Trouvez votre prochain book favori
Devenez membre aujourd'hui et lisez gratuitement pendant 30 joursCommencez vos 30 jours gratuitsInformations sur le livre
Soil Physical Measurement and Interpretation for Land Evaluation
Actions du livre
Commencer à lire- Éditeur:
- CSIRO PUBLISHING
- Sortie:
- Nov 26, 2002
- ISBN:
- 9780643099593
- Format:
- Livre
Description
*practical guidelines on the soil physical measurements across a range of soils, climates and land uses;
*straightforward descriptions for each method (including common pitfalls) that can be applied by people with a rudimentary knowledge of soil physics, and
*guidelines on the interpretation of results and integration with land resource assessment.
Soil Physical Measurement And Interpretation for Land Evaluation begins with an introduction to land evaluation and then outlines procedures for field sampling. Twenty detailed chapters cover pore space relations, water retention, hydraulic conductivity, water table depth, dispersion, aggregation, particle size, shrinkage, Atterburg limits and strength. The book includes procedures for estimating soil physical properties from more readily available data and shows how soil physical data can be integrated into land planning and management decisions.
Informations sur le livre
Soil Physical Measurement and Interpretation for Land Evaluation
Description
*practical guidelines on the soil physical measurements across a range of soils, climates and land uses;
*straightforward descriptions for each method (including common pitfalls) that can be applied by people with a rudimentary knowledge of soil physics, and
*guidelines on the interpretation of results and integration with land resource assessment.
Soil Physical Measurement And Interpretation for Land Evaluation begins with an introduction to land evaluation and then outlines procedures for field sampling. Twenty detailed chapters cover pore space relations, water retention, hydraulic conductivity, water table depth, dispersion, aggregation, particle size, shrinkage, Atterburg limits and strength. The book includes procedures for estimating soil physical properties from more readily available data and shows how soil physical data can be integrated into land planning and management decisions.
- Éditeur:
- CSIRO PUBLISHING
- Sortie:
- Nov 26, 2002
- ISBN:
- 9780643099593
- Format:
- Livre
À propos de l'auteur
En rapport avec Soil Physical Measurement and Interpretation for Land Evaluation
Aperçu du livre
Soil Physical Measurement and Interpretation for Land Evaluation - Keppel Coughlan
114.
1
Soil Physical Measurement for Land Evaluation
KJ Coughlan and NJ McKenzie
1.1 PURPOSE OF THE HANDBOOK
The primary aim of this Handbook is to encourage more physical measurement of soil in land resource survey. Such measurements provide many benefits and they are necessary inputs for computer simulation models. These models are enhancing or replacing conventional methods of land evaluation because they allow better analysis of interactions between climate and landscape processes. The Handbook is intended for several broad groups:
• Those directly responsible for making soil physical measurements in the field and laboratory;
• Land resource survey practitioners;
• Students and researchers seeking efficient methods for soil physical measurement; and
• Groups responsible for making decisions on sustainable land use.
This Handbook concentrates on soil physical measurement methods that are cost-effective and well suited to land resource survey. The majority of methods in the Handbook require only one or two visits to a site over a short period. Methods that are more expensive in both time and resources are well described in standard manuals (e.g. Klute 1986; Carter 1993; Smith and Mullins 2001; Das 2002). The Handbook focuses on practical aspects of measurement and, wherever possible, guidance is provided on the interpretation of data.
The Handbook begins with an introduction to land evaluation and then outlines procedures for field sampling. The following 20 chapters provide details on individual measurement methods. These cover pore space relations, water retention, hydraulic conductivity, watertable depth, dispersion, aggregation, particle size, shrinkage, Atterberg limits and strength. This is followed by an account of procedures for estimating soil physical properties from more readily available data. Finally, several applications are presented to show how soil physical data can be integrated and used for land planning and management decisions.
While a large effort has been devoted to routine soil chemistry in land resource survey (e.g. Rayment and Higginson 1992; Peverill et al. 1999), routine soil physical characterisation in Australia has been rare. Cost has often been cited as a reason, but this is questionable given the importance of the soil water balance in many land planning and management issues. In our view, the lack of robust methods, particularly for soil hydraulic properties, has been an impediment to routine measurement. This Handbook presents methods that overcome many past deficiencies – although significant methodological issues remain, particularly with respect to scales of measurement and interpretation.
1.2 APPROACHES TO LAND EVALUATION
A spectrum of approaches to land evaluation exists and each provides a role for soil physical measurement. It is useful to order these approaches according to the degree to which they rely on scientific principles. The least scientific uses trial and error and has no formal way of organizing the experience gained to benefit other land users. Purely empirical methods are better, but have limited applicability and can be unreliable. More scientific methods are based on models of natural processes with varying levels of complexity. The following account draws heavily on Nix (1968), Basinski (1985) and McKenzie (1991).
1.2.1 Trial and error
Trial and error is an informal system for land evaluation. It is the oldest and still the most widely used form of land evaluation. Most systems of land use in Australia were established by trial and error because reliable and useful information from land resource assessment was unavailable. However, the lack of information has led to high economic, social and environmental costs. A serious deficiency with the trial and error approach is that experience is inadequately recorded and the prospects for developing rational strategies of land use are limited, particularly when new areas are developed, untried land uses are attempted, or land managers change. Soil physical measurement is used, but on an ad hoc basis, usually to identify problems after they have developed.
