Home Page
D. BAIZE, M.C. GIRARD, J. BOULAINE,
Cl. CHEVERRY and A. RUELLAN
Translation into English by J.M. HODGSON and D. BAIZE
II. OBJECTIVES AND BASIC PRINCIPLES OF THE RÉFÉRENTIEL PÉDOLOGIQUE
III. ORGANIZATION OF THE R. P.
IV. DIFFERENT WAYS TO APPLY THE RÉFÉRENTIEL PÉDOLOGIQUE
ATTRIBUTION TO THE R.P. : A FOUR-STEP APPROACH
FIGURE # 1: ATTRIBUTION OF A SOLUM TO THE RÉFÉRENTIEL
FIGURE # 2: KINDS OF ATTRIBUTION
Since 1986 the "Association Française pour l'Étude du Sol" has undertaken a project to replace the earlier soil classification, introduced in 1967 by the "Commission de Pédologie et de Cartographie des Sols" (CPCS).
The new system presented here retains the classical morphogenetic approach, but two major innovations have been introduced :
The word "soil" is commonly used to refer to the continuous three-dimensional and natural object forming the upper layers of the land surface. In the present soil reference system we prefer to use the term SOIL MANTLE which is similarly defined.
Soil mantles are composed of mineral and organic materials, in solid, liquid or gaseous forms. These constituents are so organized to form 'structures' characteristic of the soil material. Soil mantles change continuously and thus have a further dimension : time. This is why their study needs to be based on three kinds of data :
Soil mantles are usually continuous, but they may be very thin or even absent. Moreover, they are frequently modified by human activity to a varied depth and to a more or less obvious degree.
They are heterogeneous continua ; but the variations observed from place to place do not occur at random, because soil mantles are structured systems.
Several organizational levels may be distinguished within a soil mantle. At the lowest level various methods are used to identify soil structures ranging from the electron microscope to direct visual observation. At higher levels one may distinguish :
To study soil mantles, it is necessary to auger and to dig trenches or pits for subsequent description, sampling and analytical investigation. The sites for description and sampling must be carefully chosen, using the results of a preliminary landscape analysis (geomorphology, hydrography, vegetation, etc.) and in the light of information obtained as the study progresses.
Soil mantles are affected over time by pseudo-cyclic, reversible or irreversible changes. The soil organization and some soil properties undergo periodic modification (i.e daily, seasonally, annually). Dates of observation and sampling must therefore be recorded.
In Soil Science as in other sciences, we deal with continua by dividing them into smaller units : horizons and mapping units in a spatial context, soil types in a typological one.
The horizons are recognized by the subdivision of soil mantles into volumes considered to be appropriately homogeneous. It is clear that such homogeneity is relative and corresponds to a particular scale of investigation. The recognition of horizons clearly involves the detailed study of heterogeneity : the presence of distinct peds, the nature of the varied constituents forming the groundmass and also, of course, pedofeatures.
Because of their size (a thickness measured in centimetres or decimetres), horizons can be observed by eye in the field and samples easily taken by hand. Thus horizons are the most appropriate basic unit for describing and sampling soil mantles. The Référentiel Pédologique (abbreviated as R.P.) uses horizons as basic units to identify, characterize and depict soil mantles.
Each horizon has a finite volume. We need to define its content, by describing its constituents, its organizations, properties and analytical characteristics. We must also describe the container, the horizon envelope, indicating its extent and the nature and form of its boundaries. The vertical size of a horizon can be as little as a centimetre but often horizons are some decimetres thick or even a metre in places. Laterally, they have at least a decimetre wide but often can extend to a hectometre or several kilometres. A horizon is not infinite : it either fades laterally or passes gradually into another horizon. Its spatial extension is measurable.
The upper and lower limits of a horizon are often approximately parallel to the land surface. A horizon may however also occur as lenses or tongues, and it may even be completely surrounded by another horizon. Transitions between horizons may be clear or more or less progressive. Each horizon is almost always associated closely with other horizons and is tied to them by key relations, i.e. pedogenetic (gradual slow) evolution and functional relations (daily or seasonal changes). These latter aspects are of major practical importance.
