1962). This co-deposition explains the strong inter-relationships between values of various soil attributes (Table 1), including those for which there is no physical basis, such as that between silt content and [Gamma]-K (Fig. 2b; [r.sup.2]= 0.89). The silt was mostly quartz, the strong correlation being obtained because the soil silt and clay contents were strongly correlated ([r.sup.2] = 0.96). This model also explains the relationship between [Gamma]-K and pH (Fig. 2f), this being coincidental due to the exponential decrease in both carbonate and [Gamma]-K with distance from the playa.
[Figure 3 ILLUSTRATION OMITTED]
A general feature of the bivariate relationships shown in Fig. 2 is a deterioration of the correlations when a subset of the data with [Gamma]-K values [is less than] 100 counts/100 s is analysed. The error in the [Gamma]-ray count is estimated from its standard deviation (Ward 1981). At the lowest count rate (50 counts/100 s) measured at the site, the uranium and thorium counts were also negligible; the percentage error of the [sup.40]K counts was in the order of 15%. This is the same order of magnitude as the maximum error attributable to chemical analysis. Lack of correlation at low [Gamma]-K values may be due partly to poor relationships, rather than lack of instrument sensitivity, since the soil contained [is less than] 5% clay in this range of [Gamma]-K. Thus, the argument supporting the relationship between clay content and organic matter content may not be valid.
Gauging the usefulness of [Gamma]-ray spectrometry for soil survey and management is like evaluating the worth of morphological characteristics (such as field texture, colour, and structure) in general purpose soil surveys. The fundamental assumption of soil survey is that a range of simple-to-measure field attributes are surrogates of other more expensive and difficult-to-measure attributes (Mulcahy and Humphries 1967; Butler 1980). Where this assumption is correct, and underlying relationships occur, soil surveys (and [Gamma]-ray spectrometry) can provide insights into a broad array of other soil attributes, such as soil fertility, water-holding capacity, and erodibility. Conversely, where these relationships are poor or non-existent, soil survey, and [Gamma]-ray spectrometry, will be of limited value (Gibbons 1961; Webster and Butler 1976). Thus, no increase in precision in field measurements (or instrument output) will help in improving the usefulness of the technique. Poor relationships between [Gamma]-ray spectrometric values and other soil attributes are expected where the other soil attributes (a) are strongly affected by management practices or (b) vary markedly over time.
Due to local combinations of geology and pedologic history such as those found in Jerramungup, the technique may be useful in some environments for predicting gross soil properties. Some of these relationships may be coincidental, such as that between [Gamma]-K and silt in this study, [Gamma]-ray spectrometry thus cannot be used without local calibration, as suggested for electromagnetic meters for salinity measurement by Lesch et al. (1992). Where [Gamma]-ray spectrometry is used on a regional scale, different calibrations will be required for different pedo-geomorphic zones. Further work is required to identify those soil landscapes where [Gamma]-ray spectrometry does provide useful surrogates of other soil properties.
We are grateful to the GRDC, which partly funded this work as part of project CSO 148 `Management of sustainable potassium reserves in duplex and sandplain soils of the west Australian cropping zones'. Some of the data presented here were collected under the aegis of a Grain Research Committee of WA/WA Barley Industry Research Committee postgraduate award supervised by Professor Bob Gilkes, The University of Western Australia. We are grateful to Dr S. Cook of CSIRO Land and Water for introducing us to the subject and for his kind support. All landholders are thanked for access to their land and enthusiastic support of this work.
Bettenay, E. (1962). The salt lake systems and their associated aeolian features in the semi-arid areas of Western Australia. Journal of Soil Science 13, 10-17.
Bierwirth, P. (1996). Investigation of airborne gamma-ray images as a rapid mapping tool for soil and land degradation--Wagga Wagga, NSW. Australian Geological Survey Organisation, Record 1996/22.
Birtsch, P. M., and Thomas, G. W. (1985). Potassium status of temperate region soils. In `Potassium in Agriculture'. (Ed. R. D. Munson.) pp. 131-62. (American Society of Agronomy-Crop Science Society of America-Soil Science Society of America: Madison, WI.)