1.2.2 Empirical land evaluation relying on transfer by analogy
Most programs of land resource survey have relied on transfer by analogy. This approach recognises that the results of a land use trial (e.g. farmer’s experience, field experimental results, small catchment study, etc.) are strictly applicable to that site only. To seek generality, results are transferred by analogy by assuming that all occurrences of a particular class of land (i.e. the land analogue) will respond similarly under the same use. The success of the approach relies heavily on the classification and identification of land analogues. These analogues may be defined using classification systems for land (e.g. land system, soil landscape, capability class) or soil (e.g. taxonomic class of a local or national system such as the Australian Soil Classification (Isbell 1996)). Transfer by analogy works well when the criteria used for defining and partitioning land analogues can be readily mapped and are well correlated with attributes influencing land use. Most mapping programs depend heavily on soil morphological descriptions (McDonald et al. 1990) for defining soil and land units. Unfortunately, relationships between soil morphology and physical properties are complex and sometimes poor (see Chapter 22). As noted earlier, routine soil physical measurements (apart from particle size analysis) have not been common in most Australian mapping programs.
1.2.3 Semi-empirical land evaluation
Land resource scientists in Australia have begun to realize that taxonomic distinctions made by either national classification systems or generalized land capability systems form an unreliable basis for land evaluation. This has led to the development of procedures where the potential for a nominated land use across a mapping unit is assessed using soil and land characteristics or qualities. Individual polygons mapped during a survey have attached estimates for each soil and land characteristic or quality, as well as a description of the dominant soil taxa. The assessment of land suitability for each polygon is usually based on the most limiting factor.
While many of the soil and land characteristics or qualities relate to soil physical properties (e.g. profile available water capacity, erodibility, permeability), there has been limited direct field or laboratory measurement to support such interpretations.
1.2.4 Land evaluation using process models
The best theoretical approach to land evaluation is based on process modeling. However, its practical superiority is only just starting to be evident in routine soil survey and land evaluation, despite being advocated for many years (e.g. Nix 1968, 1981). In process modeling, land performance (in terms of productivity, hazard of use or management inputs required) is related to individual soil and land characteristics or qualities, and their net effect is assessed using a model of land function. These models may portray specific processes such as water movement (e.g. Verburg et al. 1997), or they may be more comprehensive and model particular farming systems (e. g. Littleboy et al. 1989; Moore et al. 1997; McCown et al. 1996).
Process modeling recognises the complex relationships between land characteristics and utilisation, and an attempt is made to represent these explicitly. Conventional approaches to land evaluation tend to be static, and the implicit model relating land qualities to land performance is commonly stated in verbal or qualitative terms. In contrast, process modeling allows land evaluation to be quantitative, dynamic and probabilistic. In particular, interactions between soil and climate can be more fully appreciated (see Chapters 24 and 25).
Most process models suited to land evaluation require soil physical data as inputs. However, there will be many instances where soil physical properties will have to be estimated using more readily available data. This is particularly so for hydraulic properties (see Chapter 22).
Process modeling requires measurement to be undertaken at scales appropriate to the process of interest. Many contemporary natural resource management problems require a predictive capability at a range of scales (e.g. plot, paddock, farm, small catchment, region). The methods of measurement in this Handbook are most useful at finer scales. Efficient and robust physical measurement methods for characterising soil related processes at broader scales are in their infancy. Chapters 2 and 22 consider issues of scale in more detail.
1.3 SOIL PHYSICAL DATA, LAND CHARACTERISTICS AND LAND QUALITIES
The FAO Framework for Land Evaluation (FAO 1976; Dent and Young 1981; Beek 1981) provides a useful terminology for both empirical land evaluation and process modeling. The term land characteristic refers to an attribute of land that can be measured or estimated. A land quality is a complex attribute of land that acts in a manner distinct from the actions of other land qualities in its influence on the suitability of land for a specific use. Land qualities take account of interactions between relevant land characteristics. In the following sections, land qualities using soil physical data (i.e. characteristics) are summarised. Conventional practice in Australian land evaluation is summarised along with the proposed changes outlined in this Handbook.
1.3.1 Water availability
Profile available water capacity has routinely been estimated from field texture, soil structure and the occurrence of layers that limit root growth. In some cases, these estimates are supplemented by measurements of volumetric water contents at potentials of either –1.0 m or –3.3 m (notional field capacity) and –150 m (notional wilting point, but see Chapters 3 and 4). The available water capacity has been equated with the difference between the two water contents. The plant available water capacity has been calculated for the soil profile by summing the available water capacity over either the depth of rooting or a standard depth. Simple monthly water balances have sometimes been used to detect periods of limited water availability.