The position of a horizon in relation to the interface zone of the soil mantle and the atmosphere is an essential feature. It governs in effect the accumulation of organic material, the thermal or water fluxes that affect or pass through the horizon, the mass of overlying layers which press down on it, the penetration of roots and animals, etc. In fact it influences almost all the circumstances that determine its development and behaviour.
Two other concepts will be used in the Référentiel Pédologique (with definitions that vary from those given in other countries or in earlier works) : the Solum and the Profile.
The Solum is the vertical section of the soil mantle, that can be observed within a pit or a trench. If possible, the solum includes enough of the underlying rock to enable its characterization. The horizontal dimensions of the solum are of the order of a decimetre wide and a centimetre thick. The vertical dimension varies from a few cm (in LITHOSOLS) to several meters (old soil mantles developed under conditions of strong weathering).
The Profile is the sequence of information related to a solum ordered from the land surface downwards. This information includes visual properties (the structural profile) or one single variable (the calcium carbonate profile, the moisture profile, the texture profile, etc.), or even to more complex aspects such as the weathering profile.
Solum and Profile, thus defined, differ distinctly from the concept of the Pedon which is not used in the R.P.
Soil mantles are actual natural volumes. They are exploited by Man ; they are studied in situ by scientists and mapped by soil surveyors, etc. Although the dimensions of a solum are arbitrarily limited, it can be nevertheless considered to be a real volume.
World-wide soil scientists use daily conceptual horizons. These are the result of the interpretation of certain specific morphological properties of a horizon, associated to pedogenetic processes, but this interpretation takes also into account other horizons and various elements of the soil landscape 1. Those conceptual horizons are the subject of a morphogenetic classification and are given arbitrary notation with specific symbols : H, O, A, E, S, BT, BP... In the R.P. they are called Reference Horizons.
Once interpreted, the solum may be conceptualized and schematized in the form of a superposition of reference horizons in a defined order : this corresponds to the Conceptual solum. Such conceptual solums are thus abstract perceptions in the collective minds of a group of soil scientists arising as a result of the generalization of repeated observations. This conceptualization, being the result of scientific progress and individual experience, associates a specific morphology, a specific behaviour, a range of properties and a process of evolution to define categories of various kinds : morphological, pedogenetic or other.
The Référentiel Pédologique is not a traditional classification system. Its authors have aimed to establish a typological framework which is scientific yet pragmatic, precise yet flexible, and which contains only two categories :
The R.P. is conceived as a typological space with N dimensions, in which the References are located without consideration of hierarchy. When it is necessary to make regional, national or international correlations, the soil scientist identifies a solum-concept, a cartographic delineation or a mapping-unit, in relation to the References. The number of References in the R.P. will most probably increase in future. Indeed, as soon as sola sufficiently different from ones already established are recognized, a new concept can be defined and a new Reference specified. Thus the R.P. is an "open" system.
This typological framework takes into consideration, as far as possible :
The morphology of the solum (in the broad sense including analytical and mineralogical data also), forms the main basis by which a solum is attributed to References. Major emphasis is however placed on those properties that have a dominant effect on the behaviour and properties of soils (texture, depth, structural differentiation, etc.).
The behavioural properties (agronomic, sylvicultural, geotechnical) and function of soils (regime, hydrology or structural aspects...) are given emphasis when distinguishing and defining the References. This is why PÉLOSOLS, ARÉNOSOLS, VERTISOLS, PLANOSOLS, and RÉDUCTISOLS are essential elements in the system.
Pedogenetic processes are used if they are sufficiently well understood. They form an ideal basis for the overall interpretation of sola and soil landscapes. In fact, in certain cases, the morphology and other soil properties depend largely on them. In other circumstances however, pedogenetic evolution is too weak and the solum mainly reflects the parent rock properties. Where it is known that several pedogenetic cycles have occurred, emphasis is given to the most recent cycle.
The R.P. constitutes also a synthetic LANGUAGE (synthetic in the classical sense). It includes a vocabulary developed over more than 20 years, nationally and internationally. Various terms originating from other systems (PLANOSOLS, PÉLOSOLS, ARÉNOSOLS) are used. As the result of new knowledge acquired since 1967, we have modified the definition of certain older terms and created new terms to express new concepts (ALOCRISOLS, PEYROSOLS, reductic horizon, etc).
