Characterisation of soils by mineral magnetic measurements

Abstract

Processes of iron mineral authigenesis, diagenesis, and dissolution, in interaction with primary inputs of iron minerals, act to produce vertical differentiation of soil magnetic properties. Thus, magnetic iron oxides, in common with other iron forms, may both respond to and reflect soil forming processes. Investigation of the degree of detectable and persistent magnetic differentiation within different soil materials can provide insights into the nature and direction of distinct sets of soil processes. Evidence of clear association between magnetic variation and discrete types of soil environment also has applied significance, in several areas of environmental study: for example, the observed magnetic variation can be used to discriminate between individual soils and soil horizons, for the purposes of soil surveying, sediment ‘tagging’ and tracing, and empirical modelling of source-sediment linkages. This paper reports the use of mineral magnetic techniques to characterise the iron oxide assemblages within soils of varying type and provenance. The effect of different pedogenic regimes have been investigated through measurement of a range of magnetic parameters (including magnetic susceptibility, frequency dependent susceptibility, and anhysteretic and saturation remanences). Contrary to current pedological thought, but in accord with other, magnetically-based, studies (e.g., Mullins ), the presence of magnetite/maghemite (typically of superparamagnetic-single domain grain size) within soils has been found to be an extremely widespread phenomenon. Its contemporary neoformation within the soil environment is inferred. However, the input of artificially-generated ferrimagnetic material is indicated for one of the magnetically ‘enhanced’ soils examined here, and the possibility of extraneous sources of magnetite in topsoils should be considered before pedogenic, processes of magnetic enhancement are inferred. From the data presented, discrimination of industrially-derived magnetic particles appears possible, on the basis of their very low levels of frequency dependent susceptibility (χ fd ), low anhysteretic remanence ( ARM ), and dissociation from the finer, clay-sized (< 2 μ m) components of the soil. Not the formation but the active dissolution of ferrimagnetic minerals is indicated for soils affected by the processes of gleying and podsolisation. Eluvial horizons of podsolised profiles demonstrate almost total loss of detectable magnetic content; variations in the form of the iron reprecipitated in the underlying illuvial layers may reflect differences in the pedological characteristics of these horizons. Within gleyed (waterlogged) soils, processes of magnetite dissolution appear to be grain size-specific; preferential reduction of those grains of ultrafine (superparamagnetic) and fine (single domain) magnetic grain size is indicated. Where soil forms a major contributor to catchment sediment loads, this has significance in the area of source-sediment modelling. Soil derived from permanently wet or highly leached catchment areas is unlikely to contribute to peaks in the magnetic content of deposited sediments; further, only those erosional processes that operate at a faster rate than those of magnetic depletion will have the capacity to produce peaks in sediment ferrimagnetic concentrations such as have been observed within the historical magnetic record.