Agriculture Reference
In-Depth Information
It is known that many mapping units correspond with organized systems for which the distri-
bution of soil units, like pedons, is not random but follows the laws of pedological evolution.
Consequently, different soil types are not independent entities. Instead, their relative geographical
position frequently indicates the pedogenetic relationships in the arrangement of the soil mantle.
These spatially linked structures inÞuence present soil and landscape functions, such as the orien-
tation and importance of the vertical or lateral Þow of water and matter. The aim is to determine
the existence of structures at appropriate levels that may range from the basic catchment area to
the regional landscape unit. Thus we attempt to reveal the role played by this geographic arrange-
ment in the general functioning of natural systems. This has led to the notion of Ñpedological
systemsÒ or Ñsoil-systemsÒ linked, within the landscapes, to geomorphologic units formed by
toposequences and catchment areas of different size.
SHORT REVIEW OF THE CONCEPTS OF SOILSCAPE AND SOIL-SYSTEMS
From the early inception of the idea of soilscape in the 1970s, the importance of soil/landscape
relationships was brought to the fore by a number of authors, relating them to the concepts of ÑSoil
landscape systemsÒ and ÑSoil-systems,Ò and stressing the importance of modelling these arrange-
ments (Hugget, 1975). The formalization of the concept of ÑSoil CoverÒ also dates from this time
(Fridland, 1975; 1976), with different levels of Ñsoil combinations.Ò
A structured approach to the analysis of the relationships between landscapes and soilscapes,
the latter being a geographic subdivision of the former, was proposed by Hole (1978). He indicated
the advantage of this approach for structuring the whole of the existing cartographic data. Northcote
(1984) elaborated on the relationships between Soilscapes or ÑSoil-landscapes,Ò taxonomic units,
and soil proÝles.
However, the understanding of the arrangement of soils within their soilscapes raises numerous
methodological problems (Wilding, 1989). The study of the morphogenetic processes of the land-
scapes themselves should enable the prediction of soil distribution. It seems evident that the
landscapes should be described in three dimensions, including vertical and lateral changes in the
composition of parent materials (Hall and Olson, 1991). The authors appear to consider the ÑcatenaÒ
model (Milne, 1935; 1936) as the best model for this type of analysis.
This period sees the development of various models. Hewitt (1993) highlights in particular the
existence of two types of complementary models for spatial analysis: one concerning soilscapes
and the other taking into account soil properties. These modelling approaches are widely used by
Slater et al. (1994) and McSweeney et al. (1994) to introduce different types of models. They reveal
the necessity of combining pedogenetic models with models describing the overall landscape
context. The Ñsoil-landscape modelsÒ offer a genuine spatial context for the application of pedo-
genetic models. They propose a methodological and conceptual approach, consisting of several
steps, for a three-dimensional modelling of the Ñsoil-landscape continuum.Ò The subsequent soil-
landscape modelling enables the prediction of certain soil properties, particularly by the use of
appropriate indexes (Gessler et al., 1995).
FridlandÔs ideas have been further elaborated by other Russian scientists working on the structure
of the soil cover (Kozlovskiy and Goryachkin, 1994). In 1997, a comprehensive evaluation of the
approaches was made by Sommer and Schlichting (1997), who synthesized the many works carried
out on the ÑcatenaÒ concept. The authors propose here a preliminary approach to the structuring
and typology of toposequences in relation to their functioning.
The Global Soil and Terrain Database (SOTER) programme (FAO-ISRIC, 1995), whose aim
is to elaborate a worldwide geographic database, uses in part the ideas that have been mentioned.
More recently, the European program ÑGeoreferenced Soil Database for EuropeÒ is based on the
concepts of ÑSoil Body,Ò ÑSoilscape,Ò and ÑSoil RegionsÒ (EC, 1998), stressing the necessity of
describing with precision the spatial structure of the soils within mapping units.