Reference horizons form the basis of the system as they serve to define the References. R.P. defines seventy two Reference Horizons, each of them defined and described using various of the elements listed below :
A Reference Horizon is in general not diagnostic when considered on its own. Particular successions of Reference horizons however, identified together as "diagnostic sola", are used to associate the sola with their appropriate Reference.
Usually a Reference is defined by its diagnostic solum. But some References are otherwise defined :
The References are currently defined by vertical horizon sequences which have been replaced in their soil landscape. It is hoped that soon it will be possible to define "pedological systems" using the lateral organization of several sola.
In the R.P. each Reference is described in terms of the following :
The differences between References are based on observable and/or measurable properties. The present version of R.P. describes 102 References. The number may reach 150 in a future edition to deal with the soils of the whole world. The names of the References consist of one or two words which by convention are written in capital letters.
The References can be subdivided into Types by prefixing one or several Qualifiers. For example, a fluvic, vertic, clayey CALCOSOL is a Type belonging within the CALCOSOL category.
It is very useful to add as many Qualifiers as possible to specify the main properties of a solum. As examples :
An initial list of Qualifiers (adjectives, periphrases, prefixes) has been established. Each describes a property of the solum, and is defined so it has a precise meaning. The Qualifiers are always written in low case letters.
Only examples of known Types are presented in R.P. ; many more however must exist. The list of the Qualifiers is open-ended, and the number of combinations is unlimited, so the number of Types is by its very nature without limit.
Globally, the References may suffice for the exchange of information, or to convey the major geographic distribution of soil qualities. At national, regional or local level however, more details are needed to complete the information and make it easy to use.
Using the Types, it is possible to establish links between the Référentiel Pédologique and other classification systems. At this level it is also possible to extrapolate knowledge acquired at one site (the relation between the nature of soil mantle and its utilization) to other sites comparable in pedological terms. Thus, the results of an agronomic experiment on a drained, resaturated, redoxic LUVISOL DÉGRADÉ, with fragipan, developed from loam, in the Faux-Perche region, will probably be applicable to all soils of the same Type in the Paris Basin.
These refer to broad groups with well defined central concept recognized in several world classification systems (PODZOSOLS, ANDOSOLS, VERTISOLS, etc.), but which have rather ill-defined limits in relation to adjoining major groups.
MGRs have been adopted mainly to avoid unnecessary repetition in the presentation of References. They group several References which have many common characteristics, having, for example, the same Reference horizons. It has been necessary editorially to present these common properties and horizons in a single chapter.
The other more didactic advantage, is to regroup various References with the central concepts traditionally recognized as being associated. Thus seven References characterized by a podzolization process, are gathered in the PODZOSOLS MGR. In this book, several MGR are offered, but groups could be assembled in other ways.
The MGRs are not a category of the Référentiel Pédologique. Their role is only secondary and they have no hierarchical significance.
The Référentiel Pédologique enables us to attribute any solum, any mapping area or any mapping unit to one or more References. The process of attribution includes three steps :
Characterization
Ideally characterization of the soil mantle at each site requires :
Some of data are easily collected on the spot, others require special techniques (of preparation, measurement, analysis) and complementary laboratory or field studies which may take a few weeks or months.
Interpretation
Vertical or lateral horizon successions mostly help to clarify our interpretation because they do not occur at random but result from the action of pedological processes (natural or anthropic) on the original parent material.
During interpretation, a soil scientist uses available knowledge and his personal experience. He must, however, interpret and conceptualize the horizons he has described and studied in terms of Reference horizons. To do this, pedofeatures and certain morphological and/or analytical properties are attributed to pedogenetic processes ; the interpretation of a horizon cannot be made independently of the organization of other horizons in geographical space (vertical or lateral), nor independantly of various aspects of the soil landscape.
Attribution
Attribution is the process of assigning a solum to one or more References and giving it the appropriate name(s). It is made using pedological reasoning, similar to that used in interpretation of horizons.