Butler, B. E. (1980). `Soil Classification for Soil Survey.' (Clarendon Press: Oxford.)
Cannon, M. E., McKenzie, R. C., and Lachapelle, G. (1994). Soil salinity mapping with electromagnetic induction and satellite-based navigation methods. Canadian Journal of Soil Science 74, 335-43.
Carroll, T. R. (1981). Airborne soil moisture measurement using natural terrestrial gamma radiation. Soil Science 132, 358-66.
Colwell, J. D. (1963). The estimation of phosphorus fertiliser requirements of wheat in southern New South Wales by soil analysis. Australian Journal of Experimental Agriculture and Animal Husbandry 3, 190-7.
Colwell, J. D., and Grove, T. S. (1976). Assessment of potassium and sulphur fertiliser requirements of wheat in Western Australia. Australian Journal of Experimental Agriculture and Animal Husbandry 16, 748-54.
Cook, S. E., Corner, R. J., Groves, P. R., and Grealish, G. J. (1996). Use of airborne gamma radiometric data for soil mapping. Australian Journal of Soil Research 34, 183-94.
Cox, W. J. (1980). Potassium for crops and pastures in medium and low rainfall areas. Journal of Agriculture, Western Australia 21, 12-18.
Drover, D. P. (1961). The potassium status of some coarse textured soils of the south-western region of W.A. Australian Journal of Experimental Agriculture and Animal Husbandry 1, 11-14.
Duval, J. S. (1990). Modern aerial gamma-ray spectrometry and regional potassium map on the conterminous United States. Journal of Geochemical Exploration 39, 249-53.
Gee, G. W., and Bauder, J. W. (1986). Particle-size analysis. In `Methods of Soil Analysis. Part 1. Physical and Mineralogical Methods'. 2nd Edn. (Ed. A. Klute.) Agronomy Monograph No. 9. pp. 383-411. (American Society of Agronomy-Soil Science Society of America: Madison, WI.)
Gibbons, F. R. (1961). Some misconceptions about what soil surveys can do. Journal of Soil Science 12, 96-100.
Gillman, G. P., and Sumpter, E. A. (1986). Modification to the compulsive exchange method for measuring exchange characteristics of soils. Australian Journal of Soil Research 24, 61-6.
Harper, R. J. (1994). The nature and origin of the soils of the Cairlocup area, Western Australia, as related to contemporary land degradation. PhD Thesis, The University of Western Australia.
Harper, R. J., and Gilkes, R. J. (1994a). Hardsetting in the surface horizons of sandy soils and its implications for soil classification and management. Australian Journal of Soil Research 32, 603-19.
Harper, R. J., and Gilkes, R. J. (1994b). Soil attributes affecting water repellency and the utility of soil survey in predicting its occurrence. Australian Journal of Soil Research 32, 1109-24.
Harper, R. J., and Gilkes, R. J. (1995). The incidence of wind erosion in relation to the properties of some sandy surfaced soils from south-western Australia. Journal of Arid Land Studies 55, 323-6.
IAEA (1991). Airborne gamma ray spectrometer surveying. International Atomic Energy Agency, Technical Report 323.
Isbell, R. F. (1996). `The Australian Soil Classification System.' Australian Soil and Land Survey Handbook Series. (CSIRO Publishing: Collingwood.)
Lesch, S. M., Rhoades, J. D., Lund, L. J., and Corwin, D. L. (1992). Mapping soil salinity using calibrated electromagnetic measurements. Soil Science Society of America Journal 56, 540-8.
Lewis, D. C., Clarke, A. L., and Hall, W. B. (1981). Factors affecting the retention of phosphorus applied as superphosphate to the sandy soils in south-eastern South Australia. Australian Journal of Soil Research 19, 167-74.
Martz, L. W., and de Jong, E. (1990). Natural radionuclides in the soils of a small agricultural basin in the Canadian prairies and their associations with topography, soil properties and erosion. Catena 17, 85-96.
McArthur, W. M. (1991). `Reference Soils of South-western Australia.' (Australian Society of Soil Science Inc., WA Branch: Perth.)