The latter approach is supported in this Handbook, with the soil water retention curve being determined on undisturbed soil cores at small negative potentials and small aggregates or sieved soil at large negative potentials. It is recognised that water contents at –1.0 m and –150 m may not correspond with upper and lower limits attained in the field. However, for survey purposes there are few practical alternatives. In some circumstances, to gain a better estimate of the upper limit it may be possible to undertake direct measurements of volumetric water contents for soil profiles after extended periods of heavy rain. To provide estimates of water availability and, in particular, periods of water deficit, we strongly encourage the use of simulation models. Estimates of soil water availability should take into account limitations to root growth associated with poor aeration and excessive soil strength – this is achieved using the non-limiting or least limiting water range (Letey 1985, 1991; Da Silva et al. 1994; Groenevelt et al. 2001).
1.3.2 Excess wetness
Excess wetness has been estimated in routine land resource surveys using landscape position for external drainage and permeability classes for the internal drainage of soils (McDonald et al. 1990). Soil colour has also been used, particularly with respect to A2 horizons and mottles. Digital elevation models are starting to be used to estimate external drainage. There has been only very limited measurement of soil hydraulic properties, particularly saturated hydraulic conductivity, to help determine excess wetness.
This Handbook encourages the direct measurement of soil hydraulic properties (see Chapter 6). When used in conjunction with soil water retention curves (Chapters 4 and 22) and related soil profile information, such measurements can be used as inputs to simulation models to estimate excess wetness.
1.3.3 Water repellence
Water repellence (Chapter 5) has been estimated in conventional land resource survey using local experience or interpretations of standard morphological data, particularly field texture. There has been some restricted laboratory testing. In this Handbook, a simple laboratory determination is recommended for assessing the degree of water repellence.
1.3.4 Limited or excessive levels of hydraulic conductivity
Very slow hydraulic conductivity may result in waterlogging (Section 1.3.2) and, by limiting recharge of profile water, it can also cause excess runoff or problems in irrigation. In contrast, high levels of hydraulic conductivity can be detrimental in paddy rice culture and can increase the potential for mobilisation and transport of contaminants (e.g. nitrates). As with excess wetness, existing methods of assessment rely heavily on descriptions of soil morphology and landform.
Recharge estimates can be undertaken using a range of measurement methods – including tracers (Cook and Herczeg 1999) and various hydrological techniques – but these methods have not been part of routine land resource surveys and they are beyond the scope of this Handbook (see Zhang and Walker 1998). Here we encourage the direct measurement of hydraulic conductivity either using large soil cores or field methods. Rainfall simulator measurements (Chapter 11) have advantages in situations where surface seals form during rainfall but the methods demand significant resources. Standard soil dispersion measurements (Chapters 13 and 14) provide a further useful source of information.
1.3.5 Water erosion
Most land resource assessment in Australia provides estimates of hazards associated with water erosion. Existing diagnostic attributes include the occurrence of erosive rainfall, slope angle, natural slope length, local vegetative cover associated with the proposed land use, estimated soil erodibility and local experience. In this Handbook, methods are provided for improved estimation of soil erodibility parameters.
1.3.6 Wind erosion
The potential for wind erosion is estimated by most land resource assessment practitioners through an analysis of the occurrence of erosive winds, site exposure to wind, likely surface cover for the proposed land use and a measure of soil texture. Some agencies augment the estimate of soil erodibility through measurement of dry aggregate size distributions and aggregate strength. This Handbook provides a simple method for measuring dry aggregate size distribution that deserves to be more widely adopted.
1.3.7 Excessive soil volume change
Excessive soil volume change associated with variations in water content can create engineering problems (e.g. foundations, buried cables). When used for agriculture, these soils have some advantages (they can recover from compaction more readily) and disadvantages (soil-water management can be difficult). Methods of soil physical measurement and analysis have to be modified for soils with excessive soil volume change (see Chapters 3 and 4). Direct measurement of parameters relating to the potential for shrinking and swelling has not been widespread in Australian land resource surveys, although it has been a routine part of site investigations for geotechnical purposes. This Handbook provides several simple methods for measuring soil shrinkage (Chapter 18).
1.3.8 Poor trafficability and narrow soil water range for cultivation
Soils that have poor trafficability and a narrow range of water content in which cultivation is beneficial, are difficult to manage and susceptible to compaction. Assessment in conventional surveys has depended on soil type and local experience. In this Handbook, water balance modeling in conjunction with information on the plastic and liquid limit (Chapter 19) are recommended as a means for determining the frequency and length of cultivation windows (i.e. periods when cultivation is beneficial and soil compaction is avoided).
1.3.9 Excessive strength in shallow layers of the soil profile
The presence of pans and layers that impede root growth has been determined qualitatively in most land resource assessment in Australia. Diagnostic attributes indicating excessive strength have relied on field morphological observations, sometimes supplemented by measurements of bulk density and particle size. This Handbook recommends the use of micro-penetrometer measurements on soil cores at controlled potentials to establish relationships between penetration resistance and water content. While other measurement methods have advantages (e.g. shear vanes), they are more difficult to incorporate into a routine program of measurement in land resource surveys. Measurements of soil strength are used as inputs to calculate the non-limiting or least limiting water range (Letey 1985, 1991; Da Silva et al. 1994; Groenevelt et al. 2001).
