5 Soils in the landscape, scale and quality

Back to: Soil health for Victoria's agriculture - context, terminology and concepts

Table 1 - Soil modifying and soil forming processes: their relationship to spatial and temporal scale, and suggested indicators for their recognition
Table 2 Data recorded during profile characterisation (McDonald et al. 1990), classified according to data type and sensitivity to change

It is well known that soils vary locally and that this variability is often strongly related to landscape position. Local (paddock and catchment scale) assessment of soil quality and health must take note of such variability and the role that landscape plays, not only in its relationship to distribution of soils, but also in the influence it has on soil functions (particularly hydrological functions). Interactions between soil type, landscape position and land management can confound comparisons of soil condition and these interactions need to be understood in any monitoring of soil health.

On a global scale differences in soil type can be explained, to various degrees, by the soil forming factors proposed by Hans Jenny (1941, 1980). Just as there are differences in some basic physical properties and behaviour for soils of different texture, there are fundamental qualitative differences between soil types. Processes that affect the pedological constitution of a soil occur at a range of spatial and temporal scales and these have been summarised in Table 1 by MacEwan (1997). Scale and rate of change are significant aspects of the soil system and soil in the landscape that affect choice and interpretation of indicators. Suitable indicators for reporting on soil health at a regional scale are related more to land qualities than to soil type (e.g. erosion). Indicators appropriate to soil management at the paddock scale should be carefully considered in relation to soil type and inherent soil quality.

Soil characterisation routinely carried out for soil survey and mapping follows agreed data standards (McDonald et al. 1990) but most of the data collected are non–quantitative and only change subtly in response to management. While they are good indicators of inherent soil quality they are not well suited to monitoring of soil condition or soil health.

Field pedological data have been classified by MacEwan and Fitzpatrick (1996) as nominal, ordinal, interval or ratio (Table 2). Nominal data are simply named and put into non quantitative classes, e.g.horizon designations. Ordinal data can be ranked, more or less, in semi–quantitative classes, e.g. air dry consistency. Interval data can be quantified in terms of equal intervals on a scale having no zero, or only a relative zero, e.g. Munsell colour. Ratio data can be quantified in relation to a true zero, e.g. horizon thickness. Most pedological attributes are recorded in a qualitative (nominal), or semi quantitative (ordinal), way. Such data are easily obtainable but are useful as indicators only when gross differences exist between the value of a parameter prior to, and following a period of management. Inherent high variability of these properties, even within a relatively pure soil mapping unit (Wilding and Drees, 1983), means that there is unlikely to be any improvement in assessment of soil quality by finding more precise quantitative methods for most field observed pedological attributes. The attributes in Table 2 have been ranked in approximate order of their sensitivity to management and, by implication, their relative usefulness as indicators of soil health (dynamic soil attributes).

Table 1 Soil modifying and soil forming processes: their relationship to spatial and temporal scale, and suggested indicators for their recognition. (from MacEwan 1997)