Mulcahy, M. J., and Humphries, A. W. (1967). Soil classification, soil surveys and land use. Soils and Fertilizers 30, 1-8.
Nelson, D. W., and Sommers, L. E. (1982). Total carbon, organic carbon and organic matter. In `Methods of Soil Analysis. Part 2. Chemical and Microbiological Methods'. 2nd Edn. (Eds A. L. Page, R. H. Miller and D. R. Keeney.) Agronomy Monograph No. 9. pp. 539-79. (American Society of Agronomy-Soil Science Society of America: Madison, WI.)
Northcote, K. H. (1979). `A Factual Key for the Recognition of Australian Soils.' 4th Edn. (Rellim Technical Publishers: Glenside, S. Aust.)
Rayment, G. E., and Higginson, F. R. (1992). `Australian Laboratory Handbook of Soil and Water Chemical Methods.' Australian Soil and Land Survey Handbook. (Inkata Press: Melbourne.)
Roberts, F. J. (1968). The annual pattern of potassium release from organic matter in sandy soils of south-west Australia. Australian Journal of Soil Research 6, 193-202.
Robson, A. D., and Gilkes, R. J. (1981). Fertiliser responses (N, P, K, S, micronutrients) on lateritic soils in southwestern Australia--a review. In `Laterisation Processes'. pp. 381-90. (A.A. Balkema: Rotterdam.)
Smith, G. J., and Talibudeen, O. (1981). The [sup.40]K and [sup.214]Bi contents of Rothamsted soils determined by gamma spectrometry. Journal of Soil Science 32, 593-603.
Snedecor G. W., and Cochran W. G. (1978). `Statistical Methods.' 6th Edn. (The Iowa State University Press: Aimes, IA.)
Soil Survey Staff. (1987). `Keys to Soil Taxonomy.' 3rd Edn. SMSS Technical Monograph No. 6. (Cornell University: Ithaca, NY.)
Spielberg, D. (1975). Soil. In `Engineering Compendium on Radiation Shielding. Vol. II. Shielding Materials'. (Eds R. G. Jaeger E. P. Blizard, A. B. Chilton, M. Grotenhuis, A. Honig, T. A. Jaeger, and H. H. Eisenlohr.) pp. 31-3. (Springer-Verlag: Berlin.)
Toms, W. J., and Fitzpatrick, E. N. (1961). Potassium deficiency in medium rainfall areas. Journal of Agriculture, Western Australia 2, 763.
Ward, S. H. (1981). Gamma-ray spectrometry in geologic mapping and uranium exploration. Economic Geology (75th Anniversary Volume), 840-9.
Webster, R., and Butler, B. E. (1976). Soil classification and survey studies at Ginninderra. Australian Journal of Soil Research 14, 1-24.
Wilford, J. R., Bierwirth, P. N., and Craig, M. A. (1996). Application of airborne gamma-ray spectrometry in soil/regolith mapping and applied geomorphology. Australian Geological Survey Organisation Journal 17, 201-16.
Wollenhaupt, N. C., Wolkowski, R. P., and Clayton, M. K. (1994). Mapping soil test phosphorus and potassium for variable-rate fertilizer application. Journal of Production Agriculture 7, 441-8.
Manuscript received 1 May 1998, accepted 4 December 1998
M. T. F. Wong(A)(C) and R. J. Harper(B)
(A) CSIRO Land and Water, Private Bag, PO Wembley, WA 6014, Australia.
(B) CALM Science, Department of Conservation and Land Management, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia.
(C) Corresponding author; email: M.Wong@ccmar.csiro.au
|Printer friendly Cite/link Email Feedback|
|Author:||Wong, M. T. F.; Harper, R. J.|
|Publication:||Australian Journal of Soil Research|
|Date:||Mar 1, 1999|
|Previous Article:||Comparison of peroxide-oxidisable sulfur and chromium-reducible sulfur methods for determination of reduced inorganic sulfur in soil.|
|Next Article:||Using active fractions of soil organic matter as indicators of the sustainability of Ferrosol farming systems.|