1.3.10 Surface condition
Many soils in Australia have surface limitations that restrict infiltration or impair crop establishment. Hard setting occurs in lighter textured soils, while crusting and coarse dry aggregates in the soil surface occur in heavier textured soils. Existing diagnostic attributes for these problems include local experience, field morphological description (e.g. condition of surface soil, structure size, grade and type) and several methods for measuring aggregate stability. This Handbook encourages increased quantification of surface condition measurement using a suite of methods suited to different soil types. Modulus of rupture (Chapter 21) is used to estimate hard-setting, and rainfall simulation is recommended for measurement of seal formation during wetting (Chapter 11). Soil shrinkage measures are recommended for the estimation of self-mulching capacity (Chapters 18 and 29). The field methods of Tongway (1994) and Tongway and Hindley (1995) should also be considered.
1.3.11 Rockiness
The presence of large quantities of coarse fragments creates a series of limitations that relate to trafficability, cultivation and water availability. Procedures for estimating the proportion of coarse fragments are considered in Chapter 2.
1.4 METHOD CODES
Table 1.1 provides a list of codes for each soil physical measurement method presented in this Handbook. These codes must be entered into soil databases in conjunction with the actual measurements. The first digit in the code is 5 and it refers to the volume number of this Handbook. The second and third digits refer to the measurement method via the chapter number, while digits after the decimal refer to variants of each method.
The results of some soil physical properties (e.g. hydraulic conductivity, clay dispersion) depend strongly on the measurement method, and any analysis must take this into account. In our view, the lack of information on measurement methods has severely limited the capacity for reanalysis of soil physical data in Australia.
1.5 MINIMUM DATA SET
The selection of measurement methods depends on the purpose of a study and the resources available. If ample resources are available, then soil physical measurement methods that are more comprehensive than those presented in this Handbook should be considered. However, resource limitations associated with most land resource survey programs will usually constrain soil physical measurement. As noted earlier, the development of comprehensive soil physical databases generates many benefits. Comprehensive soil physical databases are most effective when consistent methods have been used for measurement. A key task is specifying and adhering to a minimum data set. The minimum data presented in Table 1.2 is for reference sites, or modal profiles, in Australian land resource surveys. It may act as a useful guide for other characterisation studies, for example, those relating to field sites used for research and monitoring (e.g. agronomic experiments, hillslope and small catchment studies, ecological studies). The minimum data set should in no way constrain collection of soil physical data using other methods presented in this Handbook or other manuals (e.g. soil engineering tests). Note that some soil properties have not been included in the minimum data set because they are specific to particular groups of soils. It may be beneficial at a later time to specify minimum data sets for well-defined groups of soils and related environments.
The minimum data set presented in Table 1.2 should be collected for at least two and preferably three or more layers in the modal profile (see Section 2.1.5).
Table 1.1: Codes for soil physical measurement methods in this Handbook
Table 1.2: Minimum soil physical data set for modal profile characterisation in Australian land resource survey.
¹ Clay+silt >25% is equivalent to the porphyric fabrics of Brewer (1979)
1.6 REFERENCES
Basinski JJ (1985) Land evaluation – some general considerations. In ‘Environmental Planning and Management.’ In ‘Proceedings of a Commonwealth Science Council workshop, Canberra 1984.’ (CSIRO Division of Water and Land Resources: Canberra).
Beek KJ (1981) From soil survey interpretation to land evaluation. Part 2. From the present to the future. Soil Survey and Land Evaluation 1, 18–25.
Brewer R (1979) Relationships between particle size, fabric and other factors in some Australian soils. Australian Journal of Soil Research 17, 29–41.
Cook PG, Herczeg AL (1999) ‘Environmental tracers in subsurface hydrology.’ (Kluwer Academic Press: Boston).
Carter MR (Ed.) (1993) ‘Soil sampling and methods of analysis.’ (Lewis Publishers: Boca Raton, Florida).
Da Silva AP, Kay BD, Perfect E (1994) Characterisation of the least limiting water range of soils. Soil Science Society of America Journal 58, 1775–1781.
Das BM (2002) ‘Soil mechanics laboratory manual (6th edn).’ (Oxford University Press: New York).
Dent D, Young A (1981) ‘Soil survey and land evaluation.’ (George Allen and Unwin: London).
FAO (1976) ‘A framework for land evaluation.’ Soils Bulletin 32 (FAO: Rome).
Groenevelt PH, Grant CD, Semetsa S (2001) A new procedure to determine soil water availability. Australian Journal of Soil Research 39, 577–598.
Isbell RF (1996) ‘The Australian Soil Classification.’ (CSIRO Publishing: Melbourne).
Klute A (1986) (Ed.) ‘Methods of soil analysis, Part 1 – Physical and mineralogical methods (2nd edn).’ Agronomy Monograph No. 9, (American Society of Agronomy / Soil Science Society of America: Madison).
Letey J (1985) Relationships between soil physical properties and crop production. Advances in Soil Science 1, 277–294.
Letey J (1991) The study of soil structure: science or art. Australian Journal of Soil Research 29, 699–708.