Attribution is a flexible process which requires careful comparisons of a solum and the References. Statistical concepts like "modes" and "mathematic distances" and methods such as "multi-dimensional analysis" and "expert systems" can be used to assess similarity.
Single, imperfect, double, multiple attributions and intergrades (Figure 2) can be distinguished.
Creation of a new Reference
If a solum is too unlike currently defined References, it is possible to define an appropriate new Reference. This is possible without change to the Référentiel Pédologique as a whole. Nevertheless, to avoid confusion, every proposed new Reference Horizon, new Reference or new Qualifier should be based on a detailed preliminary, documented study, in order to maintain overall consistence.
The pedological information can be treated into two distinct ways :
Recognition of soil types
The organization of References and Types into more general categories is left to the pedologists' choice. Thus, it is possible to group all References and all Types with a key common property to form a Cognate set.
For example, one may assemble all FLUVIOSOLS and all fluvic Types (fluvic CALCOSOLS, fluvic ARÉNOSOLS, fluvic RÉDUCTISOLS, etc.) into one Cognate set (both conceptually and geographically). Another example : all Types or References with excess water within 50 cm depth could be grouped together. The composition of Cognate sets is entirely flexible.
For some particular objectives, it may be necessary to set up a typology or a classification. The Référentiel Pédologique provides the building material for such a system. From the whole or from a part of the R.P. it is possible :
When using the R. P. at local level (regional development zone, natural region, department, or canton) more detailed terminology is needed than for a national or international system. At local level, several Qualifiers drawn from the list must be used in order to characterize the Types in detail. The nature of the parent rock in particular will have to be specified : its petrology or mineralogy, age, way of deposition, etc. Each Qualifier has a relatively precise definition, so the pedological information can be transferred without confusion from one region to another, or from a local to a national scale.
Scientists from other disciplines, with interests in soil mantles, may then use the Reference horizons to set-up another typological framework that meets their needs. They can include the elements of the soil landscape (slope, vegetation,...) to complete their own classification or reference base.
Geographical units or spatial units
The characterization of a typological unit is not necessarily associated to a spatial analysis. In contrast, cartographic units, by definition, are spatial distributions (surface areas, forms, location of mapping area within them). The mapping unit is closely dependent on the scale of the survey and the publication scale of the map, so it often groups several typological units together.
The term cartography currently has two distinct meanings :
Cartogenesis needs no pre-established overall classification. Nevertheless, any synthesis of the spatial organization of soil mantles is facilitated by the use of a language. The Référentiel Pédologique acts as a language through its Qualifiers, Reference horizons and References.
There are several questions about graphic representations :
The R.P. is very flexible and it offers a language that helps in the representation of the following items :
When representing two dimensions using closed surfaces (on maps), it seems that the References are well adapted to the interpretation of mapping areas. For the cartographic units, concepts of References, Types or Cognate sets can be used. For the representation of soil mantles on a vertical plane (soil sections) the R.P. terminology may be used in order to name the volumes studied.
When constructing a map legend, the R.P. does not impose a hierarchy. The legend can be organized on the basis of soil landscapes or by natural regions. The presentation may differ from one map to another, but maps will remain compatible if the terminology (References, Types, horizons, Qualifiers) remains the same. This is more so when all the graphic and semantic data are available in computerized data bases. The Référentiel Pédologique thus does not impose the form of cartographic representation, but it offers a common language which permits change from one form of representation to another.
The Référentiel Pédologique as currently described here may not be adequate for all regions of the world : adaptation is always necessary to meet specific conditions and practical use. Such changes should be easy because of the principles of the R.P..
The Référentiel Pédologique is not definitive : knowledge progresses and concepts mature will enable improvements. The R.P. has already foreseen ways it can be updated.
The Référentiel Pédologique cannot solve all mapping problems, particularly those of the extrapolation of a set of point data into three dimensions. We hope, however, that the R.P. terminology will help to advance various such procedures.
1 soil landscape : the whole of soil horizons and landform elements (vegetation, effect of human activities, geomorphology, hydrology, parent rocks or substrata) the spatial organization of which permits to define as a whole part or the total of a soil mantle.
|
Home Page
|
Last modification by dbz@denis-baize.fr :
|