ИСТИНА

Geochemical indexes based on the relations of bulk concentrations of selected major and minor elements became recently one of the most popular instruments for paleoenvironmental interpretation of paleosols and especially paleosol-sedimentary sequences due to the relative affordability of the method, allowing for obtaining a quantitative estimation of paleoclimatic parameters. The conclusion that these indexes could be successfully used as a paleoclimate “proxy” relies on the assumption that the behaviour of these elements in the soil systems depends upon some processes which are in turn controlled by the climatic conditions (temperature and precipitation). After acceptance of this assumption, researchers established and quantified empirical relationships between bulk geochemistry of modern soils and climatic data such as mean annual precipitation and mean annual temperature for a limited number of North American soils. These relationships were directly projected to the past. In the majority of cases, each coefficient is related to a single pedogenetic process and regarded as a “climafunctions” for paleosols. Most popular are coefficients indicative of weathering, among them are CIA, CIA-K, PWI. These indices called the “chemical index of alteration” were originally elaborated and applied to geological materials: sediments and regolith of crystalline rocks (Nesbitt, Young, 1982). They are based on the knowledge about differences in the element mobility after their release in the course of primary mineral alteration and usually are ratios between selected immobile and mobile element values. We decided to verify whether these coefficients are really closely related to the weathering status of rock and soil materials and coincide with the other evidences of mineral alteration. We tested these coefficients on the bulk chemical composition of two types of objects: 1) weathering rinds of crystalline rocks and 2) genetic horizons of the profiles of Retisol and Podzol – “central images” of the temperate humid forest pedogenesis on loamy and sandy sediments respectively. Weathering rinds have shown higher values of the coefficients in comparison with the unaltered rock from the core – this confirms that in this case they correctly indicate the weathering status. However, in the soil profiles situation was quite different. The most weathered horizons in the Retisols and Podzols according to mineralogical and micromorphological data are eluvial E horizons. However, these horizons had relatively low values of the geochemical coefficients whereas their maxima were determined in the Bt and Bs horizons – where other weathering features are less pronounced. This means that in the mature profiles of humid temperate soils, the geochemical coefficients supposed to reflect the weathering status fail to meet the expectations. We speculate that the coefficients work well in the weathering rinds because they represent a simplified model where only the weathering process is transforming the mineral mass and controls bulk chemical composition. Soil is a much more complex system where other processes are also affecting mineral mass leaving their imprint in the contents of different elements. In the particular cases of the studied soils these processes are: clay illuviation in the Retisols and illuviation of Al- and Fe- organic complexes in Podzols. These processes have their own bioclimatic requirements and rates of development which do not coincide with that of weathering. Thus, we cast doubt on the simplistic model of application of geochemical coefficients to paleoenvironmental reconstruction: coefficient value – weathering status – parameters of paleoclimate. Further limitations to the use of geochemical coefficients are imposed by the variations in the composition of sediments within the paleosol-sedimentary sequences. These variations caused by changes in the source area or by the fluctuating characteristics of a deposition agent (e.g. wind or water flow velocities) could cause major changes in granulometric, mineralogical and thus – in the bulk chemical composition. In turn, those changes could have a major influence on the coefficient values overshadowing the weathering signal. Furthermore, frequently the sedimentary layers include redeposited soil materials. In this case, their geochemical composition could be indicative of earlier weathering and pedogenesis in the original source soil – but not of the layer to which they were incorporated. An illustrative example of this phenomenon was encountered in the Pleistocene paleosol-colluvium sequence near Kursk, Russia. In this sequence, early glacial pedosediment overlying interglacial paleosol showed similar values of weathering coefficients to the paleosol Bt horizon – despite being deposited under completely different paleoclimatic conditions. The reason for this similarity is that this sediment incorporates a large amount of redeposited Bt material – that was confirmed micromorphologically. We conclude that the geochemical weathering indexes could be successfully utilized in the paleoecological interpretation of paleosols only as part of a toolkit of methods, after thorough detection of all pedogenetic and sedimentary processes which could influence this signal.