Littleboy M, Silburn DM, Freebairn, DM, Woodruff DR, Hammer GL (1989) ‘PERFECT. A computer simulation model of Productivity Erosion Runoff Functions to Evaluate Conservation Techniques.’ Queensland Department of Primary Industries, Bulletin QB89005 (Queensland Department of Primary Industries, Brisbane).
McCown RL, Hammer GL, Hargreaves JNG, Holzworth DP, Freebairn DM (1996) APSIM: a novel software system for model development, model testing, and simulation in agricultural systems research. Agricultural Systems 50, 255–271.
McDonald RC, Isbell RF, Speight JG, Walker J, Hopkins MS (1990) ‘Australian soil and land survey field handbook (2nd edn).’ (Inkata Press: Melbourne).
McKenzie NJ (1991) A strategy for coordinating soil survey and land evaluation in Australia. CSIRO Division of Soils, Divisional Report No 114.
Moore AD, Donnelly JR, Freer M (1997) GRAZPLAN: Decision support systems for Australian grazing enterprises. III. Pasture growth and soil moisture submodels, and the GrassGro DSS. Agricultural Systems 55, 535–582.
Nix HA (1968) The assessment of biological productivity. In ‘Land evaluation.’ (Ed. GA Stewart). (MacMillan: Melbourne).
Nix HA (1981) Simplified simulation models based on specified minimum data sets: The CROPEVAL concept. In ‘Application of remote sensing to agriculture production forecasting.’ (A. A. Balkema: Rotterdam).
Peverill KI, Sparrow LA, Reuter DJ (1999) (Eds) ‘Soil analysis: an interpretation manual.’ (CSIRO Publishing: Melbourne).
Rayment GE, Higginson FR (1992) ‘Australian laboratory handbook of soil and water chemical methods.’ (Inkata Press: Melbourne).
Smith KA, Mullins CE (2001) ‘Soil and environmental analysis. Physical methods (2nd edn).’ (Marcel Dekker: New York).
Tongway D (1994) ‘Rangeland soil condition assessment manual.’ (CSIRO Division of Wildlife and Ecology: Canberra).
Tongway D, Hindley N (1995) ‘Manual for soil condition assessment of tropical grasslands.’ (CSIRO Division of Wildlife and Ecology: Canberra).
Verburg K, Ross PJ, Bristow KL (1997) ‘SWIMv2.1 user manual.’ CSIRO Division of Soils Divisional Report 130, CSIRO, Australia.
Zhang L, Walker GR (1998) (Series Eds) Studies in catchment hydrology: the basics of recharge and discharge. Parts 1–8 (CSIRO Publishing, Melbourne).
2
Field Sampling
NJ McKenzie and HP Cresswell
2.1 THE SOIL TO BE SAMPLED
The utility of physical measurement depends on effective sampling. Every sample should represent a definite body or class of soil. Most commonly, the body or class will be a morphologically defined soil horizon or stratigraphic layer with clearly stated dimensions. But unless the morphological criteria used for defining the body or class correlate well with the physical measurements, the body or class forms an inappropriate basis for extrapolation.
Published correlations between horizons defined using conventional morphology and soil physical properties vary considerably (Butler and Hubble 1977; McKeague et al. 1984; Bouma 1989; van Genuchten and Leij 1992; van Genuchten et al. 1999). In some instances horizons and soil profiles are effective carriers of soil physical information (Baker 1978; Wösten et al. 1985); in others correlations between recognizable horizons and physical properties are poor (e.g. Kooistra et al. 1985; McKenzie and MacLeod 1989; McKenzie et al. 1991). This Handbook draws attention to predictive relationships, or pedotransfer functions, and Chapter 22 addresses the subject directly.
Soil physical measurements will have long-term value if they have an associated site and profile description that conforms to standards defined in the Australian Soil and Land Survey Field Handbook (McDonald et al. 1990). At some experimental sites, a single description may suffice for all the plots. However, in many instances there is too much spatial variation in soil morphology, and in those situations several descriptions of the site and profile are needed. While there are problems in correlating physical measurements with morphologically defined entities, such as horizons or taxa from national soil classifications (Butler 1980), in most instances the entities still form the only feasible means for extrapolation.
The soil unit or individual to be characterised requires specification beyond that provided by a standard site and profile description. The dimensions of the soil individual of interest should be clearly defined, and for statistical purposes should be kept constant in any quantitative investigation.
2.1.1 The soil individual and georeferencing
The methods in this Handbook all involve the extraction of a specimen from a larger body of soil or in situ measurement on part of that body. The dimensions of the body are determined primarily by the purpose of the investigation, logistic constraints and existing data. The last arises so that new data can be combined with those from earlier projects and leads to more general conclusions.
A widely used definition of a soil individual is the pedon (Soil Survey Staff 1975). A pedon is a three-dimensional body of soil that has lateral dimensions large enough to include representative variations in the shape and relation of horizons and in composition of soil. Its area ranges from 1 to 10 m², depending on the nature of the variability in the soil. Webster and Oliver (1990) note that the pedon is usually too large for practicality, and its extended definition to embrace cyclic variation is statistically quite unsatisfactory.
Table 2.1: Dimensions for the site after Speight and McDonald (1990)
In Australia, the concept of a site, as presented by Speight and McDonald (1990) has prevailed, though it is rarely implemented in a strict manner for sampling and analysis. The concept is as follows:
A site is a small area of land considered to be representative of the landform, vegetation, land surface and other land features associated with the soil observation.
The extent of a site is arbitrary, but Speight and McDonald (1990) recommend dimensions for observing attributes, and these are summarised in Table 2.1.
The definitions of the pedon and site concept are flexible and have proven useful in conventional land resource survey. However, both concepts are deficient when quantitative methods of spatial data analysis are used. These deficiencies relate to geo-referencing and the dimensions of the soil individual.
An unambiguous sampling frame is required so that inclusion probabilities of all locations can be specified (e.g. in simple random sampling, every location has an equal inclusion probability). Furthermore, spatial analysis often involves data from several sources, and registration of field observations with other sources of data (e.g. remotely sensed imagery or digital elevation models) requires accurate geo-referencing of field observations.
Real-time digital differential global positioning allows field location with sub-metre accuracy and single global positioning receivers have had an accuracy of <5 m since the removal of selective availability by the United States Government in 2000. Positioning systems continue to improve and the technology is now a prerequisite for land resource survey and soil physical characterisation.
Unless there are strong reasons to the contrary, we recommend the soil individual as having an area 25 m × 25 m square that conforms to the Australian Map Grid. In landscapes with sharp boundaries, the 25 m individual may be subdivided into 5 m cells to improve resolution, or even 1 m × 1 m. Such subdivision or stratification may be warranted when there are local patterns of soil variation (e.g. compaction by traffic). The concept of the soil individual is illustrated in Figure 2.1.
It is recommended that sampling points and any relevant site boundaries be located to within 5 m, and preferably within 1 m, of their true position. This can be achieved using Differential Global Positioning System (DGPS) or at least a single GPS unit. In some circumstances (e.g. remote areas), a local benchmark may have to be used for the DGPS base station and locations will only have the required relative accuracy. The required absolute accuracy will only be achieved when the local benchmark is tied to the standard topographic survey (geodetic) framework. The local benchmark must therefore be clearly identified so that it can be located at a later date.
Figure 2.1: Concept of the soil individual.
The selection of 25 m × 25 m is pragmatic and for the following reasons:
• it corresponds to a soil volume near the upper limit exploited by mature trees;
• it encompasses commonly repetitive scales of variation in soil (though some large gilgai may have longer length scales);
• current global positioning technology can be used to locate the boundaries of the soil individual with sufficient accuracy (i.e. within 5 m and preferably within 1 m);
• it is broadly consistent with the resolution of current technology used for spatial extrapolation (e.g. digital elevation models with this resolution can be generated for many parts of the country); and
• it demands consideration of a larger volume of soil than is often the case in soil physical investigations where short-range variability is usually an issue.
Conformity to the Australian Map Grid ensures that soil individuals are contiguous and do not overlap. While some other shapes have some small advantages (e.g. triangles and hexagons), squares are easy to mark in the field. Some may prefer to use a 20 m × 20 m soil individual because field calculations and layout are easier.
The volume of the soil individual is large as specified above, and few measurement technologies can work practically at this scale. Replicated sampling within the soil individual and estimation of relevant statistics are therefore required (see below). In the future, some geophysical methods may allow for direct measurement at the scale of the soil individual.
Deciding the depth of the soil individual is more problematic than specifying the area. Many parts of Australia have a deep and often strongly weathered regolith. In the past, most descriptions in soil and land surveys have been restricted to an arbitrary solum (i.e. A and B horizons often extending to approximately 1 m). This solum is not necessarily associated with the depth of root growth, and in many landscapes, plants exploit deeper layers (C and D horizons). Furthermore, land managers and researchers require a clear understanding of the movement of water, solutes, sediments and contaminants in the whole regolith. Unfortunately, existing databases have a large portion of censored data. This is because the depth of characterisation has been limited by the method of observation (e.g. soil augers or backhoe pits are often restricted to 1–2 m) or survey purpose (e.g. many agriculturally focused surveys were only concerned with the first metre). So record whether data are censored and if so for what reason (e.g. standard procedure, lack of time, limit of equipment, coarse fragments).
2.1.2 Locating the soil individual in the field
The particular method chosen for locating the soil individual in the field will depend on the purpose of the investigation. Sample sites are often selected as ‘typical’ of the experimental area, land unit or unique mapping area in question. This is purposive sampling and Webster and Oliver (1990) commented:
When a population is very variable and resources allow the soil at only one or two sites to be examined, this method can provide a more accurate description than others. It relies heavily on personal judgement. There is no way of knowing just how good that judgement is, or of communicating the expert's confidence in his choice. Further, there is a strong risk that an expert's choice will be biased to some extent; that is, it will give preference to some part of the population at the expense of the rest. Bias is almost always present in human judgement, and it cannot be avoided either by training or by conscious effort.
With experience, the level of bias associated with purposive sampling can be reduced. Purposive sampling should be supported by statements on:
• the resources available for sampling;
• criteria used for stratification of the study area (whether it be an experimental plot, catchment or region);
• criteria for allocating samples to strata;
• rules used for locating observations in the field (e.g. Petersen and Calvin 1986); and
• areas excluded from sampling.
However, the only sure way of avoiding bias is by probabilistic sampling. The theory is well established – see for example Cochran (1977), Yates (1981), and Webster and Oliver (1990) for designs and computation of statistics. These designs lead to unbiased estimates of population parameters (means, variances, etc.) and to the replication needed to achieve sufficient precision. A design-based sampling strategy is appropriate in most instances. In some cases, the spatial structure of soil variation within the soil individual will be of interest, and in these circumstances a model-based or geostatistical approach will be more appropriate (Brus and de Gruijter 1997; de Gruijter 1999).
A detailed discussion of statistical sampling theory is beyond the scope of this Handbook. Seek advice from a statistician because the optimal sampling strategy will vary from study to study and putting theory into practice can be problematic. Bias can enter because certain parts of the landscape are inadvertently ignored, for example, outcropping rock or areas under large trees might not be registered as potential sample sites. This will lead to biased estimates of some properties (e.g. water storage capacity of a catchment).
Probabilistic sampling can address some of these problems through well-defined protocols. These should include definition of the soil individual and rules for accepting or rejecting sites when in the field – this is necessary so that every site has a known inclusion probability. However, the logistic problems associated with some circumstances cannot be readily overcome (e.g. obtaining soil specimens from under the base of large trees).
2.1.3 Representative elementary volume
The size of a specimen can have a major influence on a physical measurement. For example, Anderson and Bouma (1973) measured the saturated hydraulic conductivity (Ks) of soil cores of varying height but with a diameter of 75 mm. They found Ks declined from 271 mm/h in cores with a height of 50 mm to 31 mm/h in cores with a height of 170 mm. The standard deviation also declined from 145 mm/h to 11 mm/h. A substantial portion of the short-range variation reported, particularly for hydraulic conductivity (Wilding and Drees 1983), is probably due to measurement methods being applied at inappropriate scales (Williams and Bonnell 1988).
The size of a specimen should be varied as a function of soil structure wherever structure has an important effect on the physical property of interest. For example, macro-structure has a major bearing on hydraulic conductivity at potentials near zero whereas its impact on water retention at large negative potentials (e.g. ψ = –150 m) is trivially small.
Where structure is important, the volume should be large enough to contain a representative number of elementary units of structure (i.e. peds and pores) or repetitive units. This volume is the representative elementary volume (REV) (Bear 1972; Wagenet 1985). Bouma (1985) suggested that some 20 elementary units of structure should be contained in a REV and evidence to support this was provided by Burke et al. (1986) and Lauren et al. (1988). The REV is a useful guide for estimating the volume required to reduce variability between measurements, however, the concept has several limitations. The definition of the elementary unit of structure is qualitative and can be difficult to apply in some soils; for example, elementary units may not be identifiable in the field in weakly structured and massive soils. Bear in mind that soil variation is inherently multiscaled (see Chapter 22), and conceptually there is no a priori reason to expect measurement variation to be less at any specified scale.
The REV of some soils is large and specimens with 20 elementary units of structure cannot be obtained easily (e.g. large columns and prisms in Sodosols and Vertosols). Williams and Bonell (1988) measured large differences in hydraulic properties between tussock and bare areas in a north Queensland woodland. Similar differences are likely between the mounds and depressions of gilgai micro-topography. An extreme case occurs in areas with outcropping rock. In some landscapes such short-range variation can have an overriding effect on the soil physical environment (e.g. Williams et al. 2001).
In these cases, the soil layer can be stratified and measurements made on separate structural features that are then integrated (Bouma et al. 1989); for example, cores could be collected from within columns and along fissures between columns in the B horizon of a Sodosol. However, there is no general solution for integrating properties when measurements are made at a scale smaller than the characteristic length of the repetitive unit (Youngs 1983; Williams and Bonnell 1988). It may be appropriate to undertake measurements on large volumes in situ (e.g. Shaw and Yule 1978; Williams and Bonnell 1988). Measurements of the emergent properties of the landscape may also be appropriate when short-scale variation appears overwhelming.
There are practical limits to the size of a specimen. For measurements apart from Ks, the diameter of a specimen should be 75 mm, and preferably 100 mm. For some purposes in which high accuracy is not required, a 50-mm diameter specimen may suffice; narrower cores should not be used in measurement of any property affected by changes in soil structure. Ideally, the length of a core should be equal to its diameter to minimise edge effects and ensure specimen strength. A long core will increase greatly the time taken to reach equilibrium water content when wetting or draining at a fixed matric potential. Short, small cores should not be used for measuring saturated hydraulic conductivity.
Cores with diameters and lengths between 100 mm and 300 mm should be used for measuring hydraulic conductivity in the laboratory. Smaller cores or clods can be obtained from the large core and used for other measurements (e.g. water retention, bulk density, COLE, etc.). Collecting large cores requires a greater effort in the field.
2.1.4 Replication and bulking
As above, design-based sampling will allow you to determine the means and variances of variables for a specified layer in a soil individual. Randomisation ensures unbiased results regardless of the structure of spatial variation – despite the common misconception to the contrary (Brus and de Gruijter 1997).
Investigators often need to know the number of replicates required to achieve a given accuracy in estimating the mean. This requires knowledge of the population variance σ², either from previous experience or an initial sample. The sample size n needed to estimate the mean µ of a soil body within the limits µ – x and µ + x is estimated by
where tα² is Student’s t at the level of probability α and s² is the variance determined on a sample. Unfortunately, the sample sizes are often larger than the investigator can afford and the following series of difficulties loom large.
Estimating the mean value for a soil body is often complicated by logistical factors. Obtaining random replicates from surface and near-surface layers is straightforward when the soil individual is defined by a 25 m × 25 m area, with the lower boundary coinciding with the weathering front. However, collecting a random sample of undisturbed soil cores from deeper layers is more expensive. Physical access in a soil pit usually forces cores and disturbed samples to be collected purposively. Table 2.2. records typical replications. It is highly likely that the replication for some measurements (e.g. Ks) is inadequate. Every effort should be made to enforce better statistical control and achieve efficiency through stratification and bulking.
Table 2.2: Preferred specimen types for soil physical determinations
a The number of replicates is indicative. The required accuracy and precision for an investigation may demand more replication.
b A single determination is normally adequate in a well run laboratory. Duplicates or standard specimens are included for quality control.
c The number of duplicates refers to the larger ring size – more determinations with the smaller rings are required.
If disturbed specimens can be used and the soil properties are additive then bulking retains the precision of replicated measurement but at a reduced cost of measurement (Webster and Burgess 1984). All specimens that form a bulked specimen must be drawn from the same sampling unit (e.g. surface horizon from the soil individual) and for ease of calculation each specimen must contribute the same amount to the composite. The bulked specimen can provide an estimate only of the mean – and not the variance (Petersen and Calvin 1986). The bulked specimen must be mixed thoroughly prior to measurement (Section 2.5).
2.1.5 Sampling the profile (Method 502.01)
It is common practice to collect a series of specimens that encompass the whole profile – that is, sampling is continuous down the profile. This contrasts with taking specimens from the centers of selected sampling intervals. Within the general guidelines below, the subdivision of the profile should be based on horizonation and not on fixed depth intervals. Sampling of these layers is continuous when disturbed specimens are collected.
A maximum sampling interval of approximately 0.10 m should be used in the upper 0.30 m of the profile – finer divisions may be required near the surface to properly characterise soil chemistry and water repellence (Chapter 5). A maximum interval of 0.30 m used between 0.30 m and 2.00 m (i.e. some thick horizons may be sampled at two or more depths). Below 2 m the sampling interval should be sufficient to characterise whatever is found. The guidelines are intentionally approximate to allow flexibility in sampling profiles that have clear horizon boundaries. For soils with gradual or diffuse boundaries, or with very thick horizonation, sampling on a standard interval of 0.10 m, 0.20 m, 0.30 m, 0.60 m, etc., is recommended.
Where only two or three layers can be selected for characterisation, emphasise layers that exert most control on the soil’s physical environment. These will normally include: the A1 horizon; the top of the B horizon, particularly if it hinders movement of water and air; and the base of the profile. Take cores and clods from the centers of the selected intervals unless there is a reason for characterising the upper or lower boundary of the layer (e.g. when a crust or pan is present).
2.1.6 Undisturbed versus disturbed specimens
The methods described in this Handbook require different types of specimen. The preferred type for each is given in Table 2.2. For some measurements the best result will be obtained by maintaining the natural structure of the soil. In others, loose material that has been broken or ground will be more appropriate. The degree of development of soil structure, soil water content and the possible effects of distortion will influence the size of specimen and necessary care in handling.
2.1.7 Excavation and drilling
Sample freshly dug pits as far as finance and time permit. Freshly dug pits are preferred because vertical and lateral variations are easily observed and large intact specimens can be obtained. They allow you to sample accurately with little contamination from one specimen to another. On the other hand, pits have the disadvantage of being costly if contractors have to be engaged or a backhoe or mini-excavator purchased. Further, some areas are too steep for machines or for other reasons may be inaccessible (e.g. wetlands).
Figure 2.2: Mini-excavator with earth chain.
Pits are best dug with a backhoe or mini-excavator. Mini-excavators are versatile and can be towed behind a standard four-wheel-drive vehicle. They can be used at various stages during field operations: for example, to open pits, to remove excess soil during sampling and push in large cores. Mini-excavators are maneuverable and have sufficient power to dig large pits to approximately 2 m in most conditions. Implements can be attached to the boom, including hydraulic hammers, jacks and earth chains. The last can be used to cut pedestals prior to coring (Figure 2.2). Hiring a backhoe and operator for field sampling can be expensive, especially if they are idle for long. Furthermore, difficulties sometimes arise when directing operators to cut back ledges corresponding to horizons. There are strict regulations on securing and reinforcing pits deeper
Avis
Avis
Ce que les gens pensent de Soil Physical Measurement and Interpretation for Land Evaluation
00 évaluations / 0 avis